Entacapone promotes cAMP-dependent colonic Cl secretion in rats


Address for Correspondence
Jin-Xia Zhu, Department of Physiology, School of Basic Medical Sciences, Capital Medical University, Beijing, China.
Tel: +86 10 8391 1494; fax: +86 10 8391 1494;
e-mail: zhu_jx@ccmu.edu.cn


Background  Entacapone is a promising drug used widely for the treatment of Parkinson’s disease (PD) as a catechol-O-methyl transferase (COMT) inhibitor. However, entacapone has gastrointestinal side effects. The aim of this study was to investigate the effects of entacapone on the epithelial ion transport in rat distal colon, and explore the underlying mechanism.

Methods  The study was performed on freshly isolated colonic mucosa-only, submucosa-only and mucosa–submucosa preparations in rat. The short circuit current (ISC) was measured to determine electrogenic ion transport, and a scanning ion-selective electrode technique (SIET) was used to directly measure Cl flux across the epithelium. The content of intracellular cAMP was measured with radioimmunoassay (RIA).

Key Results  Entacapone increased mucosal ISC in the rat distal colon. ISC was inhibited significantly by apical addition of diphenylamine-2,2′-dicarboxylic acid (DPC), a blocker of the Cl channel, basolateral application of bumetanide, an inhibitor of Na+-K+-2Cl co-transporter (NKCC), removal of Cl from the bathing solution, and pretreatment with MDL 12330A, an inhibitor of adenylate cyclase. Inhibiting endogenous prostaglandin (PG) synthesis with indomethacin, and eliminating submucosal enteric neural activity with tetrodotoxin (TTX)-inhibited entacapone-evoked ISC increases. Similar results were also obtained when Cl flux was measured with SIET. Entacapone significantly increased intracellular cAMP content, which was greatly inhibited by either indomethacin or TTX in the tissues containing submucosal plexus, and by only indomethacin in the mucosa-only preparations.

Conclusions & Inferences  Entacapone stimulates cAMP-dependent Cl secretion in the rat colon, and this process is regulated by endogenous PG and the submucosal enteric nervous system.


Catechol-O-methyl transferase


Diphenylamine-2,2′-dicarboxylic acid


4,4′-diisothiocyanostilbene-2,2′-disulfonic acid


Dimethyl sulphoxide


Short circuit current


Na+-K+-2Cl cotransporter


Scanning ion-selective electrode technique


Parkinson’s disease








Parkinson’s disease (PD) is the second most common neurodegenerative disorder, and is characterized by the progressive loss of dopaminergic nigral neurons and striatal dopamine, leading to the motor symptoms of the disease. When 80% of dopamine is lost, PD motor symptoms occur.1–3 Levodopa is a dopamine precursor that passes through the blood–brain barrier. In the early stages of PD, the therapeutic reaction to levodopa is generally excellent. Levodopa associated with a dopa-decarboxylase inhibitor (DDCI) is considered a gold standard of PD pharmacotherapy.4,5 However, the beneficial early reaction to levodopa therapy generally continues for only ∼5 years.6,7 With progression to the advanced stage of the disease, the reaction to levodopa therapy becomes abrupt and short-lived, and motor fluctuations are frequent.

Entacapone is a selective, reversible inhibitor of catechol-O-methyl transferase (COMT) and is used as an adjunct to levodopa in the treatment of PD. Adding entacapone to levodopa/DDCI preparations increases significantly the plasma half-life and bioavailability of levodopa, and provides more consistent plasma levodopa levels without deep troughs.8,9 In recent years, different animal models have been tested, and the results have suggested that entacapone is very effective for PD.10–12 In clinical trials, combined treatment with levodopa and entacapone has a notable effect on the control of dyskinesia. Entacapone can enhance the duration of the effect of each dose of levodopa, and can preferentially improve motor symptoms of PD patients.13,14 Compared with levodopa/carbidopa, treatment with entacapone/levodopa/carbidopa results in an improved quality of life in PD patients with mild, or minimal, non-disabling motor fluctuations.15–19

A good prospect with entacapone for PD is unquestionable. However, evidence suggests that entacapone-related side effects are also very frequent. Diarrhea is the most common gastrointestinal symptom, occurring with an incidence of 10% in patients receiving entacapone.20 Diarrhea occurs most often after 1–3 months of entacapone treatment, and when entacapone therapy is discontinued, diarrhea improves within a few days.21,22 The mechanism of this side effect is unknown. The colonic submucosal plexus plays an important role in the regulation of mucosal fluid and electrolyte transport.23–25 It senses mechanical or chemical information, then stimulates mucosa secretion by releasing neural transmitters, such as vasoactive intestinal polypeptide (VIP) and/or acetylcholine (ACh) receptor.26–28 We carried out the present study to investigate the effect of entacapone on colonic epithelial ion transport, and the roles of prostaglandins (PG) and the enteric nervous system. For this purpose, we measured the short circuit current (ISC), and employed the scanning ion-selective electrode technique (SIET), combined with radioimmunoassay (RIA).

Materials and methods

Chemicals and solutions

Entacapone was obtained from Orion Corporation (Espoo, Finland). Stock solutions (1 mmol L−1) of entacapone were prepared in water. Amiloride hydrochloride, bumetanide, indomethacin, tetrodotoxin (TTX), 4,4′-diisothiocyanostilbene-2,2′-disulfonic acid (DIDS), glibenclamide and MDL12330A were obtained from Sigma Chemical Company (St. Louis, MO, USA). Diphenylamine-2,2′-dicarboxylic acid (DPC) was obtained from Riedel-de Haen Chemicals (Hannover, Germany). Indomethacin, TTX, and bumetanide were prepared in dimethyl sulfoxide (DMSO). Final concentrations of DMSO never exceeded 0.1% (vol/vol). Preliminary experiments indicated that the vehicle did not alter any baseline electrophysiological parameters.

Krebs–Henseit solution (K–HS) had the following composition (mmol L−1): NaCl, 117; KCl, 4.7; CaCl2, 2.5; MgCl2, 1.2; NaHCO3, 24.8; KH2PO4, 1.2; glucose, 11.1. In Cl-free solution, NaCl, KCl, and CaCl2 were replaced by sodium gluconate, potassium gluconate, and calcium gluconate, respectively. The solution was saturated with 95% O2 and 5% CO2, pH 7.4, at 37 °C, by bubbling the gas mixture through the solution.

Tissue preparation

Animal protocols followed guidelines established by the NIH, and were approved by the Animal Care and Use Committee, Capital Medical University. Adult male Sprague–Dawley rats (Laboratory Animal Services Center, Capital Medical University) ranging in weight from 200 to 300 g had free access to standard rodent laboratory food and water until the day of the experiments. The animals were killed by cervical dislocation. The distal colon, defined as the 7-cm long segment proximal to the lymph node (typically situated 3 cm from the anus) was removed. The distal colon was divided into four segments, termed DC1 (adjacent to the lymph node), DC2, DC3, and DC4, respectively. They were cut along the mesenteric border into a flat sheet, and flushed with ice-cold K–HS solution. The tissue was pinned flat with the mucosal side down in a Sylgard-lined Petri dish containing the ice-cold oxygenated solution. The serosa, submucosa and muscularis were stripped with fine forceps to obtain a mucosa-only preparation. The serosa and muscularis propria were stripped to obtain a mucosa–submucosa preparation. The serosa, muscularis propria and mucosa-only were stripped to obtain a submucosa preparation of the colon. In colonic mucosa-only preparations, TTX (1 μmol L−1), a neuronal Na+ channel blocker, was added routinely to the basolateral side to abolish the effects of residual neural elements.

Short-circuit current measurement

Flat sheets of colonic mucosa-only or mucosa–submucosa preparations were mounted in modified Ussing chambers with a cross-sectional area of 0.5 cm2. The mucosal and serosal surfaces of the tissue were bathed with 5 mL K–HS by recirculation from a reservoir maintained at 37 °C during the experiments. The K–HS was bubbled with 95% O2 and 5% CO2 to maintain the pH of the solution at 7.4. Drugs were added directly to the apical or basolateral side of the epithelium. Responses were recorded continuously by computer. The transepithelial potential difference for every colonic sample was measured by Ag/AgCl reference electrodes (P2020S; Physiologic Instruments, San Diego, CA, USA) connected to a preamplifier that was in turn connected to a voltage-clamp amplifier VCC MC6 (Physiologic Instruments).29 Transepithelial resistance (Ω cm2) was measured by altering the membrane potential stepwise (−0.1 mV) and applying the Ohmic relationship. The changes in short-circuit current (ΔISC) were measured as an index of electrogenic ion transfer; they were calculated on the basis of the value before and after stimulation, and were normalized as the current per unit area of epithelium (μA cm−2) (The ISC value before stimulation was a mean of values measured over 15 min. The ISC value after stimulation was measured at the highest point in the ISC curve). Tissues were kept in Ussing chambers for about 30–40 min to stabilize the ISC before adding drugs. A positive ISC corresponds to the net electrogenic secretion of anions (such as Cl), or the net electrogenic absorption of cations (such as Na+). Experiments were normally repeated using colonic mucosa or mucosa–submucosal preparations from at least three rats to ensure that the data were reproducible.

Measurement of extracellular Cl-flux

The net Cl-flux was measured non-invasively, using SIET (SIET system BIO-001A; Younger USA, LLC, MA, USA).30 The tissue was kept horizontal, with the mucosa pointing downward; it was mounted in the Petri dish that was filled with K–HS, and the Cl microelectrode was placed on the basolateral side of the preparation. The principle of this method and the description of the instrument are detailed in Land and Collett.31 Measurements were performed at 37 °C. Ion-selective microelectrodes with an external tip diameter of ∼3 μm were manufactured and silanized with tributylchlorosilane. The electrodes were back-filled with 100 mmol L−1 KCl and then front-filled with a 30-μm column length of chloride ionophore I–cocktail A (Fluka, Buchs, Switzerland). The electrode was connected to a motion-controlled head stage via an Ag/AgCl wire, and the circuit was completed by placing a reference electrode (DRIFEF-2; WPI, Sarasota, FL, USA). The microelectrodes were calibrated prior to the Cl flux measurement, and only the electrodes with Nernstian slopes >−56 mV per decade were used.

The electrode was controlled to move with an excursion of 30 μm at a programmable frequency in the range of 0.3–0.5 Hz. This minimized the mixing of the bathing saline. The Cl fluxes were calculated by Fick’s law of diffusion: J = −[D × (dC/dX)], where J represents the net Cl flux (in pmol cm−1 s−1), D is the self-diffusion coefficient for Cl (in cm2 s−1), dC is the difference value of Cl concentrations between the two positions, and dX is the 30-μm excursion over which the electrode moved in our experiments. Data acquisition, preliminary processing, control of the three-dimensional electrode positioner, and stepper-motor-controlled fine focus of the microscope stage were performed with aset software (Younger Tech. Corp., LLC, Amherst, MA, USA).

cAMP measurement

Rat colonic mucosa–submucosal tissues (about 150 mg) were incubated in the chambers containing 5 mL K–HS solution for 30 min for equilibration. The tissues were pretreated with vehicle (0.9% NaCl), TTX or indomethacin for 5 min before adding entacapone (250 μmol L−1/500 μmol L−1, 15 min). At the end of the incubation, all samples in the tubes were frozen immediately in liquid nitrogen, then homogenized on ice in saline (0.9% NaCl), and centrifuged (10 620 g, 5 min). Intracellular levels of cAMP were measured by a commercial RIA kit (Beijing Huayin Institute of Biological Technology, Beijing, China).


The results are given as arithmetic mean ± standard error of the mean (SEM). n refers to the number of rats. The EC50 was calculated using the Graph Pad Prism software 4.0 package (GraphPad Software Inc., San Diego, CA, USA). Statistical analyses were performed by one-way anova, followed by the Newman–Keuls test or Student’s paired or unpaired t-test. P values <0.05 were assumed to denote a significant difference.


Entacapone-induced ISC response in colonic mucosa-only preparations

After an equilibration time of 30 min, the basal potential difference, the ISC, and transepithelial resistance (Rt) in the colonic mucosa were −1.1 ± 0.3 mV, 15.1 ± 2.2 μA cm−2 and 70.9 ± 5.8 Ω cm2, respectively (= 35). The colonic mucosa was pretreated with TTX to eliminate residual enteric neural activity before recordings began. Entacapone was applied to the apical (mucosal) side, and no effect was observed. Adding entacapone (250 μmol L−1) to the basolateral (serosal) side produced an upward deflection in ISC (Fig. 1A), which was accompanied by a significant decrease in transepithelial resistance from 71 ± 6 Ω cm2 to 47 ± 4 Ω cm2 (n = 13, < 0.01) (Fig. 1B). The increase in ISC induced by entacapone at 10, 100, 200, 500, and 1000 μmol L−1 was 16 ± 2, 32 ± 3, 48 ± 4, 78 ± 18, and 124 ± 7 μA cm−2 (n = 6), respectively (Fig. 1C). The increase was concentration-dependent, with an apparent EC50 of 250 μmol L−1 (n = 30) (Fig. 1D). The subsequent studies were, therefore, carried out with the basolateral addition of 250 μmol L−1 entacapone.

Figure 1.

 Entacapone-induced short circuit current (ISC) response in colonic mucosa-only preparations. (A) The ISC response evoked by luminal and basolateral addition of entacapone (ENT) (250 μmol L−1). (B) Transepithelial resistance (Rt) evoked by basolateral addition of ENT (250 μmol L−1) (n = 13). (C) The ISC response evoked by basolateral addition of different dosages of ENT (10, 100, 200, 500, 1000, and 5000 μmol L−1). (D) The concentration–response curve for ENT-induced total charge transferred in 15 min (μA cm−2), with an EC50 of 250 μmol L−1. Each data point was obtained from at least six individual experiments. Values are mean ± SE; **< 0.01.

Pretreatment with TTX did not significantly affect basal and entacapone-induced ISC in mucosa-only preparations (Fig. 2A), but reduced the entacapone-induced ISC response by 39%, from 51 ± 7 μA cm−2 to 31 ± 3 μA cm−2 (n = 9, < 0.01) in mucosa–submucosa preparations (Fig. 2B), suggesting that the enteric nervous system is involved in the formation of the entacapone-induced ISC in the rat distal colon.

Figure 2.

 The role of endogenous prostaglandin (PG) and channel blocker/transporter inhibitor in the entacapone-induced short circuit current (ISC) response. Effect of tetrodotoxin (TTX, 1 μmol L−1), or indomethacin (INDO) (10 μmol L−1) on the entacapone (ENT)-induced colonic ISC response in mucosa-only preparations (A), and mucosa–submucosa preparations (B) (n = 9). (C) Summary of the effects of pretreatment with the Na+ channel blocker, amiloride (10 μmol L−1), removal of extracellular Cl, the apical addition of the Cl channel blocker, diphenylamine-2,2′-dicarboxylic acid (DPC, 1 mmol L−1), and the basolateral application of bumetanide (100 μmol L−1) on the ENT-induced ISC response in colonic mucosa-only preparations (n = 7). Values are mean ± SE; *< 0.05, **< 0.01.

To investigate the role of endogenous PG in the entacapone-induced ISC response, indomethacin (10 μmol L−1), a cyclooxygenase (COX) inhibitor, was applied to the basolateral side. As shown in mucosa-only (Fig. 2A) and mucosa–submucosa preparations (Fig. 2B), pretreatment with indomethacin reduced the basal ISC by 40%, from 15 ± 2 μA cm−2 to 9 ± 2 μA cm−2 (n = 9, < 0.05), and the entacapone-induced ISC response by 49%, from 47 ± 6 μA cm−2 to 24 ± 2 μA cm−2 (n = 9, < 0.01) in mucosa-only preparations; Pretreatment with indomethacin reduced the basal ISC by 41%, from 17 ± 2 μA cm−2 to 10 ± 2 μA cm−2 (n = 9, < 0.05), and the entacapone-induced ISC response by 45%, from 51 ± 7 μA cm−2 to 28 ± 2 μA cm−2 (n = 9, < 0.01) in mucosa–submucosa preparations. These results suggest that PG is involved in the formation of basal ISC and entacapone-induced ISC in the rat distal colon.

It is well known that an upward deflection in ISC reflects a net electrogenic anion secretion (such as Cl), and/or cation absorption (such as Na+). To investigate the ionic nature of the current, and the channels/transporters that may be potentially involved, the following experiments were performed (Fig. 2C). Apical addition of amiloride (10 μmol L−1), a blocker of epithelial Na+ channels, or the Ca2+-dependent Cl channel blocker, DIDS (200 μmol L−1), did not reduce significantly the ISC response induced by entacapone. However, removal of Cl from the bathing solution, apical application of DPC (1 mmol L−1), a non-specific Cl channel blocker, or basolateral administration of bumetanide (100 μmol L−1), an inhibitor of the Na+-K+-2Cl cotransporter (NKCC), inhibited significantly the entacapone-induced ISC response by 84%, 76% and 70%, from 50 ± 2 to 8 ± 2 (n = 7, < 0.01), 12 ± 3 (n = 7, < 0.01) and 15 ± 4 μA cm−2 (n = 7, < 0.01), respectively. These results indicate that the entacapone-induced upward deflection of ISC could be the result of Cl secretion.

Entacapone-induced Cl-flux with SIET in colonic mucosa-only preparations

The scanning ion-selective electrode technique is a novel research methodology. The SIET can obtain a variety of information about ion activity with ion sensitive electrodes. We therefore chose this methodology to measure directly the entacapone-induced epithelial Cl transport.

As shown in Fig. 3, in the basal state, a very small and stable Cl-flux was recorded from Cl-sensitive electrodes positioned on the basolateral side of the colonic mucosa. Application of entacapone (250 μmol L−1) caused a drastic increase in the Cl flux into cells from the basolateral side of the colon mucosa (Fig. 3A), from 9 ± 3 to 195 ± 16 μmol cm−2 s−1 (n = 10, < 0.001). This flux was powerfully inhibited by DPC (1 mmol L−1) and bumetanide (100 μmol L−1), returning to 10 ± 3 and 17 ± 4 μmol cm−2 s−1, respectively (n = 10, < 0.001) (Fig. 3B). Similar results were obtained when the system was pretreated with DPC and bumetanide (Fig. 3C), where the entacapone-induced Cl flux was reduced significantly from 175 ± 9 to 19 ± 4 and 12 ± 3 μmol cm−2 s−1 (n = 13, < 0.001), respectively (Fig. 3D). Apical addition of the Ca2+-dependent Cl channel blocker, DIDS (200 μmol L−1), did not affect entacapone-induced Cl flux (n = 5, > 0.05).

Figure 3.

 Entacapone-induced Cl-flux measured by scanning ion-selective electrode technique (SIET). (A) Representative net Cl flux recording following the addition of entacapone (ENT) (250 μmol L−1), diphenylamine-2,2′-dicarboxylic acid (DPC, 1 mmol L−1), and bumetanide (100 μmol L−1) in colonic mucosa-only preparations. (B) Summary of the effects of DPC and bumetanide on the ENT-induced Cl flux (n = 10) in colonic mucosa-only preparations. (C) Representative recordings of the net Cl-flux induced by ENT (250 μmol L−1), and pretreatment with DPC or bumetanide in colonic mucosa-only preparations. (D) Summary of ENT-induced Cl-flux in mucosa-only tissues pretreated with DPC and bumetanide (n = 13). (E) Representative recordings of net fluxes of Cl induced by ENT (250 μmol L−1), and pretreatment with tetrodotoxin (TTX, 1 μmol L−1) or indomethacin (INDO) (100 μmol L−1) in colonic mucosa–submucosa preparations. (F) Summary of the effects of TTX and INDO on ENT-induced Cl flux (n = 15). Values are mean ± SE; ***P < 0.001.

Involvement of the enteric nervous system in entacapone-induced Cl-flux in colonic mucosa–submucosal preparations

To study the role of the submucosal enteric nervous system in entacapone-induced Cl flux, mucosa–submucosa preparations were used. As shown in Fig. 3E, F, the entacapone-evoked Cl flux in the mucosa–submucosa preparation, 296 ± 20 μmol cm−2 s−1 (n = 15), was much higher than that in the mucosa-only preparation, 195 ± 16 μmol cm−2 s−1 (Fig. 3A, n = 10). When mucosa–submucosa preparations were pretreated with TTX, the entacapone-induced Cl flux was reduced by 49%, from 296 ± 20 to 150 ± 12 μmol cm−2 s−1 (n = 15, < 0.001). Pretreatment with indomethacin inhibited the entacapone-induced Cl flux by 51%, which was similar to results obtained with ISC in mucosa-only preparations (Fig. 2A).

Involvement of a cAMP-dependent pathway in entacapone-induced Cl secretion in colonic mucosa preparations

It is well known that Cl channels can be Ca2+-dependent or cAMP-dependent. The cAMP-dependent channel plays a predominant role in mediating colonic Cl secretion in mammals. Pretreatment with the adenylate cyclase inhibitor, MDL-12330A (20 μmol L−1), inhibited the entacapone-induced ISC increase (Fig. 4A) by 71%, from 50 ± 5 to 14 ± 3 μmol cm−2 s−1 (Fig. 4B, n = 6, < 0.01) in mucosa-only preparations. Furthermore, intracellular cAMP levels were measured by RIA in mucosa-only (Fig. 4C), mucosa–submucosa (Fig. 4D), and submucosa-only preparations (Fig. 4E). The basal levels of intracellular cAMP in these preparations were 151 ± 9 pmol mg−1, 188 ± 12 pmol mg−1 and 220 ± 17 pmol mg−1, respectively (n = 6). Treatment with entacapone at 250 μmol L−1 increased significantly the cellular cAMP levels to 266 ± 39 pmol mg−1 in mucosa-only preparations, 287 ± 37 pmol mg−1 in mucosa–submucosa preparations, and 366 ± 31 pmol mg−1 in submucosa-only preparations. These levels were further increased following treatment with a higher dose of entacapone (500 μmol L−1). Pretreatment with indomethacin significantly inhibited entacapone-induced increase in intracellular cAMP in all three tissue preparations (Fig. 4C–E, n = 6). It is worth emphasizing that administration of TTX markedly reduced the entacapone-induced cellular cAMP increase in mucosa with submucosa preparations and submucosa-only preparations (Fig. 4D, E, n = 6). A stronger inhibitory effect of TTX on entacapone-induced increase in cAMP was observed at the higher dose of entacapone applied to the mucosa with submucosa preparations. However, entacapone-induced increase in cAMP in mucosa-only preparations was reduced by pretreatment with indomethacin, not TTX.

Figure 4.

 The cellular cAMP pathway involved in entacapone-induced Cl secretion. (A) Entacapone (ENT)-induced short circuit current (ISC) response in the absence and presence of MDL 12330A (20 μmol L−1). (B) Summary of the effects of MDL 12330A on the ENT-induced ISC response (n = 6). Summary of the effect of pretreatment with tetrodotoxin (TTX) and indomethacin (INDO) on the ENT-induced change in cellular cAMP content in colonic mucosa-only preparations (C, n = 6), mucosa–submucosa preparations (D, n = 6), and in submucosa-only preparations (E, n = 6), Values are mean ± SE; *P < 0.05; **P < 0.01.


The technology of ISC has long been used to quantify transepithelial ion transport.29 The technique, however, lacks chemical selectivity, and can only measure electrogenic ion transport. The SIET, a novel methodology that has overcome the shortcomings of ISC recording, can carry out measurements through specific ion-sensitive electrodes, such as Cl-sensitive electrodes.31,32 In the present study, we combined the technique of ISC recording of epithelial ion transport with the novel SIET measuring colonic Cl flux in fresh isolated colonic tissue of rats, and we demonstrated for the first time the stimulatory effect of entacapone on colonic Cl secretion.

The Cl secretion requires both the basolateral accumulation of Cl by NKCC, and apical exit through Cl channels. According to the present data from the ISC and SIET, entacapone-induced Cl secretion arises mostly from an electrogenic Cl transport mediated by the apical Cl channels and basolateral NKCC, since blocking Cl channels by DPC and inhibiting NKCC by bumetanide eliminated about 90% of the entacapone-induced Cl secretion as measured by ISC or by SIET. Moreover, entacapone-induced Cl-flux in SIET has a very similar kinetic waveform, including responses to drugs, compared with entacapone-induced Cl transport in ISC.

In the present study, entacapone-induced Cl secretion is predominantly mediated by cAMP as a second messenger, since pretreatment of mucosa–submucosa tissues with the adenylate cyclase inhibitor, MDL-12330A, inhibited entacapone-stimulated ISC increase. In addition, entacapone-induced Cl secretion was sensitive to the non-specific Cl channel blockers, DPC or glibenclamide (data not shown), but not sensitive to DIDS, a blocker of the Ca2+-dependent Cl channel.

Under normal circumstances, the colonic mucosa can secrete PG spontaneously, or by the action of exogenous stimulatory molecules. This, in turn, increases intracellular cAMP to further promote Cl secretion.33–35 Indomethacin, a cyclooxygenase inhibitor, can eliminate the role of endogenous PG in colonic secretion.36 In the present study, entacapone-induced Cl secretion was blocked by about 50% by indomethacin, either in ISC or in SIET, indicating that entacapone may also induce mucosal endogenous PG release from the mucosa with or without the submucosa. However, how entacapone causes PG release? We do not have a firm answer for it. We suspect that entacapone might stimulate the cyclooxygenase (e.g. COX-2) activity to increase PG synthesis in the rat colon. A lot of cell types in the intestine can release PGs, such as mast cells, immune cells, neurons, glial cells, epithelial cells, etc. Future study is needed to test the hypothesis.

Inhibiting COMT will reduce the degradation of local catecholamines, which will lead to an increase in dopamine and/or noradrenaline, and then will decrease the ISC response. However, entacapone, an inhibitor of COMT, did not decrease, but increased, the ISC response and Cl secretion through stimulating PG release and enteric neuronal activity. This might be one of the reasons why entacapone causes gastrointestinal disorders. We have reported previously that dopamine or noradrenaline induces ISC decrease mainly in the late distal colon (DC1, colorectum), where Na+ absorption is predominant. Entacapone, however, not only acts on the DC1, but also on the other parts, DC2, DC3 and DC4, where anion secretion rather than Na+ absorption is predominant. In the present study, entacapone-induced ISC response is unlikely related to its inhibitory effect on the COMT or/and catecholamine system.

Entacapone is able to evoke neurotransmitter release from secretomotor neurons, since pretreatment with TTX inhibited the entacapone-induced increase in Cl flux and cAMP increase in mucosa–submucosa preparations. It is reported that the largest population of submucosal neurons (about one-half) is made up of VIP-containing secretomotor neurons.37 Epithelial stimulation by VIP induces Cl secretion through increases in the cellular cAMP.38 Whether entacapone-induced cAMP-dependent Cl secretion is via VIP release from submucosal secretomotor neurons needs to be investigated further.

According to our present study and reports in the literature, the mechanism underlying entacapone-induced Cl secretion in the rat distal colon can be illustrated as in Fig. 5. Entacapone can increase intracellular cAMP level and ISC by evoking PG release and/or activating submucosal neurons. Prostaglandin can directly stimulate epithelial cAMP-dependent Cl secretion through paracrine action by binding to the PG receptors (PGE2) on the epithelial cells,37,39 and indirectly by activating the submucosal secretomotor neurons (e.g., VIP-neurons). Entacapone can also activate submucosal secretomotor neurons either directly or indirectly by stimulating other neurons that form synaptic connections with the secretomotor neurons.40,41 Activation of the secretomotor neurons (such as VIP-ergic neurons) can lead to cAMP-dependent Cl secretion.37,39,42 The present study demonstrates that at least a portion of the neuronal mediated entacapone-induced effects on cAMP does not rely on the production of PGs since TTX (inhibiting enteric neurons) caused greater inhibition on entacapone-induced cAMP production than indomethacin (inhibiting PG synthesis and release) in mucosa/submucosa preparations (Fig. 4D). Similarly, the PG-mediated entacapone-induced effects on cAMP could also happen without activating the enteric nervous system, e.g. in mucosa-only preparations (Fig. 4C). However, it has been widely proved that PGs can activate the submucosal neurons. Thus, the neural and PG-mediated entacapone effects on cAMP and ISC are not completely independent of each other. A stronger inhibitory effect of TTX on entacapone-induced increase in cAMP was observed at the higher dose of entacapone (500 μmol L−1) applied to the mucosa–submucosa preparations, which suggests that higher doses might differentially interact with the enteric nervous system and the non-neural elements with stronger effects on the neural elements.

Figure 5.

 An illustration of the working model of entacapone in the rat distal colon. Entacapone can stimulate mucosal and submucosal prostaglandin (PG) release. Prostaglandin can stimulate directly epithelial cAMP-dependent Cl secretion and/or stimulate indirectly epithelial chloride secretion by activating enteric secretomotor neurons that release neurotransmitters [such as vasoactive intestinal polypeptide (VIP)], which can lead to cAMP-dependent Cl secretion. Entacapone, as a chemical, may also activate directly or indirectly (by evoking the release of epithelial secretagogues) the submucosal neurons, which in turn cause the postsynaptic secretomotor neurons to release neurotransmitters (such as VIP), which lead to cAMP-dependent Cl secretion.

Intestinal epithelial Cl channels are considered to play an important role in the regulation of intestinal water and salt metabolism, and in maintaining normal physiological functions.43 Under the influence of a variety of factors, excessive Cl can be secreted into the intestinal tract.44 Water flux into the intestine lumen will follow the gradient caused by Cl secretion. The result can lead to secretory diarrhea. Another factor, the paracellular permeability of colonic epithelia, should also be taken into account in the mechanism of entacapone-induced side effects in PD patients, such as diarrhea, because entacapone can reduce significantly membrane resistance (Fig. 1B), especially at higher concentrations. In fact, large doses of entacapone are often taken by advanced PD patients.20

The rapid effect of entacapone on the colonic Cl secretion may not be related directly to the slow onset of diarrhea seen as a side-effect of treatment for Parkinson’s disease. However, the rapid effect of entacapone is mediated by endogenous PG and the enteric neural system in the present study. We found that entacapone stimulates colonic PG release and evokes submucosa neural activity; this observation provides certain clues for the possibility that treatment of Parkinson’s disease with entacapone can lead to diarrhea. Most patients with PD manifest constipation, but the mechanism is not clear. It is reported that PD itself can cause gastrointestinal nerve dysfunction, and anti-PD drugs can slow down bowel movements, causing constipation.3 Therefore, logically, PD patients taking entacapone may not have diarrhea immediately, but some time later. In addition, in our in vivo study (Li LS, Xu JD, Zhang Y, Zhu JX, unpublished data), the rats with intragastric administration of entacapone displayed diarrhea on the same day, but the diarrhea got worse 7 days later. There is also a report that diarrhea could occur rapidly in a few patients treated with entacapone.

Taken together, the present study demonstrates for the first time that entacapone is able to stimulate cAMP-dependent Cl secretion in the rat distal colon, in which the endogenous PG and submucosal enteric nervous system play an important role. The study provides certain scientific evidence for entacapone-induced GI disorder seen in the treatment of PD patients with entacapone.


This work was supported by the National Natural Science Foundation of China (No. 30971076, JX Zhu), Scientific Research Key Program of Beijing Municipal Commission of Education (KZ200910025003, JX Zhu) and Beijing Municipal Project for Developing Advanced Human Resources for Higher Education (PHR201007110, JX Zhu). We express our thanks to Dr. Yue Xu and Jin Song for their kindly help on the Cl-flux measurement with SIET in this study.

Conflict of interest and funding

The authors have no financial, consultant, institutional and other relationships that might lead to bias or conflict of interest.

Author contribution

JXZ designed the research study; LSL and LFZ performed the most of the research; JXZ, LSL, and JDX analyzed the data; JXZ and LSL wrote the paper; JDX, TJ, and HG contributed to the SIET experiments; YL, XFL, and YZ contributed to the part of ISC study.