CENTRAL CHEMORECEPTION IN THE NEONATAL RAT TRANSVERSE MEDULLARY SLICE PREPARATION Sasha Aleksandar Necakov A thesis submitted in conformity with the requirements for the degree of Master of Science Graduate Department of Physiology University of Toronto O Copyright bp Sasba Aleksandar Necakov 2001
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
CENTRAL CHEMORECEPTION IN THE NEONATAL RAT TRANSVERSE MEDULLARY SLICE
PREPARATION
Sasha Aleksandar Necakov
A thesis submitted in conformity with the requirements for the degree of Master of Science Graduate Department of Physiology
University of Toronto
O Copyright bp Sasba Aleksandar Necakov 2001
National Library I*I of Canada Bibliothèque nationale du Canada
Acquisitions and Acquisitions et Bibliographie SenAces services bibliographiques
395 Wellington Street 395, rue Wellington Ottawa ON K1A O N 4 OtiawaON K 1 A W Canada Canada
The author has granted a non- exclusive licence dowing the National Library of Canada to reproduce, loan, distribute or seIl copies of this thesis in microform, paper or electronic formats.
The author retains ownership of the copyright in this thesis. Neither the thesis nor substantial extracts h m it may be printed or otherwise reproduced without the author's permission.
L'auteur a accordé une Licence non exclusive permettant à la Bibliothèque nationale du Canada de reproduire, prêter, distribuer ou vendre des copies de cette thèse sous la fome de microfiche/film, de reproduction sur papier ou sur format électronique.
L'auteur conserve la propriété du droit d'auteur qui protège cette thèse. Ni la thèse ni des extraits substantiels de celle-ci ne doivent être imprimés ou autrement reproduits sans son autorisation.
ABSTRACT
Central Chemoreception in the Neonatal Rat Transverse
Medullary Slice Preparation
University of Toronto, Department of Physiology Sasha Aleksandar Necakov Master of Science (200 1 )
This study investigates several aspects of central chemoreception in the neonatal rat. transverse
rnedullary siice preparation 1 found that the frequency of bursting recorded fiom hypoglossal
nerves changed significantly with step changes in pH produced by varying the CO2 of the
bathing solution of the slice. 1 also found that application of 1 mM acetazolarnide dissolved in
DMSO to the slice preparation produced no significant alterations in these responses to CO2.
However. acetazolarnide application did provoke a significant and prolonged decrease in
hypoglossal nerve burst duration. Although the pH sensitive fluorescent dye. BCECF-AM.
perfonned satisfactorily in bench tests. when it was used to label transverse medullary slice
preparations it exhibited a faster rate of decay in fluorescence ernission. Furthemore. 1 O bserved
no change in fluorescence emission when pH was varied in the slice. either globally or locally
using a COz diffusion pipette. 1 concluded that 1 ) the transverse rnedullary slice does contain
central chemoreceptors. which elicit a respiratory response: 7) the central chemoreceptor
response to pWCOz variations is not dependent upon carbonic anhydrase. and 3) the pH sensitive
fluorescent dye BCECF-AM is not suitable for assessing regional changes of pH in the
transverse medullary siice preparation.
ACKNO WLEDGEMENTS
1 have been blessed with the support of several individuals who have provided me with
the courage and drive to complete my thesis.
Firstly. 1 would like to thank the departialent of physiology for providing me with the
fiddler award. and the respiratory research group for their input.
I would like to thank John Peever for having the patience and kindness in helping me to
complete my Master's work. John provided me w*th a strong example of what is necessary to
become a capable researcher. For his efforts 1 sincerely thank him. I would like to thank S a k
Mahamed for his undying patience and help in providing me with an understanding of the
workings of our laboratory software. His help was instrumental in rny completion of this thesis.
and I am pleased to have had the opportunity to share in his insights on research and what lies
beyond. Thanks to my parents - you were there for me on those endless nights when the goal
seemed so distant. and when I thought I could go on no longer. You have nunured rny spirits and
have licked my wounds and I am deeply grateful for al1 that you have done. Thanks to Lillian.
Eli. Andrew. Aria and Miles - y u have always been proud and supportive of me. and 1 am
happy to share this achievement with you. To Alison. you have stood beside me along the way
and have lightened my load more than you will ever know. You have given me a reason for my
struggle and are the one with whom I wish to share my path.
Lastly. 1 would like to express my heartfèit gratitude to Dr. James Dufin. Dr. Duffin.
you have never strayed fiom my side even in times when it would have been much easier for y u
to 50. You have been my fountain of inspiration and support. You have stuck with me to the end
Vibratome Preparation of the Transverse Medul lq Slice
Recording Setup
Recording Sstup t'or the Brainstem-Spinal Cord and Transverse Medullary S lice Preparations
Fluorescently Labelled Transverse Medullary Slice and Di fiusion Pipette
Cross Correlation Histogram of Left and Right Contralateral P hrenic Nerve Rootlets
Cross Correlation of Ipsilatenl Phrenic Nerve Rootlets
Mcaii Burst Frequency vs. % C02/pH in the Transverse Medullap Slice Preparation
Mean Burst Amplitude vs. % COl/pH in the Transverse Meduilary Slice Prepantion
Mean Burst Duntion vs. % C02/pH in the Transverse Medullaly SI ice Preparation
Burst Frequency vs. pH/% CO? Before. During. and Afier Application of 3 mM Acetazolarnide
Burst Amplitude vs. pW% CO, Before. Dunng. and A%er Application of 2 mM Acetazolarnide
Burst Frequency vs. pH/% CO? Before. During. and A%er Application of 5 .u 10" M Acetazolarnide
Burst Amplitude vs. pH/% COz Before. During. and Afier Application of 5 x 1 o4 M Acetazolarnide
Acetazolamide Crystals on the Transverse Medullary S lice
Burst Frequency vs. pH/% CO2 Before. Dunng. and Afier Application of 1 mM Acetazolamide in DMSO
Burst Amplitude vs. pW% COz Before. During and After Application of 1 m M Acetazolamide in DMSO
Bunt Dumtion vs. pH/% CO? Before. During. and Afier Application of 1 miM Acetazolamide in DMSO
Rate of pH Equilibration
Figure Name Page
Fluorescence Emission vs. pH of NonCleaved BCECF-AM
Fluorescence Emission of Cleaved BCECF-AM vs. pH
Cleaved BCECF-AM Fluorescence vs. Tirne
Fluorescence Emission of Cleaved BCECF-AM vs. Concentration
Fluorescence Emission of a BCECF-AM Labelled SIice vs. pH
Application of the CO2 Diffusion Pipene to a BCECF-AM Label led Transverse Medullary S 1 ice Preparation
BC ECF-AM Labelled Transverse Mrdullary Slice
BC EC F-AM Labelled Transverse Medullary Slice
CHAPTER 1
INTRODUCTION
1. I ) Overview
Respiration is controlied automaticaily largely through feedback fiom central respiratory
chmorecepton. They are located close to the ventral surface of the medulla for the most part.
but other areas such as the raphé may also act as chernosensors for [H']. 1 chose the neonatal rat
transverse meduilary slice preparation to examine several questions about these c hemoreceptors.
First. are they effective in the slice in that hypoglossal motor output is increased in response to
an increase in the Pcoz of the aCSF bathing the s lice? Second. is carbonic anhydrase essential
for the chemoreceptor tùnction? Third. do regions of high [H'] correspond to the locations of
the central chemoreceptors?
Considering the technical challenges involved in preparing medullary slices. my tùst goal
was to become proficient in rriaking viable slices that produced a respiratory rhythm fiom their
hypoglossal nerve rootlets. As a first step 1 leamed to rnake viable brainstem spinal cord
preparations (Smith et al.. 1991) Eorn which medullary stices can be prepared. by assisting John
Peever in an expetiment invo lving cross-correlation of both contralateral and ipsilateral phrenic
nerve rootlets. The hypothesis tested was that the respiratory bursting pattern Eoom lefi and nght
phrenic nerve rootlets in the neonatal rat brainstem-spinal cord preparation is synchronized as a
result of common activation. This experiment was successful and determined that lefi and right
phrenic nerve rootlets in the neonatal rat do not receive excitation f?om a common source.
contrary to the case in the adult rat (Peever et al.. 1999a).
Upon anainhg proficiency in producing transverse meduliary slices that exhibited a
hypoglossal motor output. 1 helped John Peever to test the sensitivity of the hypoglossal motor
output to global changes in C02/pH. Peevefs hypothesis was that an increase in the
concentration of COz (resulting in a decrease in pH) in the bathing medium of the slice
preparation will lead to an increase in the hypoglossal output and should thus indicate the
presence of central chemoreceptors within the preparation. The results of t his experiment were
successiùl in demonstrating that increases in the concentration of COz of the bathing solution
produced increases in the fiequency of respiratory bursting recorded kom the hypoglossal nerve
roo tlets of the transverse medullary slice preparation ( Peever et al.. 1 999b).
As a result of demonstrating the e.&tence of central chemorecepton within the transverse
medullary slice preparation 1 probed the role of the enzyme carbonic anhydrase in centrai
chemoreception based on the hypothesis of Torrance (Tonance. 1 993) that the mechanism of
centrai chemoreception relies upon the action O f carbonic anhydrase. I hypo thesized that the
inhibition of carbonic anhydrase within the slice preparation through the application of
acetazolamide to the bathing medium wouId prevent increases in hypoglossai motor output in
response to increases in the concentration of CO? in the bathing medium. The results of this
expriment did not support this hypo thesis.
To answer the third question locating the central chemoreceptors within the slice. 1 made
use of a fluorescent dye sensitive to pH (BCECF-AM). 1 planned to test a hypothesis. based on
the hdings of Ichikawa (Ichikawa et al.. 1989) in adult cats. that regions within the transverse
medullary slice exhibithg a greater decrease in pH in response to increasing CO2 in the bathing
solution indicate the location of centrai chemoreceptors. In testing the characteristics of the dye
to ascertain its appropriateness for the task. 1 found that whiie satisfactory performance was
observed for the dye alone. when used in the slice 1 found no detectable changes in the
fluorescence emission of the dye in response to variations in pH. Despite identdjmg several
regions that were differentiaily stained. the experiment was therefore unsuccessful in providing
visual information of alterations in pH levels within the slice preparation. 1 concluded fkom my
experirnents that BCECF-AM dye could not be used to test my hypothesis.
As an alternative approach to localizing central chemoreceptor regions within the slice 1
attempted to implement the C O diffusion pipette pioneered by Nattie (Li and Nattie. 1997a) that
could be used to acida focal regions of tissue within the slice preparation. 1 hypothesized that
focal acidifkation of certain areas within the siice preparation would produce an increase in the
hypoglossal motor output. similar to that seen with increases in the concentration of COz of the
bathhg solution and so locate the central chemoreceptors. However. 1 encountered technical
dificulties in applying the diffusion pipette to the slice. and was thus unable to demonstrate its
ability to acidify any regions of tissue or to prove my hypothesis.
1.2) Background
1.2. i ) Generul
It has been well established that the neural substrate for the automa .tic control of
breathing. whose purpose is to supply tissues of the body with oxygen and to rernove the
metabolic waste product carbon dioxide. is localized in the pons and medulla and that it involves
feedback fiom many sources (Bianchi et al.. 1 995). These include receptors in the ainvays that
provide information about lung volume and its rate of change. as well as chemoreceptors in the
carotid body and medulia.
The latter are an important aspect of the feedbac k control syst em of the mammalian
respiratory rhythm generator. and are categorised as peripheral and central chemoretlexes
(Du& 1990). Central chemoreception hvolves sensors for Pco2/ [H?l w i t b the brainstem
(Erlichman et al.. 1998). It should be noted that these central chernosensors are shielded from
alterations in the blood pH by the blood-brain barrier and its associated ion transport
mechanisms.
1.2.2) Classicai Erperiments
As yet. the exact location of the central chemoreceptors has not k e n determined. nor has
their mechanism of action been elucidated. this despite the fact that from the time of the initial
experirnents conducted by Leusen in the late 1950's probing central chemoreception mtil now.
much work has been completed in search of the mechanisms involved. Loeschcke (Loeschcke
and Gertz 19%) found that acid infusions into the founh cerebral ventricle produced a
substantial increase in ventilation. and established the hypothesis that there exist sites within the
brain sensitive to acidification of the cerebrospinal fluid.
Mitchell expanded upon Loeschcke's work by demonstrating that exposure of a localised
subarachnoid region of the ventro lateral medullary surface to acidic solutions produced a large
increase in ventilation (Mitchell et ai.. 1963). Through Loeschcke and Mitchell's e'xperiments. a
topographical map was made in which two areas were outlined as the putative sites of central
chemoreception (Figure 1). These two areas came to be known as the rostral chemosensitive
area (Mitchell's area). and the caudal chemosensitive area (Loeschcke's area). A third site on
the ventral meduilary surface lying between the rostral and caudal chemosensitive areas was
dûcovered by Scbefke (Schiaefke et al.. 1970). This site was named the intermediate
chemosensitive area (Schlaefke's area figure 1 ) and was not thought to be chemosensitive but
was considered important because ventilation and chemosensitivity were depressed subsequent
to its cooling (Cherniack et al.. 1979; Millhom et al.. 1982). Figure 1 below shows the three
chemosensit ive areas of the medulla (rostrd. caudal and intemediate).
C hemosensitive Areas
Ventral Aspect
Figure 1 - The diagrarn above shows the anatomical location of the chemoreceptive sites on the ventral surface of the medulla. These three sites are represented by M, S, and L. M refers to Mitchell's are% S refers to Schlaefke's area, and L refers to M. Loeschcke's area. The p i a re s associated with each site are photographs of the researchers after whom the sites are named.
For rnany years, the central chemoreceptors were believed to reside within a few hundred
rnicrometers of the ventral meduüary surface in these three areas. However. a growing body of
evidence supports the notion that central chemoreceptors are more widely distnbuted throughout
the medulla. For example. Lipscomb and Bo yarsky (Lipscomb and Bo yarsky. 1 972) suggested
that the acidic solutions applied to the ventral rnedullary sudàce could be transported deep into
the brauistem thus stimulating chemoreceptors at sites other than those proposed by the work of
Loeschcke and Mitchell. As weii. severai radiolabel studies (Keeler et al.. 1984: Nattie et al..
1988: Yarnada et al.. 1984) have indicated that both large and smail radiolabelled molecules are
transported deep Uito the meduiia when applied to the ventral meduliary surface. further
supponing the idea that surface application rnay influence deeper structures. Thus. the
groundwork was laid for the possible existence of central chemorecepton at sites within the
medulla away from the ventral surface.
1.2.3) Recent Developments
The experiments of Nattie. summarised in his review (Nattie. 1999). provided a wealth of
evidence in support of a wide distniution of central chemoreceptors throughout the meduiia.
Nattie used microinjections of acetazolamide to show that medullary chemoreceptive sites are
widespread (Coates et al.. 1993) and include the VRG (Nattie and Li. 1996). the midline raphe
(Bernard et al.. 1996: Wang et al.. 1998; Wang and Richerson 1999). as weii as the
retrotrapezoid nucleus (Akilesh et al.. 1997; Li and Nattie. 1997a: Nattie and Li. 1994). More
recently Nanie's group have used a COz diffusion pipette (Li and Nattie. 1997a): it is capable of
produchg a quickly reversible focal acidification of neurod tissue through the diffusion of CO2
through the tip of the pipette. with the extent of acidi6ication varying with the concentration of
COz circulated through the pipette. These investigators found that in anaesthetized adult rats.
focal acidification of the retrotrapezoid nucleus increased phrenic nerve amplitude. They also
determined the extent of their acidification through the use of tissue pH microelectrodes. Nattie
argues that the high solubility of CO2 in tissue and blood allows for its widespread diffusion
throughout the rnedulla. where it will determine [W. This view is in contrast to the classical
assumption that the ventral surface. because it is the initial site of exposure to the blood supply
f?om the basilar artery, is the location rnost capable of providing a rapid response to CO?/[?-f].
Several studies have used quite a dSerent approach in determining the potential sites of
central chemoreception. These studies exposed intact animak to hypercapnia and examined the
brainstem for the expression of the imrnediate early gene product c-fos (Haxhiu et al.. 1996:
Miura et al.. 1996: Miura et al.. 1994: Sato et ai.. 1992: Tepperna et al.. 1997). They showed that
COz activates neurones in widespread sites throughout the medulla and confhed the
involvement of the ventral medullary surtàce. but they do not specificaily identiQ
chemoreceptors.
The experiments of Ichikawa provided indirect support for the concept of widespread
locations of central chemoreception (Ichikawa et al.. 1989). He completed a fàscinating
experiment. which involved measuring the extraceilular pH within the medulla in vivo during
infusion of a hypercapnic solution of saline via the intravertebral artery. The experiment
demonstrated that the tissue pH of the meduiIa fo Uowing injection of the solution was not
unifiody distniuted but heterogeneous and depended on location. He postulated that sites in
the medulla e?duiiting a Io w pH in response to COr would logicdy prove to be the locations of
central chemoreceptors. and indeed they corresponded to rnany of the sites discovered by Nattie.
More recently in vitro preparations have also been used to examine the question of
central chemoreception. Decreases in pH of the solution bathing these preparations stimulate
neural activity in bot h ventral and dorsal areas of t he medulla (Fukuda et al.. 1 980: Morin-Sum
et al.. 1995; Okada et al.. 1993b). and include the locus coeruleus (Ruiz-Ortega et ai.. 1995). the
caudal midline raphé (Richerson 1995; Wang et ai.. 1998; Wang and Richerson 1999). and the
nucleus tractus solitarius (Dean et al.. 1989). For example. Dean (Dean et al.. 1989) used
medullary slices whose ventral portions were removed in order to eliminate synaptic input from
ventral medullary surface chemoreceptors to demonstrate that cells within the nucleus tractus
solitarius increase theù firing in response to increases in Pco2. and that ceils in the nucleus
ambiguus do no t. Ho wever. none of these studies specifically ident ifL chemosensitive neurones
but only neurones whose activity increases as a result of local or global changes in C 0 2 / [ q .
Nevertheless. these in vitro preparations can be used to h d central chemoreceptors.
Severai studies using the neonatal rat brainstem-spinal cord preparation have demonstrated the
existence of central chemoreceptors at sites deep within the medulla (Issa and Remmers. 1992:
Kawai et al.. 1996) and within the pre-Botzinger complex (Johnson et al.. 1998). In addition.
both extracellular recordings (Jaro lirnek et al.. 1 990; Richersoa 1 995) and intracellula
recordings (Kawai et al.. 1996: Weber-Kienitz et al.. 1998) of intrinsically [q chemosensitive
neurones have been made in these in-vitro preparations.
SirnilarIy. in a recent experirnent Okada (Okada et al.. 2000) dernonstrated that voltage
sensitive fluorescent dyes applied to the transverse medullary slice of a neonatal rat could be
used to visually identifl regions in which ceiis become active as a consequence of hypercapnic
exposure. He showed that neurones near the ventral surface as weii as at deeper sites like the
ventral respiratory group. and nucleus raphé paihdus and obscurus were activated. By blocking
svnaptic transmission Okada showed that neurones intrinsically sensitive to COr exist in the
surface layer of the ventral medulla and that theu excitation is transmitted to deeper areas
through a synaptic connection.
2.2.4) The Centrai Chemoreceptor Stimulus
Although the approximate anatomical locations of the central chemoreceptors are known.
it remains unclear whether the central chernosensitive mechanisrn is located intra- or
extraceUularly and whether the stimulus is CO? or [m. In solution COz is hydrated into
carbonic acid which itself freely dissociates to give a proton and a bicarbonate ion. This reaction
is catalysed by the enzyme carbonic anhydrase. The reaction equation is shown below:
CO2 + H z 0 u &CO3 IV + HCO;
This equation demonstrates that increases in COz cause increases in the concentration of
H'ions in solution leading to a decrease in pH. It is assumed that COz. able to diffise through
aqueous membranes Iike the blood-brain barrier fkeely. acts to stimulate the central
chemoreceptors either directly through its action or indirectly through altering pH. The high
rate of diffusion and ease of solubility ofCOz therefore rnakes it difficult to disceni whether COz
or [q is the signal for the central chemoreceptors.
Nattie (Nattie. 1999) has argued that changes in COr would be sensed equaiIy as quickly
at either an intraceUu1ar or an extracellular site due to its high rate of diffusion. but changes in
pH would not be sensed as quickly intracellulady as it would be extracellularly. Moreover.
intraceilular pH is subject to tight regulation. Therefore. an e~acellular location for the
chemosensors appears to be a more logical choice. Nevertheless. it remains undetermined
whether the central chemosensors are located ôt either intracellular. extraceilular. or both sites.
and in addition whether either COz. [m, or a combination of both is the transduced signai used
to provide information to the respiratory control centre.
For exarnple. Harada (Harada et al.. 1 M a ) has shown that responses to both COz and
[m exist in the isolated neonatal rat brainstem preparation. Rigano (Rigatto et al.. 1994) has
shown an identical response in cultured neurones taken from the neonatal rat medulla to both
COr and [m. By contrat. Neubauer (Neubauer et al.. 199 1 ) has shown in a similar experiment
that cultured neurones are responsive to changes in CO? but not to changes in [Hl produced
under isocapnic conditions. E xperiments using meduilary slice preparat ions have no t helped to
conclusively ident@ the roles of CO2 and [m either. Fukuda and Honda(Fukuda and Honda
1976) have shown that increasing [m under isocapnic conditions acts to stimulate neurones
near the ventral meduliary surface. However. Miles (Miles. 1983) aithough verifjmg this result.
fùrther demonstrated that this isocapnic [ H 1 mediated stimulation occurs with other neurones in
the medulla away from the ventral surface. equally as weii. Further contradictions exist.
Tojima (Tojima et al.. 1991) has demonstrated that a rise in [KI while PC02 is held constant acts
to stimulate certain neurones in the medulla while Fukuda (Fukuda and Honda, 1983) has show
that decreased pH caused by an increase in Pcoz with bicarbonate held constant stimulates other
neurones.
1.2.5) The Involverneni of Carbonic Anhydrase
The enzyme carbonic anhydrase. present both intracelluiarly and enracellularly in
neuronal tissues. is a zinc metdoenzyme known to catdyze the reversible hydration ofC02
(Neubauer. 1991). Histochemistry bas shown that carbonic anhydrase is found in the medula of
the cat in an area close to the ventral meduliary surface. medial to the roots of the hypoglossal
nerve. in the cell membrane of some large neurones. in the capillary endothelium. and in glial
c e k (Ridderstrale and Hanson 1985). Ridderstrale has also demonstrated that many neuronal
ce11 bodies. dendrites and axons in the medulla contain carbonic anhydrase.
Torrance (Torrance. 1 993) has &O shown that carbonic anhydrase exists within the
medulla and that it hydrates COz near the central chemoreceptors. He postdates that carbonic
anhydrase could act at the central chemoreceptors either in the intracellular space or at the ce11
surface as a membrane bound enzyme. Carbonic anhydrase could be involved in central
chemoreception by assisting the transduction of CO2/[K] into a neuronal signal (Neubauer.
1991): the presence of intraceilular carbonic anhydrase would act to accelerate the hydration of
CO2 and thus the rapid decrease in intracellular Ievels of [HT. in the face of a slower
intracellular regdation of pH. In any case. it seems likely that by allowing the rapid transduction
of a rise in COz into a decrease in pH. carbonic anhydrase rnay affect the rate of response of the
central chemoreceptors to a CO-[HT stimuIus by accelerating COd[H'] equilibration.
The reversible carbonic anhydrase inhibitor acetazolamide has k e n used in many studies
involving the de t eda t ion of the egects of carbonic anhydrase inhibition on respiration
(Torrance, 1993. Torrance. 1 996. Neubauer. 1 99 1. Natt ie. 1 999). Acetazolarnide is used in the
treatment of acute mountain sickness Ui humans because it causes COz retention and thereby
stimulates ventilation and leads to increased tissue o.xygenation (Laux and Raic hle. 1 978). With
respect to the investigation of central chemoreception. a number of studies have used
ace tazo lamide to inhi'bit car bonic anhydrase.
Adams found that application of acetazolamide to the brain tissue in the in vivo
anaesthetized adult rat preparation increases the rebreathing response to COz. implying that
central chemosensitivity provides a greater response to CO2 upon carbonic anhydrase inhibition
(Adams and Johnson. 1 990). By contnist. others found that inhibition of carbonic anhydrase
with acetazolamide produces a response of normal magnitude but of a slo wer thne course in
response to a step change in PC02 in the meduiîa (Coates et al.. 1993) and in the isolated carotid
body (Iturriaga et ai.. 1 99 1 ). Furthemore. it has ken shown that the respiratory fiequency
response to CO2 in the in vivo anaesthetized rat preparation is slowed upon the inhibition of
carbonic anhydrase with acetazolamide (Tojima et al.. 1 988).
As mentioned previously. Nattie (Coates et al.. I993: Coates et al.. 199 1 ) used focal
acetazolamide injections in the in vivo anaesthetized adult rat preparation to probe the location of
central chemoreceptors. The focal acetazolamide injections produced a local tissue acidosis that
resulted in increased phrenic nerve activity. indicating the presence of central chemoreceptors at
widespread locations within the medulla.
1.2.6) Preparations
In my experiments I utilized the in vitro brainstem-spinal cord preparation. and the
using suction electrodes. Details of the experirnents are provided below but bnefly the protocols
used were as follows:
a) To detect the presence of comrnon short-the scale synchronisation of le ft and right phrenic
nerves in the brainstem spinal cord preparation I cross-correlated the discharge recorded
tiom left and right phrenic rootlets and examined them for broad peaks at tirne zero in siu
preparations. I also cross-correlated the discharge between ipsilateral phrenic nerve rootlets
in one preparation.
b) To demonstrate the CO2 sensitivity of the transverse medullary slice preparation I recorded
the response of the rhythmic hypoglossal rootlet discharge to changes in the pH of the
bathing aCSF as weU. as the time course of the pH changes in six preparations. I used a pH
of 7.42 as a baseline and then decreased pH in steps to 7.70 and then 7.00. allowing at least 5
minutes between changes for equiliiration.
c) To test whether carbonic anhydrase is involved in the central chemoreceptive process. I
compared the response of the hypog lossal disc harge of the transverse medullary siice
preparation to pH variations in the aCSF. with and without the presence of acetazolamide (a
carbonic anhydrase inhibitor) in the aCSF bathing solution in thirteen preparations.
d) To test the applicability of the pH sensitive dye fluorescent dye BCECF-AM for visualizing
focal areas of acidificat ion wit hin the transverse rneduhry slice preparat ion I carried out a
series of experiments as follows. First. 1 measured the pH sensitivity of both the cleaved and
un-cleaved t o m of BCECF-AM. the rate of decay of fluorescence emission for cleaved
BCECF-AM. and the optimal concentration ofcleaved BCECF-AM for maximal
fluorescence emission. Then, 1 measured the changes in fluorescence of one BCECF-AM
labeiied transverse meduiiary slice preparation in response to changes in the COr content of
the bat hing aCS F.
e) To test the feasibility of ushg a CO2 diffusion pipette to focally acidw regions of tissue to
probe the location of central c hemorecept ive sites within the transverse medullary slice
preparation. 1 appiied it to various areas while recording the hypogIossai nerve rootlet
discharge in nine preparations.
2*2) The Preparations
2* 2- 1) aCSF Preparation
In order to deliver o'xygen to neurones within the brainstem-spinal cord and transverse
meduiiary siice preparations. experiments were conducted in a recording chamber (2.5 ml)
through which a solution termed 'art ificial Cerebro-S pinal Fluid' (aCSF) constant ly flowed. The
solution is named for its similarity in composition to that of the cerebro-spinal fluid to which the
neonatal rat brain is exposed in the intact anunal.
The aCSF composition was as follows: 125 mM sodium chloride (NaCi). 25 mM sodium
bicarbonate (NaHCOs). 30 mM dextrorotatory-Glucose. 3 m M potassium chloride ( K I ) . I mM
potassium dihydrogen ortho-phosphate (KH2P04). 2 rnM calcium chloride (CaCI?). and 1 mM
magnesiurn sulphate (MgSO4) in distiiled H20. Once the components were combined the aCSF
was well mived and lefi to cool in a retiigerator to a temperature of 18 OC. From this mixture ice
cubes. used to cool the aCSF bathing the preparations. were made in a standard ice tray ( 12 ice
cubes with a total volume of 300 ml).
I controiied the pH of the aCSF bathing solution by bubbling it with specük gas
mixtures. The baseline pH of the bathing solution was maintained at 7.42 by bubbling it with
carbogen (5% C02/95% Oz). To decrease the pH to approxirnately 7.20 and 7.00.1 increased
the percentage of CO? bubbled into the aCSF by 5% and 10%. respectively.
2.2.2) The Brainstem-Spinai Curd Preparation
The brainstem-spinal cord preparation implemented in the experiments was prepared as
outlined by (S unie. 1 984). Sprague-Dawley neonatal rats were anaesthetised using 3 %
Halothane pas (Halocarbon laboratones Inc.) in oxygen and were sectioned caudal to their fiont
legs. The head and t o m ofthe rat were isolated and placed into a dissection charnber filled with
cold aCSF (-12 OC) bubbled with carbogen (95% O?/ 5%C02) at a pH of 7.42 (pH meter. Hanna
Corporation). The skin covering the skuil. along with the fiontal and occipital bones. were
removed using surgical scissors. The portion of skuil covering the cerebeiium was removed.
dong with the dorsal portion of the vertebrae. taking extreme care not to rupture any spinal or
cranial nerves in the process.
The entire brain and spinai cord (cerebellum cerebnim midbrain and spinal cord) were
then excised fiom the rat by cutting al1 connecting nerves as close to their distal end as possible.
Following this excision the c e r e b m was removed through a transection at the intercollicular
leveL and the cerebeilum was removed as weil leaving the brainstem-spinal cord free for
experimentation. It is important to note t hat shorter times of complet ion provide more viable
preparations in t e m of their ability to produce a rhythrnic respiratory output. The time taken in
producing each brainstem-spinal cord preparation was approximately 8 minutes. and was
ni t t ic idy short for maintenance of viability.
Following its removal. the brainstem-spinal cord preparatim was placed in the recording
chamber on top of a nylon mesh and perfused at a rate of 20 dminu te with aCSF at a
temperature of 25.0 OC and bubbled with carbogen gas to achieve a pH of 7.42. In order to
maintain the temperature of the aCSF bathing solution within the recordkg charnber constant. an
automatic temperature controiier (mode1 TC-324B. Warner Instrumentation Corp.) was used. To
stabilize the preparation two smaii tungaen rods (diameter = 70 prn length = 1 cm) were
inserted through its caudal and rostral ends into Sylgard ( 184 Silicone Elastorner. Dow Coming)
Lining the bonom of the recording chamber.
Care was taken not to dismpt the spinal cord above the level of the 7" spinal nerve. and
the medulia caudal to the level of the pons. Once stabilized with regards to position.
temperature. and pH. the dura covering the hypoglossal and spinal nerve rootlets of the
preparation was removed using micro forceps (Dumoxel non-magnetic microforceps #5.
A.Durnont & Fils Co.) with the aid of a 40x dissection microscope (Wild Heerbrug Co.). The
spinal nerve rootlets were then carefully separated fkom one another using micro forceps in order
to clear them for electrophysiological recording. It is important to note that extreme care was
taken in order to ensure that the rootlets were not damaged.
2.2.3) The Transverse Medu f fury S k e Prepuration
The transverse meduliary s lice preparat ion used for the central chemosensitivity
experiments was that pioneered by (Smith et al.. 199 1 ). The initial steps in obtaining the slice
preparation were the same as those outlined for the brainstem-spinal cord preparation in section
2.2.2.
In preparation for making a slice. 50 ml of distiued water was heated to a boil in a 230 r d
pyrex beaker and to it was added 3.5 gram of agar (Bactom Agar. Becton Dickinson & Co.).
This agar solution was affowed to cool until hard. and was then cut to produce a small rectangle
(approxhately 10 mm x 8 mm x 4 mm). which was glued to a plastic vibratome mount using
cyanoacrylate glue (Krazy glue. Elmer's Co. ).
Once the brainstern-spinal cord preparation was obtained. Î t was removed kom the
dissection tray fîlled with bubbied aCSF using a d spatula dried carefuily with filter paper.
and its dorsal end was glued to the mounting bIock/agar rectangle with the rostral end down (see
figure 2.2.3 on the folIowing page) using the cyanoacrylate glue. The mounting block was
Vibratome Preparation of the Transverse MeduIIary SIice
Agar Block
II
- - -
Inrtia! Cut -- - - A -
Vibratome Am
Figure 2.23 - The diagram above shows the brainstem-spinal cord preparation mounted ont0 the vibratome block in preparation for sectioning to produce a transverse medullary slice.
placed into the tray of the vibratome (mode1 752M. Vibroslice. Campden Instruments). which
was Nled with cold (-12 OC), carbogen gas bubbled (pH = 7.42) aCSF. The vibratome iight was
adjusted in order to illuminate the ventral surface of the brainstem-spinal cord so that the dura
and the hypoglossal nerve rootlets underlying it were h l y visible.
An initial cut was made using the vibratome. removing the section of tissue caudal to the
spinal medullary border in order to expose the hypoglossal nerve rootlets of the preparation to
full view. Using two sets of micro forceps (Dumoxel non-rnagnetic microforceps. A.Dumont &
Fils). the dura was removed ffom the ventral surface of the preparation in order to fkther expose
the hypoglossal rootlets to plain view. Then the blade of the vibratome was positioned
approxirnately 10 pm rostral to the rostrai-most hypoglossal nerve rootlet. and the blade was
moved upwards (caudal) to the desired thickness of the slice. The slice thickness in al1
experirnents varied between 400 and 1 200 prn with a mean thic kness and error of 930 7- 24
microns.
Once the blade was positioned at the desired level. al1 hypoglossal rootlets caudal to the
location of t he intended cut were placed rostral to the edge of the blade in order to avoid their
king sectioned. and the vibratome was set in motion at its lowest speed and highest vibration
rate. As weii. the light was turned off in order to maintain a low temperature w i t h the
preparation. The blade was retracted following completion of the cut and the iight was t m e d on
in order to illumuiate the slice. The ronral portion of tissue produced was removed. and the
blade was repositioned to the previously deteniiined location that king 10 pm rostral to the
rostral-rnost hypoglossai roo tlets. Once the location of the cut was verified. the light was turned
off again and a cut was made exactly as descnid previously. The transverse meduilary siice
produced was removed fkom the vibratome chamber and placed into a 250 ml pyrex beaker füled
with cold (-12 OC) carbogen bubbled (pH = 7.42) aCSF. and aiiowed to sit for 20 minutes in
order for neuronal function to recuperate (Paton et al.. 1994).
Following the 20 minute recuperation period. the slice was placed into the recording
c h b e r as outlined in section la and s w e d by inserthg two srnail tungsten rods through the
attached agar backing and into Sylgard gel (Sylgard 1&4 elastorner. Dow Corning Co.) c o v e ~ g
the bottom of the recording chamber. ïhe slice was positioned in the chamber such that the agar
backing was oriented towards the hcoming flow in order to ensure mechanical stability of the
slice and perfusion of both its sides. Once secured. the KCL concentration within the aCSF
bathing solution was increased to 8 mM by slowly adding 2.5 ml of 1 M KCI to the 500 ml of
aCSF in the reservoir over a period of 20 minutes in order to depolarise the neurones within the
prepmion and thus facilitate their firing. it was important to add KCI to the aCSF reservoir
only after the slice was placed in the recording chamber in order to provide a slow. graded
increase in [KCI] as it was my experience that rapid step increases in [KCL] are detrimental to
the viability of the slice. As weli. the temperature of the bathing solution was slowly (- 0.5
"C/minute) increased using the temperature controiier to a temperature of 25.0 O C . The
hypoglossal nerve rootlets were prepared for electrophysiological recording exactly as described
for the phrenic nerve rootlets of the brainstem spinal cord preparation (see section 2.2.3).
2.2.4) Nerve Recordings
The activity of newe rootlets was recorded using glas suction electrodes prepared by
precise manipulation of the suction electrode tip. A 10 ml W g e attached to the microelectrode
holder was used to provide the suction necessary to pull a single phrenic nerve rootlet hto the tip
of the suction electrode. Once the nerve rootlet was secured within the suction electrode. and the
rootlet demonstrated evidence of respiratory bursting (discernible bursts seen on the
oscilioscope), it was possible to begin recording the respiratory bursting pattern of t he
preparat ion.
The recorded signal was arnpiified using a pre-amplifier (HS-IA Headstage pre-
amplifier. Axon instruments Inc.) and an amplifier (Neurolog. NL 104). The amplified signal
was then £iltered (bandpass 0.12-8 kHz) and Uitegrated ( t h e constant = 50 rns). The resulting
signals were displayed on oscilioscopes (Tektronics. Nicolet) and monitored using loudspeakers.
A thermal anay chart recorder (Graphtec. WR3600) provided a permanent record of the signals
and a digitised videotape recording (Vetter) was also made for archival purposes. Figure 2.24
on the foilowing page illustrates the recording setup.
Recording Setup
Figure 2.2.4 - The diagrarn above shows the organization of the recording equipment used in both the brainstem-spinal cord and transverse medullary slice preparations.
2.3) Data Analysk
2* 3.1) Nerve recordings
To allow for stabiiization of the preparat ion. the rhythmic respiratory output of the s lice
was recorded for a period of approxirnately 40 respiratory bursts (Because of variation in the
bursting kquency between slice preparations a period involving a set nurnber of bursts was
chosen rather than a set period of tirne). For example. using t his criterion after making changes
to the pH of the bathing aCSF. at least 5 minutes passed before measurements were made. The
chart recordings of the rhythrnic respiratory output of al1 siice preparations were analysed to
obtain mean values for burst frequency. ampiitude. and duration as foiiows:
Mean bursr frequency was measured as the inverse of the mean t h e (seconds) taken for a
siice to complete 20 respiratory bursts (Hz).
Mean burst amplirude was measured as the mean height of the peak of the integrated
signal for 20 bursts.
Mean bursr hration was measured as the mean duration of 20 bursts.
2- 3. 2) Statisticu f Ana &sis
One-way repeated measures ahalysis of variance (ANOVA) tests of significance were
used to test (Sigmastat 1 .O. Jandel Corp.) whether the dserences in respiratory burst fiequency.
amplitude and duration of the slice at diffierent pH values were statistically significant. Because
of ditferences in recording characteristics between slices comparisons were only made within the
sarne stice preparations. In cases where natisticaily sigruficant differences were found. the
S tudent Neumann-Keuls test was irnplemented for comparisons.
2.4) Cross-Correiution
I used cross-correlation (Kirkwood. 1979) to detect short t h e scale synchronisation in
the respiratory buras recorded fiom lefi and right phrenic nerve rootlets (Cl-C5). For
Electronics Inc.). Then. cross-correlation histograms were computed on-line and occasionally
f?om the digital tape recordings. I used both a hard-wired cross-correlator (Anderson and Duffin
1976) as well as a cornputer, discriminated pulses were input via an A/D i n t d c e (AT-MIO-
16XE- 10. National Instruments). where specially written software (National Instruments.
LabVI E W). simultaneously cross-correlated neurone activity to phrenic and hypoglossal
disc harpe.
Features observed in the cross-correlation and post-stimulus h iaogam were quantified
using the k-ratio (Sears and Stagg. 1 976). For peaks. the k-ratio was calculated by dividing the
peak bin count by the mean bin count: for troughs. the sum of the mean bin count and the
difEerence between the mean bin count and the nadir for the trough divided by the mean bin
count determined the k-ratio. The mean bin count was measured away fiom any features present
in the histogram. Peaks and troughs were tested for statistical significance at P < 0.0 1 level
(Graham and Du& 198 1 ): only features shown to be statistically significant are reported.
Features were dso described by theù latency to the start. and the half-amplitude width of the
feature: values are eqressed as mean I standard deviation.
2.5) Acetazolamide
To apply acetazolarnide (Sigma chernicd company) to the transverse medullary slice
preparation I used two separate reservoirs containing 500 ml aCSF and bubbled with carkgen
gas to a pH of 7-42 (See Figure 2.5 on the following page).
The first reservoir contained no acetazolamide and thus acted as the control solution. A
kno wn arnount of acetazolamide (2 mM ACZ in the fist 3 preparations. 5 x I o4 M AC2 U1 the
second 4 preparations. and I rnM ACZ in the third 6 preparations) was added to the second
reservoir. It is important to note that KCl was applied to both reservoirs to increase its
concentration to 8 mM. as detailed in section 2.2.3.
The 2 rnM and 5 x 10" M solutions of acetazolamide were prepared by mixing the
acetazolamide directly with the aCSF and dissolved by n k g vigorously for 20 minutes using a
magnetic stirrer and a stir bar. However. in the case of the 1 mM AC2 solution to enhance its
solubility it waç dissolved in DMSO (dimethyl sulfoxide) pnor to its administration to the aCSF
(0.1 1 1 1 gram of acetazolamide in 1 ml of DMSO added to a reservoir containing 500 ml of
aCSF. provided a bathing solution with a 1 mM concentration of acetazolamide). The solution
was then stirred vigorously with a rnagnetic stirrer and stir bar for 70 minutes prior to use in
order to evenly distribute the acetazolamide. In order to maintah consistency between the
control solution and the solution containing the acetazolamide. 1 ml of DMSO was also added to
the control reservoir containing 500 ml aCSF.
Three-way valves were used to direct the flow into the recording chamber fkom either
of the two reservoirs so that I could quickly alter the bathing medium of the slice between the
control aCSF and the acetazolamide containhg aCSF solution. Nevertheless. d e r any switch
the slice was leA to equilibrate for a period of 20 minutes in order to ailow t h e for wash in or
washout of the acetazolamide and for binding or unbinding to the carbonic anhydnse present
within the siice.
Recording Setup for the BrainsternSpinal Cord and Transverse Medullary Slice Preparations
Recording Chamber ---- --
Ternperatwe Robe
Note - - = direction of K S F flow
Figure 2.5 - The diagram above shows the recording setup for the transverse rnedullary slice preparation. The setup above was used for the application of acetazolamide. The same setup was used for the brainstem spinal cord preparation. the only difference being that only one aCSF reservoir was used.
2.6) BCECF-AM
2. 6. I ) hepararion
BCECF-AM ( 2 ' , 7 ' - b i s - ( 2 - c a r b o x y e t h y l ) - 5 - ( a n d - 6 ) - c a acetoxymethyl
ester) is a membrane permeable fluorescent dye that. once within a ceL has its AM group
cleaved off by intraceilular esterases to produce its highiy active form BCECF. To prepare the
dye for experimentation 50 pg of BCECF-AM dye was hs t dissolved in 50 pl of Dimethyl
Sulfoxide (DMSO) and then sonicated in an ultrasonic cleaner (Cole Palmer) for 3 minutes. Low
Light conditions were maintained in al1 steps involving the dye in order to decrease the effects of
photobleaching. The resulting solution was termed the ' BCECF-.-1 M sruck solution ' and
provided a I pg/pl concentration of the dye.
To prepare the cleaved dye 1 followed the directions provided by its manufacturer
(Molecular Probes Inc. product information sheet on acetoxymethyl (AM) asters and diacetates).
I dissolved 50 pg of BCECF-AM dye in 50 pl of DMSO. and subsequently added 50 pl of
methanol. Following this. 1 added 25 pl of 2 M KOH (this is the active reagent involved in
cleavage of the AM group). 1 sonicated the vial containing the solution for 3 minutes. and the
via1 was lefi to sit for 60 minutes for cleavage to take place. At the end of 60 minutes 1 adjusted
the pH in the vial back to approximately 7.00 by adding 25 pl of 2 M HCI. Care was taken to
prepare the entire solution under low Eght conditions in order to minimize the effects of
photobleaching on the dye. The resulting solution was tenned cleaved BCECF-.4 IV sfock
soluhm and had a volume of 150 pl and a BCECF concentration of l pg/3pl.
BCECF solutions were prepared for fluorescence measurements as foiiows. First
pluronic acid stock solution was prepared by dissolving 200 mg of pluronic acid in 1 ml of
DMSO. The pluronic acid stock solution acts as a detergent and thus aids in distnîuting the
BCECF-AM dye evenly throughout the aCSF. Then, in a 250 ml beaker. 200 ml of aCSF at a
temperature of 25 OC was bubbled with carbogen gas (95% O?. 5% CO2). carbon dioxide. and
oxygen to obtain a desired pH. 300 pl of this aCSF was placed in a clear plastic via1 with a
volume capacity of 500 pl. dong with 1 pl of pluronic acid stock solution and 8 pl of the
BCECF-AM stock solution. Since the concentration of dye in the cleaved BCECF-AM stock
solution was one third that of the BCECF-AM aock solution 24 pl. as opposed to 8 pl of the
cleaved BCECF-AM stock solution was added to the viai containhg 300 pi aCSF and 1p1 of
piuronic acid stock solution. The solution in the vial was then mixed using a 2-20 pl
micropipettor (Wheaton Socorex) and placed under the fluorescence microscope and secured
with Plastik putty (Platignum inc.). Care was taken to quickly transfer the aCSF. pluronic acid,
and BCECF-AM stock solution into the vial (4 0 seconds) and capture an image immediately in
order to assure that the pH of the aCSF was rnaintained at the desired value.
2- 6- 2) Fluorescence Image Anuiysis
1 observed the fluorescence emission of the dye using a fluorescence microscope
(Olympus BXSO Wi fluorescent microscope with a Wi BA 530 nm ernission/excitation filter). and
too k pictures with a high performance CCD carnera (Sensicam) opera~ed by cornputer so ttware
(.&on Irnaging Workbench 2.2. Axon Instruments Inc.). To maintain consistency between the
vials. the microscope was focused just beneath the surface of the solution in the %st via1 viewed
and was not changed over the course of viewing the subsequent viak. As weil. the same
exposure t h e was maintained for each vial (The exposure was set to a value at which ali of the
pixels in each region of interest were w i t h the scale of measurement intensity).
1 used the histograms provided in Avon Irnaging Workbench Eorn each image captured
to provide a value for the emitred fluorescence of BCECF-AM dye. These provided an 8-bit
value of intensity (0-255) for each pixel in a region of interest. From the histogram 1 sumrned
the nurnber of pixels at that intensity level multip lied by their intensity to give an overall value of
ernitted fluorescence for the image that 1 temedfluorescence emission.
2- 6-3) BCECF Control Tesling
To determine the sensitivity of the cleaved and un-cleaved BCECF-AM dye to changes in
pH 1 measured the fluorescence emission at varying pH values (8.10. 7.42. 7.00. and 6.70). It
was important to ver% that the non-cleaved version of the dye did emit fluorescent light upon
excitation in order to determine whether the fluorescence emission of dye residing in the
extracellular space of the BCECF-AM labelled transverse medullary slice was sensitive to pH
variation.
To determine the effects of photobleaching on BCECF-AM. 1 measured the fluorescence
emission of a via1 containing 300 pl of aCSF (25 OC. pH = 8.00). 1 pl of pluronic acid stock
solution and 8 FI of the cleaved BCECF-AM stock solution each minute for 50 minutes. An
exponential regession iine was fitted to the data (Microsof€ Excel) to obtain a value for the
decay in fluorescence emission using an order and senes of 2.
To v e m the concentration ofcleaved BCECF-AM needed to provide an optimal value
of fluorescence emission (maximal fluorescence emission with minimal occurrence of self-
quenching). I measurinp the fluorescence emission of several vials. each containing 300 pl of
aCSF at a pH of 7.42 and a temperature of 25 OC. 1 pl of the pluronic acid stock solutioa and a
varying volume of the cleaved BCECF-AM stock solution ( 2 pl. 4 pl. 8 pl. 16 pi. and 32 pl).
These values encompassed the concentration of 8 pVpg recomrnended by the manufacturer
(Molecular Probes Inc.) for optimal fluorescence emission. A gaph of the tluorescence
emission venus the concentration of cleaved BCECF-AM was plotted for each via1 (Microsoti
ExceI).
2.6.4) BCECF-A M L abelled Transverse Medultary Stice Ekperiments
In one experiment 1 measured the tluorescence response of BCECF-AM to pH in the
transverse medullary slice preparation by recording the tluorescence emission at several pH
values (7.42. 8.10. 7.42. 7.00.6.70. 7.00. 7.42. and 8.10 in that order). A 500 prn thick
transverse rnedullary slice from a single 3 day old Sprague-Dawley rat was prepared and on
removal fiom the vibratome chamber was placed in a 500 pl plastic via1 containing 300 pi of
aCSF bubbled to a pH of 7.42. 1 pl of the pluronic acid stock solution. and 8 pl of the BCECF-
AM stock solution. and lefi to sit for 10 minutes in low light conditions before use. 1 maintained
the temperature of the recording charnber at 25.0" C. and ailowed a five minutes equilibration
period between each pH variation. 1 ploned the variation of fluorescence emission with pH on a
vaph (Microsoft Excel) to obtain a tluorescence ernission venus pH curve for the BCECF-AM z
labelled transverse medullary slice preparation. 1 did not record fiom the hypoglossal rootlets to
determine if the respiratory rhythm generator was operational. because my only goal in these
experiments was to demonstrate the fluorescence emission response of the dye-labelled slice:
tluorescence emission depends only on intracellular esterase activity for AM cleavage. An
rxponential regression line was fit to the tluorescence emission venus time gaph for the above
-35-
data (Microsoft Excel) to obtain a value for the decay in fluorescence emission with the
regression set to an order and series of 2.
1.6.5) CC2 Diffusion Pipette
1 tested the feasibility of using a CO2 di ffision pipette to probe the existence of focal sites
of central chemoreception withiri the transverse medullary slice preparation. labelled wirh
BCECF in nine experiments. With the diffusion pipette in place. a BCECF-AM labelled
transverse medullary slice was placed into the recording chamber. and the di fision pipette was
insened into the tissue of the slice in order to focally increase the concentration of COz. Images
of the slice were taken to visualize focal changes in pH.
The COz diffusion pipette was made from a glass suction microelectrode with a tip
diameter of 50 um and was c o ~ e c t e d to a preamplitier headstage micropipette holder with a
side port. and mounted on a micromanipulator to allow precise positioning as descnbed in
section 2. CO2 rich aCSF bubbled to a pH of 6.00 was circulated through the microelectrode in a
unique manner as follows (see figure 2.6.5 on the following page). A 50 ml syringe was tilled
with aCSF at a pH of 6.00 and placed ont0 an injector pump (World Precision instruments Inc).
The syinge outflow was connected to the inside of the microelectrode tip through the headstage
holder side port via P-50 Intrarnedic plastic tubing (Becton Dickinson and Company). The
syringe pump injected the aCSF at 1-1 0 mI/min so that the aCSF flowed into the microelectrode
tip. with the excess draining out through the side port of the headstage micropipette holder.
Figure 2.6.5 - The diagram above shows the difision pipette setup implemented in my experiments.
CHAPTER 3
RESULTS
3.1) General Results
These statements btiefly surnmarize my main findings with respect to the hypotheses
proposed in the introduction: details for each project follow.
1 ) 1 found a significant central peak when cross-correlating the activities recordcd tiom
ipsilateral phrenic nerve rootlets. indicating that they receive excitation from a common source.
However. in experiments cross-correlating the respiratory bursting recorded from leti and nght
contralateral phrenic nerve rootlets of the neonatal rat brainstem spinal cord preparation. I found
no central peaks: indicating that left and right phrenic motoneurone pools do not receive a
common activation.
2 I found that significant increases in the burst frequency. recorded from hypoglossal
nerves of the neonatal transverse medullary slice preparation. occurred upon providing a step
increase in COz tiom 0% to 10%. and a step decrease in CO2 from 10% to 0% to Vary the pH of
the bathing solution of the slice preparation.
3) The application of acetazolamide at a concentration of either 2 mM or 5 x 1 0 ~ M to the
neonatal rat transverse medullary slice preparation resulted in covering of the slice with
acetazolamide crystals. To prevent this crystal formation I dissolved acetazolamide in DMSO in
subsequent experiments. 1 subsequently found that application of 1 rnM acetazolamide dissolved
in DMSO to the transverse medullary slice preparation produced a significant change in
hypoglossal nerve burst duration. However. I found no sipificant changes in the sensitivity of
the slice to [H']/C02. with regards to burst frequency. amplitude or duration resulted fiom such
an application of acetazolarnide.
4) In quality control tests of the pH-sensitive fluorescent dye BCECF-AM. 1 determined that
both the cleaved and non-cleaved forms of BCECF-AM dye were sensitive to changes in pH.
that the fluorescence emission decay constant &,) of cleaved BCECF-AM is 0.01 02 min-'.
and that the concentration of cleaved BCECF-AM dye for optimal fluorescence emission is 8
pgi300ml. However. experiments involving staining of the transverse medullary slice
preparation with BCECF-AM dye demonstrated a faster rate of decay in tluorescence emission
than cleaved BCECF-AM dye alone. and no response in tluorescence emission to pH variation
was found.
5 ) While application of BCECF-AM dye to the transverse medullary slice preparation did
exhibit areas of differential tluorescence emission. nevenheless. the CO2 diffusion pipette failed
to produce a visually identitiable focal change in the tluorescence emission of the BCECF-AM
labeled transverse medullary slice preparation.
3.2) Cross-Correlation Experiments
3*2* I ) Left and Right Phrenic Nerves
I cross-correlated the rhythmic respiratory output tiom left and ri& phrenic nerve
rootlets in ~ neonatal rat brainstem-spinal cord preparations. The temperature of the bathing
solutions varied from 25.0 to 26.0 OC. with a mean * SE of 25.3 * 0.01 2 OC. the pH varied fiom
7.34 to 7.42. with a mean * SE of 7.40 5 0.1 6. and the age of the rats varied tiom 2 to 5 days oid.
with a mean L- SE of 3 m0.45 days old between preparations.
-3 9-
As figure 3.2.1 demonstrates the cross-correlation histopms did not display any central
peaks. The cross-cone1oga.m~ for al1 experiments are shown in appendix 1.1.
Cross-Correlation Histogram of Left and Rlght Contralateal Phmnic Nerve Rootlets
Figure 3.2.1 - Above is shown the cross-correlation his topm oftwo contralateral phrenic nerve rootlets tiom the brainstem-spinal cord preparation of a nvo day old nt. The gaph shows no indication of a central peak.
3.2.2) Ipsiloerai Ph renie Nerve Rootlets
1 cross-correlated the rhythmic respiratory output h m ipsilateral phrenic nerve rootlets
(C-4 to C-4) in one brainstem-spinal cord preparation. The temperature and pH were maintained
at 16.0 O C and 7.40 respectively throughout the experiment. and the rat used was 3 days old. nie
cross-correlation histopun for this experiment is shown in figure 3.2.2 below and in appendix
1.1. It displays a centrai peak. The mean bin count and the peak value of the histogram were
2000 and 25 10. respectively. Thus. the k-ratio was 1.255. and proved to be statistically
230Q --
\ l em Bin C'wnt = ?ûûû
Figure 3.2.2 - Above is show a cross-correlation histo-m o f ipsilateral phrenic nerve rootlets from the brainstem- spinal cord preparation of a three day old nt. The histograrn shows cvidence of a central peak.
3.3) CO2 Sensitivity in the Neonatai Rat Transverse Medullary Slice
1 probed & transverse meduliary slice preparations for their sensitivity to C02/[H-1. The
temperature of the slice bathing solutions varied from 25.0" C to 27.7' C with a mean * SE of
25.8 * 0.43' C. the age of the rats used varied from 2 to 8 days old with a mean * SE of 4 B0.93
days old. and the thickness of the slices varied from 700 to 1000 microns with a mean * SE of
900 -54 microns between preparations. The mean burst fiequency. amplitude. and duration
were cdculated (methods section 2.3.1 ) for each pH/% CO? value.
Mean Burst Fmquency vs. % CO, pH in the Tramverse Meduihry Slke Preparation
Figure 3.3.1 - Mean burst frequency = SE of six transverse medullary slice preparations vs. the pHiO'oCOL of their bathing solutions. Significant differences between coiumns are indicated by an asterisk. The n w data for the graph appears in appendix 2.1.
The burst tiequency vs. pH/% COz p p h (Figure 3.3.1 above) demonstrates that the
mean burst frequency of the slice preparations increases in response to a decrease in pH (as a
result of an increase in % CO2) of the bathing solution. A one-way repeated measures analysis
of variance ( ANOVA) test performed on the burst fiequency vs. pH data (appendix 2.1.1 )
indicated that significant differences in the burjt fiequencies of the preparations existed between
the initial pH of 7.42 and pH 7.00 (p=0.0046. power=0.9678). and between pH 7.00 and the final
pH of 7.42 ( p=O.O 1 86. powe~0.7528). The percentage increase in the frequency of bursting
îtom pH 7.42 to 7.00 was 54.7% (mean burst frequency increased From 0.1 1 7 Hz to 0.18 1 Hz).
Conversely. the percentage decrease in the fiequency of bursting from pH 7.00 to 7.42 was
54.7% (mean burst frequency *SE decreased fiom 0.18 1 + 0.034 to 0.1 17 * 0.0 13 Hz). No
other statistically significant differences between burst tiequencies were found. Neither burst
amplitude (Figure 3.3.2) nor burst duration (Figure 3.3.3) changed significantly with pH/ %CO2.
Figure 3.3.2 - Mean burst amplitude = SE of six transverse meduIIary slice preparations vs. the pHl'O/oCO2 of their bathing solutions. No significant differences were found between columns. The raw data for the :-ph appears in appendix 2.1.
Figure 3.3.3 - Mean burst duration * SE of six transverse medufla- slice preparations vcrsus the pH/?/oC02 of their bathing solutions. No signrficant Merences were found beween columns. The raw data for the graph appears in appendk 2.1.
One-way repeated measures ANOVA tests indicated no significant differences (p>0.05)
between the burst amplitudes (appendù 2.1 2) or the burst durations (appendk 2.1 -3) when the
%C02/pH of the bathhg aCSF was varied.
3.4) Acetuzoiamide
The concentration of acetazolamide necessary to block al1 carbonic anhydrase enzyme
activity was unknown nor were any indirect effects on tissues of the transverse meduiiary slice.
1 first used 2m.M. a high concentration- thinking to ensure that aii carbonic anhydrase enzyme
activity would be blocked and tested to see if the response to COz was blocked. It was not.
However. the siice viabiiity was adverseiy aected. and so 1 used a lower concentration of 5 x
1 o4 M and tested to see if the response to CO2 was reduced. It was not. and slice viability was
still not good. I f o n d that the application of acetazolamide c a w d crystals to form on the
surface of the slices and thought that they rnight account for the reduced viability. To prevent
crystals forrning in subsequent applications of acetazo lamide I used dimethyl sulfo xide ( DMSO )
to assist in the solubilization of the acetazolamide. 1 then tested the slice response to COz with
and without L rnM acetazolamide present and the same concentration of DMSO (28.16 mM)
throughout. The detailed tïndings for this set of experiments are provided below.
S. 4* 1) 2mM Acetnzolamide
1 probed fifieen transverse meduiiary slice preparations for changes in burst îi-equency
and burst amplitude responses to variations in C02/[HT] of the bathing solution in response to the
application of 2 mM acetazolamide. Only three of these Meen slices. however. provided a
rhythmic respiratory output throughout the entire experiment. In these three slice preparations
the temperature of the bathing solutions varîed fiom 26.2O C to 26.3 O C with a mean * SE of16.2
i 0.03"C. the slice thickness varied fiom 800 to 1000 microns. with a mean i SE of 900 * 58
microns. and the age of the rats varied fiom 3 to 8 days old with a mean * SE of5 1.5 days old.
The graph of bunt fiequency vs. pH is shown in figure 3 .4.1(a) below.
Bumt Fmquency M. pHACCO2 Before, During and Amr Application of 2 mM Acebizohmide
0% CO2
pH 7 42
Pm Aceiazdamide Contra
Figure 3.4.l(a) - Burst frequency = SE of three transverse rnedullary slice preparations in response to variations in pHI0'oCO2 of their bathin2 solutions during application of 2 mM acetazolamide. SipifIcant diff'rences between conditions are indicated by an asterisk. See appendix 3 . i for the MW data.
There was a 74.1 % rise in the mean burst frequency of the slice preparations upon
decreasing the pH of the bathing solution containine acetazolarnide tiom 7.43 to 7.00. A one-
way repeated measures ANOVA statistical analysis of the data (see appendix 3.1.1 ) indicated
that a statistically significant difference exists only between the 0% COz acetazolamide treatment
and the 1 0% CO2 acetazolarnide treatrnent (p=0.026. power=O. 835).
The _maph of burst amplitude vs. pH is shown in figure 3.4.l(b) below.
Burst Amplitude M. pHR6 CO2 Before, Durfng and A b r Application of 2 mM Acetmabmide
Figure
Pre Acetazdamide C m M
3.4.1(b) - Bunt amplitude = SE of three transverse medullary slice preparations in response to variations in pH/O/6CO2 of their bathing solutions durinç application of 7 m M acetazolamide. No sigificant differences between conditions exist. See appendix 3. I for the n w data.
The results of the one way repeated mesures ANOVA analysis performed on the mean
amplitude of bursting data (appendix 3.1.2) did not provide evidence of sta~istically significant
differences in the burst amplitude of the slice preparations at varying pW0hC02 values.
3* 4.2) 5x1 06 M Acefazolamide
1 probed nine transverse medullary slice preparations for changes in burst frequency.
amplitude. and duration in response to variations in pH/%C02 of the bathing solution pior to.
during. and following the application of 5x1 O* M acetazolamide. Only four provided a rhythmic
respiratory output throughout the entire expenrnent. The temperature of the bathing solutions for
these four slice preparations varied From 25.9' C to 27.1 O C with a mean *SE of 26.3 * 0.20 O C .
the age of the rats varied fiom 7 to 5 days old with a mean * SE of 4 * 0.75 days old. and the
thickness of the slice preparations varied fiom 1000 to 1 100 microns with a mean * SE of 1050
c 39 microns between preparations.
The _mph of burst frequency vs. pH / %CO2 is s h o w in figure 3.4.2(a) below.
Figu
Burst Frequency vs. pH 1 % CO2 Before, During, and After Application of 5x1 o4 M Acetazolamlde
Siiœ 2 O Siiœ 3 O Sliœ 4
re 3.4.2(a) - Burst frequency z SE response of four transverse medullary slice preparations to variations in pH/*6C02 before. during. and afler application of 5x 1 O' M acetazolamide. The frequency of bursting of the slices at pH 7.001 !O% CO2 was sigificantly different in all cases than the bursting at pH 7.42! 1 Ooh CO2 (asterisks). See appendix 3.2 for the raw data.
The magnitude of the mean burst Frequency increase in response to changing the
pH/%C02 of the baihing solution fiom 7.4210% to 7.0011 0% was 76.0% in the pre-acetazolamide
control. 48.2% in the acetazolamide treatment. and 103.6% in the post-acetazolarnide control. A
two-way (COz & Condition) repeated measures ANOVA. perfonned on the burst fiequency vs.
pH/% CO2 data (appendix 3.2.1). indicated that the diflerence in the mean values among the
three conditions (pre-acetazolamide. acetazolamide. and post-acetazolamide) was not significant.
-48-
after allowing for the effects of differences in pH/% COz (p=0.349. Power = 0.0753). In
addition. there was no statisticall y si@ ficant interaction between acetazolamide and p W%CO?
@=O. 126. Power = 0.273).
Furthemore. the difference in the mean values arnong the 0% CO2. 10% CO?. and the 2"*
0% CO2 levels was significant afier allowing for the effects of differences in acetazolamide
(p=0.0 12). A one-way repeated measures ANOVA test of the data (see appendix 3 2 . 2 ) . using
the Student-Newman-Keuls al1 painvise multiple comparison method indicated that the
differences between the 0% and 10% CO? g-roups. and between the 10% and the 2"" 0% COz
groups were signiticant. Both the nomality and the equal variance tests for the two way
repeated rneasures ANOVA performed on the fiequency of bursting failed (p=0.020 and
p=<O.OOl. respectively). while the normality test for the one way repeated measures ANOVA
passed (p=O.Oj 1 ) and the equal variance test failed (p=0.037).
The bmph of burst amplitude vs. ~H/O/OCO~ is s h o w in Figure 3.4.2(b) below.
Burst Amplitude vs. pH 1 % CO2 Before, During, and APter Application of 5 x 1 0 ~ M Acetazolamlde
il, Figure 3.4.2(b) - Burst amplitude r SE response of four transverse meduIlan, slice preparations to variations in
~H/? /OCO~ before. during. and after application of 5 x IO" M acetazolamide. There were no significant ditrerences. See appendix 3.2 for the raw data.
- - --- . -
.A two-way (CO? & Condition) repeated measures ANOVA. performed on the burst
amplitude vs. pH/% COz data (see appendix 3.2.3). indicated no significant ditferences between
conditions (pre-acetazoiamide. acetazolamide and post-acetazolamide). afier allowing for the
ctfects of differences in pH/% CO2 (p=O.J94. Power = 0.0502). In addition. thrre was no
statistically significant interaction between acetazolamide and pH/%C02 ( ~ ~ 0 . 6 0 1. Power =
0.0500).
Furthemore. the difference in the values among the 0% CO?. 10% CO?. and the 2" 0%
CO2 levels was not sipifkant (p=0.960. Power = 0.0502).
Both the normality and equal variance tests for the rwo way repeated measures ANOVA
performed on the amplitude of bursting failed (p=<O.OO 1. and p=<O.OO 1 respectively).
The decreased viability of the slice preparations exposed to 5 .u 1 O* M acetazolamide
(only 4 preparations of 9 proved rhythrnic throughout the experiment) was not much improved
-50-
compared to that seen with the application of 2 m M acetazolamide (3 of 15)
explanation, 1 viewed a slice under 40x magnification. The acetazolamide ha
dissolved in the aCSF and acetazolamide crystals were adhering to the surfac
figure 3.4.2(c) below). 1 postulated that these densely packed acetazolamidr
hindering the diffision of oxygen from the bathing solution to the tissue of tl
decreasing the viability of the preparation.
Figure 3.4.2(c) - A 4Ox magnifieci image of a transverse meduliap- slice preparation (tf line. horizontal m w ) shows the surface covered uith acetazolamide microelecuode can be seen in the top leil corner attached to a hyoglc m w ) .
3.4.3) ImM Aceta~iarnide in DMSO
1 pro bed nineteen transverse meduiiary slice preparat ions for their sensitivity. with
respect to burst frequency. amplitude. and duration to changes in pW%C02 of the bathing
solution prior to. during. and foiiowing the application of l mM acetazolamide dissolved in
DMSO. Only & provided a rhythmic respuatory output throughout the entue experirnent. The
temperature of the bathing solutions for these six slice preparat ions varied from 25.4' C to 76.7'
C with a mean * SE of 16.2 * 0.03 O C . the age of the rats varied from 1 to 8 days old with a
mean i SE of 4 * 1 days old. and the thickness of the slice preparations varied from 800 to 1 O00
microns with a mean * SE of 970 * 30 microns. Ail six slices were viewed under a
magnification of 100 x in order to determine whether any acetazolamide cqstals were adhering
to the siice surface. however no crystals were seen upon examination in any of the preparations.
The graph of burst fiequency vs. pW%C02 is shown in figure 3.4.3(a) below.
Figure 3.13(a) - The burst m e n - of six transverse meduil- s l ia preparations in response to kariations in pHP?CQ before, during. and d e r application of 1 mM acetazolamide in DMSO. The -en- of bursting of the sliœs at pH 7.00110% C a was sigrufïcar;~ different in al1 cases than the bursting at pH 7.42110% CG- (asterisks). See appendiv 3.3 for the raw data.
A two-way (CO2 & Condition) repeated measures ANOVA. perfonned on the burst
Eequency versus pH/%C02 of the slice preparations (see appendiv 3 -3.1 ). indicated no
significant dinerences between conditions (pre-acetazolamide. acetazolamide and poa-
acetazolamide). afler allowing for the effects of differences in pW% CO? (p=0.932. Power =
0.0500). In addition there was no statistically signtficant interaction between acetazo larnide and
pW%C02 (p=0.362. Power = 0.0746).
Furthermore. the dinerences in the mean values arnong the 0% COz. 10% CO?. and the
znd 0% CO? levek was greater than would be expected by chance d e r allowing for the effects of
differences in acetazolamide (p=<0.00 1. Power = 1.000). A one-way repeated measures
ANOVA test using the Student-Newman-Keuls a11 painvise multiple cornparison method.
appiied to the mean burst fiequency at various % CO2 values (see appendiv 3.3.2). indicated that
the dserences between the 0% and 10% CO2 groups. and between the 10% and the znd 0% CO2
were significant. Both the norrnality and the equal variance tests for the two way repeated
measures ANOVA on the fiequency of bursting failed ( p=O.O l3 and p=<O.OO I . respect ively ).
while both the normaiity test and the equal variance test for the one way repeated measures
ANOVA passed (p=0 .O5 1 and p=0.027. respectively ).
The graph of burst amplitude venus pW%COz appears in figure 3.4.3(b) below.
Figure 3.4,3(b) - The burst amplitude of sis m e r s e medullary slice preparations in response to cë,uiations in pH/%C02 before. during. and after application of 1 mM acetazolamidc in DMSO. There were no signifiant ciifferences betwccn columns. Sce appendix 3 .3 for the raw data.
A two-way (CO2 & Condition) repeated measures ANOVA. performed on the bum
amplitude versus pW%COr of the siice preparations (see appendiv 3.3.3). indicated that the
differences between conditions (pre-acetazolamide. acetazolarnide and post-acetazolamide) were
not significant. after aiiowing for the effects of differences in pH/% CO2 (p=0.717. Power =
0.0500). In addition. there was no statisticaiiy significant interaction between acetazolamide and
pW%C02 (p=0.303. Power = 0.102).
Furthemore. the dEerences in the mean values among the 0% CO2. 10% COz. and the
2" 0% COz Ieveis was not sieuficant after ailowing for the effects of differences in
acetazolamide (p=<0.868. Power = 0.050). Both the normality and the equal variance tests for
the two way repeated measures ANOVA on the fiequency of bursthg failed (p=O.OO 1 and
p=<0.001. respectively).
The graph of burst duration versus pH/?/oC& appears in figure 3 -4.3 (c) below
Figure 3.4.3(c) - The burst duration of sis transverse medullary slice preparations in responsc to variations in pH/??oCQ before. during. and d e r application of 1 mM acetazolamide in DMSO. Burst durations in the pre-acetazolamide control group differ signrficantiy from the values in both the acetazolamide treatment group and the pst-acetazolarnide conml group (astcrisks). See appendis 3.3 for the raw data.
.A two-way (COz & Condition) repeated measures ANOVA. perfomd on the burst
duration versus pW%COz of the stice preparations (see appendiv 3.3.4). indicate that the
diferences between conditions ( pre-acetazolamide. acetazo lamide and post-acetazo lamide) were
significant. afier allowing for the effects of differences in pW% COz (p=<0.00 1. Power = 0.996).
However. there was no statistically significant interaction between acetazolamide and pW%COt
(p=0.660. Power = 0.0500). Post-hoc tests (appendix 3.3.5). using the Student-Newman-Keuls
all painivise multiple cornparison method. of the conditions hdicated that the dEerences
between the pre acetazolamide control and the 1 rnM acetazolamide treatment groups. and
between the pre and post acetazolamide control groups were significant.
Furthemore. the difference in the mean values among the 0% CO2. 10% COz, and the Yd
0% CO2 levels was not significant. after allowing for the effects of differences in acetazolamide
(p=0.090. Power = 0.323).
Both the normaiity and the equal variance tests for the two way repeated measures
ANOVA on the fkequency of bursting faiied (p=<0.00 1 and p=<0.00 1. respectively ). while bo th
the normality test and the equal variance test for the one way repeated measures ANOVA passed
(~-00.06 1 and ~ 0 . 5 5 7 . respectively).
3.4.4) Rute of p H Equiiibrution
In order to ensure that I had ailowed enough t h e for equilibration between pH
variations. I analyzed the rate of pH equilibration of the aCSF bathing solution upon ôoth
increasing and decreasing the %CO2 before. dtiring. and afier the application of 1 mM
acetazolamide in DMSO. In the analysis. 1 used the pH and time data obtained From the ImM
acetazolarnide in DMSO experiment (appendiu 3 d). Figure 3.4.4 below shows both the rate of
decrease and increase in pH upon changing the % COz bubbled into the bathing solution fiom
0% to 10%. The graph demonstrates that it takes approximately five minutes to reach steady
state for pH following step increases (fiom 0% to 5%) or decreases (frorn 10% to 0%) in %COz
Time (min)
1 Figure 3.4.4- The time course of pH decreasc ( r d points). in response to a nep increase from 0% to 10% of the CO2 bubbled into the bathing aCSF. and pH incrcase Blue points). in pH in rcsponse to a step decrcase from 10% to 0% of the CQ bubbled into the bathing aCSF. The pH data was coltcctcd for thc pre- acetazolrnide controt (squares). the 1 rnM acetazolamide in DMSO trcatrncnt (circlcs). and the p s t - acetazolamide control (triangles). The raw data can be found in appendis
3.5) BCECF-AM Quafity Controf Euperiments
3-52) U'ncleawd BCECF-AM Sensitivity to pH
I completed one test in which I anaiyzed non-cleaved BCECF-AM dye with respect to
changes in its fluorescence emission intensity with pH variation. The temperature of the ACSF
in each of the vials was rnaintahed at 25.0" C. and the pH of the ACSF was set at 6.70. 7.00.
7.42. and 8.10. respectively. The calculated values of fluorescence emission for each pH value
appear in appendix 4.1 .1 .
The graph of fluorescence emission vs. pH is shown in figure 3 3 . 1 below. The data
indicates that the non-cleaved form of BCECF-AM is sensitive to changes in pH.
Figure 3.5.1 - The graph above demonstrates the nse in fluorescence emission of unclcaveci BCECF-AM dye in response to increases in pH. The raw data can be found in appendix 1.1.1.
The graph demonstrates the sensitivity of uncleaved BCECF-AM to variations in pH
between 6.70 and 8.10.
3.5.2) CIeoved BCECF-AM Sensifkviiy tu p H
1 completed tests in which 1 analyzed cleaved BCECF-.Ah4 dye for changes in its
fluorescence emission htensity with pH variation. The temperature of the ACSF in each of the
Mals was maintained at 25.0" C. and the pH was set at 6.70. 7.00. 7.42. and 8.10. respectively.
The caiculated values of fluorescence emission for each pH value in both experhents. dong
with the means standard deviations. and standard errors. appear in appendix 4 bl.
The graph of fluorescence emission venus pH is shown in figure 3-52 below.
-58-
Figure 3.5.2 - The graph demonstrates the i n c m in fluorescence emission of unclmvcd BCECF-AM due in response to increases in pH. shown for wo cxperirnental trials. The nw data a n bc found in appendix 4.1 2.
The graph demonstrates the sensitivity of cleaved BCECF-AM to variations in pH
between 6.70 and 8.10.
3.5.3) CIeuved BCECF-AM Fluorescence Decay Cime
1 completed test in which I analyzed the decay in fluorescence emission of cleaved
BCECF-AM dye over a 50 minute period. The temperature and pH of the ACSF of the via1
containing the dye were maintained at 35.0 OC and at 8.00. respectively. throughout the
experirnent. The fluorescence emission data for the fifty-minute anaiysis period appear in
appendix 4 c 1.
The graph of fluorescence decay over tirne is shown in figure 3.5.3 below.
-59-
Figure 3.5.3 - The decrease in fluorescence emission intensity of cicaveci BCECF-AM dye with time (one minute intervais) over a fif& minute period. The bIue &amonds rcpresent the individual data points. and the red line represents an e-uponcntid trendlinc fit to the data. The raw data cm be found in appendix
4.1.3.
The graph demonstrates the decay in fluorescence emission over fifly minutes. An
exponential trendline fitted to the data (Microsofl Excel) has both a series and order of 2. and is
described as y=2E+OSe -0.0 102r with an R' value of 0.998 1 . From this equation the rate constant
of decay (Lo) in fluorescence emission due to photobleaching of cleaved BCECF-AM dye is
0.0 1 02 min-': a tirne constant of 98 minutes.
3.5.4) Cïeaved BCECF-AM Dye Concen fraiion Test
1 cornpleted tests were in which the fluorescence emission of cleaved BCECF-AM
dye was observed over a range of dye concentrations. The concentrations used (expressed in
pg/3OO pl) were 2.4.8. 16. and 32. Exposure times were set at 5 ms For both trials in order to
rnaintain consistency. The results of both experirnents. dong with their mean standard
deviation. and standard error. appear in appendix 4.1.4.
The graph of fluorescence emission vs. concentration of cleaved BCECF-AM dye
appears in figure 3.5.4 beiow.
Figure 3.5.4 - The graph above shows the fluorescence emission of cleaved BCECF-AM dye at various concentrations for two trials Qellow and red lines). The raw data can be found in appndiu 1.1.4.
The graph demonstrates that cleaved BCECF-AM dye produces a mvirnal fluorescence
ernission at a concentration of 8 @300 pL which f& within the optimal concentration range of
6- 10 pg/300 pl cited by the manufacturer (Molecular Probes Inc.).
3.6) The BCECF-AM Lubelled Transverse Medulhry Slice Preparaîion
3.6.2) pH Response in the BCECF-AM Labelled Transverse MeduIlary S k e Prepurafion
1 completed one experiment in which the fluorescence emission at varyhg pH levels was
tested in an 800 micron thick transverse rnedullary slice preparation fiom a 3 day old neonatal
rat. The temperature of the bathing solution was rnaintained at 25.0" C throughout the
experiment . The values of fluorescence emission at each pH value appear in appendiv 5.1 . l .
The graph of fluorescence ernission vs. pH appears in figure 2.6.1 below.
Figu re 3.6.1 - The graph above shows the fluorescence emission at various pH values of BCECF-AM labclled transverse medullary slice preparation. The time between each column is five minutes. The red Iine is an e'tponential trendline fit to the data whose equation a p p r s at the ri@ side of the graph. The raw data can be found in appendix 5.1.1.
The graph indicates that the BCECF-AM labeled slice preparation is not responding to
pH variations due to the fact that fluorescence emission decreases despite inaeases or decreases
in pH. The rate of this decrease in fluorescence emission was determineci by fitting an
exponential trendluie to the data, with a series and order of 2 (Microsofl Excel). The equation
obtained from this fit is y = = + 07e-' with an R~ value of 0.9452. However, pH was
changed at five minute intervals. Therdore the rate constant of decay (Kdaiy) of fluorescence
emission in the BCECF-AM labelleci siice preparation is
0.05 19 - min"; a time constam of 19.3 minutes.
3.62) CC2 Diffwion Pipene Application in the BCECF-AM Labeled Transverse Meduiiary Sfice Prepïmtion
Nine experiments were completed in which a CO2 diffusion pipette was used in
conjunction with BCECF-AM Iabeled transverse medullary slices in order to produce. and
visually identify. focal regions of acidification withùi the tissue of the slices. In these
preparations. t here was no visual evidence of focal reg ions of acidificat ion despite repeated
efforts. Severai preparations (n=3) demonstrated a focal region surrounding the diffision pipette
in which the fluorescence emission was reduced. However. upon closer examination. it became
evident that the focal decrease in fluorescence emission was due to rupturing of the tissue
resulting corn a high tlow rate of CO2 rich ACSF out of the tip of the diffision pipette.
Figure 3.6.2(a) below pro vides visual evidence of this.
There were no preparations in which the fluorescence ernission of a focal region of tissue
surrounding the CO2 diffision pipette was altered as a result of a variation in pH.
Application of the CO, Diffusion Pipette to a BCECF-AM La belled Transverse Medullary Slice Preparation
7 Pipette Tissue)
BCEFC-AM Labeled Slice Preparation Prior to Diffusion Pipette Insertion
Diffusion Pipette Tip (In Tissue) Dark Area Indicates Ruptured Tissue t
I i Area of Rupture k
BCECF-AM La beled Slice Preparation Following Diffusion pipette Insertion
Figure 3.6.2(a) - The figure above shows two images of a BCECF-AM labelled transverse m e d u l l ~ slice. In the top image, a CO2 difision pipette is seen pnor to insertion into the tissue of the slice. The bottom image shows the slice following insertion of the CO2 ditfusion pipette tip. The circled dark area represents the location of tissue rupture. The red dotted line represents the midline -65-
3.6.3) Photobleaching and Dijferentid Fïuorescence Emission in the BCECF-AM Labeied Transverse Medulay Slice Preparation
1 completed 2 experiments in which 1 stained a transverse medullary slice with BCECF-
AM dye. In these experiments fluorescence emission in certain regions of the labelled slice
preparation appeared to be greater than in others. As welL 1 noticed that regions of the labelled
slice e.xposed to excitation light undenvent photobleaching. Figure 3.6.2(b). s h o w below.
provides an excellent demonstration of both dnerential fluorescence emission and
photo bleaching .
In figure 3.6.2(b). image 1 is a magnified (JO x) image of an 800 pm thick BCECF-AM
labeled transverse medullary slice bom a 3-day old neonatal rat. The temperature and pH of the
slice bathing solution were 35.0 O C and 7.42. respectively. The ventral surtàce of the medulla is
located at the top of the image. and a hypoglossal nerve rootlet is clearly visible. The colour
palette at the bottom of the diagram represents the fluorescence emission intensity uit h O
( purple) representing the lowest level of emission and 255 (red) representing the highen. The
circled area near the centre of the slice is an area that viewed under 100 x rnagmfïcation. Image
2 shows the rnagnified area of interest immediately upon placing the slice in the bathing solution.
The foiiowing images (3-6) were obtained at two-minute intervals. The rnagnifled area of
interest exhibits regions with dxerent levels of fluorescence emission and it would appear that
the intensely fluorescent region in images 2-6 is the central canal.
Image 1 was the obtained 2 minutes d e r figure 6. and demonstrates the decrease in
fluorescence emission of the area of interest as a result of exposure to the fluorescence excitation
light. The dark region at the centre of figure 1 shows the extent of photobleaching in the slice.
Area of Magnification
BCECF-AM Labelleci Transverse Medullary Slice
Ventral Surface -
Hypoglossal Nerve Rootlet
Area of Magnificution (1 00 x) at Varying Time Intervals
Fig. 2 Fig. 3 Fig. 5 Fig. 6
8-Bit Fluorescence Emission Palette Figure 3.6.2(b) - The figure above shows a BCECF-AM labelled transverse medullary slice preparation.
Image 1 is a 40x magnified image of the slice preparation following exposure of the circled area to fluorescent excitation light. Images 2 to 6 show the circled area of interest at 100x magnification at 2 minute intervals. The area of increased fluorescence. corresponding to bright green. in images 2 to 6 represents the central canai.
-67-
AU of the BCECF-AM labeled transverse medullary slice preparations demonstrated
regions of increased fluorescence exnission. Figure 3.6.2(c) shows an example of this differential
fluorescence emission. The diagram shows a 600 pm thick BCECF-AM labeled transverse
medullary siice from a 2-day old rat. The temperature and pH of the slice bathing solution were
25.0" C and 7.42. respectively. Image 1 shows the whole siice at 40 x rnagnitication with its
ventral surface at the top of the image. Two areas were viewed under 100 x rnagnification to
demonstrate the dflerential fluorescence ernission. The first area (Figures 2-4) was centred
about the mid the and showed increased fluorescence emission of the central canal. and the
second area (Figures 5-71 was centred around the ventral respiratory group and showed increased
fluorescence emission of the slice tissue in a h a n d p ~ t pattern delineating the pyramidal tracts.
11 is important to note that the region centred around the rnidiine consistently demonstrates an
increased fluorescence emission.
BCECF-AM Labeled Transverse Medullary Slice
Fig. 2
Area of Magnification for Figures 5-7
Hypogiossal Rootlet Ventral
Fig. 5
Fig. 3
Dorsal Fig. 1
Area of Magnification for Figures 2-4
Fig. 6
Fig. 4 Fig. 7
Figure 3.6.2(c) The figure above shows a BCECF-AM labelled transverse medullary slice preparation. Image 1 is a 40xmagnification of the labelled slice with a hypoglossal nerve root visible in the top left corner. The black circles represent two areas magnified 100x. The top circle represents images 5 to 7. and the -69-
CHAPTER 4
DISCUSSION
4.1) Genetuf
In my study of central chemoreception 1 first determined whether the central
chemoreceptors are effective in the transverse medullary slice. modulating hypoglossai motor
output in response to changes in the Pco2 of aCSF bathing the slice. 1 found that they were. To
gain experience for the stice experiments 1 had previously used the brainstem-spinal cord
preparation to test whether le fi and right phrenic nerves in the neonatal rat are sync hronized
because they receive a cornmon activation and found that they do not. In a hnher study of the
central c hemoreceptors. I tested the hypothesis that carbonic anhydrase is an essent ial CO rnponent
of chemoreceptor function in the transverse medullary slice. 1 determined that it was not. and
based on these hdings suggest that the central chemoreceptor mechanism may be extracellular
rather than intracellular. Finally. 1 tested a new technique to detennhe whether the locations of
the central chemoreceptors in slices could be resolved through the application of a pH sensitive
dye (BCECF-AM) in conjunction with a COz difision pipette. and was unable to do so.
In my experiments. 1 utilized both the neonatal rat brainstem-spinal cord and transverse
medullary slice preparat ions. The use of these in-vitro preparations as models of respiratory
rhythm generation and transmission in intact neonatal rats is accepted by sorne investigators
(Smith et al.. 1990. Smith et al.. 1991. Telgkamp and Ramirez 1999) but not by others (Fung et
ai.. 1994. StJohn. 1998). The reduced preparations are relatively hypoxic (Okada et al.. 1993a)
and at lower temperatures (26-29°C) than intact neomtes. although in this respect. the transverse
-70-
meduiIa. slice preparation (Koshya and Smith. 1999. Ramirez et al.. 1996b. Smith et al.. 199 1 )
offers superior control of the extracellular environment compared to the brainstem-spinai cord
preparation (Ballanyi et al.. 1992. Hayashi and Lipski, 1992).
Despite such diaerences between intact and reduced preparations. 1 believe that the
reduced preparations were usehl for my investigations as they contain sufficient neural circuitry
to generate periodic bursts of neural activity on phrenic and hypoglossal nerve rootlets (Koshiya
and Smith, 1999. Ramirez and Richter, 1996b. Smith et al.. 199 1. Smith et al.. 1990. Suzue,
1984). In addition. other investigaton have observed respiratory neurones with firing patterns
similar to those of the ventral respiratory group in adult rats: including pre-inspiratory
inspiratory. expiratory and phase-spanning neurones (Onunani and H o m 1992a. Onimam et
al.. 1992b. Ramirez et al.. 1996a Smith et al.. 1990). 1 therefore suggest that the in-vitro
brainstem-spinal cord and transverse medullary siice preparations contain the appropriate neural
circuitry to generate and transmit a rhythm to respiratory motoneurones.
Furthemore. both preparations detect and respond to pH changes in a manner similar to
those seen in intact rats (Issa and Rernmers. 1992. Johnson et al.. 1997. Peever et al.. 1999b). As
weil. increasing the temperature of the bathing solution of either preparation to a more
physiological level. the Eequency and pattern of inspiratory discharge of both phrenic and
hypoglossal nerves approaches that of intact age-matched neonatal rats (Peerer et al.. 1999b).
My experirnents also demonstrated that the transverse medullary slice responded to pH changes
in the bathing solution by altering the bvsting fkquency recorded kom the hypoglossal nerve
rootlets.
It could be argued. however. that if the rhythm generated in the reduced preparations is
not eupneic in character and corresponds to gasping then it is possible that there exist separate
transmission pat hways for each mode of rhythm generat ion. I f so. the rhythm transmission and
generation pathways in reduced preparations cannot be compared with those of intact rats or
adult preparations. Whiie this possibility cannot be ruled out. 1 do not consider it Wtely. Rather.
1 believe that the reduced preparations are appropriate models for comparison with adults in
detecting connection differences that are due to developmental changes. and to implement as
valid models for the audy of central chemoreception. Nevertheless. it should be noted that my
interpretation of the results is based on this premise.
4.3) Cornmon Acriv~ar'un Between Lefl and Right Phrenic N m e Rouilets in the Neontde
1 observed no central peaks in any of the cross-correlation histograms I computed
between the 1eR and right phrenic nerve rootlets in the neonatal rat brainstem spinal cord in-vitro
preparation. It is possible that the cross-correlation technique was not working properly in this
preparation. so I teaed it by cross-correlating the phrenic nerve discharge recorded from adjacent
phrenic nerve rootlets. 1 reasoned that they would be likely to share excitation from descending
projections arborising in the motor nucleus. The resulting cross-correlograms demonstrated
broad central peaks and I concluded that the cross-correlation technique was capable of detec ting
common excitation in the neonatal rat brainstem-spinal cord preparation. It demonstrated that
ipsilateral phrenic motoneurones receive excitation fiom a common population of inspiratory
premotoneurones. The central peak was broader than those found in adults. iikely due to an
increased range of transmission times because of the reduced temperature and lack of
myelination of in-vitro preparations (Fitzgerald. 1985. Kashiwagi et al.. 1993).
1 therefore concluded that leil and right phrenic motoneurones did not receive excitation
tiom a common source (Kirkwood and Sears. 1 99 1 ). This hding constitutes the first such
examination in neonatal rats (Peever et al.. 1999a) and is contrary to the case in the adult rat
where cross-correlation histograrns computed bet ween lefi and right phrenic nerve discharges
displayed central peaks (Dufi and van Aiphen 1 995: Tian and Duf£in. 1 996) interpreted them
as evidence that both lefi and right phrenic motor pools receive excitation 6om a cornrnon
inspiratory premotoneurone population.
In the adult rat. inspiratory premotoneurones in the ventral respiratory group have axons
that bifurcate to descend both IeA and right sides of the spinal cord (Dobbins and Feldman
1994). and it is Likely that these provide the common activation. but no information is available
on bulbospinal projections for the neonatal rat. It is possible that central peaks could result fiom
axons descendhg unilaterally from brainstem premo toneurones if t hey excite mo toneurone
dendrites that cross the midline of the spinal cord. While this anatomical feature exists (Allan
and Greer. 1997. Cameron et al.. 199 1 ). Peever (Peever et al.. 1999a) believes that it is
ineffective in synchronising le fi and right phrenic nerve discharges. His conviction is based on
two observations. First. the absence of any discemîble effect on phrenic discharge when the
spinal cord is transected in the region of the phrenic motoneurones in his experiments on the
neonatal rat brainstem spinal cord preparation. If a substantial proportion of phrenic
motoneurone drive had been provided from this source. a decrease in activity and possibly
desynchronisation of le ft and right disc harges should have been O bserved. Indeed. in other
experirnents (Peever et al.. 1998). a midhe transection of the adult rat brainstem resulted in
independent rhythmic activities of left and right phrenic nerves. and there was no evidence of
le fi-right synchronisation. Skdarly. Janczewski and Ao ki (Janczewski W k 1 997) fo und that
Standard Devlcrtlcm of Amplitude 0.943 i 0.775 1 0.328 1 0.644 SEM 1 0.544 1 0.447 1 0.1 89 1 0.372
.4 .X?. 1) Two way repeuted rneasures AN0 VA on hvo factors - 5 x 1 Q% acetazolamide and pHP/oC02 with bumi frequency as the dependent variable
'lote - ACZ (N/Y/R) rekrs to the pre acetazolamide control (N). the 5 x 10" M acrtazolamidr treatment (Y). and the post acetazolamide control (R). In addition. the % CO? values b. d. and a refer to the 0% (b). 10% (d). and 2" 0% (a) CO? exposures for the prepantions.
Balanced Design
Dependent Variable: Freq. (Hz)
Normality Test: Fai led
Equal Variance Test: Failed
Source of Variation Slice Num ber .AC2 (N/Y/R) ACZ tNN!R) x Slice Number %CO2 O C 0 2 x Slice Nurnber ACZ (NiY/R) i( OhCo2 Residual Total
The difference in the mean values among the different levels of ACZ (N/Y/R) is not -mat enough to exclude the possibility that the differexe is just due to random sampling variability afler alIowing for the effects of differences in 9hC02. There is not a statistically significant difference (p = 0.349).
The difference in the mcan values among the dif'ferent levels of ?/oc02 is greater than would be expected by chance after allowing for effects of differences in AC2 (NNI R). There is a statistically significant difference ( p = 0.0 12). To isolate which group(s) differ from the others use a multiple cornparison procedure.
The effect of different levels of ACZ (NNIR) does not depend on what level of ?CO2 is present. There is not a statistically significant interaction between ACZ (NNIR) and %C02. (P = 0.136)
Power of performed test wiùi alpha = 0.0500: for ACZ (N/YIR) : 0.0753 Power of performed test with alpha = 0.0500: for %CO3 : 0.835 Power of performed test with alpha = 0.0500: for AC2 WN!R) x ?/oc02 : 0.273
Least square means for ACZ (N/YIR) Croup Mean SEM n 0.214 0.1 13 v 0.197 0.1 13 r 0.219 0.1 13
Least square means t'or ?/oCOS Group Mean SEM b 0.167 0.115 d 0.292 0.1 15 a 0.171 0.1 15
Least square means for AC2 (N/Y!R) x 9'0C02 Group Mean SEM n x b 0.165 0.1 15 n x d 0.291 O. 1 15 n x a 0.186 0.1 15 y x b 0.173 0.1 15 ! x d 0.257 O. 1 15 v x a 0.159 0.1 15 r x b 0.162 0.115 r x d 0.329 0.1 15 r x a O. 166 0.1 15
A.3.2.2) One way repeated measures ANOVA on one factor -pH?% COr with brtrst frequency as the dependent variable
Note - ACZ (NIYIR) refers to the pre acetazolarnide control (N). the 5 xlo4 M acetazolamide treatment (Y). and the post acetazolamide control (R). In addition. the % COz values b. d. and a refer to the 0% (b). 10% (d). and 2"' 0% (a) CO2 exposures for the preparations.
Normality Test: Passed (P = 0.05 1 )
Equal Variance Test: Failed (P = 0.037) Test execution ended by user request. RM ANOVA on Rank begun
Friedman Repeated Measures Analysis o f Variance on R a n h
Group N iMissing b 12 O d 12 O a 13 O Tested 12 O
Chi-square= 13.167 with 2 degrees of fkeedom. (P = 0.00 1 )
The differences in the median values among the treatment groups are greater than would be expected by chance: there is a statistically significant difference ( P = 0.00 1 )
To isolate the group or groups that differ fiom the others use a multiple comparison procedure,
All Pairwise Multiple Cornparison Procedures (Student-Newman-Keuls Method) :
Com parison Diff of Ranks p 9 P<O.OS d vs b 17.000 3 4.907 Yes d vs a 13.000 2 5.307 Yes a vs b 4.000 2 1.633 No
3.2.3) Two way repeateti measures A N 0 VA on two factors - 5 xlQ6 !M acetazolamide and p H ' C O I with burst amplitude as the dependant variable
Note - ACZ (NIYIR) refers to the pre acetazolamide control (N). the 5 .u lo4 M acetazolarnide treatment (Y). and the post acetazolarnide control (R). In addition. the % CO2 values b. d. and a
nJ O refer to the 0% (b). 10% (d). and 2 O /o (a) CO2 exposures for the preparations.
Balanced Design
Dependent Variable: Amp (cm)
Normal@ Test: Faiied ( P = ~ 0 . 0 0 1 )
Equal Variance Test: Failed t P = 10.00 1 )
Source o f Variation Slice Number AC2 (niy/r) .4CZ (n/y/r) x Slice Number O'aCO2 O'oC02 x Slice Number 4CZ (n/y/r) I( ?CO3 Residual Total
The difference in the mean values among the different levels of AC2 (n/yr) is not q a t enough to exclude the possibility that the difference is just due to random sampling variability aiter alIowing for the etTects of differences in 9hC02. There is not a statistically significant difference ( p = 0.494).
The difference in the mean values among the different levels of 9 K 0 2 is not great enough to exclude the possibility that the difference is just due to random sarnpling variability afier allowing for the effects of dittèrences in ACZ in/y/r). There is not a statisticaIly significant difference (p = 0.960).
The effect of different levels of AC2 (n/y/r) does not depend on what level of O6CO2 is present. Therc is not a statistically sigificant interaction between ACZ (n/yir) and ?/oC02. (P = 0.60 1 )
Power Power Power
of perfonned test with alpha = 0.0500: for AC2 (n/y/r) : 0.0502 of perforrned test with alpha = 0.0500: for ?/oc02 : 0.0502 of perforrned test with alpha = 0.0500: for AC2 (n/y/r) .u ?'oc02 : 0.0500
Least square means for AC2 (n/y/r) Group Mean SEM n 1.927 0.288 Y 1.6 10 0.288 r 1.646 0.288
Lest square means for ?'oc02 Group Mean SEM b 1,783 0.330 d 1.678 0.320 a 1.723 0.320
Least square means for ACZ (n/y!r) s "'oc02 Group .Mean SEM n x b 1.966 0.369 n x d 1.978 0.369 n x a 1.838 0.369 y x b 1.756 0.369 y x a 1.658 0.369 r x b 1.625 0.369 r x d 1.639 0.369 r x a 1.674 0.369 y x d 1.418 0.369
The chart below shows the burst frequency. amplitude. and duration values for six transverse medullary slice preparations prior to. duringh. and following the application of 1 mM acetazolamide in DMSO.
Sllce 5 1 Ffe~ [Htl 0.1 14 0.161 1 0.1 17 0.179 0.255 O. 168 0.1 52 f 0.244 0.156
Slice 6
.. - St. Dev. F w) SEM F IH21
Freq M MemFreaM1
Slke i
Slice 2
1
0.067 1 0.056
0.027 1 0.023
0.034
0.105
( S I O u Isecl
Slice 3
lice 4
Sllce 5
I 1
0.062 1 0.068 1 0.093 1 0.059 1 0.049
0.026 1 0.028 1 0.038 1 0.024 1 0.020
0.710
0.675 O u (sec) Du (secl D u [secl
Siice 6
Siice 1
Siice 2
0.085
0.149
0.755
1.118
- - - -- - - - -
Du (=CI Mean Du (mm) SI.Dev.Du.[ml
Note - AC2 (N/Y/R) refen to the pre acetazolarnide control (N). the 5 x lad M acetazolamide treatment (Y). and the post acetazolamide control (R). In addition. the % CO2 values b. d. and a refer to the 0% (b). 10% (d). and 2" 0% (a) CO2 exposures for the preparations.
0.064
0.026
0.715
0.465
- -
1.003
0.880
Am3 (cm1 Am0 [cm)
Siice 3
Stke 4
Sllce 5
Silce 6
Balanced Design
0.037 1 0.012
0.107 10.111
0,052
0.021
0.613
1.550
1.07010.990
0.808
0.881
0.758
0.848
Arnp (cm) Arrp (cm1 Am0 [cm1
Dependent Variable: Burst Frequency
0.728
0.758
0.197 10.385
0.975 1.020
Amp (Cm1 ~ e a n Amp (cm] St.Dev.AmO[cml
Normality Test: Failed ( P = 0.0 13)
0.019
0.158
0.633
1.210
1.153
0.980
0.760
2.270
2.555
1.015
0.610
0.285
0.940
0.600
0.543
0.573
0.240 10.281
3.830
2.195
3.420
1.650
1.115
0.505
0.585
1 .321
0.866
0.01 5
0.104
0.495
1.200
0.768
0.585
0.596
0.915
1.505
0.518
0.518
0.164
2.575
1.730
0,670
1.363
1.062
0.034
0.108
0.410
0.893
0.ô5.3
0.865
0.697
1.650
1.205
0.560
0.293
0.302
2.175
1.595
0.260 1 0.1 30
0 . M 1.417
1.360
0.085
0.169
o. 488
1 . 1 1 5
0.843
1 .O03
0.729
0.895
1 .O05
0.037
0.105
0.W
0.528
0.226
2.375
1.295
0.100
4.1 20 1.851
1.358
0.528
0.968
0.808
0.758
0.688
0.473
0.600
0.808
0.768
0.1 97
2.800 1 2.720
0.985 1 0.920
0.870
1.4bo
0.890
1.855
3.470
1 1.671
1.108
0.493
0.943
0.650
0.603
0.8&
0.243
3.1 15
2.195
0.355
0.850
1.725
0.550
0.530
1.210
0.808
0.355
3.140
1 .dm
l .IO1
0.565
1.427
1.062
0.585
1.278
0.883
0.670
1.207
0.870
Equal Variance Test: Failed ( P = <0.00 1 )
Source of Variation DF SS MS F P Slice Number 5 O. 148 0.0297 AC2 (Y/N) - 7 0.0004 16 0.000208 0.0710 0.932 ACZ (YIN) x Slice Number 1 O 0.0293 0.00293 O'oC02 2 0.032 1 0.0 16 1 19.652 <O.OO 1 Oh CO2 x S lice Nurnber 1 0 0.00542 0.000543 AC2 (Y W) x O 6 CO2 3 0.00095 1 0.000238 1.151 0.362 Residual 20 0.004 13 0.000207 Total 53 0.221 0.00417 The difference in the mean values among the different lcvels of AC2 (YRJ) is not geat enough to exclude the possibility that the difference is just due to random sampling variabilin, atler allowing for the etTects of differences in 96 COL There is not a statistically significant difference (p = 0.932).
The difference in the mean values among the different levels of '6 CO? is greater than would be expected by chance alter allowing for effects of differences in AC2 (YIN). There is a statistically sigificant difference (p = ~0.00 1 ). To isolate which group(s) differ from the others use a multiple cornparison procedure,
The et'fect of different levels of ACZ (YIN) does not depend on what level of 9% CO2 is present. There is not a statistically significant interaction beween ACZ (YM) and O 6 CO?. (P = 0.362)
Power ofpertormed test with alpha = 0.0500: for AC2 (Y.*) : 0.0500 Power of pertôrrned test with alpha = 0.0500: for 9% CO3 : 1 .O00 Power of pertôrmed test with alpha = 0.0500: for ACZ (YN) x O 6 CO2 : 0.0746
Least square means for ACZ ( Y 4) Group Mean SEM II O. 120 0.0257 Y 0.124 0.0257 R O. 127 0.0357
Least square rneans for 9'0 CO2 Group Mean SEM b 0.108 0.0239 d 0.1 58 0.0239 ;a O. IO5 0.0239
Least square means for ACZ (YU) x O b CO2 Group Mean SEM N s b 0.105 0.0263 N x d 0.148 0.0263 'I x a 0.107 0.0263
.4.3.3.2) One way repeated nteasures ANOVA on one focior - p W ? CO2 with bumi frequency as the dependent variable
One Way Repeated Measures Analysis of Variance
Nonnality Test: Passed (P = 0.102)
Equal Variance Test: Passed (P = 0.286)
Group N M issing b 18 O d 18 O a 1 8 O
Group Mean Std Dev SEM b 0.108 0.0583 0.0137 d O. 158 0.0689 0.0162 a O. IO5 0.0544 0.0128
Power of periormed test with alpha = 0.050: 1 .O00
Source of Variation DF SS M S F P Between Subjects 1 7 0. t 78 0.0 105 Between Treatments 2 0.032 1 0.0 16 1 52.005 <O.OO Residual 34 0.0 105 0.000309 Total PI 0.22 1
The differences in the mean values arnong the treatrnent groups are geater than would be expccted by chance: therc is a statisticaliy significant ditrerence ( P = <0.00 1 ). To isolate the group or groups that ditkr from the others use a muit iple comparison procedure.
All Painvise Multiple Comparison Procrdures ( Student-Newman-KeuIs Mcthod) :
Cornparisons for factor: Comparison Di tT of Means p 9 PcO.05 d vs. a 0.0529 3 12.769 Yes d vs. b 0.0505 2 12.192 Yes b v s - a 0.002391 0.577 No
.4.3.3.3) Two way repated measures ANOVA on two factors - 1 m M ocetazolamide and pH/O/oC02 with burst amplitude as the dependent variable
Note - ACZ (N/Y/R) refers to the pre acetazolarnide control (NI. the 5 x 10" M acetazolarnide treatrnent (Y). and the post acetazolamide control (R). In addition. the % COz values b. d. and a refer to the 0% (b). 10% (d). and 2" OYO (a) CO2 exposures for the preparations.
Balanced Design
Dependent Variable: amp (cm)
Normality Test: Failed (P = cO.00 l )
Equal Variance Test: Failed (P = ~0.00 1 )
Source of Variation DF SS MS F P Slice Number 5 30.339 6.068 A C Z ( N N I R ) 2 1.328 0.664 0.345 0.717
ACZ RINIR) x Slice Number 1 O 19.263 1.926 O'oCO3 2 0.0449 0.0224 0.14 0.868 0'oC03.uSliceNurnber IO 1.559 0.156 ACZ (NNIR) .u O 4 CO2 4 0.574 O. 143 1.301 0.303 Residual 30 2.204 0.1 10 Total 3 55.3 l? 1.044
The difference in the mean vaiues among the different levels of AC2 (NN'R) is not great enough to exclude the possibility that the difference is just due to randorn sampling variabilip after allowing for the etrects of differences in 9'0 CO?. There is not a statistically significant difference (p = 0.71 7).
The difference in the mean vaiues among the different levels of CO2 is not great enough to exclude the possibility that the difference is just due to random sampling variability afier allowing for the ettects of differences in AC2 (N/Y/R). There is not a statistically sigificant difference (p = 0.868).
The effect of different lcvels of AC2 INN'R) does not dcpend on what level of O'o CO2 is present. There is not a statistically sipifIcant interaction behveen AC2 (N/Y!R) and O'o CO?. ( P = 0.303)
Power of performed test with alpha = 0.0500: for AC2 (N/Y!R) : 0.0500 Power of performed test with aipha = 0.0500: for Oh CO2 : 0.0500 Power of pertomed test with alpha = 0.0500: for ACZ (N/YfR) x 9'0 CO2 : 0.102
Least square means for ACZ (N:Y /R) Group Mean SEM N 1.367 0.429 Y 1.663 0.429 R 1.304 0.429
Least square means for O'o CO2 Group Mean SEM b I .483 0.344 d 1.4 4 0.344 a 1 .437 0.344
Least square means for ACZ (N:Y>'R) x "6 CO2 Group Mean SEM N ' r b 1.321 0.445 N x d 1.363 0.445 N x a 1.417 0.445
A.3.3.4) Two way repeated measures ANOVA on two factors - I m M acetazolamide and pHP/oC02 witrh burst duration as the dependent vtzriabIe
'lote - AC2 (NIYIR) refers to the pn acetazolamide control (N). the 5 .u 10" M acetazolarnide treatment (Y). and the post acetazolamide controi (R). In addition. the % COz vaiues b. d. and a refer to the 0% (b). 10% (d). and 2" 0% (a) CO2 exPosures for the preparations.
Balanced Design
Dependent Variable: dur (mm)
Normality Test: Failed (P = <0.00 1 )
EquaI Variance Test: Failed (P = <0.00 1 )
Source of Variation DF SS MS F P Slice Number 5 2.451 0.490 AC2 (N/Y/R) 2 0.350 0.175 17.599 <O.OOI AC2 ( N I Y I R ) x Slice Number 1 O 0.0993 0.00993 O/oC02 2 0.0708 0.0354 3.088 0.090 o'oC02xSliceNumber 10 0.1 15 0.01 15 . K Z ( N N I R ) x O 6 CO2 4 0.0355 0.00887 0.6 1 O 0.660 Residual 20 0.291 0.0145 Total 53 3.412 0.0644
The ditkence in the mean values among the ditkrent levels of AC2 (NIY'R) is p a t e r than would be expected by chance after allowing for effects of differences in O h C02. Thcre is a statistically significant difference (p = €0.00 1). To isolate which goup(s) dittèr from the others use a multiple cornparison procedure.
The differcnce in the mean values among the different levels of O h CO2 is not great enough to exclude the possibility that the difference is just due to nndorn sampling variability afier allowing for the etfects of ditlierences in AC2 (NIY'R). There is not a statistically signitlcant difference (p = 0.090).
The rt'fect of dittèrent levels of ACZ W ' Y R ) does not depend on what level of Ob CO2 is present. There is not a statistically sigificant interaction between AC2 ( N N i R ) and "6 CO?. ( P = 0.660)
Power of pertormed test with alpha = 0.0500: for ACZ (NI'Y:'R) : 0.996 Power of pertbrmed test with alpha = 0.0500: for O 6 CO2 : 0.323 Power of pe rhned test with alpha - 0.0500: for ACZ (NNiR) .u O.0 CO3 : 0.0500
Lem square means for AC2 (NIYIR) Group Mean SEM N 0.873 0.0972 Y 0.684 0.0972 R 0.729 0.0972
Least square means for O 6 CO2 Group Mean SEM b 0.793 0.0975 d 0.71 1 0.0975 a 0.782 0.0975
Least square means for AC2 (NIY!R) x 96 CO2 Group Mean SEM N .u b 0.889 0.105 N .u d 0.849 0.105
rl.3.3.5) One Way Repeated Measures ANOVA on One Factor - pH/% CO2 wilh Burst Duration as the Dependent Variable
Nonnality Test: Passed ( P = 0.06 1 )
Equal Variance Test: Passed ( P = 0.557)
Group N ,Missing N 18 O Y 18 O R 18 O
Group Mean Std Dev SEM N 0.873 0.269 0.0633 Y 0.684 0.25 1 0.0592 R 0.729 0.912 0.0500
Power of pertbrmed test with alpha = 0.050: 0.997
Source of Variation DF SS MS F P Benveen Subjects 17 7.636 O. 155 Between Treatments - 3 0.350 O. 175 13.963 '0.001 Residucil 34 0.426 0.0 125 Total 53 3.412
The dit'ferences in the mean values among the treatment groups are p a t e r than would be expected by chance: there is a statistically sigiticant difference (P = ~0.00 I ). To isolate the group or groups that ditEr tiom the others use a multiple cornparison procedure.
All Painvise Multiple Comparison Procedures (Student-Newman-Keuls Methodl :
Comparîsons for factor: Comparison Diff of Means p 9 P 4 . 0 5 ;Vvs.Y 0.188 > 7.146 Yes N v s . R 0.144 Z 5.469 Yes R vs. Y 0.0442 3 1.677 No
A.3.4) Rate of pH EquiIibralion
The chart below shows the pH values during increasrs and decreases in %CO2 of the aCSF bathing solution prior to. dunng. and following the application of I mM acetazolamide in DMSO.
A.4) Appendir 4 - BCECF-AM Quality Control Erperiments
A. 4.1) Uncleaved BCECF-AM Sensitivity t o pH
The chart below shows the raw data for the fluorescence emission of uncleaved BCECF-AM dye
in response to changes in pH.
A.4.2) Cfeaved BCECF4 M Sensitivi@ to pH
PH r
6.70 7 .O0
The chart below shows the raw data for the fluorescence emission of cleaved BCECF-AM dye in
Fluorescence Emission 2730871 2 27723536
response to changes in pH.
F 1 wrexence Fluorescence 1 pH 1 Erniuion Wal 1 1 Mean Standard 1 Standard 1 Devialion
Enor 1 Ernission Trial 2
A.4.3) Ckaved BCECF-AM Fluorescence Decay
The chart below shows the fluorescence ernission of cleaved BCECF-AM dye over a period of 50 minutes
Fluorescence Decay Over Tirne I Minute
1 2
Fluorescence Ernission 231 5441 10 2281 23248
Minute 28 29
Fluorescence Emission 1 781 24620 176385559
A. 4.4) Cleavrd BCECF-A M Dye Concentration Test
The chart shows the tluorescence emission of BCECF-AM dye at varying concentrations