Advertisement
Research ArticleGASTROINTESTINAL, HEPATIC, PULMONARY, AND RENAL

Role of Constitutive Cyclooxygenase-2 in Prostaglandin-Dependent Secretion in Mouse Colon In Vitro

Wallace K. MacNaughton and Kelly Cushing
Journal of Pharmacology and Experimental Therapeutics May 2000, 293 (2) 539-544;

Abstract

The relative contributions of cyclooxygenase (COX)-1 and COX-2 in mediating prostaglandin (PG)-dependent chloride secretion were investigated in segments of mouse colon mounted in Ussing-type diffusion chambers. COX-2 mRNA and protein were constitutively expressed as shown by reverse transcription-polymerase chain reaction and Western immunoblot, respectively. COX-2 immunoreactivity was detected immunohistochemically in cells lying subjacent to the crypt epithelial cells. In segments of colon mounted in Ussing chambers, arachidonic acid caused a concentration-dependent increase in short-circuit current that was blocked by piroxicam, the COX-2 inhibitor NS-398, and the COX-1 inhibitor SC-560. Exposure to the PG-dependent secretagogue, bradykinin, also caused an increase in short-circuit current that was not blocked by piroxicam or SC-560, and only by the highest dose of NS-398. When incubated in the presence of 10 μM arachidonic acid, segments of mouse colon produced both PGE2 and PGD2. Synthesis of PGE2but not PGD2 was blocked by NS-398 and SC-560. These data demonstrate that both COX-1 and COX-2 are constitutively expressed in the mouse colon, and both contribute to PG-dependent electrolyte transport.

Prostaglandins (PGs) are produced from free arachidonic acid through the catalytic activity of two cyclooxygenase (COX) enzymes. Until recently, COX-1 was considered the constitutive isoform, being expressed in almost all tissues, whereas COX-2 was considered inducible, because its expression could be triggered by inflammatory cytokines such as interleukin-1β and tumor necrosis factor-α. This paradigm led to the belief that COX-1, which is found in abundance in the gastrointestinal tract, produced PGs that were cytoprotective and that subserved physiological functions such as control of mucosal blood flow and electrolyte secretion. In contrast, COX-2-derived PGs were considered deleterious because they were found at sites of inflammation (where COX-2 was induced) and were hence thought to trigger, participate in, or exacerbate the inflammatory response. This dogma has provided the impetus for the recent development of selective COX-2 inhibitors that would inhibit inflammatory COX-2 and spare cytoprotective COX-1 (Masferrer et al., 1994; Hawkey, 1999). Traditional nonsteroidal anti-inflammatory drugs are relatively nonselective and their chronic use in conditions such as rheumatoid arthritis is associated with a high incidence of serious adverse gastrointestinal events (Wallace and Granger, 1992).

Recently, however, evidence has mounted that clouds the “COX-1 is good, COX-2 is bad” paradigm. Several groups have reported constitutive expression of COX-2 in numerous tissues, including the kidney (Harris et al., 1999), female reproductive tract (Slater et al., 1994) and fetal membranes (Slater et al., 1999), central nervous system (Yasojima et al., 1999), and vascular endothelium (McAdam et al., 1999). We have shown in a carageenan-induced rat paw edema model that COX-2 inhibitors only exhibited anti-inflammatory effects at concentrations that have been shown to also inhibit COX-1 (Wallace et al., 1998). In addition, it has recently been shown that COX-2-derived PGs may exhibit anti-inflammatory properties (Gilroy et al., 1999). A further complicating observation is the demonstration that COX-1 can be induced under certain circumstances (Jun et al., 1999). Given these recent data, it is clear that the physiological roles of the cyclooxygenases need to be reevaluated.

In the intestinal tract, secretion of chloride ions and water into the lumen is considered, along with low epithelial permeability, part of a primary host defense mechanism preventing the translocation of luminal bacteria, bacterial products, and antigens into the mucosa (Wood, 1993). Permeability defects have been implicated in relapse of inflammatory bowel disease (Wyatt et al., 1993) and secretory function is suppressed in inflamed intestine (Goldhill et al., 1993; MacNaughton et al., 1998). It has long been known that the response of the epithelium to numerous secretagogues is regulated by PGs. PGE2, probably derived from subepithelial myofibroblasts, is a “final common mediator” of secretagogues encompassing various lipid mediators, cytokines, oxygen radicals, amines, and enzymes (Hinterleitner and Powell, 1991). It acts both directly on the enterocyte and via submucosal secretomotor neurons (Dekkers et al., 1997). In contrast, PGD2 is antisecretory, functioning to block secretion stimulated by submucosal secretomotor neurons of the enteric nervous system (Goerg et al., 1992). The balance between PGE2 and PGD2 is important in the regulation of epithelial electrolyte transport (Keenan and Rangachari, 1989). However, the COX isoform responsible for production of PGs involved in intestinal secretion is not known. Given the role of PGs in regulating this important host defense mechanism, and the increasing interest in selective COX-2 inhibitors, we sought to determine the relative roles of COX-1- and COX-2-derived PGs in mediating PG-dependent chloride secretion with a mouse model.

Experimental Procedures

Animals

Male C57Bl/6 mice (4–6 weeks of age) were obtained from Charles River (Montreal, Quebec, Canada) and were housed in wire-bottomed cages under constant temperature (22°C) and photoperiod (12-h light/dark cycle). They were allowed unrestricted access to standard mouse chow and water. Mice were allowed to acclimatize to these conditions for at least 5 days before inclusion in an experiment. All procedures involving animals were approved by the University of Calgary Animal Care Committee and were conducted according to the guidelines of the Canadian Council on Animal Care.

COX-1 and COX-2 Expression and Localization

Reverse Transcription-Polymerase Chain Reaction (RT-PCR).

Mice were euthanized by cervical dislocation. Colons were removed and rinsed in cold (4°C) Krebs buffer. Segments of tissue (50–100 mg) were placed in 1.0 ml TRIzol reagent (Life Technologies, Grand Island, NY) and homogenized with a polytron homogenizer (Brinkmann, Mississauga, Ontario, Canada). After homogenization, samples were left at room temperature for 5 min and then chloroform was added (0.2 ml/1 ml TRIzol). Samples were shaken by hand for 15 s, left at room temperature for 2 to 3 min, and then centrifuged at 12,000gfor 30 min at 4°C. A 500-μl aliquot of the aqueous phase was placed in a microfuge tube and isopropyl alcohol (0.5 ml/1 ml TRIzol) was added to precipitate the RNA. The tubes were then centrifuged at 12,000g for 20 min at 4°C to pellet the RNA. The supernatant was removed and the RNA was washed with 75% ethanol, vortexed, and centrifuged again. The ethanol was removed and the RNA pellet was air-dried before dissolving in RNase-free water. The ratio of RNA/DNA and the tRNA concentration of the sample were determined with a GeneQuant II nucleic acid analyzer (Pharmacia Biotech, Baie d'Urfé, Quebec, Canada).

The extracted RNA was then reverse transcribed by adding 2.0 μg of RNA per sample to a reaction mixture that contained 2 μl of 10× PCR buffer, 2 μl of mixed nucleotide triphosphates (10 mM), 2 μl of random nucleotide hexamers (N6; 900 pmol/μl), 0.5 μl of RNA guard (27 U), 11 μl of double-distilled water, and 1.5 μl of Superscript reverse transcriptase (300 U). Samples were placed in a thermocycler (Barnstead Thermolyne, Chicago, IL) and reverse transcribed at 42°C for 50 min.

For PCR, 2 μl of the reverse-transcribed cDNA was placed in a tube that contained 2 μl of 2 mM mixed nucleotide triphosphates, 5 μl of 10× PCR buffer, and 35 μl of double-distilled water. Two microliters of each of the 5′ and 3′ primers (Table1) were added for COX-1 and COX-2 and GAPDH. GAPDH was used as an internal control. Taq polymerase was added with a hot start to reduce nonspecific binding. The samples were cycled as follows: 94°C for 1 min, 55°C for 30 s, and 72°C for 1 min. Optimal amplification was achieved at 32 cycles for COX-1 and COX-2 and 22 cycles for GAPDH. The PCR products were run on a 1% agarose gel containing ethidium bromide. After separation the bands were visualized under UV light with a Gel Doc 2000 (Bio-Rad, Mississauga, Ontario, Canada), and analyzed with Quantity One software (Bio-Rad).

View this table:
Table 1

Sequences of primers for PCR

Western Immunoblot.

Mice were euthanized by cervical dislocation. Colons were removed and rinsed in PBS consisting of 137 mM NaCl, 8.16 mM Na2HPO4, 1.47 mM KH2PO4, and 3.22 mM KCl. Samples were homogenized in 1 ml of suspension buffer (0.1 M NaCl; 0.01 M Tris-HCl, pH 7.6; 0.001 M EDTA, pH 8.0; 1 μg/ml aprotinin; 100 μg/ml phenylmethylsulfonyl fluoride; and 1% Triton X-100) and centrifuged at 12,000g for 5 min at room temperature. Half of the supernatant was removed and added to an equal volume of sample buffer (catalog no. S-3401; Sigma Chemical Co., Mississauga, Ontario, Canada). The other half of the sample was used to determine protein concentration with a protein assay kit (Bio-Rad). Samples were boiled for 10 min and the DNA was sheared by sonication. Samples were then centrifuged at 12,000g for 5 min at room temperature and the supernatant run on an 8% acrylamide gel in a mini trans-blot electrophoretic transfer cell (Bio-Rad). Protein was transferred to nitrocellulose overnight at 30 V and 4°C. After transfer, the membrane was blocked in 5% milk for 1 h and washed briefly in PBS-Tween. The COX-1 or COX-2 primary antibodies (goat anti-mouse; Santa Cruz Biotechnologies, Santa Cruz, CA) were diluted 1:75 in 1% milk/PBS-Tween. The membrane was placed in the primary antibody for 1.5 h, and then washed five times in PBS-Tween. The secondary antibody (rabbit anti-goat IgG conjugated to horseradish peroxidase; Santa Cruz Biotechnologies) was diluted 1:10,000 in 1% milk/PBS-Tween. The membrane was placed in the secondary antibody for 1 h, and then washed five times in PBS-Tween. Luminol reagent (Santa Cruz Biotechnologies) was applied to the membrane for 2 min. The membrane was then exposed to X-ray film to visualize the proteins. The X-ray film was analyzed with a GS-710 calibrated imaging densitometer (Bio-Rad) and Quantity One software (Bio-Rad).

Immunohistochemistry.

Segments of mouse colon were removed and fixed in Zamboni's fixative overnight. They were then washed in PBS (three times) and then cryoprotected in PBS containing 20% (w/v) sucrose. Sections (12–14 μm) were cut with a cryostat (Microm GmbH, Heidelberg, Germany) and placed onto poly(d-lysine)- coated microscope slides. Sections were washed three times in PBS and were incubated with a rabbit anti-COX-2 primary antibody at 4°C overnight. Sections were then rinsed with PBS and incubated with donkey anti-rabbit IgG CY3-conjugated secondary antibody for 90 min. They were then washed three times in PBS, coverslipped, and viewed under epifluorescence with a Zeiss Axioplan microscope. Images were captured and digitized with a charge-coupled device video camera and Northern Exposure imaging software (Carsen Vision, Edmonton, Alberta, Canada).

Electrolyte Transport Studies In Vitro

Mice were sacrificed by cervical dislocation and segments of colon removed and gently flushed with cold Krebs' buffer (4°C). The segments were opened longitudinally along the mesenteric border and mounted on standard Ussing-type diffusion chambers (Navicyte Inc., Sparks, NV). Tissues were bathed on both the mucosal and serosal surfaces with Krebs' buffer containing 10 mM d-glucose. The bathing solutions were maintained at 37°C and pH 7.4 and were aerated with 5% CO2, 95% O2. The transepithelial potential difference was maintained at zero volts by applying a short-circuit current (Isc) with a voltage clamp apparatus (EVC4000; World Precision Instruments, Sarasota, FL). TheIsc was taken as the measure of net active electrolyte transport by the colonic epithelium. Tissues were allowed to equilibrate with respect to Isc for 15 to 20 min. Drugs were applied to the serosal bath.

Materials

Routine chemicals were obtained from BDH Chemicals (Toronto, Ontario, Canada). NS-398 was obtained from Research Biochemicals Inc. (Natick, MA). SC-560 was a generous donation from Dr. Frank Degner, Boehringer-Ingelheim GmbH. Piroxicam was obtained from Sigma Chemical Co. Other chemicals were obtained as indicated in the text.

Statistical Analysis

Data are expressed as the mean ± SE. Comparisons among groups were made by one-way ANOVA with post hoc Newman-Kuels test with Instat version 3.00 (GraphPad Software, San Diego, CA). A Pvalue of <.05 was considered to be significant.

Results

Expression of COX-2 in Mouse Colon.

COX-1 and COX-2 mRNA expression was observed in full-thickness segments of mouse colon (n = 3), as demonstrated by RT-PCR (Fig.1). Western blot analysis revealed COX-1 and COX-2 proteins in full-thickness segments of mouse colon (n = 3; Fig. 1).

Figure 1
Figure 1

Constitutive expression of COX-1 and COX-2 in mouse colon. A, results (reversed image) of RT-PCR assay for COX-1 (left) and COX-2 (right) mRNA expression in segments of colon from three C57BL/6 mice. B, COX-1 (left) and COX-2 (right) protein expression as shown by Western immunoblot. Positive immunoreactivity was seen in segments of colon from three mice.

COX-2 immunoreactivity (COX-2-IR) in normal mouse colon is shown in Fig. 2. COX-2-IR was not observed in the epithelium but was evident in cells subjacent to the epithelium. This microanatomical localization is consistent with, but not proof of, identification of these cells as myofibroblasts. Furthermore, COX-2-IR was observed in the muscularis externa, particularly in the longitudinal muscle layer. Routine histology indicated that there was no ongoing chronic inflammation; there was no evidence of an inflammatory cell infiltrate, mucosal or submucosal edema, or muscle thickening (data not shown). Clinically, the animals exhibited no symptoms such as diarrhea.

Figure 2
Figure 2

COX-2 immunoreactivity in mouse colon. A, full thickness cross-section of mouse colon. Mucosal staining was only observed in subepithelial cells (arrowheads) at the bases of the crypts (Cr). Positive staining also was observed in the longitudinal muscle layer. B, section of mouse colon through mucosal crypts (Cr). Staining in subepithelial cells is observed (arrowheads). Scale bar, 50 μm.

Prostaglandin-Dependent Secretion.

Concentration-response relationships were established for the PG-dependent secretagogues bradykinin (1 nM–1 μM; Musch et al., 1983) and arachidonic acid (1–100 mM; Field et al., 1984). Both caused concentration-dependent increases in Isc in unstripped segments of mouse colon mounted in Ussing chambers (Fig.3). The response to arachidonic acid occurred within 1 min, peaked within 5 min, and either remained elevated or returned to baseline only slowly. TheIsc remained elevated above baseline for the duration of the experiment. The response to bradykinin also began within 1 min, peaked within 3 min, and then returned to baseline over the next 5 min.

Figure 3
Figure 3

Isc response (ΔIsc) of isolated segments of mouse colon to serosal application of arachidonic acid (○) or bradykinin (●). Both caused a concentration-dependent increase in short-circuit current.

The effect of selective inhibitors of COX-1 and COX-2 on the responses to arachidonic acid and bradykinin were examined in vitro. Doses of the selective COX-2 inhibitor NS-398 and the selective COX-1 inhibitor SC-560 were chosen based on published pharmacological data demonstrating their selectivity in in vitro systems (Futaki et al., 1994; Smith et al., 1998). A high dose (100 μM) of the nonselective inhibitor piroxicam was used to block both COX-1 and COX-2. The selective COX-2 inhibitor NS-398 caused a concentration-dependent inhibition of the response to arachidonic acid that was almost complete at 30 μM (Figs. 4 and5). The secretory response to arachidonic acid also was blocked by the selective COX-1 inhibitor SC-560 at nanomolar concentrations and by the nonselective COX inhibitor piroxicam (Fig. 5).

Figure 4
Figure 4

A, representative trace (top) of the effects of NS-398 (30 μM) on Isc response of mouse colon to 10 μM arachidonic acid (AA), and response to 10 μM AA in the absence of NS-398 (bottom). B, NS-398 (30 μM) was less effective in reducing the secretory response to 10 nM bradykinin.

Figure 5
Figure 5

Effect of pretreatment with COX inhibitors onIsc response (ΔIsc) of isolated segments of mouse colon to 10 μM arachidonic acid (AA, top) or to 10 nM bradykinin (BK, bottom). Con, vehicle control; Piro, 100 μM piroxicam. **P < .01, ***P < .001 versus control.

Interestingly, the response to 10 nM bradykinin was not blocked by piroxicam (100 μM), NS-398 (1–3 μM), or SC-560 (0.1–1 μM). Only NS-398 at 30 μM resulted in a significant reduction in the response to bradykinin (Fig. 5).

Prostaglandin Synthesis.

To determine the degree to which the COX inhibitors blocked PG synthesis, further experiments were conducted in which whole-thickness segments of mouse colon were incubated at 37°C in Krebs buffer containing 10 μM arachidonic acid or 10 nM bradykinin plus piroxicam (100 μM), NS-398 (1–30 μM), or SC-560 (0.1–1 μM). These concentrations of arachidonic acid and bradykinin stimulated chloride secretion in Ussing chamber studies (see above). Controls were incubated with vehicle, arachidonic acid, or bradykinin only. After 10 min, the tissues were removed and processed for measurement of PGs as previously described (Wallace et al., 1990). Tissue levels of PGE2 and PGD2 were determined with commercial enzyme-linked immunosorbent assay kits (Cayman Chemical, Ann Arbor, MI). The levels of PG synthesis in the presence of the inhibitors were expressed as a percentage of the maximal response, in the presence of arachidonic acid or bradykinin, to take interassay variation into account. In the presence of arachidonic acid, piroxicam almost completely blocked PGE2 synthesis (Fig.6). NS-398 caused a concentration-dependent inhibition of PGE2 that was almost complete at 30 μM. SC-560 also reduced PGE2 synthesis in a dose-dependent manner, but did not cause more than an ∼55% inhibition (Fig. 6). In the same tissues, PGD2 synthesis was significantly inhibited by 100 μM piroxicam but not by NS-398 (1–30 μM) or SC-560 (0.1–1 μM; Fig. 6). Bradykinin did not stimulate PGE2 or PGD2 synthesis compared with unstimulated controls (Fig.7). NS-398 (30 μM) and SC-560 (300 nM) reduced PGE2 synthesis in the presence of bradykinin. Neither inhibitor significantly affected PGD2 synthesis. Piroxicam induced a significant reduction in PG synthesis compared with controls (Fig. 7).

Figure 6
Figure 6

PG synthesis in isolated segments of mouse colon incubated with 10 μM arachidonic acid for 10 min. Data are expressed as a percentage of the PG synthesis observed without addition of inhibitors. Top, PGE2 synthesis; bottom, PGD2synthesis. **P < .01, ***P < .001 versus control.

Figure 7
Figure 7

PG synthesis in isolated segments of mouse colon incubated with 10 nM bradykinin for 10 min. Data are expressed as a percentage of the PG synthesis observed without addition of bradykinin. ▪, PGE2; ■, PGD2. *P < .05, ***P < .001.

Discussion

In this study, we have demonstrated the constitutive expression of both COX-1 and COX-2 in the mouse colon, and have provided evidence that both COX-1 and COX-2 contribute to PG-dependent secretion in vitro. It has now been widely demonstrated that COX-2 is constitutively expressed in a variety of tissues, including brain, kidney, female reproductive tract, fetal tissues, and vascular endothelium. As expected, COX-1 was constitutively expressed in mouse colon. However, we now demonstrate constitutive COX-2 expression in mouse colon. In each mouse studied, COX-2 mRNA and protein were expressed. COX-2 expression was not due to ongoing inflammation because none of the mice exhibited clinical, macroscopic, or microscopic signs of colonic inflammation. Of note, we also have shown constitutive colonic expression of the inducible isoform of nitric oxide synthase in the absence of inflammation (MacNaughton et al., 1998). Based on our immunohistochemical data, it appears that COX-2 is expressed primarily in cells lying subjacent to the crypt enterocytes. The subepithelial myofibroblasts have long been known to be a source of PGs that regulate epithelial function (Hinterleitner and Powell, 1991) and are localized to this region of the mucosa. However, the isoform of COX present in these cells, and which is responsible for the production of secretory PGE2, is not known. Our data are suggestive of, but not proof of, subepithelial myofibroblasts as the cellular source of COX-2-derived PGs regulating epithelial secretory function.

Both arachidonic acid and bradykinin caused concentration-dependent increases in short-circuit current in segments of mouse colon mounted in Ussing chambers. Both of these secretagogues have been reported to stimulate chloride secretion through the synthesis of PGs (Musch et al., 1983; Field et al., 1984). Pretreatment of the tissue with the nonselective COX inhibitor piroxicam almost abolished the response to arachidonic acid, but failed to significantly inhibit the response to bradykinin. The response to arachidonic acid was significantly and concentration-dependently blocked by the COX-2 inhibitor NS-398, suggesting that a substantial portion of arachidonic acid-mediated secretion was through a COX-2-dependent mechanism. Similarly, the response to arachidonic acid was reduced by pretreatment with the selective COX-1 inhibitor SC-560. That the response could be blocked by both selective COX-1 and COX-2 inhibitors was curious. The most obvious explanation would be that one or both of these inhibitors were acting nonselectively. However, care was taken to use concentrations of NS-398 and SC-560 that have previously been shown to be selective for COX-2 and COX-1, respectively (Futaki et al., 1994; Smith et al., 1998). NS-398 at 100 μM did not block the activity of COX-1 isolated from sheep seminal vesicles, whereas the IC50 of NS-398 against COX-2 isolated from sheep placenta was 3.8 μM (Futaki et al., 1994). Although action against isolated, purified enzyme may not equate to activity in tissue, 30 μM NS-398, the highest concentration used herein, has been reported to be a selective COX-2 inhibitor in rat aortic organ culture (Hamilton et al., 1999). SC-560 demonstrated a 1000-fold higher selectivity for human recombinant COX-1 than for human recombinant COX-2 in vitro (Smith et al., 1998).

The lack of effect of COX inhibition on the secretory response to bradykinin was unexpected, given the previous reports of the PG-dependence of this response (Musch et al., 1983; Field et al., 1984). A possible explanation may be that bradykinin was added to the tissue in the presence of arachidonic acid. The availability of arachidonic acid has been shown to significantly affect the ability of COX-2 inhibitors to block COX-2 (Hamilton et al., 1999). In our study, only the highest dose of NS-398 (30 μM) significantly blocked the response to bradykinin. However, the ability of the COX inhibitors to block the response to arachidonic acid itself in these preparations argues against an effect of arachidonic acid on their ability to block the subsequent response to bradykinin. It may be that in the mouse colon, there is a substantial PG-independent component to the secretory response, and this is confirmed by the experiments showing a lack of effect of bradykinin on PG synthesis, at a concentration that elicited a change in Isc. Indeed, it has been reported that there is a substantial neural component to the response to bradykinin in rat (Perkins et al., 1988) and dog (Rangachari et al., 1993) colons. Our data support a PG-independent response to bradykinin in mouse colon, however, determining the exact mechanism of action was beyond the scope of this study.

To further characterize the isoforms of COX present in the mouse colon, experiments were conducted in which isolated, full-thickness segments of colon were incubated in the presence of arachidonic acid or bradykinin and either NS-398, SC-560, or piroxicam, with subsequent measurement of tissue levels of PGE2 and PGD2. Arachidonic acid induced PG synthesis. NS-398 caused a concentration-dependent decrease in the synthesis of PGE2, with almost complete inhibition at 30 μM. Similarly, piroxicam at 100 μM abolished PGE2synthesis. SC-560 only reduced PGE2 synthesis to ∼55% of control. This is interesting in that these same doses of SC-560 almost abolished arachidonic acid-induced chloride secretion. These data suggest that there may be a PG-independent inhibitory effect of SC-560 on arachidonic acid-induced chloride secretion in this in vitro model. Interestingly, bradykinin, at a dose that stimulated chloride secretion, did not stimulate PG synthesis. This supports our contention (see above) that bradykinin-induced secretion in the mouse colon is dependent on mechanisms other than PGE2, and represents a significant species difference compared with guinea pig and rabbit.

Although NS-398 and SC-560 inhibited PGE2synthesis, they did not significantly block the synthesis of PGD2. The synthesis of PGD2was, however, almost completely blocked by 100 μM piroxicam. The differential effect of the COX inhibitors on PGE2and PGD2 synthesis was not expected because both COX-1 and COX-2 catalyze the production of PGH2, the precursor for both PGE2 and PGD2. The results suggest that the effect of NS-398 in particular may be a combination of inhibition of COX-2 and a further downstream step in the synthesis of PGE2from PGH2. However, that NS-398 did not affect the chloride secretory response to serosally applied PGE2 (data not shown) argues against this conclusion. Further studies will be required to determine the exact mechanism of action in this regard. That these compounds can block PGE2 synthesis, although leaving PGD2 synthesis relatively untouched, has important physiological implications for mucosal function. PGE2 and PGD2 have important interactions regulating colonic secretion (Keenan and Rangachari, 1989). In particular, PGD2 is clearly antisecretory (Goerg et al., 1992), so a shift in the balance of eicosanoids in favor of PGD2 would lead to a suppression of secretion.

Constitutive expression of COX-2 in the intestinal mucosa, and its ability to produce secretory PGs, argues for an important role of this isoform of COX in the host's response to infections. Recently, Beubler et al. (1999) demonstrated that NS-398 and DFU (another COX-2 inhibitor), given at doses that are selective for COX-2, could block the large secretory response to cholera toxin in rat jejunum in vivo. Interestingly, the effect was not altered by pretreatment with dexamethasone, suggesting that constitutive, and not induced, COX-2 mediated the response to cholera toxin (Beubler et al., 1999). Given the importance of the secretory response as part of the mucosal host defense mechanism (Wood, 1993), the effects of COX-2-inhibitors on eicosanoid balance could have important clinical implications.

Acknowledgments

We thank Dr. Keith Sharkey, University of Calgary, for assistance with immunohistochemistry.

Footnotes

  • Send reprint requests to: Wallace MacNaughton, Ph.D., Department of Physiology and Biophysics, University of Calgary, 3330 Hospital Dr. NW, Calgary, Alberta, Canada T2N 4N1. E-mail:wmacnaug{at}ucalgary.ca

  • ↵1 This study was supported by a New Investigator Establishment grant from the Canadian Association of Gastroenterology and the Medical Research Council of Canada.

  • ↵2 W.K.M. is an Alberta Heritage Foundation for Medical Research Scholar.

  • Abbreviations:
    PG
    prostaglandin
    COX
    cyclooxygenase
    RT-PCR
    reverse transcription-polymerase chain reaction
    Isc
    short-circuit current
    IR
    immunoreactivity
    • Received August 2, 1999.
    • Accepted January 12, 2000.

References

  1. ↵
    1. Beubler E,
    2. Schuligoi R,
    3. Peskar BA
    (1999) Constitutive but not inducible COX-2 contributes to cholera toxin-induced fluid secretion in the rat colon (Abstract). Gastroenterology 116:A862.
  2. ↵
    1. Dekkers JAJM,
    2. Kroese ABA,
    3. Keenan CM,
    4. MacNaughton WK,
    5. Sharkey KA
    (1997) Prostaglandin E2 activation of VIP secretomotor neurons in the guinea pig ileum. J Auton Nerv Syst 66:131–137.
  3. ↵
    1. Field M,
    2. Musch MW,
    3. Miller RL,
    4. Goetzl EJ
    (1984) Regulation of epithelial electrolyte transport by metabolites of arachidonic acid. J Allergy Clin Immunol 74:382–385.
  4. ↵
    1. Futaki N,
    2. Takahashi S,
    3. Yokoyama M,
    4. Arai I,
    5. Higuchi S,
    6. Otomo S
    (1994) NS-398, a new anti-inflammatory agent, selectively inhibits prostaglandin G/H synthase/cyclooxygenase (COX-2) activity in vitro. Prostaglandins 47:55–59.
  5. ↵
    1. Gilroy DW,
    2. Colville-Nash PR,
    3. Willis D,
    4. Chivers J,
    5. Paul-Clark MJ,
    6. Willoughby DA
    (1999) Inducible cyclooxygenase may have anti-inflammatory properties. Nat Med 5:698–701.
  6. ↵
    1. Goerg KJ,
    2. Wanitschke R,
    3. Diener M,
    4. Rummel W
    (1992) Inhibition of neuronally mediated secretion in rat colonic mucosa by prostaglandin D2. Gastroenterology 103:781–788.
  7. ↵
    1. Goldhill JM,
    2. Burakoff R,
    3. Donovan V,
    4. Rose K,
    5. Percy WH
    (1993) Defective modulation of colonic secretomotor neurons in a rabbit model of colitis. Am J Physiol 264:G671–G677.
  8. ↵
    1. Hamilton LC,
    2. Mitchell JA,
    3. Tomlinson AM,
    4. Warner TD
    (1999) Synergy between cyclo-oxygenase-2 induction and arachidonic acid supply in vivo: Consequences for nonsteroidal antiinflammatory drug efficacy. FASEB J 13:245–251.
  9. ↵
    1. Harris RC,
    2. Wang JL,
    3. Cheng HF,
    4. Zhang MZ,
    5. McKanna JA
    (1999) Prostaglandins in macula densa function. Kidney Int 67:S49–S52.
  10. ↵
    1. Hawkey CJ
    (1999) COX-2 inhibitors. Lancet 353:307–314.
  11. ↵
    1. Hinterleitner TA,
    2. Powell DW
    (1991) Immune system control of intestinal ion transport. Proc Soc Exp Biol Med 197:249–260.
  12. ↵
    1. Jun SS,
    2. Chen Z,
    3. Pace MC,
    4. Shaul PW
    (1999) Estrogen upregulates cyclooxygenase-1 gene expression in ovine fetal pulmonary artery endothelium. J Clin Invest 102:176–183.
  13. ↵
    1. Keenan CM,
    2. Rangachari PK
    (1989) Eicosanoid interactions in the canine proximal colon. Am J Physiol 256:G673–G679.
  14. ↵
    1. MacNaughton WK,
    2. Lowe SS,
    3. Cushing K
    (1998) Role of nitric oxide in inflammation-induced suppression of secretion in a mouse model of acute colitis. Am J Physiol 275:G1353–G1360.
  15. ↵
    1. Masferrer JL,
    2. Zweifel BS,
    3. Manning PT,
    4. Hauser SD,
    5. Leahy KM,
    6. Smith WG,
    7. Isakson PC,
    8. Seibert K
    (1994) Selective inhibition of inducible cyclooxygenase 2 in vivo is antiinflammatory and nonulcerogenic. Proc Natl Acad Sci USA 91:3228–3232.
  16. ↵
    1. McAdam BF,
    2. Catella-Lawson F,
    3. Mardini IA,
    4. Kapoor S,
    5. Lawson JA,
    6. FitzGerald GA
    (1999) Systemic biosynthesis of prostacyclin by cyclooxygenase (COX)-2: The human pharmacology of a selective inhibitor of COX-2. Proc Natl Acad Sci USA 96:272–277.
  17. ↵
    1. Musch MW,
    2. Kachur JF,
    3. Miller RJ,
    4. Field M
    (1983) Bradykinin-stimulated electrolyte secretion in rabbit and guinea pig intestine. J Clin Invest 71:1073–1083.
  18. ↵
    1. Perkins MN,
    2. Forster PL,
    3. Dray A
    (1988) The involvement of afferent nerve terminals in the stimulation of ion transport by bradykinin in rat isolated colon. Br J Pharmacol 94:47–54.
  19. ↵
    1. Rangachari PK,
    2. Berezin M,
    3. Prior T
    (1993) Effects of bradykinin on the canine proximal colon. Regul Pept 46:511–522.
  20. ↵
    1. Slater D,
    2. Berger L,
    3. Newton R,
    4. Moore G,
    5. Bennett P
    (1994) The relative abundance of type 1 to type 2 cyclo-oxygenase mRNA in human amnion at term. Biochem Biophys Res Commun 198:304–308.
  21. ↵
    1. Slater D,
    2. Dennes W,
    3. Sawdy R,
    4. Allport V,
    5. Bennett P
    (1999) Expression of cyclo-oxygenase types-1 and -2 in human fetal membranes throughout pregnancy. J Mol Endocrinol 22:125–130.
  22. ↵
    1. Smith CJ,
    2. Zhang Y,
    3. Koboldt CM,
    4. Muhammad J,
    5. Zweifel BS,
    6. Shaffer A,
    7. Talley JJ,
    8. Masferrer JL,
    9. Seibert K,
    10. Isakson PC
    (1998) Pharmacological analysis of cyclooxygenase-1 in inflammation. Proc Natl Acad Sci USA 95:13313–13318.
  23. ↵
    1. Wallace JL,
    2. Bak A,
    3. McKnight W,
    4. Asfaha S,
    5. Sharkey KA,
    6. MacNaughton WK
    (1998) Cyclooxygenase 1 contributes to inflammatory responses in rats and mice: Implications for gastrointestinal toxicity. Gastroenterology 115:101–109.
  24. ↵
    1. Wallace JL,
    2. Granger DN
    (1992) Pathogenesis of NSAID gastropathy: Are neutrophils the culprits? Trends Pharmacol Sci 13:129–131.
  25. ↵
    1. Wallace JL,
    2. Keenan CM,
    3. Granger DN
    (1990) Gastric ulceration induced by nonsteroidal antiinflammatory drugs is a neutrophil-dependent process. Am J Physiol 259:G462–G467.
  26. ↵
    1. Wood JD
    (1993) Neuro-immunophysiology of colon function. Pharmacology 47 (Suppl 1) 7–13.
  27. ↵
    1. Wyatt J,
    2. Vogelsang H,
    3. Hübl W,
    4. Waldhöer T,
    5. Lochs H
    (1993) Intestinal permeability and the prediction of relapse in Crohn's disease. Lancet 341:1437–1439.
  28. ↵
    1. Yasojima K,
    2. Schwab C,
    3. McGeer EG,
    4. McGeer PL
    (1999) Distribution of cyclooxygenase-1 and cyclooxygenase-2 mRNAs and proteins in human brain and peripheral organs. Brain Res 830:226–236.
Back to top

In this issue

Download PDF
Article Alerts
Email Article
Citation Tools

Research ArticleGASTROINTESTINAL, HEPATIC, PULMONARY, AND RENAL

Role of Constitutive Cyclooxygenase-2 in Prostaglandin-Dependent Secretion in Mouse Colon In Vitro

Wallace K. MacNaughton and Kelly Cushing
Journal of Pharmacology and Experimental Therapeutics May 1, 2000, 293 (2) 539-544;
Twitter logo Facebook logo Mendeley logo

Jump to section

Advertisement