| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
INVITED REVIEW
1Department of Physiology and Biophysics, Environmental and Hyperbaric Cell Biology Facility, and 2Department of Community Health, Wright State University School of Medicine, College of Science and Mathematics, Dayton, Ohio 45435
 |
ABSTRACT |
|---|
 TOP ABSTRACT GlossaryOVERVIEW OF BARO-RELATED...CENTRAL EFFECTS OF PRESSURE...NARCOTIC AND TOXIC PROPERTIES...RESEARCH APPLICATION: HYPEROXIA...PERSPECTIVEAPPENDIXES A-D: HYPERBARIC...DISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
3 ATA O2) as a model for studying the cellular
mechanisms of oxidative stress in the mammalian central nervous system. anesthesia; carbon dioxide toxicity; free radicals; high-pressure nervous syndrome; membrane potential; nitrogen narcosis; oxidative stress; oxygen toxicity; polarographic oxygen electrode
1 atmosphere absolute
(ATA).1 Nonetheless,
we exploit a continuum of barometric pressure (PB) ranging from the
near vacuum of outer space, which we survive during Space Shuttle
extravehicular activity by wearing a 4.3
lb./in.2 absolute
(psia) pressure suit (30,300 ft. pressure equivalent, 0.29 ATA)
(120,
220), down to the summit of
Mt. Everest at 29,029 ft., where PB is 0.31 ATA
(221), to ocean depths as
great as 2,300 ft. of sea water (fsw), where PB increases to
70 ATA (18). These
situations are the pressure extremes of our inhabitable environment, which
only relatively few highly trained individuals have ever occupied.
Military personnel, medical personnel, and other humans, however, frequently encounter levels of hyperbaric pressure (i.e., >1 ATA) of lesser degrees in their normal work environments. Examples of moderate hyperbaric environments (e.g., <5 ATA) include the following: patients and medical attendants undergoing hyperbaric O2 therapy (HBOT) (38, 203); diving with an underwater breathing apparatus for recreational, professional (oil and salvage companies), and combat purposes (89); simulated dry and wet dives for hyperbaric research and dive training (217); and working in the compressed atmosphere of a subterranean environment (117). Abnormal work environments resulting from catastrophic accidents while at sea (32, 165) or underground (155) also can result in prolonged breathing of hyperbaric gases. For example, submariners breathe a hyperbaric atmosphere while awaiting rescue inside a disabled submarine (DISSUB) (165). Although submarine accidents rarely occur, the sinking of the Russian submarine Kursk in the Barents Sea in August 2000, in 356 fsw (11.8 ATA), with loss of all 117 crew members, is a somber reminder of the importance of preparing for DISSUB emergencies. Furthermore, the need to prepare for similar DISSUB scenarios on land, in which men are trapped under increased pressure, was demonstrated recently when nine Pennsylvania coal miners were stranded for 77 h inside a water-filled mine. Exposure to ambient pressures up to 40 fsw [1.21 atmospheres gauge pressure (ATG), or 2.21 ATA] necessitated hyperbaric medical support by medical officers and divers from the US Navy's diving and salvage community (155). Presently, the US Navy maintains agreements with >20 countries for assistance with DISSUB events (68).
Hyperbaric environments present many physiological challenges, affecting especially the lungs, hollow viscera, and nervous system. With respect to the mammalian central nervous system (CNS; mCNS), breathing air, pure O2, or mixtures of O2 and inert gases (N2, H2, and He), and air contaminated with CO2, at hyperbaric pressure for extended periods, can impair the normal functioning of neural networks (28, 89). Common neurological problems associated with hyperbaric environments included O2 toxicity, which is thought to occur through increased oxidative stress, as well as N2 narcosis (inert-gas narcosis), CO2 toxicity, and high-pressure nervous syndrome (HPNS) (18, 43, 89, 203).2
In neurophysiology, one of the commonly used experimental approaches to identify how changes in gas tension affect neural function is to conduct an electrophysiological study of single neurons in an isolated tissue preparation of the rodent CNS, such as brain slices or the neonatal brain stem-spinal cord, while manipulating the PO2, PCO2, and/or N2 partial pressure (PN2) of the perfusate at room pressure (i.e., normobaric pressure) (54, 65, 98). However, making a single-cell recording in an isolated tissue preparation of the rodent CNS, which is maintained inside a hyperbaric chamber, while increasing ambient pressure, has proven to be technically challenging due to the physical barrier imposed by the sealed pressure chamber and mechanical disruption of the recording microelectrode and neuron during tissue compression and decompression (191, 192). Consequently, relatively little is known about how hyperbaric gases and increased hydrostatic pressure affect the intracellular properties of neurons in the mCNS. Recent improvements in hyperbaric chamber designs, however, have now made intracellular recordings of mammalian neurons, maintained under hyperbaric conditions, technically feasible (58, 59, 150, 152, 153).
In 1979 and 1980, Wann and colleagues reviewed hyperbaric electrophysiological methods (215) and summarized the effects of large hydrostatic pressures (>100 ATA) on cellular excitability (213). Research at that time was based primarily on studies of large robust neurons, axons, and muscle cells in the invertebrate CNS (iCNS). Given the technical advances made in this field during the past two decades, which have increased the productivity of single-cell studies in the mCNS (58, 72-74, 150, 152-154), we have reassessed the methods used in hyperbaric electrophysiology along with their potential research applications. Unlike previous reviews of hyperbaric electrophysiology, which focused exclusively on the effects of hyperbaria exceeding 50 or 100 ATA (89, 114, 130, 213), we have emphasized the importance of studying neuronal barosensitivity and chemosensitivity to hyperbaric gases at PB <5 ATA. The various methods used to differentiate the effects on neurons of gas partial pressures vs. pressure per se are summarized (APPENDIXES A-D). Moreover, barodependent disorders of the mCNS are summarized, presenting their clinical signs and symptoms and the pressure ranges (total pressure and gas partial pressures) over which they occur, as an aid to designing physiologically relevant in vitro studies (see OVERVIEW OF BARO-RELATED DISORDERS OF THE MCNS). Specific examples are presented to demonstrate how hyperbaric pressure per se (see CENTRAL EFFECTS OF PRESSURE PER SE) and increased gas partial pressures (see NARCOTIC AND TOXIC PROPERTIES OF GASES) can alter neuronal excitability. In particular, we emphasize the use of hyperoxia under pressure [i.e., hyperbaric O2 (HBO2)] not only as an in vitro model for studying CNS O2 toxicity, but also as a new model for studying the cellular mechanisms by which acute oxidative stress [i.e., reactive O2 species (ROS)] alters neuronal activity. Accordingly, we have critically evaluated the control levels and experimental levels of tissue PO2 (PtiO2) that are used in the rat brain slice model (3, 151, 169), comparing each to levels of PtiO2 that occur in the intact mCNS when breathing normobaric air, normobaric hyperoxic gas mixtures, and HBO2 (see RESEARCH APPLICATION: HYPEROXIA AND O2 TOXICITY).
|
Glossary |
|---|
|
TOPABSTRACTGlossaryOVERVIEW OF BARO-RELATED...CENTRAL EFFECTS OF PRESSURE...NARCOTIC AND TOXIC PROPERTIES...RESEARCH APPLICATION: HYPEROXIA...PERSPECTIVEAPPENDIXES A-D: HYPERBARIC...DISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
E
G
H
)
S
V
V*
|
OVERVIEW OF BARO-RELATED DISORDERS OF THE MCNS |
|---|
|
TOPABSTRACTGlossaryOVERVIEW OF BARO-RELATED...CENTRAL EFFECTS OF PRESSURE...NARCOTIC AND TOXIC PROPERTIES...RESEARCH APPLICATION: HYPEROXIA...PERSPECTIVEAPPENDIXES A-D: HYPERBARIC...DISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
N2 Narcosis
Figure 1 shows the broad
range of ambient pressure over which the human body can function. Extended
stay at pressures exceeding 4 ATA (3 ATG, 99 fsw depth) is only
possible by reducing the fractional concentration of N2 in the
inspired air to avoid the narcotic effects of N2 (N2
narcosis). For example, nitrox is O2-enriched air, which is made by
lowering the fractional concentration of inspired N2
(FIN2). It is breathed to avoid N2
narcosis while the diver is at depth. The reduction in partial pressure of
inspired N2 (PIN2) when breathing
nitrox also reduces the incidence of decompression sickness and the time
needed for decompression during ascent. Typically, nitrox is used at depths
>100 fsw. Symptoms of N2 narcosis are significantly manifested
in most individuals when breathing air at PB of 4 ATA and then
increase insidiously but rapidly with further increasing depth and are best
characterized as a graded rapture, reduction of higher mental processes, and
impaired neuromuscular coordination, usually resembling the symptoms of
alcohol intoxication, or similarly, the early stages of anesthesia and
hypoxia. With increasing depth and PIN2, the
symptoms worsen and eventually lead to unconsciousness at depths where
PB exceeds 10 ATA
(18). To date, very little
electrophysiological research has been conducted on the cellular mechanisms
involved in N2 narcosis
(34,
35,
154).
|
CO2 Toxicity
Retention of CO2 under hyperbaric conditions, due to either increased fractional concentration of inspired CO2 (FICO2) and/or decreased alveolar ventilation, or breathing a CO2-contaminated gas mixture, impairs neurological function. It is well established that end-tidal PCO2 increases in a diver due to increased metabolic CO2 production caused by physical exertion of swimming, working, and breathing underwater, and due to reduced ventilatory efficiency (alveolar hypoventilation) caused by the added dead space of a breathing apparatus and the increased airway resistance caused by increased density of gases at increased ambient pressure (79, 147, 194). Problems associated with CO2 retention are exacerbated in experienced divers, compared with nondivers and amateur (recreational) divers, because experienced divers typically exhibit higher levels of resting end-tidal PCO2 caused by reduced ventilatory chemosensitivity to inspired CO2 at normobaric pressure (116, 217). For example, mean end-tidal PCO2 in nondivers and diving trainees was 40 Torr, whereas experienced Navy divers had a significantly higher mean end-tidal PCO2 of 46 Torr. During hyperbaria and exercise, end-tidal PCO2 in experienced divers often increased to >50-70 Torr, and, during severe exercise, end-tidal PCO2 reached >90 Torr (116, 194, 217).
CO2 retention at hyperbaric pressure can produce two types of
neurological problems. First, retention of even modest levels of
CO2 at hyperbaric pressure presumably increases nitric oxide
radical (NO·) and results in cerebral vasodilation
(101), which increases
delivery of perfusion-limited gases to neural tissue
(108,
122). Consequently, the
severity of symptoms of CNS O2 toxicity (and N2
narcosis) are enhanced during respiratory acidosis and/or begin at lower
levels of hyperbaria (5,
42,
95,
134). Second, tissue
PCO2, once reaching a sufficiently high level, either
directly due to its narcotic properties, or indirectly by the ensuing
intracellular acidification, induces CO2 toxicity without warning.
Normally, arterial PCO2 ranges from 35 to 45 Torr
(arterial pH 7.35-7.45), and small increases in arterial
PCO2 of only a few Torr stimulate both central and
peripheral CO2/H+ chemoreceptors, activating powerful
cardiorespiratory reflexes to remove CO2 from the body for
homeostasis of tissue pH (86,
156,
157,
183). In rats, going from air
to 6% CO2 in air, which stimulates breathing, causes pH of the
cerebrospinal fluid to decrease from 7.396 to 7.294 and intracellular pH
(pHi) of neural tissue to decrease from 7.044 to 6.982. Breathing a
higher level of CO2 (11% in air) decreases pH of the cerebrospinal
fluid even further to 7.190 and pHi of neural tissue to 6.910
(187). Under in vitro
conditions (brain stem slice), moderate levels of hypercapnic acidosis
decrease pHi by 0.2 pH units (control pHi is
7.24) (56,
78,
176,
177) and increase neuronal
excitability in central CO2-chemoreceptor areas of the mammalian
brain stem (54,
78,
98). This stimulation of
putative central CO2/H+-chemoreceptor neurons is
believed to underlie the reflex stimulation of ventilation, which is
interpreted as the physiological response of the respiratory control network
to hypercapnic acidosis (156,
157,
183).
As inspired CO2 increases further to 10-15% CO2, and end-tidal PCO2 exceeds 50-70 Torr, mental ability becomes increasingly impaired and is characterized by confusion, irrational behavior, drowsiness, dizziness, and impaired short-term memory (146, 217). Recovery from CO2 toxicity is often associated with severe headache (146). In some cases, inspiring 30% CO2 produces seizures (225). In rats, breathing 75-100% CO2 rapidly produces ataxia followed by loss of righting reflex and pedal reflexes, and, eventually, anesthesia (45). Divers lose consciousness when end-tidal PCO2 exceeds 90 Torr (194, 217), and prolonged exposure to >70% CO2 (balance O2) quickly leads to death in animal models due to depression of respiratory centers (53).
Despite the widespread use of extreme hypercapnia as an anesthetic and its
potential danger as an environmental toxicant, very little
electrophysiological research has been done on the cellular mechanism of
CO2 toxicity. The toxic effects of CO2, however, are
believed to be due, in part, to the large acidification that occurs when
alveolar PCO2 increases at high levels of inspired
PCO2 (PICO2). Because
CO2 is extremely soluble in lipids and water, it rapidly diffuses
into the cell membrane and through the cytoplasm, where it is hydrated to form
carbonic acid, which rapidly dissociates to form protons and bicarbonate ions.
The effects of extreme levels of hypercapnia (i.e., supercapnia, Ref.
211) on pHi in the
intact mCNS remains unknown. In the rat brain stem slice, however, extreme
hypercapnia, which also impairs neuronal excitability, decreases
pHi, on average, from 7.24 (5% CO2) to 6.71 (33%
CO2) to 6.67 (50% CO2) and to 6.49 (100% CO2
at PB 1 ATA) in locus coeruleus neurons and results in
abnormal electrical activity
(56).
O2 Toxicity
Pure O2 is breathed at >1 ATA by using the LAR V Draeger underwater breathing apparatus on combat diving operations by Special Operations personnel (US Navy SEALs and Marine Force Reconnaissance units) to prevent N2 narcosis at depth, to reduce the length of decompression stops during ascent, and to decrease the risk of decompression sickness during and after ascent (36, 37). However, breathing pure O2 or nitrox (>21% O2) also increases the risk of CNS O2 toxicity at greater depths (Fig. 1). Similarly, Fig. 1 shows that, on land, HBOT uses intermittent exposure to pure O2 at >1-3 ATA to increase arterial PO2, and thus O2 delivery to tissues, to treat a variety of clinical disorders due to diving and nondiving causes.3 Typically, several 20- to 30-min exposures to HBO2 are interspersed with 5- to 10-min air breaks to avert symptoms of CNS and pulmonary O2 toxicity (38, 203). CNS O2 toxicity manifests itself ultimately as grand mal convulsions (i.e., the "Paul Bert effect"), although other autonomic, motor, and cardiorespiratory signs and symptoms also may occur, such as bradycardia, hyperventilation, dyspnea, and altered cardiorespiratory neural reflexes (43, 51, 189, 209). Interestingly, Mulkey et al. (152) have reported that neurons in the solitary complex, an important cardiorespiratory control center in the dorsal medulla oblongata, are highly sensitive to HBO2 and this sensitivity to oxidative stress may contribute to the cardiorespiratory perturbations that occur in CNS O2 toxicity.
The inter- and intraindividual variations in susceptibility to O2 seizures make it difficult to predict who is vulnerable to O2 toxicity and when O2 seizures will occur (43). O2 seizures per se resulting from acute exposure to hyperbaric hyperoxia are not believed to be harmful. If, however, the subject is not removed immediately from the hyperoxic environment, then continued exposure to HBO2 can result, first, in permanent neurological damage and paralysis (i.e., the "John Bean effect") and eventually death with prolonged exposure (12). Presently, CNS O2 toxicity is the limiting factor in protocols employed for closed-circuit diving operations by combat divers (37) and for HBOT (38).
The pressure threshold for onset of seizures can be lowered to <3 ATA O2, however, by coexisting conditions, such as immersion (36), exercise and increased metabolic rate (5, 43, 116), and hypoventilation and respiratory acidosis (5, 42, 95, 223). For example, moderate CO2 retention (respiratory acidosis) lowers the threshold for O2 toxicity, increases the severity of seizures, and decreases survival time in human divers and animals (5, 15, 42, 134). One mechanism for this increased sensitivity to hyperbaric hyperoxia is due to CO2-induced cerebral vasodilation, which increases neural PtiO2 (101). During exposure to HBO2, cerebral blood flow initially is reduced by hyperoxia due to interruption of NO· release from S-nitrosohemoglobin (193) and increased production of superoxide, which reacts with NO· to effectively decrease biologically active NO·, thereby causing vasoconstriction (61-63). CO2 retention, however, presumably increases NO· production (101), causing cerebral vasodilation, which antagonizes O2-induced cerebral vasoconstriction and increases blood flow through neural tissues (13, 122). This, in turn, increases delivery of O2 to neural tissue (108, 122), thereby increasing PtiO2 at any given level of inspired PO2 (PIO2) (see below, Fig. 4: #2, open squares vs. #1, open circles), resulting in increased production of ROS (60, 61, 63, 70, 107, 205). Tissue PCO2 increases during HBO2 by one or more of the following mechanisms: 1) CO2-carrying capacity of venous hemoglobin is decreased because venous hemoglobin is fully saturated with O2, which leads to increased tissue and venous PCO2 (i.e., Haldane effect); 2) alveolar hypoventilation and CO2 retention (see CO2 Toxicity above); and 3) breathing CO2-contaminated gases. In addition, and independently of the increased cerebral blood flow, it is possible that there is increased production of ROS in the presence of H+ and CO2 (174, 188, 218) that renders neurons more sensitive to hyperoxia (152). In contrast to moderate hypercapnia, which enhances CNS O2 toxicity, severe hypercapnia that produces anesthesia depresses development of CNS O2 toxicity in newborn rats by an unknown mechanism (15, 211). Obviously, the interactions among CO2, pHi, and ROS during hyperbaric hyperoxia require additional study. The sensitivity of single neurons to HBO2 will be considered in more detail below (see RESEARCH APPLICATION: HYPEROXIA AND O2 TOXICITY) and in the companion paper (152).
|
He, H2, and O2 Mixtures: HPNS
Figure 1 also includes
heliox, which is a He-O2 gas mixture that is breathed at depths
>150-200 fsw; specifically, 79% N2 in air is replaced with He to
avert N2 narcosis, and the fractional concentration of inspired
O2 (FIO2) is lowered to prevent
CNS O2 toxicity
(18). Breathing heliox also
greatly ameliorates the increasing airway resistance at those depths due to
greater gas density. However, diving at depths greater than 15 ATA can
also result in HPNS, which is due to the effects of pressure per se and not to
increased He partial pressure (PHe), PO2, or
PCO2
(18). Signs and symptoms of
HPNS include muscular tremors, EEG changes, loss of coordination, nausea,
respiratory difficulties, memory deficit, and seizures (in animals), which
typically begin at 15-16 ATA or higher
(18,
89). It is likely that
seizures would also occur in humans if they were subjected to high enough
pressures while breathing heliox; however, this is neither practical nor
ethical for obvious reasons. At even greater depths (>15 ATA), a small
amount of N2 is added back to the breathing gas mixture to produce
trimix (O2-He-N2;
Fig. 1), which increases the
depth and pressure a diver can reach before onset of symptoms of HPNS
(18,
29). Similarly, H2
is substituted for N2 to produce a gas mixture of
O2-H2-He, referred to as hydreliox
(Fig. 1), which is also used
for deep dives to delay onset of signs and symptoms of HPNS
(1,
18,
29). H2, compared
with N2, is even less dense and easier than He to breathe at
pressure, yet it also affords protection from HPNS. It is well known that
anesthetic gases (e.g., hyperbaric N2 or H2) and
hydrostatic pressure have antagonistic actions on neurological function, such
that, when combined, each reduces the deleterious effects of the other
(29,
113,
114). The basis for this
interaction between anesthetics and pressure is unclear and is one of the
areas of hyperbaric electrophysiology research requiring additional study
(27,
113). The sensitivity of
single neurons to pressure per se will be considered in more detail in the
next section and in the companion paper
(153).
|
CENTRAL EFFECTS OF PRESSURE PER SE |
|---|
|
TOPABSTRACTGlossaryOVERVIEW OF BARO-RELATED...CENTRAL EFFECTS OF PRESSURE...NARCOTIC AND TOXIC PROPERTIES...RESEARCH APPLICATION: HYPEROXIA...PERSPECTIVEAPPENDIXES A-D: HYPERBARIC...DISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
We live in a sea of gas, and the weight of the Earth's atmosphere exerts a physical pressure at sea level equal to 1 ATA, to which our bodies are physiologically adapted. As we descend beneath the ocean surface, pressure continues to increase an additional 1 atm each additional 33 fsw (Fig. 1). Thus a diver at 132 fsw is exposed to 4 atm of sea water plus 1 atm of air, or 5 ATA of total pressure. Similarly, a person inside a pressure chamber (145, 203, 217), caisson, or subterranean tunnel (117) experiences an equivalent depth of sea water when the density of gas inside the sealed pressure chamber is increased.
In a hyperbaric environment, soft tissues of the body behave as a fluid and
rapidly transmit any pressure-applied force against the surface of the body to
the adjacent fluid compartments, thereby removing any pressure differential
across structures. This results in hydrostatic compression of the cerebral
spinal fluid, cerebral circulation, and extracellular and intracellular fluid
compartments of the mCNS (89,
109). The best known example
of neuronal barosensitivity is HPNS, which was already discussed above
(18,
89). The threshold pressure
for HPNS is variable and affected by the rate of compression
(18). Details regarding the
cellular mechanism (s) underlying HPNS are unclear
(114), but it appears that
both synaptic (74,
76,
85,
228,
229) and intrinsic membrane
properties are involved (72,
73,
191,
192,
216). Several excellent
reviews on HPNS exist elsewhere
(18,
100,
104,
114). When studying in vitro
neuronal activity as a model of HPNS, it is important to remember the range of
ambient pressures over which symptoms of HPNS occur (15 to 70 ATA,
Fig. 1). For example,
intracellular and extracellular recordings of neurons and axons in
invertebrates and terrestrial mammals have shown that hydrostatic compression,
ranging from 100 to 200 ATA and higher, alters neural activity
(72-74,
76,
114,
130,
191,
192,
197,
198,
216). As a test for studying
barosensitivity, particularly in the mCNS, the use of extremely large levels
of pressure is of questionable importance, because terrestrial mammals are
never exposed to these levels of hydrostatic pressure. As cautioned by Halsey
(89): "... much of the
work with excitable cells has been carried out over the maximum tolerated
pressure range, which in some cases extend up to 1,000 ATA. This is clearly
appropriate in terms of the studies per se, but it is difficult to relate such
work to the intact mammalian CNS, which would not be expected to function
above 100 ATA..." Thus, although the use of test pressures 
100 ATA
will likely increase the magnitude of the cell's pressure response, if
present, it is also important to report what happens over the pressure range
for which the symptoms of HPNS are known to occur for the species being
used.
There is evidence that lower levels of hyperbaria also affect excitability
of neurons and synapses. Excitable cells in invertebrates are barosensitive to
ambient pressures ranging from 4 to 10 ATA by using either true hydrostatic
compression or He compression
(34,
47,
48,
192). Similarly, neurons in
the rat brain stem are barosensitive to even smaller pressures of only 2.4-4
ATA He (58,
150,
153), and brief hydraulic
pressures ranging from 1.17 to 4.0 ATA can alter neuronal excitability in
dorsal root ganglion neurons
(90,
138,
181). For example,
Fig. 2A shows the
integrated firing rate response of a neuron in the solitary complex in the
caudal-dorsal medulla oblongata in a 300-µm-thick submerged slice before,
during, and after compression from 1 to 3 ATA He (PB). Notice that
input resistance (Rin) decreased during He compression
(inset, superimposed traces) concomitant with the increased firing
rate. Typically, membrane potential (Vm) depolarized by
3 mV (not shown) (58,
150,
153). This indicates that 3
ATA pressure caused an increase in membrane conductance (see
Fig. 2 legend for further
explanation). Mulkey et al.
(153) have shown that the
increased firing rate response to hyperbaria is retained during chemical
synaptic blockade, which indicates that it is an intrinsic membrane property
of certain neurons in the solitary complex. Similarly, McCarter et al.
(138) reported that a brief,
graded hydraulic pressure pulse, ranging from 0.17 to 3 ATG, depolarized
Vm and increased inward current in a graded fashion in
dorsal root ganglion neurons.
|
Figure 2B shows a similar response in another neuron in the solitary complex that was stimulated and displayed increased membrane conductance (decreased Rin) during compression to a much higher pressure, 20 ATA He. To our knowledge, this is the first example of an intracellular recording of a mammalian neuron that was maintained continuously during cyclical compression and decompression to this high level of pressure (compare Refs. 191, 192, 216). The significance of this experiment is that it demonstrates that the electrical activity of individual neurons in the mCNS can be studied over the lower range of ambient pressures at which symptoms of HPNS are first detected (Fig. 1).
The above studies indicate that neurons are sensitive to much lower levels of hyperbaric pressure than previously believed (50, 114, 130). The importance of barosensitivity to ambient pressures <5 ATA remains to be determined; however, it may, like temperature, act as another environmental stimulus that determines how the organism responds and adapts to its environment (114, 150, 153). To speculate further, neuronal barosensitivity to relatively low levels of hyperbaric pressure may be the early stage of a pressure continuum that is eventually exhibited as HPNS at higher pressures. What is interesting, however, is that not all neurons in the mCNS are equally sensitive to moderate levels of hyperbaria (82, 153). For example, initial reports indicate that hippocampal neurons do not respond to 3 ATA He or hydraulic pressure (82, 90). Similarly, not all neurons in the solitary complex are sensitive to 2-4 ATA He (153). This suggests that neuronal barosensitivity to low levels of hyperbaria is not a generalized phenomenon, but represents a specific physiological response of certain populations of neurons in the mCNS.
Intracranial Hypertension and Increased Intraocular Pressure
Another source of neuronal barosensitivity, which occurs on a much smaller
pressure scale and independently of changes in PB, is compression
of neural tissue that occurs whenever the mass of tissue inside the rigid,
closed intracranial compartment is increased by tissue edema and bleeding
subsequent to head injury or by increased production or decreased removal of
cerebral spinal fluid (4,
33,
140). Typically, normal
intracranial pressure ranges from 0 to 10 or 15 mmHg in humans
(139) and from 8 to 10 mmHg
in conscious rats (111).
Intracranial hypertension is considered as a sustained elevation in
intracranial pressure exceeding 15 or 20 mmHg and going as high as 100
mmHg (139,
140). Historically,
neurological symptoms associated with traumatic brain injury have been
attributed to cerebral ischemia and the resulting tissue hypoxia and acidosis
as intracranial pressure approaches cerebral arterial blood pressure. However,
it is likely that increased tissue pressure also contributes to neurological
dysfunction in these cases in a manner that remains to be determined. In
addition, the initial traumatic brain injury itself, which is a
fluid-percussive injury, produces a brief, transient period of hydrostatic
compression ranging from 1.4 to 2.1 ATG that results in protracted dysfunction
of electrical signaling and cellular metabolism
(64,
181). For the sake of
completeness, retinal cells, which are considered as part of the mCNS, are
exposed normally to intraocular pressures ranging from 14 to 16 mmHg,
depending on the person's age, race, and gender
(123,
173). Intraocular
hypertension and glaucoma are associated with intraocular pressures 21
mmHg, which produce long-term destruction of nerve cells and their function
(123). However, the effects
of increased intraocular pressure on the excitability and electrical signaling
of retinal cells are unknown.
Thermodynamic Reactions vs. Mechanical Forces
The preceding discussion indicates that neuronal barosensitivity occurs
over a very large range of pressure
(89,
90,
114,
130,
138,
150,
153,
192). It has been suggested
that, depending on the level of hyperbaric pressure, different cellular
mechanisms may be responsible for mediating cellular barosensitivity
(41,
94,
129,
131,
136). At extremely high
levels of compression (e.g., >100 ATA), hydrostatic pressure is a
thermodynamic intensity parameter, similar to temperature, which determines
various thermodynamic equilibria, and thus could affect various cellular
processes. It can be analyzed theoretically, based on the way in which
pressure (P) and the molar volume change (V) combine to change the free
energy (G) of a reaction
(129-131);
i.e.
 | (1) |
 | (2) |
H is the enthalpy change and is equal to
(E + P · V), E is the change in
internal energy, T is the temperature in degrees Kelvin, and S
is the entropy change. At constant T, the equilibrium constant (K),
V, and P are related by the following expression
 | (3) |
V, pressure changes can have a large effect on the equilibrium of that
reaction.
Pressure can also alter the rate of a chemical reaction by changing the
activation volume for the rate-limiting step (V*)
(50). At constant temperature,
the reaction rate constant (k), activation volume, and ambient
pressure are related by the following expression
 | (4) |
V*, pressure changes can
have a large effect on the rate of that reaction. Stated another way,
Macdonald and Fraser (129,
131) have proposed that, even
for reactions with relatively small changes in molar volume and activation
volume, very large hydrostatic pressures can significantly alter the
equilibrium and rate of the reaction, respectively. For this reason, as
indicated in Fig. 1, neural
tissue has been exposed to extremely high pressures, on the order of hundreds
of atmospheres (100 to >600 ATA), as a means of altering cellular functions
through changes in protein structure
(94) and enzymatic activity
(103). In addition, extremely
high pressures can alter membrane fluidity
(128) and cause volume
changes in cell membranes by bulk compression of lipid bilayers, thereby
decreasing the intermolecular distances between acyl chains. Extremely high
pressures can also change the hydration of lipid bilayers and proteins and the
crystalline-to-liquid-crystalline phase transitions in lipid bilayers
(127,
128).
Conversely, at lower levels of hydrostatic pressure, Macdonald and Fraser
(129,
131) have proposed that
thermodynamic and kinetic effects described by Eqs. 1-4 would be too
small to be of physiological significance, based on conventional
solute-solvent interactions, because the V and V*
terms would have to be astoundingly large for any significant change to occur
in G or k. Traditional dogma maintains that
hydro-static pressures <15 ATA have no significant effects on neuronal
function (18,
192,
197). However, as proposed in
Fig. 1 ("Cellular
Barosensitivity") and illustrated in
Fig. 2 and elsewhere
(90,
138,
150,
153,
192), neuronal
barosensitivity, as defined by changes in firing rate, Vm,
and Rin, occurs at 2.5-4 ATA and possibly as low as 100
mmHg (1.13 ATA or 0.13 ATG)
(2). Because relatively few
studies have attempted to differentiate the effects of hydrostatic pressure
from gas partial pressure in the mCNS, especially at PB <5 ATA
(89), we would argue that it
is premature to conclude that small changes in physical pressure do not alter
neuronal function in the mCNS.
In fact, Macdonald and Fraser
(131) have proposed that many
nonneuronal cells respond to micropressures, which they define as 20 kPa
(0.2 ATG or 1.2 ATA), by a mechanical process that they hypothesize to be
localized shear and strain forces resulting from the differential compression
of various adjoining cellular components, such as lipid bilayers,
membrane-bound proteins, cytoskeletal proteins, and extracellular matrix.
However, they go on to say that "... at present, there are no
distinctive candidates [cellular structures] and a practical, specific working
hypothesis is lacking"
(131). If true, however,
these same mechanical forces could be a significant factor at
small-to-moderate hydrostatic pressures (e.g., 2-5 ATA) in neurons, which, in
turn, would alter protein configurations (e.g., ion channels, postsynaptic
receptors, etc.) and thus neuronal excitability. It remains to be determined,
however, at what range of hydrostatic pressures are shear and strain involved,
if at all, in altering the function of cells and tissues, and "... at
what pressure does the putative `micro-pressure' effect saturate and orthodox
high pressure thermodynamics take over"
(131).
|
NARCOTIC AND TOXIC PROPERTIES OF GASES |
|---|
|
TOPABSTRACTGlossaryOVERVIEW OF BARO-RELATED...CENTRAL EFFECTS OF PRESSURE...NARCOTIC AND TOXIC PROPERTIES...RESEARCH APPLICATION: HYPEROXIA...PERSPECTIVEAPPENDIXES A-D: HYPERBARIC...DISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
3 ATA
breathing 100% O2), N2 narcosis (4 ATA breathing
air; i.e., 79% N2), and He narcosis [usually 100 ATA or higher
as determined in various nonmammalian tissue models
(200)].
Figure 1 shows the threshold
PB at which the symptoms or neural responses typically occur when
the requisite gas mixture is breathed, but these values serve as guidelines
only. The threshold ambient pressure at which each neurological problem
develops is not absolute and can vary considerably in either direction for an
individual and between individuals, due to preexisting physiological
conditions that affect cerebral blood flow, metabolic rate, body temperature,
and acid-base status (18,
43). We can state, in general, that the narcotic potency of a gas is directly dependent on its 1) partial pressure in the inspired gas mixture, blood, and neural tissues, which increases with increasing PB, and 2) its lipid solubility in cell membranes, which determines the level to which a gas accumulates in the cell membrane and cytoplasm at any partial pressure (18, 43). Figure 3 shows the solubility of CO2, O2, and two inert gases (He, N2) in a monolayer of egg phospholipid as a function of increasing gas pressure, as reported by Bennett et al. (19). All gases, excluding He, cause a linear increase, at differing rates (CO2 >>> O2 > N2), in monolayer film pressure as gas pressure increases (increased film pressure corresponds to increased membrane volume). Comparing Figs. 1 and 3, we can see that CO2, which has the highest solubility in cell membranes in this example (Fig. 3), likewise has significant effects on the mCNS at the lowest ambient pressure (Fig. 1) (43). Conversely, O2 and N2 have lipid solubilities, and, therefore, narcotic effects, which are intermediate to CO2 and He and are manifest only at hyperbaric pressure. He is essentially insoluble in lipid bilayers and has no noticeable narcotic effects over the range of ambient pressure that most mammals encounter (28). This is why He is used as a compression medium for many in vitro electrophysiology studies (Fig. 2) (e.g., Refs. 58, 75, 191; also see in APPENDIX A, Compression media: HBHe vs. hydrostatic compression).
|
The narcotic effects of a gas were first thought to be related to their ability to expand the hydrophobic region of the cell membrane (141). This presumption, known as the "critical volume hypothesis of anesthesia (and narcosis)," has been revised. It is now thought that the central effects of a gas are mediated by the direct and indirect actions it has on membrane-bound proteins and cytosolic proteins, which perturb ionic conductances and neurotransmission (80). Gases such as O2 and CO2 also have additional effects on neurons because they react biochemically in the extracellular space, the membrane, and intracellular space to form reactive chemical species, which are important metabolic signals in numerous processes and which, at high concentrations, are toxic to cells. For example, O2 forms a variety of highly reactive products (e.g., superoxide, hydroxyl radicals, NO·, hydrogen peroxide, and peroxynitrite), which can either modulate or disrupt normal cell function (60, 61, 63, 67, 71, 107, 205). Examples will be shown below of how HBO2 stimulates neuronal firing rate in neurons of the solitary complex and CA1 hippocampus. Recent work by our laboratory has shown that this excitatory effect of acute exposure to hyperoxia is most likely mediated by an increased production of ROS (152). Moreover, as already stated, CO2 forms H ions, which likewise either modulate or disrupt normal cell function, depending on its concentration (56, 137). Molecular O2 and CO2 will also presumably have narcotic effects on the mCNS (95), as predicted by data shown in Fig. 3 (19). The narcotic effects of O2 and CO2, however, will be difficult to discriminate from the effects of their highly reactive secondary reaction products.
|
RESEARCH APPLICATION: HYPEROXIA AND O2 TOXICITY |
|---|
|
TOPABSTRACTGlossaryOVERVIEW OF BARO-RELATED...CENTRAL EFFECTS OF PRESSURE...NARCOTIC AND TOXIC PROPERTIES...RESEARCH APPLICATION: HYPEROXIA...PERSPECTIVEAPPENDIXES A-D: HYPERBARIC...DISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
Hyperoxia as a Model of Oxidative Stress
Oxidative conditions, produced by a variety of mechanisms (e.g., an increase in exogenous or endogenous oxidizing agent or a decrease in endogenous antioxidant), have wide-ranging effects on the mCNS. Oxidation-reduction (redox) reactions are involved in normal signal transduction mechanisms (67, 71), such as O2 sensing (121, 125, 126, 172) and modulation of neuronal electrical activity (44, 152, 170). However, at high levels, oxidative stress has been implicated as a causative agent, at least in part, in various neurological disorders and diseases (67, 175).
O2 toxicity of the mCNS is one of the best known examples of how
acute exposure to an oxidative environment can disrupt neurological function
(9,
43). Normally, when animals
and humans breathe normobaric air, neural PtiO2 is
surprisingly low, ranging from 1-3 Torr up to 30-34 Torr, depending on
the area of the mCNS (71,
135,
151). Thus, for the purpose
of this review, we will define hyperoxia as any level of inspired
PO2 (in vivo) or perfusate PO2 (in
vitro) that results in a neural PtiO2 >34 Torr.
Depending on the level of PtiO2 in the brain and the
duration of exposure, the neurological response to hyperoxia is quite
variable. However, in each instance, the effects of hyperoxia on the mCNS are
attributed to the oxidative effects of ROS
(60-63,
70,
107,
205). For example, the mCNS
response to hyperoxia can range from a moderate, but reversible change in
neural activity (22,
51,
142,
150,
152), possibly with
therapeutic benefits for improved neurological function
(33,
38,
158,
203), to violent and
reversible seizures at higher levels of PtiO2
(5,
60,
63), to irreversible motor
deficits (and even death) at the highest dosages of hyperoxia
(9,
12). The use of graded levels
of hyperoxia at room pressure (normobaric hyperoxia) and increased ambient
pressure (hyperbaric hyperoxia, or HBO2), therefore, is a useful
model for studying the wide-ranging effects of oxidative stress on
neurological function, which complements more traditional in vitro models of
oxidative stress used at normobaric pressure (e.g., Refs.
169,
170). The utility of
employing hyperoxia as an oxidative stimulus is that the natural substrate,
molecular O2, is supplied to the cell, which, in turn, reacts in
one or more biochemical pathways to produce various free radicals
(9,
60,
61,
63,
67,
70,
71,
169,
170). Thus, by using graded
levels of hyperoxia, one can dissect out specific biochemical steps in redox
modulation of neuronal activity and the steps that lead to onset of oxidative
damage in brain cells.
The hyperoxic brain slice model, however, has rarely been used to study the effects of oxidative stress on neuronal activity. This is because most in vitro tissue preparations of the mCNS, including the brain slice, brain stem spinal cord, and many cell cultures use a control level of O2 that is already hyperoxic (i.e., 95% O2), which leaves little room for increasing PtiO2 further at normobaric pressure. Consequently, the only means of investigating the effects of further hyperoxia is to do so at hyperbaric pressure by using HBO2 (58, 118, 150-152).
Neural Tissue PO2 in the Intact mCNS During Normoxia and Hyperoxia
The likelihood of onset of O2-induced seizures (see above,
O2 Toxicity) is dependent on the level
of hyperoxia. The latency to seizure becomes shorter with increasing
FIO2 and/or increasing PB. In
rats, seizures occur after breathing 3 ATA O2 for 3-5 h, 5 ATA
O2 for 5-90 min, and 6 ATA O2 for 20 min
(189). Consecutive exposures
to 3-5 ATA O2 for 1 h/day produce paralysis in rats and frequently
death, after four to seven exposures
(7). Blood-gas measurements in
humans show that arterial PO2 increases from 90
Torr while normobaric air is breathed to a maximum of 1,900-2,000 Torr while
100% O2 at 3.5 ATA is breathed
(122). Polarographic
measurements of PtiO2 in animals breathing normobaric
air indicate that O2 levels in the mCNS range from <10 to 34
Torr (108,
151), but that they increase
tremendously when the animal breathes normobaric hyperoxia
(39) and HBO2
(108,
151,
210). For example, as shown
in Fig. 4, cortical
PtiO2 increases on average (±SE) from 34 ±
4 Torr in a rat breathing normobaric air (defined by the intersection of
points a and b, open symbols) to 452 ± 68 Torr when
breathing 100% O2 at 3 ATA, and to 917 ± 123 Torr while
breathing 5 ATA O2 (#1, open circles)
(108). Cortical
PtiO2 increases even further, for each level of
PIO2, when CO2 is added to the
inspired gas because of CO2-induced vasodilation of cerebral
arteries and, therefore, increased cerebral blood flow. This effect of
CO2 counteracts the vasoconstriction caused by hyperoxia, as
discussed above. For example, also shown in
Fig. 4 (#2, open squares),
cortical PtiO2 increases to 791 ± 51 Torr and
1,540 ± 94 Torr when 95% O2 + 5% CO2 at
PB of 3 and 5 ATA, respectively, is breathed.
Neural PtiO2 during normobaric normoxia, normobaric hyperoxia, and HBO2 is not homogeneous, but varies regionally within the intact mCNS. When normobaric air is breathed, regional PtiO2 is variable [see Table 1 in Mulkey et al. (151) and Erecinska and Silver (71)] with gray matter averaging 13 Torr, whereas white matter is lower and cerebral spinal fluid is higher (39, 52). During HBO2, regional differences in PtiO2, similarly, are attributed to differences in neuronal activity, metabolic rate, and cerebral blood flow (71, 162, 204, 206-208). Changes in regional cerebral blood flow during HBO2 are complex and depend on the pressure and duration of exposure (i.e., dose of hyperoxia). Breathing 3-4 ATA O2 initially decreases the effective NO· concentration (NO· is a potent cerebral vasodilator) and decreases blood flow by 26-43% in deep brain nuclei (61, 62). The mechanism for this vasoconstrictor effect of hyperoxia was discussed above (O2 Toxicity). Cerebral vasoconstriction provides protection to the mCNS when HBO2 is breathed, at least initially, by blunting the magnitude of the increase in PtiO2 and, therefore, ROS (61, 62). Longer exposures and/or larger levels of HBO2 (5-6 ATA O2), conversely, increase NO· production and augment regional blood flow (63). The mechanism for increased NO· production during prolonged hyperoxia is uncertain. It has been proposed that it results from increased NO· synthesis in the presence of excess molecular O2, increased nitric oxide synthase (NOS) activity, and increased brain L-arginine levels (63). Regardless of the mechanism, the escape of cerebral blood flow from O2-induced vasoconstriction, which results in a large increase in regional PtiO2 (I. Demchenko and C.A. Piantadosi, personal communication), is believed to be an important factor in the pathogenesis of CNS O2 toxicity, because it always precedes onset of increased EEG spiking activity (63).
Other investigators report, however, that the increase in PtiO2 remains sustained throughout the duration of HBO2 (108, 162) or exhibits a secondary decrease in neural PtiO2 that remains elevated above normoxic control level (i.e., transient increase in PtiO2 that partially recovers during hyperoxia) (9, 17, 99, 162, 210). This secondary drop in PtiO2 could merely be an artifact caused by O2 poisoning of the polarographic electrode; however, it has also been attributed to changes in body temperature during compression (17), O2-induced vasconstriction of cerebral blood vessels, increased metabolic rate, and development of diffusion impairment caused by pulmonary O2 toxicity when present (9, 206-208, 210).
Cellular Mechanism: O2-induced Free Radicals
O2 toxicity is believed to occur when the body's antioxidant defenses are overwhelmed by increased production of free radicals during the high levels of O2 reported above. Included in this list are superoxide, hydrogen peroxide, hydroxyl radicals, and peroxynitrite at high levels of PtiO2 (60-63, 70, 160, 171, 205). Excessive production and accumulation of ROS during HBO2 is thought to alter cellular components, and thus various ionic conductances that regulate cell excitability (58, 150, 152). Similarly, ROS are reported to target certain neurotransmitter and neuromodulator systems, and thus to alter chemical synaptic transmission (25, 40, 227). CNS O2 toxicity is thought to involve, or be influenced by, glutamate, GABA, and NO· signaling pathways. MK-801 antagonism of NMDA glutamate receptors (40) and inhibition of NOS have been shown to alter HBO2-induced seizure activity (25, 60, 70, 227), and the actions of O2 have been shown to decrease inhibitory (GABAergic) synaptic conductance (48). Gap junctional conductance through electrical synapses is also reduced by increased production of ROS (143, 212), but the effects of HBO2 on electrical coupling in mammalian neurons have not been investigated (56, 98). This will be important to determine, however, because recent research has shown that CO2/H+-chemosensitive neurons, which are coupled via gap junctions (56, 57, 98, 190), are highly sensitive to HBO2 and other chemical oxidants (152).
Electrophysiological and Neuroanatomic Responses to HBO2
The effects of PtiO2 and free radicals on the mCNS have been studied by using a variety of electrophysiological and neuroanatomic techniques. Cortical EEG activity in animals increases before the onset of seizure (14, 60, 63, 189), typically preceded by increased PtiO2 (210) (I. Demchenko and C. A. Piantadosi, personal communication) and increased ROS production (160, 171). Important sites in the mCNS involved in the pathogenesis of O2 seizures have been identified by using c-fos expression (6), increased 2-deoxyglucose uptake (204, 206-208), and changes in neuronal morphology indicative of lesions and tissue necrosis after extended exposures to hyperoxia (8-10). In the latter case, lesions and tissue necrosis occurred after repeated exposure to HBO2 on consecutive days, which ultimately produced symptoms associated with the John Bean effect, i.e., paralysis. According to these criteria, many areas in the mCNS are affected by HBO2 and, therefore, are involved in the onset of O2-induced seizures and paralysis [Table I in Balentine and Gutsche (10)]. Areas that were consistently affected by HBO2 include the globus pallidus, substantia nigra, superior olivary nucleus, ventral cochlear nucleus, and spinal cord gray matter (8, 10, 162). Several of these regions also endure the highest level of PtiO2 during HBO2 (162, 210). However, because prolonged exposures to HBO2 were used to identify these O2-sensitive regions, which resulted in permanent neurological deficits, these experiments do not rule out other regions of the mCNS as being involved in O2 sensitivity during acute exposure to HBO2. The advent of more sensitive assays of neuronal function, such as single-cell electrophysiology [see below, Figs. 6 and 7, and in the companion paper (152)], have shown that neuronal electrical activity in other regions of the mCNS (brain stem, hippocampus) is affected by acute exposure to HBO2, and, moreover, that not all neurons in these areas are equally sensitive to HBO2 and O2 free radicals (58, 149, 152).
|
|
|
Overall, electrophysiological data at the single-cell level are sparse concerning how HBO2 affects neurons. Refinements in hyperbaric chamber designs, however, have improved the success rate of intracellular experiments in rat brain slices (APPENDIXES A-D). For example, Fig. 5A shows the standard protocol used in our laboratory for testing neuronal sensitivity to pressure per se vs. increased PtiO2 (HBO2). By carefully adjusting the relative levels of O2 and CO2 in the high-pressure sample cylinders (APPENDIX B), a range of hyperbaric hyperoxia can be tested (58, 152). With the use of this protocol, acute exposure to HBO2 typically increases firing rate in some neurons (26, 150, 152). For example, as shown in Fig. 6A, increasing PB to 3 ATA with 100% He caused a small, but nonsignificant, increase in spontaneous firing rate, which is interpreted as being a baro-insensitive response. Isobaric exposure to 5 min of HBO2 caused a significant increase in firing rate that was reversible on lowering PO2 and repeatable during a second bout of HBO2 (152). Neuronal stimulation by acute exposure to HBO2 is not a generalized property of all solitary complex neurons; in fact, it occurs mostly in CO2/H+-chemosensitive neurons (152), which are thought to function as central cardiorespiratory chemoreceptors (54, 56, 98, 156, 157, 183). The excitatory firing rate response to HBO2 in the solitary complex is reduced or abolished by antioxidants and mimicked by using prooxidants at normobaric pressure (152). These findings indicate that acute exposure to HBO2 stimulates certain dorsal medullary neurons by increased production of free radicals and/or their reactive nonradical derivatives. Moreover, the firing rate response to HBO2 may be enhanced by hypercapnic acidosis, suggesting that intracellular acidosis may augment neuronal sensitivity to HBO2 and ROS in some neurons (152); however, further work is needed to continue testing this latter hypothesis.
|
Fig. 5B shows a variation of the above HBO2 protocol. In this protocol, the brain slice is maintained at the interface of perfusate and gas rather than completely submerged in perfusate (APPENDIX B, Fluid-gas interface brain slice preparation). In this example, PO2 and extracellular pH (pHo) were measured in the tissue bath while a small stream of 95% O2 + 5% CO2 was continuously blown across the surface of the brain slice (59). As 100% He flows into the hyperbaric chamber, PB increases, driving O2 (increases PO2) and CO2 (decreases pHo) into the perfusate, as predicted by Henry's Gas Law. After decompression, bath PO2 and pHo return to their control levels. Notice that this effect of He compression on gas tensions in the perfusate is very different from the situation in the submerged slice in which no changes in PO2 and pHo occurred during He compression due to the absence of any O2 and CO2 in the overlying atmosphere (Fig. 5A). When the test gas line was shut off at normobaric pressure (Fig. 5B, see asterisk), thereby removing the primary source of O2 and CO2 in the interface slice (54, 65), PO2 decreased and pHo increased (i.e., PCO2 decreased), and, during the ensuing He compression and decompression, there were no significant changes in PO2 and pHo. The utility of this alternative protocol for inducing hyperoxia and the precautions required when using it are discussed in APPENDIX B and elsewhere (59).
Figure 7 shows an example of a neuron in the solitary complex that was exposed to hypercapnic HBO2 during He compression by using the fluid-gas interface slice protocol presented in Fig. 5B. AsPB was increased with He, tissue PO2 increased and pHo decreased as O2 and CO2 supplied from the test gas line were driven into the brain slice, which, in turn, stimulated firing rate. We have reported that certain neurons in the solitary complex are stimulated by the combination of CO2-induced intracellular acidification and HBO2 (152). A second compression, this time using air, also stimulated firing rate, but to a lesser extent than did HBHe, which could possibly be due to the narcotic effects of increased PN2 superimposed on the hypercapnic hyperoxic stimulus (154). Finally, a third test, but this time using hypobaric (i.e., PB << 1 ATA) air, caused a decrease in firing rate. Although we have yet to measure tissue PO2 and pHo at hypobaric pressure, we would anticipate that, as PB decreases, PO2 is reduced and pHo is increased. This experiment is an interesting example that illustrates how different gas mixtures administered at equivalent total pressures have different effects on neuronal activity (HBHe vs. hyperbaric air). Furthermore, it demonstrates the technical feasibility, for the first time, of maintaining an intracellular recording during decompression to hypobaric pressures, which will be useful for studying both neuronal barosensitivity (153) and the effects of hypobaric hypoxia (97, 148).
The preceding electrophysiological examples do not differentiate between the presynaptic and postsynaptic effects of hyperoxia and ROS on mammalian neurons (152). It is also possible that O2-induced seizures occur, in part, through an imbalance of excitatory and inhibitory synaptic transmission, involving disinhibition and/or activation of excitatory neurotransmission. Studies in the iCNS suggest that the excitatory effects of HBO2 are attributed to decreased presynaptic release of glutamate (46, 47) and GABA (48). However, a portion of this effect may be due to the effect of pressure per se that occurs independently of increased PtiO2 (47). In addition, Garcia et al. (83) have shown that 3.0 ATA O2 increases the amplitude of the population spike in the rat CA1 hippocampus evoked by Schaffer collateral stimulation and frequently induces secondary spikes. Similarly, King and Parmentier (118) demonstrated that the prevolley potential and the slope of the field excitatory postsynaptic potential both increased during HBO2. However, input-output analysis showed that overall synaptic efficiency was decreased. Taken together, these observations demonstrate that HBO2 affects network activity at multiple levels, yet, ultimately, the net effect is an overall increase in network excitability (83, 118).
Neural Tissue PO2: In Vitro Studies
What range of PtiO2 should be used to mimic the
conditions that occur in vivo during normoxia and hyperoxia? We propose that
neuronal sensitivity to hyperoxia should be studied over a broad range of
PtiO2, which includes both normobaric pressure and
hyperbaric pressure, for two reasons. First, the O2 tension that a
single neuron is exposed to in vivo, at any level of
PIO2, is highly variable, as discussed above.
Second, employing a broad range of test PO2 values
enables neuronal responses to be studied at subthreshold levels of hyperoxia
(nonseizure), threshold levels of hyperoxia (onset of seizures: the Paul Bert
effect), and suprathreshold and lethal levels of extreme hyperoxia (sustained
seizures, paralysis, and death: the John Bean effect). O2 -induced
seizures are a violent neurophysiological end point. It is likely that many
neurophysiological events precede this violent end point and occur at
subthreshold levels of hyperoxia without any outward expression of clinical
signs and symptoms. If the neurophysiological events that precede
O2-induced seizures can be identified, they will provide insight as
to the neural mechanisms that are responsible for the intra- and
interindividual variability of mCNS O2 toxicity
(43,
105), as well as ways to
prevent onset of seizures. In this context, it is highly relevant that
normobaric hyperoxia (i.e., PIO2 >150 but
760 Torr), which was once thought to be innocuous to neurological
function, is now known to alter neural activity and to stimulate certain
functional networks (22,
88,
121,
144,
186,
226). Thus hyperoxia at
presumably nontoxic levels is an environmental modulator of neuronal activity
(44,
71,
135). For example, in
respiratory control, the well-known paradox of hyperoxic hyperventilation at
normobaric and hyperbaric pressures is thought to be due, in part, to
increased excitability of neurons in brain stem respiratory centers
(51,
142). In support of this
hypothesis, we have shown that HBO2 stimulates putative central
CO2/H+-chemoreceptor neurons in the dorsal brain stem
(152), which is one of
several sites of central chemoreception for breathing
(156,
157,
183). Because hyperoxic
hyperventilation serves no adaptive value to the animal in terms of
O2 homeostasis, it is more likely that the paradoxical rise in
ventilation represents the high sensitivity of the mammalian brain stem
respiratory centers, in particular, the central chemoreceptors to oxidative
environments (152).
In in vitro tissue preparations of the mCNS, which lack blood flow,
PtiO2 is affected by several factors, including region
(gray vs. white matter), position of the tissue in the bath (submerged vs.
fluid-gas interface), flow rate of oxygenated perfusate over the tissue,
tissue temperature, tissue thickness, metabolic rate, and age of the animal
(31,
110,
151,
161). Measurements of
PtiO2 in metabolically active brain slices maintained
under control conditions (95% O2 at PB 1 ATA) and
during HBO2 (98% O2 at PB 2.4-3.0
ATA) reveal several features about the brain slice model of mCNS O2
toxicity (151). First, a
significant gradient of PtiO2 occurs in the brain slice
that depends on the level of PO2 in the perfusate and
the tissue metabolic rate. Figure
8 shows that, as the polarographic electrode is advanced in
50-µm intervals toward the submerged brain stem slice, the
PO2 in the overlying perfusate begins to decrease as
O2 is consumed by cells in the upper layers of the slice. At
normobaric pressure, at the upper tissue surface, PtiO2
averages 436 Torr, and, at the middle of the 300-µm slice (150 µm
deep), PtiO2 averages 291 Torr
(151). As the polarographic
electrode passes through the bottom of the slice and downward, the
PO2 gradient due to O2 consumption by cells
in the bottom layers of the slice is evident.
During exposure to graded levels of HBO2,
PtiO2 in the slice increases significantly
(Fig. 8, top
PO2 trace), encompassing a range of values circumscribed
by the dark gray area in Fig. 4 (#3, solid symbols; data come from experiments as illustrated in
Fig. 8). The magnitude of the
PtiO2 gradient from the surface to the center of the
slice [see Fig. 7 in Mulkey et
al. (151)] is decreased
compared with the gradient at 1 ATA, which has been attributed to decreased
O2 consumption during extended exposures to HBO2 (>10
min). The range of PtiO2 that is produced in a 300-µm
brain slice exposed to 2.4-3.4 ATA of O2 is similar to
PtiO2 measured in the intact mCNS of an animal breathing
100% O2 (open circles) at PB of 5.4 to >6 ATA and 95%
O2 + 5% CO2 (open squares) at PB 3.8 to >5
ATA (range d-g on y-axis in
Fig. 4). Thus we have a good in
vitro model for studying CNS O2 toxicity. As was shown above in
Fig. 6 and elsewhere
(82,
152), acute exposure (10
min) to this level of extreme hyperoxia stimulates neuronal activity in the
brain stem and hippocampus in a reversible and repeatable fashion. The control
condition, however, which will be discussed next, is not normoxic; it is
hyperoxic and needs to be reevaluated, given the deleterious effects that
oxidative stress has on certain neurons
(67,
107).
Choosing a Control Level of PtiO2
.. .oxygen pressure in the mammalian CNS is maintained at a level which is sufficiently high to ensure undisturbed function of brain cells and sufficiently low to minimize generation of free radicals.—Erecinska and Silver (71)
The above discussion indicates that hyperoxia and increased levels of ROS
can alter neuronal excitability and synaptic transmission. It becomes
important, therefore, to reassess the control level of
PtiO2 used in most in vitro electrophysiology studies,
for it has been stated that "... it is a standard Krebs bicarbonate
buffer, pregassed with 5% CO2 and 95% O2... maybe we use
too much O2. I don't know."
(3). The use of 95%
O2 at PB of 1 ATA produces a
PO2 in the perfusate of 720 Torr
(24,
81,
110,
151), which, in turn,
produces a PtiO2 in a submerged 300-µm slice ranging
from 436 Torr (surface) to 291 Torr (center)
(151). Stated another way,
the core of a 300-µm brain slice, where PtiO2 is
lowest, is one order of magnitude or more above the normoxic level in the
intact mCNS! As shown in Fig. 4
(#3), there is a range of control PtiO2 in the brain
slice that is equivalent to the level of PtiO2 that
occurs in vivo when 2.2-2.9 ATA O2 is breathed (#1; compare
range d on y-axis to range e on x-axis in
Fig. 4). Thus the standard
control conditions used for most in vitro experiments (brain slice and brain
stem-spinal cord preparations) mimics the PtiO2 that
occurs during HBO2 (2.2-2.9 ATA), excluding the physical pressure
component. Regardless, with few exceptions
(71,
151,
186,
226), this high level of
PtiO2 is accepted by most investigators as the
appropriate level of O2 for maintaining thick and thin preparations
of the mCNS (3,
31,
110,
161).
What effects are 95% O2 at normobaric pressure having on the isolated mCNS (3), and, moreover, are neuronal responses observed during HBO2 (e.g., Fig. 6) blunted by the hyperoxic control condition that is used during slice preparation and incubation (58, 150, 152)? It is interesting that very few investigators have ever questioned the possibility that too high a control PtiO2 is used for brain slice (and isolated brain stem-spinal cord) experiments, especially because the optimum PtiO2 for cell growth and viability has been determined experimentally for neuronal cultures; for example, 9% O2 at normobaric pressure produces PtiO2 of 68 Torr, which results in maximum cell survival and synthesis of neurofilament in cultures of neonatal rat cerebral cortex (30, 112). Historically, 95% O2 was used in brain slices to produce an adequate partial pressure gradient to deliver O2 to the core of the slice to avoid tissue hypoxia and anoxia (3). Whereas an anoxic core was a concern for brain slices that ranged in thickness from >300 up to 1,000 µm (24, 110), in recent years, the trend has been for brain slices to be cut thinner, typically at 100- to 300-µm intervals (20, 69). In addition, because of widespread use in brain slices of infrared video microscopy for patch clamping and fluorescence microscopy, the tendency in recent years has been to study neurons exclusively in the outermost cell layers (55, 69, 78, 176), where PtiO2 is highest (24, 81, 110, 151). Despite the gradual reduction in thickness of the slice preparation during the past decade, electrophysiologists have continued to employ 95% O2 as the control level of PO2 with disregard for what effect increased levels of ROS might be having on baseline neuronal activity, responsiveness to the test stimulus, and slice viability in general.
In brain slice electrophysiology, very little is known about how normobaric
O2 in the range of >15% up through 95% affects excitability
of mammalian neurons. Typically, 15% O2 is used to study the
effects of hypoxia and anoxia in the brain slice model [e.g., see Refs.
65 and
110 and Table 1 in Mulkey et
al. (152)]. When a lower
level of control O2 was studied in slices, it was reported that
95-100% O2, compared with 21% O2 (control), impaired
neuronal thermosensitivity in hypothalamic neurons
(186), altered
Vm and firing patterns in hippocampal neurons
(23), and attenuated
O2 sensitivity of the peripheral chemoreceptors for
cardiorespiratory control
(121,
144). All of these authors
proposed that the changes in neuronal activity observed in 95-100%
O2 were due to the effects of increased production of ROS during
normobaric hyperoxia, which were decreased in 21% O2. This
hypothesis is supported by reports
(119,
168) that the amount of
tissue damage resulting from lipid peroxidation was significantly increased in
brain slices incubated in 95% O2 compared with 21% O2.
Similarly, there were greater levels of F2-isoprostanes and
F4-neuroprostanes in hippocampal slices maintained for up to 7 h in
perfusate equilibrated with 95% O2, compared with control slices
analyzed before incubation in 95% O2 (J. Fessel, personal
communication; Ref. 77).
Isoprostane-like compounds are reaction products derived from oxidation of
docosahexaenoic acid, a component in neuronal membranes, and are used as a
marker for increased oxidative stress. It is not clear, however, that 21%
O2 is a better level to use in all cases
(81), and a recent brain slice
study in the hippocampus suggests that the optimal control
PtiO2, as determined by electrophysiological criteria,
may, in fact, lie somewhere between 21 and 50% O2
(219).
Given the effect that oxidative stress has on neuronal activity (82, 118, 150, 152, 169, 170), future studies are needed to determine the effects of normobaric hyperoxia on brain slice metabolism and electrophysiology. Because of the characteristic PtiO2 gradient that occurs in the slice (Fig. 8), due to long diffusion distances and regional differences in O2 consumption, neurons within the slice will always be exposed to a range of PtiO2, regardless of the level of PO2 in the perfusate. However, we are proposing that the range of O2 used during slice preparation, incubation, and recording can be adjusted down toward a level that is intermediate to tissue hypoxia and hyperoxia. The use of lower control PO2 and/or addition of a suitable anti-oxidant to the medium for certain types of experiments may likewise enhance slice viability and neuronal activity as already demonstrated for cell culture preparations (30, 112), although to do so will be problematic for studies of the effects of hyperoxia. In thicker preparations like the neonatal brain stem-spinal cord, it may be advantageous to include antioxidants to reduce the effects of oxidative damage to the outermost cell layers where PtiO2 is highest. Alternatively, the perfused version of this in vitro preparation can be used, the so-called rat "working heart-brain stem preparation," in which the vascular system is perfused with oxygenated buffer. However, it will be important to lower the level of PO2 in the perfusate, which also is hyperoxic (95% O2) compared with the intact mCNS, producing an average PtiO2 of 294 Torr (222).
|
PERSPECTIVE |
|---|
|
TOPABSTRACTGlossaryOVERVIEW OF BARO-RELATED...CENTRAL EFFECTS OF PRESSURE...NARCOTIC AND TOXIC PROPERTIES...RESEARCH APPLICATION: HYPEROXIA...PERSPECTIVEAPPENDIXES A-D: HYPERBARIC...DISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
We would also propose that the utility of hyperbaric electrophysiology is not limited to just these specialized fields of environmental physiology and hyperbaric medicine. In addition, hyperbaric electrophysiology is a powerful research tool that can be used for studying the neurophysiological consequences, at the level of the single neuron, of oxidative stress (9, 43, 106, 150, 152). For example, because of the almost universal use of 95% O2 as the control level of O2 in brain slice experiments, which leaves essentially no room for elevating PtiO2 by conventional slice methods, the only means of identifying the effects of hyperoxia on mammalian neurons in slices was to do so by using hyperbaric electrophysiology. Without this novel research tool, it is unlikely that the effects of hyperoxia on the membrane properties of putative central CO2/H+-chemoreceptor neurons would have been discovered (152). Moreover, hyperbaric electrophysiology can be used to study cellular mechanisms of anesthesia; for example, hyperbaria can be used either to study the relative narcotic potencies of various gases on neuronal excitability (28), or as a tool to study pressure reversal of anesthesia (113, 115, 167). Hyperbaric electrophysiology can also be used to study other neurological questions that are associated with increased neural tissue pressure, such as intracranial hypertension that occurs during adaptation to inertial force environments (224), space adaptation syndrome (109), and high-altitude sickness (97). Similarly, studying how increased intracranial pressure alone and in conjunction with other cellular insults (e.g., anoxia and hypoxia, acidosis, osmotic imbalance, etc.) affects neuronal activity will be a useful model for studying cellular responses to traumatic brain injuries (4, 64, 140, 181).
|
APPENDIXES A-D: HYPERBARIC METHODS FOR IN VITRO ELECTROPHYSIOLOGY AND MICROSCOPY |
|---|
|
TOPABSTRACTGlossaryOVERVIEW OF BARO-RELATED...CENTRAL EFFECTS OF PRESSURE...NARCOTIC AND TOXIC PROPERTIES...RESEARCH APPLICATION: HYPEROXIA...PERSPECTIVEAPPENDIXES A-D: HYPERBARIC...DISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
4 ATA for
N2 narcosis, and from 15 to 70 ATA for HPNS
(Fig. 1). Moreover, two
conditions must be considered as possibly causing neuronal dysfunction:
increased gas partial pressures (PO2,
PN2, and PCO2; see NARCOTIC
AND TOXIC PROPERTIES OF GASES) and increased hydrostatic pressure or
"pressure per se" (see CENTRAL EFFECTS OF PRESSURE PER
SE). The following sections (APPENDIX A) describe methods for
differentiating the effects of pressure per se (barosensitivity) from those of
increased gas partial pressure (chemosensitivity and narcosis) on neurons (in
vitro). The variations used in these methods and their inherent complexity
warrant a brief summary of their key features and differences for those
unfamiliar with hyperbaric cell physiology. Appendix A: Testing Neuronal Barosensitivity
Hydrostatic compression chambers. There are two general methods of compression used to study neuronal barosensitivity. The first is hydrostatic compression of the submerged tissue preparation by using a liquid inert medium of mineral oil, paraffin, or perfusate. Hyperbaric chambers designed for hydrostatic compression typically have a smaller internal volume (milliliters) compared with hyperbaric chambers designed for cylinder gas compression (tens of liters; see next section). The small internal volume of a hydrostatic compression chamber is beneficial because it enables the investigator to rapidly increase ambient pressure when studying the biophysical effects of an extreme level of hyperbaria (184). Hydrostatic compression chambers have typically been used to impose extremely high levels of hydrostatic pressure, usually >100 ATA, on nonmammalian neurons and axons (214, 215) and large nonneuronal cells (91, 184). The tissue preparation is submerged in a static bath, and a pneumatic pump delivers compression medium to the interior of the chamber, completely displacing any gas overlying the tissue bath. Cam-driven microdrives are used to manipulate the recording microelectrode, rather than electric microdrives, because the latter are incompatible with aqueous compression medium (91, 166, 215). Typically, adiabatic temperature changes occur during rapid compression and decompression that require several minutes to dissipate, which often confounds electrophysiological data collected during the act of compressing and decompressing the chamber (91, 215). In addition, because most chamber designs use a static bath, the composition of the tissue bath cannot be manipulated by the investigator, and the small internal volume of the chamber, which facilitates rapid compression and decompression, makes it difficult to position and manipulate the recording micropipette (184).
Cylinder gas compression chambers: HBHe. The second method of compressing the tissue preparation uses cylinder gases, typically pure He. It has been the preferred method for working at moderate levels of hyperbaria and for working with in vitro preparations of the mCNS (58, 75, 191). It is the method that was used in all of the examples presented in this review article. The larger internal volume of the hyperbaric chamber and use of gaseous compression medium allow the electrophysiology equipment (microdrives, stimulating electrode, etc.) to be placed directly inside the chamber, with the brain slice for easy positioning and manipulation, when the chamber door is open at room pressure, or by remote control after the chamber door is sealed (58, 59). In addition, this style of chamber can be used to hold a custom-built, fixed-stage epifluorescence microscope, equipped with Hoffman modulation contrast optics and video-imaging capabilities, to visualize neurons in a brain slice for patch-clamp electrophysiology and/or ratiometric fluorescence imaging (see APPENDIX D) (84).
The tissue preparation is either superfused (i.e., maintained at a fluid-gas interface) or submerged in perfusate that is continuously delivered via a high-pressure liquid chromatography (HPLC) pump. Pure He flows into the pressure chamber, which, in turn, compresses the tissue bath and neural tissue. At low levels of hyperbaria, He is believed to effectively mimic the effects of hydrostatic compression (58, 192, 197, 198). Because He has the lowest solubility in lipid membranes of any of the gases (Fig. 3), it has no known narcotic or toxic effects on cells at low levels of hyperbaria (28, 215). Thus, despite increasing the PHe in the tissue during compression, any neuronal response to HBHe, at moderate pressures, is attributed to the effects of pressure per se rather than the effects of increased PHe (58, 72, 73, 150, 153, 197, 215); see below, Compression media: HBHe vs. hydrostatic compression.
As in most brain slice experiments
(3), 95% O2 + 5%
CO2 gas mixture is used to aerate a bicarbonate-buffered perfusate
at normobaric pressure before it is pumped into the hyperbaric chamber. The
use of 95% O2 at PB 1 ATA establishes a large
O2 diffusion gradient from the perfusate into the avascular brain
slice, resulting in a characteristic PtiO2 profile that
is hyperoxic at all depths in a 300-µm brain slice
(151) (Figs.
4 and
8). Similarly, a qualitatively
similar but larger PtiO2 gradient profile is measured in
the isolated neonatal rat brain stem-spinal cord
(31,
161), which has also been
used in hyperbaric studies
(197,
198). The potential problem
arising from using a hyperoxic control medium for studies of mCNS
O2 toxicity, and for in vitro studies in general, was discussed
above (Neural Tissue PO2: In Vitro Studies and
Choosing a Control Level of
PtiO2).
At normobaric pressure, medium PO2 and PCO2 and pHo remain constant as perfusate is pumped into the unpressurized hyperbaric chamber, because the solution has no opportunity to degas before reaching the tissue bath (58, 75, 191). Once at the tissue bath, however, gases dissolved in the perfusate diffuse into the overlying He atmosphere (151). The continuous flow conditions at the tissue slice prevent any significant changes in tissue PO2, PCO2, and pHo while at normobaric pressure (Fig. 5A). A common misconception is that, once He compression commences, PO2 and PCO2 in the perfusate and tissue will increase as PB increases. However, as shown above in Fig. 5A, this is not the case for two reasons. First, the hyperbaric chamber contains no source of O2 and CO2 in the overlying atmosphere to be driven into the tissue (58, 75, 197, 198); the amount of O2 lost from the perfusate to the overlying atmosphere is infinitely small compared with the volume of the hyperbaric chamber and the frequency with which it is flushed with He (59). Second, the solubility coefficients for O2 and CO2 and the negative log of the dissociation coefficient of a bicarbonate buffered solution are essentially unchanged when PB is increased to hyperbaric pressures <<100 ATA. Consequently, there are virtually no changes in dissolved gas pressures and pH of the perfusate during compression and decompression (159, 199, 200). Thus bath and tissue PO2 and PCO2 (and pHo) remain at control levels during He compression (Fig. 5A). In addition, bath and tissue temperatures can be tightly regulated by using a servo-controlled temperature regulator, despite fluctuations in atmospheric temperature during He or air compression and decompression (58).
Other variations of the cylinder gas compression protocol exist (34, 96, 191). One variation to the above method that is worth elaborating interposes a physical diffusion barrier between the overlying chamber atmosphere and the tissue bath by using either mineral oil (34) or a Plexiglas lid (96, 124). In this case, air rather than He was used to compress the tissue bath. Because neither PO2 nor PN2 increases in the perfusate, this approach provides true hydrostatic compression of the tissue preparation without any accompanying change in PO2 or PN2 (34, 96, 124). This same approach could also be used to study the effects of hydrostatic compression vs. He compression to test for any possible narcotic effects of HBHe at higher pressures, which are discussed next.
Compression Media: HBHe vs. Hydrostatic Compression. What is the evidence that HBHe can be used to mimic the effects of hydrostatic compression on the mCNS (Fig. 2)? All inert gases possess narcotic actions that are directly related to their lipid solubility (19, 28) (Fig. 3). However, the narcotic potency of He in excitable tissues, when it is detected, is reported to be only weak and of little consequence and is usually observed only at the highest levels of hyperbaria; e.g., >>100 ATA (163). For example, as is the case with other narcotic gases, HBHe (101-150 ATA) can partially reverse the deleterious effects of increased hydrostatic pressure on animal behavior and motor activity (11, 66). Less common, under certain in vitro conditions (28), effects of only 6-8 ATA He on cellular processes have been reported (87, 182). Thom and Marquis (202) have speculated that the cellular effects of HBHe, like the other narcotic gases, are caused in part by gas molecules dissolving in the hydrophobic regions of the cell membrane, causing distortion of the membrane and possibly disturbing hydrophobic interactions in proteins through non-covalent interactions.
HBHe potentially has other effects that are unrelated to its weak narcotic
nature. For example, Thom
(201) reported that low
levels of HBHe (2 ATA) caused increased production of superoxide radicals
in two different free-radical generating systems in vitro. A similar effect of
HBHe in the mCNS could potentially alter neuronal activity; however, Mulkey et
al. (153) found that neuronal
barosensitivity to HBHe (4 ATA), in solitary complex neurons, was
maintained in the presence of the antioxidants ascorbate (D. K. Mulkey and J.
B. Dean, unpublished observations) and Trolox C
(153). This suggests that, at
least in solitary complex neurons, barosensitivity to HBHe
(Fig. 2) is unrelated to
increased production of ROS.
In general, the effects of HBHe on cells, tissues, and intact organisms are
comparable to the effects of hydrostatic pressure over a range of mechanically
tolerable pressures (28,
141,
215). The few studies that
have compared the effects of both types of compression media on
electrophysiological properties of neurons and axons confirm this assertion.
Wann et al. (215) reported
that the effects of HBHe (205 ATA) applied at a rate of 6.8 ATA/s exert
an effect on Helix neurons that is indistinguishable from that
observed during hydrostatic compression
(214); both compression media
increased the duration of the action potential by slowing the peak rates of
depolarization and repolarization, decreased Rin, and
depolarized Vm. These effects on the action potential were
due primarily to decreased kinetics of Na+ and K+
activation and inactivation rather than a change in the magnitude of each
early and late current (92).
Interestingly, we saw the same effects on the action potential in solitary
complex neurons, but at much lower pressures of only 2.5-4 ATA He; however, we
have not compared these experiments to the effects of hydrostatic compression
(D. K. Mulkey and J. B. Dean, unpublished observations). HBHe and hydrostatic
pressure (35-137 ATA) also had identical effects on evoked synaptic
extracellular responses in the isolated rat superior cervical ganglion; both
compression media had no effect on the amplitudes of the preganglionic and
postganglionic action potentials, depressed synaptic transmission, and
antagonized the partial conduction block produced by halothane,
methoxyfluorane, and ethyl alcohol
(115). Bennett
(16) reported that 11 ATA He
(with normoxia) had no effect on either the brain stem auditory evoked
response or spontaneous electrical activity in the cortex in the cat, and Roth
(179) reported that 68 ATA He
had no effect on the stimulus-response curve for an isolated frog sciatic
nerve.
Comparison of the effects of He compression vs. hydrostatic compression has
also been made in intact animals and was summarized by Brauer et al.
(28). Based on quantitative
measurements of abnormal behaviors and motor deficits induced during
hyperbaria (50 to 200 ATA), in aquatic animals (newts) and
liquid-breathing mammals (mice and dogs breathing fluorocarbon), both of which
lend themselves to hydrostatic compression in the presence and absence of He,
they concluded "... that differences between the two modes of
compression were not observed." Thus, based on evidence from
electrophysiological studies of excitable cells and tissues, and physiological
and behavioral studies in intact animals, we would conclude that using He to
compress the tissue bath and brain slice effectively mimics hydrostatic
compression at PB <5 ATA (and higher), and that any changes in
neuronal excitability during exposure to HBHe, such as seen in the examples
shown in Fig. 2, reflect the
effects of pressure per se rather than increased PHe
(58,
82,
150,
153).
Appendix B: Testing Neuronal Chemosensitivity to Hyperbaric O2, CO2, and N2
Submerged tissue preparations: high-pressure sample cylinder. The
above methods enable testing of neuronal barosensitivity at controlled levels
of tissue PO2, PCO2, and pH. Once
at hyperbaric pressure, tissue PO2,
PCO2, and/or PN2 can be raised
independently of any additional increase in PB (i.e., under
isobaric conditions) by using a high-pressure sample cylinder
(58,
82,
93,
96,
151,
197). Each high-pressure
cylinder contains perfusate that is equilibrated with a gas mixture containing
O2, CO2, and/or N2, whose total pressure is
equivalent to, or just slightly greater than, the total pressure in the
hyperbaric chamber. For example, the protocol used in our laboratory for
testing neuronal responsiveness to HBO2 is isobaric replacement of
the control perfusate [PO2 420 Torr
(151)] while at hyperbaric
pressure with perfusate supplied from a high-pressure sample cylinder that has
been pressurized by using a high-dose O2 gas mixture
[PO2 1,100 Torr
(151)]. In this case, the
FCO2 is also reduced proportionately to maintain
PCO2, and thus pHo, the same as in control
perfusate (58,
82,
150-152).
With the use of the protocol shown in Fig.
5A, the effects of HBO2 vs. pressure per se
(i.e., HBHe) on neuronal activity have been tested
(58,
82,
152). Two examples of this
method were shown above in Fig.
6. In addition, PCO2 can be increased with
PO2 to study the effects of HBO2 plus
hypercapnic acidosis on neuronal activity
(152).
A comparatively simpler method to study the effects of a hyperbaric gas, which does not involve the use of a high-pressure sample cylinder or changing the perfusate, is to compress the hyperbaric chamber with the gas of interest. This approach is particularly well suited for studying the effects of hyperbaric air or N2 on neuronal function, but not HBO2, due to the risk of a chamber fire when electrical equipment is operated in a pressurized hyperoxic atmosphere (59, 185). Pressurizing the chamber with air (21% O2 + 79% N2) or pure N2 compresses the bath and tissue, while increasing the bath and tissue PN2, for investigating cellular mechanisms of N2 narcosis (34, 35, 154); as usual, O2 (95%) and CO2 (5%), at normobaric pressure, are dissolved in the perfusate and pumped into the chamber by using the HPLC pump.
In addition, Imbert et al. (102) have designed an elaborate system for saturating perfusate with different gases, including H2, to study the effects of hyperbaric gas mixtures on various in vitro tissue preparations. This system has yet to be used for electrophysiology studies, however.
Fluid-gas interface brain slice preparation. An alternative to the
submerged slice preparation is the fluid-gas interface slice preparation in
which either a control (95% O2 + 5% CO2) or test gas
mixture is blown over the surface of the brain slice maintained at the
interface between perfusate and an atmosphere of control and test gas mixture
(59,
154). Without changing the
source of perfusate, which also is aerated with 95% O2 + 5%
CO2, tissue gas tensions are manipulated rapidly by changing the
composition of the gas mixture flowing through the small test gas line. This
approach has been used at normobaric pressure to study the effects of anoxia
(0% O2 + 5% CO2 + 95% N2), hypoxia (15%
O2 + 5% CO2 + balance N2), and hypercapnic
acidosis (7-15% CO2 + balance O2) in brain slices
(54,
65). Similarly, at hyperbaric
pressure, blowing 95% O2 + 5% CO2 gas across the upper
surface of the interface slice, while compressing the hyperbaric chamber with
100% He, increases bath and tissue PO2 and
PCO2 and decreases pHo
(59).
Figure 5B shows an
example of PO2 and pHo measured in the tissue
bath while 95% O2 + 5% CO2 were continuously blown
across the brain slice surface at normobaric and hyperbaric pressures. Notice
that both PO2 and PCO2 will
increase during He compression as O2 and CO2 supplied by
the test gas line are driven into the slice and perfusate. Using pure
O2 as the test gas, rather than 95% O2 + 5%
CO2, while pressurizing the chamber with He would allow a rapid
HBO2 test without any change in pHo. However, certain
precautions must be taken, which are summarized elsewhere
(59), to prevent the slow
build-up atmospheric PO2 being supplied by the
small-diameter test gas line, which would increase the risk of a chamber
fire.
Appendix C: Comparison of Electrophysiology Recording Techniques
Pressure, like temperature, is one of the fundamental physical variables which constrain living entities. Because it is difficult to manipulate pressure as an experimental variable, however, it has not been a subject of casual study; indeed, very powerful motivation is required to pursue studies of pressure effects.—Kendig, Grossman, and Heinemann (114).
Not all electrophysiological recording methods have been equally successful when used with isolated tissue preparations during compression and decompression. The relative success of each type of electrophysiological recording in the past, however, may have been determined more by deficiencies in hyperbaric chamber designs, which limited the investigator's accessibility to the microelectrode at normobaric pressure, rather than any inherent limitation of the electrophysiological recording method per se (58, 59). The following three sections compare and contrast the various types of electrophysiological recordings and their relative merits and problems as reported in the literature to date.
Extracellular recording. Before the mid-1980s, hyperbaric electrophysiology studies employed nonmammalian and nonneuronal in vitro models exclusively. Pioneering work by several independent laboratories in the late 1980s and 1990s, however, demonstrated that in vitro preparations of the mCNS, such as the rat hippocampus and cerebellum brain slices (72-74, 76, 191, 192, 216) and neonatal rat brain stem-spinal cord (197, 198), could be manipulated under hyperbaric conditions to study HPNS. At that time, making intracellular recordings of mammalian neurons proved to be technically challenging due to the comparatively small size of the neurons, inaccessibility of the tissue preparation and microelectrode once the pressure vessel was sealed, and the challenge of maintaining mechanical stability of the electrophysiological recording while "diving" and "surfacing" (191, 192). Consequently, the majority of in vitro hyperbaric studies of the mCNS have selected more robust electrophysiological techniques, such as extracellular recording of evoked population spikes in the hippocampus (76, 118, 191, 192, 229), macropatch clamp recordings in the cerebellar cortex (72-74), and cervical nerve recordings of respiratory-related neural activity (197, 198). Compared with intracellular recordings, extracellular recordings are easier to initiate once the chamber is sealed and to maintain during the ensuing compression and decompression periods.
Intracellular recording. Intracellular recordings provide additional mechanistic information about neuronal function by quantifying changes in a variety of cellular properties, including Vm and currents, membrane conductance and Rin, integrated firing rate, evoked repetitive firing properties, and voltage waveforms (action potential, afterhyperpolarizing potential, chemical and electrical synaptic potentials). In addition, intracellular dye injection allows recovery of the physiologically identified cell for subsequent morphological and immunohistochemical analyses. Intracellular recordings, however, are extremely sensitive to mechanical disturbances. The majority of successful hyperbaric studies that used the intracellular recording technique have been conducted in robust in vitro preparations, such as large nonneuronal cells (49, 93, 96) and large invertebrate neurons and axons (34, 166), which, compared with mammalian neurons, are relatively less sensitive to mechanical disturbances.
Mammalian neurons, therefore, are more difficult to record intracellularly during manipulation of PB (191, 192, 216). Only two laboratories have studied the intracellular properties of neurons in rat brain slices, focusing on cellular mechanisms of HPNS (191, 192, 216), barosensitivity to moderate pressures (150, 153), and O2 toxicity (150, 152). Areas studied include the solitary complex in the dorsocaudal medulla oblongata (58, 150, 152, 153) and the CA1 cell layer of the hippocampus (82, 191, 192, 216) (e.g., Figs. 2, 6, and 7). Successful intracellular recording of a mammalian neuron requires a sharp-tipped microelectrode, which can easily plug or break as it tracks through the tissue and, therefore, must be frequently replaced. If the pressure chamber is too difficult to open and close, then changing the microelectrode becomes problematic and an impediment to the experiment (191, 216). The requirement for an operator-friendly door was fulfilled recently in a hyperbaric chamber design that combined a commercially available door with the best features from earlier chamber designs, which collectively, optimized mechanical stability of the microelectrode and tissue preparation (58, 59).4 Using this style of hyperbaric chamber, intracellular recordings have been made with routine success from mammalian neurons during exposure to 2-4 ATA of He, HBO2, hyperbaric air, and hypobaric air, as shown here and elsewhere (58, 150, 152, 153). Continuous recordings also have been maintained while cycling PB to and from 20 ATA of He (Fig. 2B).
Patch-clamp recording. Patch-clamp recordings of whole cell and single-channel ionic currents have been made at extremely high hydrostatic pressures in nonmammalian cells to study HPNS and the effects of large hyperbaric pressures (91, 130, 132, 184), and during exposure to HBO2 (133). This technique involves visualizing and patching onto a dissociated cell maintained in a removable bath that is attached to the stage of an inverted microscope. Once the tight seal has formed, the head stage, pipette, cell, and bath are manually transferred to a hyperbaric pressure chamber and compressed hydrostatically (see Hydrostatic compression chambers in APPENDIX A). Recently, Heinemann's group (184) demonstrated the feasibility of conducting dual-electrode voltage-clamp experiments in oocytes during hydrostatic compressions up to 600 ATA. To our knowledge, tight-seal patch-clamp recordings have not been attempted with mammalian neurons under hyperbaric conditions by using a cylinder gas-compression chamber (i.e., HBHe); however, we anticipate that these types of recordings will also be feasible given recent improvements in hyperbaric chamber design (58, 59).
Appendix D: Hyperbaric Microscopy
The application of video microscopy (55, 195) and epifluorescence microscopy (78, 178, 183) during the past decade have greatly enhanced the power of brain slice studies conducted at normobaric pressure. Measurement of proposed signaling ions and molecules in a brain slice using fluorescent probes and ratiometric fluorescence microscopy, done simultaneously with measurements of neuronal output (i.e., Vm and firing rate) by using single-cell electrophysiology, enables investigators to determine the relationship between gas-induced stimuli and changes in neuronal function (e.g., Ref. 78).
While light and epifluorescence microscopy have been used in hyperbaric research to a limited extent, neither technique has been used in conjunction with electrophysiology. Typically, a small-volume hydrostatic compression chamber, referred to as an optical pressure chamber, is adapted to the stage of an inverted microscope. The in vitro tissue preparation is illuminated through a window in the top of the chamber and visualized through a small aperture in the bottom of the chamber, which is sealed by a glass coverslip, using a high numerical aperture objective (21, 164, 180). This design, however, prohibits introducing a recording microelectrode into the tissue inside the optical pressure chamber. To overcome this problem, our laboratory (84) recently adapted the optical components required for conducting video microscopy and epifluorescence microscopy to the interior of a hyperbaric chamber. Presently, the hyperbaric microscope is being used to study the effects of HBO2 on baseline pHi and pHi recovery mechanisms in mammalian neurons. Using this new research tool, we anticipate that it also will be possible in the future to conduct electrophysiology in visualized neurons during exposure to hyperbaric gases, an approach that is now used routinely at normobaric pressure (e.g., Refs. 55, 78, 98).
|
DISCLOSURES |
|---|
|
TOPABSTRACTGlossaryOVERVIEW OF BARO-RELATED...CENTRAL EFFECTS OF PRESSURE...NARCOTIC AND TOXIC PROPERTIES...RESEARCH APPLICATION: HYPEROXIA...PERSPECTIVEAPPENDIXES A-D: HYPERBARIC...DISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTGlossaryOVERVIEW OF BARO-RELATED...CENTRAL EFFECTS OF PRESSURE...NARCOTIC AND TOXIC PROPERTIES...RESEARCH APPLICATION: HYPEROXIA...PERSPECTIVEAPPENDIXES A-D: HYPERBARIC...DISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
1 One ATA is the PB at sea level, which is equivalent to 760 Torr.
ATG is defined as chamber pressure minus room pressure; thus at sea level (1
ATA), 2 ATA inside a pressure chamber is 1 ATG, 3 ATA is equivalent 2 ATG, and
so forth. Ambient pressure increases 1 atm for every 33 fsw depth. Therefore,
1 ATA = 0 fsw, 2 ATA = 33 fsw, 3 ATA = 66 fsw, etc. The SI unit for pressure
is the kilo-Pascal (kPa), where 1 ATA = 101.3 kPa. Other commonly used
pressure equivalents for 1 ATA include 14.7 psia, 0.101 mega-Pascal (MPa), and
1.013 bars. 
2 Decompression sickness, commonly known as the "bends" and
"chokes," is not covered here because it is a problem that occurs
subsequent to decompression while diving and/or flying, and this review
focuses on neurophysiological challenges that occur while breathing hyperbaric
gases; i.e., gases in the compressed state.
3 Clinical administration of O2 at PB >1 ATA is
called HBOT and it is used clinically to help "resolve certain
recalcitrant, expensive, or otherwise hopeless medical problems"
(38). Some of the accepted
uses of HBOT for acute medical emergencies include carbon monoxide poisoning,
necrotizing fasciitis, air embolism, decompression sickness, gas gangrene, and
radiation necrosis (38). In
addition, HBOT has been used to treat a variety of neurological problems, such
as traumatic brain injury (38,
158). The application of HBOT
to neurological problems is still in its infancy, however, and debated because
of the lack of relevant cellular and clinical research on these potentially
important applications of hyperbaric medicine.
4 We recently determined that the single-bolt closure on the old Bethlehem
Hyperbaric Research Chamber (model NB-878, circa 1966) also works well. This
chamber has a maximum working pressure of 10 ATG.
|
REFERENCES |
|---|
|
TOPABSTRACTGlossaryOVERVIEW OF BARO-RELATED...CENTRAL EFFECTS OF PRESSURE...NARCOTIC AND TOXIC PROPERTIES...RESEARCH APPLICATION: HYPEROXIA...PERSPECTIVEAPPENDIXES A-D: HYPERBARIC...DISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
4 ATA) increases
membrane conductance and firing rate in the rat solitary complex. J
Appl Physiol 95:
922-930, 2003.
This article has been cited by other articles:
|
|
 |
|
  D. P. D'Agostino, D. G. Colomb Jr., and J. B. Dean Effects of hyperbaric gases on membrane nanostructure and function in neurons J Appl Physiol, March 1, 2009; 106(3): 996 - 1003. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. M. Johnson, M. A. Haxhiu, and G. B. Richerson GFP-expressing locus ceruleus neurons from Prp57 transgenic mice exhibit CO2/H+ responses in primary cell culture J Appl Physiol, October 1, 2008; 105(4): 1301 - 1311. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G.-J. Gu, Y.-P. Li, Z.-Y. Peng, J.-J. Xu, Z.-M. Kang, W.-G. Xu, H.-Y. Tao, R. P. Ostrowski, J. H. Zhang, and X.-J. Sun Mechanism of ischemic tolerance induced by hyperbaric oxygen preconditioning involves upregulation of hypoxia-inducible factor-1{alpha} and erythropoietin in rats J Appl Physiol, April 1, 2008; 104(4): 1185 - 1191. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. P. D'Agostino, R. W. Putnam, and J. B. Dean Superoxide ({middle dot}O2 ) Production in CA1 Neurons of Rat Hippocampal Slices Exposed to Graded Levels of Oxygen J Neurophysiol, August 1, 2007; 98(2): 1030 - 1041. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Zakynthinos, P. Katsaounou, M.-H. Karatza, C. Roussos, and T. Vassilakopoulos Antioxidants Increase the Ventilatory Response to Hyperoxic Hypercapnia Am. J. Respir. Crit. Care Med., January 1, 2007; 175(1): 62 - 68. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Eynan, R. Arieli, and Y. Adir Response to CO2 in novice closed-circuit apparatus divers and after 1 year of active oxygen diving at shallow depths J Appl Physiol, May 1, 2005; 98(5): 1653 - 1659. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. W. Putnam, J. A. Filosa, and N. A. Ritucci Cellular mechanisms involved in CO2 and acid signaling in chemosensitive neurons Am J Physiol Cell Physiol, December 1, 2004; 287(6): C1493 - C1526. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. B. Dean, D. K. Mulkey, R. A. Henderson III, S. J. Potter, and R. W. Putnam Hyperoxia, reactive oxygen species, and hyperventilation: oxygen sensitivity of brain stem neurons J Appl Physiol, February 1, 2004; 96(2): 784 - 791. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. K. Mulkey, R. A. Henderson III, R. W. Putnam, and J. B. Dean Hyperbaric oxygen and chemical oxidants stimulate CO2/H+-sensitive neurons in rat brain stem slices J Appl Physiol, September 1, 2003; 95(3): 910 - 921. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. K. Mulkey, R. A. Henderson III, R. W. Putnam, and J. B. Dean Pressure ( [Abstract]
[Full Text]
[PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
) or 5%
CO2 in O2 (
) at 1-6 ATA
(
) near #1.
Measurements of PtiO2 (in vivo) in rats range from
<10 to 34 Torr (point b on y-axis) while air was breathed
(point a on the x-axis; i.e., normobaric normoxia) to
) and center (
, 150-µm depth) at normobaric
pressure and during various levels of hyperbaric O2
(HBO2) are also shown in the dark shaded area (#3; data are
replotted from Ref. 
FR, integrated firing rate; Imp/sec,
impulses/second.
