To fully understand neuronal network operation, the influence of all inputs onto that network must be characterized. As in most systems, many neuronal and hormonal pathways influence the multifunctional motor circuits of the crustacean stomatogastric ganglion (STG), but the actions of only some of them are known. Therefore, we characterized the influence of the kinin peptide family on the gastric mill (chewing) and pyloric (filtering of chewed food)motor circuits in the STG of the crab Cancer borealis. The kinins are myoactive in arthropods and they occur within the arthropod central nervous system (CNS), but their CNS actions are not well characterized in any species. The pevkinins were first identified in the shrimp Penaeus vannamei,but they have yet to be studied in the STG of any species. We identified kinin-like immunolabeling (KLI) in the pericardial organs (POs) in C. borealis, but there was no KLI within the STG. The POs are a major source of hormonal influence on the STG. Pevkinin peptides activated the pyloric circuit and they caused a modest increase in the speed of ongoing pyloric rhythms. This modest influence on cycle speed resulted in part from pevkinin excitation of the lateral pyloric neuron, whose strengthened inhibitory synapse onto the pyloric pacemaker neurons limited the pevkinin-mediated increase in cycle speed. The pevkinin excitation of the pyloric rhythm was not strong enough to interfere with the previously documented, gastric mill rhythm-mediated weakening of the pyloric rhythm. Pevkinin also had little influence on the gastric mill rhythm. These results indicate that the kinin peptides have distinct and selective modulatory actions on the pyloric rhythm.
Kinin peptides were first isolated from the central nervous system (CNS) of insects, and more recently from mollusks and crustaceans(Torfs et al., 1999). Thus far, the biological actions of these peptides are best documented in peripheral structures, where they are associated with myotropic and diuretic activity (Torfs et al., 1999; Gäde, 2004). The first identified members of this family, the leucokinins (LKs), were isolated from the cockroach Leucophaea maderae(Holman et al., 1986). Since then, more than 30 kinin family members have been isolated from many insect species (Holman et al., 1990; Schoofs et al., 1992; Hayes et al., 1994; Veenstra, 1994; Blackburn et al., 1995; Predel et al., 1997; Nieto et al., 1998; Terhzaz et al., 1999). The insect kinins all share the C-terminal amino acid sequence FX1X2WGamide, where X1 is F, H, N, S or Y,and X2 is A, P or S. The non-insect members share this sequence structure but for the terminal amino acid being changed from glycine to serine in lymnokinin (snail) and alanine in some pevkinin (shrimp) family members(Cox et al., 1997; Nieto et al., 1998).
Kinin peptides are likely to act both within the arthropod CNS and as circulating hormones. Supporting the former possibility is the finding that kinin-like immunolabeling (KLI) occurs in the CNS of many arthopods, staining neuronal somata in the supraesophageal (brain), subesophageal, thoracic and abdominal ganglia (Torfs et al.,1999). In the shrimp Penaeus vannamei, pevkinin(PevK)-like immunolabeling includes a subset of neurosecretory cells in the brain and thoracic ganglion (Nieto et al.,1998). Supporting a hormonal action for kinin peptides is the finding that, in Leucophaea, leucokinin peptides influence the Malpighian tubes and the hindgut but neither structure is innervated by KLI axons (Nässel et al.,1992). Furthermore, kinins are released into the circulation during feeding in the blood-feeding bug, Rhodnius prolixus(Te Brugge and Orchard,2002).
In this report, we identify KLI in the pericardial organs (POs) of the crab Cancer borealis and characterize the pevkinin actions in the stomatogastric nervous system (STNS). The crab STNS is composed of four interconnected ganglia plus their connecting and peripheral nerves(Harris-Warrick et al., 1992; Nusbaum and Beenhakker, 2002). The four ganglia include the paired commissural ganglia (CoGs; containing∼550 neurons each) plus the unpaired stomatogastric (STG; 26 neurons) and esophageal (OG; 14 neurons) ganglia(Kilman and Marder, 1996). Most of the STG neurons are components of the gastric mill (chewing) and/or pyloric (filtering of chewed food) motor circuits. These circuits generate multiple versions of the gastric mill and pyloric rhythms due to modulatory inputs they receive from neuronal and hormonal sources(Nusbaum et al., 2001; Marder and Thirumalai, 2002; Marder et al., 2005).
Superfusion of pevkinin to the isolated STG initiated or enhanced the pyloric rhythm. These pevkinin actions included a strong excitation of the lateral pyloric (LP) neuron, whose strengthened inhibitory feedback synapse onto the pyloric pacemaker neurons limited the pevkinin-mediated increase in the pyloric cycle frequency. In contrast, this peptide neither elicited the gastric mill rhythm nor altered its ongoing activity. Because pevkinin selectively excited the pyloric rhythm, we determined whether this action altered the previously documented weakening of the pyloric rhythm during the protractor phase of the gastric mill rhythm. However, there was no evident change in this inter-circuit regulation during pevkinin application.
Some of this work has appeared in abstract form(Hertzberg et al., 2002).
Materials and methods
Male Jonah crabs (Cancer borealis Stimpson) were obtained from a commercial supplier (Commercial Lobster and Seafood, Boston, MA, USA) and from the Marine Biological Laboratory (Woods Hole, MA, USA). Before experimentation, the crabs were housed in commercial tanks containing recirculating, filtered and aerated artificial seawater (10-12°C). Before dissection, the crabs were cold-anesthetized by packing them in ice for at least 30 min. The foregut was then removed and maintained in chilled physiological saline while the STNS was dissected away from the foregut. For tissue collection, the foregut, eyestalks and dorsolateral portions of the pericardial chamber were removed. The STNS, sinus glands (SGs) and POs were subsequently dissected from these structures, respectively, in chilled(approximately 4°C) physiological saline [composition in mmol l-1 for physiology experiments: 440 NaCl, 26 MgCl2, 13 CaCl2, 11 KCl, 10 Trizma base, and 5 maleic acid, pH 7.4-7.6;composition in mmol l-1 for anatomy experiments: 440 NaCl, 26 MgCl2, 13 CaCl2, 11 KCl, 10 Hepes acid, pH 7.4 (adjusted with NaOH)]. After dissection, the tissue was pinned in a Sylgard 184 (KR Anderson, Santa Clara, CA, USA)-lined Petri dish.
All immunoprocessing was done as whole mounts. Tissue was pinned in a Sylgard-lined Petri dish and fixed in a solution of 4% paraformaldehyde in 0.1 mol l-1 sodium phosphate buffer (pH 7.4) for 12-24 h at 4°C. After fixation, tissue was rinsed five times over approximately 5 h, in a solution of sodium phosphate buffer containing 0.3% Triton X-100 (P-Triton). The tissue was then incubated for approximately 72 h in a 1:300 dilution[P-Triton with 10% normal donkey serum (NDS; Jackson ImmunoResearch, West Grove, PA, USA)] of rabbit polyclonal leucokinin I antiserum(Nässel et al., 1992). This antiserum binds with the carboxyl-terminal region of leucokinin I(Nässel et al., 1992). This region of the leucokinin I peptide is similar to the comparable region of the previously identified crustacean kinins, pevkinin (PevK) 1-6(Nieto et al., 1998; Torfs et al., 1999). The tissue incubation in primary antiserum was followed by another five rinses over approximately 5 h in P-Triton. After this set of rinses, the tissue was incubated for 12-24 h in a 1:300 dilution (as per the primary antiserum) of donkey anti-rabbit IgG conjugated to Alexa Fluor 488 (Molecular Probes,Eugene, OR, USA) or Alexa Fluor 594 (Molecular Probes). After secondary antibody incubation, preparations were rinsed five times over approximately 5 h in sodium phosphate buffer and then mounted between a glass microscope slide and coverslip using Vectashield mounting medium (Vector Laboratories, Inc.,Burlingame, CA, USA). Incubations in primary and secondary antisera were done using gentle agitation at 4°C. All rinses were done at room temperature(approximately 20°C) without agitation. Incubation in secondary antibody was conducted in the dark, as was all subsequent processing. Likewise, slides were stored in the dark at 4°C until examination. Specificity controls for the leucokinin I antiserum in Cancer tissues are described in Fu et al. (Fu et al., 2005).
Immunolabeled preparations were viewed and data collected using a Bio-Rad Radiance 2000 laser scanning confocal microscope. Descriptions of the hardware and software used for imaging on this system were documented previously(Messinger et al., 2005).
For the production of figures, Bio-Rad.pic files collected using the Radiance 2000 system were converted to .tif files using ImageJ software(freely available at http://rsb.info.nih.gov/ij/). Individual micrographs and composite figures were produced from the tif files using a combination of ImageJ and Photoshop (version 7.0; Adobe Systems Inc.,San Jose, CA, USA). The contrast and brightness of final figures were adjusted as needed to optimize the clarity of printed images. Schematic diagrams were produced using Canvas (version 8.0; Deneba Systems Inc., Miami, FL, USA).
The pevkinins were first isolated and identified in the shrimp Penaeus vannamei (Nieto et al.,1998; Torfs et al.,1999). A subset of the pevkinins (PevK-1, PevK-3, PevK-4) have the same carboxyl-terminal amino acid sequence as the insect kinins, including the leucokinins (Torfs et al.,1999). In a limited, previous study of the actions of the leucokinins on the C. borealis STNS, these peptides had a modest excitatory action on the pyloric rhythm(Blitz et al., 1995). To determine whether this modest action resulted from the use of a kinin peptide with the incorrect carboxyl-terminal sequence, we determined whether superfusion of a crustacean kinin peptide with a distinct carboxyl-terminal sequence (PevK-2: Asp-Phe-Ser-Ala-Trp-Ala-NH2) elicited a stronger response from the C. borealis pyloric circuit. In a smaller number of experiments, we also assessed the actions of a second such crustacean kinin(PevK-6: Ala-Phe-Ser-Pro-Trp-Ala-NH2). PevK-2 was synthesized by the Protein Chemistry Laboratory at the University of Pennsylvania School of Medicine, USA and at the Katholieke Universiteit Leuven, Belgium. PevK-6 was synthesized at the Katholieke Universiteit Leuven, Belgium. Each peptide was stored as a frozen stock solution (10-3 mol l-1) that was diluted to its working concentration with C. borealis saline immediately before its use. Both peptides were applied to the STNS by superfusion (7-10 ml min-1) to the desheathed STG.
Electrophysiological experiments were performed using standard techniques for this system (Beenhakker and Nusbaum,2004). In brief, the isolated STNS was pinned down in a Sylgard 184-lined Petri dish (Fig. 1). During experiments in which the dorsal posterior esophageal nerve(dpon) was stimulated to elicit the gastric mill rhythm, the CoGs remained connected with the STG. In all other experiments, the CoGs were removed to eliminate descending projection neuron influence on the STG circuits. All preparations were superfused continuously with C. borealis physiological saline (10-12°C).
Extracellular recordings were made by pressing stainless steel wire electrodes into the Sylgard alongside the nerves and isolating each area with Vaseline. These electrodes were led to a differential AC amplifier (AM Systems, Model 1700, Carlsborg, WA, USA). To facilitate intracellular recordings, desheathed ganglia were viewed with light transmitted through a dark-field condenser (Nikon, Tokyo, Japan). Intracellular recordings of STG somata were made using microelectrodes (15-30 MΩ) filled with 4 mol l-1 potassium acetate plus 20 mmol l-1 KCl or 0.6 mol l-1 K2SO4 with 20 mmol l-1 KCl. Intracellular recordings and current injections were performed using Axoclamp 2 amplifiers (Axon Instruments, Foster City, CA, USA) in single-electrode discontinuous current-clamp (DCC) mode. Sample rates during DCC were 2-3 kHz. In some preparations, the projection neuron MCN1 was stimulated extracellularly via the inferior esophageal nerve (ion;10-30 Hz) (Bartos and Nusbaum,1997). The ion was stimulated using a Grass S88 stimulator and Grass SIU5 stimulus isolation unit (Astro-Med/Grass Instruments, West Warwick, RI, USA).
Individual STG neurons were identified by their axonal pathways, activity patterns and interactions with other neurons (e.g. Fig. 1)(Beenhakker and Nusbaum, 2004; Blitz et al., 2004). Data were collected directly onto a MT-95000 or Everest chart recorder (Astromed), and were simultaneously digitized (approximately 5 KHz) and collected onto a PC computer using data acquisition/analysis tools (Spike2, Cambridge Electronic Design, Cambridge, UK).
Data analysis was facilitated with a custom-written program for Spike2 that determines the activity levels and phase relationships of neurons (freely available at http://www.neurobiologie.de/). Unless otherwise stated, each datum in a data set was derived by determining the average of 10 consecutive impulse bursts. Briefly, burst duration was defined as the duration between the onset of the first and last action potential in an impulse burst. The firing rate of a neuron was defined as the number of action potentials minus 1, divided by the burst duration. The cycle frequency of the pyloric rhythm was determined by calculating the inverse of the period between the onset of two successive pyloric dilator (PD) neuron bursts. Phase analysis was performed on data collected from extracellular recordings. Phase is defined as the latency to occurrence of an event relative to the start of a cycle, divided by the cycle period. Thus, the latency to onset and offset of activity in each pyloric neuron relative to the onset of a PD neuron burst were measured as a fraction of the total pyloric cycle duration. The gastric mill cycle period corresponded to the duration (sec)between the onset of two successive lateral gastric (LG) neuron bursts. The phase relationships among gastric mill neurons were determined relative to a gastric mill cycle as defined by the onset of one LG neuron burst (beginning of the cycle) and the onset of the subsequent LG neuron burst (end of the cycle).
Figures were made from Spike2 files incorporated into Powerpoint graphics programs (Microsoft, Seattle, WA, USA). Statistical analyses were performed with SigmaStat 3.0 and SigmaPlot 8.0 (SPSS, Chicago, IL, USA). Tests of significance were performed using either Student's t-test or repeated measures ANOVA. Data are presented as the mean ± s.d.
Kinin-like immunolabeling in Cancer borealis
A polyclonal antibody directed against leucokinin I did not label any structures of the C. borealis STNS (N=4) (see also Blitz et al., 1995). Because the STG lies within the ophthalmic artery and is thereby exposed to circulating hormones, we also determined whether there was KLI in the major neurohemal structures. There was no KLI in the SG (N=8). However, KLI was present in axons and varicosities in the POs (N=6/8)(Fig. 2). In each of the six preparations that exhibited KLI, the distribution of the labeled varicosities was patchy and was often restricted to one or more small regions in the nerve trunks that form the PO. Moreover, the specific location of KLI was variable between POs.
Pevkinin excites the pyloric rhythm
We assessed the influence of pevkinin on the STG motor circuits primarily by bath application of PevK-2 to the isolated STG. In a few experiments, where noted below, we also assessed the influence of PevK-6. We used the isolated STG in order to have a more consistent control state in which there was no spontaneous input to the STG from modulatory projection neurons. When the CoGs remain connected with the STG during saline superfusion, there is occasionally an ongoing gastric mill rhythm and always a relatively fast pyloric rhythm(cycle frequency: 1-2 Hz), due to input from spontaneously active CoG projection neurons (Fig. 1). Isolating the STG from the CoGs by transection of both superior esophageal nerves (sons) and ions eliminated the aforementioned inputs,which terminated an active gastric mill rhythm and both slowed and weakened the pyloric rhythm (cycle frequency <1 Hz) (e.g. Fig. 3). In some of these preparations, the pyloric rhythm also terminated.
Application of PevK-2 (10-6 mol l-1) to the isolated STG consistently strengthened the pyloric rhythm. This included an increase in both the speed of the ongoing pyloric rhythm and the activity level of several pyloric neurons (Fig. 3). The LP neuron activity was particularly enhanced by PevK-2. By contrast, neither the inferior cardiac (IC) nor ventricular dilator (VD) neurons was activated by PevK-2 (10-6 mol l-1; N=30) (e.g. Fig. 3). As a control to verify that the IC and VD neurons were accessible to activation, in most experiments we also stimulated the identified projection neuron MCN1 by extracellular ion stimulation (Bartos and Nusbaum, 1997). In all cases, MCN1 excited the pyloric rhythm and activated the IC and VD neurons (N>20). In the isolated STG, the VD neuron is usually silent while IC neuron activity is sometimes present. When the IC neuron was rhythmically active in saline, PevK-2 application did not change that level of activity (saline, 2.3±1.6 spikes/burst;PevK-2, 2.1±1.2 spikes/burst; N=6, P>0.05).
As mentioned above, in some preparations the pyloric rhythm terminated when the CoGs were removed. In every one of the 7 preparations where the pyloric rhythm was not active during saline superfusion, PevK-2 application(10-6 mol l-1) initiated this rhythm(Fig. 4). The mean pyloric cycle frequency and number of LP spikes per burst in these preparations (cycle frequency: 0.53±0.15 Hz, no. LP spikes/burst: 15.1±7.3; N=7) were comparable to those occurring during PevK-2 application in preparations with slow but persisting pyloric rhythms (see below).
In preparations with ongoing pyloric rhythms during saline superfusion, the fastest pyloric cycle frequency occurring during PevK-2 application(10-6 mol l-1) was 1.03 Hz. However, the mean cycle frequency that occurred during these PevK-2 applications was considerably slower (saline, 0.59±0.2 Hz; PevK-2, 0.71±0.14 Hz; N=30, P<0.0001). As shown by the scatter plot in Fig. 5, within the range of control values examined, PevK-2 (10-6 mol l-1) usually increased the pyloric rhythm speed. It is noteworthy, however, that not all of the preparations having a relatively fast pyloric rhythm (>0.5 Hz) during saline superfusion exhibited an increased cycle frequency in the presence of PevK-2 (Fig. 5). This inconsistent result of PevK-2 at faster control levels might have been a consequence of the PevK-2 excitation of the LP neuron (see below). We did not assess the influence of PevK-2 on pyloric rhythms faster than 1 Hz because this situation rarely occurred in the isolated STG during saline superfusion.
In contrast to the ability of PevK-2 to consistently increase the speed of the pyloric rhythm, superfusion of PevK-6 (10-6 mol l-1)was not as effective at changing the speed of this rhythm (saline,0.47±0.33 Hz; PevK-6, 0.80±0.18 Hz; N=6, P>0.05). This was the case even when they were applied to the same STG (N=6). Despite this discrepancy, PevK-6 application was as effective as PevK-2 in enhancing LP neuron activity (saline, 3.6±2.1 spikes/burst; PevK-2, 11.6±3.6 spikes/burst, N=40; PevK-6,14.4±3.2 spikes/burst; N=6; PevK-2 and PevK-6 relative to saline: P<0.001; PevK-2 versus PevK-6: P>0.05) (Fig. 6). Unlike the response of the pyloric cycle frequency to PevK-2 application(Fig. 5), the increased number of spikes per burst in the LP neuron in response to PevK-2 occurred in every preparation, across the entire range of control values(Fig. 6). The PevK-mediated excitation of the LP neuron also included a significant increase in its intra-burst firing frequency. This LP neuron parameter nearly doubled during PevK-2 superfusion (saline, 11.3±4.6 Hz; PevK-2, 21.6±4.9 Hz, N=35, P<0.001).
The LP neuron directly inhibits the PD neuron and, through electrical coupling between the PD and AB neurons, it also inhibits the AB neuron(Eisen and Marder, 1982). Thus, during PevK superfusion, strengthened LP neuron activity resulted in an increased amplitude and frequency of LP-mediated IPSPs in the electrically coupled pyloric pacemaker neurons (e.g. PD neuron in Fig. 3; see also below). Additionally, the LP neuron burst duration was increased in the presence of PevK-2 (saline, 0.26±0.1 s.; PevK-2, 0.47±0.15 s., N=30; P<0.0001). This increased burst duration resulted in an increased duration of LP-mediated inhibition in the pyloric pacemaker neurons (e.g. Fig. 3). In contrast, PevK-2 superfusion did not alter the PD neuron burst duration(saline, 0.29±0.15 s.; PevK-2, 0.24±0.08 s., N=7; P>0.05). However, PevK-2 (10-6 mol l-1)superfusion did increase both the number of PD neuron spikes/burst (saline,3.6±1.6 spikes; PevK-2, 5.0±1.7 spikes, N=7, P<0.05) and its intra-burst firing frequency (saline,8.4±3.3 Hz; PevK-2, 17.0±7.3 Hz; N=7, P<0.05).
Application of PevK-2 to the isolated STG did not change the relative progression of rhythmic, repeating activity in the pyloric neurons PD, LP and PY. However, it did change the fraction of a normalized pyloric cycle during which each of these neurons was active (`duty cycle'). Specifically, there was an increase in the PD and LP neuron duty cycles and a decreased duty cycle in the PY neurons (Fig. 7). The increased PD neuron duty cycle resulted in an increase in the cycle phase(latency from cycle onset to measured parameter divided by the cycle period)at which its burst terminated (Fig. 7). PD neuron burst onset defines the start and end of each pyloric cycle, precluding there being any change in the cycle phase at which its burst begins. The increased LP neuron duty cycle resulted from its mean burst onset being phase advanced and its burst offset being phase delayed relative to saline controls (Fig. 7). The decreased PY neuron duty cycle resulted largely from the phase delay in its burst onset relative to saline controls(Fig. 7). There was also a smaller but significant phase delay in its mean burst offset.
We also assessed the influence of PevK-2 on the pyloric rhythm when applied at different concentrations. Surprisingly, we found that the PevK-2 actions on the pyloric rhythm cycle frequency and LP neuron activity had different thresholds (Fig. 8). Specifically, the threshold for PevK-2 excitation of the LP neuron was 10-7 mol l-1, whereas its ability to increase the pyloric cycle frequency had a threshold of 10-6 mol l-1(Fig. 8A,B). However, the PevK-2 excitation of the LP neuron was stronger at 10-6 mol l-1 than at the lower concentration(Fig. 8B).
LP neuron regulation of the pyloric rhythm during pevkinin application
As indicated above, when the control pyloric rhythms were faster than 0.5 Hz, PevK-2 application did not increase the speed of the rhythm in all preparations (Fig. 5). We therefore assessed whether this variability might have resulted from the strengthened LP neuron inhibition of the pyloric pacemaker group limiting the increase in cycle speed in these preparations. The pyloric rhythmicity originates largely from the intrinsic properties of the anterior burster (AB)neuron, which is a conditionally active endogenous oscillator(Miller and Selverston, 1982). There are two sources of pyloric circuit regulation of the intrinsic rhythmicity of the AB neuron, including its electrical coupling to the PD neurons and the feedback inhibition that it receives through that electrical coupling, as a result of LP neuron inhibition of the PD neuron(Hooper and Marder, 1987; Ayali and Harris-Warrick,1999). To determine whether the strong PevK-2 excitation of the LP neuron was limiting the ability of this peptide to increase the speed of the rhythm, we compared the pyloric rhythm cycle frequency during PevK-2(10-6 mol l-1) superfusion with the LP neuron active and inactive. LP neuron activity was readily suppressed by hyperpolarizing current injection. As shown in Fig. 9A,suppressing LP neuron activity did not alter the speed of the pyloric rhythm during saline superfusion. By contrast, performing the same manipulation during PevK-2 superfusion did indeed increase the ongoing pyloric cycle frequency (Fig. 9B). Note that,whenever the LP neuron activity was suppressed, the LP-mediated IPSPs in the PD neuron were eliminated and the trough of the PD neuron membrane potential oscillations were depolarized. This effect on pyloric cycle frequency was consistent across preparations (saline with LP active, 0.49±0.15 Hz;saline with LP silent, 0.55±0.13 Hz, N=9, P>0.05;PevK-2 with LP active, 0.68±0.13 Hz; PevK-2 with LP silent,0.96±0.12 Hz, N=9, P<0.001)(Fig. 10).
Pevkinin minimally alters the VCN-elicited gastric mill rhythm
Despite consistently altering the pyloric rhythm, PevK-2 (10-6mol l-1) superfusion never activated the gastric mill rhythm in the isolated STG (N=40). However, in these preparations, bath application of this peptide did consistently and reversibly elicit rhythmic bursting in one gastric mill retractor motor neuron, the dorsal gastric (DG) neuron(Fig. 4). In general, the DG neuron was silent during saline superfusion (N=22/25 preparations)(e.g. Fig. 4). In three of these preparations, this neuron instead exhibited spontaneous rhythmic bursting, with an erratic cycle period, during saline superfusion. In 23 of these same 25 preparations, superfusion of PevK-2 or PevK-6 (10-6mol l-1) elicited regular DG neuron bursting that persisted for the duration of peptide application (DG burst duration: 11.06±4.0 s; DG intraburst firing frequency: 9.27±1.3 Hz; DG cycle period:39.5±12.5 s; N=23). This rhythmic DG neuron activity was independent of the activity pattern of all other STG neurons (e.g. Fig. 4). The activity of none of the other gastric mill neurons was altered by PevK superfusion (data not shown).
As is the case for the pyloric rhythm, there are several distinct versions of the gastric mill rhythm (Blitz and Nusbaum, 1997; Beenhakker and Nusbaum, 2004; Blitz et al.,2004; Wood et al.,2004). Unlike the pyloric rhythm, the gastric mill rhythm is usually not spontaneously active in the isolated STNS. However, it can be readily activated by stimulation of the appropriate pathway(s). To determine whether pevkinin also influenced the gastric mill rhythm, we focused our experiments on gastric mill rhythms activated by extracellular stimulation of the mechanosensory ventral cardiac neurons (VCNs)(Beenhakker et al., 2004). We selected this gastric mill rhythm because a relatively brief VCN stimulation(1-2 min) triggers a long-lasting gastric mill rhythm that is stable for tens of minutes. This gastric mill rhythm results from the VCNs triggering a persisting activation of two identified projection neurons, including modulatory commissural neuron 1 (MCN1) and commissural projection neuron 2(CPN2), in the commissural ganglia(Beenhakker and Nusbaum, 2004). These two projection neurons in turn drive the gastric mill rhythm. The VCN-elicited gastric mill rhythm is characterized by rhythmic bursting of the LG, GM, IC and MG protractor phase neurons alternating with the retractor phase neurons Int1, DG, VD and AM(Beenhakker et al., 2004; Beenhakker and Nusbaum, 2004)(see Fig. 11).
The VCN-elicited gastric mill rhythm was only minimally altered by the presence of PevK-2 (10-6 mol l-1; N=7; Figs 11, 12). For example, peptide superfusion did not alter either the speed of the gastric mill rhythm or the activity of the gastric mill neurons (Fig. 12). This included no change in the activity of the DG retractor neuron, which as indicated above was consistently activated to burst rhythmically by PevK application in the absence of the gastric mill rhythm. For example, neither the DG neuron burst duration nor intraburst firing frequency was altered by PevK-2 (10-6 mol l-1)superfusion during the VCN-elicited gastric mill rhythm (DG burst duration:saline, 2.8±0.8 s; PevK-2, 3.6±1.6 s; DG intraburst firing frequency: saline, 16.22±3.9 Hz; PevK-2, 18.6±3.8 Hz; N=7, P>0.05) (Fig. 12). Similarly, PevK superfusion caused no change in the LG protractor neuron burst duration or its intraburst firing frequency (LG burst duration: saline, 6.88±3.3 s; PevK-2, 6.21±2.5 s; LG intraburst firing frequency: saline, 15.78±4.6 Hz; PevK-2, 15.98±4.3 Hz; N=7, P>0.05) (Fig. 12). There was, however, a small but significant phase delay in the termination of the DG neuron burst when PevK-2 was superfused during the VCN-elicited gastric mill rhythm (Fig. 12).
Previous studies showed that the gastric mill rhythm elicited by selective stimulation of MCN1 causes a slowing and weakening of the pyloric rhythm during the gastric mill protractor phase(Bartos and Nusbaum, 1997). This action results from the rhythmic presynaptic inhibition of the STG terminals of MCN1 by the LG protractor neuron, which reduces MCN1 excitation of the pyloric rhythm. After finding that pevkinin excited the pyloric rhythm but did not alter the VCN-elicited gastric mill rhythm, we tested the hypothesis that superfusion of this peptide reduced or eliminated the gastric mill protractor-timed slowing and weakening of the pyloric rhythm. We found that PevK-2 superfusion (10-6 mol l-1) did not rescue the pyloric rhythm from the inhibitory influence of the gastric mill rhythm. For example, as shown previously for the MCN1-elicited gastric mill rhythm,the pyloric cycle frequency was consistently slowed during the gastric mill protractor phase when the preparation was superfused with saline (pyloric cycle frequency: retraction phase, 1.26±0.05 Hz; protraction phase,1.14±0.06 Hz; N=7, P<0.0001). In the same preparations, there was a similar slowing during PevK-2 superfusion(retraction phase, 1.25±0.08 Hz; protraction phase, 1.14±0.08 Hz; N=7, P<0.0001).
Previous studies showed that kinin peptides have actions in the periphery,at the level of neuromuscular systems and effector organs(Coast, 1998; Hayes et al., 1989; Te Brugge and Orchard, 2002; Te Brugge et al., 2002; Terhzaz et al., 1999; Torfs et al., 1999). In this paper, we show that pevkinin also modulates motor circuit activity within the crustacean nervous system. Specifically, PevK application to the isolated crab STG routinely excited the pyloric motor rhythm. This PevK action was relatively specific insofar as it had little influence on the gastric mill rhythm, which is also generated in the STG.
Kinins in C. borealis
We found KLI in the pericardial organs, a major neuroendocrine structure in the crab, but not within the STNS. A previous study of kinins in C. borealis, using the same antileucokinin antiserum used in the present study, had revealed no KLI in the POs(Blitz et al., 1995). This discrepancy might have resulted from alterations in the immunoprocessing methodology. For example, in the earlier work, incubations with antiserum were maintained on a refrigerator shelf instead of undergoing continuous agitation. Alternatively, the KLI in the POs might have been overlooked in the original study because, although found consistently across preparations, the immunolabeling was patchy and restricted to small areas of the POs. Mass spectrometric analysis also supports the presence of PevK-2 in the POs,although the amino acid sequence for the peptide represented by PevK-2-like peak has yet to be definitively identified in this structure(Huybrechts et al., 2003).
Given that kinins are likely to reach the STG via hormonal release, it was surprising that the actions of PevK superfusion did not have a lower threshold (Jorge-Rivera and Marder,1996; Nässel,2002). For example, previously studied neuroactive peptides native to C. borealis, including those that are localized to the POs but not the STNS, were found to have thresholds for excitation of the pyloric rhythm that were lower than those for PevK-2 by several orders of magnitude(Skiebe, 2001). This discrepancy in threshold of action suggests that PevK-2 might not be the kinin native to C. borealis. It may well be that the amino acid sequence of the C. borealis kinin (cabkinin) differs slightly from that of the studied pevkinins and that this difference influences the effectiveness of receptor binding. Previous studies of the pyloric rhythm response to several peptide family members, in each case including the peptide(s) native to C. borealis, have shown qualitatively consistent effects on the pyloric rhythm but quantitative shifts in the dose-response relationship(Nusbaum and Marder, 1988; Weimann et al., 1993; Christie et al., 1997). Alternatively, cabkinin might be identical to PevK-2 and the unusually high threshold concentration for exciting the pyloric rhythm might result from this peptide being effectively degraded by extracellular peptidase activity within the STG (Coleman et al., 1994; Wood et al., 2000; Wood and Nusbaum, 2002).
Leucokinin, an insect kinin family member, was shown previously to increase both the pyloric cycle frequency and the number of LP spikes per burst, and to elicit rhythmic bursting in the DG neuron(Blitz et al., 1995). Although leucokinin shares these actions with PevK-2, it differs from PevK-2 in that it neither increased the PD neuron duty cycle nor did it phase advance LP neuron burst onset. A more extensive comparison between the actions of the current and previous results with these kinins is not possible, however, because the previous study had a more limited scope(Blitz et al., 1995).
Pevkinin actions on the pyloric circuit
Pevkinin application to the isolated STG consistently excited the pyloric rhythm, either activating the rhythm or enhancing slow rhythms (<0.5 Hz). However, whereas all slow pyloric rhythms exhibited an increased cycle frequency, only a subset of the faster rhythms were sped up during PevK-2 application. By contrast, all preparations exhibited a considerable increase in the level of LP neuron activity during PevK-2 application. The LP neuron is well-positioned to regulate the pyloric cycle frequency because it makes the only inhibitory feedback synapse onto the pyloric pacemaker neurons(Eisen and Marder, 1982). Our results support the hypothesis that the strengthened LP neuron activity is at least partly responsible for the speed of some of the faster control pyloric rhythms being unaffected by PevK-2 application. Similarly, the lower threshold for PevK-2 excitation of the LP neuron relative to the PevK-2 enhancement of the pyloric cycle frequency might be a secondary consequence of the strengthened LP neuron inhibition of the pyloric pacemaker neurons. Although this feedback inhibition is effective during PevK-2 application, the effectiveness of this synapse is state-dependent. For example, despite the ability of other applied modulators to increase LP neuron activity,suppressing this LP neuron activity often has no impact on the speed of the pyloric rhythm (Hooper and Marder,1987; Ayali and Harris-Warrick,1999; Thirumalai et al.,2006). This was also the case in our control experiments. The mechanism underlying the enhanced ability of the LP neuron to regulate the pyloric rhythm speed during PevK-2 superfusion, however, awaits further study.
Pevkinin actions on the gastric mill circuit
As shown previously for leucokinin(Blitz et al., 1995), the only gastric mill circuit neuron affected by PevK-2 is the DG neuron. We did not determine whether this PevK action was direct, but the resulting rhythmic bursting in the DG neuron appeared to result from the activation of intrinsic oscillatory properties in this neuron. This conclusion was drawn on the basis of the DG neuron bursting pattern being independent from every other STG neuron. However, we did not eliminate the possibility that PevK application was indirectly influencing the DG neuron by eliciting transmitter release from the STG terminals of one or more projection neurons. These terminals remain functional in the acutely isolated STG, despite the elimination of their somata and arborizations in the CoGs(Nusbaum et al., 1992). In C. borealis, comparable independent rhythmic bursting in the DG neuron is also elicited by application of other neuromodulators and by selective stimulation of either an identified proprioceptor neuron or a CoG projection neuron (Katz and Harris-Warrick, 1989; Kiehn and Harris-Warrick, 1992; Coleman and Nusbaum, 1994; Blitz et al., 1995).
Despite the ability of PevK-2 to elicit rhythmic bursting in the DG neuron,it neither elicited nor altered the gastric mill rhythm. It might seem surprising that PevK-2 application did not at least indirectly influence other gastric mill neurons via its excitation of the DG neuron. However,this neuron has only inhibitory synaptic actions within the STG and its synaptic targets are not activated upon rebound from this inhibition(Norris et al., 1994; Beenhakker et al., 2005; Stein et al., 2005).
Neuromodulators can also potentially influence inter-circuit interactions. In the C. borealis STG, the pyloric and gastric mill circuits influence each other. For example, the pyloric rhythm is slowed and weakened during the protraction phase of the gastric mill rhythm. This effect results from the presynaptic inhibition of the STG terminals of the projection neuron MCN1 by the gastric mill protractor neuron LG, thereby reducing MCN1 excitation of the pyloric rhythm (Bartos and Nusbaum, 1997). This raised the possibility that, by its selective excitation of the pyloric rhythm, PevK-2 would interfere with the gastric mill weakening of the pyloric rhythm by substituting its own excitation of the pyloric rhythm for that contributed by MCN1. However, PevK-2 did not alter the gastric mill regulation of the pyloric rhythm. Although we did not explicitly determine why this was so, it may result from the fact that the pyloric rhythm cycle frequency during the gastric mill protractor phase is comparable to that during PevK-2 superfusion(Bartos and Nusbaum, 1997).
Using the well-characterized circuits in the STG, we have shown that kinin peptides can act within the arthropod CNS to change the output of a rhythmically active motor circuit. Given the previously documented peripheral actions of kinin peptides in other arthropods(Coast, 1998; Hayes et al., 1989; Te Brugge and Orchard, 2002; Te Brugge et al., 2002; Terhzaz et al., 1999; Torfs et al., 1999) and its neurohemal organ localization in C. borealis, it may well be that the PevKs further modify the pyloric motor system via peripheral actions. The C. borealis neuromuscular system is known to be extensively modulated, both pre- and post-synaptically(Jorge-Rivera and Marder,1996; Jorge-Rivera et al.,1998; Birmingham et al.,2003; Messinger et al.,2005). Similarly, despite having no central actions on the gastric mill circuit, the PevKs may well also alter the strength and/or pattern of gastric mill muscle contractions. Given the presence of many different neuroactive substances in the POs and in the terminals of projection and sensory neurons that innervate the STG neuropil(Nusbaum et al., 2001; Skiebe, 2001; Marder et al., 2005; Billimoria et al., 2005), it will be informative to determine which of these substances are normally co-released with the native PevK-like peptide in C. borealis as well as to determine the consequences of that co-release.
Cancer borealis kinin
central nervous system
commissural projection neuron 2
discontinuous current clamp
dorsal gastric nerve
dorsal posterior esophageal nerve
inferior esophageal nerve
inhibitory postsynaptic potential
lateral gastric nerve
lateral ventricular nerve
modulatory commissural neuron 1
medial ventricular nerve
pyloric dilator nerve
superior esophageal nerve
stomatogastric nervous system
ventral cardiac neuron
This work was supported by National Institute of Neurological Disorders and Stroke grant NS29436 (M.P.N.), the University of Washington (A.E.C.), Fonds voor Wetenschappelijk Onderzoek (FWO) grant G.0580.06 (L.S.), and a FWO postdoctoral fellowship to J.H.