| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Physiology and Biophysics, College of Science and Mathematics, Wright State University School of Medicine, Dayton, Ohio 45435
 |
ABSTRACT |
|---|
 TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
We developed a hyperbaric chamber for intracellular recording in rat brain stem slices during continuous compression and decompression of the tissue bath with the inert gas helium. Air, rather than helium, was also used as the compression medium in some cases to increase tissue nitrogen levels. An important feature is the chamber door, which opens or closes rapidly at 1 atmosphere absolute (ATA) for increased accessibility of the microelectrode. The door also closes and seals smoothly without disrupting the intracellular recording. Hyperbaric oxygen was administered during helium compression using a separate pressure cylinder filled with perfusate equilibrated with 2.3-3.3 ATA oxygen. Measurements of tissue/bath PO2 and pH confirmed that the effects of compression using helium or air could be differentiated from those due to increased PO2. One hundred and thirteen neurons were studied during 375 compression cycles ranging from 1 to 20 ATA (mode 3.0 ATA). We conclude that it is technically feasible to record intracellularly from the same mammalian neuron while changing ambient pressure over a physiologically important range. These techniques will be useful for studying how various hyperbaric environments affect neurophysiological mechanisms.
diving medicine; hyperbaric medicine; inert gas narcosis; oxygen toxicity; reactive oxygen species
| |
INTRODUCTION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
THE MAMMALIAN CENTRAL NERVOUS system (mCNS) has evolved to function normally while inspiring air at a barometric pressure (PB) of approximately one atmosphere absolute (~1 ATA, normobaric pressure).1 Deviations from normal levels of arterial PO2 and PCO2 are sensed by central and peripheral chemoreceptor cells, which evoke powerful cardiorespiratory reflexes for maintaining tissue oxygenation and pH homeostasis (13, 31, 38).2 At PB ~1 ATA, increased arterial PO2 and increased arterial PN2 are believed to have no deleterious effects on the mCNS, whereas an increase in arterial PCO2 above normal levels results in CO2 toxicity of the CNS (7). At hyperbaric pressure (PB >1 ATA), the partial pressure of alveolar and arterial gases increases, often with dramatic consequences on neurological functions (2, 7, 22, 29, 45). Depending on the composition of the inspired gas mixture, level of hyperbaria, and the duration of exposure to hyperbaria, the central effects of a particular gas mixture can be either beneficial, as in the case of hyperbaric O2 (HBO2) therapy (26, 45), or harmful, as in cases of O2 toxicity, inert gas narcosis (i.e., N2 narcosis), and CO2 toxicity (2, 7).
Hyperbaria also exposes the mCNS to increased hydrostatic compression,
or pressure per se, which has an additional effect on neurological
function that occurs independently of increased gas partial pressures.
In the absence of inert gas narcosis and O2 toxicity,
hydrostatic compression of the mCNS, resulting from exposure to
ambient pressure >15 ATA, produces a constellation of neurological
symptoms known collectively as high-pressure neurological syndrome
(HPNS) (2,
22).3
Recently, we reported that hyperbaric pressure of 
4 ATA also alters
neuronal function under in vitro conditions (35). It is
unknown, however, whether neuronal barosensitivity (i.e., sensitivity to hydrostatic pressure) at smaller levels of hydrostatic compression describes a neuropathological process (20) or, rather, a
normal cellular response to increasing ambient pressure [e.g., see
review by Macdonald and Fraser (34)].
Because of its relevance to environmental medicine (2, 7, 29) and hyperbaric medicine (26, 45), much is known regarding the neurophysiological effects of extended exposure to hyperbaric pressure and hyperbaric gases in humans and other mammals. In contrast, comparatively little is known about the effects of hydrostatic pressure and/or hyperbaric gases on membrane properties of mammalian neurons as measured by intracellular recording (ICR). The first ICRs in the mCNS were made in slices of rat hippocampus (41, 42, 49). These investigators, however, were unable to routinely maintain an ICR from the same neuron while changing ambient pressure. Because of this technical obstacle, they could not determine if a neuronal response to hyperbaric pressure was reversible on decompression to 1 ATA (41). Furthermore, an inability to continuously record membrane potential (Vm) from a neuron during repeated cycles of compression and decompression while manipulating the extracellular milieu has precluded defining the cellular mechanisms of barosensitivity and sensitivity to hyperbaric gases based on changes in Vm, voltage-gated conductances, and input resistance (Rin).
ICR from mammalian neurons exposed to hyperbaric conditions has been limited by the technical challenges of initiating an ICR in an in vitro tissue preparation, such as a rat brain slice, which is maintained inside a sealed hyperbaric chamber. Two major obstacles in earlier studies were 1) the inaccessibility of the recording microelectrode once the pressure chamber was sealed, which made it difficult to replace a broken microelectrode, and 2) mechanical disturbances that disrupted ICR when sealing the chamber and changing ambient pressure (41, 42). In contrast, sharp-microelectrode ICRs have been conducted routinely in the invertebrate CNS (iCNS) (4, 6, 11, 23, 48, 49), muscle cells (8, 9, 19), and oocytes (12, 39) while ambient pressure was increased. The success of nonmammalian and nonneuronal experiments was most likely due to the robust nature of larger cells, compared with the smaller neurons in the mCNS (41, 42).
We felt that it was important to record intracellularly from mammalian neurons with routine success and to maintain a stable recording while changing ambient pressure so that the various neuropathological and neurophysiological mechanisms that alter function in the mCNS under hyperbaric conditions could be studied. Achieving this goal in a rat brain stem slice would also establish a mammalian model to compare with the intracellular research that used various in vitro models of the iCNS (4, 6, 8, 9, 11, 23). Accordingly, the purpose of our study was threefold: 1) to develop a hyperbaric chamber to be used for intracellular electrophysiology in mammalian brain slices, with a readily accessible interior, which maintained recording stability even while the door was being sealed and ambient pressure was changed; 2) to determine the technical feasibility of measuring tissue and bath PO2 and pH during compression and decompression; and 3) to confirm by continuous measurement of PO2 and pH that the effects of pressure per se on membrane properties of neurons can be differentiated from those of increased gas partial pressure, in this case PO2 and PN2.
| |
MATERIALS AND METHODS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Brain Slices
Brain slices were prepared from juvenile and adult Sprague-Dawley rats (100-300 g) of both sexes as previously described (13). Brain slices (300 µm thick), cut in the transverse plane, encompassed an area beginning at the level of obex in the caudal medulla oblongata and continued rostrally for ~1,500 µm. Animals were killed by rapid decapitation without anesthesia. Anesthesia was not used because of the depressant actions these agents have on neurons and, furthermore, because of the mutually antagonistic actions exerted by anesthesia and hydrostatic pressure on neural activity (49). ICRs were made primarily in the solitary complex (SC), a cardiorespiratory control area comprising the nucleus tractus solitarius and dorsal motor nucleus of vagus. Neurons in the SC were selected for our initial experiments because cardiorespiratory control mechanisms are altered by increased hydrostatic pressure (43, 44) and hyperbaric gases (40, 46); we hypothesized that neurons in the SC would likewise be affected by these conditions (35).Fresh brain stem tissue slices were incubated at room temperature
(22-25°C), submerged in control artificial cerebral spinal fluid
(aCSF) composed of (in mM) 125 NaCl, 5.0 KCl, 1.3 MgSO4, 26 NaHCO3, 1.24 KH2PO4, 2.4 CaCl2, and 10 glucose (300 mosM). Control aCSF has a pH of
7.48, which produces an intracellular pH in various medullary neurons
ranging from 7.32 to 7.50 after equilibration with 5%
CO2
95% O2 gas mixture (37°C) at 1 ATA4 (38). After
30-60 min of incubation, the first brain slice was transferred to
the brain slice recording chamber (14) and held down
against a fine nylon mesh using a small piece of bridal veil mesh (see
Fig. 3). The brain slice was submerged in ~2.0 mm of aCSF
(36.5-37°C). Tissue slices were typically held in the incubation
chamber for up to 5-6 h and recorded from for up to 8 h,
following rapid decapitation.
Hyperbaric Chamber
A detailed description of the hyperbaric chamber and sample cylinders (Figs. 1 and 2), including materials and dimensions, requirements for pressure testing, and certification by the American Society of Mechanical Engineers, is presented elsewhere.5 The steel chamber has an internal volume of ~72 liters, a maximum working pressure of 67.3 ATA, and a hydrostatic test pressure of 101.7 ATA. The inner dimensions of the chamber are ~33 × 14 in., which provides adequate space and clearance for two electronic microdrives (used for Vm and PO2 electrodes), one manual manipulator (used for the pH electrode), brain slice chamber, and various related equipment items (Fig. 2A). The closure mechanism (Fig. 2B), moveable equipment sled (Fig. 2A), and high-pressure sample cylinders (Fig. 2C) were important features that we adapted from several previous designs (10, 24, 25, 42); however, this was the first time these exact features were combined in a chamber specifically designed for single cell recordings in tissue slices of mCNS (16, 37, 42).
|
|
The chamber is easily closed and sealed (or opened) in ~20 s, at 1 ATA, using a horizontal double-bolt yoke closure. The two halves of the
yoke, which clamps the door to the front lip of the horizontal pressure
vessel, are drawn together using a manually operated chain and sprocket
drive attached to two horizontal yoke bolts (Fig. 1, item a;
Fig. 2B).6 The
equipment sled, constructed from an aluminum optical bench plate,
slides 14 in. out of the pressure chamber, allowing easy access to the
recording microelectrode, brain slice, and other probes (Fig.
2A). An externally mounted zoom stereoscope (Meiji EMZ-5TR),
equipped with a ×0.3 supplementary lens (Meiji MA530), increases the
total working distance of the stereoscope to 23.5 cm for visualizing
the brain slice and probes inside the pressure chamber (Fig. 1,
item d; Fig. 3). Total
magnification ranges from ×2.1 to ×13.5 with ×10 eyepieces and from
×4.2 to ×27 with ×20 eyepieces. The interior of the chamber is
illuminated with an externally mounted, 250-W tungsten halogen Canty
lamp and light pipe (model HYL 250-LS; Amron International Diving
Supply), equipped with flexible, dual-armed fiber optics.
|
Three thermocouples measure temperature in the aCSF immediately beneath
the brain tissue slice, in the outer bath of the brain slice chamber,
and in the atmosphere of the hyperbaric chamber (Fig.
4). Tissue temperature is
servo-controlled at 37 ± 0.4°C using a thermoelectric Peltier
assembly and a custom-built temperature controller. The outer bath of
the brain slice chamber is regulated at 40 ± 0.2°C
(14). Pressure-dependent changes in air temperature inside
the pressure chamber are reduced but not eliminated [see Fig. 6;
however, notice that aCSF temperature (TaCSF) is
unchanged] with a heat exchanger made from 0.25-in. OD copper tubing
that penetrates the wall of the pressure chamber. Water (38°C) is
pumped through the heat exchanger using an Isotemp immersion circulator (model 730, Fisher Scientific) (Fig. 4).
|
Media and Compression Gases
Control perfusion medium. Figure 4 shows a schematic overview of the gas and fluid lines used for compressing the tissue slice and delivering aCSF to the brain slice. aCSF (see Brain Slices, above) was pumped to the brain slice (2-5 ml/min) using one of two high-pressure liquid chromatography (HPLC) pumps: model 100A solvent metering system (Beckman Instruments, Fullerton, CA) and model AA-100-S precision metering pump, equipped with a piston wash kit (Eldex Laboratories, Napa, CA). A high-pressure solenoid valve (General Valve, Fairfield, NJ) allows selection between two different experimental media. The outflow port of the solenoid valve connects to a thermoelectric peltier heater assembly, which is used to make final adjustments to TaCSF (14). The dead space volume from the solenoid valve, where the two medium lines converge, to the brain slice bath is 1.8 ml. Waste aCSF is drawn from the inner bath of the brain slice chamber to the outer bath via a filter paper wick (Fig. 3A). The outer bath of the slice chamber drains by gravity into a waste reservoir, which can be emptied to the outside by opening a ball valve whenever the hyperbaric chamber is pressurized.
Compression gases.
The bath overlying the brain slice is hydrostatically compressed, at a
rate of 1-2 atm/min, by flowing either 100% helium or air (79%
N221% O2) into the pressure chamber
(Fig. 4). Helium has no narcotic effects on the mCNS at ambient
pressures <200 ATA (2, 3, 5).
However, air, due to the increased PN2 at
hyperbaric pressure, has graded narcotic effects on neurons (2, 6).
HBO2 administered through the aCSF. A high-pressure sample cylinder (1-liter volume, Figs. 2C and 4), partially filled with aCSF, was compressed to 2.4-3.4 ATA total pressure using CO2-O2 gas mixtures to produce normocapnic-hyperoxic perfusion medium. A similar device has been used in other studies to elevate the partial pressure of various gases in medium perfusing an in vitro tissue preparation (24, 25). Levels of CO2 and O2 in gas mixtures used to compress aCSF in the sample cylinders were measured using an S-3A/I O2 analyzer and CD-3A CO2 analyzer (AEI Technologies). The sample cylinder was prepared at least 60 min in advance to allow equilibration of the aCSF with the gas mixture. Calculated PO2 of the aCSF ranged from 1,787 to 2,532 Torr, and PCO2 ranged from 35 to 40 Torr. A small pressure gradient (0.06-0.4 atm) from the sample cylinder to the 72-liter chamber was sufficient to drive hyperbaric oxygenated medium, after the HPLC pump was turned off, to the brain slice. A second high-pressure solenoid valve was used to select aCSF from either the HPLC pump or sample cylinder (Fig. 4).
Electrophysiology
Head stage.
ICRs were made using the Axoclamp 2A microelectrode clamp (head stage
gain = 0.1×), which, as previously reported, is pressure insensitive at 65 ATA and easily modified for recording at ambient pressures 
65 ATA (42). Our own tests, which used the
model cell supplied with the head stage, confirmed that the head stage was indeed pressure insensitive up to at least 59 ATA (not shown).
Intracellular microelectrode.
The recording microelectrode was made from borosilicate glass
(TW100F-6, World Precision Instruments) using a one-stage pull with a
Flaming-Brown P-87 microforge. The microelectrode was partially filled
with 3 M potassium acetate and had a tip direct current (DC) resistance
ranging from 90 to 150 M
(13). The microelectrode was
held with a standard patch-clamp micropipette holder containing an
Ag-AgCl wire (E. W. Wright, Guilford, CT). The pressure port of
the micropipette holder was always left open to ambient pressure inside
the hyperbaric chamber. The Axoclamp 2A head stage mounted directly to
the microdrive (Marzhauser-Wetzlar DC3-K micromanipulator with
x-y-z-axes of movement; Stoelting, Wood Dale, IL). As in previous studies (35, 41, 42,
49), this model of electronic microdrive performed well
during compression and decompression. However, the three motors driving
the x-, y-, and z-axes of movement are
of the brush type, which produces electrical sparks during operation.
Therefore, a high-dose O2 gas mixture was not used for
compression of the chamber atmosphere because of the potential fire
hazard (see Safety issues below).
Reference electrode.
An agar salt bridge and Ag-AgCl wire combination electrode was used as
the indifferent electrode for measurements of
Vm; the same reference electrode was also used
for measurements of pH and PO2 (see pH
and PO2 Electrodes, below). The reference
electrode was made with 2% agar solution in 50 mM KCl100 mM
potassium gluconate injected into a nonheparinized microhematocrit
capillary tube (~40 mm long × 1.5 mm OD × 0.2 mm ID). The
silver wire of the patch-clamp pipette holder (E. W. Wright) was
chlorided (1.5 V D-cell, Ag wire as the cathode) in 3 M NaCl solution
for 3-5 min until black and inserted into the agar reference
capillary tube.
Data collection.
Neurons were considered healthy if they had a resting
Vm of at least 40 mV and action potentials
that passed through 0 mV at 1 ATA. Rin was
measured as previously described (13). Current injection
protocols were delivered and their results captured with the use of
pCLAMP software (Axon Instruments). Vm was
filtered at 1 kHz using the Axoclamp 2A.
pH and PO2 Electrodes
Extracellular pH was measured in brain slices or aCSF with a needle electrode, adapted from Baumgartl et al. (1), consisting of a platinum wire (Medwire PT3T), proton-selective glass (Clark Electromedical Instruments, PH100-15), and an Ag-AgCl reference (the same reference as used for PO2 and Vm measurements). The tip had an outer diameter of ~30 µm. The ME-2 head stage of the Axoclamp-2A was used (head stage gain = 0.0001×). Each pH electrode was calibrated before and after the experiment with the use of standard buffers of pH 6.0 and 8.0. The electrodes have a voltage response of58.1 mV/pH
unit (n = 27, r2 = 0.94)
over a pH range of 6-8 units.
Tissue slice or aCSF PO2 was measured in our
first series of experiments with a glass-insulated platinum needle
electrode with a tip outer diameter of ~30 µm connected to a
polarographic amplifier (A-M Systems, model 1900) (18). A
polarization voltage of 0.6 V was used. At this potential, current
output of the electrode is proportional to dissolved O2 in
the medium. Only electrodes that showed a 3.5- to 4.0-fold increase in
current going from 95% air to 95% O2 were used. In
addition, recent experiments used a carbon fiber electrode with a tip
outer diameter of ~10 µm as described by Jiang et al.
(28). These electrodes had a sensitivity of 8.26 pA/Torr
over a PO2 range of 722-2,234 Torr
(n = 15, r2 = 0.98). For
both types of O2 electrodes, a reference electrode was
placed in the aCSF, as described above. O2 electrodes were calibrated in aCSF equilibrated for 2 h with 95% O2
and 5% CO2 or air and 5% CO2.
Analysis
Various signals were stored on magnetic tape (Vetter PCM recorder model 400, Rebersburg, PA) and analyzed off-line using pCLAMP, Axoscope, and Origin software programs. Up to four signals were stored on tape at any one time, including Vm, microelectrode potential in aCSF, integrated firing rate (1- or 10-s bins), tissue PO2, tissue pH, aCSF pH, TaCSF, chamber air temperature, and ambient pressure inside the hyperbaric chamber. Neuronal responses to hyperbaric gases were classified as described elsewhere (35).| |
RESULTS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Mechanical Stability and Accessibility of the Chamber's Interior
ICR during compression and decompression.
A total of 113 neurons in the SC were impaled at 1 ATA and studied
while ambient pressure was changed (resting
Vm = 54 ± 6 mV;
Rin = 124 ± 11 M). Three hundred
and seventy five compression tests ranging from 1.3 to 59 ATA were made
on 113 neurons. Of these, 300 hyperbaric tests were conducted on 78 neurons at ambient pressures from 1 to 4 ATA, and 75 hyperbaric tests
were conducted on 35 neurons at PB > 4 ATA. Most of the
experiments used ambient pressures ranging from 1 to 4 ATA (mode = 3 ATA maximum pressure during compression test). Of the 35 neurons
tested at >4 ATA, 10 neurons were studied continuously over the range
of 1-8 ATA using helium and/or air (36). The greatest
PB from which we were able to completely cycle pressure up
and down, beginning at 1 ATA, without losing the ICR, was 20 ATA
(n = 1 neuron). The greatest PB achieved
during ICR, beginning at 1 ATA, was 59 ATA, but the ICR was lost at
~47 ATA during decompression (n = 1 neuron).
Door opening and sealing: accessibility.
Because the chamber door is typically opened and closed 5-20 times
per experiment to access the microelectrode and tissue slice, it is
extremely important that this process be as easy as possible. At room
pressure, the door could be unsealed/sealed in <20 s by turning the
handle (30 revolutions) to engage the chain and sprocket drive (Fig. 1,
item a). The entire process of mounting the microelectrode
and positioning its tip over the nucleus, sliding the equipment sled
into the chamber, and sealing the pressure chamber required only
2-3 min. During an experiment, the microelectrode was replaced
after its resistance became unacceptable (<90 M) or after
terminating an ICR. Once this occurred, the microelectrode was
retracted from the brain slice, and the chamber door was opened. The
chamber door could be opened, once at 1 ATA, and the microelectrode
removed in ~1 min.
Stability of the closure mechanism.
We also established that the chamber could be sealed during ICR without
disrupting the impalement (Fig. 5). With
the equipment sled positioned inside the chamber, the door was left
open until an ICR was initiated. In these cases, the cell was
hyperpolarized at values from 60 to 90 mV to prevent spontaneous
firing. The pressure vessel was then sealed in ~5 min by slowly
turning the handle of the chain and sprocket drive at a rate of six
revolutions per minute. The hyperpolarizing current was removed, and
control neuronal activity was noted at resting
Vm for several minutes before compression was
commenced.
|
Operating the Hyperbaric Chamber
Cell penetration at 1 ATA followed by pressurization.
Once the hyperbaric chamber was sealed, and if an ICR was not already
underway, the microelectrode was advanced, at an angle of ~30°, by
remote control until its tip made contact with the surface of the brain
slice. It was then advanced in 4- to 6-µm intervals into the SC. The
tip was oscillated for ~2 ms once every 8-24 µm of
advancement. After a cell was impaled and a stable Vm was achieved, several minutes of control data
were collected at 1 ATA. Chamber pressure was then increased to the
desired level at a rate of 1-2 atm/min. Once the cell's response
to hyperbaric pressure and HBO2 was determined, the slice
was decompressed at a rate of 1 atm/min.
Gas bubbles. Bubbles in the aCSF were avoided because they contract and expand during compression and decompression, respectively, possibly causing mechanical instability of the tissue slice and microelectrode. Gas bubbles in the aCSF were minimized by 1) purging the HPLC pumps of gas at the onset of each experiment, 2) warming aCSF to 37°C while aerating with O2-CO2 gas at 1 ATA, and 3) avoiding rapid decompression from hyperbaria. Bubbles were rarely a problem when the 1-liter sample cylinder was used during HBO2 exposures (Fig. 2C).
Safety issues. A pressure relief valve prevents compression of the hyperbaric chamber beyond its maximum working pressure of 67.3 ATA, and two pressure-warning devices prevent opening the door while the chamber is pressurized. The other major safety concern was the high potential for fire when working with a high fractional concentration of O2 in a hyperbaric environment. Consequently, the hyperbaric chamber was pressurized only with helium or air. High-dose O2 (i.e., >21% O2) was never used to pressurize the hyperbaric chamber atmosphere. Instead, HBO2 was delivered to the brain slice by increasing PO2 of the aCSF with the sample cylinders.
Temperature Control
Atmospheric temperature inside the chamber will change during compression and decompression due to adiabatic heating and cooling. Such temperature swings, if transmitted to the brain slice, will introduce error in Vm measurements. If the temperature of the reference electrode changes, the reference potential also changes and is superimposed on the Vm record (21). In addition, many neurons are highly thermosensitive (14, 21), and simultaneous changes in tissue temperature and pressure would make it to difficult to discern which parameter, temperature or hydrostatic pressure, was affecting neuronal activity.Figure 6 shows the results of an
experiment to determine the effects of changing pressure on ambient
temperature (Ta) inside the hyperbaric chamber and on
TaCSF beneath the submerged brain slice. During the first
compression from 1 to 3 ATA in air, Ta inside the pressure
vessel increased by ~6°C; changes in air temperature were transient
because the steel chamber walls and copper heat exchanger act as heat
sinks, helping to return air temperature toward 22-24°C. During
decompression back to 1 ATA, Ta decreased transiently by
~5°C. Despite these changes in Ta, notice that TaCSF remained unchanged during compression and
decompression to and from 3 ATA. The second compression to 3 ATA used
helium, which has a greater thermal conductivity than air.
Consequently, pressure-induced changes in Ta were larger,
increasing transiently by ~14°C. Likewise, Ta decreased
transiently by ~11°C during decompression. A third, rapid
compression to 13.6 ATA in helium produced an even larger transient
deviation in Ta. Regardless, TaCSF remained
essentially constant (36.8 ± 0.2°C) at all pressures tested.
Measurement of air temperature inside the brain slice chamber showed
that it likewise remained constant during changes in PB
(not shown).
|
Pressure Artifacts of the Recording Probes
Membrane potential.
Changes in Vm during compression or
decompression were occasionally confounded by pressure-induced
artifacts of the recording system. These artifacts, however, were
abolished or minimized with the use of certain precautions and are
discussed elsewhere.5 The second trace in Fig. 6 shows
microelectrode voltage (VME) with the tip placed
in aCSF. In this example, VME varied by 3 mV
over time and showed small transient fluctuations during compression and decompression in air or helium. This example was shown because it
represents a pressure-induced DC shift that we sometimes observed; however, most ICRs were free of such voltage artifacts. One possible explanation is that small voltage fluctuations are due to an
unidentified thermosensitive component of the electrophysiology system,
which manifests itself periodically. In these cases, we look for
concomitant changes in Rin, firing rate, and
action potential waveform to corroborate changes in
Vm. In addition, the microelectrode is withdrawn
from the neuron at the end of the experiments and checked with a short
compression test to determine if it is pressure sensitive.
Tissue and bath PO2 and pH. Custom-built O2 and pH electrodes were highly sensitive to PO2 and H ions, respectively, but they were difficult to work with. The main problem with both types of electrodes was maintaining a stable baseline, which tended to drift slowly over time (1-2 h). In addition, newly constructed PO2 electrodes of the platinum type were often pressure sensitive until after several helium compressions, but it was unclear why this occurred. Recent experiments using the carbon fiber PO2 electrodes have shown these to be more stable and less electrically noisy, compared with the platinum PO2 electrodes. Measuring PO2 and pH in the bath, compared with the brain slice, was generally easier. Tips of the polarographic electrode and pH electrode became increasingly coated with tissue debris after several penetrations of the brain slice (the tips were always retracted from the tissue before the equipment sled was moved).
Tissue and Bath PO2 and pH During Hyperbaria
Figure 7A shows that the effects of pressure per se [increased helium pressure (PHe)] can be tested independently of changes in tissue PO2 and pH. Likewise, the effects of HBO2 can be tested independently of changes in ambient pressure and tissue pH. It also shows that VME, PO2, and pH can be measured simultaneously during a compression cycle. Pure helium was flushed into the hyperbaric chamber at 1 ATA, to remove room air, and then compressed to 2.3 ATA. As ambient (helium) pressure increased, tissue PO2 and pH exhibited only small transient fluctuations. When the perfusion medium was changed from control medium (aCSF saturated with 95% O25% CO2 at
PB ~1 ATA room pressure: aCSF
PO2 = 722 Torr and PCO2 = 38 Torr) to HBO2 medium
(aCSF saturated with 98% O22% CO2 at
PB = 2.4 ATA inside the sample cylinder: aCSF
PO2= 1,787 Torr and
PCO2= 36 Torr), tissue slice
PO2 increased from 420 to 1,100 Torr, but pH
was essentially unchanged. A second helium compression to 3 ATA,
likewise, did not significantly alter tissue PO2 and pH. Figure 7B shows another
experiment in which pH was measured in the tissue bath. This shows
again that pH was unchanged during helium compression to 3 ATA.
Hypercapnic acidosis at 1 ATA, however, caused an extracellular
acidosis of ~0.35 pH units (13, 38),
indicating that the pH electrode was indeed sensitive to changes in
H+ concentration.
|
O2 measurements made with the platinum electrode, in the SC, indicated that aCSF saturated with 95% O2 at 1 ATA (calculated aCSF PO2 ~722 Torr) resulted in tissue PO2 to range from 420 to 580 Torr at an ~50- to 100-µm depth in the brain slice (n = 8 slices). Measurements made with the carbon fiber electrode under similar conditions, but at ~150-µm depth in the brain slice, showed tissue PO2 to average 311 ± 21 (SE) Torr (n = 7 slices). During HBO2, with PB ranging from 2.4 to 3.4 ATA, tissue PO2 increased to ~780-1,400 Torr at ~50- to 100-µm depth (platinum electrodes) and 980-2,055 Torr at ~150-µm depth (carbon fiber electrodes).
Electrophysiology: Technical Feasibility and Research Applications
Hyperbaric helium at constant PO2 and pH.
Figure 8A shows an ICR of a
neuron in the SC that was stimulated by helium compression from 1 to 4 ATA. Spontaneous firing rate increased from 1.2 ± 0.2 impulses
per second (mean imp/s ± SD) at 1 ATA to 2.5 ± 0.3 imp/s at 4 ATA helium. Increased firing rate occurred with a
concomitant reduction of Rin from 159 to 133 M (Fig. 8, A and B1) and a small membrane
depolarization of ~3 mV (Fig. 8B2). In addition, the
afterhyperpolarizing potential decreased by ~4 mV during helium
compression (Fig. 8B3). On decompression back to 1 ATA,
firing rate partially recovered (1.9 ± 0.2 imp/s). The effects of
pressure on Vm, Rin, and
the afterhyperpolarization reversed after decompression to 1 ATA.
|
HBO2 vs. hyperbaric helium (pressure per se).
Figure 9A shows another
segment of the ICR made from the same SC neuron presented in Fig. 8. In
this example, a smaller compression to 3 ATA helium caused a smaller
increase in firing rate (1.0-2.1 imp/s) (Fig. 9, A and
B). Although Vm did not change
significantly during compression (Fig. 9, B and
C1), the amplitude of the afterhyperpolarization was
reduced by ~3 mV again (Fig. 9C2, trace 1 vs.
trace 2). Rin also decreased again
from 153 to 142 M (Fig. 9B). At PB = 3 ATA helium, switching from control O2 to hyperoxic aCSF
(HBO2) increased the firing rate (Fig. 9A;
2.1-4.1 imp/s). Vm did not decrease appreciably (Fig. 9C1), but the amplitude of the
afterhyperpolarization increased (Fig. 9C2, trace
2 vs. trace 3), whereas its duration decreased, causing
Vm to reach threshold sooner for action
potential generation (not shown). Rin increased
(Fig. 9C3; trace 2 vs. trace 3) from
142 to 148 M, concomitant with increased firing rate. A second
exposure to HBO2 caused a similar excitatory response (Fig.
9A). After return to control aCSF at 3 ATA and then
decompression to 1 ATA, firing rate, Rin, and
Vm returned toward control levels.
|
|
Hyperbaric air.
Flushing the hyperbaric chamber with air (21% O279%
N2) at 1 ATA, followed by compression with air, enables the
study of the effects of increased PN2 on
neuronal activity (6, 36). Figure
11A shows a record of tissue
PO2 during compression with air to 3 ATA. The
PN2 in the chamber atmosphere was calculated to
be 1,801 Torr at 3 ATA. Conversely, in the aCSF, aerated with 95%
O25% CO2 at 1 ATA,
PN2 would be 
600 Torr. Thus N2
diffused into the brain slice at a rate determined by the flow rate of aCSF over the slice and the solubility of N2 in aCSF and
tissue (3, 33). In contrast, 21%
O2 in the chamber atmosphere (PO2 of ~480 Torr) did not diffuse into the tissue slice because the PO2 of the aCSF was greater
(PO2 = 720 Torr).
|
| |
DISCUSSION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
General Findings and Implications
We designed and built a hyperbaric chamber for making continuous ICRs from neurons in rat brain stem tissue slices (see report by J. B. Dean, D. K. Mulkey, and J. D. Arehart5). Using this new hyperbaric chamber design, we demonstrated the technical feasibility, for the first time, of obtaining a continuous ICR from a mammalian neuron during compression from 1 ATA up to at least 4 ATA, followed by decompression back to 1 ATA (see below). We also demonstrated the technical feasibility of continuously measuring bath and tissue slice PO2 and pH over the same pressure range. Using this chamber design, we could easily manipulate an intracellular microelectrode, or quickly replace a broken microelectrode, inside a sealed or opened hyperbaric chamber. In contrast, construction of PO2 and pH electrodes, and their use inside a hyperbaric chamber, was comparatively more difficult but possible.There are several benefits to using a pressure chamber that allows continuous measurement of Vm, PO2, and/or pH while compressing a mammalian brain slice. First, neuronal activity (e.g., membrane conductance, firing rate, voltage-sensitive membrane properties, synaptic potentials) can be correlated with changing pressure, as well as steady-state pressure, under controlled conditions of PO2, pH, and temperature. Likewise, neuronal activity can be correlated with real-time changes in PO2 or pH at hyperbaric pressure. It was previously thought that measuring PO2 and pH (or PCO2) of perfused solution, directly under pressure or following decompression, would be too difficult (43, 44). However, by doing so, we confirmed previous assertions that PO2 and PCO2/H+ concentration in the aCSF set at 1 ATA outside the hyperbaric chamber remained constant during helium compression and decompression as long as no additional O2 and CO2 were admitted to the chamber atmosphere (16, 25, 42, 43).7
Continuous ICR during compression and decompression also enables
determination of the reversibility of neuronal responses to pressure
per se and hyperbaric gases. In addition, maintaining an ICR during
repeated cycles of compression permits study of neuronal activity under
different hyperbaric conditions. For example, by using different gases
to pressurize the chamber (helium vs. air) and to aerate perfusion
media (control O2 at 3 ATA helium vs. ~3 ATA
O2), we were able to identify different types of
barosensitivity and O2 chemosensitivity in the same neuron
(Figs. 8, 9, and 11). These data were presented to demonstrate the
technical feasibility of making such measurements from the same neuron
while changing gas partial pressures and ambient pressure from 1 to 4
ATA. They were also presented to illustrate potential research
applications pertaining to CNS barosensitivity, CNS O2
toxicity, and inert gas narcosis. However, at this time, the few
examples presented in this study should not be interpreted as
representative examples of how all neurons in the SC respond to
pressure per se, HBO2, and hyperbaric air. The effects of
these conditions on SC neurons require further study; other reports
(35, 36) are also available that address
these issues.
Hyperbaric Electrophysiology
A major obstacle to making an ICR in an in vitro tissue preparation that is maintained inside a pressure chamber is the inaccessibility of the microelectrode once the chamber is sealed (42). Robust cell preparations, such as muscle cells (8, 9, 19), oocytes (12, 39), and invertebrate axons and neurons (4, 6, 11, 23, 48, 49), are better suited for ICR under hyperbaric conditions because of the relative ease with which they can be impaled and recorded from with the use of lower resistance microelectrodes. Because lower resistance microelectrodes break less readily, they do not have to be changed as often, and the chamber interior does not have to be accessed as frequently. In contrast, making ICRs from the comparatively smaller neurons in the mCNS requires higher resistance microelectrodes, which plug and break more readily and must be replaced frequently.ICR in the mCNS, compared with in the iCNS, also requires greater
mechanical stability of the tissue preparation and microelectrode (42). Previous attempts to maintain a stable ICR of a
mammalian neuron in a brain slice, while changing ambient pressure,
have been largely unsuccessful. Southan and Wann (41,
42) were able to successfully record
Vm at steady-state pressure, but they were
unable to consistently maintain the ICR while changing pressure. Most
ICRs were lost during compression from 1.5 ATA, their control PB, to 
3 ATA. Likewise, if the cell was impaled at
hyperbaric pressure, the recording was usually lost shortly after
decompression was commenced. They concluded that maintaining an ICR of
a mammalian neuron, during either compression or decompression, was
"impracticable," and that "future experiments must work towards
obtaining continuous intracellular recordings from individual neurons
during compression from atmospheric pressure up to the high
pressures" (41). Consequently, they studied different
groups of CA1 neurons exposed to different pressures rather than the
effects of different pressures on the same neuron (41,
42).8
It is not surprising, therefore, that most hyperbaric studies in isolated tissue preparations of the mCNS have utilized more robust electrophysiological techniques, such as extracellular recordings (16, 17, 49), whole nerve recordings (43, 44), and loose macropatch-clamp recordings (15). These types of recordings are easier to initiate and maintain inside a sealed pressure chamber compared with an ICR. However, in some studies (15, 43, 44), even the maintenance of a stable extracellular recording during decompression has been difficult, presumably because of bubble formation and mechanical disturbances of the tissue preparation.
Important Design Features for ICR
Hyperbaric chambers used for mammalian brain slice electrophysiology, although generally suitable for extracellular recording, lacked the correct combination of features that optimized mechanical stability for ICR and improved accessibility of the microelectrode (15, 16, 37, 42). Our hyperbaric chamber incorporates features and equipment used in several previous horizontal, large-capacity chamber designs (10, 24, 25, 42). The primary feature that made our experiments feasible, however, was the horizontal double-bolt yoke closure mechanism. This style of door is well known, and it has often been employed in chambers used for whole animal research.9 However, this is the first report in which the horizontal double-bolt yoke closure was used in a hyperbaric chamber designed for mammalian in vitro electrophysiology (16, 37, 42). Such a door improved accessibility to the electrophysiology equipment inside the chamber and enabled us to close the door manually and seal the hyperbaric chamber after a neuron was impaled, without disrupting the ICR. This finding alone greatly increased the feasibility of our experiments and suggests that other electrophysiology techniques, which require frequent changes of the recording micropipette (e.g., blind whole cell recording), could also be used in a pressure chamber with the same door. Thus, although the various features of our chamber are not new to hyperbaric research, we believe that they are "new" for a chamber designed specifically for electrophysiology work using mammalian brain slices.Test Pressures: Total Pressure (PB = PHe) and Gas Partial Pressures (PO2, PN2)
Hyperbaric helium.
Beginning at 1 ATA, the maximum pressure we compressed to, and
decompressed from, was 20 ATA; however, we rarely explored this range
of pressure (35, 36). We chose to work
primarily at lesser levels of PHe (4 ATA) because our
current interests focus on how HBO2 and reactive
O2 species affect brain stem neurons (35), and
CNS O2 toxicity is typically manifested at PB
~3 ATA when a high fractional concentration of O2 is
breathed (7). Moreover, this range of PHe
corresponds to ambient pressures that humans encounter during routine
diving (22), compressed air work (29), and
hyperbaric medical treatments (26, 45).
Because helium is inert at <200 ATA, neural responses to hyperbaric
helium are believed to be due to pressure per se, rather than the
effects of increased PHe (2, 3,
5, 22).
3 ATA,
in mammalian brain slices (41, 42), we
propose that our hyperbaric chamber design, combined with the rat brain
stem slice and conventional ICR technique, will be a useful in vitro model of the mCNS for studying cellular mechanisms of neuronal barosensitivity, CNS O2 toxicity, and inert gas narcosis.
HBO2. Control conditions in the vast majority of brain slice studies, including this study, are conducted at normobaric pressure using perfusate aerated with 95% O2, which makes the PO2 of control aCSF already hyperoxic (aCSF PO2 ~720 Torr at PB ~1 ATA) compared with neural tissue in vivo (27). The high level of control PO2 is used to drive O2 into the center of the avascular brain slice and avert tissue hypoxia and anoxia (28). The only way to significantly increase tissue PO2, to study the effects of an additional increase in PO2, is to increase PB. Previous studies have done this by compressing the hyperbaric chamber with pure O2 (4, 8-10, 30). However, this was not an option with our hyperbaric chamber design because of the electrical equipment used and the potential danger for a chamber fire in a high-dose O2 environment.5 Thus we adapted the high-pressure sample cylinder reported elsewhere (24, 25). On the basis of our measurements of tissue slice and bath PO2, our findings show that the sample cylinder is an effective tool for exposing a mammalian brain slice to hyperoxia (27, 47). Jamieson and Van den Brenk (27) reported that in the intact rat cerebral tissue PO2 increased, on average, from ~34 Torr when air is breathed at 1 ATA, to ~452 Torr when 100% O2 is breathed at 3 ATA, and to ~917 Torr when 100% O2 is breathed at 5 ATA. Tissue PO2 increases even more when 5% CO2 in O2 was breathed at hyperbaric pressure (due to increased cerebral blood flow caused by vasodilation, and thus increased O2 delivery), e.g., tissue PO2 of ~1,540 Torr at 5 ATA ambient pressure. Thus our PO2 measurements made in the tissue slice during HBO2, which ranged from ~780 to ~2,055 Torr, include this range of values. It is possible that our measurements of tissue PO2 made during exposure to HBO2 were underestimated due to a small DC potential that may have interfered with the polarizing voltage of the electrode (32).
Hyperbaric nitrogen. Inert gas narcosis is manifested over a range of ambient pressures (2) and will likely need to be studied under in vitro conditions using a broad range of air pressure. In contrast to HBO2, hyperbaric air can be safely used to compress the chamber atmosphere to increase PN2. The utility of this method is that it provides a rapid means of administering N2 to the brain slice, without having to exchange perfusate, which takes several minutes longer, so that the effects of hyperbaric air vs. hyperbaric helium can be studied on the same neuron. Although tissue PN2 was not determined in our study, we believe that using air as the compression medium increased the PN2 in aCSF and the brain slice during compression (3, 6, 33). Our preliminary experiments suggest that equivalent pressures of hyperbaric helium and hyperbaric air (3-8 ATA) have different effects on neuronal activity in the SC, which are likely due to increased PN2 (36).
Conclusion
We have designed, built, and tested a hyperbaric chamber for making continuous ICRs in SC neurons of rat medullary tissue slices. We incorporated features and equipment items from several previous chambers used for both invertebrate and mammalian electrophysiology, focusing on those aspects that we felt would maximize mechanical stability of the tissue preparation and microelectrode while sealing the chamber, as well as improve accessibility of the microelectrode. Our findings show that recording intracellularly from a large sample of neurons in the mCNS is no longer an impracticable experiment, as once believed, but can now be done with essentially the same degree of difficulty as conducting ICRs using a conventional rat brain slice setup or an isolated tissue preparation of the iCNS. Mechanical stability of the ICR, during compression and decompression, was presumably due to the high quality of the microelectrode and, hence, neuronal impalement. Replacing a plugged or broken microelectrode as needed increased our chances for establishing a high-quality ICR that remained healthy and presumably stable during any mechanical disturbances that may have occurred during compression and decompression. Continuous measurement of Vm (and related cellular properties), PO2, and/or pH, under changing hyperbaric conditions, should prove useful for studying neurophysiological mechanisms in the mCNS of adaptation to hyperbaric environments and neuropathological mechanisms that impair neuronal signaling under various hyperbaric conditions.| |
ACKNOWLEDGEMENTS |
|---|
We acknowledge the assistance and advice of the late Earl Hoffman of Tube Turns Technologies, Inc., Louisville, KY. Likewise, we thank Eugene Reiter, also of Tube Turns Technologies, Inc., for assistance in designing, building, and certifying the pressure vessel. We also thank James Arehart for outstanding services during the design and construction phases of the pressure vessel; Wayne Massey and Scott Hawkins for technical assistance; Steve Hayden, Mike Hall, and Bob Miller for designing and building the chamber's electrical system and temperature controller; George Gohring and Jerry Morris for insightful discussions and permitting us to tour the Naval Medical Research Institute in Bethesda, MD; Dr. Richard Henderson III for insightful discussions on CNS O2 toxicity and tissue PO2; and Scott Kissell for making photographs of the chamber.
| |
FOOTNOTES |
|---|
Funding for the development and testing of the pressure chamber was provided in part by Wright State University (WSU) and included National Institutes of Health (NIH) Biomedical Research Support Grant NSS 2S07-RR-05794-16, Research Initiation Grant Program, Research Challenge Program, School of Medicine, College of Science and Mathematics, Dr. P. K. Lauf and Department of Physiology and Biophysics. D. K. Mulkey was supported by the WSU Biomedical Sciences PhD Program and NIH. Materials and animals were supported in part by NIH Grants R29 HL-46308 and R01 HL-56683, WSU Alpha Grant Program (Kettering Foundation), Medical Multiplex, Inc., and the Kettering Hyperbaric and Wound Care Center at Wright-Patterson Air Force Base, Dayton, OH.
Address for reprint requests and other correspondence: J. B. Dean, Dept. of Physiology and Biophysics, Rm. 160 Biological Sciences Bldg., 3640 Colonel Glenn Hwy., Wright State Univ., Dayton, OH 45435 (E-mail: jay.dean{at}wright.edu).
1
One atmosphere absolute (ATA) is equivalent to 760 Torr (sea level). Other commonly used pressure equivalents for 1 ATA
include 1.01 bar, 10.1 m seawater and 14.7 lb/in.2 (psi).
The SI units for pressure are the pascal (Pa = N/m2,
where 1 Pa = 1.02 × 10
5 ATA), kilopascal
(101.3 kPa = 1 ATA), or megapascal (1 MPa = 10.13 ATA);
however, our data are reported using ATA.
4 Room pressure or PB in Dayton, Ohio, was typically 0.98-1.02 ATA, which we have rounded off to 1 ATA for this study.
5 For the report "Details on building a hyperbaric chamber for intracellular recording in brain tissue slices," by Jay B. Dean, Daniel K. Mulkey, and James D. Arehart, order NAPS Document 05566 from NAPS c/o Microfiche Publications, 248 Hempstead Turnpike, West Hempstead, NY 11552. Supplemental material can also be viewed at http://jap.physiology.org/cgi/content/full/89/2/807.
6 The horizontal double-bolt yoke closure is available in a range of sizes and pressure ratings from Tube Turns Technologies, Inc. (2900 W. Broadway, Box 32160, Louisville, KY 40232).
7 Unpublished observations (D. K. Mulkey, R. A. Henderson, and J. B. Dean) in our laboratory showed that blowing warmed, humidified gas mixture of 95% O2-5% CO2 over the surface of the slice bath from a small n diameter gas line, while compressing the chamber with pure helium, increased PO2 and decreased pH in both the aCSF and tissue. Thus previous studies in which O2-CO2 gas mixtures were accessible to hippocampal slices during helium compression also likely involved some degree of hyperoxia and hypercapnia (30, 42).
8 Similar difficulties were reported by another laboratory using the same style of hyperbaric chamber. This anecdotal report on ICR at hyperbaric pressure was published in the Axon Instruments newsletter, Axobits (Roberts MG and McPhie GI, Intracellular recording at high barometric pressures, Axobits 10: 11, 1992). It is worthwhi
FR; impulses per second (imp/s)], and
Vm. The thick dark band in the
Vm record is many intracellular action
potentials compressed together in time. Downward voltage deflections
are electronic potentials (B1) resulting from repetitive
hyperpolarizing current injection (
