Midline section of the medulla abolishes inspiratory activity and desynchronizes pre-inspiratory neuron rhythm on both sides of the medulla in newborn rats

Abstract

Each half of the medulla contains respiratory neurons that constitute two generators that control respiratory rhythm. One generator consists of the inspiratory neurons in the pre-Bötzinger complex (preBötC); the other, the pre-inspiratory (Pre-I) neurons in the parafacial respiratory group (pFRG), rostral to the preBötC. We investigated the contribution of the commissural fibers, connecting the respiratory rhythm generators located on the opposite side of the medulla to the generation of respiratory activity in brain stem-spinal cord preparation from 0- to 1-day-old rats. Pre-I neuron activity and the facial nerve and/or first lumbar (L1) root activity were recorded as indicators of the pFRG-driven rhythm. Fourth cervical ventral root (C4) root and/or hypoglossal (XII) nerve activity were recorded as indicators of preBötC-driven inspiratory activity. We found that a midline section that interrupted crossed fibers rostral to the obex irreversibly eliminated C4 and XII root activity, whereas the Pre-I neurons, facial nerve, and L1 roots remained rhythmically active. The facial and contralateral L1 nerve activities were synchronous, whereas right and left facial (and right and left L1) nerves lost synchrony. Optical recordings demonstrated that pFRG-driven burst activity was preserved after a midline section, whereas the preBötC neurons were no longer rhythmic. We conclude that in newborn rats, crossed excitatory interactions (via commissural fibers) are necessary for the generation of inspiratory bursts but not for the generation of rhythmic Pre-I neuron activity.

the parafacial respiratory group (pFRG) around the ventrolateral part of the facial motor nucleus and the pre-Bötzinger complex (preBötC) located ventrolateral to the subcompact part of the ambiguous motor nucleus are the two generators capable of pacing respiratory rhythm. The rhythmogenic circuit of the pFRG is formed of the pre-inspiratory (Pre-I) neurons that fire before inspiration, coincidentally to active expiration (Onimaru and Homma 2003a). Pre-I neurons (via interneurons) provide expiratory drive to the first lumbar (L1) root, which supplies the abdominal muscles (Janczewski et al. 2002), and an expiratory-activity component to the upper-airway, mixed (inspiratory-expiratory) nerves, such as the facial nerves (Onimaru et al. 2006). Most of the ventral pFRG neurons express a paired-like homeobox 2b transcription factor and are CO₂ sensitive (Onimaru et al. 2008; Thoby-Brisson et al. 2009). The rhythmogenic circuit of the preBötC is formed of inspiratory, glutamatergic interneurons, derived from precursors expressing the homeobox gene Dbx1 in the developing brain. The axons of most (>50%) preBötC Dbx1 neurons project to the contralateral preBötC (Bouvier et al. 2010). The preBötC neurons, via premotor neurons, supply the diaphragm and provide an inspiratory component of upper-airway nerve activity (Dobbins and Feldman 1994, 1995; Koizumi et al. 2013; Koshiya et al. 2014). Both generators interact, and each one can initiate the respiratory cycle (Janczewski and Feldman 2006; Janczewski et al. 2002; Mellen et al. 2003).

Neurons forming the pFRG and preBötC generator in each half of the medulla are connected by commissural fibers (Bouvier et al. 2010; Kashiwagi et al. 1993; Koshiya and Smith 1999). Crossed fibers multiply the number of excitatory connections and double the number of neurons involved in burst formation. A midline section through the medulla severs all crossed connections, and depending on the species, age, and experimental conditions, either eliminates inspiratory bursts or results in desynchronized bursting (see discussion). In neonatal rats, preBötC-driven inspiratory nerve activity ceases after a midline section of the medulla (Janczewski and Aoki 1997, 1998; McLean and Remmers 1994; Peever and Duffin 2001). Here, we investigated the significance of the crossed connections to Pre-I neuron-driven activity and re-examined their role in the generation of inspiratory bursts. We severed crossed fibers rostral to the obex by making a midline sagittal section and examined the effects on Pre-I-like facial (or L1) nerve activity and inspiratory spinal or hypoglossal (XII) activity. Our results suggest that in newborn rats, bilateral interactions (via commissural fibers) are necessary for the generation of inspiratory bursts but not for the generation of rhythmic Pre-I neuron activity. Some of these findings have been presented in an abstract form (Janczewski et al. 1999; Onimaru and Homma 2003b).

METHODS

Preparations.

Brain stem-spinal cords from 0- to 1-day-old rats (n = 44) were isolated under deep ether anesthesia, as described previously (Onimaru and Homma 2003a; Suzue 1984). Experimental protocols were approved by the Animal Research Committee of Showa University, which operates in accordance with Law No. 105 of the Japanese government for the care and use of laboratory animals. To monitor facial nerve activity, the right half or both sides of the pons were retained. Inspiratory activity corresponding to phrenic nerve activity was monitored from the fourth cervical ventral root (C4). Nerve activities from the XII and L1 roots were also recorded in some experiments. Preparations were superfused continuously at 2.5–3.0 ml/min in a 2-ml chamber with the following standard solution (mM): NaCl, 124; KCl, 5.0; KH₂PO₄, 1.2; CaCl₂, 2.4; MgCl₂, 1.3; NaHCO₃, 26; and glucose, 30, equilibrated with 95% O₂ and 5% CO₂ at 25–26°C, pH 7.4 (Suzue 1984). [d-Ala2, N-Me-Phe4, Gly5-ol]-Enkephalin (DAMGO; Sigma-Aldrich, St. Louis, MO), as a μ-opiate agonist, was dissolved in the standard solution and applied by superfusion for 10–15 min. The medulla was incised with a stainless-steel blade (0.5 mm wide) attached to a manipulator, while nerve activity was monitored. We used the level of the caudal end of the facial nucleus as the zero level (see Fig. 1) to describe the extent of a section (Ruangkittisakul et al. 2008). The preBötC was reported to be ∼0.3 mm long in its rostrocaudal dimension (Smith et al. 1991) and centered at −0.5 mm; therefore, we placed the preBötC from −0.35 to −0.65 mm in our reference system (Ruangkittisakul et al. 2008). The pFRG extended from −0.2 to +1 mm (Fig. 1). Our incisions started at −1 mm (i.e., caudal to the preBötC) and either ended at −0.3 mm, just rostral to the preBötC, or continued through the pFRG to +1.2 mm (i.e., rostral to the pFRG). We referred to the shorter incision as the “preBötC incision” and the longer one, “preBötC + pFRG incision.”

Fig. 1.

Facial (VII), 4th cervical ventral root (C4), and 1st lumbar (L1) nerve activities after midsagittal section of the medulla. A: respiratory nerve activities from the VII, C4, and L1 nerves before midsagittal section in pons-medulla-spinal cord preparations that were used to record right (r) and left (l) VII and L1 nerve activities. B: respiratory nerve activity after application of 1 μM [d-Ala2, N-Me-Phe4, Gly5-ol]-enkephalin (DAMGO; 10 min). Note the increase in burst rate. C: activity after midsagittal section to the rostral medulla (at the level of the VII nucleus) from the caudal medulla [at the level of the pre-Bötzinger complex (preBötC)]. Rhythmic VII nerve activities (but not C4 and L1 activity) were preserved. D: r-L1 and l-L1 (but not C4) activities recovered at 15 min after sectioning. Voltage calibration bars denote 0.1 mV in A and B; 0.05 mV in C and D. A' and B': averaged burst activity of A (n = 7 cycles) and B (n = 10 cycles). Periods between vertical, dotted lines denote the inspiratory phase. Note the pre-inspiratory (Pre-I) and postinspiratory activity in VII and L1 nerve records. Inset: approximate level of midsagittal section [red line, preBötC + parafacial respiratory group (pFRG) incision]. AICA, anterior inferior cerebellar artery; FN, facial nucleus; IX, X, XII, cranial nerves.

Nerve activities were recorded via glass capillary suction electrodes through a high-pass filter with a 0.3-s time constant. The burst rate of respiratory-related activity (bursts/min) was calculated from the mean C4 (or facial/L1) burst activity for 3–5 min. Values are shown as means ± SD. The significance of differences (P < 0.05) was determined by Student's t-test for paired samples.

Whole-cell recordings.

Membrane potentials of putative Pre-I neurons in the rostral ventrolateral medulla, just caudal (within 200 μm) to the caudal end of the facial nucleus and overlapping the caudal part of the pFRG (Onimaru and Homma 2003a), were recorded by a blind, whole-cell patch-clamp method (Onimaru and Homma 1992). The electrodes, which had an inner-tip diameter of 1.2–2.0 μm and a resistance of 4–8 MΩ, were filled with the following pipette solution (mM): K-gluconate, 130; EGTA, 10; HEPES, 10; Na₂-ATP, 2; CaCl₂, 1; and MgCl₂, 1, adjusted with KOH, pH 7.2–7.3. For histologic analysis of the location of recorded cells, the patch electrode tips were filled with 0.5% Lucifer yellow (LY; lithium salt; Sigma-Aldrich). Membrane potentials were recorded with a single-electrode voltage-clamp amplifier (CEZ-3100; Nihon Kohden, Tokyo, Japan) after compensation for series resistance (20–50 MΩ) and capacitance.

Optical recordings.

To record facial nerve activity as the triggering signal for optical recordings, this experiment was performed in preparation, in which the right half of the pons was retained. The detailed procedure for making optical recordings using voltage-sensitive dye has been described previously (Onimaru and Homma 2003a). In brief, after midsagittal sectioning of the medulla, brain stem-spinal cord preparations (n = 3) were incubated in modified Krebs solution (described above) containing a fluorescent voltage-sensitive dye (50 μg/ml Di-2-ANEPEQ; Molecular Probes, Eugene, OR) for 40 min. For control imaging, the same type of preparation was used without midsagittal sectioning (n = 3). After staining, the preparation was placed with the ventral surface facing up in a perfusion chamber mounted on an upright fluorescence microscope stage. Neuronal activity in the preparation was detected as changes in the fluorescence of the voltage-sensitive dye by means of an optical recording apparatus (MiCAM01; BrainVision, Tokyo, Japan). Recordings were made with an acquisition time (i.e., sampling clock) of 20 ms. The optical imaging data were averaged using facial nerve activity as the trigger signal. Fluorescence signals for 13.6 s/trial, including 6.7 s before the initiation of the facial nerve burst, were averaged for 40 trials.

Histology.

For histologic verification of the incision level and location of recorded neurons, preparations were fixed for >48 h at 4°C in Lillie solution (10% formalin in phosphate buffer, pH 7.0). Transverse, 100 μm sections were then cut with a laboratory-made, vibrating-blade tissue slicer and stained with neutral red. LY-labeled neurons were reconstructed with the aid of a camera lucida attached to a fluorescence microscope (BH2; Olympus, Tokyo, Japan).

RESULTS

Desynchronization of right and left facial and L1 nerve activity after midsagittal sectioning.

To record right and left facial nerve activity, both sides of pons-attached preparations were used in this experiment (n = 8). In five of these eight preparations, the lumbar cord was also retained to record L1 activity. The burst rate of C4 and facial nerves of pons-attached preparations (2.0 ± 1.1 bursts/min; n = 8) was significantly lower than that of medulla (or half pons-attached medulla)-spinal cord preparations (typically five to six bursts/min under the present experimental conditions; see below), due to the effects of the remaining bilateral pons, which have been suggested to send inhibitory synaptic inputs to the medullary respiratory neurons (Hilaire et al. 1989). In the present experiment, we applied DAMGO (0.5–1 μM, 15–20 min) before midline sectioning to facilitate the C4 respiratory rate (Tanabe et al. 2005). With DAMGO treatment, the burst rates of C4 and facial nerves increased to 8.0 ± 1.1 bursts/min before sectioning. After a midsagittal incision of the medulla encompassing the rostrocaudal extent of the preBötC and pFRG (Fig. 1), C4 inspiratory activity was silenced, and L1 activity diminished significantly. The burst rates of facial nerves after sectioning were 11.3 ± 4.5 bursts/min (left) and 10.6 ± 3.7 bursts/min (right). Rhythmic L1 nerve activity reappeared within 10 min of sectioning. Simultaneous recordings from the facial and L1 nerves indicated the independent rhythm of left and right facial and L1 nerve activities, in which the facial nerve activity was synchronized with contralateral L1 activity (Fig. 2).

Fig. 2.

Synchronization of contralateral VII and L1 nerve activities after sectioning. The traces are a faster sweep representation from Fig. 1D. Note the independent rhythm of r-VII and l-VII and r-L1 and l-L1 nerve activities. VII nerve activity was synchronized with the contralateral L1 activity.

In the above experiments, we used DAMGO pretreatment. Although pretreatment (or post-treatment) with DAMGO was useful in such sectioning experiments for stabilizing rhythmic, Pre-I-like activity in the facial (or L1) nerves, this treatment is not always necessary (Onimaru et al. 2006). We also examined the effects of midline section on C4 and L1 in the medulla-lumber cord preparation without pons (n = 6). After sectioning at the same level as the above experiments (preBötC + pFRG incision), C4 activity disappeared, but rhythmic L1 nerve activity recovered spontaneously at 30–40 min after midsagittal section without DAMGO treatment (Fig. 3). At 40 min after sectioning, the L1 nerve burst rate was 6.4 ± 1.4 bursts/min (n = 6) compared with 5.8 ± 1.2 bursts/min in control (not significant).

Fig. 3.

Spontaneous recovery of rhythmic L1 nerve activity. C4 and L1 nerve activities were recorded in medulla-lumbar cord preparation without pons. After midsagittal section at the level from preBötC to rostral pFRG (preBötC + pFRG incision), rhythmic L1 nerve (but not C4 nerve) activity recovered at 30 min without DAMGO treatment.

In five experiments, XII nerve activity was recorded simultaneously with the facial and C4 nerve activity in the half pons-attached preparation. After a midsagittal incision of the medulla encompassing the rostrocaudal extent of the preBötC, XII nerve and C4 activity disappeared, whereas the facial nerve burst continued (Fig. 4).

Fig. 4.

Effects of midsagittal section on VII, hypoglossal (XII), and C4 activity. Right VII, XII, and C4 nerve activities were recorded in half pons-medulla-spinal cord preparation. A: nerve activities before sectioning. Right: averaged burst pattern (n = 5 cycles) from each nerve activity. Periods between vertical dotted lines denote inspiratory phase. B: after midsagittal section at the level of preBötC (0.7 mm length), XII and C4 activity disappeared, but rhythmic VII nerve activity was preserved (without DAMGO treatment). Inset: approximate level of midsagittal section (red line, preBötC incision).

In three preparations with both sides of the pons, we examined the effects of midsagittal section at the preBötC level (the same as the experiments of Fig. 4) on synchronization of right and left facial nerve activity. This section resulted in the disappearance of C4 activity (see Fig. 4), but right and left facial nerve activities remained synchronous (data not shown).

Optical recordings of rhythmic activity after midsagittal sectioning.

In the above experiments, C4 inspiratory activity disappeared, but rhythmic facial nerve activity could be recorded after midsagittal sectioning. Optical recordings with voltage-sensitive dye can reveal active regions correlated to respiratory nerve outputs using nerve activity as a triggering signal. To detect respiratory-related neuron activity optically in the ventral medulla after midline sectioning (preBötC + pFRG), we used facial nerve activity as a triggering signal. Figure 5A shows a control image in which facial nerve activity was used as a triggering signal in the half pons-attached preparation. Several active regions were clearly detected in the ventral medulla in the pFRG expressing Pre-I and inspiratory neuron activity and preBötC and upper cervical cord expressing mainly inspiratory neuron activity. In contrast, after midline sectioning, activity was located predominantly in the rostral ventrolateral medulla, ipsilateral to the recorded facial nerve, overlapping the area containing pFRG/Pre-I neurons (Fig. 5B) (Onimaru and Homma 2003a). Similar results were obtained in three experiments.

Fig. 5.

Neuronal activity detected in the pFRG by optical recording. Results are the average of 40 respiratory cycles in which right VII nerve activity was used for triggering signal in half pons-attached preparation. A: a control image before sectioning. Several active regions were clearly detected in the ventral medulla in pFRG expressing Pre-I and inspiratory activity and preBötC and upper cervical cord (C1) mainly expressing inspiratory activity. B: optical recording after midsagittal section (red line, preBötC + pFRG incision) of the medulla from another preparation. Left: summated optical neuronal activity during 1 s from the onset of the VII nerve burst (corresponding to the period between vertical bars at right), superimposed on the ventral surface of the medulla. Right: time course of fluorescence changes at the spot indicated at left. Fluorescence decrease (i.e., depolarization) is upward. The activity was located predominantly in the rostral ventrolateral medulla ipsilateral with the VII nerve, overlapping the area containing pFRG/Pre-I neurons.

Intracellular recordings of putative pre-I neurons after midsagittal sectioning.

As suggested by the above optical recordings, facial nerve activity after midsagittal sectioning is presumed to be derived from Pre-I neurons in the pFRG. The firings of Pre-I neurons under control conditions were characterized by a biphasic burst pattern with inspiratory-phase inhibition, as shown in Fig. 6A' (Onimaru and Homma 1992). After midsagittal sectioning, similar to transverse sectioning between the preBötC and pFRG (Onimaru et al. 2006), Pre-I neurons were presumed to lose their biphasic burst pattern due to the absence of synaptic inhibition from inspiratory neurons (see discussion). After a midsagittal preBötC + pFRG incision, we recorded membrane potentials from putative Pre-I neurons (n = 4) in the caudal pFRG at the level of 0 to −0.2 mm in our reference system (see methods). They showed burst activity synchronized with ipsilateral facial nerve activity of half pons-attached preparations. A representative example is shown in Fig. 6. The locations of recorded neurons stained with LY are shown in Fig. 6B. The membrane potential was −48.5 ± 1.7 mV, and input resistance was 382 ± 92.5 MΩ.

Fig. 6.

Burst-generating neuron in the rostral medulla after midsagittal section of the medulla. A: the burst activity was synchronized with the ipsilateral VII nerve activity, indicating a putative Pre-I neuron. Note the singular “throughout” burst pattern, due to lack of inspiratory inhibition. Vm, membrane potential. A': a typical example of the Pre-I neuron burst pattern recorded in another preparation. Note the biphasic burst pattern in which action-potential discharges appear during the pre- and postinspiratory phases due to inspiratory inhibition. Inset: approximate level of midsagittal section from preBötC to rostral pFRG (preBötC + pFRG incision). B: camera lucida drawing of lucifer yellow-stained cell recorded in A. The locations correspond to the caudal part of the pFRG. CST, corticospinal tract; IO, inferior olivary nucleus; RFN, retrofacial nucleus; STN, spinal trigeminal nucleus.

DISCUSSION

We found that pFRG rhythm generation was preserved after a midsagittal section encompassing the entire rostrocaudal extent of both the preBötC and pFRG, whereas bilateral synchrony was lost; bursts of Pre-I neuron activity were synchronized with the motor activity of the ipsilateral facial nerve and contralateral L1 root. We conclude that the Pre-I neurons on either side of the medulla are normally synchronized via excitatory commissural axons (Kashiwagi et al. 1993; Thoby-Brisson et al. 2009) and that the Pre-I neuron activity is transmitted to the ipsilateral facial and contralateral L1 motoneurons (Janczewski et al. 2002; Onimaru et al. 2006). An agonist of μ-opiate receptors (DAMGO) facilitated the maintenance (or reappearance) of the pFRG-driven rhythm that could otherwise be transiently depressed by the sections. DAMGO was also useful for stabilizing pFRG activity after transverse sections between preBötC and pFRG (Onimaru et al. 2006). The facial or L1 nerve burst rate tended to be high after midline sectioning, possibly due to postsectioning excitation. We presume that DAMGO inhibits μ-opiate-sensitive, excessive excitatory effects that were produced by the sectioning and might disturb burst generation of Pre-I neurons.

In contrast to the motor output driven by the Pre-I neurons, the inspiratory motor activity of the XII and C4 root ceased after midsagittal sectioning, consistent with in vivo and in vitro studies on rats of the same age (McLean and Remmers 1994; Peever and Duffin 2001). One possibility is that preBötC remained rhythmically active, whereas inspiratory bursts of the C4 and XII roots stopped because the contralateral projections to the motoneurons were severed by the midsagittal section (Peever and Duffin 2001). The other possibility is that inspiratory rhythm was eliminated at its source. In neonatal and adult rats, the majority of excitatory preBötC neurons has axon collateral projections to the ipsilateral XII premotoneurons (Dobbins and Feldman 1995; Koizumi et al. 2013; Koshiya et al. 2014; Li et al. 2003); therefore, the interruption of the pathway between the source of the rhythm and the motoneurons cannot be the sole reason for the cessation of XII nerve activity. In adult rats, the descending pathway between the preBötC and C4 phrenic motoneurons is 50% ipsilateral (Dobbins and Feldman 1994); however, this proportion may be different in newborn rats (Peever and Duffin 2001). We optically recorded activity of preBötC neurons and demonstrated that they are silenced by a midsagittal section (compare Fig. 5, A with B). We have previously demonstrated that Pre-I neuron activity is typically biphasic, with one burst before and one burst after inspiration (Janczewski et al. 2002; Onimaru et al. 2006). Inactivation of the inspiratory neurons that normally inhibit the Pre-I neurons is necessary to convert a biphasic burst into a single “throughout” burst (Janczewski et al. 2002; Onimaru et al. 2006). All neurons had a throughout pattern when we surveyed the pFRG after a midline section (see Fig. 6A). These observations indicated that in the newborn rat, commissural connections were necessary for generation of preBötC inspiratory bursts. However, under conditions in which the excitability of neurons was increased by elevated, extracellular K⁺ concentration, the preBötC could produce inspiratory bursts in the unilateral medulla (Johnson et al. 2001). Asynchronous, left-right inspiratory preBötC bursts are observed when preBötC axons that normally form connections with the contralateral preBötC are rerouted to the ipsilateral preBötC in mutant mice (Bouvier et al. 2010).

Species differences, age, and the level of excitation are critical for the outcome of midsagittal sectioning in vivo. Janczewski and Aoki (1997, 1998) examined the effects of midsagittal sectioning on inspiratory activity of spontaneously breathing, halothane-anesthetized, neonatal (2–9 days old) and older (16–22 days old) rats. They reported that midsagittal sectioning irreversibly eliminated inspiratory activity in rats younger than postnatal day (P)9 and resulted in left-right diaphragm desynchronization in rats older than P16 under identical experimental conditions. We propose that to produce a burst, preBötC neurons require more recurrent, excitatory connections in neonates than they do in adults. An alternative explanation is that there are age-dependent differences in the transmission of respiratory activity from the medullary generator to motoneurons, with contralateral projections being predominant in neonates and ipsilateral projections developing later (Li et al. 2003; Peever and Duffin 2001; Peever et al. 1998).

Comparative in vivo studies on adult monkeys, cats, rabbits, and rats (species with a different proportion of crossed respiratory pathways) demonstrated that apart from the obvious consequences of interrupting pathways between the source of the rhythm and motor output, special experimental conditions (hypercapnia, hypoxia, a low level of anesthesia or decerebration, electrical stimulation, respiratory-stimulant drugs) were required for desynchronized rhythm generation (Eldridge and Paydarfar 1989; Gromysz and Karczewski 1981, 1982; Janczewski and Aoki 1997, 1998; Janczewski and Karczewski 1984, 1990; Kubin et al. 1987; Peever et al. 1998). In rabbits, rats, and cats, an increase in artificial ventilation or in the level of anesthesia and other changes in experimental conditions that were inconsequential before the section resulted in the cessation of rhythm after the section (Gromysz and Karczewski 1981, 1982; Janczewski and Aoki 1997, 1998; Janczewski and Karczewski 1984, 1990). These observations indicate that to sustain rhythm generation, external drives need to compensate for the loss of the crossed excitatory connections.

The only previous study on the desynchronization of the pFRG was performed in embryonic day 14.5 mouse embryos (Thoby-Brisson et al. 2009). Independent rhythms were observed in the left and right embryonic pFRG after the blocking of glutamatergic transmission, whereas a cocktail of bicuculline and strychnine was not effective. Desynchronization was also observed in a transverse slice containing embryonic pFRG, suggesting that commissural axons, coupling both pFRGs, cross the midline at the level caudal to the pFRG (Thoby-Brisson et al. 2009). The present study is the first report on the left-right desynchronization of the expiratory activity. A comparable study in adults may be difficult to perform, because the pFRG-driven, active expiration in rats is variable and requires high respiratory drive or special experimental conditions (Abdala et al. 2009; de Almeida et al. 2010). Indeed, studies in adult cat in vivo preparation (Bainton and Kirkwood 1979; Sears 1977) demonstrated that rhythmic discharges of expiratory bulbospinal units in the medulla were observed to change into a tonic firing pattern. Active expiration was induced effectively by unilateral optogenetic stimulation of the pFRG and a neighboring retrotrapezoid nucleus (Abbott et al. 2011; Pagliardini et al. 2011), but interactions between pFRGs on the stimulated and contralateral side were not investigated.

Although the unilateral Pre-I neuron network in the pFRG can produce rhythmic activity, the left and right pFRGs are synchronized under normal conditions. Axonal connections between left and right Pre-I neurons contribute to the synchronization (Kashiwagi et al. 1993). In some experiments, the section did not extend through the pFRG but stopped just rostral to the preBötC (see Fig. 4). Such shorter sections eliminated C4 activity, but facial nerve activities on the left and right side remained synchronous. After these shorter sections, neurons in the pFRG lost their biphasic characteristics, indicating that they no longer received rhythmic inhibition from the preBötC inspiratory neurons. The left-right synchrony of the pFRGs indicates that rhythmic-inhibitory synaptic inputs from inspiratory neurons to Pre-I neurons are not critical for bilateral synchronization of the facial nerve motor activity. In Fig. 7, we summarize the interaction between Pre-I and inspiratory rhythm-generating networks before and after midsagittal section, with consideration to previous and present findings.

Fig. 7.

A hypothetical model of neuronal connections between medullary rhythm generators in brain stem-spinal cord preparation. Left: before midline sectioning. Pre-I rhythm generators in the pFRG synchronize by mutual connections, and inspiratory (Insp) rhythm generators in the preBötC also synchronize by mutual connections between right and left medulla (Bouvier et al. 2010; Funk et al. 1993; Kashiwagi et al. 1993; Li et al. 2003; Thoby-Brisson et al. 2009). Inspiratory neuron activity in the medulla is transmitted via mainly contralateral projections to motoneuron pools in the neonate (Peever and Duffin 2001). Outputs from Pre-I rhythm generators are sent to the ipsilateral VII nerve and contralateral L1 nerve (Janczewski et al. 2002; Onimaru et al. 2006), as well as to the inspiratory rhythm generator to trigger inspiratory burst activity (Ballanyi et al. 2009; Mellen et al. 2003; Onimaru and Homma 2003a). Right: after midline sectioning from preBötC to rostral pFRG. Inspiratory rhythm disappeared after midline sectioning. Pre-I rhythm generators produce independent rhythm between the right and left medulla. This shows the simplest model, whereas the networks may function differently depending on the level of the sectioning.

In conclusion, rhythmic Pre-I neuron activity can be produced without crossed interaction (via commissural fibers), whereas crossed interaction is more important in the generation of inspiratory bursts in newborn rat preparation.

GRANTS

Support for this study was provided by Grants-in-Aid for Scientific Research from the Ministry of Education, Science and Culture of Japan. Research performed by W. A. Janczewski in Japan was supported by the Japanese Society for the Promotion of Sciences.

DISCLOSURES

The authors declare no competing financial interests.

AUTHOR CONTRIBUTIONS

Author contributions: H.O. and W.A.J. conception and design of research; H.O., K.T., and W.A.J. performed experiments; H.O., K.T., Y.N., and W.A.J. analyzed data; H.O., Y.N., and W.A.J. interpreted results of experiments; H.O., K.T., and W.A.J. prepared figures; H.O., K.T., Y.N., and W.A.J. drafted manuscript; H.O. and W.A.J. edited and revised manuscript; H.O., K.T., Y.N., and W.A.J. approved final version of manuscript.