Aplysia Bag Cells Function as a Distributed Neurosecretory Network

Nathan G. Hatcher, Jonathan V. Sweedler


The anatomical organization of many neuroendocrine systems implies multiple sites of hormone release in areas mediating multiple aspects of physiology and behavior, yet this neurosecretory complexity has not often been verified. Here we probe the well-characterized hormonal model, the reproductive bag cell neuroendocrine system of the sea slug Aplysia californica. The bag cell neurons of Aplysia mediate egg-laying behavior through the coordinated secretion of a suite of peptides derived from a single gene product, the egg-laying prohormone (proELH). Although the majority of bag cell neurons are located within two major abdominal bag cell clusters, smaller groups of egg-laying hormone-expressing cells have been observed in specific pleural and cerebral ganglia regions, some of which have been reported to be electrically connected to the abdominal bag cell clusters. Releasates are sampled from discrete locations within the Aplysia CNS before and during stimulation of afterdischarge activity and subsequently analyzed with matrix assisted laser desorption/ionization time-of-flight mass spectrometry. Site-specific release profiles are observed at bag cell cluster, pleural, and genital ganglion sites after site-specific electrophysiological activation of bag cell afterdischarges. These data demonstrate that the bag cell network has multiple neurohemal release sites, exhibits descending activation that travels from the cerebral and pleural ganglia down to the abdominal bag cell clusters, and releases spatially distinct profiles of proELH-derived peptides within the Aplysia nervous system. Such distributed neurosecretory organization may be a common feature of neuroendocrine systems across higher order organisms linking multiple behavioral aspects to a single neuronal network.


Neuroendocrine networks consisting of multiple neurosecretory elements distributed throughout the CNS are thought to mediate multiple aspects of related behaviors. Typically, neurosecretory systems, like those comprising the hypothalamic pituitary axis, secrete peptides in a hormonal fashion within defined neurohemal regions. Interestingly, it is also common to find small numbers of neurosecretory neurons, processes, and related peptide receptors within other brain regions, suggesting that these systems participate in alternate networks and influence other behaviors (Elde and Hökfelt 1979). Although it has been suggested that these networks secrete their peptide contents in multiple locations and, in some instances, may release site-specific complements of peptide messengers (Sossin et al. 1990), such data are difficult to observe in living neuroendocrine tissues. Here we assay site-specific neurosecretory complements of the bag cell neurons of Aplysia californica to determine the spatial, temporal, and chemical dynamics of peptide secretion within this model distributed neuroendocrine network.

The bag cells mediate egg-laying behavior in Aplysia through the release of a series of peptides derived from the processing of a single prohormone, proELH (egg-laying prohormone) (Chiu et al. 1979; Kupfermann 1967, 1970; Scheller et al. 1982, 1983). Bag cell peptides are secreted in response to an all-or-none afterdischarge during which electrically coupled bag cells fire a coordinated train of action potentials that ultimately induces peptide secretion into the neurohemal region within the surrounding connective sheath (Dudek et al. 1979; Kupfermann and Kandel 1970; Newcomb and Scheller 1990; Stuart et al. 1980). Although peptide secretion does depend on elevation of free intracellular calcium, release of calcium from intracellular stores appears sufficient to cause persistent egg-laying hormone (ELH) secretion independent of afterdischarge electrical activity (Jonas et al. 1997; Wayne et al. 1998a). Although elevation of intracellular calcium is an important component mediating peptide secretion in bag cells, release persists well after intracellular calcium concentration returns to baseline levels (Michel and Wayne 2002; Wayne et al. 1998a).

The majority of bag cells are present in two discrete abdominal clusters that contain several hundred neurons each. Although bag cells are most numerous in these paired abdominal clusters, neurons expressing bag cell peptides have been identified within the pleural and cerebral ganglia with neuritic processes extending through the pedal ganglia and periphery (Chiu and Strumwasser 1981; Li et al. 1998; Painter et al. 1989). Secreted bag cell peptides act on the ovotestis and presumably other brain regions to mediate egg-laying and related stereotypic behaviors such as cessation of locomotion, inhibition of feeding, and head-waiving (Strumwasser et al. 1980). Investigations of bag cell activity and their interconnections have suggested a distributed organization whereby the cerebral or pleural bag cells activate the abdominal clusters through descending activation (Brown et al. 1989; Haskins and Blankenship 1979; Painter et al. 1988). Additionally, extensive transport of bag cell peptides occurs throughout the Aplysia CNS (Li et al. 1998). ELH-related peptides have been shown to be differentially packaged into discrete peptidergic vesicles with one vesicle class containing the N- and the other the C-terminal proELH regions specifically (Fisher et al. 1988; Garden et al. 1998; Sossin et al. 1990). These vesicles show a slight differential distribution in the processes of the bag cell neurons extending throughout the surrounding abdominal and pleural abdominal connective (PAC) nerve sheath.

Here we investigate the stimulation and secretion parameters of Aplysia bag cells using a combination of electrophysiology, solid phase extraction (SPE) bead sampling and peptide detection using mass spectrometry (MS) (Hatcher et al. 2005). Site-specific releasates are selectively obtained with bead probes placed at discrete locations in the CNS (Jakubowski et al. 2006; Jing et al. 2007). Bag cell activity is both stimulated and monitored through extracellular recordings of the interganglionic connective nerves while releasates are collected with bead probes. Secreted peptides are subsequently detected and sequenced with matrix assisted laser desorption/ionization time-of-flight MS (MALDI TOF MS). Following this methodology, profiles consisting of multiple ELH-related peptides are obtained from the bag cell cluster after stimulation of an afterdischarge. Additionally, neurohemal secretion is observed at multiple extra-abdominal sites including robust secretion near the pleural ganglia. Release of bag cell peptides from distal pleural regions is activated in a descending fashion that must be initiated at pleural or cerebral areas as activation of abdominal bag cells alone does not concurrently activate pleural bag cell neurons. Last, bag cell releasate profiles are observed to differ in the specific proELH peptides, demonstrating that the Aplysia bag cells secrete site-specific profiles of neuropeptides in localized regions of the CNS. These data demonstrate the inherent coordination of peptidergic signaling by a relatively simple neuroendocrine network mediating multiple aspects of reproduction.



The MALDI matrix, α-cyano-4-hydroxycinnamic acid (CHCA), was mixed in 70% acetonitrile (ACN) with 0.01% trifluoro acetic acid. SPE bead elution used 70% ACN/H2O. All reagents were obtained from Sigma (St. Louis, MO).


Specimens of Aplysia californica weighing 150–200 g were obtained from Mr. Charles Hollahan of Vivida Biological Supply (Monterey, CA). Animals were maintained in circulating seawater at 14°C.


Animals were anesthetized by injection of isotonic MgCl2 equivalent to 1/3 of the animal weight. Ganglia were removed with connectives intact and pinned in a silicone elastomer (Sylgard)-lined recording dish containing artificial seawater (ASW, containing, in mM: 460 NaCl, 10 KCl, 10 CaCl2, 22 MgCl2, 26 MgSO4, and 10 HEPES at a pH of 7.8) at room temperature. The recording dish consisted of two asymmetric chambers, containing 0.5 and 5 ml, respectively. The abdominal ganglion was typically placed in the 0.5-ml chamber while the remaining interconnected CNS was positioned in the larger 5-ml chamber in those experiments that used multiple ganglia. The PAC and in one set of experiments the abdominal-genital connective were positioned through an elevated connection that maintained ASW around the nerve but limited perfusion between chambers. To prevent the diffusion of bag cell peptides from abdominal bag cell clusters to the remaining CNS regions, a directional perfusion of 50 μl/min of ASW was employed from the larger CNS chamber to the smaller abdominal-containing chamber.

Extracellular recording and stimulation of connective nerves were performed with suction electrodes pulled from polyethylene tubing connected to a differential AC amplifier (Model 1700; A-M Systems, Sequim, WA) and isolated pulse stimulator (Model 2100; A-M Systems). Bag cell afterdischarges were stimulated (20–25 V, 12-s pulse train, 2.5 × 10−2-s pulse duration, and 0.1-s interpulse interval) through the PACs, specifically at pleural or abdominal regions, unless otherwise stated. Electrophysiological data were acquired with a Digidata 1200 system (Axon Instruments/Molecular Devices, Union City, CA) and recorded and analyzed with the accompanying pClamp 8.0 computer software.

Mass spectrometry

MALDI TOF MS was performed with either a Voyager DE-STR (Applied Biosystems, Framingham, MA) or an Ultraflex II TOF/TOF (Bruker Daltonics, Bremen, Germany) mass spectrometer. Mass spectra were obtained with external calibration in either linear or reflectron mode, depending on the amount of analyte present, with typical mass accuracies within 100 ppm. Releasate analytes were initially identified by mass. In releasate collections from the abdominal cluster region, bag cell neuropeptides were collected in sufficient amounts to perform sequencing via tandem MS (MS/MS), verifying the identity of the observed peptides. MS/MS sequencing was performed with the Ultrafex II TOF/TOF mass spectrometer in “LIFT” mode. MS analyses were performed with the accompanying Flex Analysis and BioTools software package (Bruker Daltonics). MS/MS sequence predictions were searched against an in-house version of the MASCOT database of Aplysia peptides. Peptide sequence verification with MS/MS required the matching of at least three sequential amino acids.

Selective sampling of secreted neuropeptides

Single SPE beads, 30–50 μm in diameter, were isolated from Resin “D” (International Sorbent Technology, Hengoed, UK), “wetted” in 50% ACN and equilibrated in ASW as described previously (Hatcher et al. 2005). For sample collection, individual beads were placed on the ganglia and connective tissue surface for 10-min intervals unless otherwise indicated. Samples were taken both before and during stimulation; fresh beads were placed in the same position for each collection period. After sample collection, beads were removed, washed in Millipore filtered H2O (Milli-Q filtration system; Millipore, Bedford, MA), transferred to MALDI TOF target plates, and eluted with saturated CHCA/70% ACN matrix solution for subsequent off-line MALDI MS analyses as described previously (Hatcher et al. 2005; Jakubowski et al. 2006; Jing et al. 2007). Stimulation-dependent analytes were readily distinguished from nonspecific background compounds by comparing mass spectra obtained from prestimulation and stimulation bead samples.


Monitoring activity-dependent secretion of peptides from the bag cell cluster region

The data presented here were generated from >30 preparations of the Aplysia CNS over the course of 3 yr. In every case, abdominal bag cell afterdischarge activity resulted in secretion of bag cell peptides from the abdominal cluster region. As reported previously, the most robust examples of releasates were acquired during summer seasons, similar to seasonal reproductive behaviors (Wayne 2001) as well as seasonal variations in second-messenger pathways mediating ELH release (Wayne et al. 1998b).

The image in Fig. 1 A shows single beads placed at discrete regions on the surface of the bag cell clusters and nearby regions of an isolated Aplysia abdominal ganglion. Bead samples were obtained before and after electrophysiological stimulations of bag cell afterdischarges with bead probes replaced between sampling intervals and repositioned in the same locations. Releasates were readily distinguished by comparing mass spectra from corresponding prestimulation and stimulation bead samples. Figure 1B shows an afterdischarge recorded extracellularly in the PAC nerve with prestimulation and stimulation sampling intervals delineated by solid lines. The corresponding mass spectra from bead samples before and during afterdischarge are shown in Fig. 1C; assays of bag cell releasates, including the time-course studies below, were repeated in a total of 26 preparations of abdominal ganglia isolated from the Aplysia CNS.

FIG. 1.

General approach for the mass spectrometry (MS) survey of Aplysia bag cell releasates secreted from the abdominal neurohemal region. A: bead probes placed on the surface of the abdominal ganglion bind compounds present in the extracellular environment for subsequent mass spectrometric analyses (scale bar = 1.0 mm). B: brief electrical stimulation of the pleural abdominal connective (PAC) nerve triggers a stereotypic afterdischarge volley of action potentials recorded from bag cell terminals within the PAC abdominal neuritic cuff area. Separate beads are positioned on the ganglion in intervals before and during afterdischarge events. C: direct comparison of mass spectra obtained from prestimulation (top) and stimulation (bottom) beads clearly distinguish the presence of stimulation-dependent analytes with masses corresponding predominantly to egg-laying prohormone (proELH)-derived peptides (listed in Table 1). Peaks labeled (*) are present in prestimulation beads. The image in A is an illustration and is not the same ganglion used to obtain these data.

The prestimulation mass spectrum (top) is nearly identical to control mass spectra from beads not exposed to tissues (not shown). Specifically, this spectrum shows three main peaks as corresponding to known masses from the CHCA MALDI matrix; these same peaks are observed from control beads not exposed to tissue. Other than these low mass CHCA artifacts, no signal is observed beyond baseline noise in the prestimulation mass spectrum. In marked contrast, bead samples acquired during PAC-mediated stimulation of afterdischarge bursting collected multiple peptides derived from proELH at the site of the abdominal bag cell clusters. As the height of each peak compared between these peptides varies over many orders of magnitude, only secreted bag cell peptides exhibiting prominent mass peaks are labeled directly in the figures. The entire profile of observed bag cell peptide releasates is listed in Table 1. Releasates obtained after proximal PAC stimulation repeatedly include the known fully processed bag cell peptides as well as some R3-14 peptides. The inset in Fig. 1C shows an expanded view of the releasate mass spectrum, demonstrating the monoisotopic distribution of secreted ELH (prestimulation–top; stimulation–bottom). The releasates listed in Table 1 were observed as such monoisotopic distributions. Peptides are listed by their name, observed and theoretical mass-to-charge ratios, and corresponding amino acid positions in the proELH.

View this table:

Neuropeptides observed in releasates after bag cell afterdischarge

We performed MS/MS to validate our assignments of bag cell peptides detected in releasates. Figure 2 A shows a representative MS/MS spectrum obtained directly from bead-collected acidic peptide (AP). Figure 2B illustrates the observed b and y ion series matching those predicted for the amino acid sequence of AP, thus verifying its identity. The assignments of other secreted peptides that were present in sufficient amounts were similarly verified (Table 1). Of course, the detection of multiple peptides resulting from the proteolytic processing of an individual prohormone adds confidence in our assignments similar to the way the detection of multiple peptides from tryptic processing of a protein improves the confidence of the protein identification (Zhang and Chait 2000).

FIG. 2.

Sequence verification of secreted bag cell peptides. A: tandem MS (MS/MS) spectrum of a releasate analyte matching the mass-to-charge ratio of acidic peptide (AP, 2959.5 m/z). B: b- and y-series fragment ion assignments confirm the identity of this peptide as AP.

We monitored bag cell release collected at 5-min intervals over the course of 1 h after afterdischarge in isolated abdominal ganglia (n = 3) to observe the secretion of the entire complement of bag cell releasates over this period. Wayne and coworkers (Wayne and Wong 1994) originally reported persistent release of ELH long after cessation of afterdischarge activity. Briefly, beads placed on the abdominal bag cell surface and proximal PAC areas, as shown in Fig. 1A, were exposed to the tissue preparation for 5-min intervals. At the end of each collection interval, beads were removed and the tissue chamber (0.5 ml in volume) was rinsed by means of ASW perfusion at a rate of 2 ml/min for 10 min to remove residual peptides present in the bath. New beads were repositioned in the same locations for consecutive collection and wash cycles. Figure 3 A shows the releasate mass spectra obtained for each 5-min collection interval for one experiment; as in Fig. 1C, only the prominent mass peaks are labeled. As an internal reference to normalize releasate peak intensities between separate releasate collection intervals, 1.0 pmol of arginine-vasopressin (AVP) was added to each MALDI target sample. After stimulation of an afterdischarge within the abdominal bag cell clusters, bag cell peptides were clearly present in bead collections throughout the first 5 min of activity. These same peptides are present with decreasing intensities in releasates collected at subsequent collection intervals. Plots of normalized peak intensities for the N-terminal β bag cell peptide and α bag cell peptide 1–9 (Fig. 3B) as well as C-terminal AP and ELH (Fig. 3C) illustrate persistent secretion of multiple bag cell peptides after stimulation of afterdischarge events.

FIG. 3.

Time course of bag cell peptide secretion from the abdominal neurohemal release site. A: mass spectra obtained from beads exposed to an abdominal bag cell cluster surface for 5-min exposures over the course of 1 h after afterdischarge-mediated peptide release (*, 1.0 pmol AVP synthetic peptide spiked into each sample spot for peak height normalization). B: normalized peak heights of α bag cell peptide 1–9 and β bag cell peptide; C: AP and ELH secreted over time (n = 3, error bars indicate SD).

Determining broad spatial differences in releasates between uncoupled abdominal bag cell clusters

Stimulation of afterdischarge activity in the intact abdominal ganglion preparations described in the preceding text typically results in secretion from both abdominal bag cell clusters as the abdominal bag cell clusters are loosely electrically coupled by neuritic processes passing through the adjoining connective tissue region. Bisection of this connective through the lower ganglionic region uncouples bag cell clusters without preventing the afterdischarge capability in an individual cluster (Kupfermann and Kandel 1970).

We repeated this experiment, while simultaneously monitoring activity in the PAC nerves and collecting releasates with beads placed on both abdominal bag cell cluster regions, to test if we could detect differences in releasates between two relatively closely positioned abdominal bag cell clusters in the same experiment. As shown in Fig. 4, we observed that electrical stimulation of the right PAC nerve triggered unilateral bursting in the right bag cell cluster; however, no activity resembling an afterdischarge was seen in the left bag cell cluster (not shown). Differential secretion was observed between these two abdominal structures with a complete complement of bag cell peptides observed from beads located on the right cluster, ipsilateral to stimulation, while no secreted peptides were observed from beads positioned on the left, unstimulated cluster.

FIG. 4.

Uncoupled secretion between bag cell clusters. Bisection of the connective tissue between bag cell clusters uncouples bag cell activation and subsequent secretion. Stimulation of the right PAC nerve triggers secretion in the right (top) but not the left (bottom) bag cell neurohemal release sites.

Assaying releasates in CNS sites known to exhibit differential transport of bag cell peptides

Bag cell peptides are differentially packaged into two distinct types of dense core vesicle; one containing peptides derived from the peptides N-terminal to the furin cleavage site in the ELH prohormone and a second vesicle class containing the peptides C-terminal to the furin site (Fisher et al. 1988; Sossin et al. 1990). Using immunogold staining with antibodies to appropriate sequences in the prohormone, combined with electron microscopy of abdominal and PAC-derived bag cell neurites, the vesicles demonstrated a statistically significant differential distribution of N- and C-terminally derived bag cell neuropeptides in the innervation throughout the connective sheath surrounding the bag cell cluster and nearby abdominal regions (Sossin et al. 1990). To determine whether the observed differential peptide distribution leads to a similar difference in the composition of peptide releasates, we positioned the collection beads throughout the medial and distal abdominal bag cell cluster/PAC regions reported to contain differential distributions of these vesicles before and during stimulation of afterdischarges at the abdominal bag cell clusters. In >30 collections of secreted bag cell peptides using either isolated abdominal or entire CNS preparations obtained over the course of this study, no difference in the release profiles within these abdominal bag cell cluster regions was resolved. Bag cell releasates collected with beads along the distal to medial regions of the abdominal cluster always contained the full complement of both N- and C-terminal bag cell peptides after afterdischarge. This lack of differences in releasate composition specifically across the surface of the abdominal bag cell clusters may reflect that there is no differential release at these areas or the subtle differences reported in vesicle populations may simply be too difficult to detect via the semi-quantitative MS approach used here.

To accentuate potential differences in peptide releasate content, locations further removed from the bag cell clusters were examined next. Li and coworkers (1998) reported relatively greater differences between amounts of N- and C-terminal bag cell peptides further up the PAC near the pleural ganglia as well as in distal portions of the genital nerve. In their study, C-terminal peptides, ELH and AP were detected in the upper PAC nerves near the pleural ganglia with little to no N-terminal peptides detected; the opposite was observed in distal portions of the genital nerve where only N-terminal peptides were detected.

To test if the composition of releasates within these regions reflected this differential transport, beads were placed at the site of the pleural PAC and genital ganglion regions, respectively. After en passant stimulation of an afterdischarge at the PAC nerve region immediately adjacent to the abdominal bag cell cluster, bag cell releasates were observed at the bag cell clusters, weakly at the genital ganglion, and contrary to expectation, not at the PAC near the pleural ganglion. These data suggest that activation of the abdominal bag cell cluster does not ascend the PAC nerve to stimulate pleural bag cell neurons (see below).

Releasates from the abdominal bag cell cluster and genital ganglion are compared in Fig. 5 A; the full complement of bag cell peptides were secreted at the abdominal cluster, but only N-terminal α and β bag cell peptides were detected in releasates collected from the genital ganglion. Releasates obtained from the genital ganglion were typically less abundant than those observed at the abdominal bag cell clusters. Figure 5B shows expanded views of spectra for two N-terminal and two C-terminal peptides (their locations in the ELH prohormone are also shown in Fig. 5). Specifically, the release from abdominal (top) versus genital (bottom) for β bag cell peptide, α bag cell peptide 1–9 and AP and ELH are shown. The intensities of releasate mass spectra from the genital ganglia are magnified by a factor of 10 to better visualize the less intense mass spectral peaks. Both abdominal and genital regions secrete β bag cell peptide and α bag cell peptide 1–9. Neither AP nor ELH is observed in genital releasates. The absence of AP implies that significant C-terminal peptides are not secreted in this region as AP typically produces the most prominent peak in releasate mass spectra obtained from the other regions. Therefore the absence of AP in genital ganglion releasates likely reflects actual differences in N- and C-terminal peptide release. In support of this interpretation, Li and colleagues (Li et al. 1998) reported a similar differential distribution—a lack of C-terminally-derived proELH peptides—in distal portions of the Aplysia genital nerve. Here the secretion of N-terminal peptides at the genital ganglion was observed in three of five preparations; no releasates were detected at the genital ganglion in two preparations.

FIG. 5.

Differential release of N-terminal bag cell peptides at the site of the genital ganglion. A: releasate collections obtained from the site of the genital ganglion show only N-terminally derived bag cell peptides after evoked afterdischarge events. B: expanded views of releasates obtained at the abdominal bag cell cluster (abd) and genital sites (gen), respectively. Prohormone processing diagram is modified from (Garden et al. 1998). The rectangle is a schematic representation of the pro-ELH with narrow vertical lines representing basic processing sites with several of the biologically active peptides labeled. These allow the regions of the prohormone that encode for the specific detected peptides to be visualized.

Determining distributed neurohemal release of bag cell peptides throughout the Aplysia CNS

Prior immunohistochemical staining of the Aplysia CNS enumerated small clusters of bag cell peptide-expressing neurons in the pleural and cerebral ganglia (Chiu and Strumwasser 1981; Painter et al. 1989). In one study, it was reported that these “ectopic” bag cell neurons in the pleural region of the right PAC nerve were robust activators of afterdischarges in the abdominal clusters (Brown et al. 1989). We isolated these same neurons, pictured in Fig. 6 A, for MALDI TOF MS analyses to verify their identity as bag cell neurons containing ELH-derived peptides. The mass spectrum in Fig. 6B shows that these pleural bag cell neurons are biochemically identical to bag cell neurons within the abdominal clusters in so far as they express the same proELH-derived peptides. No other peptides were observed in these cells.

FIG. 6.

Single-cell MS analyses of pleural bag cell neurons. A: image showing the presence of 3 pleural bag cell neurons in the pleural region of the right PAC nerve (scale bar = 100 μm). B: MS of a single isolated cell from the image in A reveals the same peptides as in the abdominally located bag cell and biochemically verifies these cells as bag cell neurons.

As the pleural bag cell neurons were established as triggers of afterdischarge events (Brown et al. 1989), we stimulated these cells with an extracellular suction electrode positioned en passant over this region (n = 4) as well as in separate experiments stimulating the cerebral pleural connective (n = 2). In both types of stimulation, electrical stimulation of these nerves initiated afterdischarges in the abdominal bag cell clusters. Figure 7, A–G, overlays positions of electrophysiological stimulation, nerve recording, releasate collection, and the resulting mass spectra over an illustration of the Aplysia CNS. Beads were positioned at locations of bag cell neuron innervation; these regions included all circumesophageal ganglia and connective nerves. Prestimulation bead samples were always devoid of bag cell peptides. Stimulation of the right pleural PAC region activated pleural bag cell neurons as evidenced by the detection of bag cell peptide release at this region, Fig. 7A. Activation descended along the stimulated PAC nerve (depicted with arrows) to both abdominal bag cell clusters causing afterdischarge bursting, Fig. 7E, as well as bag cell peptide secretion from both abdominal clusters, Fig. 7, C and D. Release was not observed at the pleural region of the PAC nerve contralateral to stimulation, Fig. 7F, suggesting that activation did not pass beyond the abdominal ganglion. Bag cell peptides were not observed to be secreted at other regions, including beads placed on the center region of the right PAC nerve, Fig. 7B; these beads served as spatial release controls acting as potential indicators of peptide diffusion between ganglia. Additionally, beads placed on the cerebral ganglia showed no evidence of bag cell release, although beads placed on the anterior tentacular and upper labial nerves resulted in the detection of Aplysia insulin peptides, Fig. 7G, after pleural PAC stimulation.

FIG. 7.

Distributed bag cell release sites within the Aplysia CNS. Electrophysiological stimulation of the right pleural PAC region (solid arrow) triggers afterdischarge in the abdominal bag cell clusters (illustrated here at E with an en passant extracellular recording of the PAC nerve adjacent to the contralateral bag cell cluster). Bead collection sites at the right PAC pleural neuritic cuff region (A), medial right PAC nerve region (B), right abdominal bag cell cluster (C), left abdominal bag cell cluster (D), and left PAC pleural region (F) show bag cell releasates at discrete locations after stimulation at the site of the right PAC pleural region; major peptides are labeled. These electrophysiological and releasate data suggest that bag cell network activation (illustrated with dotted arrows) travels down the stimulated PAC nerve to the ipsilateral abdominal cluster, stops with the activation of the contralateral abdominal bag cell cluster, and does not ascend the contralateral PAC nerve. G: beads placed at the anterior tentacular neurohemal region of the cerebral ganglion show the presence of Aplysia insulin peptides in releasates indicating the activation of F-cluster neurons innervating this location. BuG, buccal ganglia; CeG, cerebral ganglion; PlG, pleural ganglion; PeG, pedal ganglion; AbG, abdominal ganglion; bcc, bag cell cluster.


We have studied the activity of a model neurohormonal system by combining electrophysiology, spatially resolved neuropeptide sampling and MS. Specifically, we have used the well-known Aplysia bag cell neuronal system to explore peptide secretion dynamics in stimulation-specific contexts. The novel approach of peptide collection using individual bead probes has a number of advantages for this and other peptide studies: they can be placed separately at discrete locations, they concentrate compounds present in the immediate extracellular region, and they provide an optimal method for desalting and preparing sampled peptides essential for quality MALDI MS. As such, the data presented here replicate and expand previous studies of bag cell peptide release from semi-intact preparations. Most notably, we are able to observe the release of multiple peptides in the same experiment with spatial and temporal information on secretion. We have verified the extended release of multiple bag cell peptides over time, data that follow previous measures of persistent ELH secretion outlasting bag cell afterdischarge activity (Wayne and Wong 1994; Wayne et al. 1998a). Most importantly, new insights have been gained into the functional organization and chemical output of a distributed neuroendocrine system.

By obtaining spatially distinct releasate profiles within the nervous system, it is possible to follow the activity of multiple elements. For example, after stimulating the PAC, other peptidergic neurons, such as the abdominal R3-14 neurons and the Aplysia insulin-secreting cerebral F-cluster cells, often exhibit peptide release as well. It is unlikely that these cells are linked to the bag cells as no prior electrophysiological evidence exists suggesting connections among R3-14, F-cluster, and bag cell neurons. The excitation and subsequent release from these other peptidergic neurons is likely a result of nonspecific depolarization via the broad stimulation of entire connective nerves. Support for this assumption is evident as neuritic processes for R3-14 neurons innervate a similar region as the abdominally located bag cell neurons, yet these same R3-14 peptides are not observed in releasates when an afterdischarge is triggered by distal stimulation of the PAC nerve well beyond the site of R3-14 innervation.

In their ground-breaking work, Sossin and coworkers (1990) stressed that differential transport of bag cell vesicles challenged the principle derived from Dale's postulate of metabolic unity whereby a single neuron secretes the same neurotransmitter(s) at each terminal release site (Eccles 1986). Examples of such exceptions to Dale's principle are now numerous, particularly in terms of peptidergic, neurohormonal cells. More recent interpretation has modified this postulate to exclude such peptidergic neurons by limiting application of this principle to neurons expressing the fast neurotransmitters (e.g., glutamate), although reports have emerged suggesting colocalization and co-release of these fast transmitters as well (Sulzer and Rayport 2000). In any case, a consensus exists in neuroscience that typically excludes peptidergic neurons from Dale's principle on the basis of the diversity of colocalized peptide releasates within single neurons. Spatially distinct releasates have not been demonstrated directly but rather have been inferred through differential packaging and transport of N- and C-terminally derived bag cell peptides. Following these transport studies for bag cell peptides in Aplysia, we have been able to verify that differential bag cell peptide release occurs at the site of the genital ganglion.

It is not known if comparative neuroendocrine systems secrete spatially distinct peptide complements. For example, oxytocin, vasopressin, and associated neurophysin cells extend from the paraventricular nucleus to both the nucleus of the solitary tract and the neurohypophysis where these peptides are secreted (White et al. 1986b); little is known of the composition of releasates within these two areas beyond that which is inferred from histochemical studies. Although we have confirmed here that differential release of bag cell peptides occurs, likely through differential transport, alternative mechanisms of differential release are possible. As an example, intracellular processing of endogenous peptides can also alter releasate composition between regions. Region-specific processing of proenkephalin is known to occur and has been suggested to produce region-specific release of distinct processed forms of enkephalin peptides (White et al. 1986a). In this case, specific localization of processing enzymes could determine spatially distinct releasates perhaps targeted to structures or networks mediating distinct but related physiological processes.

Site-directed release of different bag cell peptides derived from a common precursor suggests a remarkable coordination of packaging, localization, and targeting at the level of the individual neuron. Our previous work determined that releasates collected from separate neuritic processes of a single bag cell in culture contained the same complement of bag cell peptides (Hatcher et al. 2005). Thus differential transport and resultant spatially distinct release profiles of bag cell peptides observed in these studies may require contextual cues present in the CNS that direct this process. As of yet, there is little understanding of the mechanism determining such differential transport in any neuronal system, much less the functional significance of such spatially tuned peptide messengers in the Aplysia CNS. It is certainly intriguing to speculate if specific stimulation protocols can regulate release profiles; this is an area of ongoing research.

Although the differential transport of C-terminal bag cell peptides has also been reported to occur up the PAC nerve toward the pleural ganglia (Li et al. 1998), the analyses of releasates obtained from these areas do not reflect this differential transport. We observe the complete bag cell peptide complement at pleural regions under site-specific stimulation of these pleural regions (as depicted in Fig. 7). Potential differential release in these regions may be masked by the separate innervation of the pleural bag cell neurons. From previous immunohistochemical studies and our current single-cell MS analyses, it is clear that small clusters of 1–10 bag cells are present in the posterior region of the pleural ganglia. Processes from these cells form a neuritic cuff in this same region consistent in structure to a neurohemal release site. The releasates observed here demonstrate that these pleural bag cell neurons, in fact, do secrete bag cell peptides at the pleural site into the surrounding hemolymph concurrently with release at the abdominal bag cell cluster region.

Interestingly, initiation of release at these distal locations requires electrophysiological stimulation at the pleural PAC nerve region or the higher pleural cerebral connectives. This distal stimulation triggers concurrent afterdischarge and secretion at the abdominal bag cell clusters as well. In contrast, en passant stimulation of afterdischarge at the site of the abdominal PAC nerve adjacent to the bag cell cluster triggers secretion at the bag cell clusters only, with no bag cell peptide secretion observed at the pleural or cerebral sites. These data reflect and refine the circuit organization established in previous electrophysiological studies. Brown and colleagues (1989) demonstrated that intracellular stimulation of a single pleural bag cell neuron robustly activated afterdischarge in the abdominal bag cell neurons. Moreover, these cells were found to be differentially coupled, favoring descending activation. Additionally, electrophysiological studies of bag cell cluster neurons have revealed coordinated bursting, linking the activity of the two abdominal clusters with comparatively little electrical output ascending the PAC nerves (Haskins and Blankenship 1979).

Here we demonstrate this descending organization through site-specific stimulation and releasate analyses. Unilateral stimulation of the pleural region of one PAC nerve triggers bag cell release in this same region followed by afterdischarge bursting and release in both abdominal bag cell clusters. In the same experiment, bag cell releasates are not observed at the contralateral pleural PAC nerve region or at the cerebral ganglion, suggesting that bag cell activation does not ascend the connective nerves to these regions. These observations are in agreement with the concept that the bag cells distributed throughout the CNS follow a descending order of activation whereby activation of cerebral or pleural initiation sites triggers activation throughout the bag cell network. Stimulation of the network at lower abdominal regions is not sufficient to activate “higher” pleural and possibly cerebral bag cell neurons. By combining the electrical connectivity with releasate data, a more complete understanding of the bag cells emerges in which a distributed network, consisting of cells of a single phenotype, exhibits multiple and sometimes spatially distinct release sites. These distinct intercellular signals may ultimately contribute to the coordination of multiple aspects of reproductive behavior in Aplysia.

From a comparative perspective, it is interesting to speculate as to the commonality of this organization in neuroendocrine systems of other organisms. A fitting mammalian comparison is found in the gonadotrophin-releasing hormone (GnRH) cells of the hypothalamus. Like the bag cells, GnRH neurons participate in neurohormonal regulation of reproduction. GnRH cells are diffusely distributed throughout the hypothalamus, mostly within the preoptic area, with terminals that converge in the median eminence (Elde and Hökfelt 1979). Moreover, analyses across multiple vertebrate animals have characterized separate clusters of GnRH neurons distributed within other brain areas such as the midbrain and regions near the olfactory bulb in fish (Wayne et al. 2005) and some mammals (Schwanzel-Fukuda and Silverman 1980). Like the neuroendocrine bag cells, GnRH neurons release their peptide content in concert into the hypophyseal portal blood system linking the hypothalamus and anterior pituitary organ. Although it has long been suggested that distributed release of hypothalamic peptides contributes to multiple aspects of neuronal function (Elde and Hökfelt 1979), little evidence has since emerged in which distributed release has actually been assayed directly. Measuring peptide release with spatial information in intact neuroendocrine networks has been challenging, although the approaches used here provide a feasible methodology for these investigations. Thus we expect to be able to confirm if peptidergic systems such as distributed GnRH cells exhibit a similar functional organization to the Aplysia bag cell network. Such assays of release dynamics will be of conceptual importance in understanding the varied roles a neuroendocrine network may play in coordinating complex aspects of physiology and behavior.


This material is based upon work supported by the National Institutes of Health Grant R01 NS-031609 and Grant P30 DA-018310 to the University of Illinois at Urbana-Champaign Neuroproteomics Center.


  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.


View Abstract