The Kolliker-Fuse nucleus (KF), which is part of the parabrachial complex, participates in the generation of eupnoea under resting conditions and the control of active abdominal expiration when increased ventilation is required. Moreover, dysfunctions in KF neuronal activity are believed to play a role in the emergence of respiratory abnormalities seen in Rett syndrome (RTT), a progressive neurodevelopmental disorder associated with an irregular breathing pattern and frequent apnoeas. Relatively little is known, however, about the intrinsic dynamics of neurons within the KF and how their synaptic connections affect breathing pattern control and contribute to breathing irregularities. In this study, we use a reduced computational model to consider several dynamical regimes of KF activity paired with different input sources to determine which combinations are compatible with known experimental observations. We further build on these findings to identify possible interactions between the KF and other components of the respiratory neural circuitry. Specifically, we present two models that both simulate eupnoeic as well as RTT-like breathing phenotypes. Using nullcline analysis, we identify the types of inhibitory inputs to the KF leading to RTT-like respiratory patterns and suggest possible KF local circuit organizations. When the identified properties are present, the two models also exhibit quantal acceleration of late-expiratory activity, a hallmark of active expiration featuring forced exhalation, with increasing inhibition to KF, as reported experimentally. Hence, these models instantiate plausible hypotheses about possible KF dynamics and forms of local network interactions, thus providing a general framework as well as specific predictions for future experimental testing.

Cortical and basal ganglia circuits play a crucial role in the formation of goal-directed and habitual behaviors. In this study, we investigate the cortico-striatal circuitry involved in learning and the role of this circuitry in the emergence of inflexible behaviors such as those observed in addiction. Specifically, we develop a computational model of cortico-striatal interactions that performs concurrent goal-directed and habit learning. The model accomplishes this by distinguishing learning processes in the dorsomedial striatum (DMS) that rely on reward prediction error signals as distinct from the dorsolateral striatum (DLS) where learning is supported by salience signals. These striatal subregions each operate on unique cortical input: the DMS receives input from the prefrontal cortex (PFC) which represents outcomes, and the DLS receives input from the premotor cortex which determines action selection. Following an initial learning of a two-alternative forced choice task, we subjected the model to reversal learning, reward devaluation, and learning a punished outcome. Behavior driven by stimulus-response associations in the DLS resisted goal-directed learning of new reward feedback rules despite devaluation or punishment, indicating the expression of habit. We repeated these simulations after the impairment of executive control, which was implemented as poor outcome representation in the PFC. The degraded executive control reduced the efficacy of goal-directed learning, and stimulus-response associations in the DLS were even more resistant to the learning of new reward feedback rules. In summary, this model describes how circuits of the dorsal striatum are dynamically engaged to control behavior and how the impairment of executive control by the PFC enhances inflexible behavior.

Interestingly, bifurcation analysis that you really wait for now is coming. It's significant to wait for the representative and beneficial books to read. Every book that is provided in better way and utterance will be expected by many peoples. Even you are a good reader or not, feeling to read this book will always appear when you find it. But, when you feel hard to find it as yours, what to do? Borrow to your friends and don't know when to give back it to her or him.

Central pattern generators (CPGs), specialized oscillatory neuronal networks controlling rhythmic motor behaviors such as breathing and locomotion, must adjust their patterns of activity to a variable environment and changing behavioral goals. Neuromodulation adjusts these patterns by orchestrating changes in multiple ionic currents. In the medicinal leech, the endogenous neuromodulator myomodulin speeds up the heartbeat CPG by reducing the electrogenic Na+/K+ pump current and increasing h-current in pairs of mutually inhibitory leech heart interneurons (HNs), which form half-center oscillators (HN HCOs). Here we investigate whether the comodulation of two currents could have advantages over a single current in the control of functional bursting patterns of a CPG. We use a conductance-based biophysical model of an HN HCO to explain the experimental effects of myomodulin. We demonstrate that, in the model, comodulation of the Na+/K+ pump current and h-current expands the range of functional bursting activity by avoiding transitions into nonfunctional regimes, such as asymmetric bursting and plateau-containing seizure-like activity. We validate the model by finding parameters that reproduce temporal bursting characteristics matching experimental recordings from HN HCOs under control, three different myomodulin concentrations, and Cs+ treated conditions. The matching cases are located along the border of an asymmetric regime away from the border with more dangerous seizure-like activity. We found a simple comodulation mechanism with an inverse relation between the pump and h-currents makes a good fit of the matching cases and comprises a general mechanism for the robust and flexible control of oscillatory neuronal networks. SIGNIFICANCE STATEMENT Rhythm-generating neuronal circuits adjust their oscillatory patterns to accommodate a changing environment through neuromodulation. In different species, chemical messengers participating in such processes may target two or more membrane currents. In medicinal leeches, the neuromodulator myomodulin speeds up the heartbeat central pattern generator by reducing Na+/K+ pump current and increasing h-current. In a computational model, we show that this comodulation expands the range of central pattern generator's functional activity by navigating the circuit between dysfunctional regimes resulting in a much wider range of cycle period. This control would not be attainable by modulating only one current, emphasizing the synergy of combined effects. Given the prevalence of h-current and Na+/K+ pump current in neurons, similar comodulation mechanisms may exist across species.

Pathology in neural circuits that control the expression of goal-directed and habitual behaviors is hypothesized to be a major contributing factor to addiction. In this study, we investigate cortico-striatal circuitry involved in learning and how cortical interactions with specific striatal subregions are involved in the emergence of inflexible behaviors such as compulsive drinking. Specifically, we develop a computational model of cortico-striatal interactions that performs concurrent goal-directed and stimulus-response learning. The model accomplishes learning by distinguishing between the dorsomedial striatum (DMS)—where dopamine release encodes reward prediction error—and the dorsolateral striatum (DLS)—where dopamine release encodes motivation or salience. These striatal subregions each operate on unique cortical input: the DMS receives input from the prefrontal cortex (PFC), which represented outcomes and the DLS receives input from the premotor cortex which determines action selection. Following an initial learning of a two-alternative forced choice task, we subjected the model to reversal learning, reward devaluation, and punishment learning. Behavior driven by stimulus-response associations in the DLS resisted goal-directed learning of new reward feedback rules despite devaluation or punishment, indicating the expression of habit. We repeated these simulations after the loss of executive control, which was implemented as poor outcome representation in the PFC. Following this manipulation, no detectable of reward devaluation was observed, however, the efficacy of goal-directed learning was reduced, and stimulus response associations in the DLS were even more resistant to the learning of new reward feedback rules. In summary, this model provides a mechanism that describes how the loss of executive control could contribute to the emergence of inflexible behavior. Introduction Advanced stages of addiction are hypothesized to coincide with the transition in the control of behavior from neural circuits that are optimized for flexible responding to those that are optimized for inflexible responding. It is thought that this transition involves the abnormal activation of neural circuitry in the basal ganglia that is devoted to habit and automaticity (Graybiel, 2008; Lipton et al., 2019). Molecular changes resulting from chronic substance abuse alter the role of this circuitry from enabling the performance of routine tasks without attention or cognitive load to driving compulsive drug seeking (Lüscher and Janak, 2021; Lüscher et al., 2020). Clearly articulating the computational processes that unfold across these circuits and how they are altered in addiction is critical for understanding this disease and identifying novel treatment strategies. Here, we focus on two structures within the basal ganglia—specifically, the dorsomedial striatum (DMS) and dorsolateral striatum (DLS)—that have been implicated in the pathophysiology of inflexible behavior and drug addiction (Corbit and Janak, 2016; Lipton et al., 2019). In an alcohol-seeking operant task, seeking behavior can be disrupted by inactivation of the DMS but is insensitive to inactivation of the DLS during early training (Corbit et al., 2012). After extensive training, the specificity of these manipulations becomes reversed: alcohol seeking is disrupted by inactivation of the DLS and is insensitive to inactivation of the DMS. Similarly, cocaine-seeking behavior is sensitive to disruption of the DMS during early learning and disruption of the DLS after extensive training (Murray et al., 2012). In summary, these partitions of the dorsal striatum have distinct temporal contributions to drug-seeking behavior where DMS is initially critical and eventually transitions to DLS. The distinct roles of the DMS and DLS in inflexible behavior are derived from differences in their computational properties. The DMS and DLS are respectively involved in goal-directed and habitual behavior (Schwabe and Wolf, 2011). Manipulations that destroy or disrupt the DMS and DLS during instrumental behavioral tasks reveal their respective roles in goal-directed learning and the formation of habit (Yin and Knowlton, 2006). The inactivation of the DMS decreases sensitivity to reward devaluation (Yin et al., 2005), and the destruction of the DLS increases sensitivity to reward devaluation and abolishes habitual seeking in the absence of a reward (Yin et al., 2004). There is evidence that it takes more time to engage learning mechanisms in the DLS; plasticity in the DLS occurs relatively slowly compared to DMS as training progresses (Yin et al., 2009). Differences in the computational function of the DMS and DLS are supported by differences in their synaptic plasticity mechanisms and differences in their cortical inputs. In the basal ganglia hypothesis for reward-based learning, action selection is gated at the striatum by dopamine-mediated synaptic plasticity of cortical projections to medium spiny neurons (Frank, 2005; Graybiel, 2008). In classical reinforcement learning, dopamine release encodes reward prediction error (RPE). However, different compartments of the striatum receive partition-specific nigro-striatal projections that encode different information (Lerner et al., 2015; Matsumoto and Hikosaka, 2009). Classical RPE-encoding nigrostriatal neurons project to the DMS, and salience-encoding neurons project to the DLS (Lerner et al., 2015). Moreover, dopamine release in the basal ganglia acts on cortico-striatal synapses, and cortical input to the striatum is topographically organized across the DMS and DLS. Inputs to the DS exhibit a clear topographic bias where the medial portion of the striatum is more likely to receive input from the prefrontal cortex (PFC), and the lateral portion more likely to receive input from somatosensory and motor regions (Hunnicutt et al., 2016; Pan et al., 2010; Peters et al., 2021). The input from PFC to the DMS is of particular interest. In rodents, the PFC is important for executive function and goal-directed learning (Barker et al., 2015; Hart et al., 2018a, 2018b; Kesner and Churchwell, 2011; Ostlund and Balleine, 2005; Tran-Tu-Yen et al., 2009). Anatomically comparable structures in humans and non-human primates are involved in cognitive and executive function as well as goal-directed behavior (Balleine and O’Doherty, 2010; Donahue and Lee, 2015; Kennerley et al., 2009; Laubach et al., 2018; Perry et al., 2011; Tsutsui et al., 2016). Critically, these regions are impaired following prolonged alcohol use (Schacht et al., 2013) and this impairment corresponds to increased responding to alcohol (Crews and Boettiger, 2009; Myrick et al., 2004). In recent work, we investigated how the computational properties of the PFC are altered in a rodent model of excessive alcohol consumption and neuronal activity in the medial PFC fails to appropriately code for intent to drink and seeking behavior (Linsenbardt et al., 2019; Timme et al., 2021). Taken in combination with observations about the DMS and DLS, we leverage these results to illustrate a hypothesis for how impairment in the PFC contributes to the emergence of inflexible behavior. Here we present a new computational model of cortico-striatal learning that incorporates goaldirected learning in the DMS and stimulus-response learning in the DLS. We derive this implementation from a theory of reinforcement learning based on dopamine-mediated plasticity of cortico-striatal projections to medium spiny neurons (MSNs) (Graybiel, 2008). Dopamine induces long-term potentiation and long-term depression in D1 and D2 receptor expressing MSNs respectively, and these changes are hypothesized to configure the basal ganglia to selectively disinhibit thalamocortical relay neurons in the context of ongoing behavioral tasks (Frank, 2005). This hypothesis has been incorporated in computational models that include reward-based learning that is based on RPE in the striatum (Frank, 2005; Kim et al., 2017; Mulcahy et al., 2020). However, stimulus-response or habit was previously implemented as a consequence of Hebbian learning in cortico-cortical projections (Kim et al., 2017; Mulcahy et al., 2020). Recent publications have investigated contemporaneous goal-directed and habitual learning in a model of reinforcement learning (Miller et al., 2019) and the consequences of the spatial distribution of dopamine release in the medial-lateral axis of the dorsal stratum (Hamid et al., 2021). To understand the mechanism of how behaviors transition from goal-directed to habitual, a computational model is required that is capable of articulating how changes in cortico-striatal plasticity support both goal-directed learning in the DMS and stimulus-response learning in the DLS, which we provide here for the first time. In the present study, we simulate two-alternative forced choice behavioral tasks to investigate the interaction of the DMS and the DLS. Following an initial learning session, we challenge the model with reward devaluation, reward reversal, and punishment learning. These different behavioral tasks are implemented by manipulations to the magnitude, action contingency, and valence of the reward feedback. We challenge the model again in scenarios characterized by the loss of executive control. In these simulations, neural activity of the PFC fails to appropriately code for action-selection. Our results demonstrate how the loss of executive function could reduce the efficacy of goal-directed learning and emphasize the expression of stimulus-response or habitual behavior. Results Organization of cortico-striatal partitions. In this study, we extend our previous model of the basal ganglia to implement a neural network that combines learning in the DMS and learning in the DLS to perform a two-alternative forced-choice decision-making task (Fig. 1A). This new model implements the pattern of choice-specific channels for neuronal circu

Cardio-ventilatory coupling refers to a heartbeat (HB) occurring at a preferred latency before the onset of the next breath. We hypothesized that the pressure pulse generated by a HB activates baroreceptors that modulates brainstem expiratory neuronal activity and delays the initiation of inspiration. In supine male subjects, we recorded ventilation, electrocardiogram, and blood pressure during 20-min epochs of baseline, slow-deep breathing, and recovery. In in situ rodent preparations, we recorded brainstem activity in response to pulses of perfusion pressure. We applied a well-established respiratory network model to interpret these data. In humans, the latency between HBs and onset of inspiration was consistent across different breathing patterns. In in situ preparations, a transient pressure pulse during expiration activated a subpopulation of expiratory neurons normally active during post-inspiration; thus, delaying the next inspiration. In the model, baroreceptor input to post-inspiratory neurons accounted for the effect. These studies are consistent with baroreflex activation modulating respiration through a pauci-synaptic circuit from baroreceptors to onset of inspiration.

Acute ethanol (EtOH) intoxication results in several maladaptive behaviors that may be attributable, in part, to the effects of EtOH on neural activity in medial prefrontal cortex (mPFC). The acute effects of EtOH on mPFC function have been largely described as inhibitory. However, translating these observations on function into a mechanism capable of delineating acute EtOH's effects on behavior has proven difficult. This review highlights the role of acute EtOH on electrophysiological measurements of mPFC function and proposes that interpreting these changes through the lens of dynamical systems theory is critical to understand the mechanisms that mediate the effects of EtOH intoxication on behavior. Specifically, the present review posits that the effects of EtOH on mPFC N-methyl-d-aspartate (NMDA) receptors are critical for the expression of impaired behavior following EtOH consumption. This hypothesis is based on the observation that recurrent activity in cortical networks is supported by NMDA receptors, and, when disrupted, may lead to impairments in cognitive function. To evaluate this hypothesis, we discuss the representation of mPFC neural activity in low-dimensional, dynamic state spaces. This approach has proven useful for identifying the underlying computations necessary for the production of behavior. Ultimately, we hypothesize that EtOH-related alterations to NMDA receptor function produces alterations that can be effectively conceptualized as impairments in attractor dynamics and provides insight into how acute EtOH disrupts forms of cognition that rely on mPFC function. This article is part of the special Issue on 'Neurocircuitry Modulating Drug and Alcohol Abuse'.

Specialized oscillatory circuits, central pattern generators (CPGs), control rhythmic motor behaviors such as locomotion and breathing. To accommodate a motor pattern to environmental changes and behavioral goals, neuromodulators adjust the dynamics of CPGs by orchestrating changes in various ionic currents in a wide range of their biophysical parameters to expand temporal characteristics of the functional pattern. Recent studies provide evidence that the Na+/K+ pump contributes to the dynamics of CPGs and is controlled by neuromodulation [1– 4]. In the leech heartbeat CPG, the neuropeptide myomodulin reduces the period of bursting activity by increasing the hyperpolarization‐activated (h)‐current and decreasing the Na+/K+ pump current [4]. The application of myomodulin reduces the period of oscillatory activity by 17%. Application of Cs+, which is an h‐current blocker, increases the period of bursting by 24% relative to control. The application of myomodulin along with Cs+ decreases the period by 12% relative to treatment with Cs+ [4]. Here we investigate how the period of a bursting can be controlled in a wide range while functional bursting is maintained with a focus on the role of the Na+/K+ pump. We also investigate the risk associated with multistability in neuronal dynamics.We optimized a model of the leech heart interneuron (HN), which includes the Na+/K+ pump current and intracellular Na+ dynamics [1] and investigated the activity regimes of pairs of mutually inhibitory coupled HNs forming half‐center oscillators (HCOs). HCOs form the kernel of the leech heartbeat CPG. We investigated eight model variants representing combinations of three experimental treatments: the blockade of chemical synapses representing the application of bicuculline, the blockade of h‐current representing the application of Cs+, and the enhancement of the h‐current and inhibition of the Na+/K+ pump current representing the application of myomodulin. The model captures the qualitative trends in change of cycle period observed in experiments with myomodulin and Cs+. We found ranges of parameters where neurons showed functional bursting. The coordinated changes of maximal conductance of h‐current (Gh) and maximal pump activity (IPumpMax) increases the range of period of functional bursting as well as the range of parameters Gh and IPumpMax. We hypothesize that myomodulin co‐modulates h‐ and pump currents to expand the domain of the functional activity.Support or Funding InformationSupported by NINDS 1 R01 NS085006 to RLC and 1 R21 NS111355 to RLC and GSC.

These four Excel files contain the values for Z (Normalized Center of Mass Position), P (Period), A (Amplitude), and D (Distance) in each cycle of a recording for four different cats. The cat names are encoded: MO, NO, TA, and WE, labeled in the Cat column. Recordings include the 04:04 m/s condition, the 04:06 m/s condition, the 04:08 m/s condition, and the 08:04 m/s condition, labeled in the Left : Right Speed Ratio column. Recordings with no anesthesia have a 0 and recordings with anesthesia have a 1 in the Anesthesia column. The cycles in this file are all of the cycles in the recording before removal of outliers and the cycles at the beginning and end of the recording, by the description in the manuscript.

The expiratory neurons of the Bötzinger complex (BötC) provide inhibitory inputs to the respiratory network, which, during eupnea, are critically important for respiratory phase transition and duration control. Herein, we investigated how the BötC neurons interact with the expiratory oscillator located in the parafacial respiratory group (pFRG) and control the abdominal activity during active expiration. Using the decerebrated, arterially perfused in situ rat preparations, we recorded the neuronal activity and performed pharmacological manipulations of the BötC and pFRG during hypercapnia or after the exposure to short-term sustained hypoxia – conditions that generate active expiration. The experimental data were integrated in a mathematical model to gain new insights in the inhibitory connectome within the respiratory central pattern generator. Our results reveal a complex inhibitory circuitry within the BötC that provides inhibitory inputs to the pFRG thus restraining abdominal activity under resting conditions and contributing to abdominal expiratory pattern formation during active expiration.

Variability in blood pressure has become an important metric to consider as more is learned about the link between excessive blood pressure variability and adverse health outcomes. In this study using slow deep breathing in human subjects, we found that heart rate and pulse pressure variations have comparable effects on the amplitude of blood pressure waves, and it is the common action of the two that defines the phase relationship between respiration and blood pressure oscillations.

Our previous study of cat locomotion demonstrated that lateral displacements of the centre of mass (COM) were strikingly similar to those of human walking and resembled the behaviour of an inverted pendulum (Park et al. 2019 J. Exp. Biol.222, 14. (doi:10.1242/jeb.198648)). Here, we tested the hypothesis that frontal plane dynamics of quadrupedal locomotion are consistent with an inverted pendulum model. We developed a simple mathematical model of balance control in the frontal plane based on an inverted pendulum and compared model behaviour with that of four cats locomoting on a split-belt treadmill. The model accurately reproduced the lateral oscillations of cats' COM vertical projection. We inferred the effects of experimental perturbations on the limits of dynamic stability using data from different split-belt speed ratios with and without ipsilateral paw anaesthesia. We found that the effect of paw anaesthesia could be explained by the induced bias in the perceived position of the COM, and the magnitude of this bias depends on the belt speed difference. Altogether, our findings suggest that the balance control system is actively involved in cat locomotion to provide dynamic stability in the frontal plane, and that paw cutaneous receptors contribute to the representation of the COM position in the nervous system.

Motor adaptation to perturbations is provided by learning mechanisms operating in the cerebellum and basal ganglia. The cerebellum normally performs motor adaptation through supervised learning using information about movement error provided by visual feedback. However, if visual feedback is critically distorted, the system may disengage cerebellar error-based learning and switch to reinforcement learning mechanisms mediated by basal ganglia. Yet, the exact conditions and mechanisms of cerebellum and basal ganglia involvement in motor adaptation remain unknown. We use mathematical modeling to simulate control of planar reaching movements that relies on both error-based and non-error-based learning mechanisms. We show that for learning to be efficient only one of these mechanisms should be active at a time. We suggest that switching between the mechanisms is provided by a special circuit that effectively suppresses the learning process in one structure and enables it in the other. To do so, this circuit modulates learning rate in the cerebellum and dopamine release in basal ganglia depending on error-based learning efficiency. We use the model to explain and interpret experimental data on error- and non-error-based motor adaptation under different conditions.

In this study, we explore the functional role of striatal cholinergic interneurons, hereinafter referred to as tonically active neurons (TANs), via computational modeling; specifically, we investigate the mechanistic relationship between TAN activity and dopamine variations and how changes in this relationship affect reinforcement learning in the striatum. TANs pause their tonic firing activity after excitatory stimuli from thalamic and cortical neurons in response to a sensory event or reward information. During the pause striatal dopamine concentration excursions are observed. However, functional interactions between the TAN pause and striatal dopamine release are poorly understood. Here we propose a TAN activity-dopamine relationship model and demonstrate that the TAN pause is likely a time window to gate phasic dopamine release and dopamine variations reciprocally modulate the TAN pause duration. Furthermore, this model is integrated into our previously published model of reward-based motor adaptation to demonstrate how phasic dopamine release is gated by the TAN pause to deliver reward information for reinforcement learning in a timely manner. We also show how TAN-dopamine interactions are affected by striatal dopamine deficiency to produce poor performance of motor adaptation.

The retrotrapezoid nucleus (RTN) contains chemosensitive cells that distribute CO2-dependent excitatory drive to the respiratory network. This drive facilitates the function of the respiratory central pattern generator (rCPG) and increases sympathetic activity. It is also evidenced that during hypercapnia, the late-expiratory (late-E) oscillator in the parafacial respiratory group (pFRG) is activated and determines the emergence of active expiration. However, it remains unclear the microcircuitry responsible for the distribution of the excitatory signals to the pFRG and the rCPG in conditions of high CO2. Herein, we hypothesized that excitatory inputs from chemosensitive neurons in the RTN are necessary for the activation of late-E neurons in the pFRG. Using the decerebrated in situ rat preparation, we found that lesions of neurokinin-1 receptor-expressing neurons in the RTN region with substance P-saporin conjugate suppressed the late-E activity in abdominal nerves (AbNs) and sympathetic nerves (SNs) and attenuated the increase in phrenic nerve (PN) activity induced by hypercapnia. On the other hand, kynurenic acid (100 mM) injections in the pFRG eliminated the late-E activity in AbN and thoracic SN but did not modify PN response during hypercapnia. Iontophoretic injections of retrograde tracer into the pFRG of adult rats revealed labeled phox2b-expressing neurons within the RTN. Our findings are supported by mathematical modeling of chemosensitive and late-E populations within the RTN and pFRG regions as two separate but interacting populations in a way that the activation of the pFRG late-E neurons during hypercapnia require glutamatergic inputs from the RTN neurons that intrinsically detect changes in CO2/pH.

Coordination of respiratory pump and valve muscle activity is essential for normal breathing. A hallmark respiratory response to hypercapnia and hypoxia is the emergence of active exhalation, characterized by abdominal muscle pumping during the late one-third of expiration (late-E phase). Late-E abdominal activity during hypercapnia has been attributed to the activation of expiratory neurons located within the parafacial respiratory group (pFRG). However, the mechanisms that control emergence of active exhalation, and its silencing in restful breathing, are not completely understood. We hypothesized that inputs from the Kölliker-Fuse nucleus (KF) control the emergence of late-E activity during hypercapnia. Previously, we reported that reversible inhibition of the KF reduced postinspiratory (post-I) motor output to laryngeal adductor muscles and brought forward the onset of hypercapnia-induced late-E abdominal activity. Here we explored the contribution of the KF for late-E abdominal recruitment during hypercapnia by pharmacologically disinhibiting the KF in in situ decerebrate arterially perfused rat preparations. These data were combined with previous results and incorporated into a computational model of the respiratory central pattern generator. Disinhibition of the KF through local parenchymal microinjections of gabazine (GABA A receptor antagonist) prolonged vagal post-I activity and inhibited late-E abdominal output during hypercapnia. In silico, we reproduced this behavior and predicted a mechanism in which the KF provides excitatory drive to post-I inhibitory neurons, which in turn inhibit late-E neurons of the pFRG. Although the exact mechanism proposed by the model requires testing, our data confirm that the KF modulates the formation of late-E abdominal activity during hypercapnia. NEW & NOTEWORTHY The pons is essential for the formation of the three-phase respiratory pattern, controlling the inspiratory-expiratory phase transition. We provide functional evidence of a novel role for the Kölliker-Fuse nucleus (KF) controlling the emergence of abdominal expiratory bursts during active expiration. A computational model of the respiratory central pattern generator predicts a possible mechanism by which the KF interacts indirectly with the parafacial respiratory group and exerts an inhibitory effect on the expiratory conditional oscillator.