rss
Gut 61:1708-1715 doi:10.1136/gutjnl-2011-301626
  • Original articles
    • Colon

5-aminosalicylic acid inhibits cell cycle progression in a phospholipase D dependent manner in colorectal cancer

  1. Gijs R van den Brink1,2,6
  1. 1Tytgat Institute for Liver & Intestinal Research, Academic Medical Center, Amsterdam, the Netherlands
  2. 2Department of Gastroenterology & Hepatology, Leiden University Medical Center, Leiden, the Netherlands
  3. 3Department of Pathology, VU University Medical Center, Amsterdam, the Netherlands
  4. 4Department of Medical Oncology, Academic Medical Center, University of Amsterdam, Amsterdam, the Netherlands
  5. 5Department of Gastroenterology and Hepatology, Erasmus MC, University Medical Center Rotterdam, Rotterdam, the Netherlands
  6. 6Department of Gastroenterology & Hepatology, Academic Medical Center, Amsterdam, the Netherlands
  1. Correspondence to Gijs R van den Brink, Department of Gastroenterology & Hepatology, C2, Academic Medical Center, Meibergdreef 9, 1105 AZ Amsterdam, the Netherlands; g.r.vandenbrink{at}amc.nl
  1. Contributors BB designed and performed experiments, analysed data and wrote manuscript. AAD designed and performed experiments, analysed data and wrote manuscript. EH designed and performed experiments and analysed data. CLB designed and performed experiments and analysed data. PWV performed experiments and analysed data. PJK designed and performed experiments and contributed reagents. MEW performed experiments and analysed data. VM contributed reagents and analysed data. JH contributed reagents and commented on the manuscript. DJR provided supervision and commented on the manuscript. JCHH provided supervision and wrote manuscript. DWH provided supervision and wrote manuscript. HWV provided supervision, wrote manuscript and contributed reagents. MPP designed experiments, provided supervision and wrote manuscript. GRvdB designed experiments, analysed data, wrote the manuscript and supervised and mentored all work.

  • Accepted 23 November 2011
  • Published Online First 20 December 2011

Abstract

Background 5-aminosalicylic acid (5-ASA) may protect against the development of inflammation-associated colorectal cancer. In vitro data suggest that, in colorectal cancer cells, 5-ASA induces cell cycle arrest, but the molecular mechanism leading to this arrest remains to be determined.

Aim To dissect the signal transduction events that lead to 5-ASA mediated inhibition of proliferation of colorectal cancer cells, focusing on mammalian target of rapamycin (mTOR), a regulator of cell cycle progression.

Methods The influence of 5-ASA on mTOR signalling was examined in a panel of colorectal cancer cell lines. The effects of 5-ASA on the pathways that control mTOR activity were studied in detail in two different colorectal cancer cell lines, using western blot, siRNA, a phospholipase D (PLD) activity assay, proliferation assays and cell cycle analysis. The phosphorylation status of mTOR and its downstream target, ribosomal protein S6, was studied in colorectal cancers before and after topical 5-ASA treatment.

Results Treatment of colorectal cancer with 5-ASA inhibited mTOR signalling in vitro and in vivo. 5-ASA had no effect on any of the pathways that regulate the activity of the tuberous sclerosis complex in colorectal cancer cells. Both proliferation and mTOR activity depended on PLD, an enzyme that generates phosphatidic acid (PA). 5-ASA treatment inhibited PLD activity and proliferation; these effects could be rescued with exogenous PA.

Conclusion 5-ASA interferes with proliferation of colorectal cancer cells via inhibition of PLD-dependent generation of PA and loss of mTOR signalling.

Significance of this study

What is already known on this subject?

  • 5-aminosalicylic acid (5-ASA) is an anti-inflammatory drug commonly used to treat patients with ulcerative colitis.

  • 5-ASA may protect against the development of colitis-associated colorectal cancer (CRC).

  • In addition to, but independent of its anti-inflammatory actions, 5-ASA has been shown to affect CRC development via effects on cell cycle progression and apoptosis.

  • Although a number potential targets have been identified at the level of signal transduction, the exact signalling intermediates via which 5-ASA induces cell cycle arrest remain to be determined.

What are the new findings?

  • Treatment of colorectal cancer with 5-ASA inhibits mammalian target of rapamycin (mTOR) signalling in vitro and in vivo.

  • 5-ASA is a potent inhibitor of phospholipase D (PLD) and PLD/phosphatidic acid (PA) dependent mTOR signalling.

  • The effects of 5-ASA on proliferation and mTOR signalling can be reversed by the addition of exogenous PA. 5-ASA induced cell cycle inhibition is therefore dependent on inhibition of PLD.

Introduction

5-Aminosalicylic acid (5-ASA or mesalamine) is the active moiety of sulfasalazine and a cornerstone in the induction and maintenance of remission of patients with ulcerative colitis. In addition to its anti-inflammatory activity 5-ASA may protect against colitis-associated colorectal cancer development.1 Previous work from us and others2–5 shows that 5-ASA can inhibit cell cycle progression of colorectal cancer (CRC) cells, at high concentrations that are achievable in vivo. Different potential targets have been identified at the level of signal transduction. Inhibition of cell proliferation seems to be independent of effects on Cox-2.3 5-ASA has been found to interfere with epidermal growth factor receptor phosphorylation in HT-29 cells6 but this was not responsible for the cell cycle arrest. In addition, 5-ASA was found to inhibit Wnt signalling, but only in APC mutant and not in β-catenin mutant cells,7 whereas 5-ASA inhibited the proliferation of both cell types. 5-ASA may also target peroxisome proliferator-activated receptor-gamma (PPARγ),8 but a PPARγ antagonist4 or dominant negative PPARγ9 did not rescue the anti-proliferative effects of 5-ASA. Thus, the exact molecular mechanism by which 5-ASA induces cell cycle arrest remains to be determined.

Mammalian target of rapamycin (mTOR, also called FRAP-1)10 is a 289-kDa Ser/Thr kinase which is highly conserved from yeast to man and plays a critical role in regulating basic cellular functions, which include cell growth, survival, mobility and angiogenesis. mTOR exists in at least two complexes, mTORC1 and mTORC2, that differ in their composition and sensitivity to rapamycin. The effect of mTOR on protein synthesis and cell proliferation is largely achieved by controlling the translation initiation via direct phosphorylation of two downstream targets: p70 S6 kinase (p70-S6K) and 4E binding protein 1 (4E-BP1).11 When phosphorylated by mTOR, p70-S6K becomes catalytically active and phosphorylates several targets, including S6 ribosomal protein (S6), a major regulator of the expression of proteins involved in translation and cell cycle progression. 4E-BP1 is also activated by mTOR-dependent phosphorylation, causing it to dissociate from eIF-4E, thereby allowing cap-dependent mRNA translation. The activity of mTOR requires GTP-bound Rheb.12 GTP-bound Rheb is converted to inactive GDP-bound Rheb by the tuberous sclerosis complex (TSC), which consists of two proteins, TSC1 and TSC2.13 Mitogenic stimuli engaging PI3 kinase-Akt, Raf-MEK-ERK and/or IKKβ signalling all seem to converge on the TSC and inhibit the TSC by phosphorylation of either one of these two TSC proteins. This reduces their GTP-exchange activity and results in accumulation of GTP-bound Rheb and subsequent activation of mTOR.13 The regulation of mTOR activity can also occur via TSC-independent mechanisms; for example, mTOR can be inhibited by proline rich Akt substrate 40 (PRAS40) which prevents binding of mTOR to its substrates.14 Finally, in order for mTOR to be active there seems to be a requirement for phosphatidic acid (PA) which directly binds to mTOR.15 ,16 PA can be generated from phosphatidylcholine by phospholipase D (PLD).

Here we investigate the molecular mechanism via which 5-ASA induces cell cycle arrest. We find that 5-ASA is a potent PLD inhibitor and inhibits cell cycle progression through PLD/PA-dependent mTOR signalling. Importantly, the effects of 5-ASA on proliferation and mTOR signalling can be reversed by the addition of exogenous PA.

Materials and methods

Cell lines

Human cancer cell lines originating from the colon were obtained from the ATCC (Manassas, Virginia, USA). Culture conditions are described in the supplementary information.

Reagents, siRNA and antibodies

Information on siRNA, antibodies and reagents can be found in the supplementary information.

Cell stimulation

For 5-ASA treatment, 7.5×10e5 cells/well were seeded in 6-well plates (Greiner). After 16 h, cells were washed twice with phosphate-buffered saline and serum-deprived in Dulbecco modified Eagle medium (DMEM) for 5 h before treatment with serum-free DMEM, 5–40 mM 5-ASA or 40 mM mannitol for the times indicated. For each experiment, 5-ASA and mannitol were freshly dissolved in serum-free DMEM and the pH was adjusted to 7.5 with 10 M NaOH. Additional information can be found in the supplementary methods.

Western blot

Cell lysis and western blotting were performed using standard protocols (see supplementary data for a detailed protocol).

Patient study and immunohistochemistry

For immunohistological analysis, tumour biopsies collected as an extension (PJK and HWV, unpublished data) of a previously described study17 were used. In this study, patients with carcinoma of the rectum or sigmoid colon were asked to take one Salofalk enema containing 4 g 5-ASA/60 g (Dr Falk Pharma, Freiburg, Germany) every evening before bedtime for 14 consecutive days while awaiting surgery. The study was performed with informed consent from the patients, obtained according to the guidelines of the Medical Ethics Committee of Leiden University Medical Centre (protocol P172/94). Tumour biopsies were taken endoscopically at entry to the study and after the 14 days of enema treatment. The biopsies were fixed in 4% buffered formalin, routinely processed and embedded in paraffin wax. Immunohistochemistry was performed using standard protocols (see supplementary methods). The staining intensity was scored by two independent observers that were blind to the treatment.

PLD assay

Cells were plated, grown and treated with 5-ASA or mannitol for 30 min as described above. Cells were then washed twice in ice cold phosphate-buffered saline and lysed on ice in cold PLD lysis buffer (10 mM Tris/HCl, pH 7.6, 100 mM NaCl, 20 mM CaCl2, 2 mM EGTA and 1% Triton X-100) supplemented with EDTA-free Complete Protease Inhibitors (Roche, Woerden, the Netherlands), 1 mM sodium orthovanadate and 10 mM sodium fluoride. Lysates were centrifuged at 13 000 rpm for 5 min at 4°C; 50 μg samples of protein were used to determine PLD activity with the Amplex Red PLD assay kit (A12219, Invitrogen, Breda, the Netherlands), according to the manufacturer's protocol.

siRNA transfections

Cells were transfected with 100 nM siRNA per well using lipofectamine, as detailed in the supplementary data.

Proliferation assay and cell cycle analysis

Cellular proliferation was determined by colorimetric BrdU ELISA or 3H thymidine incorporation assay after 24 h of treatment. Cell cycle analysis was performed by flow-cytometric determination of PI staining of cells after 20 h of treatment. For detailed protocols see the supplementary methods.

Results

5-ASA inhibits mTOR signalling in vitro and in vivo

To examine the molecular mechanism of 5-ASA-induced cell cycle arrest, we focused on an important regulator of cell cycle progression: mTOR. We studied the effect of 5-ASA on the phosphorylation of mTOR at Ser2448 using a phospho-specific antibody in DLD-1 and SW480 colorectal cancer cell lines. Phosphorylation of this residue is correlated with mTOR activation and therefore useful as a determinant of mTOR activity.18 ,19 When we exposed DLD-1 and SW480 cells to different 5-ASA doses we found that 5-ASA reduced mTOR phosphorylation on Ser2448 as compared to the osmolarity control mannitol (figure 1A,B). Depending on the experiment, the effect of 5-ASA on mTOR phosphorylation was observed at a dose of around 10–40 mM, which corresponds well with the dose at which cell cycle inhibition is observed in our previous experiments and those of others.2–5 Timecourse analysis showed that inhibition of mTOR Ser2448-phosphorylation was observed as early as 30 min after exposure to 40 mM 5-ASA (figure 1C,D) and was stable over time until at least 48 h (not shown). In order to examine the consequences of loss of phosphorylation of mTOR at this residue we analysed the phosphorylation status of two of signal transduction arms that are downstream of mTOR kinase activity, 4E-BP1 and p70S6K and its target S6. In accordance with loss of mTOR Ser2448-phosphorylation, 5-ASA treatment inhibited the phosphorylation of all three proteins (figure 1C,D). We subsequently exposed several other colorectal cancer cell lines (HT-29, HCT116, RKO) to 5-ASA (figure 1E) and found that phosphorylation of mTOR and its downstream targets was inhibited in all cell lines examined, indicating that the effect is unlikely to depend on the genetic make-up of the cancer cell line. We next examined whether 5-ASA similarly inhibited mTOR-Ser2448 phosphorylation and downstream mTOR signalling activity in human patients in vivo. Therefore colorectal cancer biopsies taken before and after 5-ASA treatment17 were stained for phospho-mTOR and phospho-S6 as a readout of downstream mTOR activity. 5-ASA treatment of these colorectal cancer patients led to a significant reduction in staining intensity with both the phospho-mTOR and phosphor-S6 antibodies in cancer biopsies (figure 2).

Figure 1

5-Aminosalicylic acid (5-ASA) inhibits mammalian target of rapamycin (mTOR) signalling in vitro. (A, B) Western blots of DLD-1 and SW480 colorectal cancer cells treated with the indicated concentrations of 5-ASA or mannitol as an osmolarity control. (C, D) Western blots of DLD-1 and SW480 colorectal cancer cells which were treated for 60 min with 40 mM 5-ASA or mannitol. (E) Western blots of colorectal cancer cell lines HT-29, HCT116 and RKO exposed to 40 mM 5-ASA or mannitol for 60 min.

Figure 2

5-Aminosalicylic acid (5-ASA) inhibits mammalian target of rapamycin (mTOR) signalling in vivo. (A) Immunohistochemistry for phospho-mTOR in two representative colorectal carcinomas before and after a 14-day treatment with 5-ASA enemas. (B) Phospho-mTOR staining intensity for all individual patients (*p<0.05, Mann–Whitney test). (C) Immunohistochemistry for phospho-S6 in two representative patients before and after 5-ASA treatment. (D) Staining intensity of p-S6 staining in all individual patients (*p<0.05, Mann–Whitney test).

5-ASA-induced inhibition of mTOR is not dependent on PPARγ

Since it has previously been shown that 5-ASA is a ligand for PPARγ and that some actions of 5-ASA are due to its agonist properties towards PPARγ,8 we examined the effect of rosiglitazone (PPARγ agonist) and GW9662 (PPARγ antagonist) on proliferation of DLD-1 and SW480 colorectal cancer cells (figure 3A). In DLD-1 cells no inhibition of proliferation was observed with either rosiglitazone or GW9662. In SW480 cells a very modest reduction in proliferation was observed with both rosiglitazone and GW9662. Treatment with GW9662 did not diminish the effect of 5-ASA on proliferation in either DLD-1 or SW480 cells. We next examined whether effects of 5-ASA on mTOR phosphorylation were dependent on PPARγ. We found that rosiglitazone did not influence the level of phosphorylation of mTOR at Ser2448, in contrast to 5-ASA (figure 3B). Also, effects of 5-ASA on mTOR did not seem dependent on PPARg signalling as they were unaffected by pretreatment with the PPARg antagonist GW9662 (figure 3B) and by PPARg specific siRNA (figure 3C).

Figure 3

5-Aminosalicylic acid (5-ASA) induced inhibition of mammalian target of rapamycin (mTOR) and cell proliferation is peroxisome proliferator-activated receptor-gamma (PPARγ) independent. (A) DLD-1 and SW480 colorectal cancer cells were treated with the 5-ASA and/or PPARγ agonists rosiglitazone or the PPARγ antagonist GW9662; the effect on proliferation was assessed after 24 h using a colorimetric BrdU ELISA (fold change in proliferation ± SE). *p<0.05, *** p<0.005; ns, not significant, in one-way ANOVA with Tukey's multiple comparison test. (B) SW480 and DLD-1 cells were treated with 5-ASA (ASA), rosiglitazone (rosi) or GW9662 (GW) and analysed by western blot. (C) Western blots of DLD-1 cells transfected with scrambled control (Scrbl) or PPARγ-specific siRNA 72 h before serum starvation and 1-h 5-ASA treatment.

5-ASA does not act at the level of the TSC

The classical mTOR activating pathways inactivate the TSC complex by phosphorylation of either TSC1 or TSC2. To our surprise however, experiments in DLD-1 and SW480 cells showed that 5-ASA treatment did not affect the phosphorylation status of crucial components of the major pathways known to regulate TSC phosphorylation (see figure 4A). 5-ASA had no effect on phosphorylation of PI3 kinase-Akt signalling pathway components such as phosphatase and tensin homolog (PTEN), phosphoinositide-dependent kinase-1 (PDK1) or Akt. Phosphorylation of AMP-activated protein kinase (AMPK) as well as MAP/ERK kinase (MEK) and extracellular signal-regulated kinase (ERK) was also unaffected. Similarly, no effects were observed on the phosphorylation of protein kinase C (PKC) using a pan-phospho-PKC antibody. Indeed, 5-ASA did not affect the phosphorylation status of Ser1254 of TSC2 (figure 4B), which is the target of the pathways mentioned above. Furthermore, when the TSC was inactivated by siRNA-mediated knockdown of TSC2, both the basal mTOR phosphorylation status and its sensitivity to 5-ASA were unaffected (figure 4C). Together these data suggest that 5-ASA may inactivate mTOR signalling independently from the TSC.

Figure 4

5-Aminosalicylic acid (5-ASA) does not act at the level of the tuberous sclerosis complex (TSC). (A) Western blots of DLD-1 and SW480 colorectal cancer cells treated with 5-ASA or mannitol. (B) Western blots of DLD-1 cells treated with 5-ASA or mannitol. (C) Western blots of DLD-1 cells transfected with scrambled control (scrbl) or TSC2-specific siRNA 48 h before a 1-h treatment with 5-ASA or mannitol.

Proliferation and mTOR signalling are dependent on PA and PLD in colorectal cancer cells

Alternative, TSC-independent regulation of mTOR activity can occur via PLD-dependent generation of phosphatidic acid (PA).15 ,16 Primary alcohols compete with water as the hydroxyl donor in the hydrolysis of phospholipids by PLD, resulting in the generation of phosphatidylalcohol rather than PA. The activity of the two PLD isoforms 1 and 2 can therefore be inhibited by treatment of cells with the primary alcohol 1-butanol, whereas tertiary alcohols, such as t-butanol, are not utilised by PLD and serve as negative controls. We found that 1-butanol but not t-butanol inhibited proliferation of both DLD-1 and SW480 colorectal carcinoma cells (figure 5A). To examine whether PLD signalling is required to maintain mTOR phosphorylation at Ser2448 in colorectal carcinoma cells, we treated DLD-1 and SW480 cells with 1-butanol and t-butanol and examined the effects on mTOR phosphorylation. We found that 1-butanol but not t-butanol inhibited mTOR phosphorylation in both cells (figure 5B). To confirm the requirement of PLD signalling in the maintenance of mTOR signalling, we treated DLD-1 cells with specific siRNA against PLD1 and PLD2 and found that mTOR signalling depends on PLD2 and not PLD1 in these cells (figure 5C). Together these data suggest that maintenance of mTOR phosphorylation depends on PLD2 signalling.

Figure 5

Mammalian target of rapamycin (mTOR) signalling is dependent on phosphatidic acid (PA) and phospholipase D (PLD) in colorectal cancer cells. (A) DLD-1 and SW480 colorectal cancer cells were treated with primary (1-BtOH) or tertiary butanol (t-BtOH), as a control. The effect on proliferation was assessed after 24 h using a 3H-thymidine incorporation assay (counts per minute ± SE). *p<0.05, **p<0.01 in one-way ANOVA with Tukey's multiple comparison test; * vs DMEM. (B) Western blots of DLD-1 and SW480 colorectal cancer cells treated with primary (1-BtOH) or tertiary butanol (t-BtOH) for 6 h. (C) Western blots of DLD-1 cells 48 h after transfection with scrambled control (scrbl), PLD1- or PLD2-specific siRNAs.

5-ASA inhibits PLD activity and exogenous PA rescues 5-ASA effects on mTOR signalling and proliferation of colorectal cancer cells

Several observations suggest a role for PLD in the inhibition of mTOR by 5-ASA. First, 5-ASA inhibits mTOR signalling independent of the TSC in DLD-1 and SW480 colorectal cancer cells (see figure 4). Second, PLD is known to regulate mTOR below the level of the TSC.20 Finally, mTOR phosphorylation depends on PLD signalling in DLD-1 and SW480 cells under our experimental conditions (see figure 5). Therefore, we examined whether 5-ASA inhibits PLD/PA-dependent signalling. We used an enzymatic assay to measure PLD activity and found that 5-ASA significantly inhibits PLD activity in both DLD-1 and SW480 cell lysates (figure 6A). Inhibition of PLD activity was dose-dependent and maximal at doses between 20 and 40 mM (figure 6B). Western blot analysis of the same cell lysates that were assayed for PLD activity indicated that the dose dependent effect of 5-ASA correlated well with inhibition of phosphorylation of mTOR and S6 (figure 6B). Furthermore, with 5-ASA from two independent additional sources, including clinical grade 5-ASA (a kind gift of Ferring), we found similar dose-dependent effects on both inhibition of the mTOR signalling pathway and PLD activity (see supplementary figure S1), suggesting that the observed effects are dependent on 5-ASA and not likely to be caused by potential contaminants.

Figure 6

5-Aminosalicylic acid (5-ASA) inhibits phospholipase D (PLD) activity and exogenous phosphatidic acid (PA) rescues the 5-ASA effect on mammalian target of rapamycin (mTOR) signalling and proliferation of colorectal cancer cells. (A) DLD-1 and SW480 colorectal cancer cells were treated with 5-ASA or mannitol and analysed for PLD activity (mean relative PLD activity relative to DMEM ± SE). ***p<0.005 in one-way ANOVA with Tukey's multiple comparison test. (B) PLD activity (mean ± SE) and western blots of DLD-1 cells treated with DMEM or increasing concentrations of 5-ASA (in mM). (C, D, E) DLD-1 cells were treated with 5-ASA or mannitol (Mann) (C), primary (1-BtOH) or tertiary butanol (t-BtOH) (D), or the indicated concentrations of rapamycin (RAPA) (E) diluted in DMEM with 5% FCS (DMEM); cell cycle distribution was assessed after 24 h by flow-cytometric determination of PI staining of live cells (percentage of live cells per cell cycle phase ± SE). */# p<0.05, **/## p<0.01, ***/### p<0.001 in two-way ANOVA with Bonferroni post-test; * versus DMEM and # versus mannitol in (C) and versus t-BtOH in (D). (F) Western blots of DLD-1 cells treated with 5-ASA for 30 min before addition of vehicle (−) or PA (+) for 1 h. (G) DLD-1 cells were treated with 5-ASA in the presence or absence of PA and the effect on proliferation was assessed after 24 h using a colorimetric BrdU ELISA (fold change in proliferation ± SE). ***p<0.005 in one-way ANOVA with Tukey's multiple comparison test.

PRAS40 is a negative regulator of mTOR that can be inhibited by phosphorylation at Thr264 in a PLD/PA-dependent manner.16 Indeed, we found that the reduction of PLD activity by 5-ASA correlated very well not only with reduction of phosphorylation of mTOR but also with that of PRAS40 (figure 6F). The PLD/PA dependence of PRAS40 phosphorylation was further substantiated by the observation that 1-butanol but not t-butanol inhibited PRAS40 phosphorylation (see figure 5B).

Both 5-ASA and inhibition of PLD activity with 1-butanol inhibited cellular proliferation (as shown in figures 3 and 5). To examine the cell cycle arrest in more detail we analysed the cell cycle distribution of DLD-1 cells treated for 24 h with 5-ASA, 1-butanol or appropriate controls. In accordance with the data presented above, both 5-ASA and 1-butanol induced a relatively similar cell cycle arrest, with reduced G1 phase and an accumulation in mostly G2/M-phase compared to untreated and control cells (figure 6D and supplementary figure S2). This finding was confirmed with the two additional high-grade 5-ASA preparations (see supplementary figure S3). In contrast, inhibition of mTOR with high dose (80 nM) rapamycin led to a G1 arrest (figure 6E and supplementary figure S2), suggesting that the downstream effects of 5-ASA induced PLD/PA inhibition might not be solely mediated through this kinase. However, we noticed that when lower concentrations (40 and 20 nM) of rapamycin were used, we found cell cycle distributions very similar to those induced by 1-BtOH and 5-ASA (see figure 6E and supplementary figures S2 and S3). This suggests that perhaps specific (rapamycin-sensitive) mTOR containing complexes are more readily affected by reduced PA availability, induced by 5-ASA or 1-butanol treatment. Similar findings have previously been reported,16 and thus provide an alternative explanation for the apparent discrepancy found between high dose Rapa and 5-ASA induced cell cycle arrest.

To study the functional relevance of the inhibitory effect of 5-ASA on PLD activity, we examined the effect of bypassing the need for PLD activity on both mTOR signalling and cellular proliferation by adding exogenous PA to the culture medium. We found that exogenous PA was able to rescue the effect of 5-ASA on phosphorylation of mTOR, S6 and PRAS40 (figure 6F) without affecting PLD2 protein levels. PA addition had little effect on 4E-BP1-phosphorylation (figure 6F). Importantly, exogenous PA completely rescued the inhibitory effect of 5-ASA on cancer cell proliferation (figure 6G). In conclusion, we not only find that 5-ASA is a potent PLD inhibitor, but further show that both inhibition of colorectal cancer cell proliferation as well as downstream mTOR signalling can be reversed by supplying the cells with exogenous PA. These results suggest that the mechanism of action of 5-ASA on cellular proliferation is via inhibition of PLD-dependent PA production and is, at least partly, mediated through reduction of mTOR signalling.

Discussion

Although 5-ASA has been a cornerstone in the treatment of patients with ulcerative colitis for the last 30 years, its mechanism of action remains poorly defined. Observational studies suggested that 5-ASA may have chemopreventive activities, and it has been shown by several groups using different colorectal cancer cell lines that 5-ASA induces cell cycle arrest.2–5 The molecular mechanism that underlies this 5-ASA induced cell cycle arrest has, however, not been resolved. Here we show that 5-ASA is an inhibitor of PLD and downstream mTOR signalling in CRC cell lines. We find that inhibition of mTOR signalling by 5-ASA occurs independently of the TSC. Our data show that 5-ASA inhibits PLD and thereby limits the availability of PA which is required to maintain mTOR activity. Bypassing the need for PLD activity by adding exogenous PA rescued both mTOR signalling and the effects of 5-ASA on colon cancer cell proliferation.

5-ASA is a drug that is used as both an oral and topical treatment. Oral treatment is 2–4 g/day. Intracolonic luminal concentrations of 5-ASA in patients in remission that are on an oral dose of 2 g/day of 5-ASA are around 20 mM,21 and concentrations in the left sided colon and rectum can be significantly enhanced by combining oral with topical treatment.22 The luminal concentrations reached in patients correlate well with the dose at which various groups have observed, among others, inhibition of leucocyte migration,23 mast cell activation24 and inhibition of colon cancer cell proliferation in vitro.2–5 In our hands the effects of 5-ASA on colorectal cancer cells start at a dose of around 20 mM and are maximal at a dose of 40 mM, at which we observe inhibition of PLD activity and reduced mTOR phosphorylation. The effects of 5-ASA on colon cancer cell proliferation correlated with inhibition of downstream mTOR signalling. However, to our surprise we observed no changes in the phosphorylation status in any of the pathways that act upstream of the TSC complex, nor did knockdown of one of the components of the TSC complex affect either mTOR phosphorylation or the effect of 5-ASA on mTOR phosphorylation. This suggests not only that the effect of 5-ASA on mTOR is TSC-independent but also that, in colorectal cancer cells, mTOR may be activated below the level of the TSC.

Alternative TSC-independent mTOR activation can occur via PLD-generated PA. PA is an important signalling molecule that has been suggested to activate mTOR by direct interaction with the same domain that is targeted by the mTOR inhibitor rapamycin.15 ,16 Furthermore, PLD itself has also been found to transduce signals via direct interaction with and lipase activity-dependent activation of mTOR.25 A role for PLD-dependent mTOR signalling in colorectal cancer cells was suggested by our finding that both downstream mTOR signalling and proliferation were inhibited by either treatment with 1-butanol (which results in the generation of phosphatidylalcohol rather than PA) or knockdown of PLD2. The observations that 5-ASA inhibited PLD enzymatic activity and that the effects of 5-ASA on mTOR-signalling and cancer cell proliferation could be rescued by the addition of exogenous PA, further suggest a role for PLD/PA in the mechanism of action of 5-ASA.

mTOR,15 PRAS4016 and p70S6K26 are all direct cellular targets of PA which can therefore activate the mTOR pathway at multiple levels, showing the potential of PLD/PA signalling to activate the mTOR pathway. It is important to realise that our finding that PA can reverse the effects of 5-ASA suggests that these effects are dependent on inhibition of PLD but not necessarily on mTOR signalling. Targets of PA, other than members of the mTOR signalling pathway, are likely to exist and their inhibition may contribute to the effects downstream of 5-ASA mediated PLD inhibition. For example, PLD2-generated PA has been found to be required for epidermal growth factor induced Ras activation.27 Additionally, although we found the effect of 5-ASA on mTOR phosphorylation and proliferation to be PPARγ-independent (see figure 3), the protective effect of 5-ASA on CRC development could additionally be regulated via this nuclear receptor and a PLD2-derived second messenger. A recent study showed that an alternative second messenger produced by PLD2, cyclic PA, was a direct inhibitor of PPARγ.28 Therefore, 5-ASA-treatment of CRC patients could potentially have parallel beneficial effects by increasing the anti-inflammatory action of PPARγ, combined with the simultaneous inhibition of mTOR and consequent reduction of cellular proliferation reported here; both via reduction of PLD-dependent (cyclic)PA production.

In contrast to the large amount of data that support an important role for mTOR signalling in cancer development,10 ,29 relatively little is known about the role of PLD signalling in cancer initiation or progression in vivo. PLD is activated by mitogenic stimuli, therefore it is likely to play an important role in oncogenic signalling in cancer.30 In vivo, increased PLD2 expression has been reported in a number of different cancer types, including colon cancer.31 In addition, immunohistochemical staining for PLD2 as presented in the Human Protein Atlas (http://www.proteinatlas.org), shows high PLD2 expression in the majority of colorectal cancer specimens. Although the role of PLD/PA in the 5-ASA induced mTOR inhibition in vivo is not formally proven in this study, these observations, combined with our data, suggest that PLD signalling may significantly contribute to colon cancer cell proliferation. Therefore, a role for PLD-dependent mTOR signalling in colon cancer development should be further investigated. Unfortunately, we were unable to do determine PLD activity and/or PA concentration in CRC tumour samples due to a lack of availability of sufficient quantities of biopsy material. Despite these caveats, we feel that the data presented in this study provide substantial evidence that the inhibitory action of 5-ASA on colon cancer cell proliferation is achieved via inhibition of PLD/PA-dependent mTOR signalling. It would be interesting to examine whether this same pathway plays a role in the anti-inflammatory actions of 5-ASA in the treatment of ulcerative colitis.

Acknowledgments

We would like to thank Annie van der Zon for technical assistance.

Footnotes

  • The authors BB and AAD contributed equally.

  • Funding The research leading to these results was sponsored by an unrestricted grant from Ferring BV.

  • Competing interests None.

  • Ethics approval Medical Ethical Committee of University of Leiden.

  • Provenance and peer review Not commissioned; externally peer reviewed.

References


Free sample
This recent issue is free to all users to allow everyone the opportunity to see the full scope and typical content of Gut.
View free sample issue >>

Don't forget to sign up for content alerts so you keep up to date with all the articles as they are published.