UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Respiratory pattern formation in the bullfrog (Rana castebieana) Chatburn, Jonathon William 2004

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2004-0338.pdf [ 7.52MB ]
Metadata
JSON: 831-1.0091584.json
JSON-LD: 831-1.0091584-ld.json
RDF/XML (Pretty): 831-1.0091584-rdf.xml
RDF/JSON: 831-1.0091584-rdf.json
Turtle: 831-1.0091584-turtle.txt
N-Triples: 831-1.0091584-rdf-ntriples.txt
Original Record: 831-1.0091584-source.json
Full Text
831-1.0091584-fulltext.txt
Citation
831-1.0091584.ris

Full Text

R E S P I R A T O R Y P A T T E R N F O R M A T I O N IN T H E  B U L L F R O G (RANA  CASTEBIEANA)  By Jonathon William Chatburn B.Sc. Hon., Bishops University, 2001 A T H E S I S S U B M I T T E D IN P A R T I A L F U L F I L L M E N T O F THE REQUIREMENTS FOR THE D E G R E E OF M A S T E R OF SCIENCE in T H E F A C U L T Y OF G R A D U A T E STUDIES Department of Zoology  We accept this thesis as conforming to thp'required standard  T H E U N I V E R S I T Y O F BRITISH C O L U M B I A A P R I L 2004 © Jonathon W . Chatburn, 2004  Library Authorization  In presenting this thesis in partial fulfillment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission.  Name of Author (please print)  ¥itleofThesis:  Date {dd/mm/yyyy)  ty»r^b*y  Degree:  Year:  Department of  Lc>  The University of British Columbia v Vancouver, BC  Canada  ^/ (  J^Do  </  ABSTRACT This study examined the stability o f the bullfrog in vitro brainstem-spinal cord preparation over time, and the influence of the midbrain on fictive breathing, chemosensitivity, and respiratory pattern formation. Fictive breathing was measured over an 8-hour recording period from both the in vitro brainstem-spinal cord preparation (midbrain intact) and the isolated medulla preparation (midbrain removed). Preparations were exposed randomly to aCSF equilibrated at three different p H levels (8.0, 7.8, 7.6) prior to and following removal o f the midbrain. The fictive breathing pattern was analyzed using Poincare pot distributions following progressive rostro-caudal transections within the midbrain and rostral medulla of the in vitro brainstem-spinal cord preparation. Following setup, the fictive breath frequency progressively declined over the first 2-3 hours, leveled off and stayed constant for the remainder of the recording period in both the in vitro brainstem-spinal cord and the isolated medulla preparations. Lowering the p H o f the a C S F caused significant increases in breath frequency prior to and following removal o f the midbrain, but elicited a larger chemoresponse in preparations without the midbrain intact. Following transection at the midbrain medulla border the fictive breath frequency decreased, the average inter-breath interval (IBI) length increased, and the average breath duration increased. Fictive episodic breathing was not eliminated following removal o f the midbrain, however, the distances between breaths within episodes, and between episodes dramatically increased in size. The spatio-temporal coordination o f breaths within the breathing pattern became less precise following removal o f the midbrain, resulting in episodes that were less discrete and inconsistent in size and occurrence. Preparations that exhibited consistent episodic breathing produced Poincare distributions on which the inter-breath intervals fell into three distinct size groupings. These groupings corresponded to the distance between breaths, episodes, and episode clusters within the breathing pattern. These preparations also produced harmonic distribution patterns on the Poincare plots where the inter-breath interval groupings that occurred were related to each other in size by whole number ratios. These findings suggest that transections through the brainstem-spinal cord initially stimulate breathing, and that there is a period of time following in which neural activity stabilizes. The results also suggest that the midbrain contains a site, or sites that moderate the chemoresponse in  vitro, while independently supplying a source of tonic drive that increases the number of breaths per unit time, and increases the spatio-temporal coordinating o f breaths within the breathing ii  pattern. W e feel that the central pattern generator (CPG) that is responsible for the anuran respiratory pattern continuum could consist o f a multiple-coupled oscillator network. T w o o f the oscillators may lie bilaterally within the medulla, while one lies bilaterally within the caudal midbrain. W e propose that it is the synchronizing interactions o f these oscillators that are responsible for producing discontinuous breathing patterns in which breaths, episodes, and episode clusters occur with precise regularity.  iii  T A B L E OF CONTENTS ABSTRACT  ii  TABLE OF CONTENTS  iv  LIST O F FIGURES  vi  ACKNOWLEDGMENTS  viii  1.0 I N T R O D U C T I O N  1  1.1 Anuran lung ventilation  3  1.1.1 Neuromuscular events and fictive ventilation  5  1.1.2 Central rhythm generation and the respiratory pattern continuum 1.1.3  7  Role o f peripheral and central sensory inputs  1.1.4 Role o f the midbrain in pattern formation 1.2 Conserved adaptive mechanisms and emerging concepts  8 11 13  1.3 Expression o f the pattern continuum in vitro  15  1.4 Hypotheses  17  2.0 M E T H O D S  18  2.1 Animals  18  2.2 Surgical procedures  18  2.3 The bullfrog in vitro brainstem-spinal cord preparation  19  2.4 Measurements o f fictive breathing  20  2.5 Experimental Protocols 2.5.1 2.5.2  20  Stability o f the in vitro preparation over time Influence of the midbrain on fictive breathing and central chemosensitivity  2.5.3  21 21  Effect o f progressive transections on the fictive breathing pattern  22  2.6 Data analysis and statistics  22  3.0 R E S U L T S  25  3.1 Fictive breathing in vitro  25  3.2 Stability o f the in vitro preparation over time  25  3.3 Influence o f the midbrain on fictive breathing and chemosensitivity  28  3.4 Effect o f progressive transections on the fictive breathing pattern  30  3.4.1  Effect o f transection on fictive episodic breathing  34  3.4.2  Influence o f the midbrain in pattern  formation in vitro  38  4.0 D I S S C U S I O N  41 iv  4.1 Stability of the bullfrog in vitro preparation over time  42  4.2 Influence o f the midbrain on fictive breathing and chemosensitivity  43  4.3 Effect o f progressive transections on the fictive breathing pattern  46  4.3.1  47  Effect o f transection on episodic breathing  4.3.2 Influence of the midbrain on pattern  formation in vitro  50  4.4 Multiple-coupled oscillating networks  52  5.0 C O N C L U S I O N S  61  REFERENCES  64  APPENDIX 1  69  v  LIST O F FIGURES  Figure 1.  Schematic diagrams illustrating anuran lung ventilatory mechanics  4  Figure 2.  The anuran breathing pattern continuum  9  Figure 3.  Schematic diagram o f the bullfrog brain indicating brain structures, nerves, and levels of transections  Figure 4.  23  Fictive breathing showing episodes o f variable sizes, recorded from the trigeminal nerve of the bullfrog in vitro brainstem-spinal cord preparation  Figure 5.  Figure 6.  Figure 7.  26  Changes in the fictive breathing pattern over time o f the bullfrog in vitro brainstem-spinal cord preparation  27  Changes in the fictive breathing over time of the bullfrog in vitro brainstem-spinal cord and the isolated medulla preparations  28  The effect o f removing the caudal portion o f the midbrain on fictive breathing  Figure 8.  29  The effect o f lowering p H of the a C S F on the fictive breathing frequency prior to and following removal o f the caudal midbrain  Figure 9.  The effect o f progressive rostral-caudal transections on fictive breathing of the in vitro bullfrog brainstem-spinal cord preparation  Figure 10.  32  The effect o f progressive transections on the fictive breathing pattern of the in vitro bullfrog brainstem-spinal cord preparation  Figure 12.  31  The effect o f progressive rostral-caudal transections on breath duration  Figure 11.  30  33  The fictive episodic breathing pattern and resulting Poincare plot from an in vitro bullfrog brainstem-spinal cord preparation prior to removal of the midbrain  Figure 13.  34  The fictive episodic breathing pattern and resulting Poincare plot from an in vitro bullfrog brainstem-spinal cord preparation following removal o f the midbrain  Figure 14.  35  Poincare plots showing episodic and non-episodic distribution patterns from an in vitro bullfrog brainstem preparation prior to removal o f the midbrain  .:  vi  36  Figure 15.  Poincare plots showing episodic and non-episodic distribution patterns from an in vitro bullfrog brainstem preparation following removal o f the midbrain  Figure 16.  37  The percentage of preparations that exhibited episodic and non-episodic fictive breathing patterns prior to and following removal o f the midbrain  Figure 17.  38  Poincare plot and fictive breathing traces o f an in vitro bullfrog brainstem-spinal cord preparation illustrating interval groupings and episode clustering  Figure 18.  39  Poincare plot and corresponding ficitve breathing traces from an in vitro bullfrog brainstem-spinal cord preparation that exhibited a harmonic distribution pattern  Figure 19.  40  Schematic diagrams illustrating the general location o f the oscillators within the bullfrog brainstem and spinal cord and the synaptic configuration o f the ring circuit  Figure 20.  Diagram illustrating the oscillatory behavior o f a recurrent inhibitory ring-circuit in which the oscillators are arranged in a hierarchal order  Figure 21.  54  55  Diagram illustrating how the ventilatory C P G composed of three oscillators produces motor output patterns that are intrinsically episodic ...57  Figure 22.  Diagrams illustrating how the hypothetical discharge periods produced by the interactions o f a three-oscillator system can produce the anuran respiratory pattern continuum  Vll  60  ACKNOWLEDGEMENTS Right off the top I must thank my supervisor Dr. B i l l Milsom. B i l l , there are many things I am appreciative of.. .1 would like to thank you first for the mayonnaise, the margarine, the coffee, the cream and sugar, the beer and wine, the conferences, ski trips, and dinners. I had a great time. Thanks for taking a chance on a football jock with very little research experience. Thanks for your patience when listening to my philosophical nonsense and half-baked ideas. I learned more from listening, talking, and exchanging ideas with you than any papers I read the entire time I was here. Thank you for taking the time to engage me. I admire the manner in which you conduct both your personal and professional life, and the diligence of your science and your commitment to knowledge. Thanks for being a role model I will spend the rest of my life trying to live up to, though I must admit that the very thought of it utterly exhausts me! I must thank the members o f the M i l s o m lab that have endured part o f this experience with me. The early ones: Beth, Glenn, Charissa, Lisa, K i p , and Angie. The latter ones: Catalina, Andrea, Lieneke, Joanna, and Gladwin. Thank you all for your support and for putting up with my terribly immature sense of humor. Good luck with your future endeavors. Live long and prosper. I must thank m y committee members: Vanessa A u l d and Colin Brauner I would like to thank my family: M o m , Dad, K i m , Stacey, Carlie, Crislee, B i l l , Liam, and Gabe. None of the experiences I had in Vancouver over these last three years would have been possible without your unconditional support, I love you all. I must thank Megan Polk. Your care over these years has been vital to all my successes. Thank you for believing in me when I doubted myself. Thanks for sticking by me and putting up with my peculiarity. I love you. I would like to dedicate this work to my late grandmother, Gertrude Smith. Without her love, financial support, and endless prayers, I may have never made it this far. Thanks Grams.  viii  1.0 Introduction Over time organisms have evolved to survive in an oxygen rich atmosphere that is composed o f roughly 21% oxygen. Arguably, one o f the most vital physiological processes for animal life is the sequestering o f oxygen from the environment in order to power cellular respiration. For oxidative metabolism to proceed, a constant supply o f oxygen and continuous removal o f the waste product, carbon dioxide, are absolutely essential. Accordingly, animals have evolved a variety o f efficient gas exchange mechanisms to adequately meet these demands. The O2 requirements and CO2 production o f an organism increases as a function of mass, but the rate o f gas transfer across the body surface is related to, and limited by its surface area. Through evolution, as animals increased in size and complexity, diffusion alone became insufficient for complete gas exchange prompting the development and elaboration of specialized respiratory surfaces that increased surface area and the rate o f gas transfer. A s size and metabolic rate further increased, simple inactive diffusion o f the respiratory gases across the exchange surfaces and between body tissues also became inadequate. A s a result, the circulatory system developed to efficiently transport respiratory gases between the tissues and the gas exchange organ, along with the respiratory system that enabled active ventilation o f the exchange surface, circumventing stagnation o f the medium and greatly increasing the exchange capacity. It is therefore not surprising that some o f the most significant physiological adaptations allowing animals to exploit a variety of environments and dramatically increase in size, complexity, and diversity, have been in the respiratory system and its control. The vertebrate respiratory system is o f no exception in this regard. When considering the vast array o f vertebrate species, environments they occupy, and remarkable structures by which they accomplish gas exchange, it often appears difficult to draw comparative insight. However, a large number o f commonalties in the respiratory system exist between vertebrate  1  groups, not only superficially in their anatomical structures, arrangement, and functions, but also within the fundamental control systems that govern the respiratory processes. The vertebrate respiratory system functions to maintain relatively constant levels of arterial O2 and CO2 by movement o f water or air over the gas exchange surfaces. It is generally understood that the production, integration and optimization o f the respiratory movements in all vertebrates are the result o f central processing of a wide variety o f sensory inputs from different receptor groups. This "respiratory control system" is thought to generate and modulate the respiratory rhythm, integrating the respiratory movements with other bodily functions, and optimizing the breathing pattern as respiratory demands change. For most mammals, the necessity to maintain arterial gas homeostasis within a narrow range requires continuous rythmicity in the breathing pattern, usually exhibited as uninterrupted inspiration, coupled to immediate expiration. In contrast, reptiles and amphibians, having relativity low metabolic demands, tolerate wider homeostatic ranges, and therefore tend to exhibit respiratory behaviors that are often discontinuous in nature. Discontinuous respiratory behaviors are characterized by breathing patterns in which ventilatory efforts are separated by significant non-ventilatory periods termed apneas. This type of ventilatory behavior is subdivided into categories o f periodic breathing that are described as either episodic or non-episodic. Episodic breathing describes patterns where ventilation occurs in discrete breath clusters with sizable apneas between them, while a non-episodic pattern describes breathing patterns in which breaths occur as single events, each followed by an apneic span. Both episodic and non-episodic periodic patterns are often further classified into rhythmic or arrhythmic episodic patterns, where breaths or episodes either occur with a consistent cadence or are randomly distributed. Although discontinuous ventilatory patterns are most often encountered in amphibians, air breathing fish, and reptilian species, variability in breathing pattern can be seen within most 2  vertebrate groups. Some species offish such as the carp and the bullhead display episodic breathing patterns when metabolic needs are lowered or oxygen availability is high (Shelton et al., 1986). Hibernating mammals such as the golden-mantled ground squirrel often alternate between bouts of episodic breathing and continuous breathing as their metabolic demands are lowered and diving mammals are known to breathe episodically, often on land or during sleep when their metabolism is depressed (Milsom et al., 1996). This similarity in breathing pattern response to alterations in metabolic demands, suggest that vertebrates may share common respiratory control mechanisms. In recent years there has been increasing interest in the adaptive origin of the respiratory control system among vertebrate groups, with specific interest in the neural mechanism generating the respiratory rhythm and the influencing factors that modulate breathing pattern. Anuran amphibians, such as the bullfrog, serve as an ideal model for studying the respiratory control system and the neural origin of pattern generation. From an evolutionary standpoint, ancestral amphibians represent one link between obligate water breathing and terrestrial air-breathing vertebrates. Thus, extant amphibians with semi-terrestrial lifestyles serve in providing valuable insights into the physiological, morphological, and neurological adaptations that were required for the transition to life on land. More importantly, present day anurans display respiratory behavior that is by and large discontinuous in nature. They are, in fact, capable o f exhibiting all breathing patterns, from random single breaths to continuous breathing, making them a superlative animal for studying the neural mechanism that produce and control these phenomena.  1.1 Anuran lung ventilation Gas exchange in adult anurans occurs primarily through lung ventilation, diffusion across the skin, and to a lesser extent through the skin of the buccopharyngeal cavity (Burggren and 3  West, 1982). Anurans typically ventilate their buccopharyngeal cavity continuously and the lungs intermittently, both of which are accomplished by the coordinated muscular events of the glottis, nares and the "buccal force pump" (De Jongh and Gans 1969; West and Jones 1975; and Vitalis and Shelton, 1990). There are two motor output patterns responsible for the ventilatory movements in anurans. The first, termed buccal ventilation, describes the continuous oscillatory movements of the buccal floor while the nares are open and the glottis is closed, and introduces fresh air to the buccopharyngeal cavity only. The second, termed lung ventilation, periodically interrupts buccal oscillations and serves to empty and introduce fresh air into the lungs. Figure (1A) illustrates the general airflow and structures involved in lung ventilation of anuran amphibians.  DEFLATION  B. (ml)  5.0-  Flow (mis")  Out 2.50In 2.5,. CloseOpen.:. CloseOpen-I l l !  1S  (Pa)  Figure 1. (A) Diagrammatic cross section of a bullfrog showing the general airflow and structures involved in lung ventilation (adapted from de Jongh and Gans, 1969). (B) Lung ventilation of Rana pipiens. Diagram showing the volume changes and airflow at the nares, and pressure changes in the buccal cavity and lungs during the four phases of an inflation and deflation breath cycle (Vitalis and Shelton, 1990).  4  There are multiple types of lung ventilatory motor patterns exhibited by anurans. The typical ventilatory cycle most often displayed by a resting, unrestrained animal is the balanced lung breath (deJongh and Gans, 1969; Vitalis and Shelton, 1990). This breath cycle is characterized by a larger than normal inhalation of air into the buccopharyngeal cavity by expansion of the buccal floor after which the glottis opens allowing the lungs to partially deflate. The nares then rapidly close and the lungs are forcibly inflated by compression of the buccal floor. The glottis then closes prematurely, trapping the remaining air in the lungs, and maintaining them inflated with a roughly similar volume of air to that expired. A n inflation breath can result when the buccal pump is contracted earlier in this cycle allowing less time for air to escape the lungs, trapping a larger volume of air in the lungs than is expired.  Conversely, a lung deflation breath results when the buccal pump contracts less  vigorously, the nares remain open longer, and glottal closure is slightly delayed in this phase allowing more air to escape than is forced back into the lungs. Under conditions of elevated respiratory drive, multiple inflation breaths may occur in rapid succession without allowing sufficient time for the lungs to empty. This event results in a progressive increase in air pressure and volume within the lungs and is termed a lung inflation cycle (West and Jones, 1975; Macintyre and Toews, 1976; Viltalis and Shelton, 1990; Sanders and Milsom, 2000). Likewise, a lung deflation cycle consists of a series of multiple deflation breaths where the timing of glottal closure and buccal elevation allows more air to escape the lung during each breath than is pumped back in, resulting in a progressive decrease in pressure and volume (Vitalis and Shelton, 1990).  1.1.1 Neuromuscular events andfictiveventilation The variety of breath types and ventilatory patterns exhibited by anurans reflects the precise coordination of the neuromuscular events associated with the buccal force pump, nares, 5  and glottis (West and Jones, 1975; Vitalis and Shelton, 1990). The buccal depressors are responsible for expansion of the buccal floor and are innervated by the sternohyoid branch of the hypoglossal nerve (Hsh). The buccal levetors facilitate compression of the buccal floor and are innervated by the mandibular branch of the trigeminal nerve (Vmd) and the main branch of the hypoglossal nerve (Hm). The muscles that facilitate movements of the glottis are the glottal dilators and constrictors, both innervated by the laryngeal branch of the vagus nerve (XI) (Sakaibara, 1984; Kogo et al. 19941). Balanced breaths, lung inflation breaths, and lung deflation breaths are composed of four slightly different neuromuscular phases, the synchronization of which defines the type of breath cycle. Figure (IB) illustrates the temporal arrangement of the neuromuscular phases with the mechanical events and resulting pressure and airflow changes in a deflation and inflation lung breath cycle. During typical lung ventilation, in phase 1 (buccal inspiration) discharge from Hsh activates the buccal depressors causing the floor of the buccal cavity to lower and draws air in through the nares, while firing of XI activates the glottal constrictors sealing the off the lungs. In phase 2 (lung emptying and expiration 1), while the nares remain open, discharge from XI causes glottal dilation allowing the air in the lungs to escape and the pressure between the lungs and the buccal cavity to equalize. In phase 3 (lung filling), while the glottis is still open and the nares are closed, discharge from H m , and V m d causes buccal elevation, forcing the air from the buccal cavity into the lungs. In phase 4 (expiration 2), discharge from XI causes activation of the glottal constrictors, trapping the air and maintaining lung inflation while the remaining air in the buccal cavity passes out the nares. Ultimately, it is the volume of air expired and subsequently pumped backed into the lungs during the second and third neuromuscular phases of this cycle that determines whether the breath is inflation, deflation, or balanced in nature (Vitalis and Shelton, 1990).  6  Neural discharge can be recorded from the nerves innervating the respiratory muscles and has been shown to closely coincide with the muscular and mechanical events responsible for buccal and lung ventilation in the intact animal (Sakaibara, 1984; Kogo et al. 19941). When this activity is recorded from the  in situ paralyzed, decerebrate, unidirectionally ventilated ( U D V )  animal, or the in vitro isolated brainstem-spinal cord preparation, it is considered to represent "fictive attempts" at ventilatory movements (deJongh and Gans, 1969; Sakakibara, 1984). The neural discharge pattern associated with buccal oscillations occurs as high-frequency, lowamplitude reciprocating bursts from V m d and H m , while the neural discharge associated with lung ventilation occurs as low-frequency, high-amplitude, simultaneous bursts from V m d , H m , and XI (Kogo et al. 19941). While the muscle groups recruited to produce buccal oscillations and lung breaths are essentially the same, the corresponding neural discharge patterns are considered to be distinct ventilatory rhythms, and are thought to arise from separate but coordinated rhythm generators (Wilson et al., 2002).  1.1.2 Central rhythm generation and the respiratory pattern continuum Rhythm generation responsible for the production of the motor output underlying buccal and lung ventilation in anurans appears to originate in a region located in the rostral medulla (Torgenson et al. 2001a; Kogo et al., 1994; M c L e a n et al., 1995a; Perry et al., 1995; Reid et al., 2000). This location has been suggested to contain at least two-bilateral oscillating neural networks. The more rostral site, located between the VHIth and LXth cranial nerves, appears to initiate lung ventilation, while the more caudal region, in the area o f the X t h cranial nerve, appears to be responsible for initiation of buccal oscillations (McLean et al., 1995b; Wilson et al., 2002). Whether these sites constitute anatomically distinct locations or are part of a large diffuse network is not yet clear. They do however appear to be coupled and their oscillating activities are highly coordinated (McLean et a l , 1995b; Kimura et al., 1997; Wilson et al., 2002). 7  Thus, the motor output responsible for the overall respiratory pattern in anurans appears to be a consequence of the interaction between two endogenous rhythms that, in turn, are produced by two connected oscillating neural networks. The normal pattern of lung ventilation exhibited by resting anuran amphibians consists of randomly distributed single or doublet breaths. When drive is slightly elevated due to hypoxia or hypercapnia, the frequency o f lung ventilation increases and a pattern of evenly spaced single breaths emerges. When drive is further increased, evenly spaced breaths give way to distinct multi-breath clusters (episodes) that are separated by considerable apneic spans (Boutilier and Towes, 1976; Kinkead and M i l s o m 1994). If drive is raised enough, the pattern eventually shifts to high frequency continuous lung ventilation with very few apneas that are brief in duration (Milsom, 1991; Kinkead, 1997). This range in expression of the respiratory rhythm in anurans represents a breathing pattern "continuum" in which breathing can span from no or little breathing at one end, to episodes of variable length, to continuous lung ventilation as respiratory drive progressively increases. Figure (2) illustrates the expression of the anuran respiratory pattern continuum in response to progressively increasing tonic drive. The selective expression of the respiratory rhythm within this continuum is highly influenced by peripheral and central sensory inputs, affecting both the type of breath exhibited and the overall breathing pattern (Kogo et al., 1994; Kinkead, 1994; Kinkead and Milsom, 1997; Reid et al., 2000; Sanders and Milsom, 2001). The ability to precisely manipulate lung ventilatory frequency and produce numerous respiratory patterns suggests that anurans possess a well-developed respiratory control system that responds to changes in metabolic demands.  1.1.3 Role ofperipheral and central sensory inputs There are a number of sensory inputs that contribute to the overall drive to breath in anurans. CC^-sensitive pulmonary stretch receptors within the lung relay information, via the 8  Figure 2. Traces illustrating expression of the anuran respiratory pattern continuum under progressively increasing tonic drive recorded in situ from the trigeminal nerve rootlet (V). (A) Buccal oscillations with no lung ventilation. (B) Evenly-spaced single lung breaths. (C) Single and doublet lung breaths. (D) Small episodes of variable length. (E) Large episodes of variable length. (F) Continuous lung ventilation.  9  vagus, to the central nervous system indicating the degree o f lung inflation and its gas composition (Kogo et al., 1994; Kinkead and Mislom, 1994; Kinkead and Milsom, 1997; Wang et al., 1999; Reid et al., 2000; Sanders and Milsom 2001). There are peripheral chemoreceptors in the carotid labyrinth and the pulmocutaneous arch that provide information about blood gas fluctuations (Ishii et al., 1985), while central chemoreceptors, thought to be distributed throughout the brainstem in the area of the fourth ventricle, supply information on the acid-base status of the cerebrospinal fluid (Branco et al. 1991; Smatresk and Smits, 1991; Branco et al., 1993; Torgerson et a l , 1997). It is now well understood that in anurans i f the respiratory drive is sufficiently reduced by maintaining blood gases near normoxic or at hypocapnic levels, no breathing will occur (West et al., 1987; Kinkead and Milsom, 1994). It is not until O2 levels are allowed to drop or CO2 levels increase that breathing is stimulated. Then, depending on the level o f drive the breathing pattern can range from random single and doublet breaths to episodes of variable length. It was formerly thought that fluctuations in blood gas levels were solely responsible for the drive to breath in these animals; it was the combination o f central and peripheral afferent feedback alone that determined the breathing pattern. Episodic patterns were therefore thought to be the result of phasic receptor feedback that stimulated lung ventilation when O2 levels fell and CO2 levels in the blood increased, and inhibited ventilation when normal blood gas levels were restored. If, however, pulmonary pressure and blood gas fluctuations are prevented, providing only a tonic Tevel of afferent feedback, the full repertoire of breath types and the entire pattern continuum can still be produced (West et al., 1987; Kinkead and Milsom, 1997; Sanders and Milsom, 2001). This ability to promptly alter both the breath type and breathing pattern in response to changes in tonic respiratory related drive again suggests that anurans possess an advanced respiratory control system. More importantly, it implies that the respiratory pattern continuum is an intrinsic part of the respiratory control system itself, and that the selective expression of a given pattern 10  within the continuum is influenced by the total respiratory drive, not simply the phasic receptor feedback associated with each ventilatory effort. Presently, there is substantial understanding o f the respiratory processes and ventilatory mechanics in anurans amphibians. However, far less is known about the neural mechanisms governing the respiratory control system or the means through which central and peripheral sensory inputs are integrated to modulate the breathing pattern. The advent o f reduced preparations, such as the  in situ decerebrate, unidirectionally ventilated ( U D V ) animal, and the in  vitro isolated brainstem-spinal cord preparation, have made it possible to study the peripheral and central components o f the respiratory control system independently, and have provided valuable insights into the neural mechanism responsible for both rhythm generation and respiratory pattern formation in these animals.  1.1.4 Role of the midbrain in pattern formation Numerous studies have suggested that the centers producing the endogenous respiratory rhythm responsible for the generation o f ventilatory motor output in anurans are located within the rostral medulla (Langendorff, 1887; Kinkead et al., 1994; K o g o et al., 1994; M c L e a n et al., 1995; Reid and M i l s o m 1998). It has also been suggested that respiratory pattern formation is the result o f descending inputs from structures within the midbrain acting on the rhythm generating centers within the medulla (Milsom et al., 1999; Reid et al., 2000; Gargaglioni et al., submitted). The midbrain in anuran amphibians, as in all vertebrates, contains regions responsible for processing sensory information that allow integration and coordination o f multiple organ systems in order to produce the appropriate responses. M a n y o f these responses require modulation o f the breathing pattern, and not surprisingly a number o f regions in the midbrain have been found to exert influence over the respiratory rhythm. Recent in vitro and in  situ studies have identified areas in the midbrain of the bullfrog (Rana catesbeiana) that influence buccal and lung ventilation, and have also suggested that sites within the midbrain are 11  responsible for production of episodic breathing patterns (Reid et al., 2000; Gargaglioni et al., submitted). In a study in which the midbrain was removed  in vitro, at a level just rostral to the optic  nerve roots, expression of the episodic pattern was compromised; transforming it to one of evenly spaced breaths with increased frequency (Milsom et al., 1997; Reid et al., 2000). The same transection  in situ produced similar results, however in this case the overall breath  frequency was reduced and amplitude was increased (Gargaglioni et al., submitted). In both studies episodic patterns reappeared over time or were induced by manipulation of respiratory drive (Reid et al., 2000; Gargaglioni et al., submitted). When transections were made at the optic tectum-medulla border, removing the midbrain completely, the episodic pattern was irreversibly altered (Oka, 1958; Reid et al., 2000; Torgersson et al., 2000; Gargaglioni et al., submitted). The remaining respiratory discharge produced by the isolated medulla occurred as relatively evenly spaced breaths that were lower in frequency and longer in duration, a pattern that was assumed to be the intrinsic cadence of the rhythm generating centers in the medulla (Reid et al., 2000). These data suggest that sites within the midbrain not only provide tonic respiratory drive that influences breath frequency and modulates breath shape but are also responsible for producing episodic breathing patterns. It was concluded that sites within the caudal portion of the midbrain likely provided positive and negative inputs to the rhythm generating centers in the medulla in an alternating fashion that turned breathing on and off to produce episodes (Reid et al., 2000). The nucleus isthmii are mesencephalic structures in the anuran brain, located bilaterally between the roof o f the midbrain and cerebellum (structurally similar to the pons in mammals) that have been implicated as centers involved in providing the modulation that produces episodes. In the bullfrog, bilateral chemical lesioning of this area  in situ changed the episodic  pattern to one of evenly-spaced breaths concomitant with a reduction in both the overall frequency of breathing and the responsiveness of the preparation to increases in CO2 related 12  drive. Episodic patterns did return, however, when respiratory drive was significantly increased (Kinkead, et al. 1997). These findings suggested that the nucleus isthmii are not solely responsible for producing episodic patterns but provide tonic drive and serve to enhance or integrate chemoreceptor information and facilitate episode formation. A s mentioned, episodic breathing patterns are part of a larger respiratory pattern continuum that spans from occasional single breaths, to episodes, to continuous breathing. Expression o f a given pattern within the continuum is highly dependent on the overall level o f drive and the balance between mechanoreceptor and chemoreceptor input (Kinkead and Milsom, 1994; 1996; 1997). The bullfrog  in situ preparation responds in a similar fashion to the intact  animal to increases in respiratory drive by increasing the overall breathing frequency via increasing the frequency o f episodes, and number o f breaths per episode (Kinkead et al., 1994; Kinkead and Milsom, 1996 1997). However, when the midbrain is removed in situ, changes in lung volume (vagal input) or chemoreceptor input no longer elicit these responses (Gargalioni et al., submitted). Vagal feedback has been shown to be a large source of drive for pattern modulation in these animals, enhancing both chemosensory responsiveness and episode formation considerably (Kinkead et al., 1994; Kinkead and Milsom 1994; 1996). While the afferent pathways that relay vagal input enter through the medulla, it appears that their full expression requires the midbrain. This suggests that these afferent pathways ascend to the midbrain and then descend into the medulla to influence the rhythm generating areas, as they do in mammals, or require modulation by the midbrain to be properly expressed (Milsom et al., 1999).  1.2 Conserved adaptive mechanisms and emerging concepts Although vertebrate species possess different metabolic requirements and a variety o f structures through which respiration is accomplished, there appears to be a high degree o f 13  homology within the vertebrate respiratory control system and neural mechanism that govern it (Milsom et al., 1997). Recent studies involving isolated brainstem spinal-cord preparations have revealed considerable similarities between mammals and anurans, both structurally and mechanistically, with regards to the origins of respiratory rhythm and pattern formation. It is currently thought that respiratory rhythm generation in mammals is initiated within a neural network localized to the rostral medulla (Smith et al., 1991), and, similar to amphibians, sites within the midbrain of mammals have been suggested to modulate aspects of this respiratory rhythm (Paydarfar and Eldridge, 1987). Sites within the midbrain and rostral medulla, such as the pons, have been found to provide tonic input to more caudal regions, influencing both breathing frequency and amplitude (Hilaire et al., 1989; 1997). The cumulative data from previous studies in the bullfrog have implied that the neural components required to generate episodic breathing patterns are located in the caudal midbrain, and that the medulla lacks the capability to produce them when isolated (Oka, 1958; Reid et al., 2000; Gargaglioni et al., submitted). However, the most recent works with the neonatal rat in vitro preparation have also shown that the isolated medulla, during recovery from hypothermia, is capable of generating bouts of episodic breathing without the midbrain and pons present implying that inputs from regions rostral of the medulla are not entirely necessary for episode formation and may only serve an integrative or modulatory role in these animals (Mellen et al., 2002; Zimmer and M i l s o m 2004). If we assume that episodic breathing is a conserved adaptive trait in the vertebrate respiratory control system, and that the neural mechanisms that produce the respiratory rhythm and modulate breathing pattern are homologous in structure and function throughout, we then need to suggest an alternative scenario. Is it possible that, similar to mammals, sites in the anuran midbrain only supply a source of tonic drive, or the descending input required to convey and enhance sensory feedback that is essential to modulate the breathing pattern? Could the isolated medulla of anurans contain 14  mechanisms that generate episodic breathing but is simply unable to express it because sensory inputs are missing, or the overall drive level is insufficient to promote a shift in the pattern continuum?  1.3 Expression of the pattern continuum in vitro In the freely behaving animal, central chemoreception provides the major source o f respiratory drive and is highly influential in pattern formation in situ, when peripheral inputs are intact (Branco et al. 1991; Smatresk and Smits, 1991; Branco et al., 1993). The bullfrog in vitro brainstem-spinal cord, with the midbrain intact, also has central chemoreceptors intact but is completely devoid o f all peripheral input. In the absence o f exogenous respiratory drive (hyperoxic superfusate) it is still capable of generating random single breaths, evenly-spaced single breaths, and episodes o f variable length. Changing chemoreceptor drive through altering the p H / C 0 2 o f the superfusate does not elicit extremely large changes in breath frequency or pattern (Kinkead et al., 1994; MacLean et al., 1995: Reid et al., 2000), suggesting that central chemoreception is much less o f a source of drive for pattern modulation  in vitro.  In anurans, episodic breathing patterns are associated with relativity high levels o f respiratory drive. It would be expected that in the absence o f peripheral input and with reduced central chemoreceptor drive, little or no fictive breathing should occur at all  in vitro. This of  course is not the case and implies that the brainstem-spinal cord itself contains some minimal amount o f intrinsic drive that promotes fictive breathing. This source o f extraneous drive is unknown, although it has been suggested that the ionic composition o f the superfusate may be a potential source o f excitatory drive (Milsom et al., 1999). This has yet to be investigated.  It is  also possible that this source o f excitatory drive reflects an artifact o f the in vitro brainstemspinal cord preparation itself, resulting from transections or the elimination o f peripheral inputs.  15  In producing the bullfrog brainstem-spinal cord preparation cranial nerves are severed near their exit point from the skull and transections are required to remove the lower spinal-cord and rostral midbrain and forebrain regions. There is little doubt that transections through the brain destroy a large number of neurons and axons o f passage causing massive local depolarization. Transections could therefore act as excitatory input to sites upstream or downstream of the cut and hence produce a shift in the breathing pattern exhibited by the preparation. Cranial nerves not only relay efferent motor output to muscles, but afferent sensory information back to the central nervous system as well. Pulmonary vagotomy has been shown to have a significant effect on breathing frequency and amplitude in the bullfrog (Kinkead and Milsom, 1996; 1997). Similar to transections, removal of such afferent pathways may not simply eliminate peripheral sensory input, but could possibly cause transient over-excitation o f the pathways themselves, acting as excitatory or inhibitory inputs to sites that modulate breathing and thus promote a shift in the pattern continuum. The Active breathing pattern that is produced by the isolated medulla in vitro occurs as evenly-spaced breaths that are relativity low in frequency (Reid et al., 1999). A s mentioned, the in vitro preparation lacks peripheral inputs and as a result does not respond particularly well to changes in respiratory related drive (Kinkead et al., 1994). If a transection that removes the midbrain potentially causes excitation to regions downstream o f the cut, it is unclear whether the breathing pattern produced by the isolated medulla is the result o f removing the neural mechanisms essential for the production o f episodic patterns, sites that provide tonic drive and are required to enhance pattern modulation, or are the result of the transection itself. The endogenous respiratory rhythm produced by the isolated medulla and the influence of the midbrain in pattern formation have not been studied extensively in anurans; hence the overall objective o f this study was to define the role that the midbrain plays in pattern 16  modulation with three questions in mind: 1) How does the respiratory motor output of the bullfrog in vitro brainstem-spinal cord, and the isolated medulla preparations change over time? 2) H o w does the midbrain in anurans influence central chemosensitivity in vitro?  3) H o w does  the midbrain influence respiratory pattern formation in these animals? Anurans typically breathe discontinuously, therefore analysis of breathing in these animals usually considers the production of the respiratory rhythm, the breathing pattern and the burst (or breath) pattern. Burst pattern formation refers to the manner in which the motor output is turned on and off with each breath, constituting the overall shape of the individual respiratory effort. The respiratory rhythm refers to the manner in which the endogenous rhythm produced by the oscillating networks in the medulla is expressed to produce the uniform cadence with which breaths occur within an episode. The breathing pattern then, refers to the manner in which this rhythm is selectively expressed (turned on and off) to produce the overall motor output to the respiratory muscles. The ultimate result of these separate components, influenced by the total respiratory drive is what produces the variety of breathing patterns exhibited by these animals. This study was designed to investigate how the midbrain influences each of these components to produce the anuran breathing pattern continuum.  1.4 Hypotheses 1. The transections required to produce the bullfrog in vitro brainstem-spinal cord and isolated medulla preparations temporarily provide an excitatory stimulus to sites down-stream of the cut that alter the fictive breathing pattern. Such effects are only present following transection, and since transient, will subside over time.  2. Sites within the caudal midbrain enhance central chemosensitivity. will reduce the chemo-responsiveness of the preparation. 17  Removal of these sites  3. The mechanisms that generate episodic breathing patterns are intrinsic to the medulla of the bullfrog. Following isolation, the medulla is still capable of generating fictive episodic breathing with time.  2.0 Methods 2.1 Animals Experiments were performed on 32 adult American bullfrogs  (Rana catesbeiana), mass  (303 ± 2 1 g ) of undetermined sex that were obtained from a commercial supplier (Boreal) through the University of British Columbia. Animals were maintained indoors in fiberglass tanks supplied with a constant flow of dechlorinated water at 1 8 - 2 0 ° C and free access to heating lamps and basking platforms. Photoperiod was maintained at 12 hours o f light and 12 hours o f dark. Animals were fed live crickets twice per week. Food was withheld for at least two days prior to experiments.  2.2 Surgical procedures Bullfrogs were anesthetized by immersion in an aqueous solution of ethyl-maminobenzoate ( M S 222;  0.5gr\  Sigma St. Louis. M O . U S A ) until eye blink and toe pinch  responses were eliminated. Using scissors, a midsaggital incision was made in the skin of the head from between the eyes to just anterior to the shoulder blades. Using a D r e m m e l ™ tool a small hole was drilled into the skull just rostral to the optic lobes. Full access was gained to the cranial cavity using ronguers to clip away excess bone and expose the caudal forebrain. Using small micro-scissors the brain was transected between the forebrain and optic lobes, and the entire forebrain was sucked out of the cranium by aspiration. The frontal cranium was packed tightly with cotton covered in Vaseline to prevent excess cerebro-spinal fluid (CSF) or blood 18  loss. The hole in the cranium was then sealed with dental dam affixed to the top of the skull with cyanoacryleate glue. The skin was closed with sutures and the animals were placed into an empty but moist holding tank and monitored until full recovery from anesthesia. Following a 24hour recovery period all animals were spontaneously active.  2.3 The bullfrog in vitro brainstem-spinal cord preparation The in vitro brainstem-spinal cord preparation was prepared in a similar fashion to that described by M c L e a n et al. (1995a). Experiments were performed at least 24 hours post-surgery at which time the animal was weighed and placed in crushed ice for approximately one hour to reduce metabolic rate. Upon removal from the ice the animal was spinalectomized and the brain case was removed, placed in a Sylgard coated perfusion tray, and superfused continually with oxygenated artificial cerbral-spinal fluid (aCSF). The brainstem and upper spinal cord were removed from the braincase by dissecting the cranial nerves (CN) close to their exit points from the skull and severing the spinal cord just caudal to the second spinal nerve ( S N II). The brainstem-spinal cord was then transferred to a recording chamber where the meninges (dura, and arachnoid) were removed to free the cranial nerve roots. The midbrain was then transected through the optic tectum at the level of the third cranial nerve ( C N III) to allow access of the a C S F to the third and fourth ventricles. The brainstem-spinal cord was then pinned ventral side up on a thin stainless steal mesh within the recording chamber, and continually superfused with oxygenated a C S F delivered at a rate of approximately lOml/min, maintained at a p H of 7.8 and temperature of 1 8 - 2 0 ° C unless otherwise indicated. A p H of 7.8 was selected as it approximates the p H of amphibian plasma at these temperatures (West et al., 1987; M c L e a n et al., 1995a) The artificial cerebrospinal fluid (aCSF) used was similar to that described by Kinkead et al. (1994), and consisted of a solution of distilled water with the following ionic concentrations: (in mmol I T ) N a C l , 75.0; KC1, 4.5; M g C l , 1.0; N a H P 0 , 1.0; N a H C 0 , 40.0; C a C l , 2.5; 2  2  19  4  3  2  glucose, 5.0; p H 7.8. The p H o f the superfusate was monitored continuously using a commercial digital p H meter (Hana Instruments) and controlled by altering the amount o f and CO2 gas bubbled through the a C S F reservoir by a digital flow meter (Cameron Instruments). Temperature within the bath was monitored with a thermocouple (Sensortek) and maintained by passing the perfusate through a counter-current exchanger (Lauda, model R C 6 ) just prior to it entering the recording chamber.  2.4 Measurements offictivebreathing Utilizing a micromanipulator, a glass suction electrode fashioned from thin walled capillary glass (inner diameter 1mm) was positioned over one o f the severed tips of either the mandibular branch o f the trigeminal (Vmd), the laryngeal branch o f the vagus (XI), or the main branch ( X l l m ) o f the hypoglossal nerve rootlet. The severed rootlet was gently aspired into the glass suction electrode creating a tight seal around the nerve in order to record whole-nerve motor discharge. Nerve activity recorded from the suction electrodes was amplified [filter setting: 200Hz (high pass) and 1kHz (low pass)], full waved rectified and integrated for subsequent analysis. Nerve signals were monitored visually using an oscilloscope (Tektronic 5111 A ) and acoustically with an audio monitor (Grass A M 8 ) . The electroneurogram was recorded on a computer using commercial software (Windaq v l .32 data acquisition system, DI720. D A T A Q Instruments) sampling at a frequency o f 2000Hz.  2.5 Experimental protocols 2.5.1 Stability of the in vitro preparation over time Two series o f experiments were designed to determine the robustness and stability o f the bullfrog in vitro brainstem-spinal cord and isolated medulla preparations over time. The first series (n=6) was conducted using the in vitro bullfrog brainstem (caudal to C N III) and rostral 20  spinal cord (transected caudal to C N XII) preparation, superfused with a C S F at a p H o f 7.8 and temperature o f 1 8 ° C . Once stable levels o f neural discharge were observed from a preparation, approximately 10 minutes, motor output was recorded for a period o f 20 hours. The second series was conducted using both the in vitro brainstem-spinal cord (n=9), and the isolated medulla preparation (n=6), where the midbrain was removed by a transection between the optictectum and cerebellum prior to securing it in the recording chamber. Because no changes were observed after the first 8 hours o f recording in the first series o f experiments, motor output was recorded under similar experimental conditions (pH 7.8, 1 8 ° C ) for a period o f only 8 hours.  2.5.2 Influence of the midbrain on fictive breathing and central chemosensitivity Two series o f experiments were designed to investigate the effect of removal of the midbrain on fictive lung ventilation and the central chemosensitivity o f the in vitro brainstemspinal cord preparation. In the first series, preparations (n=9) were exposed to a C S F equilibrated to a p H o f 7.8 for approximately 60 minutes. A transection was then made between the optic tectum and cerebellum, isolating the medulla, in order to compare fictive breathing prior to and following removal of the midbrain in the same preparation. After stable neural activity was observed from the isolated medulla, motor output was recorded for an additional 60 minutes. In the second series, in vitro brainstem-spinal cord preparations (n=5) were exposed to a C S F equilibrated to three different p H levels (7.6, 7.8, and 8.0), which were determined by changing the gas composition bubbled through the a C S F reservoir as follows: 2.0%, 5.0%, 7.0% CO2, balance O2. A l l preparations were exposed to each p H level in random sequence for 30 minutes. Following this, a transection was made between the optic tectum and the cerebellum removing the midbrain from the medulla, and after stable neural activity was observed, the preparations were re-exposed to the same p H levels, again in random sequence for 30 minutes each.  21  2.5.3 Effect ofprogressive transections on the fictive breathing pattern This series o f experiments was designed to determine the effect o f multiple progressive transections through the brainstem-spinal cord on the fictive breathing pattern  in vitro.  Commencing with the in vitro brainstem-spinal cord, a series o f progressive transections were made to each preparation (n=8), as illustrated in Fig. 3B, moving in a rostral-caudal direction. The first transection was made between the optic tectum and cerebellum, removing the midbrain and isolating the medulla (midbrain O F F ) . The second transection was made at a level just caudal to the trigeminal nerve rootlets ( C N V O F F ) , and the third transection was made at a level just caudal to the acoustic nerve rootlets ( C N VIII O F F ) . Once stable neural discharge was observed following a transection, motor output was recorded for 2 hours under constant conditions (pH 7.8, 18°C) before the next transection was made.  2.6 Data analysis and statistics In some preparations, high frequency, low amplitude bursts o f neural activity reflecting fictive buccal oscillations were present, however, all preparations exhibited low frequency, high amplitude bursts o f neural activity representing fictive lung ventilation (Kinkead et al., 1994; M c L e a n et al., 1995). Neural activity was considered a fictive lung breath if the integrated burst amplitude was greater than twice that o f the average buccal burst amplitude. The absolute breath frequency was calculated as the total number o f fictive lung breaths occurring within a given period o f analysis. Lung burst duration was measured from the onset o f each fictive breath at the baseline to the termination o f the breath at the baseline of the integrated neural trace. Lung burst amplitude was measured from the baseline o f the integrated neural trace to the peak o f each individual fictive breath.  Forebrain  Midbrain  Hindbrain  I  I  I  a  b  e  I  d  Figure 3. Schematic diagram of the bullfrog brainstem-spinal cord showing a dorsal view (A) and lateral view (B). In both figures roman numerals indicate location of the cranial nerves (CN). Brain structures appear on the left of the dorsal view (A). Brain regions are indicated above the lateral view (B). In the lateral view (B) the dashed lines indicate the levels at which the brainstem was transected: (a) just caudal to C N III, (b) optic tectum-medulla border, (c) just caudal to C N V , and (d) just caudal to C N VIII.  23  Integrated area o f the burst was the area under the integrated neural trace measured from the The inter-breath interval was measured from the peak o f one fictive lung breath to the peak o f the following fictive lung breath. In the bullfrog in vitro brainstem-spinal cord preparation with the midbrain intact the resting breathing pattern can be extremely variable often exhibiting a consistent episodic pattern less than 50% o f the time (Reid and Milsom, 1998). Given the inconsistency in pattern between any two in vitro preparations, or in any one preparation over time, fictive breathing patterns were analyzed graphically using Poincare plots. Preparations were designated as either episodic or. non-episodic, subjectively based on the collective distribution pattern o f inter-breaths intervals that were measured over a 2-hour recording period. A description o f this analysis, and pattern classification based on the Poincare plots are provided in Appendix 1. The data are presented as mean values ± 1 standard error o f the mean ( S E M ) . Mean values for fictive breathing variables for each preparation were used to calculate group means for each time period analyzed. Results were statistically analyzed using one-way or two-way repeated measures A N O V A ( P O . 0 5 ) followed by Dunnett's multiple comparison tests. The values before and following transections were compared using paired /-tests. When parametric test assumptions were violated the data were analyzed using a Kruskal-Wallis A N O V A on ranks test followed by the Dunnett's multiple comparison method or signed ranks Mest. In the presentation o f fictive breathing variables, an asterisk (*) denotes a significant difference from the control group. A plus sign (+) denotes a significant difference for the values before transection compared with values following transection. A l l statistical testing, including determinations o f normality, were performed using commercial software (Sigmastat; Jandel Scientific, 1997).  3.0 Results 3.1 Fictive breathing in vitro  Figure 4 illustrates the resting fictive breathing pattern of a bullfrog in vitro brainstem-spinal cord preparation recorded from the trigeminal nerve rootlet (V) at pH 7.8. This pattern is similar to breathing patterns that can be observed in less reduced preparations and the intact animal under conditions of similar drive (Kinkead and Milsom 1996; 1997). Figure 4A illustrates a trace of fictive breathing over a 60 second time scale, demonstrating the inconsistency in size and occurrence of discrete episodes that frequently occur in vitro. Figure 4B illustrates the temporal relationship of the fictive breaths occurring within an episode over a 20 second time scale. The ramp-like bursts of incrementing and decrementing motor output, indicative of an individual fictive lung breath, are illustrated in figure 4C.  3.2 Stability of the in vitro preparation over time  Experiments were designed to examine the robustness of the bullfrog in vitro brainstemspinal cord and isolated medulla preparations over time. Figure 5 A illustrates the changes that occurred in the fictive breathing pattern of one preparation over a 20-hour recording period. There was a significant reduction in the overall fictive breath frequency following the setup of the in vitro brainstem spinal-cord preparation, as shown for all preparations (n=6) in figure 5B. In the first four hours of recording the overall frequency progressively declined from 14.87 ± 4 . 1 breaths/min, reaching a minimum of 4.5 ± 1 . 6 breaths/min. Frequency then leveled off at 6.5 ± 2.8 breaths/min and remained stable for the following 14 hours of recording (PO.05).  25  A.  f  ^n^»|iii.M  mi ^"f'fff""! f"f P"f^ " f f >l  >  >  l  B  Figure 4. (A) A n example of fictive breathing showing episodes of variable size recorded from the trigeminal nerve root (V). (B) A n expanded trace illustrating the temporal relationship of breaths within the breathing pattern. (C) A further expanded trace illustrating the incrementing and decrementing shape of the neural discharge associated with each fictive lung breath. In each panel, upper trace represents the raw signal while the lower trace represents the integrated signal  (fv).  Figure 6 A illustrates the changes that occurred in the fictive breathing pattern of one in vitro brainstem preparation with the midbrain removed (isolated medulla) over an 8-hour recording period. Figure 6 B shows the effect of time on the overall fictive breathfrequencyof the in vitro brainstem-spinal cord preparation with the midbrain intact (midbrain ON) and 26  removed (midbrain O F F ) over an 8-hour recording period. In preparations with the midbrain on, fictive breath frequency progressively declined over the first three hours, from 14.87 ± 3.4 breaths/min to 6.4 ± 2 . 1 breaths/min, then leveled off and remained stable for the following 5 hours of recording (P<0.05). In preparations with the midbrain off, the overall breath frequency progressively declined within the first two hours, from 6.1 ± 1.5 breaths/min to 2.0 ± 0.8 breaths/min, leveled off and remained stable for the following 6 hours o f recording (P<0.05). The trend for frequency to initially decline following setup was similar in both preparations, however, the change in frequency over this time was larger in preparations with the midbrain on. Although not significant here, the fictive breath frequencies over the entire recording periods were consistently lower in preparations with the midbrain removed (Fig. 6B).  0  10  20  30  40  50  60  70  80  90  100 110 120  Tirre(Sec)  "  Time(Hrs)  Figure 5. Changes infictivebreathing over time in the bullfrog in vitro brainstem-spinal cord preparation. (A) Traces illustrating the fictive breathing pattern at different times over a 20-hour recording period. (B) Changes in the overall fictive breathing frequency (breaths/min) at each hour over the 20-hour recording period. The data are shown as the mean ± 1 S.E.M. An asterisk (*) denotes a significant difference from the first hour of recording.  27  Figure 6. Changes in fictive breathing over time in the bullfrog in vitro brainstem-spinal cord (midbrain ON) and the isolated medulla (midbrain OFF) preparations. (A) Traces illustrating the fictive breathing pattern of the isolated medulla at different times over an 8-hour recording period. (B) Changes in the overall fictive breathing frequency (breaths/min) of both preparations over a 8hour recording period. The data are shown as the mean ± 1 S.E.M. An asterisk (*) denotes a significant difference from the first hour of recording in each preparation.  3.3 Influence of the midbrain onfictivebreathing and chemosensitivity These experiments were designed to investigate the influence o f the midbrain on fictive lung ventilation and chemosensitivity o f the in vitro brainstem-spinal cord preparation. Figure 7 shows the effect that removal o f the midbrain had on fictive lung ventilation. Removal o f the midbrain significantly reduced the average fictive breath frequency (Fig. 7A), while simultaneously producing an increase in the average inter-breath interval (Fig. 7B). Removal o f the midbrain did not alter the burst amplitude, but caused a significant increase the burst duration (Fig. 7C) and the average integrated area under the burst (Fig. 7E). Removal o f the midbrain significantly reduced the fictive breath frequency at p H levels o f 8.0 (P=0.042) and 7.8 (P=0.043), but not at 7.6 (Fig. 8). A s a consequence, while lowering the  28  p H o f the a C S F caused significant increases in breath frequency both with the midbrain on and off (P<0.05), a similar drop in p H from 8.0 to 7.6 caused a 14-fold (P=0.002) increase in frequency with the midbrain removed compared with only a 2-fold (P=0.012) increase with the midbrain intact.  MMbrainON  Midbrain OFF  MMbrainON Midbrain OFF  Figure 7. The effect of removing the caudal portion of the midbrain via transection at the optic tectum-medulla border on (A) the fictive breath frequency (breaths/min), (B) the average interbreath interval (sec), (C) the average lung burst amplitude (Volts), (D) the average lung burst duration (sec), (E) the average integrated area under the burst (Volts ). The data are shown as the mean ± 1 S.E.M. A plus sign (*) denotes a significant difference from values before transection. 2  29  7.6  7.8  8.0  pH  Figure 8. The effect of lowering pH of the aCSF on the fictive breathing frequency prior to (Midbrain ON), and following removal of the caudal midbrain (Midbrain OFF). The data are shown as the mean ± 1 S.E.M. An asterisk (*) denotes a significant difference from pH of 8.0, while a plus sign (+) denotes a significant difference at that pH from values recorded before transection.  3.4 Effect ofprogressive transections on the fictive breathing pattern This experiment was designed to determine whether the isolated medulla contains the intrinsic mechanisms for generating episodic breathing and to investigate the effect o f progressive rostral-caudal transections on the fictive breathing pattern in vitro. Figure 9 shows the effect o f progressive rostral-caudal transections on fictive lung ventilation o f the bullfrog in vitro preparation. Transections at the level between the optic tectum-medulla border (midbrain off) reduced the average fictive breath frequency from 6.9 ± 1 . 0 breaths/min to 3.8 ± 0.7 breaths/min (P=0.026) (Fig. 9A), while transections within the medulla, (illustrated in Fig. 3B), did not significantly reduce the average fictive breath frequency further. The average fictive breath duration significantly increased following transection at the optic tectum-medulla border 30  and following transection just caudal to the roots of C N VIII (Fig. 9B). Figure 10 are traces illustrating the effect that progressive rostral-caudal transections had on lung burst duration.  The  fictive breath amplitude was not significantly altered by removal o f the midbrain or transections within the medulla (Fig. 9C)  Figure 9. The effect of progressive rostral-caudal transections through the in vitro bullfrog brainstem-spinal cord preparation on (A) the fictive lung breath frequency (breaths/min), (B) lung burst duration, and (C) lung burst amplitude. The data are shown as the mean ± 1 S.E.M. A n asterisk (*) denotes a significant difference from values prior to removal of midbrain, while a plus sign (+) indicates a significant difference from previous transection values.  31  Midbrain O N  /  flfl-Tv.  »  0. 8 Sec  Midbrain O F F  '  U  —u u - s  rv  J  -1 1 .5 Sec  ^\  CN V OFF  *  r-i- .Pi.ri-nrL ..  fKj|  J  l ^qrijirAn .M L  «  •  2.3 Sec  Ijj.  C N VIII O F F  ruinlfrnr/\ 4  1  hi *  4.4 Sec  Figure 10. Traces illustrating the effect of progressive rostral-caudal transections on the lung burst duration of the in vitro bullfrog brainstem-spinal cord preparation. Horizontal lines represent the baseline of the neural trace, while vertical lines indicate the initiation and termination points of each breath.  32  Figures 11A-D are Poincare plots illustrating the effect that progressive transections had on the size and distribution o f inter-breath intervals (IBI) within the fictive breathing pattern. Prior to removal of the midbrain, the minimal inter-breath intervals occurred tightly clustered between 1-2 seconds (Fig. 11 A ) . Transection at the optic tectum-medulla border (midbrain off) not only increased the length o f these intervals, but also decreased the consistency in spacing between breaths within the pattern, causing them to occur roughly between 5-15 seconds (Fig. 1 IB). Further transections within the medulla did not alter the minimal IBIs or their distribution within the pattern further (Fig. 11C and 1 ID).  A.  B.  Midbrain On  10  Midbrain OFF  30  20  _  10  4  0  2  4  6  8  10  10  D.  CN V OFF  30  20  20  30  CN VIII OFF  30  20  o <u  S  "•V*  L  • 10  10  m  0  10  20  30  0  10  20  30  IBI-1 (Sec)  IBI-1 (Sec)  Figure 11. The effect of progressive transections on the fictive breathing pattern of the in vitro bullfrog brainstem-spinal cord preparation. Poincare plots show inter-breath intervals in seconds (IBI) plotted against the preceding inter-breath intervals (IBI-1) measured over a 2-hour recording period.  33  3.4.1 Effect of transection on fictive episodic breathing Figure 12 illustrates the distribution o f inter-breath intervals from an in vitro bullfrog brainstem preparation that exhibited a consistent fictive episodic breathing pattern over a 2-hour recording period prior to removal of the midbrain. Figure 12A shows a traces demonstrating the episodic pattern recorded from this preparation over 30-second and 300-second time scales. Fictive episodic breathing patterns were characterized on the Poincare plot by an L-shaped distribution o f IBIs with a segregated cluster at the origin (Fig. 12B).  O n these plots, IBIs in the  origin fell between 1.5-2.0 seconds, indicating breaths in the pattern that were spaced closely together, and fell scattered parallel along both axes, reflecting breaths that were either preceded by (y-axis) or followed by (x-axis) a longer IB I. Thus, intervals falling parallel to the axis are indicative o f the first and last breaths occurring within an episode, while those within the cluster at the origin correlate to the breaths within an episode.  i  1  1  i  1  1  1  0  30  60  90  120  150  180  T  210  1  1  1  240  270  300  Time (Sec)  1  1  0  1  '  '  20  40  60  '  1  80  100  IBM (sec)  Figure 12. (A) Traces illustrating a consistent fictive episodic breathing pattern exhibited by an in vitro bullfrog brainstem-spinal cord preparation prior to removal of the midbrain. The upper trace in the panel is an expanded time scale showing the temporal relationship of fictive breaths within the pattern. (B) Poincare plots showing the distribution of inter-breath intervals from this preparation measured over a 2-hour recording period. The upper plot is the same data from the lower plot over a restricted range. 34  Following removal o f the midbrain, by transection at the optic tectum-medulla border, both the IBI length and distribution were altered (Fig. 13). Figure 13A are traces demonstrating the resulting fictive breathing pattern over 100-second and 600-second time scales. Following transection, the IBIs maintained an L-shaped distribution, however, the IBI cluster at the origin increased to between 5-15 seconds, and the arrangement o f intervals within this cluster and along the axis became less discrete, indicating more variability in the distance between breaths (Fig. 13B). Note also, that following the removal o f the midbrain, the inter-breath interval between breaths within an episode increased, as did the IBIs between episodes.  B.  A.  UJJlUil  10  20  30  40  50  140 120 100  80 U3  60 40 20 0  300  0  20  40  60  80  100  120  140  IBI-1 (Sec)  Time (Sec)  Figure 13. (A) Traces illustrating a fictive episodic breathing pattern exhibited by an in vitro bullfrog brainstem-spinal cord preparation following removal of the midbrain. The upper trace in the panel is an expanded time scale showing the temporal relationship of fictive breaths within the pattern. (B) Poincare plots showing the distribution of inter-breath intervals from this preparation measured over a 2-hour recording period. The upper plot is the same data from the lower plot over a restricted range.  The occurrence of fictive breathing episodes was inconsistent  in vitro, as illustrated in  Figure 4. Figure 14 shows the Poincare plots from an in vitro bullfrog brainstem-spinal cord  35  preparation with the midbrain intact that exhibited a consistent fictive episodic pattern (Fig. 14A) compared with one that did not (Fig. 14B). Figure 15 shows the resulting Poincare plots of an in vitro bullfrog brainstem-spinal cord preparation with the midbrain removed that exhibited an episodic pattern (Fig. 15 A ) compared with one that did not exhibit consistent episodes (Fig 15B). A l l preparations that exhibited consistent episodic patterns, regardless of the transection level, displayed an L-shaped distribution pattern with a segregated cluster o f intervals at the origin. Those preparations that did not exhibit consistent episodes produced Poincare plots in which IBIs either occurred randomly, without an L-shape distribution, or without a segregated cluster at the origin.  Non-episodic  Episodic  A.  B.  100  80  80  O  40  CO —  60  60  (/) ^  100  40  t.  20  20 0  »••••••• • •« 0  20  40  60  80  100  10  0  20  40  60  80  100  0  2  4  6  8  10  10 8 6  o  r  4  —  2  CO  if  4 2 0  0  2  4  6  8  10  IBI-1 ( s e c )  IBI-1 ( s e c )  Figure 14. Poincare plots showing the inter-breath interval distribution within the fictive breathing pattern of in vitro bullfrog brainstem preparations prior to removal of the midbrain. (A) Inter-breath interval distribution of a preparation that exhibited a consistent fictive episodic breathing pattern (Episodic). (B) Inter-breath interval distribution of a preparation that did not exhibit a consistent episodic pattern (Non-episodic). In both panels the lower plots are the same data from the upper plot over a restricted range. 36  Non-episodic  Episodic  B.  A . , fj  CD  1  60  & 40 CD — 20  0  20  40  60  0  80 100  20  20  15  15  tt> 10  10  CD  *  5  20  40  60  80 100  5 0 0  5  10  15  20  0  5  10  15  20  IBI-1 ( S e c )  IBI-1 ( S e c )  Figure 15. Poincare plots showing the inter-breath interval distribution within the fictive breathing pattern of in vitro bullfrog brainstem preparations following removal of the midbrain. (A) Inter-breath interval distribution of a preparation that exhibited a consistent fictive episodic breathing pattern (Episodic). (B) Inter-breath interval distribution of a preparation that did not exhibit a consistent episodic pattern (Non-episodic). In both panels the lower plots are the same data from the upper plot over a restricted range.  Due to the inconsistency in episode occurrence, fictive breathing patterns were classified as either episodic or non-episodic based on the results o f the Poincare plots (see Appendix 1). Since the average fictive breath frequency and the inter-breath interval distribution patterns appeared to be unaffected by transections within the medulla (Fig. 9 A , 11), for this analysis all preparations following removal o f the midbrain were collectively grouped as midbrain off. In preparations with the midbrain intact, consistent episodic patterns were exhibited 66% o f the  37  time, while preparations with the midbrain removed only exhibited consistent episodic patterns 36% of the time (Fig. 16).  Figure 16. The percentage of preparations prior to, and following removal of the midbrain that exhibited episodic and non-episodic fictive breathing patterns.  3.4.2 Influence of the midbrain in pattern formation in vitro  When in vitro brainstem-spinal cord preparations with the midbrain intact exhibited consistent fictive episodes, the IBIs running parallel to the axis in the Poincare plots generally occurred in distinct groups (Fig. 17A). These groups consisted of short IBIs that occurred between 1.5-15 seconds in length, medium intervals between 20-65 seconds in length, and longer intervals between 70-120 seconds in length. Figure 17B shows a series of fictive breathing traces from this preparation over increasing time scales that illustrates the spatio-temporal relationship of these IBIs within the episodic pattern. In preparations with the midbrain intact that consistently exhibited an episodic breathing pattern, the spatio-temporal arrangement of inter-breath intervals on the Poincare plot often 38  00  Q o -  oo  a a 3 QJ g. „ o A  —  22  Ow  .5  o  CD  CO  CD  E  "  o  ob  £  03 t-  cs  0  01 Sri c u. — .— X> . __ 00 £ £  o  ""° cs  £ a)  o  00 x B  2 & u s _  5  * I g s§ 3 a > •° °« 2 2 Is  £ -s « * M  *t "fi < H te > « _ o -3 •** Jo 5 B in  * a &  *i  e . 2 2 ; -5 $ .2 & o g . « a «• CB 3 c _S c o « o> S » « o S •a is c u« js £ — .2 "S II E trc s °^£ S2" 55 c ~c . 2 0  0  £ s « u " H s2 §  —  •=  r-  JO  ^  if O c/3  U  ~  c«  I I §I & * - =s « §  8 8  5  =  8  g 2 E •5 -o o OD.2 S .E E o-  l  S  O  J3  . £  0  ^  a * ' " t« T S ^ <£ o 03 o —  - °  "S •£ u § H- O CO * S < u o — < —' — vu o * I H  7 !  «  t  1) <u O w fa  "O  i  1) -O  oO . 3  C  .S ^ °* S ^-s  O  .22 u  «J  _o >  «I  .5  -H  < "§.•- 3 <u '  O.  g  r~; o «• *> £ £ is £ c  A  o  s « s ;  <  8  8  8  8  (o©s) iai 39  M —  U  O  *-< _ d  i  u  *  ft  o>  IU  0  2  4  6  8  10  IBI-1 (Sec)  Figure 18. (A) Poincare plot from an in vitro bullfrog brainstem-spinal cord preparation prior to removal of the midbrain that exhibited a harmonic distribution pattern. (B) A trace illustrating the fictive breathing pattern exhibited by this preparation. X indicates where a breath is not expressed in the pattern. (C) Poincare plot showing the same data over an expanded range. (D) A trace illustrating the fictive breathing pattern that was exhibited by this preparation following removal of the midbrain. (E) Poincare plot from this preparation following removal of the midbrain, (n) indicates the minimal distance between breaths within the pattern, while (2n), (3n), and (4n) represent whole number ratio of this value.  40  produced harmonic distribution patterns. This phenomenon was characterized by a distribution pattern in which discrete interval clusters occurred that were related to each other in length by whole number ratios (n, 2n, 4n) (Fig. 18A). Figure 18B is a trace illustrating the fictive breathing pattern that was exhibited by this preparation. Figure 18D is a trace demonstrating the resulting breathing pattern exhibited by this preparation following removal of the midbrain. With the midbrain intact, IBIs occurred in discrete clusters that formed multiple L-shaped distributions that were separated by whole number ratios (Fig. 18D). Following removal of the midbrain, both the size and variability of IBI's increased, and the spatio-temporal arrangement of breaths within the pattern, although occurring with regularity, did not exhibit a harmonic distribution (Fig. 18E).  4.0 Discussion The present study investigated the role of the midbrain in respiratory pattern formation in anuran amphibians. The collective data suggest that structures within the caudal portion of the midbrain provide descending inputs to the medulla that influence the burst shape, and the burst rhythm of motor output responsible for fictive lung ventilation, and are essential in providing precise spatio-temporal coordination of breaths within the pattern. Numerous studies have suggested that in anurans, the endogenous rhythm that is responsible for respiratory motor output is generated by centers in the medulla (Langendorff, 1887; Kinkead et al., 1994; Kogo et al., 1994; M c L e a n et al., 1995; Reid and Milsom 1998; Wilson et al., 2002). It has also been suggested that respiratory pattern formation, specifically the formation of episodic patterns that are distinctive of anuran lung ventilation, are the result of descending inputs from sites within the midbrain acting downstream on the rhythm generating centers within the medulla (Milsom et al., 1999; Reid et al., 2000; Gargaglioni et al., submitted). The present investigation supports these conclusions, however, it also provides evidence to 41  suggest that the isolated medulla is capable of generating episodic breathing patterns and provides a clearer picture of how the midbrain contributes to respiratory pattern formation in these animals.  4.1 Stability of the bullfrog in vitro preparation over time The final step in the production of the bullfrog in vitro brainstem-spinal cord or the isolated medulla preparation involves severing the cranial nerves and transecting the brain at either the level of C N III, through the optic tectum, or at the optic tectum-medulla border as well as transecting the spinal cord below C N XII. In all preparations the overall fictive breath frequency following setup progressively declined over the first 2-3 hours o f recording, then stabilized and remained relatively constant for the rest of the recording period (Fig 6B). Transections through the brainstem inevitably destroy cell bodies and axons of passage, most likely causing massive local depolarization and neurotransmitter release. These data suggest that transections through the brainstem-spinal cord initially stimulated sites downstream of the cut. This stimulus appears to either influence sites within the midbrain that descend to the medulla, or influence the rhythm generating sites within the medulla directly, acting as an extraneous source o f drive that increases the fictive breathing frequency. Alternatively, these transections could have possibly removed inhibitory sites located upstream of the cut, thereby removing descending influences that act downstream to inhibit breathing. If this were the case, one would predict that the breathing frequency, following transection, would have remained elevated for the duration of the recording period. However, since transections were carried out at two different levels, and had nearly similar effects that were transient in all preparations, the elevated frequency at the onset of recording appears to be caused by excitation that subsides over time, rather than removal of inhibitory mechanisms. Based on the results of this study, the time required for the fictive breathing pattern to stabilize following a transection could take anywhere from 1-3 hours. 42  To our knowledge this is the first study to document stabilization o f the fictive breathing pattern produced by the bullfrog in vitro preparation. In the absence o f peripheral input, the brainstem-spinal cord is capable o f producing respiratory motor output and a variety o f fictive breathing patterns (Kinkead et al., 1997; Reid and Milsom, 1998; Reid et al., 2000). Appropriately, this preparation is often used as a tool to investigate the origin o f respiratory rythmogenesis and the central mechanisms involved in breathing pattern formation. However, there is no standard time period allotted for stabilization of neural discharge before commencement o f experimental protocols or recording in studies that utilize this preparation. Our findings indicate that there is a period, following a transection, in which breathing frequency and pattern stabilize (Fig. 5-6). Thus, results obtained within this period must be interpreted with caution, as they may not reflect the preparation's authentic response or base level of respiratory drive.  4.2 Influence of the midbrain on fictive breathing and chemosensitivity Removal o f the midbrain caused a significant decrease in the fictive breathing frequency while producing a concomitant increase in the average inter-breath interval length (Fig. 7A,B). The average burst duration and the integrated area under the burst increased following removal of the midbrain, while burst amplitude was unaffected (Fig, 7 C - E ) , confirming the observations made by Reid et al. (2000). These data confirm that the midbrain supplies a source o f tonic excitatory drive to the rhythm generating centers in the medulla that raise the breathing frequency, while at the same time providing descending inhibitory input that limits the duration and shape o f the motor output associated with the individual breath. Lowering the p H o f the a C S F caused significant increases in the fictive lung breath frequency, prior to and following removal o f the midbrain (Fig. 8). The central chemosensitivity of the bullfrog in vitro brainstem-spinal cord preparation has been documented previously 43  (Kinkead et al., 1994; M c L e a n et al., 1995; Reid and Milsom, 1998; Torgersson et al., 2001; Morels and Hedrick, 2002), and our observations in the in vitro preparation with the midbrain intact are entirely consistent with these investigations. In anurans, the receptors responsible for central chemosensitivity are thought to be distributed throughout the rostral medulla, in the area o f the fourth ventricle (Torgersson et al., 2001). Since the transections in this study removed sites within the midbrain believed to be necessary to enhance chemoreception (Kinkead et al., 1997), but did not remove central chemoreceptors, it was expected that the chemoresponse o f the isolated medulla would be reduced in comparison to preparations with the midbrain intact. This was not the case. Although the general chemoresponse was similar in both preparations, the change in breath frequency over the entire p H range examined (pH 8.0-7.6) was actually much greater following removal o f the midbrain. The greater chemoresponse elicited by the isolated medulla reflects increased chemosensitivity imparted by the removal of sites within the caudal midbrain. It would appear that in the absence o f tonic drive supplied by these sites, and when p H / C 0 2 related drive is low, the isolated medulla is permitted to reduce breath frequency to much lower levels than would otherwise be produced in the intact in vitro preparation. This suggests that sites within the midbrain do in fact influence the central chemoresponse  in vitro, but instead of enhancing central  chemosensitivity as predicted, they seem to have a moderating effect on the chemoresponse itself, narrowing the range in breath frequencies the preparation is capable o f producing. Thus, the midbrain appears to provide respiratory drive to the medulla that is independent and above what is provided by central chemoreceptors. This drive maintains the average breath frequency at a higher rate, even when central pH/CC>2 related drive is low or absent. The purpose of this tonic drive is unclear, however, it is.possible, that in the intact animal, it serves to excite or prime  44  the rhythm generators in the medulla in order to sustain them closer to threshold, permitting a greater and perhaps more timely response to central and peripheral inputs. When peripheral inputs are intact in the freely behaving animal or in situ, central chemoreception provides the major source o f respiratory drive (Bronco et al. 1991; Smatresk and Smits, 1991; Bronco et al., 1993). However, varying central chemoreceptor drive through altering the pH/CC>2 of the superfusate does not elicit the same magnitude o f response in vitro (Kinkead et al., 1994; MacLean et a l , 1995: Reid et a l , 2000), suggesting that central chemoreception is dependent upon peripheral feedback. In contrast, the  in situ preparation still  receives a host o f peripheral and central sensory input, yet when the midbrain is removed or sites within it are destroyed, chemosensitivity is markedly reduced, implying that sites within the midbrain either enhance or convey the information that is supplied by these inputs (Kinkead et al., 1997, Gargalioni et al., submitted). Taken collectively, the evidence implies that a hierarchy exists within the anuran respiratory control system, in the order in which central and peripheral inputs are integrated to produce the overall drive to breathe. Within this hierarchy, it would appear that central chemosensitivity is most influential when peripheral inputs, and the sites that relay or enhance them (within the midbrain), are intact. Without peripheral feedback, central chemosensitivity appears to be partially masked by the tonic drive that is supplied by the midbrain; the range in response that can be elicited by central chemoreception is reduced without peripheral input. When the midbrain is removed, however, the range in response that can be elicited by the central chemoreceptors again increases. This suggests that in the intact animal peripheral feedback promotes an increase in the chemoresponse range over which central chemoreception is the most influential input. This also implies that the caudal midbrain contains sites that are crucial in integrating and perhaps amplifying the response to central and peripheral sensory input in order to produce appropriate breathing patterns.  45  4.3 Effect ofprogressive transections on the fictive breathing pattern Transections at the level between the optic tectum and medulla border significantly reduced the fictive breath frequency, while transections at a level just caudal to roots of C N V and just caudal to the roots of C N VIII, did not alter the frequency further (Fig. 9A). The average fictive breath duration significantly increased following transection at the optic tectummedulla border and following transection just caudal to the roots of C N VIII (Fig. 9B). Figure 10 shows traces illustrating the effect that progressive rostral-caudal transections had on lung burst duration. Following transections at the optic tectum-medulla border, the minimal interbreath interval length increased and the spatio-temporal distribution of intervals became less precise (Fig. 11 A , C ) . Transections within the medulla did not alter the size or distribution of inter-breath intervals further (Fig 1 I D , E), suggesting that the reduction in breath frequency, resulting from an increase in the average inter-breath interval distance was due to removal of the midbrain, and not an artifact of the transections themselves. This evidence again suggests that the midbrain supplies a source of tonic drive that serves to increase the frequency of breathing. However, it also suggests that there are sites within the caudal midbrain and rostral medulla that contribute to shaping the motor output associated with the individual breath. Regions responsible for rythmogenesis in the bullfrog are thought to lie bilaterally in an area of the rostral medulla, extending from just caudal to the roots of C N VIII and C N LX (McLean et al., 1995; Wilson et al., 2002). These studies also indicate that there are sites just rostral to this region, in the area between C N V and C N VIII, which seem to modulate respiratory related activity. The results of the present study support the previous observations, since removing the midbrain altered the breathing pattern by reducing frequency, and progressive rostral-caudal transections within the medulla, to a level just caudal of C N VIII, increased the breath duration but did not eliminate respiratory activity. Collectively, these data also suggest that the neural mechanisms that produce and shape respiratory motor output in 46  anurans may be arranged in columns that run bilaterally through the brainstem-spinal cord, extending from just caudal to C N X and projecting into the caudal midbrain, as they do in mammals (Rekling and Feldmen, 1998).  4.3.1 Effect of transection on episodic breathing O n the Poincare plots each inter-breath interval (IBI) is plotted against the preceding inter-breath interval (IBI-1) for a given recording period. The distribution of intervals on the Poincare plot is, therefore, a graphical representation of the spatio-temporal arrangement of breaths within the breathing pattern during a particular recording period. The Poincare plot of a bullfrog in vitro preparation with the midbrain intact that displayed a consistent fictive episodic breathing pattern was characterized by an L-shape distribution pattern with a segregated cluster of IBIs at the origin (Fig. 12). Intervals within this cluster are the minimal distances between breaths within an episode that a preparation produced. Intervals outside this cluster, falling parallel to the y-axis, are distances prior to the first breath in an episode, while the intervals falling parallel to the x-axis are the distance immediately following the last breath in an episode (see Appendix 1). Following transection at the optic tectum-medulla border, both the length and distribution of inter-breath intervals on the Poincare plots were altered (Fig. 13). Although the general pattern of intervals was still L-shaped with a segregated cluster at the origin, as is characteristic of fictive episodic breathing, the minimal interval distances within and between episodes increased, and the spatio-temporal arrangement of breaths within the pattern became less precise. Episodic breathing patterns are the hallmark of anuran ventilatory behavior and are highly influenced by peripheral and central sensory feedback (West et al., 1987; Kinkead et al., 1996, 1997; Kinkead and Milsom, 1994). However, some of the mechanisms producing these patterns are known to be intrinsic to the brainstem-spinal cord and function independently o f 47  sensory inputs (Kinkead et al., 1994; Reid and Milsom, 1998, Reid et al., 2000). In the bullfrog in vitro preparation the occurrence o f episodes is highly unpredictable, occurring less that 50% of the time (Reid and Milsom, 1998). In the present study not all preparations displayed consistent episodic breathing patterns either. If a preparation did not exhibit discrete episodes, or produced episodes in which the inter-breath intervals were inconsistent in length, the Poincare plots did not produce an L-shape distribution pattern (Fig. 14). Episodic distribution patterns were displayed in 66% of the preparations prior to removal of the midbrain, and in only 35% of the preparations with the midbrain removed (Fig. 16). Taken together, these data suggest that removal of the midbrain does not eliminate breathing in episodes, but dramatically increases the minimal distances between breaths within an episode, while decreasing the preciseness in the spatio-temporal relationship of breaths within the pattern. Although the episodic nature of breathing was not eliminated by removal o f the midbrain, providing evidence that mechanisms capable o f producing episodes are present in the medulla, the occurrence and consistency of discrete episodic patterns was reduced. Previous studies have suggested that episodic patterns in anurans are the result o f descending inputs from sites within midbrain that cluster breaths into episodes (Reid et al., 2000; Gargaglioni et al., submitted). In these studies transections at the optic tectum-medulla border converted the episodes into a slower frequency pattern of evenly spaced breaths. The remaining rhythm, presumed to be the underlying cadence produced by the rhythm generators in the medulla, was slower in frequency than that seen within the episodes prior to transection. It was thus concluded that the midbrain not only provided a source of tonic drive that increased breath frequency, but also provided positive and negative modulation to the rhythm generators in an alternating fashion that produced episodic patterns. In the present study, following removal of the midbrain, the average inter-breath interval distance increased (Fig. 7B), also producing a decrease in the average fictive breath frequency 48  (Fig. 7A), and implying a loss o f tonic drive. However, analysis o f the Poincare plots also suggests that the isolated medulla is capable o f generating fictive episodic breathing patterns without the influence o f higher brain centers, albeit the size, consistency, and uniformity of episodes are greatly affected, and do not closely resemble episodes produced by preparations with the midbrain intact (Fig. 12, 13). Mechanisms within the caudal midbrain appear to be essential to tightly cluster breaths together to form discrete, well-defined episodes and increase the frequency o f breaths within them. Thus, removing the midbrain does not simply remove mechanisms that selectively allow the respiratory rhythm to become episodic, nor does it only promote a shift in the pattern continuum from episodes to lower frequency evenly-spaced single breaths via a reduction in intrinsic drive, but rather appears to do both simultaneously. The results o f the present study, therefore, only partially agree with the conclusions o f Reid et al. (2000) and Gargalioni et al. (submitted) with regards to how descending inputs from the midbrain influence the respiratory rhythm to produce episodic breathing patterns. The discrepancy partially lies in the manner in which episodes are defined within each study. The definition of an episode in previous studies was two or more breaths in sequence with no more than the distance of one (or two) breath cycles between them (Kinkead et al., 1994; M i l s o m et a l , 1999; Reid et al., 1998, 2000). This definition is subjectively adequate for describing episodic patterns in which the frequency o f breathing is relatively high and breaths tend to occur relatively close together. However, this definition cannot encompass discontinuous breathing patterns in which breaths are obviously occurring in episodes but are separated within the episode by distances greater than one or two breath cycles. Previous studies have also analyzed data obtained promptly following a transection or over a relatively short recording period, often between 10-30 minutes. A s the results o f this study have already suggested, transections through the brainstem may act as an excitatory stimulus to sites downstream o f the cut, artificially stimulating breath frequency for some time 49  after (Fig. 6). If removing the midbrain eliminates a large source of tonic drive, slowing breath frequency by increasing the minimal distance that breaths can occur in succession within the pattern, then transections that promote an increase in frequency would certainly give the impression of slower frequency patterns consisting of evenly spaced breaths. Without time for the pattern to stabilize, due to relativity short recording or analysis periods, longer and slower episodic patterns would be completely undetectable. The findings of this study suggest that episodes are not being entirely eliminated following removal of the midbrain, but are greatly increased in size and decreased in consistency.  Taken together the evidence implies that the midbrain does indeed play an essential  role in proper episode formation but is not solely responsible for the production of episodic patterns. This would then suggest that simple positive and negative modulation of the rhythm generators in an alternating fashion, as suggested by Reid et al., 2000, does not adequately explain how episodic patterns are produced.  4.3.2 Influence of the midbrain on pattern formation in vitro Poincare plots provide a graphical representation of the spatio-temporal relationship of breaths within the breathing pattern. In the present study, preparations that exhibited a consistent episodic breathing pattern produced distribution patterns in which inter-breath intervals fell into distinct groupings (Fig. 17A), suggesting that intervals occurred with sufficient regularity within the breathing pattern, over a 2-hour recording period, to produce segregated groups on the Poincare plot. The groups with the shortest intervals on the plot corresponded to intervals between breaths within an episode, while the medium sized group corresponded to the intervals between episodes, and the largest o f these group corresponded to intervals between what we term "episode clusters" (Fig. 17B).  50  The spatio-temporal arrangement of inter-breath intervals on the Poincare plots often displayed harmonic distribution patterns in which intervals were preceded or superceded by intervals that were related to each other by whole-number ratios (Fig. 18). This implies that even when a fictive lung breath does not occur, or is not expressed within the pattern, the spatiotemporal coordination of breaths is still being precisely maintained (Fig. 18B), such that when the next breath does occur, it is temporally expressed at a ratio value of the minimal interval between breaths within an episode (Fig. 18 A ) . Removal of the midbrain disrupted this distribution pattern. Breaths, although occurring with some periodicity, were much less precise in their spatial-temporal arrangement and did not display a tight harmonic distribution pattern (Fig. 16C). A s illustrated in Figure 2, the anuran respiratory pattern exists as a continuum that can span from random single breaths, to episodes of increasing length, and finally to continuous breathing as respiratory related drive progressively increases (Milsom et al., 1999). The present findings suggest that breaths within the pattern continuum are not only expressed as single events, or as multi-breath events (episodes), but also as clusters of multi-breath events (episode clusters), all of which appear to be maintained with extremely precise spatio-temporal coordination. Although these patterns are produced by neural mechanisms that are intrinsic to the brainstem and spinal cord, it is not precisely clear how are they are generated or maintained. The endogenous respiratory rhythms that are responsible for buccal and lung ventilation in the bullfrog are thought to be the product of multiple coupled oscillating neural networks located in the medulla (Wilson et al., 2002).  It is common knowledge in the physical sciences  that connected oscillators often produce harmonic patterns as the result of synchronizing interactions between their independent frequencies that are related in period by whole number ratios. Our data appears to be consistent with this notion. Since removal of the midbrain resulted in the abolishment of the harmonic distribution pattern by disrupting the spatio-temporal 51  arrangement of breaths, but did not remove episodes or the underlying periodicity in pattern (Fig. 13, 15), it would appear that the caudal midbrain is either required to synchronize the medullary oscillators, or potentially contains a separate oscillator itself that is interacting with those downstream to produce these phenomena.  4.4 Multiple-coupled oscillating networks In the vertebrate nervous system, oscillating neural networks responsible for rhythmic motor patterns are thought to arise segmentally during development. Thus, rhythmic motor output associated with breathing movements appears to originate from regions that extend the length of the hindbrain and into the midbrain (Lumsden and Keyens, 1989), suggesting the involvement of multiple rhombomeric oscillators that are intersegmentally coordinated (Fortin et al., 1995). Multiple brainstem oscillators have been found to play a role in the generation of respiratory motor movements in the lamprey (Thompson, 1985), the chick embryo (Fortin et al., 1995, 1999), the frog (Wilson et al., 2002) and the embryonic mouse (Abadie et al., 2000). In mammals, respiratory rhythm generation is thought to occur in the PreBotzinger complex; a region in the ventrolateral medulla (Smith et al., 1991), but developing evidence suggests that rythmogenesis may also involve sites that are slightly rostral to this area (Ballanyi, 1999; Mellen et al., 2003; H o m m a et al., 2003). There is also data, although unsubstantiated, proposing that the pons in mammals, a structure that is homologous with the caudal midbrain in anurans, contains a separate oscillator capable of rythmogenesis and is essential for normal breathing (St John, 1996). Based on this evidence, and the findings of the present study, we propose a model of ventilatory rythmogenesis in anurans that is the product of synchronized interactions between three neural oscillators; two that have been suggested by Wilson et al. (2002), that lie bilaterally within the medulla, in the area between C N VIII and C N X , and one that lies bilaterally within 52  the caudal midbrain, between C N III and the midbrain-medulla border (Fig. 19A). Figure 19B is a synaptic model illustrating a possible spatial arrangement and the coupling dynamics between oscillators within the brainstem-spinal cord. In this system, each oscillator is distinct but coupled in a recurrent cyclical inhibitory ring circuit. The first oscillator in the circuit supplies excitatory input onto the second, while the second oscillator supplies excitatory input onto the third. The third oscillator then provides inhibitory input back onto the second, and the second provides inhibitory input back onto the third. This three-oscillator circuit represents the ventilatory central pattern generator (CPG) that is responsible for generating the anuran respiratory pattern continuum. Output from the C P G is relayed via a dominant pathway that runs from oscillator within the midbrain to the motor neuron ( M N ) pools and then to the respiratory muscles. However, since removing the midbrain altered the fictive breathing pattern but did not eliminate respiratory motor output, a redundant pathway must exist that runs from the second oscillator directly to the M N pools. Similarly, in studies where transections separated the medullary oscillators, both the rostral and caudal brainstem sections were capable of producing motor output (Reid et al., 2000; Wilson et al., 2002), suggesting the existence of yet another redundant pathway that runs from the third oscillator to the M N pools. Although the oscillators are distinct in this network and possess their own connection with the M N pools, because o f the dominant pathway that runs from the midbrain, the motor output pattern that is expressed at the respiratory muscles is the result of the interoscillator coupling within the C P G . Figure 20 illustrates how the discharge properties of the three separate oscillators within the C P G coalesce to produce respiratory motor output that is intrinsically episodic. There is a hierarchal organization within the network so that expression of the rhythm is dependent upon the properties o f the other oscillators in the circuit. In this system, increasing the respiratory drive serves in raising the first oscillator closer to its threshold level (TH). Once the first  53  54  O £ O cn t£S co cu  "9  v-  CL.  L.-S  @  o  cS g , cn  o o  a  -a <»  '5 IS 2 cn  P-  tr,  o  f  «  cn  .2  O  cj cn O  00 . _. c_  CO  » -c S3 2  >  00  ~— .S  ° S 5 S3 .2 O  JS  cu  CO  I "2  g CU cu _ 5 ~ 5= cu cn Q H • —  — ° •S  .2 o rS o£ 00 S  X  ca t> .2 2 - a j o co cu CU ~ T3 c E ° cu O -f= CL  8 £ c o o e fa  3  Cu  * J  * J  S °- 5 CL,  §1  1I 3 — s  fa  O  a  O  > i  CU  '—  ™ CU  CS *o £ c H  °  « o» eu cn  E a o c Jj o o  cfc  §  g  •a .2  cn cu fa X cu cu 00 J5  c  o  3  cn cn cu CL X U  .2 cn rs cn Sg 2 .2 fe C L  cn _c cu * 3 <-  cu JS  cu cn fa CU j a cu  s CO  2 CL.  fa 00  CU  o  CO fa.  CO  ro o + ra CN "o + to  o  CD  1o + " t_  to  O  55  o  o  I y  * cu  >  _  t/3  .  3  -fa  tj  cu  C  5  o  fa cu O c/5 O»  N cn • — cn > fa fe co JU  §-§ O  3  3 — — fa -o 3 53 00 o o fa cn  o  cu  CL,  o  fa.  CU  C cu •— « - O CO cu  •o cu cu a c 5  •5 Q  CO  X X X  o>  S f S >i 3 CU  1H CQ  -fa  CL.  fa• cn  O  g 3  «3u 2  "S 2  -S  £ 2 g cu 4s  oscillator has surpassed its threshold level it begins to discharge (Fig. 20C).  Because the first  oscillator directly inputs onto the second oscillator in the network, its discharge serves in raising the second closer to its own threshold. When the second oscillator reaches and surpasses its threshold level, it too begins to discharge. However, since the circuit is arranged in a ring, and the second oscillator's period is shorter than the first, the inhibitory feedback from the second oscillator, onto the first only allows expression o f the discharge when both oscillators are above their respective threshold levels (Fig. 20B).  In turn, the summated discharge from the first and  second oscillators serves in bringing the third oscillator closer to its threshold level. Once the third oscillator has surpassed its threshold it begins to discharge, however, its period is shorter still than the first and second, thus expression of the overall discharge as motor output can only occur when all three oscillators are above there respective threshold levels at the same time (Fig.  20A). Figure 21 is a diagram illustrating how the theoretical interactions between oscillators in the C P G form motor output that is expressed as a single event (breath), a multi-breath event (episodes), and a multi-episode event in succession (episode cluster). Due to the dynamics o f the ring circuit, even though each oscillator within the circuit possesses a distinct discharge period, determined in part by its own threshold (TH) properties, it is also influenced by the discharge periods o f the other oscillators within the network. Thus, the overall motor output pattern that is expressed is contingent upon the summation and degree o f synchronization between the discharges produced by three separate oscillators. When the discharge periods of the oscillators are in phase with each other (synchronized), that is, when they are all above their respective thresholds simultaneously, expression o f each oscillator is precisely reinforced. Because the third oscillators period is the shortest, expression of the motor output can only occur briefly when its threshold (TH) has been breached. This oscillation represents the motor output associated with a single fictive breath. However, as long 56  S —i o § > S  O  w  •—  b o o . o 3 "«» cj cd o  o  i l '  CO  co  U5  s  '3  ss  0  1e <3.S l  l  ©9 CJ S  o. X  <U  a*  g -s  •S <» =-  a> •a 3 .C£ J3 -S3 • «  to  N  «  .—i  ,0  X to 0>  P o  ? co  O  B 2 o jo » 2 a co  o .2  1 -s 2  |  a  03 <D cn  .S  ca  8 2g S  80  08 o  CO  ° J3 -s o CO  _ oO  57  iri  as the second oscillator remains above its threshold level the discharge of the third oscillator is expressed in succession, thus producing motor output that appears to occur episodically. Thus, the oscillation that underlies the expression of an episode is produced by the discharge period of the second oscillator. Likewise, as long as the first oscillator remains above its threshold level, its discharge impinges on the second oscillator, which impinges on the third and thus expression of the motor output pattern occurs as breaths within episodes, within episode clusters. Thus, the oscillation that underlies the expression of episode clusters is produced by the discharge period of the first oscillator. This is one possible network configuration of the neural mechanisms that would generate breathing patterns in which breaths, episodes, and episode clusters would occur with precise regularity. Thus, we believe it is the interactions between the oscillators, when they are synchronized, that results in the harmonic distribution patterns observed on the Poincare plots from highly episodic preparations. In this model, the oscillator within the caudal midbrain has the shortest discharge period. It must, therefore, be the last through which the respiratory rhythm is expressed, and serves to increase the overall discharge frequency while producing a temporally coordinated pattern in which breaths, and episodes occur closer together. When the caudal midbrain was removed in this study, via transection at the optic tectummedulla border, fictive respiratory discharge was still capable of being expressed although the frequency was reduced; the average interval between breaths and episodes increased, and the spatio-temporal arrangement of breaths within the pattern became less precise. Based on the model, i f this transection removed the oscillator with the shortest discharge period, it would result in a slower-frequency pattern of motor output that would be longer in duration. This is exactly what was observed following removal of the midbrain and portions of the rostral medulla (Figs. 9,10). However, simply removing the midbrain oscillator does not intuitively explain why the spatio-temporal coordination of breaths within the pattern became less precise. 58  Since the production of harmonics within coupled oscillating systems results from the synchronization of multiple frequencies, there is the possibility that removal of an oscillator that was a part of a coordinated network could result in temporary de-synchronization between the remaining parts of the circuit. Because the only preparations that produced harmonic distribution patterns were those that exhibited consistent episodes, it suggests that these patterns result from fairly precise phase synchronization between oscillators within the network. Thus, desynchronization between the oscillators may account for why fewer preparations exhibited a consistent episodic pattern following removal of the midbrain. This event would most likely compromise the spatio-temporal relationship of breaths within the pattern that remained, perhaps producing episodes that appeared to be inconsistent in size and occurrence. Although hypothetical, a three-oscillator network model with this configuration is theoretically possible and can account for many of the observations made in this study. One unique feature of this model is the ability to alter the period and threshold levels (amplitude) of each oscillator independently while observing the effect it would theoretically have on the other oscillators within the network and the overall motor output pattern that would be expressed. Through manipulation of the discharge period and threshold levels of each oscillator, the degree of synchronization between oscillators can be altered, and any breathing pattern within the entire respiratory pattern continuum of anurans can be replicated (Fig. 22). This model suggests a more complex mechanism for adjusting motor output to produce a respiratory pattern continuum through simultaneously modulating the discharge period and threshold properties of multipleinteracting network oscillators that exist in a ring circuit configuration, rather than simple selective expression of one respiratory rhythm. If this theoretical configuration were accurate, it would also generate the definable hierarchical organization of sensory inputs within the respiratory control system that have already been experimentally determined. Thus, certain inputs would alter the discharge period or 59  clu ers  cu  cn  X  par  cj lo  c  CJ  cCO  cn  tem c llator ex resse^ oc urrin  00  cn  CJ  CJ  cn  cn  cn  O CJ  -a 3  !c  —  CO _CJ CJ X I s- CO •r"  a thr mplit ) No of vai  O CJ CJ  00 _c  J3  eng  eac  O -o 3  JS  cn  >, •j a CJ 3  cr  ods fre  o.  CO  CJ CJ J =  n. CJ 00 s_ CO  CJ  CJ  n. X  n  Q CJ U J  o ncn  cn  J=  CO  CJ 00 3 O . x>  3  —i  O  J3 CJ  CO CJ  eract ions betw pat tern inu um below leve • (QR gul c o CJ  c  CO  Oo  CJ  u.  IH o  CJ  cn  00 3 JS _n _C "cn  _c  ob c Ic  no  X>  CJ  O X  he disc man mot  a.  *•*  c CJ cj  cn  no p p u  -C  <  brea  Q  c  ssed. pisod  uced  >> CO x> -a  di  o c  CJ s —  !—  cn  3 O  O  CJ  —  cn  fa  CO u. 3  -O CJ - o 1—  CO  CJ . 3  ms illus esp licat the a le rep sent bre isio al si gle ar oc nsiste  1 « -g CJ Jo 3 00  c C  00  ~£  CO  (5  c  'In  4-*  C  n o CJ c  O  cn  CJ  o 'C O  CJ  cn  si o CO c/3  o m  pa  CN CJ - o - C °— CJ 3 J - 12 _0p CJ  60  CO CJ _N  po  N i-i o  CJ  "a  X)  no  c  o '5.  threshold properties of one oscillator more than another, depending on their relative importance in adjusting ventilation to meet metabolic demand, and thereby contributing more (or less) to manipulating the overall breathing pattern. Rings of coupled oscillators are known to play roles in a variety of physiological functions. Recurrent cyclic inhibitory circuits, like the one proposed is this study, are thought to be vital for the coordinating the motion of locust wings during flight (Robertson and Pearson, 1985), swimming movements in the leech (Friesin, 1989), and respiratory activity in snails (Bulloch and Syed, 1992). Recently, ring circuit oscillators have been utilized in modeling central pattern generation (CPG) that coordinates the muscular activity of locomotion in vertebrates (Grillner et al., 1995). A n d ring oscillator systems that involve recurrent cyclic inhibition with similar properties to those outlined in the present study have been shown to be capable o f selectively entraining to produce a continuum of different motor output patterns (Canavier et al., 1997; Dror et al., 1999). The model in the present study is based on the accumulated evidence of numerous investigations and presents a possible configuration for the ventilatory C P G within anurans that is theoretically capable of generating breathing patterns that range from no breathing (Fig 22A), to even single or double breaths (Fig. 22B), to various episodic patterns Fig. (22C-E), to continuous breathing (Fig. 22F). Although theoretical, models such as this may be useful in providing a conceptual platform for investigating how respiratory rhythm is generated and how sensory inputs are integrated and contribute to modulating the breathing pattern in vertebrate species.  5.0 Conclusions Taken collectively, the data from the present study suggest that sites within the caudal and rostral medulla provide a variety of inputs that are essential in modifying the burst pattern, and burst rhythm for proper respiratory pattern formation. 61  Burst Pattern Following removal o f the midbrain and rostral portions o f the medulla, the duration o f neural discharge associated with each lung breath significantly increased. This implies that sites within the caudal portion of the midbrain and rostral medulla are responsible for providing inhibitory input to the medullary oscillators that limits the duration o f motor output associated with each lung burst.  Burst rhythm Following removal of the midbrain, fictive lung breath frequency decreased. This implies that sites within the caudal portion o f the midbrain provide a source o f excitatory tonic drive to the medullary oscillators that increase the overall frequency o f lung bursts per unit time.  Breathing pattern Following removal o f the midbrain, although episodic breathing patterns were not eliminated, the spatio-temporal arrangement o f breaths within the breathing pattern became more variable, producing episodes that were less discrete and inconsistent in size and occurrence. This evidence implies that the medulla possesses neural mechanisms that are capable o f generating episodic patterns, but that sites within the caudal midbrain influence the medullary oscillators in order to produce a more consistent, and temporally coordinated episodic pattern.  Modulation A larger central chemoresponse was elicited following removal o f the midbrain. This evidence implies that in the absence o f peripheral feedback, sites within the caudal midbrain moderate central chemosensitivity by providing tonic input that reduces the overall range in  chemoresponse in vitro. Multiple coupled oscillators The endogenous respiratory rhythm responsible for lung ventilation in anuran amphibians originates in the medulla (Langendorff, 1887; Kinkead et al., 1994; K o g o et al., 1994; M c L e a n et 62  a l , 1995; Reid and M i l s o m 1998; Wilson et al., 2002). It has also been suggested that respiratory pattern formation is modulated by descending inputs from structures within the midbrain acting downstream on the rhythm generating centers within the medulla (Milsom et al., 1999; Reid et a l , 2000; Gargaglioni et al., submitted). Our new data suggest that respiratory pattern formation in anurans may be the consequence of multiple oscillating neural networks working in concert and arranged in a bilateral column within the brainstem. Two of these oscillators appear to lie within the medulla in the area between C N X and C N VIII (Wilson et al., 2000), while the present study suggests that the third lies within the midbrain between C N III and the optic tectum-medulla border. In this system, the oscillators are arranged in a recurrent inhibitory ring circuit with a hierarchical organization that produces patterns of motor output that are intrinsically episodic. Further investigation is certainly required to elucidate how these oscillators influence each other and how they are modulated through sensory inputs to produce the anuran respiratory pattern continuum.  63  References Abadie, V . , J. Champagnat, and G . , Fortin (2000). Branchiomotor activities in the mouse embryo.  Neuro. Repor. 11.1: 141-145.  Ballanyi K , Onimaru, H , Ffomma I (1999). Respiratory network function in the isolated brainstem-spinal cord of newborn rats.  Progr. Neurobiol. 59: 583-634.  Boutilier, R . G . , and D . P . Toews (1976). The effect of progressive hypoxia on respiration in the toad  Bufo marinus. J. Exp. Biol. 68: 99-107.  Branco, L . G . S . , L . Mogens, and A . Hoffman (1991). Central chemoreceptor drive to breathing in the unanaesthetized toads,  Bufo paracnemis. Respir. Physiol. 87: 195-204.  Branco, L G . S., M . L . Glass, T. Wang, A . Hoffman (1993). Temperature and central chemoreceptor drive to ventilation in the toad,  Bufo paracnemis. Resp. Physiol.  93.3: 337-346. Bulloch, A . G . M . , and N . I. Syed (1992). Reconstruction of neural networks in culture.  Trends Neurosci. 15:422-427. Burggren W . W . and N . West (1982). Changing respiratory importance of gills, lungs. and skin during metamorphosis in the Bullfrog Rana  catesbeiana. Respir. Physiol. Al:  151-164. Canavier, C . C . R.J. Butera, R . Q . Dror, D . A . Baxter, J . W . Clark, and J.H. Byrne (1997). Phase response characteristics of model neurons determine which patterns are expressed in an ring circuit model of gait generation.  Biol Cybern. 11: 367-380.  Dror, R . O . , C . C . Canavier, R.J. Butera, J . W . Clark, and J.H. Byrne. A mathematical criterion based on phase response curves for stability in a ring of coupled oscillators.  Biol Cybern. 80: 11-23.  de Jongh, H . J . , and C . Gans (1969). O n the mechanism of respiration in the Bullfrog,  Rana catesbeiana: a reassessment. J.Morph.  127:259-290.  Fortin, G . , F . , Kato, A . Lumsden and J. Champagnat (1995). Rhythm generation in the segmented hindbrain of chick embryos. J. Physiol. 486.3: 735-744. Fortin, G . , F . , S. Jungbltuh, A . Lumsden, and J. Champagnat (1999).  Segmental  specification of G A B A e r g i c during development of hindbrain neural networks.  Nat. Neuro. 2.10: 873-887. Friesen, W . O . (1989). Neuronal control of leech swimming movments. In: Jacklet, J . W . (ed).  Neuronal and cellular oscillators. Marcel Dekker, New York, pp 269-315.  Gillner, S., T. Deliagina, O . Ekeberg, A . E l Manira, R . H . H i l l , A . Lasnser, G . N . Orlovsky, P. Wallen (1995). Neural networks that coordinate locomotion and body orientation in the lamprey.  Trend Neurosci. 6: 270-279  64  Hilaire, G . , Bou, C . , Monteau, R. (1997). Maturation o f the mammalian respiratorysystem.  Phyiol. Rev. 79(2): 325-360.  Hilaire, G . , Monteau, R., Errichidi, S. (1989). Possible modulation o f the medullary respiratory rhythm generator by the noradrenergic A 5 area: an in vitro study in the new  born rat. Brain Res. 485: 325-332 Kimura, N . , S.F. Perry, and J.E. Remmers (1997). Strychnine eliminates reciprocation and augmentation of the respiratory burst in the  in vitro frog brainstem. Neuroscience.  225: 9-12 Ishii, K . , K . Ishii, and T. Kusakabe (1985). Chemo- and Baroreceptor innervation o f the aortic trunk o f the toad  Bufo vulagris. Respir Physiol. 60:365-375.  Kinkead, R., and W . K . M i l s o m (1994). Chemoreceptors and the control o f episodic breathing in the bullfrog  (Rana catesbieana). Respir. Physiol. 95: 81-98.  Kinkead, R., and W . K . M i l s o m (1996). C02-sestivie olfactory and pulmonary receptor modulation o f episodic breathing in bullfrogs. Am. J. Physiol. 270: R134-R144. Kinkead, R., and W . K . M i l s o m (1997). The role of pulmonary stretch receptor feedback in control o f episodic breathing in the bullfrog. Am. J. Physiol. 270: R497-508 Kinkead, R. (1997). Episodic breathing in frogs: Converging hypothesis on neural control o f respiration in air breathing vertebrates. Am. Zool. 37: 31-3-40. Kinkead, R., W . G . Filmyer, G.S. Mitchell, and W . K , M i l s o m (1994). Vagal input enhances responsiveness o f respiratory discharge to central charges in pH/C02 in  bullfrogs. J. Appl. Physiol. 77: 2048-2051. Kinkead, R., M . B . Harris, W . K . M i l s o m (1997). The role o f the nucleus isthmi in respiratory pattern formation in bullfrogs.  Exp. Biol. 200: 1781-1789.  Kogo, N . , S.F. Perry, and J. Remmers (1994). Neural organization o f the ventilatory activity in the Bullfrog, Rana catesbeiana. I.  J. Neuro. 25: 1067-1079.  Kogo, N . , S.F. Perry, and J. Remmers (1994). Neural organization o f the ventilatory activity in the Bullfrog, Rana catesbeiana. II.  J. Neuro. 25: 1080-1094.  Langendorff, O. (1887). Die Autmatie des Atemzentrums.  Arch Anat Pysiol. 285-295.  Liao, G.S., L . Kubin, R.J. Galante, A . P . Fishman, and A . I . Pack (1996). Respiratory activity in an in vitro brainstem of tadpole,  Rana catesbeiana. J. Physiol. 492(2): 529-  544. Lumsden, A . , R. Keynes (1989). Segmental patterns o f neuronal development in the chick hindbrain. Nature. 337: 424-428.  65  Macintyre, D . H . , and D . P . Toews (1976). The mechanics of ventilation and the effects of hypercapnia in  Bufo marinus. Can. J. Zool. 54: 1364-1374.  M c L e a n , H . A . , N . Kimura, N . Kogo, S.F. Perry, and J.E. Remmers (1995a). Fictive respiratory rhythm in the isolated brainstem of frogs. J. Comp. Physiol. 176:703-731.  McLean, H . A . , N . Kogo, S.F. Perry, and J.E. Remmers (1995b). Two regions in the brainstem of the frog that modulate respiratory-related activity. J. Comp. Physiol.  177:  133-144. Mellen, N . M . , Milsom, W . K . , Feldmen, J . L . (2002). Hypothermia and recovery from arrest in neonatal rat in vitro brainstem preparation. Am. J. Physiol. 282: R484-491. Mellen, N . M . , W . A . Janczweski, C M . Bocchiaro, J.L. Fledman (2003). Opioid induced quantal slowing reveals dual networks for respiratory rhythm.  Neuron. 37: 821-826  Milsom, W . K . , M . Castellini, M . Harris, J. Castellini, D . Jones, R. Berger, S. Bharma, L . Rea and D . Costa (1996). Effects of hypoxia and hypercapnia on patterns of sleepassociated apnea in elephant seal pups.  Am J. Pysiol. 271 :R1017-R1024.  Milsom, W . K . , M . B . Harris, and S.G. Reid. (1997). D o descending influences alternate to produce episodic breathing?  Respir. Physiol. 110: 307-317.  M i l s o m W . K . , S. G . Reid, J.T. Meier, ad R. Kinkead (1999). Central pattern generation in the bullfrog,  Rana catesbeiana. Comp. Biochem. And Physiol. 124: 253-264.  M i l s o m W . K . , (1991). Intermittent breathing in vertebrates. Annu.Re. Pysiolo.  53,87-  105. Morales, R . D . , Hedrick, M . S . (2002). Temperature and p H / C 0 2 modulate respiratory  activity in the isolated brainstem of the bullfrog (Rana catesbieana). Comp. Biochem. Physiol. Part A 132: 477-487 Oka, K . The influences of the transection of the brain upon respiratory movement of the  frog. (1958). J Physiol. Soc. Japan 20: 513-519 Paydarfar, D . , Eldridge, F . L . (1987). Phase resetting and disrhythmic responses of the  oscillator. Am. J. Physiol. 252: R55-R62. Perry, S.F., H . A . M c L e a n , N . Kogo, N . Kimura, H . Kawasaki, M . Sakurai, E . A . Kabotyanski, and J.E. Remmers (1995). The frog brainstem preparation as a model for studying the central control of breathing in tetrapods. Braz. J. Med. Biol. Res. 28: 1-8. Reid S . G . , and W . K . M i l s o m (1998). Respiratory pattern formation in the isolated bullfrog  (Rana catesbieana) brainstem-spinal cord. Respir. Physiol. 114: 239-255.  66  Reid S.G., J.T. Meier, and W . K . M i l s o m (2000). The influence o f descending inputs on breathing pattern formation in the isolated brainstem-spinal cord. Resp. Physiol. 120: 197-211. Rekling, J . C . , and J.L. Feldman (1998). Prebotzinger complex and pacemaker neurons: Hypothesized site and kernel for respiratory rhythm generation. Annu. Rev. Physiol. 60: 385-405. Roberston, R . M . , and K . G . Pearson (1985). Neural circuits in the flight systems of the locust.  J. Neruophisol. 53: 110-128  Sakakibara,Y. (1984). The pattern o f respiratory nerve activity in the bullfrog. Jap. J.  Physiol. 34: 827-838. Sanders, C . E . , and W . K . M i l s o m (2001). The effects o f tonic lung inflation on ventilation on the American bullfrog Rana  catesbeiana shaw. J. Exp. Biol. 204: 2647-  2656. Shelton, G . , D . R . Jones, and W . K . M i l s o m (1986). Control o f breathing in ectothermic  Handbook ofphysiology, Sect. 3, The respiratory system, V o l . II, Control of breathing, Part II, A . P . Fisherman, N . S . Cherniack, J . G . Widdicombe, and  vertebrates. In:  S.R, Geiger, eds. American Physiological Society, Bethesda, Maryland, pp. 857-909. Smatresk, N . J . , and A . W . Smits (1991). Effects o f central and peripheral chemoreceptor stimulation on ventilation in the marine toad. Respir. Physiol. 83: 223-38. Smith, J . C . , Elleberger, H . H . , Ballanyi, K . , Ritcher, D . W . , Feldmen, J . L . , (1991). PreBotzinger Complex: a brainstem region that may generate respiratory rhythm in  mammals. Science 254: 726-729. St. John, W . M . (1996). Medullary regions for neurogenesis o f gasping: neoud vital or  neouds vitalis? J. Appl. Phyiol. 81.5: 1865-1877. Thompson, J.K. (1985). Organization o f inputs to motoneurons during fictive ventilation in the isolated lamprey brain.  J. Comp. Pysiol. 157A: 291-302  Torgerson, C . S . , M . J . Gdovin, R. Brandt, and J.E. Remmers (2001). Location of central respiratory chemoreceptors in the developing tadpole. Am. J. Physiol. 280: R921-R928. Vitalis, T . Z . and G . Shelton (1990). Breathing in Rana pipiens: the mechanism of  ventilation. J. Exp. Biol. 154: 537-556 Wang, T . , E . W . Taylor, S . G . Reid, and W . K . M i l s o m (1999). Lung deflation stimulates ventilation in decerebrate and unidirectionally ventilated toads. 3): 181-191.  67  Respir. Physiol. 118(2-  West, N . H . , and D . R. Jones (1975). Breathing movements in the frog  Rana pipiens. I.  The mechanical events associated with lung and buccal ventilation. Can. J. Zool. 53 332-344. West, N . H . , Z . L . Topor, and B . N . V a n Vilet (1987). Hypoxemic threshold for lung ventilation in the toad.  Respir. Physiol. 70: 377-399  Wilson, R. J. A . , K . Vasilkos, M . B . Harris, C . Straus, and J. E . Remmers (2002). Evidence that ventilatory ryhthmogenisis in the frog involves two distinct neuronal oscillators.  J. Pysiol. 540(2): 557-570.  68  Appendix 1: Poincare plot pattern analysis  Pattern anaylsis: In the Poincare plot method o f pattern analysis used in this study inter-breath intervals were measured between every breath over a 2-hour recording period. The IBIs in seconds were then plotted against the preceding inter-breath interval (IBI-1) for the recording period. Consequently, the distribution o f intervals on the Poincare plot provides a graphical representation o f the spatio-temporal arrangement o f breaths within the breathing pattern over this time period.  Continuous breathing: If the fictive breathing pattern produced by a preparation occurred as continuous uninterrupted breathing, the distribution o f intervals on the Poincare plot occurred tightly clustered around the average interval occurring between breaths.  Episodic breathing: If the fictive breathing pattern produced by a preparation was episodic, where groups of breaths were separated by definable apneas, the intervals on the Poincare plot occurred with an L-shaped distribution along both axis with a tightly clustered, segregated group at the origin. The intervals within this cluster correlate to the average intervals occurring between breaths within an episode. Intervals outside this cluster, falling parallel to the y-axis, correlate to the distances prior to the first breath in an episode, while the intervals falling parallel to the x-axis correlate to the distances immediately following the last breath in an episode.  Random breathing: If the fictive breathing pattern produced by a preparation was irregular, where breaths or groups o f breaths occurred at irregular distances away from each other, the distribution o f intervals on the Poincare plot occurred as an indefinable pattern that appeared random.  Pattern Classification: Preparations were classified a either episodic or non-episodic based on the results o f their Poincare plots. Preparations were deemed episodic i f the intervals occurred with an L-shaped distribution parallel to both axis with a segregated cluster at the origin. The pattern was deemed non-episodic i f it did not display an L-shape distribution, or did not have a defmably segregated cluster o f intervals at the origin.  69  70  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0091584/manifest

Comment

Related Items