INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

CCK IS A PEPTIDE that is released into extracellular space by I-cells of the small intestinal mucosa. Although it is the product of a single gene, CCK is released in forms ranging from 8 to 58 amino acids in length (14, 18, 22). CCK promotes the delivery of bile into the small intestine, stimulates pancreatic enzyme secretion, inhibits gastric emptying, promotes intestinal peristalsis, and induces satiation (30). Many of these effects of CCK, including stimulation of pancreatic exocrine secretion (28), inhibition of gastric emptying (23, 36, 42), and satiation (43, 46), are mediated by activation of CCK-A receptors on vagal afferent fibers (1, 4, 28, 36, 54, 56). Although plasma CCK concentrations are elevated after entry of fats and proteins into the small intestine (5, 30), mounting evidence suggests that CCK mediates some responses to intestinal stimulation without entering into the circulation. Hence, several investigators have proposed that some responses to CCK are mediated via a paracrine action of the peptide to stimulate vagal afferent nerve endings in the intestinal mucosa (5, 15, 21, 45).

Vagal responsiveness to CCK has been demonstrated by recordings from vagal afferent fibers (3, 10, 12, 45) and nodose ganglion cells (29) in vivo. In addition to these direct actions, CCK has been shown to enhance vagal responses to mechanical stimuli in vivo (12, 45). Finally, several reports indicate that chronic elevation of CCK levels reduces sensitivity to acute elevation of CCK (8, 11) and that chronic exposure to high-fat or high-protein diets, which release CCK, also reduces sensitivity to the satiation effects of acute CCK (9). Unfortunately, the cellular and biochemical basis for CCK-induced activation of vagal afferent fibers remains unappreciated, and thus the cellular basis for these interactions and adaptive responses is unknown.

A CCK-induced increase in cytosolic calcium could play a major role in the vagal afferent response to CCK and in adaptive and synergistic actions of CCK with other vagal stimuli. Therefore, in the present report we have used cultured nodose ganglion neurons to test the hypothesis that increases in intracellular calcium constitute a part of the vagal afferent response to CCK. We found that sulfated CCK-8 induced an increase in cytosolic calcium in 40% of the vagal afferent neurons tested. This increase in calcium was triggered by CCK action at the low-affinity site of the CCK-A receptor and was dependent on extracellular calcium. Finally, a few cells were found to respond to nonsulfated CCK-8, indicating that CCK-B receptors also can trigger increases in cytosolic calcium in a small subpopulation of vagal afferent neurons.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Cell isolation and culture. Nodose ganglia were excised from anesthetized male Sprague-Dawley rats (ketamine 25 mg/100 g, xylazine 2.5 mg/100 g) under aseptic conditions and placed in ice-cold Ca²⁺- and Mg²⁺-free HBSS supplemented with antibiotic (penicillin-streptomycin, 100 U/ml and 100 µg/ml, respectively). The ganglia were then cleaned of connective tissue and desheathed. For each rat both left and right ganglia were combined into a single isolation. Nodose cells were dissociated according to a procedure described by Lancaster et al. (25). Briefly, the desheathed ganglia were washed once in HBSS, placed in ~3 ml of digestion buffer (1 mg/ml dispase II and 1 mg/ml collagenase type Ia in HBSS), sliced into ~1-mm³ fragments using a sterile scalpel blade, and then placed in an incubator at 37°C to digest for 90 min. At the end of the digestion the cells were dispersed by gentle trituration through Pasteur pipettes coated with Sigmacote. The dispersed cells were then washed two times in HEPES-buffered DMEM (HDMEM), supplemented with 10% FCS and antibiotic (penicillin-streptomycin). After the final resuspension the cells were plated onto poly-L-lysine-coated coverslips (100 µg/ml polylysine for 30 min) and grown in HDMEM with 10% FCS at 37°C in an atmosphere of 5% CO₂-95% air. In some isolations the procedure of Lankisch et al. (26) was followed. Briefly, dispase II was replaced with trypsin II, the digest proceeded for only 30 min, and the digest reaction terminated with soybean trypsin inhibitor. All experimental procedures were conducted within 2 days of collection of the nodose ganglia.

Calcium measurements. Intracellular calcium levels were determined by ratiometric measurements with fura 2. All manipulations and experiments were performed at room temperature in a physiological salt solution (in mM: 140 NaCl, 5 KCl, 2 CaCl₂, 1 MgCl₂, 6 glucose, and 10 HEPES with the pH adjusted to 7.4 with NaOH) (standard bath) unless noted. Any ionic adjustments to the standard bath were made such that the adjusted solution remained isosmotic (i.e., Ca²⁺ replaced with Mg²⁺ and Na⁺ replaced with K⁺). Cells on coverslips were washed in standard bath and then allowed to take up fura 2-AM for 30 min (1 µM with 0.05% pluronic acid), followed by a 30-min wash period. Coverslips were then mounted in a closed chamber (~0.3 ml vol) through which solutions were perfused by gravity (~3 ml/min). Solution changes were made by switching inflow lines through a common manifold (new solutions reach cells ~15 s after switch). Chambers were then mounted on the stage of a Nikon Diaphot inverted microscope and examined with a ×40 oil immersion lens. Cells were alternatively exposed to 340- and 380-nm light (800- and 400-ms exposures, respectively), and images were collected through a 510-nm filter with a CCD camera (SenSys camera, Photometrics, Tucson, AZ). Image pairs were captured every 6 s, and ratios of fluorescent intensities at the two excitation wavelengths were obtained from regions over individual cells. Data collection and manipulations were performed with MetaFluor Software (Universal Imaging, West Chester, PA). Calcium concentrations were then determined by comparing ratio values to a standard curve obtained in a bath containing 10 µM fura 2, 130 mM KCl, 10 mM MOPS buffered to pH 7.4 with KOH, 10 mM EGTA, and various concentrations of CaCl₂ (0 to 10 mM). Free calcium concentrations were calculated using the computer program EQCAL (Biosoft, Ferguson, MO).

Cell selection. Nodose neurons were easily identified and selected based on their large, round cell bodies. Nonneuronal cells had spindle or filamentous shapes. In most cases basal calcium levels were below 200 nM; however, in some neurons a challenge with CCK resulted in a large calcium increase that did not completely recover. Whether this lack of recovery was due to the neuron entering a prolonged activated state or was due to its inability to regulate calcium because of cellular damage could not be determined. Thus we only analyzed those neurons in which calcium increases reversed and that maintained basal calcium levels of <400 nM throughout the experiment. Finally, at the end of every experiment, cells were depolarized with saline containing 55 mM K⁺. Any cells that did not exhibit a calcium increase of at least 40 nM in response to K⁺ were excluded from further analysis. A total of 465 neurons studied from 35 separate isolations met these criteria.

Experimental protocols. Challenge solutions were applied for 1 min or until the cytosolic calcium signal stopped increasing, whichever was longer. Challenges of 55 mM KCl were for 30 s or until the calcium response stopped increasing, whichever was longer. Repeated challenges were typically separated by 2-3 min unless recovery from a previous response required more time. Administration of CCK in combination with CCK-A receptor antagonist was preceded and followed by administration of CCK alone in standard bath. This approach ensured that attenuation of calcium signal in the presence of the antagonist was not the result of desensitization. In general, a similar design was used to examine the effect of calcium-free bath on CCK-induced responses. However, for some of the calcium-free bath experiments the first CCK exposure was made in the absence of bath calcium and then followed by a second and third challenge with CCK after the return of bath calcium. Once again, we took this approach to be sure that any attenuation of the CCK response observed under calcium-free conditions could not be ascribed to desensitization. Accordingly, with the exception of the experiments employing thapsigargin (see below), results from experiments involving multiple CCK exposures were analyzed only when neurons responded to CCK alone after all other experimental manipulations were complete. Thapsigargin was used to test the possibility that intracellular stores of calcium were involved in the CCK-induced calcium response. Because thapsigargin is irreversible, a challenge after washout of thapsigargin could not be performed. In experiments comparing the sensitivity of the neurons to nonsulfated CCK vs. sulfated CCK-8, or CCK-8 and JMV-180, the order of exposure to the various agents was randomized to ensure that the exposure to the first challenge was not responsible for the lack of response to the second challenge. For each experimental protocol, nodose neurons from at least three separate isolations were used.

Data analysis. Calcium responses to CCK and/or other challenges are expressed as the change in cytosolic calcium from a basal value (calcium). The basal value was sampled over the 30 s just before exposure to the challenge solution. Summarized results are expressed as the average calcium ± SE. For experimental protocols that involved three challenges (control, experimental condition, recovery), a repeated-measures two-factor ANOVA was performed. One factor was the experimental condition and the other was the individual neuron. This design was selected because the absolute magnitude of the control responses could vary from one neuron to the next. As mentioned above, for some of the calcium-free bath experiments, the calcium-free condition was tested first, followed by two control challenges under standard conditions. In these cases, the first response to CCK in the presence of standard calcium condition was used for the control condition. If ANOVA revealed a significant effect, subsequent multiple comparisons were made using a Newman-Keuls procedure. For thapsigargin experiments a paired t-test was used to evaluate statistical significance of the results. Comparisons of cell size, basal calcium levels, and KCl-induced calcium responses between populations (CCK responsive vs. CCK nonresponsive) were performed by an unpaired t-test. In all statistical tests P < 0.05 was considered significant.

Chemicals. In all experiments the sulfated form of CCK-8 was used unless indicated. Culture media and FCS were obtained from Life Technologies (Grand Island, NY). Lorglumide was obtained from Research Biochemicals International (Natick, MA). Dispase II was obtained from Boehringer Mannheim (Indianapolis, IN). Fura 2 pentapotassium salt and fura 2-AM were obtained from Molecular Probes (Eugene, OR). JMV-180 was obtained from Research Plus (Bayonne, NJ). All other chemicals were obtained from Sigma (St. Louis, MO).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

General characteristics of responses to CCK. Of the 465 cultured nodose neurons challenged with CCK-8 (10 or 100 nM), 184 (or 40%) responded with a transient increase in cytosolic calcium. Most neurons exhibited a relatively stable baseline calcium level with fluctuations no greater than 5 nM in amplitude. For these neurons to be categorized as responding to CCK-8, we required that CCK-8 induced a calcium increase greater than 20 nM above baseline. On the other hand, a few neurons exhibited spontaneous calcium oscillations. For these neurons to be categorized as responding to CCK-8, we required that the CCK-8-induced calcium response be at least twice the magnitude of any spontaneous oscillation observed before the CCK-8 challenge. When the maximal calcium response to CCK-8 was considered, regardless of concentration of CCK-8, we found that the average ± SE increase in intracellular calcium was 131 ± 9 nM (n = 184), the largest increase we observed was 865 nM, and the median increase was 94 nM.

Concentration dependence of responses to CCK. We performed a series of studies to examine the dependence of the calcium responses on CCK-8 concentration (10 pM to 300 nM). Some neurons exhibited a monotonically increasing calcium response to increasing concentrations of CCK-8 (Fig. 1A). However, summarizing concentration-response characteristics was complicated by several problems. First, in many neurons, but not all, the response to CCK-8 exhibited desensitization after repeated challenges with CCK-8 (Fig. 1B). This desensitization could be striking in that ~20% of the neurons tested (24 of 116 given repeated challenges) responded to CCK-8 only one time (for example, Fig. 2A). Furthermore, some neurons responded repeatedly to CCK-8, but their responses were small, making expression of the responses as a percentage of a standard response susceptible to large errors (Fig. 2B). We also observed wide variation in the threshold for responses to CCK-8. In a series of concentration-response experiments, 6 of 41 neurons that eventually responded to 10 nM CCK-8 responded to 10 pM CCK-8 (15%), and in a study on a separate group of neurons, 6 of 44 neurons that eventually responded to 10 or 100 nM of CCK-8 responded to 100 pM CCK-8 (14%). On the other hand, of 30 responsive neurons tested with a series of CCK-8 concentrations up to or greater than 100 nM, 6 (20%) required at least 100 nM CCK-8 before they responded. Thus not only was there a large variation in the magnitude of responses, but the threshold CCK-8 concentration necessary to induce a response varied over 10,000 fold.

View larger version (18K): [in this window] [in a new window]   Fig. 1.   A: example of concentration-response characteristics of calcium responses induced by CCK-8 in an isolated nodose neuron. Trace represents the calcium signal from a single nodose neuron (experiment no. at lower right). Bars over trace represent when inflow line was switched to a test solution. In this and subsequent figures, when a bar is labeled with a number, it indicates the concentration of CCK-8 (in nM) in the test solution. Bars labeled with K indicate when cell was challenged with 55 mM K+. B: example of modest desensitization of the calcium response to repeated CCK-8 challenges.

View larger version (16K): [in this window] [in a new window]   Fig. 2.   A: example of a neuron showing complete desensitization after a single CCK-8 challenge. B: example of a neuron showing small but reproducible calcium responses to CCK-8. See Fig. 1 legend for further description of bar labels.

Role of bath calcium in responses to CCK. We next examined whether calcium responses to CCK-8 were dependent on extracellular calcium. When calcium was removed from the bathing media, the basal cytosolic calcium concentration dropped slightly (Fig. 3). In 13 of 16 neurons challenged in the absence of calcium, the response to CCK-8 was completely eliminated (for example, Fig. 3A). In the remaining three neurons, the increase in cytosolic calcium was greatly reduced, but a small residual calcium response could still be detected (for example, Fig. 3B). In these three neurons the calcium response in the absence of bath calcium was 48 ± 13% of the control response (average ± SE). When all 16 neurons tested were considered, the average calcium response after removal of bath calcium was significantly reduced (Fig. 3C). After return of calcium to the bathing media, the CCK-8-induced calcium response was significantly greater than the response in the absence of bath calcium (Fig. 3C). These results show that in most, if not all, nodose neurons the calcium response to CCK-8 is due entirely to the influx of calcium from the extracellular milieu but that recruitment of calcium from intracellular stores might occur in a few neurons.

View larger version (25K): [in this window] [in a new window]   Fig. 3.   Effect of removing bath calcium on the calcium response to CCK-8. A: example of a neuron that demonstrated a complete lack of calcium response to CCK-8 when bath calcium was removed (Ca). Arrow indicates lack of response. B: example of a neuron that could still partially respond even though bath calcium was removed. C: summary of results from removing bath calcium during the CCK exposure. Results are expressed as change in cytosolic calcium (calcium) from basal levels (averages ± SE, n = 16). Con, response to 1st control CCK exposure; no Ca, response to CCK in the absence of bath calcium; Rec, response to a 2nd exposure of CCK in the presence of bath calcium.. * P < 0.001 compared with 1st CCK response. ** P < 0.001 compared with response in absence of bath calcium and P < 0.05 compared with 1st CCK challenge. See Fig. 1 for further description of bar labels.

Role of intracellular calcium stores in responses to CCK. The conclusion that extracellular calcium is the main source of calcium for CCK-8-induced calcium responses is further supported by the observation that thapsigargin, an irreversible inhibitor of the smooth endoplasmic reticular Ca-ATPase (51), failed to abolish the calcium response to CCK-8. Thapsigargin by itself caused a slow increase in calcium levels in all nine neurons tested (Fig. 4A; average increase ± SE in response to thapsigargin alone was 84 ± 20 nM, P < 0.005), indicating that the Ca-ATPase was functional in these neurons. In eight of the nine neurons, CCK-8 still induced a calcium response after thapsigargin. Indeed the calcium response after thapsigargin was 108 ± 23% of the pre-thapsigargin response (minimum 63%, maximum 244%; summarized in Fig. 4B).

View larger version (13K): [in this window] [in a new window]   Fig. 4.   A: effect of thapsigargin (50 nM; bar labeled Tg) on basal calcium levels and the calcium response to CCK-8. B: summary of results from thapsigargin experiments. Results are expressed as the change in cytoplasmic calcium (calcium) from basal levels (averages ± SE, n = 9). Con, response to CCK during the 1st CCK exposure; +Tg, response to CCK after thapsigargin exposure. Treatment with thapsigargin had no effect on CCK-induced calcium responses (P = 0.845). See Fig. 1 for further description of bar labels.

Responses to sulfated vs. nonsulfated CCK-8. Sulfated CCK-8 activates both CCK-A and CCK-B receptors (38), while nonsulfated CCK-8 activates CCK-B receptors but not CCK-A receptors (16). Overall, only 5% of neurons (3/62) challenged with nonsulfated CCK-8 responded. Because so few neurons responded to nonsulfated CCK-8, these responses were not further characterized. However, in 16 neurons that had positive responses to 10 or 100 nM sulfated CCK-8, 14 had no response to an equal concentration of nonsulfated CCK-8 (for examples, see Fig. 5, A and B) while two neurons responded to both forms of CCK-8. In these two neurons, the calcium response to nonsulfated CCK-8 averaged just 34% of the response to sulfated CCK-8 (data not shown). Finally, one additional neuron exhibited a large robust response to nonsulfated CCK-8 but did not respond at all to sulfated CCK-8 (Fig. 5B).

View larger version (17K): [in this window] [in a new window]   Fig. 5.   Examples of responses to sulfated and nonsulfated CCK-8 (CCK-8 analog indicated). A: CCK-8 concentration was 10 nM for both analogs. B: responses from 2 neurons within the same image are illustrated. CCK-8 concentration was 100 nM for both analogs. See Fig. 1 for further description of bar labels.

Effect of CCK-A receptor antagonist on responses to CCK. In 14 of 14 neurons tested, lorglumide, a potent and selective CCK-A receptor antagonist (33), completely abolished responses to CCK-8 (for example, Fig. 6A; results summarized in Fig. 6C). The average response in the presence of lorglumide was just 2 ± 1% of the response before lorglumide. After lorglumide was washed from the neurons, the response to CCK returned to 53 ± 7% of the response before lorglumide. This represented a statistically significant recovery of the response (Fig. 6C).

View larger version (18K): [in this window] [in a new window]   Fig. 6.   A: example of effect of the CCK-A antagonist lorglumide (10 µM, applied as indicated) on CCK-8 response. B: effects of JMV-180 on CCK-8 responses. Cells were exposed to JMV-180 as indicated (JMV-0.1, 0.1 nM JMV-180 was applied; JMV-100, 100 nM JMV-180 was applied). Break in x-axis was when data were not collected. C: summary of results from experiments with lorglumide. Results are expressed as calcium from basal levels (averages ± SE, n = 14). Con, the response to the 1st exposure to CCK; +Lorg, response to CCK in the presence of lorglumide; Rec, response to CCK after lorglumide was washed from the neurons. * P < 0.001 compared with 1st CCK exposure. ** P < 0.005 compared with the response in the presence of lorglumide and P < 0.001 compared with 1st control response. D: summary of results from experiments with JMV-180. Results are expressed as calcium from basal levels (averages ± SE, n = 8). Con, response to the 1st exposure to CCK; +JMV, response to CCK in the presence of 100 nM JMV-180; Rec, response to CCK after JMV-180 was washed from the neurons. * P < 0.001 compared with 1st CCK exposure. ** P < 0.025 compared with CCK response in the presence of JMV-180 and P < 0.05 compared with the 1st CCK response. See Fig. 1 for further description of bar labels.

Effect of JMV-180 on responses to CCK. We next used the CCK analog JMV-180 to investigate the role of the high- and low-affinity sites of the CCK-A receptor in the calcium response. In the rat, JMV-180 is an agonist for the high-affinity site (K_d ~70 pM) and an antagonist for the low-affinity site (K_d ~10 nM) (49, 58). In a first set of experiments, we first challenged neurons with CCK-8 to determine if a neuron would respond to CCK. If a neuron responded to CCK-8, it was then challenged with a concentration of JMV-180 (100 pM) that would be adequate to activate the high-affinity site but would have minimal actions on the low-affinity site of the CCK-A receptor. In some neurons the order of the first challenge was reversed to ensure that prior CCK-8 exposure was not interfering with any JMV-180 responses. Neurons were then challenged with a higher concentration of JMV-180 (100 nM) that would block most low-affinity sites. In the continued presence of JMV-180, the neurons were then rechallenged with CCK-8 (10 or 100 nM, concentration matching the first CCK-8 challenge). After washing JMV-180 from the neurons, the neurons were then challenged a final time with CCK-8 alone. In 16 neurons identified as CCK-8 responsive, neither 100 pM nor 100 nM JMV-180 induced any response by itself (see example in Fig. 6B). In eight of these neurons, we could obtain no calcium response to CCK-8 in the presence of JMV-180 or after JMV-180 was washed from the neurons (data not shown). In the other 8 neurons, the calcium response to CCK-8 in the presence of JMV-180 (100 nM) was only 52 ± 12% of the first response to CCK-8 (for example, Fig. 6B; summarized results in Fig. 6D), a statistically significant decrease. In these cells the response to the final challenge with CCK-8 alone returned to 79 ± 9% of the first response, which represented significant recovery of the response from that in the presence of 100 nM JMV-100 (Fig. 6D). Using 100 or 10 nM CCK-8 for the standard challenge did not produce significantly different results (data not shown). These results suggest that the low-affinity site of the CCK-A receptor is responsible for the calcium signal induced by CCK-8 in cultured nodose neurons.

View larger version (17K): [in this window] [in a new window]   Fig. 7.   Examples of responses to CCK-8 and JMV-180 in cells isolated by a trypsin digest rather than with dispase (concentrations in nM are indicated). In A-D, the thin line indicates when 10 pM CCK-8 was applied and the thick bar indicates when the bathing solution was switched to 10 nM CCK, or other indicated agents. See Fig. 1 for further description of bar labels.

CCK-responsive vs. CCK-insensitive neurons. A final issue we addressed was whether some property of the neurons that responded to CCK-8 was different from CCK-8-insensitive neurons. We examined neuron size, basal calcium levels, and responses to KCl. Neuron size, estimated from phase contrast pictures of neurons, varied >10-fold (smallest 79, largest 908; reported in arbitrary units based on number of pixels in the image covered by the neuron). The size of CCK-8-responsive neurons was not significantly different from CCK-8-insensitive neurons [responsive, 349 ± 23 (n = 41); insensitive, 360 ± 25 (n = 48); averages ± SE, P = 0.76, 2-tailed t-test]. Basal calcium levels were also not significantly different [responsive, 166 ± 4 nM (n = 184); insensitive, 169 ± 3 nM (n = 281); averages ± SE, P = 0.58, 2-tailed t-test], nor were responses to KCl [responsive, 412 ± 22 nM (n = 184); insensitive, 406 ± 18 nM (n = 281); averages ± SE, P = 0.83, 2-tailed t-test].

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

General characteristics of the CCK-8-induced calcium response in nodose neurons. The results presented in this report demonstrate that CCK-8 triggers an increase in cytosolic calcium in neurons cultured from adult rat nodose ganglia. In some neurons calcium responses could be observed with concentrations as low as 10 pM, but some neurons required at least 100 nM CCK-8 to elicit a response. This signal depends on the presence of extracellular calcium. In contrast to the study by Lankisch et al. (26), our results suggest that this response is mediated primarily by activation of the low-affinity site of the CCK-A receptor. In a few neurons we found evidence that CCK-B receptors also are capable of inducing a calcium response. Overall we found that ~40% of the cultured nodose neurons tested responded to CCK-8. This percentage is consistent with the 33% of nodose neurons reported to express CCK-A receptor mRNA (6) and the percentage of nodose neurons reported to express a calcium response to 1 nM CCK-8 (45%; 26). Broberger et al. (6) also reported that just 9% of nodose neurons expressed mRNA coding for the CCK-B receptor, a number in close agreement with the 5% of the cultured neurons we observed that responded to nonsulfated CCK-8.

Validity of cultured nodose neuronal model. Although it would be ideal to examine changes in cytosolic calcium in the peripheral terminals of vagal afferent neurons, these structures are too small to allow for convenient biochemical analysis of their responses. Furthermore, examination of vagal afferent fibers themselves necessitates inclusion of end organ tissues that they innervate. These tissues may release neuroactive substances in response to CCK, making it difficult to distinguish direct vagal responses to CCK from responses to CCK-induced release of other vagal stimulants. Because the cell bodies for all vagal afferent fibers reside within the nodose ganglia, a method that removes these limitations is to examine the actions of CCK on neurons dissociated from the nodose ganglia and maintained in short-term culture. Previous studies have established that vagal afferent neurons from the nodose ganglia retain their electrophysiological characteristics in short-term culture (27, 39). Furthermore, receptor autoradiographic studies indicate that cell bodies of neurons within the nodose ganglia normally express CCK receptors (20, 37, 55). Therefore, it is likely that CCK-induced responses in cultured nodose neurons reflect the nature of responses of vagal afferent neurons in vivo.

Calcium responses in nodose neurons vs. other cell types. CCK induces an increase in cytosolic calcium in a variety of different cells; however, the nature of the CCK-induced calcium signal varies depending on the preparation under study and concentration of CCK used. These CCK-induced calcium signals range from an oscillating signal independent of extracellular calcium (52, 57), to a single spike of calcium from intracellular sources (19), to a spike and plateau pattern with the spike from intracellular sources and the plateau from extracellular sources (13, 24, 31, 34, 41, 44, 47, 48), to a single spike of calcium that is totally dependent on extracellular calcium (35).

Lack of role for the high-affinity site of the CCK-A receptor in nodose calcium responses. An apparent discrepancy between our results and previous studies (26, 29) is that we did not observe convincing responses to JMV-180, an agonist at the high-affinity site of the CCK-A receptor. An explanation for this discrepancy is not immediately obvious. We replicated the conditions of Lankisch et al. during the process of revising this report. However, we could find little evidence for a unique oscillatory response to low concentrations of CCK-8 or the high-affinity site agonist JMV-180. We did observe that concentrations of CCK-8 as low as 10 pM could activate a small minority of nodose neurons, but these responses were not qualitatively different from those activated at higher concentrations. Even Lankisch et al. (26) found that the responses to low concentrations of CCK-8 or CCK-OPE were qualitatively similar to those they observed with higher concentrations of CCK-8 in that all responses were dependent on extracellular calcium. This would suggest that these high- and low-dose responses are fundamentally similar rather than a result of activating a unique oscillatory mechanism as occurs in pancreatic acinar cells. Furthermore, we would submit that while nodose neurons exhibit a wide range in sensitivity to CCK, most data are consistent with the interpretation that vagal afferent calcium responses are all dependent on activation of a low-affinity site. Finally, a speculation that is consistent with all observations of the oscillatory nature of the CCK-8-induced calcium signal reported by us and by Lankisch et al. (26) is that the nodose neurons are oscillating into and out of spiking behavior with resulting rounds of calcium influx through voltage-dependent calcium channels. The weaker depolarizing effect of low concentrations of CCK-8 may make it more likely that oscillations are observed. Additional work is needed to test this hypothesis.

Role of CCK-B receptor in CCK-8-induced responses. While sulfated CCK-8 exhibits much higher affinity for the CCK-A receptor than nonsulfated CCK-8, both the sulfated and nonsulfated peptides bind with equal affinity at the CCK-B receptor. Thus responses to nonsulfated CCK-8 usually are indicative of CCK-B receptor activation. Although we did observe a few calcium responses to nonsulfated CCK-8, these responses occurred so infrequently that it was not possible to characterize them pharmacologically. CCK-B receptors reportedly are expressed by nodose neurons (6, 7, 32, 37); however, to date, no physiological functions have been linked to CCK-induced activation of CCK-B receptors on vagal afferent fibers.

Variation in threshold response to CCK. We observed that the threshold response to CCK-8 in cultured vagal afferent neurons could vary as much as 10,000-fold between individual neurons. In vivo electrophysiological experiments also have revealed variations in CCK sensitivity in vagal afferent fibers innervating the gastrointestinal tract. For example, Blackshaw and Grundy (3) concluded that most vagal afferents innervating the gastric wall were only indirectly sensitive to CCK by virtue of its effects on gastric smooth muscle (100-200 pmol, by near arterial infusion). On the other hand, vagal afferents that innervated the antral or duodenal mucosa were directly excited by CCK at near arterial doses ranging from 3 to 200 pmol. Data such as these suggest that vagal afferent sensitivity to CCK varies widely, depending on the target of vagal afferent innervation. In addition, they suggest that even neurons innervating the same organ exhibit a range of CCK sensitivities. Despite this range, when one considers the relatively small volume of distribution for in vivo near arterially administered CCK, even low picomole doses are likely to result in local concentrations in the nanomolar range. Such concentrations are likely to be similar to those that produced calcium responses in our cultured nodose neurons.

Paracrine actions of CCK. Because systemic plasma levels of CCK are reported to be in the low picomolar range (5, 30), one could conclude that only responses we observed at subnanomolar concentrations are physiologically relevant. On the other hand, vagal afferent nerve endings terminate in the small intestinal lamina propria, in close proximity to the basal pole of the enterocytes and CCK-secreting I cells (2, 53). The concentrations of CCK that are achieved in this interstitial area are not known, but they are likely to be much higher than either portal venous or systemic venous levels. In support of this idea, Brenner et al. (5) found inhibition of gastric emptying and reduction of food intake after intestinal infusion of some nutrients produce little or no elevations of plasma CCK, yet these phenomena are abolished by CCK-A receptor antagonists, suggesting that plasma CCK may not be an appropriate indicator of CCK acting on vagal afferent neurons. Indeed, several investigators have suggested that some of the effects of CCK may be paracrine in nature (5, 15, 21, 45). On the other hand, two recent studies of the immunohistochemical localization of CCK-A receptor were unable to find labeling in vagal afferent nerve endings (40, 50). Whether the failure to find CCK-A receptors in these terminals is due to the absence of the receptors, or that the antigenic regions of the receptor were somehow masked in these structures, is unknown. The fact remains that reduction of food intake and inhibition of gastric emptying by exogenous CCK are mediated by the low-affinity CCK-A receptor site and involve vagal activation. This observation supports the notion that vagal afferent processes may be adapted to respond to relatively high local concentrations of paracrine CCK in vivo. In this regard, it seems appropriate that afferent neuron cell bodies whose terminals have adapted to respond to high local concentrations of a peptide might not respond to low circulating concentrations of the peptide.

Perspectives

Both reduction of food intake (43, 46) and inhibition of gastric emptying (23, 36, 42) by CCK are mediated by vagal afferent neurons. These actions of CCK are abolished by antagonists for the low-affinity site of the CCK-A receptor (1, 4, 36, 54, 56), as are the calcium responses that we observed in cultured neurons. Likewise, antagonists for the low-affinity site of the CCK-A receptor abolish activation of mesenteric afferent fibers, which are probably vagal, by exogenous CCK-8 (17). Thus the specific site on the receptor mediating vagal afferent responses in vivo is the same as that which mediates the calcium responses in cultured nodose cells.

Although we cannot rule out the possibility that cellular processes involved in the response to CCK differ qualitatively between the afferent terminals and their cell bodies, our results to date are consistent with the supposition that cultured nodose neurons represent a valid model to gain cellular insights into how vagal afferent fibers are activated. As a better understanding of the mechanisms by which CCK activates vagal afferent neurons is developed, it will not only enhance our understanding of how vagal afferent neurons monitor the internal environment of the gut, but it will also provide the insights necessary to determine whether transduction processes in afferent neuron terminals differ from those currently observed in their cell bodies.