The Kölliker-Fuse nucleus gates the postinspiratory phase of the respiratory cycle to control inspiratory off-switch and upper airway resistance in rat

The Kolliker-Fuse nucleus gates the postinspiratory phaseof the respiratory cycle to control inspiratory off-switch andupper airway resistance in rat

Mathias Dutschmann1 and Horst Herbert21Department of Neuro and Sensory Physiology, Georg August University of Gottingen, Humboldtallee 23, 37073 Gottingen,Germany2Graduate School of Neural & Behavioural Sciences, IMPRS, University of Tubingen, Osterbergstr. 3, 72074 Tubingen, Germany

Keywords: apneusis, eupnoea, larynx, pneumotaxic centre, upper airway, vocalization

Abstract

Lesion or pharmacological manipulation of the dorsolateral pons can transform the breathing pattern to apneusis (pathologicalprolonged inspiration). Apneusis reflects a disturbed inspiratory off-switch mechanism (IOS) leading to a delayed phase transitionfrom inspiration to expiration. Under intact conditions the IOS is irreversibly mediated via activation of postinspiratory (PI) neuronswithin the respiratory network. In parallel, populations of laryngeal premotoneurons manifest the IOS by a brief glottal constrictionduring the PI phase. We investigated effects of pontine excitation (glutamate injection) or temporary lesion after injection of a GABA-receptor agonist (isoguvacine) on the strength of PI-pool activity determined from respiratory motor outputs or kinesiologicalmeasurements of laryngeal resistance in a perfused brainstem preparation. Glutamate microinjections into distinct parts of thepontine Kolliker-Fuse nucleus (KF) evoked a tonic excitation of PI-motor activity or sustained laryngeal constriction accompanied byprolongation of the expiratory phase. Subsequent isoguvacine microinjections at the same loci abolished PI-motor or laryngealconstrictor activity, triggered apneusis and established a variable and decreased breathing frequency. In summary, we revealed thatexcitation or inhibition of defined areas within the KF activated and blocked PI activity and, consequently, IOS. Therefore, weconclude, first, that descending KF inputs are essential to gate PI activity required for a proper pattern formation and phase controlwithin the respiratory network, at least during absence of pulmonary stretch receptor activity and, secondly, that the KF contains largenumbers of laryngeal PI premotor neurons that might have a key role in the regulation of upper airway resistance during reflex controland vocalization.

Introduction

Over the past 10 years, pontine aspects of the respiratory networks,compared with the rhythm-generating circuits in the medullaoblongata, have received less attention in studies concerned withthe neuronal control of breathing. Nevertheless, in the pons threemajor cell groups have been identified that are closely linked to thecontrol of breathing. These areas include – from caudal to rostral –the noradrenergic A5 cell group, the intertrigeminal region (ITR)and the parabrachial ⁄ Kolliker-Fuse complex (PB ⁄ KF; for a reviewsee Alheid et al., 2004). The A5 group was demonstrated to exertrespiratory-related activity (Guyenet et al., 1993) that might haveimportant functions in the modulation of respiratory frequency (seeHilaire et al., 2004). Other studies suggested a role of the A5 groupin the mediation of respiratory phase transitions (Jodkowski et al.,1997) or in the adaptation and plasticity of the breathing pattern inresponse to hypoxia and hypercapnia (Coles & Dick, 1996; Hsiehet al., 2004). The ITR is even less well defined regarding itsfunction, yet it was demonstrated that stimulation of cell bodies

within the ITR evokes transient cessation of respiratory activity(Chamberlin & Saper, 1998), which was attributed to a potentialrole for reflex control of respiration in response to airway stimuli(Chamberlin, 2004). The PB ⁄ KF was demonstrated to be a majorkernel for the adaptation of breathing activity in response tonociception (Jiang et al., 2004), to nasopharyngeal stimulation(Dutschmann & Herbert, 1998; Shannon et al., 2004), to chemo-reflex (Koshiya & Guyenet, 1994) and to Hering-Breuer reflex(Shaw et al., 1989; Ezure et al., 1991). Owing to the processing ofmultimodal sensory inputs related to respiration, we recentlysuggested that the PB ⁄ KF might be a command centre requiredfor the adaptation of the breathing pattern to changes in environ-ment and behaviour (Dutschmann et al., 2004; see also Song &Poon, 2004).Although a role of the PB ⁄ KF for the processing of sensory inputs

and therefore the secondary modulation of the respiratory rhythm orpattern is widely accepted, its role as an integral part of therespiratory network required for the generation of an eupnoeicbreathing pattern is still controversial (see St-John, 1998; St-John &Paton, 2004). Lumsden (1923) and several follow-up studiesdemonstrated that the dorsolateral pons including the PB ⁄ KFreflects a pneumotaxic centre, given that following ablation or

Correspondence: Dr M. Dutschmann, as above.E-mail: [email protected]

Received 1 March 2006, revised 11 May 2006, accepted 19 May 2006

European Journal of Neuroscience, Vol. 24, pp. 1071–1084, 2006 doi:10.1111/j.1460-9568.2006.04981.x

ª The Authors (2006). Journal Compilation ª Federation of European Neuroscience Societies and Blackwell Publishing Ltd

pharmacological manipulation of this area the eupnoeic breathingpattern is switched to apneusis (for a review see St-John & Paton,2004). Apneusis is seen as a disturbed inspiratory off-switch (IOS)mechanism that reflects a key element in the transition from theinspiratory to expiratory phase in all mammalian species. Never-theless, a fundamental role for the dorsolateral pons in respiratoryrhythm generation was questioned by several studies. First, the signalfor IOS can be provided by the activation of slowly adaptingpulmonary stretch receptors during end inspiration (Richter, 1982;von Euler, 1983) and lesion of the dorsolateral pons in vagi-intactand awake animals showed only modest changes in breathing pattern(Baker et al., 1981; Caille et al., 1981). These studies gave rise to astill common view that the pons provides a failsafe mechanism(Song & Poon, 2004) for the IOS and, indeed, apneusis after pontinelesions occurs only in vagotomized animals. Another influentialstudy by Segers et al. (1985) demonstrated that only a minority ofpontine respiratory-related units interacted mono-synaptically withthe medullary respiratory centres.Recent publications have reinforced a discussion on the import-

ance of pneumotaxic mechanisms for the generation of eupnoea(Rybak et al., 2004; St-John & Paton, 2004), although its role,particularly in rats, still remains unclear (compare Monteau et al.,1989; Morrison et al., 1994) Recent studies have shown respiratoryactivities in an area corresponding to the KF area in the neonatal en-bloc preparation of rat (Kobayashi et al., 2005). However, the role ofrespiratory activities remains unclear and might be associated withthe respiratory-related modulation of upper airway calibre (Laraet al., 2002).In the present study, we re-analysed pontine respiratory functions in

a perfused brainstem preparation with a particular emphasis on thecontrol of upper airway calibre. To do so, we recorded phrenic nerveactivity (PNA) and upper airway-related activities, such as recurrentlaryngeal nerve discharge or subglottal pressure changes, to analysethe response pattern to chemical stimulation and inhibition of neuronswithin the PB ⁄ KF.

Methods

Experiments were performed with juvenile rats (Wistar, postnatal days21–26, 70–100 g) of either sex. We used the intra-arterially perfusedbrainstem preparation as described previously (Paton, 1996). Allexperimental procedures were performed in accordance with Europeancommunity and NIH guidelines for the care and use of laboratoryanimals. The study was approved by the ethics committee of the GeorgAugust University Gottingen.Rats were deeply anaesthetized with isoflurane (1-chloro-2,2,2-

trifluoroethyl-difluoromethylether, Abbott, Wiesbaden, Germany).Once respiration was depressed severely and the animal failed torespond to noxious pinch of tail or toe, it was transected below thediaphragm. In chilled Ringer’s solution, gassed with 95% O2 and5% CO2 (carbogen), rats were decerebrated at the precollicular leveland cerebellectomized. After these initial procedures, the prepar-ation was transferred to a recording chamber (custom made). Thedescending aorta was cannulated and perfused using a peristalticpump (Watson & Marlow, Rommerskirchen, Germany) via adouble-lumen catheter with carbogen-gassed Ringer’s containingFicoll (1.25%) at 31 �C and at a flow rate of 28–32 mL ⁄ min. Theperfusate was filtered and passed through bubble traps (custommade) to remove gas bubbles and dampen both perfusion pump andcardiac-generated pulses. The perfusate leaking from the preparationwas collected and re-circulated after its re-oxygenation. Two to

5 min after the start of perfusion, rhythmic contractions of thediaphragm resumed. In some preparations, respiratory-relatedmovements were abolished by using vecuronium bromide(0.3 lg ⁄ mL).The perfusate contained (in mm): 125 NaCl, 24 NaHCO3, 2.5

CaCl2, 1.25 MgSO4, 4 KCl, 1.25 KH2PO4, 10 d-glucose and Ficoll(1.25%; Sigma, Taufkirchen, Germany) to maintain colloid osmoticpressure. The osmolarity of the ACSF was 298 ± 5 mosmol ⁄ L and ongassing with carbogen the pH was 7.35 ± 0.05.

Recording of cardiovascular and respiratory parameters

Perfusion pressure within the aorta was monitored via one port of adouble-lumen catheter. This lumen was connected to a pressuretransducer and the pressure was set between 70 and 90 mmHg byadjusting the flow rate. The transient lesions within the dorsolateralpons caused occasional changes in perfusion pressure, which could beinterpreted as changes in the resistance of the vascular system.However, these changes were small (1–4 mmHg change in perfusionpressure) and heterogeneous within the experimental groups and werenot further analysed.In all experiments the left phrenic nerve was cut at the level of the

diaphragm and the inspiratory discharge was recorded from its centralend using a glass suction electrode. Rhythmic phrenic nerve dischargepersisted for 5–6 h. In 14 experiments, we additionally recorded fromthe central end of the recurrent laryngeal nerve using a second glasssuction electrode. Peripheral nerve activity was amplified (Neurolog100), filtered (8 Hz to 3 kHz, Neurolog modules 104 and 125) andintegrated (time constant 100 ms).In ten experiments we measured respiratory-related changes in

glottal resistance using subglottal pressure recordings. Here, weperfused the larynx with a constant stream of warmed, humidifiedcarbogen gas in the expiratory direction via a cannula placed with itstip distal to the larynx and pointing towards the buccal cavity(Dutschmann & Paton, 2002a). Subglottal pressure was recorded froma side arm connected to this cannula. Increases and decreases insubglottal pressure were indicative of constriction and dilatation,respectively, thereby providing a direct index of the respiratorymodulation of glottal resistance. The exact rate of airflow differedbetween preparations (20–35 mL ⁄ min) and was preset to showinspiratory dilation and early expiratory constriction. In preparationswhere subglottal pressure was measured, no neuromuscular blocker(see above) was used.All signals were displayed on a computer using a MacLab 8s

interface and Chart software (ADInstruments, Sydney, Australia).Data analysis was performed off-line.

Experimental protocols for microinjection studies

Subsequent pressure microinjections of glutamate (10 mm, Sigma),the GABA receptor agonist isoguvacine (10 mm, Sigma) andpontamine sky blue (Sigma) into the dorsolateral pons were performedwith a micromanipulator-driven multibarreled micropipette (tip diam-eter 30–40 lm). The injected volumes were measured by observingthe movement of the meniscus through a binocular microscope fittedwith a calibrated eyepiece graticule. As a first step, we mapped thedorsolateral pons ipsilaterally for respiratory modulation in responsesto injection of the excitatory neurotransmitter glutamate (20–30 nL).Effective injection sites were found at the following coordinates: 0.2–0.5 mm caudal to the caudal end of the inferior colliculus and 1.8–2.5 mm lateral to midline, approximating to the rostral pontine

1072 M. Dutschmann and H. Herbert

ª The Authors (2006). Journal Compilation ª Federation of European Neuroscience Societies and Blackwell Publishing LtdEuropean Journal of Neuroscience, 24, 1071–1084

respiratory group region in this species. After identification ofmodulatory injection sites (e.g. transient apnoeas or tachypnoeas),we injected isoguvacine (40–50 nL) to inactivate restricted neuronalpopulations via GABA receptor-mediated hyperpolarization. After4–5 min recording time, the position of the pipette was markedby injection of 40–50 nL pontamine sky blue. The pipette was thenmoved to the same coordinates at the contralateral site to injectisoguvacine followed by pontamine sky blue. At the contralateralsite, we did not inject glutamate, as previous experiments revealedthat given the already disturbed breathing pattern caused by ispilat-eral isoguvacine injections, the glutamate effects were hardly inter-pretable.

At the end of the experiment the brainstem was removed and fixedfor 1–2 days in 4% paraformaldehyde containing 20% sucrose. Foranatomical verification of injection sites, we cut series of 50-lm-thickcoronal sections through the pons using a freezing microtome(Reichert and Jung). Subsequently, the sections were stained withneutral red. The locations of microinjections were documented onsemischematic drawings of coronal sections through the dorsolateralpons.

Data analysis

From PNA recordings, we analysed three basic respiratory parameters:the duration of the respiratory cycle length (Ttot), the duration of theinspiratory phase (Ti) and the time of PNA quiescence (Te) during therespiratory cycle. In rat PNA recordings, there is usually no clearlyidentifiable postinspiratory (PI) activity. Therefore, we classified theentire expiratory interval including the postinspiratory and lateexpiratory phase as Te. If simultaneous recording of PNA andrecurrent laryngeal nerve activity (RLNA) or subglottal pressure(SGP) were performed, we analysed the postinspiratory (Tpi) and lateexpiratory time (Te2). For RLNA recordings, postinspiratory activitywas quantified and identified via the decrementing discharge of therecurrent laryngeal nerve (RLN) following the inspiratory activity,which was identified from parallel recording of PNA. In otherexperiments, we identified postinspiratory activity from fluctuations ofSGP levels over the respiratory cycle. Thereby, inspiratory laryngealabductor and postinspiratory adductor activity were indicated by adecrease (abductor) or increase (adductor) in SGP. To quantifychanges in activity, we used the mean SGP level during late expiration200 ms before the onset of the laryngeal abductor activity as areference point. During the reference period, the slope of the SGPtraces was low and therefore was considered to reflec a quasi steady-state level. The mean SGP levels during inspiration (dilation) andpostinspiration (constriction) were then expressed as an increase ordecrease in relation to the reference measurement during lateexpiration (E2).

All values analysed from PNA, RLNA (Ttot, Ti, Tpi, Te2, and Te)and SGP (abductor and adductor activity) were taken before and afterisoguvacine injection and at recovery. All data were analysed over60 s prior and 2–3 min and 40–60 min after the injection for eachindividual respiratory cycle. Statistical tests were performed withrepeated anova followed by a Tukey HSD post-hoc test. Allstatistical tests were performed with Systat 8 software (SPSS Inc.,Chicago, IL, USA). Finally, a potential quantal slowing of PNAfrequency or quantal relation of Ti, Tpi or Te before and after KFblockade was analysed. Here, the data were normalized to the mediancontrol activity prior to isoguvacine injection. All data are expressedas mean ± SEM. Differences were taken as significant at the 95%confidence level.

Neuron recordings

In 15 experiments, respiratory neurons were recorded extra- andintracellularly from the dorsolateral pons using fine tipped glassmicroelectrodes. For intracellular recordings, the electrodes were filledwith 2 m NaCl or with potassium methyl sulphate (3 m) and the Ca2+

chelator BAPTA [5 mm, 1,2-bis(2-aminophenoxy)ethane-N,N,N¢,N¢-tetraacetic acid; Sigma], which assists neuron recovery after impale-ment. The resistance of the microelectrodes used varied from 35 to65 MW. Extracellular recordings were performed with electrodes filledwith 2 m NaCl (4–10 MW). Neuronal activity was amplified (NPI SEC10) and, via a MacLab interface, both displayed and stored on acomputer running the MacLab software Chart (version 5.0). Micro-electrodes were held in a three-dimensional micromanipulator anddriven into the dorsolateral pons via the dorsal medullary surface usinga stepper system (Burleigh inchworm). The step width ranged between5 and 10 lm. The surface of the pons was cleared of the pia mater toallow entry of microelectrodes into the tissue. Under the visualguidance of a binocular microscope, microelectrode tips werepositioned on the dorsal surface of the pons at the followingcoordinates: 0.2–0.5 mm caudal to the caudal end of the inferiorcolliculus and 1.8–2.5 mm lateral to midline, approximating to therostral pontine respiratory group region in this species. Respiratoryneurons were recorded at depths between 1 and 1.5 mm below thedorsal surface on the ipsi- and contralateral side to the recordedphrenic nerve. The subtypes of neurons were classified according totheir discharge pattern in relation to PNA.

Results

We analysed the physiological significance of the dorsolateral pons forrespiratory pattern formation and upper airway control in the ratperfused brainstem preparation. We used 40 juvenile rats (postnataldays 25–31) for microinjection studies with multibarreled micropi-pettes in order to excite (glutamate) or inactivate (isoguvacine) locallypontine neuronal populations. The coordinates for the microinjectionsor cellular recordings were aimed in particular at the KF, representingan integral part of the pontine respiratory group.

Excitation and suppression of PI activity following microinjectioninto intermediate parts of the KF

In 24 preparations, we simultaneously recorded PNA and RLNA(n ¼ 14) or SGP changes (n ¼ 10). Both RLNA and SGP served asindex for upper airway-related respiratory activity. The RLN displayeda characteristic discharge pattern consisting of an inspiratory compo-nent (in coincidence with PNA), followed by decreasing activityduring the postinspiratory period and a lack of discharge during thelate stages of the expiratory interval (Fig. 1A). Thus, the inspiratorymotor discharge is related to laryngeal abductor activity, while thepostinspiratory discharge drives laryngeal adductors. The alternatingactivity of laryngeal abductor and adductors during a respiratory cyclewas indirectly revealed from SGP recordings. Laryngeal abductoractivity was illustrated by a decreased SGP during inspiratory PNAand was always followed by a sharp increase of SGP immediately afterthe end of PNA. The latter is indicative of laryngeal abductor activity(see Fig. 1B).

Glutamate injections

In 14 experiments injection of 20–30 nL glutamate into the dorsolat-eral pons evoked a transient prolongation of the expiratory phase (Te),

The Kolliker-Fuse nucleus gates the postinspiratory phase 1073


from 2.14 ± 52 to 5.59 ± 1.1 s (P < 0.01), leading to an increase inthe total respiratory cycle length (Ttot) from 2.91 ± 0.51 to6.36 ± 0.93 s (P < 0.01). In eight preparations, the Te prolongationwas accompanied by an increase of integrated postinspiratory RLNAduring the expiratory interval (from 0.32 ± 0.06 to0.727 ± 0.10 VÆ100 ms, P < 0.01; see Fig. 2B, Table 1 and Fig. 3:injection sites 1–8). In six other preparations, the Te prolongation wasaccompanied by a massive and permanent increase in SGP, whichreached significantly higher levels compared with the postinspiratorypeak values recoreded during control cycles (2.31 ± 0.34 vs.3.12 ± 0.47 mmHg, P < 0.05; Fig. 3B). Both types of experimentssuggested that the increase in Te prolongation following glutamateinjection was predominantly mediated by a prolongation of thepostinspiratory phase leading to glottal constriction.

Isoguvacine injections

Subsequent ipsilateral injection of isoguvacine (40 nL) at the sameinjection sites altered the pattern of PNA (Figs 2C and 3C). In PNArecordings, we observed Ti prolongation from 0.77 ± 0.12 to1.76 ± 0.48 s (n ¼ 14; P < 0.01, Fig. 4A) and, consequently, anincrease in integrated activity from 165 ± 87 to 297 ±67 mVÆ100 ms (P < 0.05). The changes in PNA pattern wereaccompanied by increased Ttot from 2.91 ± 0.51 to 4.85 ± 0.92 s(n ¼ 14; P < 0.05). Contralateral injection of isoguvacine (40 nL) atthe same coordinates further increased Ti to 3.45 ± 1.1 s(P < 0.001), integrated PNA to 587 ± 187 mVÆ100 ms (P < 0.001)and Ttot to 13.68 ± 2.61 s (P < 0.001; Fig. 4A) leading to themanifestation of an apneustic breathing pattern (Figs 2D and 3D).Forty to 50 min after the bilateral injections, PNA recovered: Ti andintegrated PNA returned to 1.12 ± 101 s and 231 ± 121 mVÆ100 ms(both P < 0.05), while Ttot returned to 3.45 ± 2.34 s (P < 0.01;Figs 2E, 3E and 4A).

Effect of isoguvacine on RLNA

In addition to changes in the inspiratory pattern, the simultaneousRLNA recordings (n ¼ 8) revealed changes in the two phases ofexpiration (postinspiration and E2) following isoguvacine injections(Fig. 1). We observed that integrated PI-related RLNAwas suppressedin strength (from 386 ± 86 to 197 ± 107 mVÆ100 ms, P < 0.05) andduration (1.25 ± 0.36–0.51 ± 0.47 s, P < 0.05) following ipsilateralinjections (Figs 2C and 4B). The reduction in postinspiratory activityand increased Ttot (see above) led consequently to a prolonged E2phase (0.50 ± 0.13 vs. 1.72 ± 0.37 s, P < 0.05). Bilateral isoguvacineinjections manifested an apneustic breathing pattern that wascharacterized by a synchronization of PNA and RLNA (Fig. 2D) asa result of a virtual absence of PI-related activity. After bilateralinjection, we observed a further decrease of integrated postinspiratoryRLNA to 27 ± 8 mVÆ100 ms (P < 0.001), and the duration of thepostinspiratory phase was shortened to 0.1 ± 0.07 s (P < 0.001,Fig. 4B). During apneustic breathing, the expiratory interval wasdominated by the E2 phase, which now lasted for 11.23 ± 3.64 s(P < 0.001). Forty to 50 min after the bilateral injections, postinspir-atory RLNA recovered and returned to 296 ± 76 mVÆ100 ms(P < 0.001) and 1.23 ± 0.54 s (P < 0.001) and, consequently, theE2 phase returned to a duration similar to that of controls(0.64 ± 0.30 s, P < 0.001, Figs 2E and 4A).

Effect of isoguvacine on SGP

In six experiments, we recorded SGP to relate the effects of thetemporary lesions with isoguvacine on the dynamic adjustment ofglottal resistance. To analyse changes in SGP following ipsi- andbilateral isoguvacine injections, we took the SGP level of the last thirdof the expiratory interval as baseline glottal resistance (¼ 0). Peakvalues of SGP were measured for the inspiratory and postinspiratoryphases and were expressed as negative values for inspiratory glottal

Fig. 1. Analyses of respiratory parameters. (A) Simultaneous recordings from the phrenic nerve (PNA, upper trace) as an index for inspiratory activity andrecurrent laryngeal nerve (RLNA, lower trace) as as index for the laryngeal motor outputs. Note the biphasic discharge pattern of the RLNA. The shading highlightsthe three phases of breathing as we defined and used it for data analysis: dark grey, inspiration (I); light grey, postinspiration (PI); and no shading for the lateexpiration (E2). (B) PNA and subglottal pressure recordings (SGP). The SGP trace can be used as an index for laryngeal resistance during the respiratory cycle. Theshading highlights the three phases as in A. During inspiration, SGP always decreases owing to the activity of laryngeal abductor muscles while duringpostinspiration, SGP is increased owing to activity of laryngeal adductor muscles. From both traces, we analysed respiratory parameters such as total respiratorycycle length (Ttot), time of inspiration (Ti), time of postinspiration (Tpi), time of late expiration (Te2), and total expiratory duration (Te) including Tpi and Te2.



Fig. 2. Excitation or inhibition of postinspiratory discharge in the recurrent laryngeal nerve. (A) Pattern of phrenic (PNA) and recurrent laryngeal nerve activity(RLNA) during the control situation prior to microinjections. Note the ramp-like discharge of PNA and the three-phase pattern of respiration displayed in RLNA,consisting of increasing discharge during inspiration, decreasing discharge in postinspiration, and lack of discharge in late expiration. (B) Microinjection ofglutamate evoked transient cessation of PNA accompanied by prolonged postinspiratory discharge in the RLN. This effect was defined as transientpostinspiratory apnoea. (C) Unilateral injection of isoguvacine at the same injection site resulted in a decrease in breathing frequency and changes in PNA andRLNA. Following isoguvacine injection PNA duration increased while postinspiratory discharge of the RLN was reduced compared with control activity (A).(D) Subsequent injection of isoguvacine in the contralateral site using the same coordinates as for the ipsilateral injection disrupted the eupnoea-like controlactivity and established an apneustic breathing pattern that was characterized by a further decrease in breathing frequency and a four-fold longer inspiration, asindicated by PNA. The synchronized RLNA was indicative of a virtually abolished postinspiratory discharge of the RLNA. (E) Forty minutes after thecontralateral injection, a recovery of the breathing pattern, as compared with control levels, was observed. (F) Comparison of rhythmic PNA before (uppertrace) and after bilateral injection of isoguvacine (lower trace). Compared with the control frequency, the frequency during apneustic breathing apparentlyshowed no consistent quantal slowing (decrease of frequency in a 1 : 2, 1 : 3, etc., relationship), as indicated by the arrows that mark the onset of PNA duringcontrol.



Fig. 3. Excitation or inhibition of postinspiratory laryngeal adductor activity illustrated with subglottal pressure recording. (A) Control activity illustrating thedynamic changes of upper airway patency during the respiratory cycle. Owing to laryngeal abductor activity the upper airway resistance is decreased duringthe inspiratory phase (indicated by PNA). After inspiration, upper airway resistance was transiently increased due to the activity of laryngeal adductors. During theremaining expiratory phase, SGP levels gradually decreased to baseline level immediately prior to the start of the next inspiration. (B) Effect of glutamate injectioninto the ipsilateral KF (injection site shown in F). Note the massive increase of SGP over the prolongation of the expiratory interval following glutamate injection.(C) Subsequent injection of isoguvacine at the same loci caused prolongation of PNA duration and a decrease in frequency. While abductor activity duringinspiration was apparently not affected, the postinspiratory laryngeal adductor activity was decreased compared with control. (D) Isoguvacine injection at thecontralateral site triggered an apneustic breathing pattern accompanied by absence of any postinspiratory-related upper airway constriction. (E) Illustration of apartial recovery of the control activity prior to isoguvacine injections. (F) Photomicrograph illustrating histological verification of the injection site. Note that thecontralateral injection was placed slightly more ventral than the ipsilateral injection. (G) Superimposed traces of the SGP during control (black) and after bilateralisoguvacine injection (red) underlining the absence of laryngeal adductor activity after transient lesion of the KF.



dilation and positive values for postinspiratory glottal constriction (seeFig. 1). Ipsilateral isoguvacine injection evoked no change in theinspiratory SGP values ()2.28 ± 0.33 vs. )2.01 ± 0.29 mmHg, n.s.),whereas the postinspiratory values were reduced from +2.31 ± 0.31 to+1.15 ± 0.54 mmHg (P < 0.05; Figs 3C and 4C). Following bilateralisoguvacine injections, the SGP value for the inspiratory phaseremained largely unchanged ()2.282 ± 0.338 vs. )1.92 ±0.28 mmHg, n.s.; Fig. 4C). By contrast, SGP values during postin-spiration decreased to +0.19 ± 0.02 mmHg (P < 0.01), suggesting acomplete absence of glottal constrictor activity during apneusticbreathing. After 40–50 min, the postinspiratory SGP values partiallyreturned to control levels (+1.41 ± 0.48 mmHg, P < 0.05, Fig. 4C).Comparison of SGP values during late expiration between control andafter bilateral isoguvacine injection revealed no significant differences.

The latter suggested that the bilateral injections had no significanteffects on the baseline glottal resistance.

Quantal slowing following isoguvacine injections

A recent study has demonstrated that a strong decrease of respiratoryfrequency following pharmacological manipulations reveal a quantalslowing of the rhythm. This was interpreted as transmission failuresbetween two coupled oscillators for respiratory rhythm generation(Mellen et al., 2003). In our study, the apneustic breathing patternevoked by bilateral injection of isoguvacine into the KF was alwaysassociated with a significant slowing of rhythmic motor outputs (seeabove). However, data analysis of a quantal relation between thecontrol rhythm and during apneustic breathing (minutes after bilateralisoguvacine injection) revealed no quantal relation of Ti, Te or Ttot

Fig. 4. Histograms illustrating the effects of unilateral and bilateral isoguvacine injection into the intermediate KF on respiratory parameters. (A) Statisti-cal summary of the effects of temporary lesion of the intermediate KF (n ¼ 14, see stars in Fig. 5) on phrenic nerve activity (PNA). (B) Summary of the effects ofmicroinjections on postinspiratory and expiratory intervals analysed from recurrent laryngeal nerve activity (n ¼ 8 experiments). (C) Summary of the effects ofmicroinjections into the intermediate KF on subglottal pressure changes indicative of the activity of laryngeal abductors or adductors during the respiratory cycle.*P < 0.05, **P < 0.01, ***P < 0.001.



(data not shown). Compared with control values, parameters such asTi, Te and Ttot oscillated randomly in duration following injections(see Fig. 2F).

Histological verification of the injection sites that modulatedpostinspiratory activity

All 14 ispilateral injection sites causing excitation and inhibition ofpostinspiratory activity or glottal constrictors were clustered in theintermediate parts of the KF (Fig. 5). The subsequent isoguvacineinjections caused prolongation of Ti and Ttot accompanied by areduction of postinspiratory discharge in RLN (injection nos. 1–8;Fig. 5, Table 1) or postinspiratory SGP values (injection nos. 9–14;Fig. 5, Table 1). The contralateral injections that finally manifested theapneustic breathing pattern were all clustered in the intermediate partsof the KF as well.An occasional mismatch within the pairs of injection sites was due

to lack of a precise stereotaxic approach in the perfused brainstempreparation. In some cases, the ponto-medullary brainstem wasslightly bent following the removal of forebrain and cerebellum. Thiscertainly had an impact on the location of the injection sites, althoughthe same injection coordinates were used for both.

Effects of glutamate and isoguvacine injections outside theintermediate KF

Caudal injection sites

Glutamate injections (n ¼ 5) into the caudal aspects of the KF and intothe intertrigeminal region (injection sites 15–19; Fig. 5, Table 1) causedtransient apnoeas (prolongation of the expiratory interval by122 ± 21%). These apnoeas were characterized by complete absenceof PI discharge in RLN during the prolonged expiratory interval(Fig. 6A) or by SGP levels remaining at end expiratory valuesthroughout the entire expiratory interval (Fig. 6A). Thus, we classifiedthese apnoeas as E2 apnoeas. Isoguvacine injections at these loci evokedonly a modest Ti prolongation and the subsequent contralateralinjections led to an overall increase in Ti by 90.4 ± 13.6% (see Table 1).

Rostral injection sites

Rostral glutamate injections (n ¼ 5) evoked transient tachypnoealasting 6–14 s (increase of PNA frequency by 71.3 ± 12.8%,P < 0.001), whereas the effects on RLNA and SGP were hetero-geneous (Table 1). Histological analysis revealed that the injectionswere placed in the rostrodorsal parts of the KF (injection sites 21 and23–24, Fig. 6B) and in lateral aspects of the parabrachial complex(injection sites 20 and 22; Fig. 5, Table 1). Ipsilateral isoguvacineinjections into the rostral KF sites evoked a moderate prolongation ofTi; in the lateral PB, injections were even less effective (see Table 1).Subsequent injection at contralateral sites further increased Ti(Table 1), although Ti prolongation was never similar to that followinginjections in the intermediate KF.

Respiratory neurons of the dorsolateral pons

To characterize cellular activity of the dorsolateral pons, we recordedrespiratory neurons in 15 perfused preparations at the followingcoordinates: 0.2–0.5 mm caudal to the caudal end of the inferiorcolliculus and 1.8–2.5 mm lateral to midline. Although we did notverify the precise location of the recording sites histologically, wedetected dorsoventral differences in neuronal activities. Most of thesuperficially recorded neurons (n ¼ 35, recording depth 0.5–1.5 mm)

exerted tonic and not respiratory-related discharges or had burstingactivity unrelated to PNA (data not shown). Nevertheless, 12 of thesuperficially recorded neurons showed a weak inspiratory modulationcharacterized by a slight increase in discharge frequency during PNA(n ¼ 12). At a recording depth of 1.5–2.5 mm, we recorded fromrespiratory neurons that had a clear phase-locked activity in relation toPNA. In total, we identified 19 inspiratory, 17 postinspiratory, sevenlate expiratory and seven phase-spanning neurons (Fig. 7).

Discussion

Excitation or inhibition of discrete neuronal populations within theventral aspects of the intermediate KF led to profound modulation ofbreathing activity. Excitation evoked transient cessation of phrenicnerve activity (sustained IOS) that was accompanied by prolongedpostinspiratory activity in laryngeal motor outputs resulting in glottalclosure. Inhibition of the same neuronal population with the GABAreceptor agonist isoguvacine had severe consequences for ongoingrespiratory activity. Bilateral injections transformed the eupnoea-likebreathing pattern into apneusis, characterized by prolonged inspirationindicating a disturbed IOS mechanism. A novel finding of our studywas that apneusis was characterized by a virtual absence ofpostinspiratory laryngeal motor activity or laryngeal adductor activity.The latter two activities were always present during the ongoingundisturbed respiratory activity. Our data indicated that the interme-diate KF might reflect the pneumotaxic centre in rats, which isapparently involved in the gating of postinspiratory activity within therespiratory network. The latter is of particular importance for asufficient termination of the inspiratory phase during absence ordesensitization of afferent inputs arising from pulmonary stretchreceptors. The strong modulation of postinspiratory upper airwayactivity by KF neurons implies that the pneumotaxic centre potentiallyoverlaps with a major source of laryngeal premotoneurons.

Technical considerations

The present study was performed using an in situ perfused brainstempreparation (Paton, 1996), which provides the advantage to studyrespiratory pattern formation under non-anaesthetized and largelyintact network conditions. It was demonstrated that the perfusedbrainstem preparation exerts an eupnoea-like pattern of respiratoryactivity (Paton, 1996), as was reported for in vivo preparations, butincludes advantages of an in vitro milieu owing to the short set-uptime and advanced mechanical stability (no respiratory movements,largely reduced cardiac pulsations). However, there is one importantdifference to the in vivo situation. The perfused brainstem preparationlacks any sensory feedback from pulmonary stretch receptors (PSRs),as the lungs are collapsed or even removed from the preparation.Under fully intact conditions, the sensory feedback of PSRs is thoughtto play a crucial role in initiation of the postinspiratory phase, resultingin termination of the inspiratory phase (summarized in Richter, 1996).Therefore, in future experiments it will be necessary to determine therole of the pontine respiratory group for the generation of a eupnoeicbreathing pattern under conditions of rhythmic feedback from PSR.

The pneumotaxic centre in rat

The pneumotaxic centre or apneustic centre in rats has been acontroversial issue over the last 10 years. Published studies eitherverified the existence (Wang et al., 1993; Morrison et al., 1994) orprovided evidence that rodents might lack such a centre (Monteau



Fig. 5. Semi-schematic line drawings of coronal sections through the dorsolateral pons to illustrate the location of glutamate and isoguvacine injection sites. Theline drawings illustrate the anatomical location of the injection sites in the PB ⁄ KF complex in the dorsolateral pons, from rostral (top) to caudal (bottom). Weillustrate the ipsilateral injections sites of glutamate and isoguvacine. The symbols depict the type of glutamate-evoked breathing modulation (stars: postinspiratoryapnoea, diamonds: expiratory apnoea, squares: tachypnoea). The numbers allow for the correlation of the pairs of ipsi- (left) and contralateral (right) injection sitesfor isoguvacine. With reference to Table 1, the effects of unilateral and bilateral isoguvacine on PNA, RLNA or SGP can be precisely determined. Note that the pairsof injections sites 1–14 were the most effective in producing apneusis accompanied by either loss of postinspiratory RLNA or laryngeal adductor activity asdetermined from SGP recordings. Abbreviations: CnF, cuneiform nucleus; Dll, dorsal nucleus of the lateral lemniscus; ITR, intertrigeminal region; KF, Kolliker-Fusenucleus; LC, locus coeruleus; ll, lateral lemniscus; Mo5, motor trigeminal nucleus; Pr5, principal sensory trigeminal nucleus. scp, superior cerebellar peduncle.Subnuclei of the parabrachial complex: el, external lateral parabrachial nucleus; exl, extreme lateral parabrachial nucleus; il, internal lateral parabrachial nucleus;m, medial parabrachial nucleus; s, superior lateral parabrachial nucleus; v, ventral lateral parabrachial nucleus; w, waist area of the parabrachial nucleus.



et al., 1989). In the present study, we were able to demonstrate thatthe potential location of a pneumotaxic centre in rat might be moreventral than expected. Topographical analysis of the injection siteswithin the dorsolateral pons revealed that only injection sites placedin the ventral aspects of the intermediate KF led to a maximalexpression of an apneustic breathing pattern. By contrast, injectionsplaced in the lateral or medial parabrachial nucleus were significantlyless effective, although glutamate injections also modulated respir-atory activity. In our study, the inspiratory phase was maximallyprolonged by 400%, which is in accordance with the effectsdescribed for in vivo non-anaesthetized decerebrated rats (Wanget al., 1993; Morrison et al., 1994). The latter experimentalconditions match the situation of the perfused brainstem preparation.Furthermore, our data suggest that the ITR does not contribute topneumotaxic mechanisms. Although the transient cessation of PNAafter glutamate injections into the ITR was similar to that observedin the KF, we never observed a clear apneusis after the subsequentblockade of this region. Simultaneous recordings of the RLN orglottal resistance indicated that the transient apnoeas evoked fromthe ITR were associated with silence of laryngeal adductor activityduring the prolonged expiration. This suggests that the ITR might beinvolved in the modulation of respiratory activity, which requires thespecific activation of late expiratory neurons, for instance duringlocomotor entrainment (Potts et al., 2005).Our recordings of phasic respiratory-related activity in the

dorsolateral pons support recent findings in neonatal rats (Kobayashiet al., 2005), although the cell types recorded were different.However, this discrepancy should be not overestimated and might bedue to differences in experimental approaches and is potentiallyheavily influenced by postnatal age. Overall, the existence ofrespiratory-related neurons in the pons is also in line with severalreports from cats (Cohen & Wang, 1959; Bertrand et al., 1973; Shawet al., 1989; Dick et al., 1994; Cohen & Shaw, 2004) and our results

suggest similar neuronal activities in rats. Anatomical tract tracingstudies showed that the KF of rat is densely reciprocally connectedwith the ventral respiratory groups, including the pre-Botzingercomplex (Dobbins & Feldman, 1994; Ellenberger & Feldman, 1990;Nunez-Abades et al., 1993; Gaytan et al., 1997). This connectivityof the KF with important medullary structures required for respir-atory rhythm generation provides two different views on the roleof the KF in respiratory rhythmogenesis or pattern formation.Respiratory activity of the dorsolateral pons could ‘simply’ be seenas an afferent copy of medullary activities. This would be reasonableunder the assumption that the KF is involved, in particular, in theprocessing of multimodal afferent inputs in order to adapt breathingto changes in environment or behaviour (Dutschmann et al., 2004).If the KF is seen as such an integrative centre, an afferent copy ofthe ongoing respiratory activity would be essential to co-ordinate theafferent inputs with the medullary rhythm generators. Nevertheless,local neuronal oscillators that generate the pontine respiratorydischarges might be feasible (Bertrand et al., 1974; St-John &Bledsoe, 1985). Of particular interest in this regard is the fairly highdensity of specific phase-spanning types of respiratory neurons thatare apparently not present, or at least not numerous, in medullaryrespiratory centres. In particular, these phase-spanning types ofneurons might be a unique feature of pontine respiratory activity, afact that weakens the idea of a simple afferent copy of medullaryactivities. Indeed, the phase-spanning type might control therespiratory phase transition, as has been demonstrated withcomputational models (Rybak et al., 2004).

State-dependent pontine gating of postinspiratory activity

The breathing cycle can be divided into three principal phases:inspiration, postinspiration (also termed passive expiration or decre-menting expiration) and expiration (active expiration or stage II

Table 1. Characterization of the physiological effects that correspond to the anatomical location of the injection sites illustrated in Fig. 5

Injection site

Response to glutamate injectionChange in the duration of theinspiratory phase (Ti)

Effect on PNA Effect on RLNA Effect on SGP Ipsilateral Bilateral

1 Apnoea Tonic PI – +201% +383%2 Apnoea Tonic PI – +181% +398%3 Apnoea Tonic PI – +165% +301%4 Apnoea Tonic PI – +134% +345%5 (Fig. 2) Apnoea Tonic PI – +121% +445%6 Apnoea Tonic PI – +148% +378%7 Apnoea Tonic PI – +156% +391%8 Apnoea Tonic PI – +81% +294%9 Apnoea – Constriction +66% +298%10 Apnoea – Constriction +107% +413%11 Apnoea – Constriction +157% +296%12 Apnoea – Constriction +79% +245%13 (Fig. 3) Apnoea – Constriction +149% +378%14 Apnoea – Constriction +153% +285%15 Apnoea Silence – +27% +112%16 Apnoea Silence – +61% +151%17 Apnoea Silence – +46% +83%18 Apnoea – No constriction +74% +101%19 Apnoea – No constriction +25% +41%20 Tachypnoea PI – +39% +32%21 Tachypnoea PI – +74% +164%22 Tachypnoea PI – +12% +23%23 Tachypnoea – Constriction +65% +151%24 Tachypnoea – Constriction +34% +123%

PI, postinspiratory activity; PNA, phrenic nerve activity; RLNA, recurrent laryngeal nerve activity; SGP, subglottal pressure.



expiration), according to peripheral and central activities. Of crucialimportance for the formation of the breathing motor pattern is thecentral mediation of the phase transitions between inspiration andexpiration or, vice versa, between expiration and inspiration. In thisregard, the neuronal mechanism controlling the IOS, as a hallmark ofthe transition of the respiratory cycle from inspiration into expiration,was investigated in detail. The IOS is mediated by the activity of twodistinct cell types of the respiratory network. It is initiated by theactivity of late inspiratory neurons (Cohen et al., 1993; Haji et al.,2002) and irreversibly mediated by the activation of postinspiratoryneurons (Ezure & Manabe, 1988; Manabe & Ezure, 1988; Hayashiet al., 1996; Krolo et al., 2005). The timing and efficacy of the IOS iscontrolled by afferent feedback from PSRs (von Euler, 1983) anddepend on glutamatergic neurotransmission involving NMDA recep-tors (Pierrefiche et al., 1998).

In other motor behaviours, such as locomotion, pattern generation isconsidered to depend on gating mechanisms that are attributed tosensory interneurons, but also to premotor interneurons that arerhythmically active during locomotion (Sillar, 1991). Because theperfused brainstem preparation is lacking the crucial sensory feedbackof PSRs, the activity of postinspiratory medullary neurons, which areessential for respiratory pattern generation, may rely on a potentially

excitatory drive arising form distinct populations of KF neurons.However, because under physiological conditions both mechanismsare functional, it still remains unclear if there is a hierarchicalorganization of the convergent gating mechanisms. Recent publica-tions suggest that sensory feedback from PSRs might be desensitizedunder certain conditions (Siniaia et al., 2000). This situation mightpredominate during normal, unforced breathing and therefore the KFmight be the crucial component for postinspiratory gating under theseconditions. However, under conditions that require forced breathingactivity, such as during exercise, the PSR component becomesdominant with increasing lung volumes. We conclude that theconvergent gating mechanisms for the postinspiratory phase of thebreathing cycle might not be hierarchical, but state-dependent. In thiscontext, it is important to note that postinspiratory activity is not onlyrequired for the primary pattern generation of the respiratory activitybut is also crucial for many respiratory-related behaviours, inparticular for vocalization (see below).

KF premotor populations that control upper airway calibre

The control of upper airway calibre and its dynamic modulationover a breathing cycle was suggested to be a main feature of

Fig. 6. Respiratory motor responses and changes in subglottal pressure (SGP) or recurrent laryngeal nerve activity (RLNA) following glutamate microinjectionsinto the dorsolateral pons. (A) Injections placed into the intertrigeminal region (see Fig. 5, Table 1) evoked transient expiratory apnoea, as indicated by prolongationof the expiratory interval in PNA recordings. However, in contrast to glutamate injection as illustrated in Figs 2 and 3, the apnoea was never accompanied bypostinspiratory RLNA or any notable laryngeal adductor activity. (B) Transient tachypnoea following glutamate injection into rostral KF or lateral parabrachialnuclei (see Fig. 5, Table 1). Because the response patterns of RLNA or SGP during the evoked tachypnoea were heterogeneous, we illustrate two different patterns.In the left-hand recording, postinspiratory adductor activity was enhanced, as indicated by the increased SGP levels. On the right-hand side, postinspiratory RLNAadapted to the increased breathing frequency while amplitude or peak discharge frequency was not enhanced.



eupnoeic breathing (Dutschmann & Paton, 2002a). A main findingof the present study is that stimulation or inhibition of theintermediate KF activated or abolished postinspiratory laryngealadductor activity. Although the function and connection of theneuronal population of the intermediate KF is presumably

heterogeneous and the descending projections target almost theentire medullary respiratory column, our results suggest thepresence of specific laryngeal adductor premotoneurons in the KF.These neurons may transmit excitatory drive to primary motoneur-ons located in the medullary nucleus ambiguus (Jordan, 2001), an

Fig. 7. Extra- and intracellular recordings of different types of respiratory neurons in the rat dorsolateral pons. (A) Extracellular recording of an inspiratory neuronwith augmenting discharge pattern (I-aug, left-hand side) and intracellular recording of an inspiratory neuron with a decreasing discharge (I-dec, right-hand side). (B)Extracellular recording of an expiratory neuron with a decreasing discharge (E-dec, left-hand side) and intracellular recordings of a postinspiratory neuron (PI, right-hand side). (C) Extracellular recordings of phase-spanning- type neurons. The neuron shown on the left-hand side discharged throughout the late expiratory andinspiratory phase (E-2 ⁄ I) but was consistently inhibited during early expiration. The neuron illustrated on the right-hand side spiked during inspiratory and very earlyexpiratory phase but displayed inhibition during the remaining expiratory interval. (D) Extracellular recording of an augmenting expiratory neuron (E-2).



assumption that is supported by anatomical findings showing thatKF neurons indeed project heavily to the nucleus ambiguus (Nunez-Abades et al., 1990). Nevertheless, it is not unlikely that theselaryngeal premotoneurons contact via axon collaterals also otherneurons of the medullary respiratory networks (such as bulbospinalneurons; see Zheng et al., 1992) that are required for respiratoryrhythm and pattern generation.

More importantly, the dynamic control of laryngeal adductors isessential for various respiratory-related behaviours, such as the vitalairway sneeze and cough reflexes (Shiba et al., 1999) or protectiveapnoea following nasal stimulation (Dutschmann & Paton, 2002b).Besides these vital reflex adaptations, laryngeal adductor activity isrequired to co-ordinate respiration with other behaviours, includingswallowing during food intake (Jean, 2001). Recently publishedresults have suggested that the KF might be involved in thepremotor control of hypoglossal activity (Kuna & Remmers, 1999;Gestreau et al., 2005), underlining the potentially pivotal role of theKF in the co-ordination of more complex respiratory-associatedbehaviours. Last but not least, laryngeal adductor control is ofoutstanding importance for vocalization (Farley et al., 1992;Jurgens, 2002). Consequently, premotor control of these musclegroups links respiration to higher functions such interspeciescommunication.

Acknowledgements

The study was supported by the Deutsche Forschungsgemeinschaft (SFB406 ⁄ C12). We thank C. Kuhlbach, A. M. Bischoff and H. Zillus for technicalassistance.

Abbreviations

E2, late expiration; IOS, inspiratory off-switch; ITR, intertrigeminal region; KF,Kolliker-Fuse; PB, parabrachial; PI, postinspiratory activity; PNA, phrenicnerve activity; PSR, pulmonary stretch receptor; RLN, recurrent laryngealnerve; RLNA, recurrent laryngeal nerve activity; SGP, subglottal pressure; Te,the time of PNA quiescence; Te2, late expiratory time; Ti, the duration of theinspiratory phase; Tpi, postinspiratory time; Ttot, duration of the respiratorycycle length.

References

Alheid, G.F., Milsom, W.K. & McCrimmon, D.R. (2004) Pontine influences onbreathing: an overview. Respir. Physiol. Neurobiol., 143, 105–114.

Baker, T.L., Netick, A. & Dement, W.C. (1981) Sleep-related apneic andapneustic breathing following pneumotaxic lesion and vagotomy. Respir.Physiol., 46, 271–294.

Bertrand, F., Hugelin, A. & Vibert, J.F. (1973) Quantitative study of anatomicaldistribution of respiration related neurons in the pons. Exp. Brain. Res., 16,383–399.

Bertrand, F., Hugelin, A. & Vibert, J.F. (1974) A stereologic model ofpneumotaxic oscillator based on spatial and temporal distributions ofneuronal bursts. J. Neurophysiol., 37, 91–107.

Caille, D., Vibert, J.F. & Hugelin, A. (1981) Apneusis and apnea afterparabrachial or Kolliker Fuse N. lesion; influence of wakefulness. Respir.Physiol., 45, 79–95.

Chamberlin, N.L. (2004) Functional organization of the parabrachial complexand intertrigeminal region in the control of breathing. Respir. Physiol.Neurobiol., 143, 115–125.

Chamberlin, N.L. & Saper, C.B. (1998) A brainstem network mediating apneicreflexes in the rat. J. Neurosci., 18, 6048–6056.

Cohen, M.I., Huang, W.X., Barnhardt, R. & See, W.R. (1993) Timing ofmedullary late-inspiratory neuron discharges: vagal afferent effects indicatepossible off-switch function. J. Neurophysiol., 69, 1784–1787.

Cohen, M.I. & Shaw, C.F. (2004) Role in the inspiratory off-switch of vagalinputs to rostral pontine inspiratory-modulated neurons. Respir. Physiol.Neurobiol., 143, 127–140.

Cohen, M.I. & Wang, S.C. (1959) Respiratory neuronal activity in pons of cat.J. Neurophysiol., 22, 33–50.

Coles, S.K. & Dick, T.E. (1996) Neurones in the ventrolateral pons are requiredfor post hypoxic frequency decline in rats. J. Physiol., 497, 79–94.

Dick, T.E., Bellingham, M.C. & Richter, D.W. (1994) Pontine respiratoryneurons in anesthetized cats. Brain Res., 636, 259–269.

Dobbins, E.G. & Feldman, J.L. (1994) Brainstem network controllingdescending drive to phrenic motoneurons in rat. J. Comp. Neurol., 347, 64–86.

Dutschmann, M. & Herbert, H. (1998) NMDA and GABAA receptors in the ratKolliker-Fuse area control cardiorespiratory responses evoked by trigeminalethmoidal nerve stimulation. J. Physiol., 510, 793–804.

Dutschmann, M., Morschel. M., Kron, M. & Herbert, H. (2004) Developmentof adaptive behaviour of the respiratory network: implications for the pontineKolliker-Fuse nucleus. Respir. Physiol. Neurobiol., 143, 155–165.

Dutschmann, M. & Paton, J.F.R. (2002a) Glycinergic inhibition is essential toco-ordinate cranial and spinal motor outputs of the central respiratorynetwork in the neonatal rat. J. Physiol., 534, 643–653.

Dutschmann, M. & Paton, J.F.R. (2002b) Influence of nasotrigeminal afferentson medullary respiratory neurons and upper airway patency in the rat. Pflug.Arch., 444, 227–235.

Ellenberger, H.H. & Feldman, J.L. (1990) Brainstem connections of the rostralventral respiratory group of the rat. Brain Res., 513, 35–42.

von Euler, C. (1983) On the central pattern generator for the basic breathingrhythmicity. J. Appl. Physiol., 55, 1647–1659.

Ezure, K. & Manabe, M. (1988) Decrementing expiratory neurons of theBotzinger complex. II. Direct inhibitory synaptic linkage with ventralrespiratory group neurons. Exp. Brain Res., 72, 159–166.

Ezure, K., Otake, K., Lipski, J. & She, R.B. (1991) Efferent projections ofpulmonary rapidly adapting receptor relay neurons in the cat. Brain Res.,564, 268–278.

Farley, G.R., Barlow, S.M. & Netsell, R. (1992) Factors influencing neuralactivity in parabrachial regions during cat vocalizations. Exp. Brain Res., 89,341–351.

Gaytan, S.P., Calero, F., Nunez-Abades, P.A., Morillo, A.M. & Pasaro, R.(1997) Pontomedullary efferent projections of the ventral respiratoryneuronal subsets of the rat. Brain Res. Bull., 42, 323–234.

Gestreau, C., Dutschmann, M., Obled, S. & Bianchi, A.L. (2005) Activation ofXII motoneurons and premotor neurons during various oropharyngealbehaviors. Respir. Physiol. Neurobiol., 147, 159–176.

Guyenet, P.G., Koshiya, N., Huangfu, D., Verberne, A.J. & Riley, T.A. (1993)Central respiratory control of A5 and A6 pontine noradrenergic neurons. Am.J. Physiol., 264, R1035–R1044.

Haji, A., Okazaki, M., Yamazaki, H. & Takeda, R. (2002) Physiologicalproperties of late inspiratory neurons and their possible involvement ininspiratory off-switching in cats. J. Neurophysiol., 87, 1057–1067.

Hayashi, F., Coles, S.K. & McCrimmon, D.R. (1996) Respiratory neuronsmediating the Breuer-Hering reflex prolongation of expiration in rat.J. Neurosci., 16, 6526–6536.

Hilaire, G., Viemari, J.C., Coulon, P., Simonneau, M. & Bevengut, M.(2004) Modulation of the respiratory rhythm generator by the pontinenoradrenergic A5 and A6 groups in rodents. Respir. Physiol. Neurobiol.,143, 187–197.

Hsieh, Y.H., Siegel, R.E. & Dick, T.E. (2004) Pontine GABAergic pathways:role and plasticity in the hypoxic ventilatory response. Respir. Physiol.Neurobiol., 143, 141–153.

Jean, A. (2001) Brain stem control of swallowing: neuronal network andcellular mechanisms. Physiol. Rev., 81, 929–969.

Jiang, M., Alheid, G.F., Calandriello, T. & McCrimmon, D.R. (2004)Parabrachial-lateral pontine neurons link nociception and breathing. Respir.Physiol. Neurobiol., 143, 215–233.

Jodkowski, J.S., Coles, S.K. & Dick, T.E. (1997) Prolongation in expirationevoked from ventrolateral pons of adult rats. J. Appl. Physiol., 82, 377–378.

Jordan, D. (2001) Central nervous pathways and control of the airways. Respir.Physiol., 125, 67–81.

Jurgens, U. (2002) Neural pathways underlying vocal control. Neurosci.Biobehav. Rev., 26, 235–258.

Kobayashi, S., Onimaru, H., Inoue, M., Inoue, T. & Sasa, R. (2005)Localization and properties of respiratory neurons in the rostral pons ofthe newborn rat. Neuroscience, 134, 317–325.

Koshiya, N. & Guyenet, P.G. (1994) Role of the pons in the carotid sympatheticchemoreflex. Am. J. Physiol., 267, R508–R518.

Krolo, M., Tonkovic-Capin, V., Stucke, A.G., Stuth, E.A., Hopp, F.A., Dean, C.& Zuperku, E.J. (2005) Subtype composition and responses of respiratoryneurons in the pre-botzinger region to pulmonary afferent inputs in dogs.J. Neurophysiol., 93, 2674–2687.



Kuna, S.T. & Remmers, J.E. (1999) Premotor input to hypoglossalmotoneurons from Kolliker-Fuse neurons in decerebrate cats. Respir.Physiol., 117, 85–95.

Lara, J.P., Dawid-Milner, M.S., Lopez, M.V., Montes, C., Spyer, K.M. &Gonzalez-Baron, S. (2002) Laryngeal effects of stimulation of rostral andventral pons in the anaesthetized rat. Brain Res., 934, 97–106.

Lumsden, T. (1923) Observations on the respiratory centers in the cat.J. Physiol., 57, 152–160.

Manabe, M. & Ezure, K. (1988) Decrementing expiratory neurons of theBotzinger complex. I. Response to lung inflation and axonal projection. Exp.Brain Res., 72, 150–158.

Mellen, N.M., Janczewski, W.A., Bocchiaro, C.M. & Feldman, J.L. (2003)Opioid-induced quantal slowing reveals dual networks for respiratory rhythmgeneration. Neuron, 37, 821–826.

Monteau, R., Errchidi, S., Gauthier, P., Hilaire, G. & Rega, P. (1989)Pneumotaxic centre and apneustic breathing: interspecies differencesbetween rat and cat. Neurosci. Lett., 99, 311–316.

Morrison, S.F., Cravo, S.L. & Wilfehrt, H.M. (1994) Pontine lesions produceapneusis in the rat. Brain Res., 652, 83–86.

Nunez-Abades, P.A., Morillo, A.M. & Pasaro, R. (1993) Brainstem connectionsof the rat ventral respiratory subgroups: afferent projections. J. Auton. Nerv.Syst., 42, 99–118.

Nunez-Abades, P.A., Portillo, F. & Pasaro, R. (1990) Characterisation ofafferent projections to the nucleus ambiguus of the rat by means offluorescent double labelling. J. Anat., 72, 1–15.

Paton, J.F.R. (1996) A working heart-brainstem preparation of the mouse.J. Neurosci. Meth., 65, 63–68.

Pierrefiche, O., Haji, A., Foutz, A.S., Takeda, R., Champagnat, J. & Denavit-Saubie, M. (1998) Synaptic potentials in respiratory neurones during evokedphase switching after NMDA receptor blockade in the cat. J. Physiol., 508,549–559.

Potts, J.T., Rybak, I.A. & Paton, J.F.R. (2005) Respiratory rhythm entrainmentby somatic afferent stimulation. J. Neurosci., 25, 1965–1978.

Richter, D.W. (1982) Generation and maintenance of the respiratory rhythm.J. Exp. Biol., 100, 93–10.

Richter, D.W. (1996) Neural regulation of respiration: rhythmogenesis andafferent control. In Greger, R. & Windhorst, U. (Eds), ComprehensiveHuman Physiology. Springer, Berlin, pp. 2079–2095.

Rybak. I.A., Shevtsova, N.A., Paton, J.F., Dick, T.E., St-John, W.M., Morschel,M. & Dutschmann, M. (2004) Modeling the ponto-medullary respiratorynetwork. Respir. Physiol. Neurobiol., 143, 307–319.

Segers, L.S., Shannon, R. & Lindsey, B.G. (1985) Interactions between rostralpontine and ventral medullary respiratory neurons. J. Neurophysiol., 54,318–334.

Shannon, R., Baekey, D.M., Morris, K.F., Nuding, S.C., Segers, L.S. & Lindsey,B.G. (2004) Pontine respiratory group neuron discharge is altered duringfictive cough in the decerebrate cat. Respir. Physiol. Neurobiol., 142, 43–54.

Shaw, C.F., Cohen, M.I. & Barnhardt, R. (1989) Inspiratory-modulated neuronsof the rostrolateral pons: effects of pulmonary afferent input. Brain Res., 485,179–184.

Shiba,K., Satoh, I.,Kobayashi,N.&Hayashi, F. (1999)Multifunctional laryngealmotoneurons: an intracellular study in the cat. J. Neurosci., 19, 2717–2727.

Sillar, K.T. (1991) Spinal pattern generation and sensory gating mechanisms.Curr. Opin. Neurobiol., 1, 583–589.

Siniaia, M.S., Young, D.L. & Poon, C.S. (2000) Habituation and desensitiza-tion of the Hering-Breuer reflex in rat. J. Physiol., 523, 479–491.

Song, G. & Poon, C.S. (2004) Functional and structural models of pontinemodulation of mechanoreceptor and chemoreceptor reflexes. Respir. Physiol.Neurobiol., 143, 281–292.

St-John, W. & M. (1998) Neurogenesis of patterns of automatic ventilatoryactivity. Prog. Neurobiol., 56, 97–117.

St-John, W.M. & Bledsoe, T.A. (1985) Genesis of rhythmic respiratory activityin pons independent of medulla. J. Appl. Physiol., 59, 684–690.

St-John, W.M. & Paton, J.F.R. (2004) Role of pontile mechanisms in theneurogenesis of eupnea. Respir. Physiol. Neurobiol., 143, 321–332.

Wang, W., Fung, M.L. & St-John, W.M. (1993) Pontile regulation ofventilatory activity in the adult rat. J. Appl. Physiol., 74, 2801–2811.

Zheng, Y., Barillot, J.C. & Bianchi, A.L. (1992) Intracellular electrophysio-logical and morphological study of the medullary inspiratory neurons of thedecerebrate rat. Brain Res., 576, 235–244.



Documents

The Kölliker-Fuse nucleus gates the postinspiratory phase of the respiratory cycle to control inspiratory off-switch and upper airway resistance in rat