Adrenergic activation of electrogenic K+ secretion in guinea pig distal colonic epithelium: desensitization via the Y2-neuropeptide receptor

Jin Zhang, Susan T. Halm, Dan R. Halm


Adrenergic activation of electrogenic K+ secretion in isolated mucosa from guinea pig distal colon was desensitized by peptide-YY (PYY). Addition of PYY or neuropeptide-Y (NPY) to the bathing solution of mucosae in Ussing chambers suppressed the short-circuit current (Isc) corresponding to electrogenic Cl secretion, whether stimulated by epinephrine (epi), prostaglandin-E2 (PGE2), or carbachol (CCh). Neither peptide markedly inhibited the large transient component of synergistic secretion (PGE2 + CCh). Sustained Cl secretory Isc was inhibited ∼65% by PYY or NPY, with IC50s of 4.1 ± 0.9 nM and 9.4 ± 3.8 nM, respectively. This inhibition was eliminated by BIIE0246, an antagonist of the Y2-neuropeptide receptor (Y2-NpR), but not by Y1-NpR antagonist BVD10. Adrenergic sensitivity for activation of K+ secretion in the presence of Y2-NpR blockade by BIIE0246 was (EC50s) 2.9 ± 1.2 nM for epi and 13.3 ± 1.0 nM for norepinephrine, approximately fourfold greater than in the presence of PYY. Expression of mRNA for both Y1-NpR and Y2-NpR was indicated by RT-PCR of RNA from colonic mucosa, and protein expression was indicated by immunoblot. Immunoreactivity (ir) for Y1-NpR and Y2-NpR was distinct in basolateral membranes of columnar epithelial cells in the crypts of Lieberkühn as well as intercrypt surface epithelium. Adrenergic nerves in proximity with crypts were detected by ir for dopamine-β-hydroxylase, and a portion of these nerves also contained NPYir. BIIE0246 addition increased secretagog-activated Isc, consistent with in vitro release of either PYY or NPY. Thus PYY and NPY were able to suppress Cl secretory capacity and desensitize the adrenergic K+ secretory response, providing a direct inhibitory counterbalance against secretory activation.

  • epinephrine
  • norepinephrine
  • ATP
  • adenosine
  • Y1-neuropeptide receptor
  • enteric nerves
  • β1-adrenergic receptor
  • β2-adrenergic receptor

the composition of luminal fluid in the gastrointestinal tract is modified by epithelial secretion to aid in the process of digestion and to maintain the system until the arrival of the next meal (7). As digestion progresses, the requirements on luminal composition alter with the type of nutrients present. In the colon, the primary nutrients available are short-chain fatty acids produced by bacterial fermentation of cellulose and other nondigested carbohydrates. A mode of ion secretion particularly evident in the distal colon is electrogenic K+ secretion, which can be stimulated via β-adrenergic activation (23, 27, 42, 52) from the sympathetic nervous system (17, 26, 45). This modulatory mode of secretion occurs with minimal sustained Cl secretion, which distinguishes it from the well-studied flushing mode that exhibits high rates of sustained electrogenic Cl secretion (25, 33, 39). A sympathetic response in the distal colon therefore includes a sustained electrogenic K+ secretion in concert with the slowing of motility and vasoconstriction that diverts blood away from the gut. The consequent slowing of the overall digestive process will be associated with a difference in luminal composition as electrogenic K+ secretion becomes the dominant secretory mode.

A minor constituent of intestinal epithelia is the L cell population of enteroendocrine cells that release various signaling peptides, with peptide-YY (PYY) being most prominent at distal sites (10, 15). These L cells sense luminal composition, particularly nutrient molecules such as short-chain fatty acids, leading to release of PYY. Also, in L cells of the distal colon, β-adrenergic activation releases PYY (2, 5, 37). Cleavage of this PYY by dipeptidyl-peptidase-4 produces PYY(3-36), one of the signals for satiety (48). Thus these distal L cells are a point of integration between digestion and sympathetic tone, leading toward a cessation of feeding and the digestive process. In addition, PYY inhibits Cl secretion by epithelial cells (10, 15), thereby limiting any further increases to the volume of luminal contents. The action of PYY on electrogenic K+ secretion is important because the modulatory mode is the secretory character of the sympathetic response in the distal colon, and this action of PYY is largely undefined.

The enteric nervous system is comprised of an extensive network of nerves that influence all aspects of gastrointestinal activity (8, 17). In the colon, the myenteric and submucosal enteric plexuses are centers of reflex circuits that integrate the control of motility, blood flow, and transepithelial fluid flow. Input from the central nervous system is relayed through these nerve plexuses but also can signal directly to effectors such as blood vessels. Central influence on sympathetic responses first occurs at nerve cells residing in the prevertebral ganglia that have processes extending into the wall of the colon (26, 45). Those sympathetic fibers projecting into the mucosa often are near vascular elements such that excitation limits blood flow via vasoconstriction. Another major action of this sympathetic input to the colon is to inhibit secretomotor nerves in the submucosal plexus such that the rate of epithelial secretion is reduced.

Mucosal nerve fibers with varicosities containing norepinephrine (norepi) are in close proximity to epithelial cells such that direct sympathetic action on epithelial cells is possible (17, 22). The norepi-containing nerves of the sympathetic ganglia also often contain other signaling peptides, generally either neuropeptide-Y (NPY) or somatostatin (45). Therefore, a sympathetic response will lead not only to norepi release near epithelial cells but likely also NPY release from nerves and PYY release from L cells. Because NPY and PYY can activate the same neuropeptide receptors (NpR), these peptides would contribute to stimulation of the same signaling event (4, 9, 10, 36). Specifically, activation of these NPY receptors leads to inhibition of Cl secretion, indicating that a sympathetic response will have an inhibitory action directly on the secretory epithelial cells as well as the inhibitory action on the secretomotor nerves in the submucosal plexus (9).

The present study examines the action of the neuroendocrine peptides PYY and NPY on the adrenergic activation of electrogenic K+ secretion to further define the modulatory mode of secretion produced by sympathetic stimulation. This modulatory mode was studied in isolation by removing all of the underlying muscle layers so that neural and paracrine release of secretagog substances would not confound the action of these peptides. Adrenergic activation of electrogenic K+ secretion occurs through β-adrenergic receptors present on the epithelial cells (52), and the presence of NpRs on these cells would support an intracellular signaling cascade with cAMP production as a key component such that β-adrenergic receptor action increases cAMP and the Y-NpRs would decrease cAMP (1, 4, 36). Such a link in secretory control has implications for pathophysiology of the gut because changes in PYY release likely contribute to the progression of several gastrointestinal disorders and in some cases are a primary feature of the disorder (14). The results presented here demonstrate the expression of mRNA and protein for NpRs as well as functional evidence for the action of these receptors on the activation of electrogenic K+ secretion in addition to electrogenic Cl secretion.


Male guinea pigs (500–800 g body wt, Hartley strain; Hilltop Laboratory Animals, Scottdale, PA and Harlan Laboratories, Indianapolis, IN) received standard chow and water ad libitum and were housed on site at least 2 wk before experiments. The large majority of the experiments reported were with guinea pigs from Hilltop Laboratory Animals, with only a small number of experiments conducted with guinea pigs from Harlan Industries. Guinea pigs were euthanized with an animal decapitator (Harvard Apparatus, Holliston, MA) in accordance with a protocol approved by the Wright State University Laboratory Animal Care and Use Committee. Colonic mucosa was isolated as described previously (52). These isolated colonic mucosal sheets were used for measurement of transepithelial electrical parameters, RNA isolation, and protein detection by immunoblot.

Detection of mRNA and proteins.

Total RNA was extracted by RNeasy Mini Kit (Qiagen, Valencia CA) from EDTA-released epithelial cells as described previously (52). Briefly, cDNA was amplified by PCR: initial denaturing at 95°C (10 min), 40 cycles denaturation at 92°C (1 min), annealing at 64°C (1 min), and extension at 72°C (8 min). Primers specific for NPY receptors were designed from those in a previous report (19). After alignment of those primers with nucleotide sequences for Y1-NpR (GenBank accession AF135061, guinea pig Npy1r) and Y2-NpR (GenBank accession AF072821, guinea pig Npy2r), bases were changed to correspond to the guinea pig sequence, in positions where the nucleotide sequences were not homologous. Primers for Y1-NpR were (forward primer) 5′-gct tcc tct ctg ccc ttc atg-3′(502–522) and (reverse primer) 5′-tgt ctc ata gtc atc atc ccg-3′(1,1321,152). Primers for Y2-NpR were (forward primer) 5′-aaa tgg gtc ctg tcc tct gcc-3′(353-373) and (reverse primer) 5′-tgc ctc cgc tgg tgg tag tgg-3′(774-794). Proteins were isolated from colonic mucosa as described previously (33, 52). Briefly, after disruption by sonication in a buffered solution containing protease inhibitors, epithelial cell lysates were centrifuged to obtain a membrane sample. Following SDS-PAGE and transfer to polyvinylidene difluoride membranes, incubation with specific primary antibody and then with horseradish peroxidase-conjugated secondary antibody allowed detection of specific proteins.

Tissue fixation and immunolocalization.

Colonic tissues were fixed after isolation, as described previously (33, 52). Briefly, isolated mucosal sheets were chemically fixed followed by dehydration, sectioning, and mounting on gelatin-coated slides. Sections were permeabilized, blocked, and then incubated for 48 h (4°C) with primary antibody. Antibodies for NPY receptors were obtained from Santa Cruz Biotechnology (Santa Cruz, CA) and were polyclonal rabbit-anti-Y1-NpR [4 ng/μl; residues 181-271 (H-91) of human Npy1r] and polyclonal rabbit-anti-Y2-NpR [4 ng/μl; residues 124-270 (H-147) of human Npy2r]. Antibodies for dopamine-β-hydroxylase (DβH) and NPY were obtained from Abcam, Cambridge, MA and were polyclonal rabbit-anti-DβH [1:1,000; full length (ab63939) of bovine Dbh] and polyclonal sheep-anti-NPY [1:800; residues 30-64 (ab6173) of human NPY]. Secondary antibodies to detect immunoreactivity (ir) (2 h, room temperature) were obtained from Invitrogen (Carlsbad, CA) and were donkey-anti-rabbit IgG antibody, conjugated to AlexaFluor 488 (4 ng/μl) and a donkey-anti-sheep IgG antibody, conjugated to AlexaFluor 568 (10 ng/μl). Sections were washed, mounted in Vectashield (Vector Laboratories, Burlingame, CA), and fluorescence was visualized with an Olympus BX60 epifluorescence microscope. Double-labeled sections were visualized with an Olympus FluoView FV300 confocal microscope in the Microscopy Core Facility of the Comprehensive Neuroscience Center. Images were acquired using identical confocal aperture, background, and gain settings.

Transepithelial current measurement.

Isolated mucosal sheets were used for measurement of transepithelial current and conductance as described previously (24, 52). Mucosal sheets were mounted in Ussing chambers (0.64 cm2 aperture), supported on the serosal face by Nuclepore filters (∼10 μm thick, 5 μm pore diameter; Whatman, Clifton, NJ). Bathing solutions (10 ml) were circulated by gas-lift through water-jacketed reservoirs (38°C). Standard Ringer's solution contained (in mM): 145 Na+, 5.0 K+, 2.0 Ca2+, 1.2 Mg2+, 125 Cl, 25 HCO3, 4.0 H(3-X)PO4X−, and 10 d-glucose. Solutions were continually gassed with 95% O2 and 5% CO2, which maintained solution pH at 7.4. Automatic voltage clamps (Physiologic Instruments, San Diego, CA) permitted measurement of short-circuit current (Isc) and calculation of transepithelial conductance (Gt) from current responses to voltage pulses imposed across the mucosa (± 5 mV). Isc was referred to as positive for cation flow across the epithelium from mucosal to serosal side.

Mucosal responses to physiological secretagogs and to inhibitors were examined after producing a quiescent basal condition. This basal state was produced by suppressing the neural and paracrine activators persisting in the isolated colonic mucosa (24, 52). The mucosal preparation removes influences from nerves in the underlying muscle layers. Prostanoid production within the isolated mucosa was suppressed by the cyclooxygenase (COX)-1 inhibitor SC560 (1 μM) and the COX-2 inhibitor CAY10404 (1 μM) added to both bathing solutions. Other compounds released into the bathing solutions were reduced in concentration (∼8,000-fold) by replacing the solutions three times after mounting the mucosa. Amiloride (10 μM) was added to the mucosal bath to inhibit electrogenic Na+ absorption.

PGE2, CAY10404, and SC560 were obtained from Cayman Chemical (Ann Arbor, MI); BIBP3226, BIIE0246, BVD10, CGP20712A, ICI-118551, SR59320A, and TTX were from Tocris Bioscience (Ellisville MO); NPY, PYY, PYY(3-36), and pancreatic polypeptide were from Bachem Americas (Torrance, CA); epinephrine (epi) was from Elkins-Sinn (Cherry Hill, NJ); P32/98 was from BioMol (Plymouth Meeting, PA). All other chemicals were obtained from Sigma Chemical (St. Louis MO). Drugs were added in small volumes from concentrated stock solutions. PGE2 was prepared in an ethanol stock solution that added 0.03% ethanol at 3 μM of PGE2; additions of 1% ethanol alone did not alter transepithelial measures of K+ or Cl secretion (24).

Data analysis.

Responses of Isc and Gt to secretagogs and antagonists were obtained from adjacent mucosae in each colon to permit direct comparisons. Recordings of Isc were digitized at 10-s intervals to examine the secretory time course. Concentration responses of Isc and Gt were fit to Henri-Michaelis-Menten binding curves using a nonlinear least-squares procedure (40). Results are reported as means ± SE with the number of animals (N) or the number of mucosal tissues (n) indicated. Statistical comparisons were made using a two-tailed Student's t-test for paired responses, with significant difference accepted at P < 0.05.


Suppression of Cl secretion by PYY.

Three modes of electrogenic ion secretion (11, 25, 33, 49, 52) were stimulated by secretagogs to examine the action of the neuroendocrine-peptide PYY, epinephrine for modulatory mode, PGE2 for flushing mode, and PGE2 + carbachol (CCh) for synergistic mode. Adding PYY to the serosal bathing solution during the basal quiescent state reduced the negative Isc toward zero and decreased Gt (Fig. 1, A and B). Epi stimulation produced a transient positive change in Isc associated with electrogenic Cl secretion and a sustained negative Isc indicative of electrogenic K+ secretion (39). Concurrently, Gt increased monotonically to a higher level. The transient positive Isc was dramatically reduced by pretreatment with PYY, leaving a larger sustained negative Isc and a smaller sustained Gt. PYY added during epi stimulation produced a small negative Isc deflection such that the final level was similar to the mucosa pretreated with PYY. Similarly, Gt decreased to the level of the PYY-pretreated mucosa. Together, the PYY-induced changes in Isc and Gt supported the concept that the primary action of this neuroendocrine peptide was inhibiting the Cl secretory component of the epi response.

Fig. 1.

Each mode of secretory activation was sensitive to peptide-YY (PYY). Isolated mucosae were stimulated sequentially by adding epinephrine (epi) (5 μM), prostaglandin-E2 (PGE2) (3 μM), and carbachol (CCh) (10 μM) to the serosal bath, from the standard basal condition (see materials and methods). Short-circuit current (Isc) and transepithelial conductance (Gt) were measured in 4 adjacent mucosae (◊,▵, ▿, •) as PYY (1 μM) was added to the serosal bath during secretory activation (*). The time course was split into separate panels to aid viewing of each secretory mode, and the mucosae receiving PYY before secretagog stimulation (◊) was shown throughout. Gt was normalized by subtracting the basal value prior to stimulation (δGt), except for the mucosae pretreated with PYY (◊) for which the normalization was made to the preaddition level. The basal Gt for the 4 mucosae was 9.47 ± 38 mS/cm2. A and B: modulatory mode was activated by epi with one mucosa having PYY added before epi (◊) and the other after epi activation (▵). Each of the control mucosae (▵, ▿, •) had a transient positive Isc at the onset of epi activation. The peak transient ΔIsc and ΔGt for epi addition alone were significantly larger than in the presence of PYY (P < 0.05, N = 10), respectively, +119 ± 13 μA/cm2 and +20 ± 7 μA/cm2 as well as +2.78 ± 28 mS/cm2 and +0.93 ± 12 mS/cm2. Before epi addition, the changes in Isc and Gt attributable to PYY addition compared with control were significantly different from zero (P < 0.05, N = 10), respectively, +19 ± 3 μA/cm2 and −1.12 ± 0.17 mS/cm2. C and D: flushing mode activated by PGE2 was shown in the presence (◊) and absence (•) of PYY, and a third mucosa was shown (▿) with PYY added ∼12 min after PGE2 activation. E and F: synergistic mode was activated by adding CCh in the continued presence of PGE2. The action of PYY was assessed further by comparing the response in the presence of PYY (◊) to the mucosa receiving PYY ∼26 min after CCh activation (•). G and H: PYY-sensitive components of Isc (PYYΔIsc) and Gt (PYYΔGt) for each mode of secretory activation were calculated as the difference between mucosae with PYY absent and those with PYY present; these positive ΔIsc and ΔGt likely represented electrogenic Cl secretion. The responses were representative of 6 similar experiments.

Stimulating the flushing mode of secretion with PGE2 reversed the orientation of the Isc from that produced by epi (Fig. 1, C and D), resulting in a large positive overshoot of Isc that declined toward zero. Previous measurements of Cl and K+ unidirectional fluxes indicate that the sustained PGE2-activated Isc is the result of nearly equal amounts of Cland K+ secretion such that the Isc is near zero (39). The high rate of secretion in this mode of activation was evident from the large increase in Gt. Mucosae pretreated with PYY had a much smaller Isc overshoot, and sustained Isc was suppressed to negative values consistent with electrogenic K+ secretion as the primary ion secretory component. Adding PYY during the PGE2 interval decreased Isc to the epi-activated level followed by a gradual return to a higher but still negative value. Recovery of Gt from PYY inhibition during flushing mode supported the concept that Cl secretion was inhibited and then recovered partially.

Synergistic stimulation produced by CCh addition in the presence of PGE2 resulted in a large positive Isc with a transient component lasting ∼10 min before a slowly declining plateau phase became evident (Fig. 1, E and F). Mucosae pretreated with PYY had a similar early transient, but the following phase declined more rapidly. Adding PYY during the late phase of synergistic activation rapidly inhibited Isc and Gt with a small rebound suggestive of a minor recovery in Cl secretion as occurred during the flushing mode (Fig. 1, C and D). Addition of PYY to the mucosal bathing solution did not alter any of the secretory modes (data not shown) supporting a localization of the NpR involved to the basolateral membrane.

The PYY-sensitive component of the stimulated Isc (PYYΔIsc) was positive for all secretory modes and was mirrored by similar PYYΔGt, consistent with Cl secretion (Fig. 1, G and H). None of this inhibitory suppression of Cl secretion by PYY was sensitive to TTX (data not shown), indicating a lack of involvement in these inhibitory responses by remnant mucosal nerve processes. Each mode of secretory activation had a distinct size and time course, but three general phases of stimulation were apparent, a brief positive transient lasting ∼2 min, a longer positive transient lasting 5–15 min, and a sustained plateau. All three phases were sensitive to PYY during epi and PGE2 activation although the size of PYYΔIsc was considerably smaller with epi. Sensitivity during the synergistic mode was limited to the brief transient and the sustained plateau, with the intermediate transient phase largely resistant to PYY inhibition. This suppression of Cl secretion by PYY left electrogenic K+ secretion as the major secretory event during the modulatory and flushing modes.

Secretory sensitivity to neuropeptides.

The sensitivity of secretory suppression by neuropeptides was examined during the sustained phase of stimulation. Adding increasing concentrations of PYY and NPY reduced secretory Isc to a maximal extent of ∼65% (Fig. 2); Gt was reduced to a similar extent (data not shown), supporting inhibition of an electrogenic process. The IC50 for PYY was 4.1 ± 0.9 nM, similar to the values obtained for inhibition of Isc in rat distal colon (43) and VIP-stimulated cAMP production in rat jejunal crypt cells (20). The shortened form of PYY, PYY(3-36), that results from peptidase activity in situ also produced a potent suppression. Because PYY(3-36) preferentially binds to Y2-NpR, the ability of PYY(3-36) to suppress secretory Isc indicated an action that worked via Y2-NpR. Addition of P32/98 (10 μM), an inhibitor of the ectopeptidase dipeptidyl-peptidase-4, to the serosal bath did not alter the response to PYY (data not shown), suggesting that in situ conversion to PYY(3-36) was not responsible for the observed results. Pancreatic polypeptide (1 μM) did not inhibit the secretory Isc activated by PGE2 or CCh (data not shown), which, on the basis of agonist selectivity (4, 36), suggested that Y4-NpR was not involved. The receptor type involved in this action was further defined using selective antagonists for the Y1-NpR and Y2-NpR, BVD10 (Y1-NpR) and BIIE0246 (Y2-NpR). The ability of BIIE0246 to inhibit the action of PYY and lack of effect by BVD10 supported the involvement of Y2-NpR.

Fig. 2.

Neuroendocrine peptides inhibited secretory Isc with high efficacy. Isolated mucosae were stimulated by sequential additions of epi (5 μM), PGE2 (3 μM), and CCh (10 μM) as in Fig. 1. Neuroendocrine peptides were added at increasing concentrations to the serosal bath in adjacent mucosa either during PGE2 (as in Fig. 1B) or CCh activation (as in Fig. 1C). Inhibitory responses were normalized to the Isc before addition compared with the sustained Isc during epi activation. Fractional inhibition was similar for both secretory modes so that the results were combined, neuropeptide-Y (NPY) (⧫, n = 3), PYY (•, n = 13), PYY(3-36) (○, n = 4). Antagonists to neuropeptide receptors [BVD10, Y1-neuropeptide receptor (Y1-NpR); □, n = 3] and (BIIE0246, Y2-NpR; ▿, n = 4) also were added (1 μM) for some mucosae before secretory activation and PYY addition. The fit of the data with a single binding site [NPY and PYY, solid line; PYY(3-36), dashed line; PYY + BVD10, gray dashed line] yielded IC50s of 4.1 ± 0.9 nM (PYY), 6.2 ± 1.9 nM [PYY(3-36)], and 9.4 ± 3.8 nM (NPY); maximal fractional inhibitions were 0.67 ± 0.03, 0.70 ± 0.05, and 0.58 ± 0.10, respectively. Neither the IC50s nor the maximal fractional inhibitions were significantly different among these peptides (P < 0.05). The Y1-NpR antagonist BIBP3226 (1 μM) also did not inhibit the action of PYY (data not shown).

Identification of NPY receptors in distal colonic mucosa.

The presence in distal colonic epithelial cells of mRNA for Y1-NpR and Y2-NpR was detected using RT-PCR (Fig. 3). Both PCR products were purified and sequenced, which verified identity with the reported sequence for these receptor proteins. The presence of Y1-NpR and Y2-NpR also was examined by immunoblot of the epithelial cell membrane fraction (Fig. 4). Immunoreactive bands consistent with Y1-NpR and Y2-NpR were detected (13, 47). The band observed at ∼41 kDa for Y1-NpR was similar to the anticipated size of the monomeric form, 44 kDa. The larger band at ∼94 kDa was similar to those found in cells transfected to express hY1-NpR (13, 47), consistent with posttranslational modification and undissociated oligomerization observed previously (4, 12, 16). The ∼55-kDa and ∼61-kDa bands observed for Y2-NpR were similar to the immunoreactive bands in cells transfected to express hY2-NpR (13, 47), both of which were slightly larger than the anticipated monomeric form, 42 kDa. These results support the presence of both Y1-NpR and Y2-NpR in the colonic mucosa.

Fig. 3.

NPY receptor mRNA detected by RT-PCR. RNA isolated from distal colonic mucosa was used to amplify Y1-NpR and Y2-NpR products by RT-PCR. Products were obtained at 546 base pairs (bp) for Y1-NpR and 442 base pairs for Y2-NpR as predicted from the position of the forward and reverse primers (indicated by asterisks), and amplification of GAPDH served as a positive control for RNA isolation. The negative control obtained by not including reverse transcriptase indicated the lack of contamination by genomic DNA. The faint band smaller than 100 bp likely was due to unused primers from the PCR.

Fig. 4.

NPY receptor proteins detected by immunoblot. Protein isolated from distal colonic epithelial cell membranes was immunoblotted with antibodies against the Y1-NpR and Y2-NpR proteins. Immunoreactive bands occurred at 43 kDa and 92 kDa for Y1-NpR and 55 kDa and 60 kDa for Y2-NpR (arrowheads), consistent with monomeric and possible oligomeric forms. Use of the secondary antibody alone eliminated all bands (data not shown), indicating that the primary antibodies were necessary for the observed results.

Epithelial localization of NPY receptors.

Ir for Y1-NpR (Fig. 5) and Y2-NpR (Fig. 6) proteins was detected in a mucosal location consistent with the plasma membrane of colonic epithelial cells, similar to localization of Y1-NpR in human colon (34). Prominent labeling was seen in the lateral membrane of crypt and surface epithelial cells. The luminal margins of epithelial cells were not labeled, indicating an absence from the apical membrane, consistent with the observed action of PYY (NPY) only from the serosal bath. The uniform lateral labeling in crypts supported the possible presence of Y1-NpR and Y2-NpR in goblet cells as well as in columnar cells. No other structures in the mucosa had distinct Y1-NpRir or Y2-NpRir labeling, indicating that mucosal actions of the neuropeptides NPY and PYY would likely occur at the epithelial cell.

Fig. 5.

Y1-NpR proteins localized in the colonic epithelium. Y1-NpR was detected by immunofluorescence (anti-Y1-NpR) in distal colonic mucosa. A and B: surface epithelial cells had prominent immunoreactivity (Y1-NpRir) labeling of lateral membranes (arrowheads) without marked labeling of basal membranes (b). Labeling for apical membranes (a) was not apparent. Apically located goblet granule masses were apparent as dark voids (G). C and D: crypts showed distinct lateral membrane Y1-NpRir labeling (arrowheads), whereas the basal membrane lacked distinct labeling. Use of the secondary antibody alone eliminated all membrane labeling (data not shown), indicating that the primary antibodies were necessary for the observed results. Scale bars = 10 μm.

Fig. 6.

Y2-NpR proteins localized in the colonic epithelium. Y2-NpR was detected by immunofluorescence (anti-Y2-NpR) in distal colonic mucosa. A and B: surface epithelial cells had prominent Y2-NpRir labeling of lateral membranes (arrowheads) without marked labeling of basal membranes (b). Labeling for apical membranes (a) was not apparent. Apically located goblet granule masses were apparent as dark voids (G). C and D: crypts showed distinct lateral membrane labeling for Y2-NpRir (arrowheads), and luminal margins did not show labeling (lumen, L). Use of the secondary antibody alone eliminated all membrane labeling (data not shown), indicating that the primary antibodies were necessary for the observed results. Scale bars = 10 μm.

Purinergic activation of secretion.

Adrenergic nerves often release ATP as well as norepi and NPY (17, 26). Adding ATP (100 μM) to the serosal bath during the basal quiescent state produced a transient positive Isc followed by a sustained negative Isc (Fig. 7A); similarly, Gt was stimulated (data not shown). Because mucosal addition also has been shown to stimulate Isc in rat and mouse colon (28, 35), ATP was added to the mucosal bath, which resulted in a brief negative transient that subsided before the onset of a slow stimulation to a negative Isc of similar magnitude. Previously, distal colon from aldosterone-treated guinea pig was shown to produce a transient positive Isc with either mucosal or serosal ATP addition at 1,000 μM (50). In this study, the sustained negative Isc response with either mucosal or serosal ATP addition was smaller than the epi response. Addition of the ATPase apyrase (10 U/ml) to both baths did not alter Isc activation by epi (data not shown), suggesting that ATP did not contribute to the epi response. Adding UTP (100 μM) to either the serosal or mucosal bath did not alter Isc (data not shown), suggesting that P2Y receptors (30) were not coupled to secretory activation. The neuropeptide sensitivity of ATP stimulation was examined using PYY to activate suppression and the Y2-NpR-specific antagonist BIIE0246 to blunt in situ neuropeptide responses (Fig. 7B). PYY suppressed a positive transient Isc component, leaving the negative Isc essentially unaltered. Subsequent stimulation with epi produced Isc (Fig. 7B) similar to responses in paired mucosae without ATP pretreatment (data not shown). Thus the serosal ATP response had the same transient and sustained components and PYY sensitivity that occurred with epi.

Fig. 7.

ATP and adenosine activated the modulatory mode of secretion. Isolated mucosae were stimulated by secretory agonists, from the standard basal condition as in Fig 1. A: Isc was measured in 3 adjacent mucosae with ATP (100 μM) added to the serosal (▾) or mucosal (▵) bath, or epi (shaded circles, 1 μM) added to the serosal bath. The responses were representative of 3 similar experiments (agonistΔIsc: transient ATPmucosal, −22 ± 2 μA/cm2; sustained ATPmucosal, −13 ± 4 μA/cm2; sustained ATPserosal, −31 ± 2 μA/cm2; sustained epi, −72 ± 9 μA/cm2; each response was significantly different from zero, ATP responses were significantly smaller than epi, and sustained ATPserosal response was significantly more negative than sustained ATPmucosal response, P < 0.05). B: Isc was measured in 2 adjacent mucosae as ATP (100 μM) and epi (1 μM) were added sequentially to the serosal bath with pretreatment either by PYY (○, 0.3 μM) or BIIE0246 (BIIE) (•, 1 μM). Difference in Isc between mucosae (▵) revealed the PYY-sensitive components. The responses were representative of 3 similar experiments (sustained ATPΔIsc for serosal addition was not significantly different with BIIE0246 compared with PYY, P < 0.05). C: Isc was measured in 3 adjacent mucosae as ATP (100 μM) was added either to the serosal (▾) or mucosal (▵) bath with pretreatment by the adenosine receptor antagonist CGS15943 (5 μM) (shaded circles, no addition control). The small positive change in Isc during the initial addition of CGS15943 may have resulted from blunting of a response to adenosine present in the bath (CGSΔIsc: +6 ± 1 μA/cm2, significantly different from zero, P < 0.05). The responses were representative of 3 similar experiments (agonistΔIsc: sustained ATPmucosal, −1 ± 1 μA/cm2; sustained ATPserosal, −4 ± 1 μA/cm2; ATPmucosal response was not significantly different from zero, ATPserosal response was significantly different from zero, but also significantly smaller than without CGS15943, P < 0.05). D: Isc was measured in 3 adjacent mucosae with adenosine (100 μM) added to the serosal (▾) or mucosal (▵) bath, or epi (shaded circles, 1 μM) added to the serosal bath. The responses were representative of 4 similar experiments (agonistΔIsc: mucosal adenosine, −12 ± 2 μA/cm2; serosal adenosine, −25 ± 8 μA/cm2; epi, −79 ± 13 μA/cm2; each response was significantly different from zero, and adenosine responses were significantly smaller than epi, P < 0.05).

The serosal ATP response was largely eliminated by pretreatment with the adenosine receptor antagonist CGS15943 (Fig. 7C), suggesting that the ATP response acted via adenosine receptors after in situ hydrolysis by ectonucleotidases. Similarly, sustained negative Isc produced by mucosal ATP was inhibited by CGS15943, but the brief negative transient was unaltered. Secretory responses to epi, PGE2, and CCh were not altered markedly by the presence of CGS15943 (data not shown). Adding adenosine to the serosal or mucosal bath reproduced the responses observed with ATP, except for the brief negative Isc transient with mucosal ATP (Fig. 7D). These results suggested that ATP released in the serosal vicinity of colonic epithelial cells was converted to adenosine, which stimulated Cl and K+ secretion, whereas mucosal ATP produced transient K+ secretion via a nucleotide receptor (P2R) and sustained K+ secretion via an adenosine receptor (AdoR).

Localization of mucosal adrenergic nerves containing NPY.

Ir for DβH was used to identify adrenergic nerves in the mucosa and NPYir to identify specific peptidergic nerves containing NPY (Fig. 8). The results confirmed previous studies of adrenergic and peptidergic innervation in the colon (17, 22). Double labeling for DβHir and NPYir allowed the determination of which adrenergic nerves also contained NPY. Distinct double labeling was apparent in the mucosa near crypt epithelial cells (Fig. 8, AC). Prominent DβHir and NPYir labeling also was seen around blood vessels in the submucosa (Fig. 8D) and in the myenteric plexus (Fig. 8E), with colocalization only apparent adjacent to blood vessels, which confirmed previous chemical coding of these nerves. The absence of colocalized labeling for apparent nerve fibers within muscle layers and the myenteric plexus indicated a separation between DβHir and NPYir such that the adrenergic nerves lacked NPY and the NPY containing nerves were nonadrenergic. Although morphologically identifiable elements of the submucosal plexus were not detected, a nerve bundle on the inner surface of the circular muscle layer did not show colocalization (Fig. 8D).

Fig. 8.

Adrenergic and NPY-containing nerve types localized in the mucosa. Adrenergic and peptidergic nerves were detected in distal colon by ir for dopamine-β-hydroxylase (DβH) and NPY (anti-DβH, anti-NPY). A, B, and C: prominent immunoreactive labeling for DβH (green) and NPY (red) was apparent in the mucosa and muscularis mucosa with numerous points of colocalization (C, yellow). An apparent nerve bundle (arrowhead) crossed the muscularis mucosa into the mucosa. Nonspecific labeling of immune cells apparent in the interstitium were marked with asterisks. Scale bar = 40 μm. D: submucosa situated between the muscularis mucosa (mm) and the circular muscle had prominent points of DβHir and NPYir labeling with colocalization distinct around blood vessels (bv). Other nerve bundles did not show colocalization (arrowhead). Scale bar = 40 μm. E: myenteric plexus had prominent DβHir and NPYir labeling without colocalization. Scale bar = 40 μm. F: apparent nerve bundle lay between the muscularis mucosa and the base of crypts and extended between crypts (arrowhead) with points of colocalized DβHir and NPYir labeling seen throughout the nerve bundle course. Another nerve bundle was apparent crossing the muscularis mucosa (asterisk). Scale bar = 40 μm. G: crypt from C is shown enlarged with apparent nerve fibers parallel to the crypt axis. Scale bar = 20 μm. HK: nerve fibers were apparent often parallel to the crypt axis with either colocalization of DβHir and NPYir labeling or only labeling for one of these markers (arrowheads). Muscularis mucosa is at the bottom of each panel with the crypt axis vertical; scale bar = 20 μm. L: region near the yellow puncta indicated in K (arrowhead) was enlarged. The z-axis profiles through the puncta confirmed colocalization. A line scan of fluorescence intensity across this puncta (inset) indicated overlap of the peaks for both wavelengths, as well as a neighboring red puncta with only NPY labeling that lacked overlap. Use of each primary antibody alone, together with both secondary antibodies led to loss of labeling from the unmatched secondary antibody (data not shown), indicating that the primary antibodies were necessary for the observed labeling and that overlap did not occur between the fluorescence channels. Labeling of apparent immune cells (asterisks, C) was nonspecific because it occurred in the absence of primary antibodies.

Adrenergic innervation of the mucosa was apparent as fibers running along blood vessels through the muscularis mucosa and into the mucosa near the crypts (Fig. 8, C and F). Colocalized DβHir and NPYir labeling was apparent in many of these nerve bundles in the mucosa as closely spaced puncta, particularly in areas near muscularis mucosa crossovers where multiple fibers were seen. In other mucosal areas, single fibers were apparent, some exhibiting colocalization and others having either DβHir or NPYir labeling alone, possibly running along blood vessels in the lamina propria (Fig. 8, GL). These results support the likely stimulation of colonic epithelial cells by release of norepi and NPY from sympathetic nerves.

Endogenous release of neuropeptides.

The secretory response varied between guinea pigs, with respect to the appearance of the brief transient positive Isc component during epi stimulation and the size of the sustained Isc during PGE2 or synergistic stimulation. Those mucosae lacking a noticeable epi-stimulated transient Isc also generally had lower sustained Isc during PGE2 stimulation and a more rapid decline in Isc during the late phase of synergistic stimulation. Together, these observations when compared with the action of PYY (Fig. 1) suggested that PYY or NPY may have been released in vitro to a varying extent. This concept was supported by the increase in Isc resulting from addition of the Y2-NpR antagonist BIIE0246 (Fig. 9A). In paired mucosae, the one having BIIE0246 before synergistic activation also had a higher Isc. Adding BIIE0246 to the untreated mucosa of the pair led to a rapid increase in Isc and Gt to the value of the pretreated mucosa, consistent with removal of the suppression that had been induced by either PYY or NPY released into the serosal bath.

Fig. 9.

Secretory Isc suppressed by endogenous sources. Isolated mucosae were stimulated by secretory agonists, from the standard basal condition as in Fig 1. A: Isc was measured in 2 adjacent mucosae after secretory activation by CCh (10 μM) followed by PGE2 (3 μM) with either no pretreatment (○) or BIIE0246 (1 μM) preaddition to the serosal bath (shaded diamond); time was from the addition of CCh. BIIE0246 was added at asterisk. Differences of Isc between these mucosae (▵) revealed the Y2-NpR-dependent component. The ΔIsc and ΔGt responses to BIIE0246 during synergistic secretion were +80 ± 28 μA/cm2 and +1.63 ± 0.36 mS/cm2 (significantly different from zero; P < 0.5, N = 6, n = 13). B: Isc was measured in 3 adjacent mucosae (shaded circle, ◊) after secretory activation by epi (5 μM) followed by PGE2 (3 μM). One mucosa (◊) was pretreated with veratridine (10 μM), and two (shaded circle, ◊) had tetrodotoxin (TTx) added at 1 μM (*); time was from the addition of veratridine. PYY (1 μM) was added subsequently to all mucosae. Difference of Isc between mucosae (▵) revealed the neural-dependent component. The veratridine-inhibited ΔIsc and ΔGt in the presence of TTx were +45 ± 2 μA/cm2 and +2.36 ± 0.69 mS/cm2 (significantly different from zero; P < 0.5, N = 4). C: Isc was measured in 2 adjacent mucosae (shaded circle, ◊) after secretory activation by epinephrine (5 μM) followed by PGE2 (3 μM). One mucosa was pretreated with veratridine (10 μM) (◊); time was from the addition of veratridine. After TTx (1 μM) addition (shaded circle, ◊), BIIE0246 (1 μM) was added (*) to the veratridine-treated mucosa (◊). Difference of Isc between mucosae (▵) revealed a Y2-NpR-dependent component. The responses were representative of 3 similar experiments (response to BIIE0246 was significantly different from zero, P < 0.05).

To examine whether remnant nerve processes in the isolated mucosa could be the source of this neuropeptide, veratridine, which activates the voltage-sensitive Na+ channels found in nerves, was added to stimulate release of neurotransmitters (Fig. 9B). Addition of TTX during PGE2 stimulation in the veratridine-treated mucosa rapidly reduced Isc to below the control PGE2-stimulated level, but the Isc of untreated mucosa was unaltered by TTX. This result suggested that veratridine had activated release of excitatory and inhibitory neurotransmitter substances from remnant nerve processes that acted on Cl secretion. Inhibition of voltage-activated Na+ channels by TTX eliminated the stimulus such that the excess Cl secretion was stopped, revealing that persistent inhibitory factors also had been released, possibly NPY. Subsequent addition of PYY reduced Isc in all mucosae to the same negative value, indicating that this Isc was subject to neuropeptide control. Adding BIIE0246 reduced the difference between control and veratridine-stimulated mucosae, supporting the concept that a portion of these inhibitory factors acted via Y2-NpR (Fig. 9C). Together, these results indicated that an inhibitory tone was present in vitro for colonic mucosa, which varies between guinea pigs depending on neural and paracrine influences.

The results obtained with guinea pigs supplied by Hilltop Laboratory and Harlan Laboratories were essentially identical, except that the epi response in Harlan guinea pigs rarely exhibited a transient positive Isc and Hilltop guinea pigs generally exhibited a noticeable transient. Pretreatment with BIIE0246 generally led to an observable transient positive Isc in Harlan guinea pigs, suggesting that in vitro release of PYY occurred more readily for colons from these animals. Only a few experiments were accomplished to examine this difference in supplier, before Harlan Laboratories discontinued North American production of guinea pigs. This distinction between guinea pigs from various suppliers is of note since most of the previous studies from this laboratory (24, 25, 32, 33, 39, 52) were with guinea pigs obtained from Harlan Laboratories, and this supplier had been a common source for many other published studies of intestinal function. Interestingly, the guinea pig colonies at Hilltop Laboratory and Charles River Laboratories (Wilmington, MA) were begun with Hartley strain guinea pigs from Harlan Laboratories.

Sensitivity to epi and norepi.

The sensitivity of secretory activation to epi and norepi was determined by measuring Isc during increases in agonist concentration (Fig. 10A). Responses to epi and norepi were obtained in adjacent mucosae from each animal pretreated either with BIIE0246 or PYY. These paired EC50s for epi and norepi were different between pretreatment groups with PYY significantly increasing epiEC50 and neEC50 (P < 0.05). The maximal Isc response was more negative with PYY than with BIIE0246, likely attributable to inhibition of a small sustained Cl secretory component by PYY. The markedly lower EC50s in the presence of BIIE0246 (Fig. 10B) compared with the sensitivity measured previously without pretreatment (52) were consistent with the presence of in vitro-released neuropeptides. Variability observed between guinea pigs in the magnitude of the epi-activated transient positive Isc during neuropeptide blockade with BIIE0246 suggested that another endogenous inhibitor may be released in vitro depending on the state of the animal.

Fig. 10.

PYY altered the secretory sensitivity to epi and norepi. Isolated mucosae were stimulated by increasing concentrations of either epi or norepi, from the standard basal condition as in Fig 1. A: 4 mucosae from each animal (N = 5) were paired for pretreatment with either BIIE0246 (◊, ○; 1 μM) or PYY (⧫, •; 0.3 μM), and then Isc was recorded during addition of either norepi (◊, ⧫) or epi (○, •). The fit of the data with a single binding site (epi, black lines; norepi, gray lines; PYY, solid lines; BIIE0246, dashed lines) yielded EC50s of 2.9 ± 1.2 nM (BIIE/epi), 13.3 ± 1.0 nM (BIIE/norepi), 12.8 ± 1.3 nM (PYY/epi), and 39.8 ± 8.4 nM (PYY/norepi). Maximal secretory Isc values were −119 ± 12 μA/cm2 (BIIE/epi), −123 ± 13 μA/cm2 (BIIE/norepi), −133 ± 9 μA/cm2 (PYY/epi), and −144 ± 10 μA/cm2 (PYY/norepi). The EC50s for epi and norepi in the presence of PYY were significantly larger than those with BIIE0246 (P < 0.05), respectively, +9.9 ± 2.9 nM and +26.5 ± 8.7 nM. The EC50s for epi compared with norepi were significantly larger with either BIIE0246 or PYY (P < 0.05), respectively, +10.4 ± 1.6 nM and +27.0 ± 9.5 nM. Maximal Iscs for epi and norepi in the presence of PYY were significantly more negative than those with BIIE0246 (P < 0.05), respectively, −14 ± 3 μA/cm2 and −21 ± 6 μA/cm2. Maximal Iscs for epi compared with norepi were not significantly different with either BIIE0246 or PYY (P < 0.05). B: EC50s of the Isc responses to norepi and epi in the presence of BIIE0246 and PYY were compared with those from untreated guinea pig distal colonic mucosa (shaded circle; 52), rabbit distal colonic mucosa (shaded square; 42), and the three β-adrenergic receptors (▵,▿, ◊; 41).

Subtype selective antagonism of epi response.

Sensitivity of the secretory response to epi was examined using the β-adrenergic subtype-selective antagonists CGP20712A (β1-AdrR), ICI-118551 (β2-AdrR), and SR59320A (β3-AdrR) to reexamine the synergistic antagonism observed previously (52). Variation in epi sensitivity induced by endogenous neuropeptide tone was addressed by using the Y2-NpR antagonist BIIE0246 and PYY to produce consistent and definable conditions. Responses were obtained in four adjacent mucosae from each colon with each group of four pretreated either with BIIE0246 or PYY. Neither CGP20712A nor SR59320A at 1 μM produced a large shift in epiEC50, indicating only minor antagonism (Fig. 11, A and B). ICI-118551 produced a shift in epiEC50 consistent with a Kd of ∼130 nM, which was 10- to 100-fold less efficacious than expected for antagonism at β2-AdrR (6, 29, 46). The relative shifts in epiEC50 were similar whether BIIE0246 or PYY was present, indicating that antagonist sensitivity was not influenced by Y2-NpR activation with PYY.

Fig. 11.

β1 and β2 subtype selective antagonists inhibited Isc synergistically. Isolated mucosae were stimulated by increasing concentrations of epi as in Fig 9. A: 4 mucosae from each animal were pretreated with PYY (0.3 μM) and a β-adrenergic antagonist (1 μM) either CGP20712A (▿, N = 10), ICI-118551 (◊, N = 10), SR59320A (▵, N = 4), or no antagonist control (•, N = 10). The epiEC50s were 13.7 ± 2.5 nM (control), 18.6 ± 5.4 nM (CGP20712A), 104 ± 12 nM (ICI-118551), and 17.7 ± 5.2 nM (SR59320A). The epiEC50 with ICI-118551 was significantly larger than control (P < 0.05). Inhibition constants (Kd) were calculated from epiEC50s obtained with antagonists, assuming competitive antagonism (40); Kds were 2800 ± 690 nM (CGP20712A), 152 ± 31 nM (ICI-118551), and 439 ± 89 nM (SR59320A). B: 4 mucosae from each animal were pretreated with BIIE0246 (1 μM) and a β-adrenergic antagonist (1 μM) as in A. The epiEC50s were 4.4 ± 0.9 nM (control, N = 10), 3.6 ± 0.9 nM (CGP20712A, N = 10), 44.7 ± 3.8 nM (ICI-118551, N = 10), and 6.9 ± 2.4 nM (SR59320A, N = 4). The epiEC50 with ICI-118551 was significantly larger than control (P < 0.05). The corresponding Kds were 109 ± 9 nM (ICI-118551), 1,230 ± 420 nM (SR59320A), and for CGP20712A was indeterminate but likely >3 μM. C: 4 mucosae from each animal (N = 6) were pretreated with PYY (0.3 μM) and β-adrenergic antagonists (1 μM) either CGP20712A (▿), ICI-118551 (◊), CGP20712A + ICI-118551 (○), or no addition control (•). The epiEC50s were 18.0 ± 5.5 nM (control), 22.0 ± 7.4 nM (CGP20712A), 115 ± 15 nM (ICI-118551), and 535 ± 65 nM (CGP + ICI). The predicted epiEC50 (black dotted line) for a single binding site including competitive antagonism from CGP20712A and ICI-118551 (40) using the Kds determined from A was 121 ± 13 nM. The measured epiEC50 for combined addition of CGP20712A and ICI-118551 was significantly higher (4.41 ± 0.99 fold, P < 0.05) than this predicted value for these 2 competitive antagonists acting at a single site. D: 4 mucosae from each animal (N = 6) were pretreated with BIIE0246 (1 μM) and antagonists as in C. The epiEC50s were 3.9 ± 1.1 nM (control), 2.8 ± 0.9 nM (CGP20712A), 46.0 ± 5.6 nM (ICI-118551), and 205 ± 66 nM (CGP + ICI). This measured epiEC50 for combined addition of CGP20712A and ICI-118551 was significantly higher (4.33 ± 1.24 fold, P < 0.05) than the predicted value for these 2 competitive antagonists acting at a single site, 47.4 ± 9.9 nM.

CGP20712A and ICI-118551 also were added in combination to quantify the synergistic antagonism (Fig. 11, C and D). The presence of both antagonists before stimulation resulted in a large shift in epiEC50, which was greater than expected from the simple combined action of these two antagonists acting at a single site in the receptor monomer (40). In the presence of BIIE0246 (Fig. 11, B and D), Isc reached a minimum at 0.1–0.3 μM epi and increased slightly at higher epi concentrations. This increase in Isc likely represented stimulation of a small sustained Cl secretion because Gt also increased at these concentrations (data not shown), and PYY suppressed this Isc response (Fig. 11, A and C). ICI-118551 alone at 1 μM eliminated this increase in Isc, indicating likely involvement of β2-AdrR; the apparent epiEC50 was ∼1 μM. The transient positive Isc also was eliminated by ICI-118551 at 1 μM but not by CGP20712A (data not shown). These results with subtype-selective antagonists suggested that two sites existed such that each antagonist could bind simultaneously to produce the synergistic action observed for activation of electrogenic K+ secretion.


The modulatory response characteristic of sympathetic activation in the distal colonic epithelium (33, 39, 52) includes sustained electrogenic K+ secretion and a transient Cl secretion involving stimulation by three signaling molecules (Fig. 12), norepi, ATP, and NPY (PYY). Both norepi and ATP likely are coreleased by sympathetic nerves in the mucosa because these two molecules are generally stored in the same secretory granules (17, 26). Once released, ATP can be converted to adenosine by ectonucleotidases (30), and adenosine produced a secretory Isc similar to the epi (norepi) response but with a smaller sustained rate (Fig. 7). Thus the presence of both agonists would not fundamentally alter the final secretory state. The third signaling molecule arrives via two routes, either release as NPY from nerves or as PYY from L cells (9, 10, 15, 17). Neuropeptide release is not tightly coupled to norepi release because NPY is stored within the nerve varicosities in separate granules that have a different sensitivity for stimulation of exocytosis (17). In addition, β-adrenergic activation of PYY release (5, 37) interposes the sensitivity factors of the L cells between the originating sympathetic signal and the ultimate ion secretory response. Arrival of PYY at the epithelial cells in the guinea pig distal colon resulted in a suppression of the Cl secretory component of the modulatory response (Figs. 1A and 7B). From these points, the sympathetic response can be seen to consist of signaling via two stimulatory types (norepi/epi and adenosine) and one inhibitory type (NPY/PYY). The magnitude of the response will be set by the concentration of these three agonist types, with the neuropeptides NPY and PYY primarily suppressing the transient Cl secretory component, leaving the sustained K+ secretory rate to be set by norepi/epi and adenosine concentrations. Because the sensitivity to epi was lowered by PYY acting via Y2-NpR (Fig. 10), sympathetic release of NPY (PYY) would lead to a need for higher concentrations of epi to elicit the maximal K+ secretory response.

Fig. 12.

Neural activation of modulatory mode electrogenic ion secretion. The colonic epithelium is shown schematically with an L cell (PYY releasing) of the enteroendocrine cell population between two columnar cells. The ion transport proteins necessary for electrogenic K+ and Cl secretion are indicated together with the receptor types examined in this study (β1-AdrR, β2-AdrR, Y2-NpR, A2B-AdoR, P2R). The upper columnar cell is secreting K+ and Cl, as illustrated by shading of the channels. Receptors that can activate this modulatory K+ and Cl secretion are shaded. The lower columnar cell is secreting only K+ because of the activation of Y2-NpR that suppresses Cl secretion. Apical membrane A2B or P2 receptors also activate K+ secretion without accompanying Cl secretion. Nerve processes with neurotransmitter-containing varicosities near the epithelial cells are shown from three populations, sympathetic inhibitory neurons (norepi, NPY, ATP), sympathetic secreto-inhibitory neurons (norepi), and cholinergic secretomotor neurons (ACh, NPY). Other ion transport proteins and receptors are present but were omitted for clarity.

Sympathetic control of ion secretion.

Sympathetic innervation of the distal colon is primarily via the superior and inferior mesenteric prevertebral ganglia (17, 26, 45). Neurons within these ganglia have distinct targets in the colon, which is reflected in the combination of neurotransmitters present. The secretomotor neurons in the submucosal plexus of the distal colon receive sympathetic innervation from neurons containing norepi and somatostatin that inhibit firing such that Cl secretion is reduced (8, 17). Concurrently, blood flow is reduced by sympathetic neurons containing norepi and NPY, which cause constriction of arterial vessels (26). The routes of blood vessels crossing into the mucosa were seen in the distal colon as routes for sympathetic nerves, some of which had a chemical coding for NPY as well as norepi (Fig. 8, C and F). These apparent mucosal nerves colocalizing norepi and NPY also were present in close proximity to crypt epithelial cells, which are likely sites of ion secretion.

Within the lamina propria surrounding the crypts, three populations of apparent nerves were observed, those exhibiting colocalization of norepi and NPY, those exhibiting only norepi, and those exhibiting only NPY (Figs. 8 and 12). Nerves with only NPY likely were cholinergic secretomotor nerves, and the other two types were sympathetic nerves (8, 9, 17). Norepi-containing nerves in the mucosa have been defined previously (17), but the present result indicated a distinction within this group. Those nerves with norepi and NPY would be capable of stimulating electrogenic K+ and Cl secretion as well as suppressing sustained Cl secretion, which would be added to their known action in vasoconstriction. The nerves with only norepi also would be able to stimulate electrogenic K+ and Cl secretion but would inhibit Cl secretion only through the norepi-stimulated release of PYY from L cells (5, 37). This second group of norepi containing nerves also likely contained either somatostatin or no peptide transmitter, similar to those innervating the submucosal and myenteric plexuses, respectively. Although sympathetic nerves are a minority of the nerves present, the actions on epithelial ion secretion would involve both an inhibition of secretomotor activity via submucosal ganglionic neurons and a direct inhibition of Cl secretion by norepi/NPY nerves. The resulting sympathetic state would not be secretory quiescence but rather a predominance of the modulatory secretory mode consisting of sustained electrogenic K+ secretion.

Purinergic activation of electrogenic K+ secretion.

Although the release of ATP likely occurs in response to a variety of stimuli including sympathetic nerve activity (17, 30), the resulting increase in negative Isc with ATP or adenosine (Fig. 7) also was interesting in regard to the group of stimuli that activate electrogenic K+ secretion. This common functional outcome suggests that similar intracellular signals may be involved for each of these receptor types. In guinea pig distal colon, the primary action on secretion appeared to occur through an extracellular processing of ATP to adenosine both at the apical and basolateral side of the cell (Figs. 7 and 12). The response to serosally added adenosine was modulatory secretion with a sustained Isc smaller than with epi. In the colonic tumor cell line T84 (3, 44), serosal adenosine produces Cl secretion via the A2B adenosine receptor (A2B-AdoR), but the absence of K+ secretion suggests that coupling to key K+ secretory elements is lacking in these cells. In contrast, the response to ATP in rat colon is Cl secretory via ATP acting directly on P2Y receptors without evidence for conversion to adenosine (31). Mucosally added ATP had two components in guinea pig distal colon, both consistent with electrogenic K+ secretion (Fig. 7, C and D). The sustained component of this negative Isc response was via adenosine action, which differs from the Cl secretory response of T84 cells to mucosal adenosine (3). Similar to mouse and rat colon (28, 35), the transient component in guinea pig distal colon apparently was activated by unmodified ATP (Fig. 7C). The differences among these responses likely reflects unique physiological needs of each species. Importantly, epithelial cells studied ex vivo from several species were capable of producing electrogenic K+ secretion in response to ATP.

Suppression of ion secretion by NPY receptors.

The neuroendocrine peptides NPY and PYY activate the family of NPY receptors, Y1-NpR, Y2-NpR, Y4-NpR, and Y5-NpR (4). The Y-NpRs occurring in colonic epithelial cells are Y1, Y2, and Y4, as demonstrated by RT-PCR and immunofluorescence localization (9, 21, 34). The present results indicated Y1-NpR and Y2-NpR on epithelial cells in guinea pig distal colon both by RT-PCR (Fig. 3) and immunofluorescence localization (Figs. 5 and 6). Immunoblots suggested that these Y-NpRs may have undergone posttranslational modification including oligomerization (Fig. 4) as has been seen in other cell types (12, 16). In mouse colon, the direct inhibitory action of NPY (PYY) on Cl secretion operates via Y1-NpR (9). Perhaps an example of species variability, the guinea pig inhibitory action clearly was via Y2-NpR, on the basis of selective antagonists as well as the selective Y2-NpR agonist PYY(3-36) (Fig. 2). Because Y1-NpR is not found in rat colonic epithelial cells by RT-PCR (21), the inhibitory action in rat (43) also may operate via a Y-NpR other than Y1-NpR. The present result in guinea pig distal colon, indicating the lack of effect with pancreatic polypeptide, suggested that, even if present, Y4-NpR was not involved in the inhibition of Cl secretion. In addition to actions on epithelial cells, NPY (PYY) inhibits neuronal activity via Y1-NpR and Y2-NpR such that excitatory neural inputs to epithelial cells are reduced (9, 10).

The secretory state of the colonic epithelium depends on an integration of excitatory and inhibitory inputs. In addition to the primary balance between activation by the enteric nervous system and inhibition by the sympathetic nervous system, the neuropeptide tone of the mucosa serves an important role in adjusting the rate and character of the secretion produced (8, 9). Even in the isolated mucosa, L cells and remnant nerve fibers are present, and effects consistent with in situ release were apparent. Actions on both secretagog-associated (Fig. 9A) and neural-associated (Fig. 9B) secretion suggested that PYY and/or NPY were present in the bathing solution, likely released from nerve processes and L cells. Although in vitro changes in PYY or NPY concentrations may not be large (2, 5), an increase in concentration near the epithelial cells may have been sufficient to produce these Isc changes. A portion of the NPY and PYY released will be cleaved by ectopeptidases, with dipeptidyl-peptidase-4 producing NPY(3-36) and PYY(3-36), which are selective for Y2-NpR (15, 48). This extracellular processing has important consequences for digestion beyond the action on epithelial secretion because PYY reaching the brain through the circulation promotes feeding via Y1-NpR as opposed to PYY(3-36), which promotes satiety via Y2-NpR (48).

Adrenergic sensitivity of electrogenic ion secretion.

A distinguishing feature of the adrenergic response in distal colon is the relative insensitivity to epi and norepi (42, 52). Each of the β-AdrRs has a distinct sensitivity to epi compared with norepi (6, 46) so that cells with these receptors will have an inherent preference for sympathetic signaling through neuronal release of norepi vs. blood-borne epi. The low and equal sensitivity of the distal colon to epi and norepi (Fig. 10B) supported the concept that the sensitivities of the β-AdrRs involved have been altered by posttranslational modification and/or interaction with auxiliary signaling components. Sensitivity to antagonists also appeared to have been reduced (Fig. 11). In addition, activation of Cl secretion appeared ∼200-fold less sensitive to epi than K+ secretion, as indicated by the increased Isc observed at high concentrations, which may involve only β2-AdrR (Fig. 11B). The ability of the Y2-NpR antagonist BIIE0246 to sensitize the adrenergic response by roughly fourfold over the PYY-activated state indicated that Y2-NpR signaling specifically controlled β-AdrR sensitivity to epi and norepi. With this regulatory feature, the adrenergic response potentially could be adapted to varying physiological requirements.

The ability of PYY to suppress Cl secretion coincides with the primary mode of signaling by β-AdrRs and Y-NpRs, controlling intracellular cAMP. With adenylyl cyclase as the key intermediary, β-AdrRs would lead to activation via Gαs and Y2-NpR would lead to inactivation via Gαi. Interestingly, in guinea pig distal colon, the adenylyl cyclase stimulator forskolin produces a negative Isc consistent with K+ secretion but leads to Cl secretion only at higher concentrations (32). This difference together with the lower sensitivity for β-AdrR activation of Cl secretion compared with K+ secretion could indicate that coupling to intracellular signaling cascades was dependent on receptor occupancy, as suggested for other cell types (1, 53). The largely transient nature of the Cl secretory component also was consistent with the β-AdrR responses seen in other cell types. Especially in cardiac cells, β-AdrR responses are brief because of inactivation and internalization of the receptor (1, 6, 53). The sustained nature of the K+ secretory component was distinct from these other β-AdrR actions because it apparently lacked these response-limiting controls that can be exerted on the receptor. Interestingly, the colonic tumor cell line T84 responds to epi in a β-AdrR manner with a transient Isc consistent simply with Cl secretion (51). The features of β-AdrR signaling in the distal colon that permit electrogenic K+ secretion to persist after termination of Cl secretion remain to be elucidated.

AdrR activation of electrogenic K+ secretion.

A definition of which β-AdrR is coupled to a physiological response requires a means to link the receptor to the function. The array of available pharmacological agonists and antagonists for β-AdrRs was used in this study to provide that link with sustained electrogenic K+ secretion. Although indications of β-AdrR involvement instead of α-AdrR are clear (52), the β-AdrR subtype involved remains less conclusive. Because both β1-AdrR and β2-AdrR antagonists inhibited the epi-induced K+ secretory response (52), the simplest interpretation would be to choose the antagonist that had the highest efficacy. However, neither the β1-AdrR nor the β2-AdrR antagonist inhibit at the low concentrations expected for these receptor subtypes (6, 29, 46), and the greatest inhibition occurs when both are present (52). Importantly, the decreased sensitivity to epi produced by PYY was not accompanied by a change in antagonist efficacy, as indicated by the similar actions in the presence of PYY or the Y2-NpR antagonist BIIE0246 (Fig. 11). The Kd of the β2-AdrR antagonist ICI-118551 (∼130 nM) was near the value expected for its action at β1-AdrR, and the Kd of the β1-AdrR antagonist CGP20712A (≥3 μM) was near the value expected for its action at β2-AdrR. If this antagonist sensitivity had been attributable to the action of two antagonists acting at a single binding site on a receptor, then the combined response can be calculated from the individual responses. However, the antagonism produced in the presence of both was roughly fourfold larger than this prediction (Fig. 11, C and D), implying instead that two sites were present to allow for this synergistic antagonist interaction. Because both CGP20712A and ICI-118551 are well-characterized competitive antagonists, the possibility of the interaction occurring between the orthosteric epi site and a putative allosteric site seems unlikely. Given that G protein-coupled receptors have been shown to form homodimers, heterodimers, and other oligomers (18, 38) as well as the apparent β-AdrR oligomers present in colonic epithelial cells (52), a more likely possibility for the two sites would be heterodimers of β1-AdrR and β2-AdrR on the colonic epithelial cells.

Evidence for functional β1-AdrR/β2-AdrR heterodimers has been shown previously as differences in antagonist binding constants and in activation of adenylyl cyclase (29). The observed synergistic antagonism of electrogenic K+ secretion (Fig. 11, C and D) implied that simultaneous activation of both β1-AdrR and β2-AdrR was required to produce appropriate intracellular signaling. An expectation of this possibility might be that the slope of the epi concentration dependence would be steeper than that for a single-site interaction, remembering the caveat that this would occur only if the two sites exhibited positive cooperativity. With that point in mind, the observation that the slope was consistent with a single site cannot be taken as evidence for the lack of interaction between β1-AdrR and β2-AdrR. A two-site receptor system (dimer) could have the observed slope (Figs. 10 and 11) if the two sites were on separate monomers and epi binding was independent, as long as the conformational changes leading to G protein activation in the two receptor monomers were linked. Because the receptor response in this study was inferred from the functional endpoint of ion secretion, any subtle differences anticipated among these kinetic schemes may have been obscured. Ultimately the adrenergic response of epithelial cells in the distal colon would be a combination of multiple signaling components, with the present results indicating that signaling to produce modulatory secretion could be adjusted at several points to affect the amount of Cl and K+ secretion.


This study was supported by a grant from the National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases (DK65845).


View Abstract