Synaptically activated burst-generating conductances may underlie a group-pacemaker mechanism for respiratory rhythm generation in mammals

Abstract

Breathing, chewing, and walking are critical life-sustaining behaviors in mammals that consist essentially of simple rhythmic movements. Breathing movements in particular involve the diaphragm, thorax, and airways but emanate from a network in the lower brain stem. This network can be studied in reduced preparations in vitro and using simplified mathematical models that make testable predictions. An iterative approach that employs both in vitro and in silico models argues against canonical mechanisms for respiratory rhythm in neonatal rodents that involve reciprocal inhibition and pacemaker properties. We present an alternative model in which emergent network properties play a rhythmogenic role. Specifically, we show evidence that synaptically activated burst-generating conductances-which are only available in the context of network activity-engender robust periodic bursts in respiratory neurons. Because the cellular burst-generating mechanism is linked to network synaptic drive we dub this type of system a group pacemaker.

Figure 1

The neural control of breathing studied at multiple levels. A) Hierarchy of model systems from breathing studied in vivo to reduced preparations in vitro, which includes en bloc brain stem-spinal cords and slices, and finally to in silico mathematical models. B) A cartoon of the slice preparation from neonatal rodents in vitro. C) Whole-cell patch-clamp recording of a preBötC neuron with XII respiratory motor output in a slice preparation.

Figure 2

Pacemaker properties in preBötC neurons are a natural byproduct of heterogeneity. The distribution of g_NaP vs. g_K-Leak in the model proposed by Butera and colleagues in 1999 demonstrates that pacemaker properties are expected in 5–25% of constituent neurons. The white area of the plot denotes tonically active neurons; the gray area reflects bursting pacemaker activity; and finally the black area indicates a quiescent state. Superimposed experimental data points were obtained from whole-cell recordings of preBötC neurons in neonatal rats. The two-tuple (g_NaP, g_K-Leak) was added to the existing g_NaP vs. g_K-Leak plot from the model, and the points were color-coded based on whether the recorded neurons showed pacemaker activity after synaptic isolation. The data matched the theoretical distribution, indicating that the ratio of g_NaP/g_K-Leak governs whether intrinsic bursting is possible in any given neuron. This suggests that pacemaker properties are not specialized, but rather a natural byproduct of heterogeneity in a network of cells that all express g_NaP and g_K-Leak.

Figure 3

Riluzole blocks I_NaP. Steady-state IV curves in control and 20 µM RIL (lower left) obtained from slow voltage-ramp commands (upper left). The effects of 1–200 µM RIL on action potentials evoked by 3-ms step current commands (upper right). A dose response curve shows the effects of RIL on I_NaP and on spike amplitude (lower right). Data modified and reprinted with permission from Del Negro, Morgado-Valle, and Feldman (2002) Neuron, 34, 821–30.

Figure 4

Pharmacology of intrinsic bursting in preBötC neurons. A) 10 µM RIL blocks I^NaP-dependent bursting in a representative preBötC neuron synaptically isolated using low Ca²⁺ / high Mg²⁺ artificial cerebrospinal fluid (ACSF). B) 10 µM FFA blocks I_CAN-dependent bursting in a representative preBötC neuron. 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX), 2-amino-5-phosphonopentanoic acid (APV), picrotoxin (PTX), and strychnine hydrochloride (STR) are excitatory synaptic receptor antagonists used to synaptically isolate the neuron in B. In both A and B the baseline membrane potential was ramped from −70 to −40 to test (unsuccessfully) for the presence of voltage-dependent bursting activity. Some plotted data were modified with permission from Del Negro et al. (2005) J Neurosci, 25, 446–53.

Figure 5

The effects of RIL on respiratory motor rhythm and inspiratory preBötC neurons recorded in vitro. A) Motor activity recorded from the hypoglossal nerve (XII) in a cumulative dose-response experiment. RIL concentration is displayed above each trace. Raw and integrated traces (XII) are shown. B) Substance P (SP) at 2 µM restores respiratory rhythm in a mouse slice exposed to 20 nM TTX. This inspiratory neuron hyperpolarized by 7 mV in TTX, therefore +250 pA was applied (I_app) to restore baseline V_M to −60 mV. After the rhythm stopped, 2 µM SP revived it. The amplitude of the cellular drive potentials and XII amplitude recovered in SP, but spike discharge did not occur with baseline VM at −60 mV. Data have been modified with permission from Del Negro et al. (2002) Neuron, 34, 821–30; and Del Negro et al. (2005) J Neurosci, 25, 446–53.

Figure 6

Respiratory rhythm in the presence of 100 µM FFA and 20 µM RIL. Continuous segments of the experiment showing control, FFA, recovery, and then FFA + RIL co-application, and finally FFA + RIL + 2 µM SP conditions in a mouse slice. SP rescued the rhythm after its cessation due to FFA + RIL. Examples of inspiratory bursts and XII output are shown with greater time resolution (lower panels). Data have been modified with permission from Del Negro et al. (2005) J Neurosci, 25, 446–53.

Figure 7

Microinjection of RIL into the preBötC does not perturb respiratory rhythm. Bilateral microinjection of muscimol is a positive control to ensure the correct placement of the pipettes in the preBötC. After recovery from muscimol, 10 µM RIL microinjection has no effect on XII motor amplitude or frequency for >30 min. Original data are displayed in the figure, but similar experiments, and a more complete set of pharmacological conditions, can be found in: Pace et al. (2007) J Physiol, 580, 485–96.

Figure 8

Microinjection of RIL into the midline Raphé region stops respiratory rhythm in vitro. Plotted data have been modified with author permission from Pace et al. (2007) J Physiol, 580, 485–96.

Figure 9

Intrinsic properties of NK1R⁺ preBötC neurons. Inspiratory drive latency is plotted versus whole-cell capacitance (C_M) in the lower graph. Histogram (top) shows the frequency of NK1R⁺ neurons with C_M separated into 10-pF bins. Note that small (i.e., low C_M) NK1R⁺ neurons predominate in the preBötC. A fast sweep of a typical inspiratory burst is shown at right with a diagram of characteristic measures (inspiratory drive latency and drive amplitude). Depolarization block of intraburst spike frequency is also indicated, which may have functional significance (see Fig. 16). Some data have been modified with permission from Hayes and Del Negro (2007) J Neurophysiol, 97, 4215–24.

Figure 10

Role of I_CAN in generating inspiratory bursts: chelation of intracellular Ca²⁺ by BAPTA attenuates inspiratory drive potentials. Patch-electrode filling solution contained 30 mM K₄-BAPTA as well as Lucifer yellow. The perforated-patch configuration (left cartoon and photo) is confirmed by the failure of the Lucifer yellow in the patch-pipette solution to dialyze the neuron. The whole-cell configuration (right cartoon and photo) allows Lucifer yellow to fill the neuron. Control conditions in perforated patch are shown at 5 and 35 min. BAPTA is introduced into the cytosol via patch rupture and causes a progressive attenuation of the inspiratory burst. Subsequent bath-application of 100 µM FFA has no additional attenuating effects even after 15 min of exposure to the drug. Baseline membrane potential was −60 mV throughout the experiment. Dotted line is drawn to facilitate comparison of the size of the inspiratory drive potential at different time points during the experiment. Data have been modified with author permission from Pace et al. (2007) J Physiol, 582, 113–25.

Figure 11

Role of I_CAN in preBötC neurons: Ca²⁺ uncaging by ultraviolet illumination of the photolyzable chelator DM-Nitrophen. I_CAN was measured at a range of membrane potentials in voltage clamp. I_CAN reversed at ~0 mV, consistent with a mixed monovalent cation current. Similar data were obtained in current clamp (not shown).

Figure 12

A high dose of FFA (300 µM) blocks inspiratory activity in 70% of slices (top traces). 350-µM FFA blocks the respiratory rhythm in the other 30% (bottom traces). 1-µM substance P (SP) does not restart the rhythm in 350 µM FFA, but rhythmic activity does recover after > 1 h of washout. Baseline membrane potential was −60 mV for traces in the top. Data have been modified with author permission from Pace et al. (2007) J Physiol, 582, 113–25.

Figure 13

Evidence for TRPM4 and TRPM5 channels in the preBötC. A) The mRNA coding for TRPM4 and TRPM5 is expressed in the preBötC region. Total RNA was extracted from preBötC and positive control tissues, and then reverse transcribed. Amplified products of the expected sizes were obtained for TRPM4 (301 bp), TRPM5 (483 bp) and GADPH (452 bp). Negative control reactions were performed without reverse transcriptase and amplified nothing. B) Immunohistochemistry from adult rats. Top panels (positive controls) show TRPM4 and TRPM5 in the compact nucleus ambiguous (cNA) taken from the ventral medulla immediately rostral to preBötC. Lower panels show TRPM4 and TRPM5 labeling in the preBötC at increasing magnification. Data in A were re-plotted with author permission from Crowder et al. (2007) J Physiol, 582, 1047–58.

Figure 14

AMPA receptor-mediated depolarization triggers I_CAN activation via Ca²⁺ channels. A) FFA (300 µM) significantly attenuates the AMPA receptor depolarization. B) N,N,N,-Trimethyl-5-[(tricyclo[3.3.1.13,7]dec-1-ylmethyl)amino]-1-pentanaminiumbromide hydrobromide (IEM-1460) (100 µM), a Ca2+- permeable AMPA receptor antagonist, does not perturb the AMPA receptor depolarization. C) Cd²⁺ attenuates the AMPA receptor depolarization (200 µM). Plotted data have been modified with author permission from Pace et al. (2008) Eur J Neurosci, 28, 2434–42.

Figure 15

Role of mGluR5 in inspiratory burst generation. A) Bath application of the selective mGluR5 antagonist MPEP (10 µM) attenuates inspiratory bursts in the context of ongoing network rhythm. B) Downstream of mGluR5, IP₃ receptors cause intracellular Ca²⁺ release and lead to I_CAN activation. Perforated-patch recordings serve as control. The IP₃ receptor antagonist xestospongin (1 µM), applied intracellularly by patch rupture, reduces inspiratory bursts. Subsequent application of MPEP causes no further change in the drive potential. Data have been modified with author permission from Pace et al. (2007) J Physiol, 582, 113–25.

Figure 16

The group-pacemaker hypothesis realized in an explicit mathematical model. A) On-cell patch recording in the preBötC with inspiratory motor discharge (XII). Below, whole-cell recording of the same cell as above after rupture of the patch, also with XII output. Baseline membrane potential was −60 mV. B) Data from another preBötC neuron in whole cell recording is plotted with the model behavior. C) Fast sweep of a burst in the model from B (above) coded with numerals to illustrate the sequence of physiological steps in one model burst in the Na⁺-Ca²⁺ phase plane, which is also displayed. Specific captions are superimposed with the circle-enclosed numbers to encapsulate model dynamics in sequence. See text for more detail. These data were modified with permission from Rubin et al. (2009) Proc Natl Acad Sci USA, 106, 2939–44