| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
NEUROSCIENCE |
Department of Physiology and Pharmacology and Neurological Sciences Institute
1 , Oregon Health & Science University, Portland, OR 972393098, USA
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
(Received 14 March 2007;
accepted after revision 11 May 2007;
first published online 18 May 2007)
Corresponding author T. W. Bailey: Department of Physiology and Pharmacology, Oregon Health & Science University, Portland, OR 972393098, USA. Email: bailey.ohsu{at}gmail.com
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
We previously identified marked differences in synaptic organization and transmission properties in two separate groups of NTS projection neurons (Bailey et al. 2006a). Using retrograde tracers, we identified NTS neurons that projected either to the paraventricular hypothalamus (PVN) in the forebrain or to the caudal ventrolateral medulla (CVLM), an area within the lower brainstem (Bailey et al. 2006a), two independent pools of NTS projection neurons (Hermes et al. 2006). The CVLM is a critical component of the baroreceptor reflex and contains GABAergic neurons that inhibit rostral ventral lateral medullary neurons and reduce sympathetic tone (Aicher et al. 1995, 1996; Jeske et al. 1995). PVN contains a highly heterogeneous population of neurons. Some PVN neurons regulate neuroendocrine function through projections to the posterior pituitary or median eminence while others project to brainstem and spinal cord to regulate sympathetic tone (Toney et al. 2003; Guyenet, 2006) In horizontal brainstem slices, activation of the solitary tract (ST) evoked only monosynaptic excitatory responses to CVLM-projecting NTS neurons, but generally triggered only polysynaptic responses in PVN-projecting NTS neurons. These results suggested fundamentally different synaptic organization along specific intra-NTS pathways associated with different projection targets.
NTS neurons express a broad complement of voltage-dependent and -independent currents and the differential expression of these intrinsic membrane properties is an important component of NTS neuron heterogeneity (Dekin et al. 1987; Dekin & Getting, 1987; Sundaram et al. 1997; Bradley & Sweazey, 1990). Since intrinsic properties of neurons strongly impact on the translation of synaptic inputs into action potential outputs, the present studies focused on a systematic assessment of whether properties of NTS projection neurons contributed to differential processing of ST afferent signals. In neurons characterized by their ST synaptic responses, voltage step protocols and pharmacological channel blockers identified a large, transient A-type potassium current (IKA) that markedly attenuated the generation of action potentials to both depolarization and to ST synaptic activation in PVN-projecting NTS neurons but not in the CVLM-projecting group. PVN-projecting NTS neurons polysynaptically coupled to ST had much larger somatic areas with complex dendritic morphology compared with the smaller simpler CVLM group. Interestingly, shifts in membrane voltage altered the success rate and timing characteristics of ST-triggered action potentials in a manner consistent with the kinetic and voltage-dependent properties of IKA in PVN but not CVLM output pathways. These intra-NTS mechanisms dynamically and differentially tune the frequency response characteristics of these two pathways and profoundly impact central integration of afferent information at sites very early in the overall reflex process.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Retrograde tract tracing identifies specific projection neurons in NTS
All recorded neurons were prepared according to the procedures previously described in detail (Bailey et al. 2006a). Retrograde tracers rely on somewhat different uptake mechanisms and individual tracers have different physical properties including diffusivity and differences in procedure that could influence labelling efficiency and detection of neurons. To control for these differences, three different retrograde tracers were used to identify NTS projection neurons: 1,1'-dioctadecyl-3,3,3',3'-tetramethyindocarbocyanine perchlorate (DiI, 0.5% in ethanol, Molecular Probes, Eugene, OR, USA), red RetroBeads (rhodamine; LumaFluor Inc., Naples, FL, USA) or FluoroGold (2% in saline; Fluorochrome Inc., Englewood, CO, USA).
CVLM injections
Rats (230–270 g) were anaesthetized with isoflurane (4% for induction, 2% for maintenance) and tracers were pressure injected (Picospritzer II, General Valve Inc., Fairfield, NJ, USA) as previously described (Hermes et al. 2006; Bailey et al. 2006a). CVLM was approached from the dorsal surface with stereotaxic coordinates 1.0 mm rostral, 1.9–2.0 mm lateral and 1.9–2.0 mm ventral from calamus scriptorius. Tracer volumes were 50–100 nl for FluoroGold and 50–120 nl for rhodamine.
PVN injections
In one group of rats (40–70 g), DiI was injected iontophoretically using positive current (0.1–1 µA; 10 s on/off cycles; 1.5–3 min) as previously described (Bailey et al. 2003; Doyle et al. 2004). Briefly, these small rats were anaesthetized with intramuscular injections (1 ml (kg body wt)–1) of a cocktail (ketamine, 56 mg ml–1, xylazine, 6 mg ml–1 and acepromazine, 1 mg ml–1); and mounted in a stereotaxic apparatus. To accommodate these small rats, the incisor bar was 5 mm above horizontal and stereotaxic arms canted 10 deg lateral from vertical. Left PVN was targeted at coordinates of 0.4 mm rostral, 1.65 mm lateral and 6.75–6.8 mm ventral to bregma in these DiI rats. In much larger rats (230–270 g) anaesthetized with isoflurane, the stereotaxic incisor bar was set 9 mm below horizontal and the arms were vertical. For these rats, the PVN coordinates were 1.9 mm caudal, 0.6 mm lateral and 7.6 mm ventral from bregma. Either rhodamine (115–320 nl) or FluoroGold (150–200 nl) was pressure injected into the left PVN as previously described (Hermes et al. 2006; Bailey et al. 2006a).
Tracer and electrophysiological identification in slices
Slices for study were sectioned 1–10 weeks following tracer injection. Since results from the different tracers were similar, findings were combined for reporting. Patch electrodes, 1.8–3.5 M
, were guided to cell bodies containing retrograde tracer under infrared illumination using differential interference contrast microscopy (40x, Axioskop FS2+; Zeiss, Thornwood, NY, USA). Electrodes were filled with a solution composed of (mM): 10 NaCl, 130 potassium gluconate, 11 EGTA, 1 CaCl2, 2 MgCl2, 10 Hepes, 1 NaATP, 0.1 NaGTP; pH 7.3; 295 mosmol. In some cases, 0.5% biocytin was included in the recording pipette for post-experiment examination of the structure of the recorded neuron. Whole-cell voltage clamp recordings were made with an Axoclamp 2A or Multiclamp 700B amplifier (Axon Instruments, Foster City, CA, USA). Tip potentials were minimal (
4 mV) and data were not corrected. Signals were sampled at 50–100 kHz and filtered at 5 kHz via an Axon 1325 A/D converter and pCLAMP9 software (Axon Instruments).
ST synaptic characterization
A concentric bipolar stimulating electrode (200 µm outer diameter, F. Haer, Bowdoinham, ME, USA) was placed on the ST remote from the recording site (1–5 mm). Initially, the relationship between ST stimulus intensity and response amplitude was evaluated in each neuron (Bailey et al. 2006b). For afferent driven protocols, ST shock intensity was suprathreshold (2–5x threshold) and shocks were delivered via an isolated, programmable stimulator (Master-8, AMPI, Jerusalem, Israel) using bursts of five shocks at 50 Hz repeated each 3 s. Latency and jitter of the first ST-evoked EPSC in each burst (Doyle & Andresen, 2001) were measured within each neuron. Synaptic jitter in each neuron was calculated as the S.D. of latency to repeated ST shocks (n > 20) and used to classify neurons as either ST monosynaptic (jitter < 200 µs, i.e. second-order neurons) or ST polysynaptic pathways (jitter > 200 µs, i.e. higher-order NTS neurons). Following synaptic characterization, voltage clamp studies of potassium currents and/or current clamp tests of evoked action potential patterns were performed.
Potassium current measurements
To assess outward currents, voltage clamp protocols included command voltage steps to evoke transient and steady outward currents. The voltage dependence of the activation of these currents was assessed by initially conditioning at –90 mV for 500 ms followed by longer activation test steps (1200 ms) to voltages ranging from –100 to 0 mV. Voltage-dependent inactivation was tested with conditioning steps (500 ms) ranging from –100 to 0 mV that were followed by a longer test step (1200 ms) to –10 mV. Steady currents were measured near the end of the extended test step in each protocol. The peak of the transient outward currents was measured after capacitative transients had subsided at 3–15 ms following the initiation of the long step. TEA (10 mM) and 4-aminopyridine (4AP, 5 mM) were obtained from Sigma-RBI (Natick, MA, USA) and were dissolved in extracellular solution on the day of experiment.
Action potential responses
To measure the encoding capacity of projection neurons, action potential responses were tested under current clamp to two excitation protocols: (1) injections of depolarizing currents, and (2) ST activation of synaptic inputs. In both cases, hyperpolarizing steps of current were used to condition neurons by removing voltage-dependent inactivation from IKA. These conditioning steps hyperpolarized neurons to –90 mV for 450 ms. ST stimulation protocols used bursts of ST shocks identical to those used for synaptic characterization under voltage clamp. Success rates of action potential generation to ST shocks were expressed as an input–output ratio calculated in each neuron by dividing the number of action potentials evoked by the number of ST shocks delivered. Fractional input–output ratios (I–O) for ST shocks within the burst (i.e. 1–5) were calculated at a fixed low, intra-burst frequency (5 Hz) but also presented as an aggregate measure averaged over the entire burst for high-frequency, intra-burst inputs (50 Hz). Note that the I–O ratio is distinct from the synaptic failure rate (absence of an EPSC) (Bailey et al. 2006a) and instead measures the success of synaptic activation being translating into an action potential, a process that reflects postsynaptic properties of each neuron's ‘safety factor’. Thus, action potential successes clearly reflect the integrated excitability of the postsynaptic neuron.
Statistical analyses
All data are presented as averages ± S.E.M. Statistical comparisons were made using either Student's unpaired t test, repeated measures (RM) ANOVA, or one-way ANOVA followed by Fisher's PLSD, or paired t test post hoc analysis where appropriate (see individual results; Statview 4.57, Abacus Concepts). P values < 0.05 indicated significant differences. Except where noted, all tests included an initial ANOVA analysis followed by post hoc difference testing and P values cited reflect the results of the final post hoc test.
Histochemistry
Neurons were filled with 0.5% biocytin during electrophysiological experiments and then slices were processed histologically to facilitate visualization. After recording, brainstem slices were fixed rapidly by microwave exposure in 4% paraformaldehyde–1% gluteraldehyde in 0.1 M phosphate buffer (PB, 23 mM NaH2pO4*H2O + 77 mM Na2HPO4) at 800 W for 5 s. Fixed slices were then rinsed in 0.1 M PB and incubated for 30 min in 1% sodium borohydride in 0.1 M PB before being transferred to a 0.1 M Tris-saline solution containing 1% H2O2 (to block endogenous peroxidase) for 30 min. Slices were then rinsed and incubated in a solution containing an avidin–biotin complex (Elite Vectastain ABC kit; Vector Laboratories, Burlingame, CA, USA) and 0.25% Triton in 0.1 M Tris-saline for 24 h at 4°C. Slices were rinsed in 0.1 M Tris-saline and the avidin–biotin complex was visualized by placing the slice in 0.1 M Tris-saline containing diaminobenzidine (DAB, 0.5 mg ml–1) and 0.05% H2O2 for 5–8 min. Slices were rinsed in 0.1 M PB, mounted on gelatin-coated slides, dehydrated in alcohols and xylenes then coverslipped with DPX (distyrene dibutylphthalate; Sigma-Aldrich, MO, USA) mounting media.
Neurolucida
Filled cells were traced using Neurolucida (5.05.4, MicroBrightField, Inc., Williston, VT, USA), Lucivid (MicroBrightField, Inc.) and a Nikon Eclipse E600 microscope (Melville, NY, USA). To minimize bias, observers blinded to the type of cell performed all tracings. Soma and all visible processes including spines were traced. Morphometric analyses were performed with Neural Explorer (5.05.4, MicroBrightField, Inc.). The resulting collective measurements were then decoded, tabulated and summarized.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Using retrograde tracer injections targeted for CVLM or PVN, injection sites typically were limited to the region encompassing and immediately surrounding the area of interest (Fig. 1A). Horizontal slices were cut to preserve neurons within NTS together with a long section of intact ST fibres (Fig. 1B, top panel). Successful tracer injections labelled projection neurons throughout the ipsilateral NTS (Fig. 1B, lower panels). Individual injection sites were examined and cases were discarded in which injections missed the area of interest (i.e. did not cover a substantial component of the targeted region) or extended beyond the targeted region by more than 500 µm.
|
Retrograde tracers injected into either CVLM or PVN illuminated individual NTS projection neurons (Fig. 2A). We studied a total of 28 CVLM projection neurons from 20 separate slice preparations and 30 total PVN projection neurons from 26 separate slice preparations. These neurons received distinctly and characteristically different ST synaptic responses depending upon their projection target (Fig. 2B and see Bailey et al. 2006a). In voltage clamp recordings in CVLM-projecting NTS neurons (n = 28), ST shocks activated large-amplitude (258 ± 56 pA), nearly invariant EPSCs (synaptic jitter < 200 µs) that were consistent with second-order neurons directly contacted by cranial visceral afferents (Fig. 2B, left). Note that EPSC amplitude rapidly declined within a short duration burst of five shocks (50 Hz) and this form of frequency-dependent depression was prominent in all CVLM-projecting NTS neurons. In contrast in NTS neurons projecting to PVN (n = 24), identical ST stimulus bursts (Fig. 2B, right) generally evoked EPSCs with smaller amplitudes (81 ± 9 pA), high latency variability (high synaptic jitter > 200 µs) and little depression – characteristics associated with higher-order neurons that were polysynaptically connected to the ST (Bailey et al. 2006a). As with unlabelled neurons in naïve slices, a small percentage of dye-filled neurons did not respond to ST stimulation. Such neurons were excluded from further study (PVN, n = 4; CVLM, n = 9) as they may represent neurons with ST connections damaged by slice preparation. Depolarizing voltage steps evoked distinctly different whole-cell currents in these two groups of neurons (Fig. 2C). In second-order CVLM-projecting NTS neurons, steps to progressively more depolarized levels activated relatively steady outward currents (Fig. 2C, left). In contrast, identical command steps evoked substantial, early transient currents in the typical higher-order, PVN-projecting NTS neuron (Fig. 2C, right). With sustained depolarization, the peak outward currents in higher-order, PVN-projecting neurons decayed over several hundred milliseconds to a lower steady level. Such results indicate a systematic difference in potassium channels expressed in these two groups of anatomically and synaptically defined NTS projection neurons.
|
Overall, the mean current–voltage relations indicated distinctly different outward currents in second-order, CVLM-projecting neurons compared with higher-order, PVN-projecting neurons (Fig. 3A, n = 20 CVLM, n = 18 PVN). The transient components of the currents activated at membrane potentials more depolarized than –60 mV were significantly greater in higher-order, PVN-projecting neurons than the CVLM-projecting group (Fig. 3A, top panel, P < 0.003). No differences were found in the steady currents between the two groups of neurons (Fig. 3A, lower panel, P > 0.2). The magnitude of the peak transient outward current varied greatly across individual higher-order, PVN-projecting neurons but was uniformly low across the population of CVLM-projecting neurons (Fig. 3B). Thus, not only were these currents different across these two groups but the higher-order, PVN-projecting neuron group displayed great heterogeneity of K+ channel expression.
|
|
|
Action potentials are required to convey information from NTS projection neurons to sites beyond NTS. To test whether the differences in IKA and IKV might be associated with different spike encoding capacities, we measured action potential generation in response to depolarization under current clamp conditions (Fig. 6). Typically, from resting potential (–60 mV), moderate depolarizing current injections elicited multiple action potentials with minimal delay in both CVLM- and PVN-projecting NTS neurons (Fig. 6Aa and Ba). Brief conditioning by hyperpolarization (–30 pA current injection) to –90 mV, however, induced a delay in the excitatory response to current injection in PVN-projecting NTS neurons (Fig. 6Bb) but not in CVLM-projecting NTS neurons (Fig. 6Ab). Thus, conditioning at potentials that remove IKA inactivation (potentials hyperpolarized compared with resting levels) transiently reduced discharge rates to depolarization in PVN-projecting but not CVLM-projecting NTS neurons.
|
Activation of excitatory synapses generates EPSPs that can potentially trigger action potentials but the safety factor for this process depends on many factors beyond the amplitude of the EPSP. To test the effectiveness of the translation of ST shocks into action potentials in NTS projection neurons, we stimulated ST and recorded action potentials (i.e. output) triggered in current clamp at resting potential (–58.1 ± 1.8 mV, PVN, n = 10; –57.8 ± 1.8 mV, CVLM, n = 9) in the two groups of NTS projection neurons. Dividing the number of output spikes by the number of ST shocks delivered (i.e. input) yielded an I–O ratio. In each neuron, we stimulated the ST pathway in bursts of five shocks at a moderate frequency (5 Hz, Fig. 7A, upper) and a higher one (50 Hz, Fig. 7A, lower). In CVLM-projecting NTS neurons (single representative CVLM projection neuron in Fig. 7A, left) the bursts of 5 Hz typically evoked stimulus-synched action potentials that faithfully encoded the ST inputs for nearly every shock and ST shocks late in the burst were only modestly less likely to trigger an action potential (Fig. 7A, left). In contrast, the identical bursts of ST inputs rarely activated an action potential in typical higher-order, PVN-projecting neurons (a single representative PVN-projection neuron in Fig. 7A, right). Increasing the intra-burst frequency to 50 Hz decreased the number of successfully evoked action potentials in both of these representative projection neurons (Fig. 7A, lower). As might be predicted, the latency to action potential was always substantially longer than the EPSC latency in the same neuron with greatly increased jitter for that action potential latency (Fig. 7A), but the CVLM-projecting second-order neurons nonetheless remained tightly synced to ST shocks compared with higher order PVN-projecting NTS neurons.
|
On average, the I–O ratio was close to 1.0 for CVLM-projecting NTS neurons (n = 9) for the initial shock within the burst and declined slightly but significantly for later shocks within the burst, even at the 5 Hz frequency (Fig. 7B, left, RM ANOVA, P < 0.01). Strong frequency-dependent depression at these primary afferent synapses probably contributes to this decline in I–O throughout the burst of ST shocks. In contrast, the mean I–O ratio at 5 Hz for higher-order, PVN-projecting NTS neurons (n = 10) was quite low, approximately 0.2 for the very first shock and held steady with no use-dependent decline (Fig. 7B, left). At all positions within the 5 Hz ST burst sequence, CVLM-projecting NTS neurons successfully translated more ST shocks to action potentials than PVN-projecting NTS neurons (Fig. 7B, left; P < 0.001). Presented as an aggregate, mean response across the entire burst, the I–O ratio for the CVLM-projecting NTS neurons was more than three times greater at 5 Hz than for the higher-order, PVN-projecting group (Fig. 7B, right; P = 0.017). Increasing the intra-burst ST shock frequency to 50 Hz depressed this aggregate I–O ratio in the CVLM-projecting NTS neurons (P = 0.003), but higher-order, PVN-projecting neurons remained substantially lower at < 0.2 (Fig. 7B, right). Thus, CVLM-projecting NTS pathways convey more afferent information than do higher-order PVN pathways regardless of activation frequency. However, because IKA is partially activated at resting potentials these throughput assays test the combined contribution to pathway throughput of afferent pathway synaptic strength together with IKA and its decay kinetics.
IKA modulates throughput in PVN-projecting neurons
The substantial IKA expression of PVN-projecting NTS neurons (Fig. 3) has a strong potential to affect the translation of EPSPs into action potentials (I–O ratio). To specifically assess the contribution of IKA to pathway throughput and I–O ratio, neurons were tested in two conditioning protocols before delivering each ST burst: first, at –50 mV where IKA is largely inactivated, and second, under conditions in which IKA inactivation was minimized by conditioning neurons at –90 mV. Without hyperpolarizing pre-conditioning, ST activation produced greater I–O ratio in CVLM pathways compared with PVN projecting neurons (n
= 5 and 4, respectively; Fig. 8A and B). However, in PVN-projecting NTS neurons, hyperpolarizing pre-conditioning substantially reduced I–O ratios at the first two ST shocks (i.e. the first 200–280 ms following release from pre-conditioning) within 5 Hz bursts (Fig. 8B, n
= 4, P < 0.02) but had no effect on responses of CVLM-projecting NTS neurons (n
= 5, Fig. 8A, P > 0.12). Burst aggregate I–O means were higher and unaltered by hyperpolarizing preconditioning in CVLM-projecting neurons (Fig. 8B, right; P > 0.25) but I–O was significantly reduced at both 5 and 50 Hz in PVN-projecting neurons (Fig. 8B, right, P < 0.017). Thus, in the absence of IKA, powerful direct afferent pathways to CVLM-projecting neurons drive higher throughput with greater afferent fidelity. Interestingly, in PVN pathways with the lower I–O ratios and weaker, indirect afferent pathways, IKA activation further attenuates the transfer of ST afferent initiated signals into action potentials. IKA expression thus offers a mechanism of dynamic modulation of afferent information transfer by virtue of its voltage-dependent and time-dependent characteristics.
|
All CVLM-projecting NTS neurons were directly innervated by ST afferents and had uniformly low IKA expression. PVN-projecting NTS neurons, however, were generally only indirectly linked to ST afferents and had substantial IKA. A small subset of PVN-projecting NTS neurons had second-order ST synaptic characteristics (Fig. 9A). In these second-order, PVN-projecting NTS neurons (n = 6), step depolarization evoked large IKA currents (Fig. 9A and B) that were identical in all respects (P > 0.16) to IKA in higher-order, PVN-projecting NTS neurons (n = 18) and both PVN groups exceeded the mean IKA in CVLM-projecting NTS neurons (n = 20, P < 0.0001). Thus, the expression of large IKA currents was uniformly associated with the PVN pathway whether synaptically organized within NTS as directly or polysynaptically linked to ST afferents.
|
In the process of visually identifying neurons for recording in slices (Fig. 2A, and Bailey et al. 2006a), the CVLM- and PVN-projecting neurons differed clearly in their morphology. To more accurately examine cellular morphology, biocytin was included in the recording pipette in some recordings in order to recover structural information from functionally characterized neurons (Fig. 10). Histological examinations of the cellular profiles of these recorded neurons were performed using neurolucida (Fig. 10; Table 1). While it is possible that some portions of neurons were lost during slice preparation, consistent differences were seen between CVLM-projecting neurons and higher-order PVN-projecting NTS neurons. The cell bodies of CVLM-projecting NTS neurons (Fig. 10A) were, on average, approximately half the size of the cell bodies of average PVN-projecting neurons (Fig. 10B; Table 1). The total dendritic length of CVLM-projecting NTS neurons was also significantly shorter than the total dendritic length of PVN-projecting neurons (Table 1). Higher-order PVN-projecting NTS neurons had more than 10-fold the number of dendritic spines than the second-order CVLM-projecting NTS neurons. The greater density of spines may represent increased convergence of various inputs beyond ST linked pathways to PVN-projecting NTS neurons, but generally correlates with their respective cell body size and total dendritic length. Thus, the morphological features of these PVN-projecting neurons are consistent with a potentially greater structural framework to receive inputs distributed across more elaborate postsynaptic cell architecture. The morphological structures of these two neuron classes correlate well with the unique intra-NTS synaptic organization for these two groups of neurons with PVN-projecting neurons receiving complex synaptic inputs and CVLM-projecting NTS neurons commonly receiving relatively simple synaptic inputs (Fig. 2 and Bailey et al. 2006a).
|
|
In previous work, we observed a correlation between IKA expression and ST afferent subtype using the susceptibility of synaptic responses to the TRPV1 agonist, capsaicin (Doyle et al. 2002; Jin et al. 2003). IKA expression was generally greater in second-order NTS neurons receiving unmyelinated, capsaicin-sensitive afferents (Bailey et al. 2002). In a subset of our present studies of NTS projection neurons, capsaicin (200 nM) blocked ST-evoked EPSCs in some second-order NTS neurons (2 of 6 CVLM-projecting neurons and 3 of 4 PVN-projecting NTS neurons). No partial capsaicin blockades (reductions in EPSC amplitude) were observed, a finding consistent with segregation of myelinated from unmyelinated ST afferent contacts on single, second-order NTS neurons (Doyle et al. 2002; Jin et al. 2003). Capsaicin also blocked ST-evoked polysynaptic EPSCs in most (5 of 7) higher-order, PVN-projecting neurons. Presumably, capsaicin acts on ST transmission at the intervening second-order neuron within these polysynaptic intra-NTS pathways. In this subset of neurons tested with capsaicin, PVN-projection neurons expressed significantly larger IKA than CVLM-projection neurons (1070 ± 201 pA versus 62 ± 36 pA; n = 11 and 6, respectively; P = 0.002). Within both CVLM and PVN NTS projection neuron populations, when grouped solely by capsaicin sensitivity (CVLM + PVN), those with capsaicin-sensitive ST-EPSCs expressed far larger IKA compared with capsaicin-resistant (1052 ± 234 pA versus 231 ± 134 pA; n = 10 and 7, respectively; P = 0.01). Thus, regardless of projection target or ST-synaptic order, ST myelinated afferents drive intra-NTS pathways that are associated with small or no IKA.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
|
The first step in information transfer from visceral afferents to the CNS is ST synaptic transmission. Sensory synaptic transmission within the brainstem can be of uncommonly high fidelity (Trussell, 1999) and cranial visceral afferent synapses in NTS are no exception (Andresen et al. 2001; Doyle & Andresen, 2001). Our previous studies (Bailey et al. 2006a) detailed a substantial and target-specific difference in synaptic amplitude and reliability as well as convergence patterns between CVLM-projecting NTS neurons and those sending axons to PVN. ST synaptic currents were uniformly large and rarely failed in all second-order projection neurons. Conversely as a result of the indirect, polysynaptic pathway from ST afferent to most PVN-projecting neurons, ST-synced EPSCs were much smaller and failed to occur nearly 20% of the time. Thus, pathways to such higher-order neurons might include at least one interneuron before reaching the projection neuron (see neuron B in Fig. 11). The presence of marked convergence of ST-driven inputs to these polysynaptically coupled PVN-projecting neurons reflects very different, intra-NTS synaptic organization of afferent pathways compared with second-order neurons and this contributed to relatively indistinct afferent signalling at the level of synaptic currents. Low I–O success rates, it should be noted, occurred in spite of the high likelihood of activation of multiple, parallel excitatory inputs to single ST shocks. These convergent polysynaptic inputs reveal that single ST shocks simultaneously activated multiple ST afferent axons (see neuron C in Fig. 11). This local network convergence was not found in CVLM-projecting NTS neurons (Fig. 11). Afferent signals received at PVN via these polysynaptic intra-NTS paths were greatly attenuated, a finding consistent with the diffuse quality of the responses to baroreceptor activation found in PVN neurons (Kannan & Yamashita, 1983) and in a subgroup of PVN neurons projecting to spinal cord (Chen & Toney, 2003).
IKA expression is featured in particular projection pathways
Synaptic current amplitude is one factor generating the action potentials that convey afferent messages beyond NTS. The ‘safety factor’ for that process includes not only the EPSP amplitude, but also reflects properties of postsynaptic neurons including cytoarchitecture and the passive/active membrane properties that depend on specific ion channel expression (membrane potential, action potential threshold, membrane conductance, etc.). Such intermediary factors add several-fold to the response latency variability within neurons that increases substantially from measures of EPSC timing compared with action potential timing variability (Fig. 7A). Subclasses of NTS neurons have been suggested on the basis of differences in ion channel expression – particularly K channels – and neuron morphology (Dekin & Getting, 1987; Dekin et al. 1987; Paton et al. 1993). As these classifications overlap substantially, their uniqueness or function has been uncertain. Even relative to a specific visceral target within the gastrointestinal system, for example, cell morphology of vagal premotor NTS neurons did not predict synaptic contacts or distinct action potential firing properties (Glatzer et al. 2003). Conversely, nearby dorsal motor nucleus neurons projecting to different targets across the stomach universally expressed IKA (Browning et al. 1999). In NTS projection neurons, the present studies identified a specific association between NTS neuronal phenotype and two prominent CNS targets. Second-order, CVLM-projecting neurons within the medial NTS were substantially smaller in somatic area and had quantitatively simpler dendritic arbor patterns compared with the much more complicated dendritic arbors of higher-order, PVN-projecting NTS neurons. Importantly, all CVLM neurons had little to no IKA expression. However, PVN-projecting neurons, whether higher-order (Fig. 11, pathways B and C) or second-order (Fig. 11, pathway A), expressed substantial IKA currents. Interestingly, the myelinated subclass of ST afferents was associated with a paucity of IKA whether the projection neurons were connected to CVLM or PVN. Thus, at the earliest stages of afferent processing within NTS, differences in pathway structure and key potassium channel expression were closely related to the CNS destination.
Fidelity of afferent signals preserved in simple, direct ST–NTS–CVLM pathway
The more convoluted, polysynaptic pathway to higher-order, PVN-projecting NTS neurons generated action potentials to fewer than 25% of ST shocks. This attenuation of spike generation by afferent excitation compared with a roughly 75% success rate for the CVLM pathway at the same, 5 Hz ST input frequency. Complex arborization itself can strongly influence such I–O coupling (Schaefer et al. 2003), although our experiments do not allow us to assess directly the contribution of neuronal architecture differences to overall performance (Hoffman et al. 1997). The powerful direct synaptic inputs on the simpler and more compact somatodendritic structure of CVLM-projecting neurons were paired with an absence of IKA and produced a consistently higher fidelity translating ST shocks into conducted action potentials.
Dynamic modulation of throughput by IKA
A-type potassium currents are steeply voltage dependent and a portion of these channels are active at resting membrane potentials in NTS neurons (Schild et al. 1993; Bailey et al. 2002). The sharp inflection in the activation characteristics of IKA is close to the average resting potential of these neurons and, as a result, small shifts in membrane voltage will have large effects on IKA activation and excitability. This voltage-dependent aspect of IKA activation plus its rapid inactivation mean that IKA will act as a strong but transient brake on neuron excitation. In the present studies, IKA further dampened synaptically driven I–O ratio by more than 50% in ST–NTS–PVN pathways but had no detectable impact on ST–NTS–CVLM pathways with little IKA. Thus, the time- and voltage-dependent properties of IKA offer a powerful mechanism for dynamic modulation of afferent information destined for forebrain PVN. Since both directly and indirectly coupled PVN-projecting NTS neurons expressed IKA, its potential participation was independent of intra-NTS pathway organization but related to output destination. Virtually any signal that transiently changes membrane voltage in NTS neurons may be responsible for dynamic modulation of IKA. For example, GABAergic synaptic inputs may hyperpolarize and produce a bias for IKA activation and attenuation of throughput. Alternatively, subtle depolarizations though glutamatergic synaptic inputs may dynamically inactivate IKA and augment throughput. However, IKA was larger in pathways initiated by Cfibre afferents regardless of whether targeting CVLM or PVN so that there appears to be a broad co-association of IKA expression with unmyelinated cranial visceral afferents. Both the activation characteristics and expression of IKA together with the natural discharge characteristics of unmyelinated afferents (Andresen et al. 2004) raise the possibility of patterned activity-dependent changes in membrane voltage that modulate IKA and produce dynamic changes in I–O ratio in these specific intra-NTS pathways. Myelinated visceral afferents tend to fire in bursts of high frequencies reaching interspike intervals of a few milliseconds and thus paired with low IKA such trains would tend to be translated into outgoing action potentials attenuated primarily by depression associated with neurotransmitter depletion (Doyle & Andresen, 2001; Bailey et al. 2006b) whereas the low frequency and intermittent discharge of unmyelinated afferents would be attenuated by the presence of IKA except under the strongest stimulation (Kunze & Andresen, 1991). Interestingly, IKA and its inhibitory impact on firing properties are reduced in NTS neurons from hypertensive rat strains (Sundaram et al. 1997) and renal wrap hypertensive rat models (Belugin & Mifflin, 2005). Thus, similar cellular mechanisms may contribute to disease states in which altered autonomic processes contribute to long-term changes in homeostatic regulation.
Functional implications
We focused largely on medial NTS, a region of dense innervation by arterial baroreceptors (Mendelowitz et al. 1992; Andresen & Kunze, 1994) and critical to blood pressure regulation (Weston et al. 2003). The response properties detailed here for CVLM-projecting NTS neurons are consistent with their role in the pathway responsible for sympathoinhibition in baroreflexes. CVLM neurons are probably GABAergic interneurons that provide rapid inhibition of sympathetic premotor neurons within the rostral parts of the ventrolateral medulla during the baroreflex (Aicher et al. 2000; Schreihofer & Guyenet, 2003). The direct connections between ST afferents and CVLM-projecting NTS neurons predict powerful, consistent afferent activation of CVLM neurons. Such high fidelity transmission was evident in the tight coupling of such neurons in vivo to the presumed discharge of afferents during pulsatile increases in arterial pressure recorded in baroreceptor-activated CVLM neurons (Jeske et al. 1993; Schreihofer & Guyenet, 2003). Interestingly, however, stressful conditions even when accompanied by marked hypotension robustly activate PVN and this depends on baroreceptor inputs to NTS (Dampney et al. 2003). The mixture of direct and indirect intra-NTS pathways to PVN that we have identified may contribute to highly divergent and even negatively correlated relationships between afferent activation and target activity in regulation of the autonomic neurons that project to medullary and spinal cord regions (Stern, 2001; Chen & Toney, 2003) and neuroendocrine systems (Sawchenko et al. 1996). These patterns of innervation may result in differential afferent control across the heterogeneous PVN population (e.g. parvocellular neurons versus magnocellular, vasopressinergic neurons projecting to median eminence).
Our results show that the afferent pathways to PVN are heterogeneous and this is consistent with diverse PVN function. In contrast, those pathways to CVLM are remarkably uniform despite the fact that CVLM neurons carry signals from a diverse complement of afferent modalities supporting similarly diverse homeostatic processes including cardiovascular, gastrointestinal and respiratory (Loewy, 1990; Guyenet, 2006). Thus, despite this systems level diversity for CVLM function, afferent information bound for CVLM is relayed through simple NTS pathways with uniformly high fidelity. Thus, pathway differences may be a repeated feature of brainstem organization that depends upon CNS target, afferent modality or specific reflex process.
Many homeostatic reflexes engage rather compact processing pathways within limited central networks. Our results demonstrate that these pathways are distinct in their earliest intra-NTS portions in their local circuit organization, morphology and intrinsic properties and differ even among closely adjacent neurons. This compact nature of brainstem pathways that utilize heterogeneous neurons presents a very different afferent processing strategy than that more common in highly redundant, often fairly uniform cortical networks organized for massively parallel processing.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Aicher SA, Milner TA, Pickel VM & Reis DJ (2000). Anatomical substrates for baroreflex sympathoinhibition in the rat. Brain Res Bull 51, 107–110.[CrossRef][Medline]
Aicher SA, Saravay RH, Cravo S, Jeske I, Morrison SF, Reis DJ & Milner TA (1996). Monosynaptic projections from the nucleus tractus solitarii to C1 adrenergic neurons in the rostral ventrolateral medulla: Comparison with input from the caudal ventrolateral medulla. J Comp Neurol 373, 62–75.[CrossRef][Medline]
Andresen MC, Doyle MW, Bailey TW & Jin Y-H (2004). Differentiation of autonomic reflex control begins with cellular mechanisms at the first synapse within the nucleus tractus solitarius. Braz J Med Res 37, 549–558.
Andresen MC, Doyle MW, Jin Y-H & Bailey TW (2001). Cellular mechanisms of baroreceptor integration at the nucleus tractus solitarius. Ann N Y Acad Sci 940, 132–141.
Andresen MC & Kunze DL (1994). Nucleus tractus solitarius: gateway to neural circulatory control. Annu Rev Physiol 56, 93–116.[Medline]
Bailey TW, Hermes SM, Andresen MC & Aicher SA (2006a). Cranial visceral afferent pathways through the nucleus of the solitary tract to caudal ventrolateral medulla or paraventricular hypothalamus: target-specific synaptic reliability and convergence patterns. J Neurosci 26, 11893–11902.
Bailey TW, Jin Y-H, Doyle MW & Andresen MC (2002). Vanilloid sensitive afferents activate neurons with prominent A-type potassium currents in nucleus tractus solitarius. J Neurosci 22, 8230–8237.
Bailey TW, Jin Y-H, Doyle MW, Smith SM & Andresen MC (2006b). Vasopressin inhibits glutamate release via two distinct modes in the brainstem. J Neurosci 26, 6131–6142.
Bailey TW, Nicol GD, Schild JH & DiMicco JA (2003). Synaptic and membrane properties of neurons in the dorsomedial hypothalamus. Brain Res 985, 150–162.[CrossRef][Medline]
Belugin S & Mifflin SW (2005). Transient voltage-dependent potassium currents are reduced in nts neurons isolated from renal wrap hypertensive rats. J Neurophysiol 94, 3849–3859.
Bradley RM & Sweazey RD (1990). In vitro intracellular recordings from gustatory neurons in the rat solitary nucleus. Brain Res 508, 168–171.[CrossRef][Medline]
Browning KN, Renehan WE & Travagli RA (1999). Electrophysiological and morphological heterogeneity of rat dorsal vagal neurones which project to specific areas of the gastrointestinal tract. J Physiol 517, 521–532.
Chen QH & Toney GM (2003). Identification and characterization of two functionally distinct groups of spinal cord-projecting paraventricular nucleus neurons with sympathetic-related activity. Neuroscience 118, 797–807.[CrossRef][Medline]
Craig AD (2002). How do you feel? Interoception: the sense of the physiological condition of the body. Nat Rev Neurosci 3, 655–666.[Medline]
Dampney RAL, Polson JW, Potts PD, Hirooka Y & Horiuchi J (2003). Functional organization of brain pathways subserving the baroreceptor reflex: studies in conscious animals using immediate early gene expression. Cell Mol Neurobiol 23, 597–616.[CrossRef][Medline]
Dekin MS & Getting PA (1987). In vitro characterization of neurons in the ventral part of the nucleus tractus solitarius. II. Ionic basis for repetitive firing patterns. J Neurophysiol 58, 215–229.
Dekin MS, Getting PA & Johnson SM (1987). In vitro characterization of neurons in the ventral part of the nucleus tractus solitarius. I. Identification of neuronal types and repetitive firing properties. J Neurophysiol 58, 195–196.
Doyle MW & Andresen MC (2001). Reliability of monosynaptic transmission in brain stem neurons in vitro. J Neurophysiol 85, 2213–2223.
Doyle MW, Bailey TW, Jin Y-H & Andresen MC (2002). Vanilloid receptors presynaptically modulate visceral afferent synaptic transmission in nucleus tractus solitarius. J Neurosci 22, 8222–8229.
Doyle MW, Bailey TW, Jin Y-H, Appleyard SM, Low MJ & Andresen MC (2004). Strategies for cellular identification in nucleus tractus solitarius slices. J Neurosci Methods 37, 37–48.[CrossRef]
Duan Y-F, Kopin IJ & Goldstein DS (1999). Stimulation of the paraventricular nucleus modulates firing of neurons in the nucleus of the solitary tract. Am J Physiol Regul Integr Comp Physiol 277, R403–R411.
Glatzer NR, Hasney CP, Bhaskaran MD & Smith BN (2003). Synaptic and morphologic properties in vitro of premotor rat nucleus tractus solitarius neurons labeled transneuronally from the stomach. J Comp Neurol 464, 525–539.[CrossRef][Medline]
Guyenet PG (2006). The sympathetic control of blood pressure. Nat Rev Neurosci 7, 335–346.[Medline]
Hermes SM, Mitchell JL & Aicher SA (2006). Most neurons in the nucleus tractus solitarii do not send collateral projections to multiple autonomic targets in the rat brain. Exp Neurol 198, 539–551.[CrossRef][Medline]
Hoffman DA, Magee JC, Colbert CM & Johnston D (1997). K+ channel regulation of signal propagation in dendrites of hippocampal pyramidal neurons. Nature 387, 869–875.[CrossRef][Medline]
Jeske I, Morrison SF, Cravo SL & Reis DJ (1993). Identification of baroreceptor reflex interneurons in the caudal ventrolateral medulla. Am J Physiol Regul Integr Comp Physiol 264, R169–R178.
Jeske I, Reis DJ & Milner TA (1995). Neurons in the barosensory area of the caudal ventrolateral medulla project monosynaptically on to sympathoexcitatory bulbospinal neurons in the rostral ventrolateral medulla. Neuroscience 65, 343–353.[CrossRef][Medline]
Jin Y-H, Bailey TW, Doyle MW, Li BY, Chang KSK, Schild JH, Mendelowitz D & Andresen MC (2003). Ketamine differentially blocks sensory afferent synaptic transmission in medial nucleus tractus solitarius (mNTS). Anesthesiology 98, 121–132.[CrossRef][Medline]
Kannan H & Yamashita H (1983). Electrophysiological study of paraventricular nucleus neurons projecting to the dorsomedial medulla and their response to baroreceptor stimulation in rats. Brain Res 279, 31–40.[CrossRef][Medline]
Kubin L, Alheid GF, Zuperku EJ & McCrimmon DR (2006). Central pathways of pulmonary and lower airway vagal afferents. J Appl Physiol 101, 618–627.
