- Email: [email protected]
Development of chemosensitivity of rat medullary raphe neurons
Pergamon PII:
Neuroscience Vol. 90, No.3, pp. 1001–1011, 1999 Copyright 䉷 1999 IBRO. Published by Elsevier Science Ltd Printed in Great Britain. All rights reserved 0306-4522/99 $20.00+0.00 S0306-4522(98)00505-3
DEVELOPMENT OF CHEMOSENSITIVITY OF RAT MEDULLARY RAPHE NEURONS W. WANG* and G. B. RICHERSON†‡ *Yale University, New Haven, CT 06510, U.S.A. †Department of Neurology, Veteran’s Affairs Medical Center, West Haven, CT 06516, U.S.A.
Abstract—In many neonatal mammals, including humans and rats, there is a developmental increase in the ventilatory response to elevated pCO2. This maturation of central respiratory chemoreception may result from maturation of intrinsic chemosensitivity of brainstem neurons. We have examined age-related changes in chemosensitivity of neurons from the rat medullary raphe, a putative site for central chemoreception, using perforated patch-clamp recordings in vitro. In brain slices from rats younger than 12 days old, firing rate increased in 3% of neurons and decreased in 17% of neurons in response to respiratory acidosis (n 36). In contrast, in slices from rats 12 days and older, firing rate increased in 18% of neurons and decreased in 15% of neurons in response to the same stimulus (n 40). A tissue culture preparation of medullary raphe neurons was used to examine changes in chemosensitivity with age from three to 74 days in vitro. In cultured neurons younger than 12 days in vitro, firing rate increased in 4% of neurons and decreased in 44% of neurons in response to respiratory acidosis (n 54). In contrast, in neurons 12 days in vitro and older, firing rate increased in 30% of neurons and decreased in 24% of neurons in response to respiratory acidosis (n 105). In both types of chemosensitive neuron (“stimulated” and “inhibited”), the magnitudes of the changes in firing rate were greater in older neurons than in young neurons. These results indicate that the incidence and the degree of chemosensitivity of medullary raphe neurons increase with age in brain slices and in culture. This age-related increase in cellular chemosensitivity may underlie the development of respiratory chemoreception in vivo. Delays in this maturation process may contribute to developmental abnormalities of breathing, such as sudden infant death syndrome. 䉷 1999 IBRO. Published by Elsevier Science Ltd. Key words: respiration, medulla, chemoreception, tissue culture, brain slice, sudden infant death syndrome.
At the time of birth, control of breathing is sufficiently well developed in mammals to maintain homeostasis. However, in some species the ventilatory response to a CO2 challenge increases in the postnatal period. For example, sensitivity of the respiratory system to CO2 continues to increase after birth in premature and in term human infants. 11,27,44 A postnatal increase in chemosensitivity has also been described in premature monkeys, 13 rats 1 and dogs. 35 The postnatal development of chemoreception appears to be, in part, a result of an increase in central chemosensitivity. This is suggested by the fact that homeostasis is highly dependent on peripheral chemoreceptors in some neonatal mammals, 3,8,18 and peripheral chemoreceptors are generally regarded as essential for the neonate.28 This dependence on peripheral chemoreceptors indicates ‡To whom correspondence should be addressed. Abbreviations: AHP, afterhyperpolarization; AP-5, ( ^ )-2amino-5-phosphonopentanoic acid; CI, chemosensitivity index; CNQX, 6-cyano-7-nitroquinoxaline-2,3-dione; DIV, days in vitro; EGTA, ethyleneglycolbis(b-aminoethyl ether)-N,N,N 0 ,N 0 -tetra-acetate; FBS, fetal bovine serum; HEPES, N-2-hydroxyethylpiperazine-N 0 -2-ethanesulfonic acid; ISI, interspike interval; MEM, modified Eagle’s medium; P, postnatal day.
that central chemoreception is not fully developed in the fetus and undergoes maturation during the neonatal period. 22,23 This appears to be particularly true in altricial species such as rats and humans, which are relatively immature at birth. In those species in which postnatal maturation of the hypercarbic ventilatory response occurs, virtually nothing is known about the cellular mechanisms of development of central chemoreception. The most likely explanation for the increase in CO2 sensitivity of the system as a whole is that the individual neurons responsible for chemotransduction undergo maturation of their cellular properties. Thus, examining development of the response of individual neurons to respiratory acidosis is likely to be a fruitful approach in defining these mechanisms. The medullary raphe is a putative site for central chemoreceptors. We have shown previously that rat medullary raphe neurons are chemosensitive in brain slices 41 and in tissue culture. 49 During these studies, we were interested in using brain slices from young [postnatal day (P) 1–8] rats, because: (i) they are resistant to hypoxic/ischemic and traumatic damage during slice preparation; (ii) it is easier to visualize neurons prior to development of myelin; and (iii) recordings are generally more stable than in adult
1001
1002
W. Wang and G. B. Richerson
tissue. However, we observed that, in the medullary raphe of young slices, most neurons were not chemosensitive. This observation led us to examine the relationship between age and chemosensitivity in rat medullary raphe neurons. Here, we used recordings from brain slices and in tissue culture, 49 and investigated the chemosensitivity of raphe neurons with increasing age during the first few postnatal weeks. EXPERIMENTAL PROCEDURES
Materials Picrotoxin, EGTA, HEPES, cytosine b-d-arabinofuranoside hydrochloride, papain, cysteine, trypsin inhibitor, bovine serum albumin, poly-l-ornithine, laminin, streptomycin, penicillin, fibroblast growth factor-5, and all salts and chemicals not otherwise listed were purchased from Sigma Chemical Co. (St Louis, MO). (^)-2-Amino-5-phosphonopentanoic acid (AP-5) and 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX) were purchased from Research Biochemicals Incorporated (Natick, MA). Culture media, including fetal bovine serum (FBS), basic fibroblast growth factor, Neurobasal medium and B27 supplement, were purchased from Gibco BRL Products (Gaithersburg, MD). Modified Eagle’s medium (MEM; No. 56419) and F12 supplement were purchased from JRH Biosciences (Lenexa, KS). Brain slices Brain slices were prepared using methods described previously. 41,43 Briefly, Sprague–Dawley rats (Charles River Labs, Wilmington, MA) were anesthetized with chloral hydrate (1 g/kg, i.p.), then perfused through the aorta with 50 ml of ice-cold neuroprotective solution (in mM: choline chloride 135, KCl 1, CaCl2 0.5, MgCl2 20, choline bicarbonate 20, NaH2PO4 1.2, dextrose 10, kynurenic acid 1; pH 7.4 at 5% CO2) to cool the brain and decrease extracellular NaCl and CaCl2 concentrations. The medulla was rapidly removed, and thin (100 mm) transverse slices were cut on a Vibratome in the same neuroprotective solution. Slices were placed in a recording chamber on a fixedstage upright microscope (Zeiss Axioskop FS). Recordings from neurons in brain slices were included in analysis only if they were located within the medullary raphe. Localization was determined using visualization at high power ( × 400) with Nomarski DIC optics, and at low power ( × 25). The location of the medullary raphe (raphe pallidus, raphe magnus and raphe obscurus) was determined using the rat brain atlas of Paxinos and Watson. 38 Cell culture Neurons of the ventromedial medulla were dissociated and maintained in primary tissue culture using a protocol described previously. 49 Briefly, neonatal Sprague–Dawley rats (P1–P3) were decapitated and the medulla was removed using an aseptic technique. All efforts were made to minimize animal suffering and to reduce the number of animals used. The medullary raphe and immediately surrounding tissue was isolated under a dissecting microscope. Dissected tissue was digested, triturated and plated on poly-lornithine- and laminin-coated 12-mm round coverslips at a density of 0.5–1 × 10 5 cells per ml. Neurons were fed with glial-conditioned medium (10% FBS in MEM, or 10% FBS in 60% MEM ⫹ 40% Neurobasal/B27; with penicillin/ streptomycin; conditioned for one day by glial cultures obtained from the ventromedial medulla). In most cases, basic fibroblast growth factor (0.1–1 ng/ml) and/or
fibroblast growth factor-5 (1–10 ng/ml) were added to the culture medium to enhance survival. 31 Cells were first fed on days 4–7 with culture medium to which cytosine b-darabinofuranoside (3 mM) was added to inhibit glial growth, and then fed approximately once per week. Recordings were made after cells were grown in culture for three to 74 days. Electrophysiological recordings Neurons were maintained at room temperature and continuously superfused at a rate of 3–4 ml/min with oxygenated Ringer solution (in mM: NaCl 124, KCl 3, CaCl2 2, MgCl2 2, NaHCO3 26, NaH2PO4 1.3, dextrose 10; pH 7.4 at 5% CO2). For recordings in tissue culture, antagonists of ionotropic GABAergic and glutamatergic receptors were included in the Ringer solution (100 mM picrotoxin, 50 mM AP-5 and 10 mM CNQX). Acid/base changes were made by altering pCO2, while maintaining constant NaHCO3 concentration (i.e. “respiratory” acid/base change). 41,49 Control conditions were always at 5% CO2 and pH 7.4. Bath pH was continuously measured with a pH electrode (MI-414; Microelectrodes, Inc., Londonderry, NH) in the inflow to the recording chamber. All recordings in both brain slices and cell culture were made using the amphotericin perforated-patch technique, 40 except as noted. Electrodes (4–10 MV; borosilicate glass, Corning 7052) were filled with intracellular solution containing (in mM): potassium methanesulfonate 135, KCl 10, HEPES 5 and EGTA 1 (pH 7.2; osmolarity 275 ^ 10 mOsm). Neurons were considered healthy with resting potential ⱕ⫺45 mV and action potential height ⱖ 60 mV. Recordings were amplified (Axopatch 1D, Axon Instruments, Foster City, CA), filtered (10 kHz low-pass) and acquired at 10 kilosamples per second with a computerized data acquisition system (AT-MIO-16F-5, National Instruments, Austin, TX) using custom-written software. Data were stored simultaneously on tape (Neurocorder DR-484, Neuro Data Inst. Corp., New York, NY) and on magnetic storage media. All recordings were made in current-clamp mode. In brain slices, all recordings were made without current injection. In culture, if a cell did not have spontaneous firing, it was depolarized to induce firing to near 1 Hz, and changes in firing rate were measured in response to changes in CO2/pH. In culture, the mean firing rate of neurons that fired spontaneously was 1.67 ^ 1.52 (n 72), while the mean firing rate of neurons depolarized to induce firing was 1.46 ^ 1.23 (n 43) after depolarization. Since these values were comparable, we have reported the results together in subsequent analyses. If a neuron required continuous hyperpolarization to decrease firing rate or stabilize membrane potential, that neuron was not used to study chemosensitivity. Membrane capacitance and input resistance were calculated from the current induced by a ⫺10 mV voltage-clamp pulse from a holding potential of ⫺70 mV. Mean membrane potential, spike threshold, afterhyperpolarization (AHP) level and firing rate were measured during the first 5 min of stable recording. The relative interspike interval (relative ISI) was used as a measure of the regularity of firing. 49 For the entire period of stable recording for each neuron, the relative ISI was calculated as the ratio of the ISI before each spike to the ISI after that spike. The standard deviation of the relative ISI (S.D. of relative ISI) was then calculated from all spikes for each neuron. Neurons with a monotonously regular firing pattern are typical of serotonergic neurons, 32 and have a low S.D. of relative ISI. Definition and analysis of chemosensitivity Recordings were analysed using methods described previously. 41,49 Neurons were defined as chemosensitive if they demonstrated changes in firing rate in response to
Development of chemosensitivity in the raphe
1003
Table 1. Electrophysiological properties of medullary raphe neurons in brain slices: changes with age of rat Age (days)
Capacitance (pF)
Input resistance (MV)
Membrane potential (mV)
Firing rate (Hz)
Age ⬍ 12 days in vitro Mean 9.4 S.D. 1.6 n 41
69.9 54.6 11
1246 841 11
⫺ 55.5 6.3 33
2.2 1.6 30
Age ⱖ 12 days in vitro Mean 14 S.D. 2.3 n 45
96.4* 58.2 15
2097 1911 19
⫺ 58.2 8.6 33
2.3 1.9 31
*P ⬍ 0.05.
changes in CO2 that were: (i) reproducible and reversible on two or more exposures to acid/base changes (i.e. four transitions between pH levels); (ii) consistent in time-course for each stimulus; (iii) if exposed to both acidosis and alkalosis, the response was opposite in sign; and (iv) there was a statistically significant change in firing rate of at least 20% per change in pH of 0.2 for at least four transitions. To compare the degree of chemosensitivity of different neurons, the chemosensitivity index (CI) was calculated as the average firing rate (as percentage of control) induced by a decrease in pH of 0.2 units: CI 100% × 10logFR9 =FR3 0:2= pH3 ⫺pH9 ; where FR9 mean firing rate (percentage of control) at 9% CO2, FR3 mean firing rate (percentage of control) at 3% CO2, pH9 mean pH at 9% CO2 and pH3 mean pH at 3% CO2. The CI, which has been described previously, 49 provides a normalized value for the degree of chemosensitivity. For example, a neuron that increased its firing rate to 200% of control when pH decreased from 7.4 to 7.2 would have a CI of 200%, while one that increased its firing rate to 200% of control when pH increased from 7.4 to 7.6 would have a CI of 50%. The CI was calculated for each transition, then the mean CI for each neuron was calculated as the average of all transitions. All data presented in the form x ^ y are mean ^ S.D. unless noted otherwise. Statistical significance was tested using the non-parametric Mann–Whitney–Wilcoxon test, unless specified otherwise.
RESULTS
Development of electrophysiological properties of medullary raphe neurons Recordings from brain slices were divided into two age groups: ⬍12 days old and ⱖ12 days old. Within these two groups, the basic cellular electrophysiology of medullary raphe neurons was similar (Table 1). There was a significant increase in whole-cell membrane capacitance in older neurons (P ⬍ 0.05), consistent with an increase in cell size. There was no difference in input resistance (P ⬎ 0.1), mean resting potential (P ⬎ 0.1) or mean firing rate (P ⬎ 0.1). The mean age of the younger group of neurons was 9.4 days, and none of the recordings were from slices less than five days old. Thus, it is possible that neurons in brain slices
from animals younger than five days old would have shown developmental differences. Development of the basic electrophysiological properties of neurons was also studied in tissue culture. Using the whole-cell patch-clamp recording mode, 15 the cell capacitance and input resistance of neurons grown for 3 ^ 1 days in vitro (DIV; n 18) were compared to those grown for 13 ^ 2 DIV (n 18). There was an increase in cell capacitance from 54 ^ 20 to 100 ^ 35 pF (P ⬍ 0.025) and a decrease in input resistance from 745 ^ 375 to 485 ^ 241 MV (P ⬍ 0.025) in older neurons. This was consistent with an increase in size and neurite extension with maturation. Basic electrophysiological properties were then compared in neurons grown in culture for ⬇6.5 DIV with those grown in culture for ⬇21.5 DIV, and there were no differences in membrane potential, AHP level or spike threshold (Table 2). Older neurons tended to have a slightly lower firing rate and a more regular firing pattern, but these differences were not statistically significant. Older neurons were also slightly more likely to not spontaneously fire (45%) compared to younger neurons (27%). When neurons were separated into those inhibited by CO2 and those with no response to CO2, there were also no age-related differences in any of the parameters except capacitance and input resistance within each class. When each of these parameters was plotted vs age, rather than dividing neurons into two large age groups, there were still no differences seen with age in culture. Most recordings were made from neurons grown in culture for more than four days, and no recordings were made prior to 3 DIV. Thus, there may have been changes that occurred prior to this age. These results indicate that the basic electrophysiological characteristics of medullary raphe neurons were established by at least 3–5 DIV. Examples of two neurons grown in culture for four days are shown in Fig. 1, which display the two basic types of firing pattern described previously in neurons cultured from the medullary raphe. 49 Based solely on firing pattern, firing rate, membrane
1004
W. Wang and G. B. Richerson Table 2. Electrophysiological properties of rat medullary raphe neurons in tissue culture: changes with age in culture Membrane potential (mV)
Spike threshold (mV)
AHP level (mV)
Firing rate (Hz)
S.D. of relative ISI
Age ⬍ 12 days in vitro Mean ⫺ 52.1 S.D. 6.5 n 48
⫺ 42.3 5.0 48
⫺ 57.5 6.4 48
1.18 1.33 45
1.15 0.5 21
Age ⱖ 12 days in vitro Mean ⫺ 51.1 S.D. 7.5 n 101
⫺ 42.3 4.7 101
⫺ 56.6 6.6 101
0.74 1.46 77
0.99 0.58 37
AHP level, minimum afterhyperpolarization potential; S.D. of relative ISI, standard deviation of interspike interval. Firing rate was mean firing rate without current injection, in Ringer solution with AP-5, CNQX and picrotoxin. Mean ages: 6.9 ^ 2.4 DIV (n 49) vs 21.5 ^ 9.8 DIV (n 101) for membrane potential, spike threshold and AHP; 6.5 ^ 2.2 DIV (n 46) vs 22.2 ^ 10.9 DIV (n 78) for firing rate; 6.6 ^ 2.6 DIV (n 21) vs 21.4 ^ 11.8 DIV (n 37) for S.D. of relative ISI.
Fig. 1. Baseline activity of two types of cultured neurons. The basic electrophysiological properties of these neonatal neurons were indistinguishable from those of more mature neurons in culture. 49 (A) Young neuron (4 DIV) with a highly regular firing pattern. The S.D. of the relative ISI in this neuron was 0.45. (B) Young neuron (4 DIV) with an irregular firing pattern. The S.D. of the relative ISI in this neuron was 1.89.
potential or other basic electrophysiological properties, it was not possible to distinguish these recordings from those of medullary raphe neurons
grown in culture for more than two weeks, 49 in brain slices prepared from two-week-old animals 41 or from adult animals in vivo. 32
Development of chemosensitivity in the raphe
1005
Fig. 2. The percentage of raphe neurons that were stimulated by respiratory acidosis increased with age in brain slices and in tissue culture. (A) Brain slices: percentage of neurons stimulated, inhibited or with no response to respiratory acidosis (n 36 for cells ⬍12 days old; n 40 for cells ⭓12 days old). (B) Tissue culture: percentage of neurons with each type of response to respiratory acidosis (n 54 for cells ⬍12 DIV; n 105 for cells ⭓12 DIV). *P ⬍ 0.05, Chi-square test.
Development of chemosensitivity in brain slices In the medullary raphe of brain slices, during the first two weeks of age there was an increase in the percentage of neurons that were stimulated by respiratory acidosis. In brain slices prepared from animals younger than 12 days old, only 3% of neurons (1/36) were stimulated by respiratory acidosis. In contrast, in slices from animals 12 days and older, 18% (7/40) of neurons were stimulated (P ⬍ 0.05, Chi-square test; Fig. 2A). There were no differences in the percentage of neurons inhibited by respiratory acidosis in the two age groups.
The paucity of chemosensitive neurons in the medullary raphe of brain slices from young animals did not make it feasible to use animals less than 10 days old in our earlier study. 41 It is also difficult to use medullary brain slices from animals older than about 18 days for patch-clamp recordings from visualized neurons (see Introduction). Thus, recordings from brain slices were limited to a relatively narrow time window, making it difficult to completely define developmental changes in the degree of chemosensitivity of medullary raphe neurons in slices. We therefore used tissue culture to further examine maturation of chemosensitivity.
1006
W. Wang and G. B. Richerson
Fig. 3. Maturation of the response to CO2-induced acid/base changes in stimulated medullary raphe neurons in tissue culture. Firing rate vs time as CO2 was altered to produce acidosis or alkalosis is shown in both panels. (A) Young stimulated neuron (7 DIV). Note that the firing rate increased each time CO2 increased, but it was not obvious that the changes in firing rate were caused by the CO2 changes, rather than random variations. (B) Older stimulated neuron (13 DIV). The response in this neuron was more robust, and was clearly due to the stimulus. A simultaneous recording of pH is also shown in B.
Development of chemosensitivity in tissue culture In tissue culture, as in brain slices, older neurons were more likely to be stimulated by respiratory acidosis (Fig. 2B). Only 4% (2/54) of neurons cultured for less than 12 days were stimulated, whereas 30% (31/105) of neurons cultured for 12 or more days were stimulated (P ⬍ 0.005, Chisquare test). There was a lower percentage of inhibited neurons in the older age group, but this difference was not statistically significant. In brain slices, the percentage of neurons that were unresponsive to respiratory acidosis was greater than in tissue culture. This may have been due, in part, to the difficulty in recording from neurons
older than 18 days in brain slices. Alternatively, regrowth of dendrites in culture may have resulted in an increase in chemosensitivity, or ischemic damage 43 may have blunted the response in brain slices. Although two of 54 young neurons in culture met the criteria for a stimulated neuron, their responses were not as obvious as most older neurons. An example of one of these young stimulated neurons is shown in Fig. 3A. In response to three periods of exposure to an increase in CO2 to 9%, the firing rate of this neuron (7 DIV) increased each time. An example of an older stimulated neuron is shown for comparison (Fig. 3B). The response of the older neuron was more robust and clearly evident.
Development of chemosensitivity in the raphe
1007
Fig. 4. Relationship between chemosensitivity and age in culture. (A) Histograms of CI for neurons separated into two age groups. Black bars: neurons younger than 12 DIV. Shaded bars: neurons 12 DIV and older. Note that most neurons ⬍ 12 DIV had a CI between 60% and 120%. In contrast, a relatively large percentage of neurons ⱖ 12 DIV had a CI that was below 40% or above 120%, indicating that there was a greater incidence of neurons with either a stimulated or a strong inhibited response to acidosis. (B) Mean CI vs age for inhibited neurons with a CI ⬍ 80%. Note that, with increasing age, the mean CI became smaller, indicating a greater degree of inhibition by acidosis. (C) Mean CI vs age for stimulated neurons with a CI ⬎ 120%. Note that, with increasing age, the mean CI became larger, indicating a greater degree of stimulation by acidosis. Numbers above points in B and C are numbers of neurons for each data point. Error bars indicate ^ S.E.M.
Such a clear and reproducible response is common in older cultured medullary raphe neurons. 49 One of the advantages of using tissue culture for studying chemosensitive neurons is the enhanced ability to quantify responses. We previously developed an approach to quantify the degree of chemosensitivity of neurons in tissue culture. 49 In the current study, we applied this approach to study developmental changes in chemosensitivity. The CI was calculated for 122 neurons. With increasing age, there was an increase in the number of cells
with a CI ⬎ 120%, consistent with the observed increase in the percentage of neurons classified as stimulated. Interestingly, there was also an increase in the number of neurons with a CI ⬍ 30%. Five percent (1/19) of young (⬍12 DIV) inhibited neurons had a CI less than 30%, whereas 50% (11/22) of older (ⱖ12 DIV) inhibited neurons had a CI less than 30% (P ⬍ 0.005, Chi-square test). The mean CI for all young (⬍12 DIV) inhibited neurons was 61 ^ 20%, whereas that for all older (ⱖ12 DIV) inhibited neurons was 33 ^ 21%.
1008
W. Wang and G. B. Richerson
To further compare the distribution of the CI for neurons of different ages, histograms were plotted of the CI for neurons within two age groups (Fig. 4A). Among neurons 12 DIV and older, there was a large percentage of neurons with a CI greater than 200% or less than 40%. The difference in distribution of CI in young neurons and older neurons was statistically significant (P ⬍ 0.05, Chi-square test). Thus, although our classification of neurons revealed no difference in the percentage of inhibited neurons in older neurons, there was an increase in the degree of chemosensitivity of those neurons that were inhibited. Development of chemosensitivity did not occur in a stepwise fashion at the age of 12 DIV, but instead the degree of chemosensitivity of neurons increased progressively over the age range studied. The relationship between the CI and age was determined for inhibited neurons (Fig. 4B) and stimulated neurons (Fig. 4C). With an increase in age above 12 DIV, there was a decrease in the mean CI for inhibited neurons (P ⬍ 0.01, ANOVA) and a trend toward an increase in the mean CI for stimulated neurons (not significant, ANOVA), consistent with an increase in the degree of chemosensitivity of both types of neuron.
DISCUSSION
Development of basic electrophysiology in vivo Neurons of the medullary raphe have a stereotypical firing behavior that has been observed in tissue culture, 49 in brain slices 41,47 and in vivo. 20,32 The developmental time-course for this firing behavior has not been defined, but the adult electrophysiological phenotype is apparently present at a very young age. In the present study, we found that there were no major age-related changes in brain slices or in culture. Thus, it appeared that the basic firing properties which distinguish raphe neurons, a highly regular firing pattern, prominent AHP and ramp depolarization between spikes, were nearly fully developed by P3–P5. The current study does not provide evidence for when raphe neurons first begin to produce their characteristic firing behavior. It is interesting that these properties were present shortly after birth, suggesting that they develop in utero. This may indicate that normal activity of raphe neurons is important for homeostasis at the time of birth, consistent with the proposed role of raphe neurons in regulation of a variety of autonomic functions in vivo. 21 The anatomical development of the medullary raphe has been studied previously. 20 However, little is currently known about prenatal development of the basic electrophysiological properties of raphe neurons. Thus, it will be important in the future to define the developmental sequence of the electrophysiology of raphe neurons.
Development of respiratory chemoreception in vivo It is not known when the central chemoreceptors first begin to function during development. Early work suggested that there was no central response to CO2 in the fetus, 17 but this work was performed in anesthetized and/or exteriorized animals. In later experiments performed in unanesthetized, chronically instrumented fetal lambs between 126 and 136 days gestation, varying HCO⫺ 3 concentration of the cerebrospinal fluid via ventriculocisternal perfusion altered the incidence of fetal breathing movements. 19 Thus, central chemoreceptors are functional even in the fetal lamb, apparently developing in utero. Although central chemoreception exists early in development, the sensitivity of the respiratory system to CO2 continues to increase after birth in some species, 1,13,35 including man. 11,27,44 It is not clear how much of this postnatal development of chemosensitivity is a result of an increase in the central vs peripheral chemosensitivity. However, in general, the central chemoreceptors are more important for the response to changes in CO2, while the response to hypoxia is dominated by the peripheral chemoreceptors. The peripheral chemoreceptors are thought to be inactive in the fetus, but undergo rapid resetting immediately after delivery. 22 In kittens, robust peripheral chemoreceptor responses to CO2 occur shortly after birth, at a time that responses to hypoxia are less than those of the adult. 50 However, in this species, development of the peripheral response to CO2 during hypoxia continues for more than eight weeks, which is longer than development of the response to hypoxia itself. 4 After denervation of peripheral chemoreceptors in three- to nine-day-old piglets, ventilation decreases, arterial pCO2 rises and mortality rate is increased. 8 Similar results are seen in two-day-old lambs. 3 In rats younger than 21 days old, peripheral chemoreceptor denervation results in an abnormal respiratory pattern and mortality rates near 50%. 18 Not all species show this dependence on peripheral chemoreceptors in the neonatal period. 24 However, the dependence of homeostasis on peripheral chemoreceptors in some neonatal mammals, and the postnatal increase in the ventilatory response to hypercapnia, suggest that the central chemoreceptor response undergoes maturation during the neonatal period in some mammals, including rats and humans. If raphe neurons are respiratory chemoreceptors, and they do not develop the cellular property of chemosensitivity until they are older than 12 days postnatal, then it would be expected that there would be a developmental increase in the respiratory response to CO2 over the same age range in the same strain of rats. This has been found to be the case. There is little response in ventilation to inspiration of 5% CO2 in Sprague–Dawley rats of three and
Development of chemosensitivity in the raphe
eight days of age, but ventilation approximately doubles at 19 and 34 days. 1 This similar dependence on age does not prove a causal link, but it is consistent with the hypothesis that development of chemosensitivity of raphe neurons plays a key role in development of the systems level response to CO2. Maturation of cellular chemosensitivity in vitro The cellular mechanisms of chemoreception have not been defined in adult mammals, making it difficult to directly study the mechanisms of maturation of the systems level response to hypercapnia. However, it is generally assumed that central chemoreception is due to intrinsic chemosensitivity of specialized neurons in brainstem nuclei linked to respiratory control. Neurons from a variety of brainstem regions are chemosensitive in adult brain slices, including the ventrolateral medulla, 12,25,37 nucleus of the solitary tract, 6 locus coeruleus, 39 hypothalamus 7 and medullary raphe. 41 It is unclear whether any of these neurons are involved in respiratory control, although each of them have been proposed as sites for central respiratory chemoreception. 2,5,16,30,33,36,39,45,46 Changes in chemosensitivity with age have not been described for neurons in any of these brainstem regions. When neurons from the medullary raphe are isolated in tissue culture, they display a high degree of chemosensitivity due to intrinsic cellular properties. 49 Here, we have used tissue culture to provide the first description at the cellular level of agerelated changes in chemosensitivity for any CNS region. These results indicated that, although chemosensitivity was present in neurons of the medullary raphe postnatally, the percentage of chemosensitive neurons and the degree of chemosensitivity both increased with age. The degree of chemosensitivity of stimulated and inhibited neurons both increased with age, with the largest change during the second postnatal week. These changes in chemosensitivity occurred over a time period during which the basic cellular properties of raphe neurons did not undergo significant agerelated changes other than an increase in capacitance and decrease in input resistance. This suggests that chemosensitivity is a cellular property that develops relatively late in these neurons, compared to other basic neuronal properties. Our results confirm our previous impression that brain slices from rats younger than 10 days of age, although technically superior for patch-clamp recording, are not a suitable model system for studying cellular mechanisms of chemosensitivity in adult rats. By analogy, other in vitro preparations using tissue from neonatal rats also may not demonstrate responses to CO2 that are directly comparable to those seen in adult tissue. Maturation of chemosensitivity was very similar in tissue culture and in brain slices, lending important
1009
validation of the dissociated cell culture preparation for studying chemosensitivity. Since recordings in slices are made acutely from neurons that have developed in situ, the developmental sequence of neurons in slices is likely to be an accurate reflection of what occurs in vivo. Thus, we believe that development of raphe neurons in culture is probably very similar to what occurs in vivo, and tissue culture will be an important and useful method for studying chemosensitivity of these neurons. However, despite the similarity in development of chemosensitivity in slices and culture, we have not yet established a oneto-one correlation between age in slices and days in vitro in culture, making it impossible to make a direct quantitative comparison between data obtained using the two preparations. Relationship between maturation of cellular chemosensitivity and development of respiratory chemoreception The relationship between maturation of cellular chemosensitivity in vitro and development of respiratory chemoreception in vivo is unclear. Within the medullary raphe, chemosensitivity at the cellular level in vitro has not been definitively linked with respiratory chemoreception in vivo. However, there is strong in vivo evidence to suggest that the medullary raphe is an important site for central chemoreception. Raphe neurons in cats increase their firing rate with hypercapnia. 48 Electrical stimulation of the raphe pallidus in cats 29 and acetazolamide microinjections in the raphe in rats 2 stimulate respiration. Chemical lesions of serotonergic neurons of the raphe of rats with 5,7dihydroxytryptamine result in an increase in baseline CO2 and a decrease in the response to hypercapnia, 34 and injections of lidocaine or ibotenic acid into the raphe lead to blunting of respiratory chemoreception. 9 Thus, it is likely that the increase in chemosensitivity of medullary raphe neurons with age contributes to the increase in central respiratory chemoreception that occurs postnatally. The similarity of the time-course of maturation of chemosensitivity in brain slices and in culture indicates that defining age-related changes of medullary raphe neurons in tissue culture is relevant to development of these neurons in vivo. The human homolog of the medullary raphe has been reported to be abnormal in a subset of infants who have died of sudden infant death syndrome, 26 and may also be involved in other developmental breathing disorders such as Ondine’s curse and Haddad syndrome. 10,14,42 Neurons of the medullary raphe are involved in cardiorespiratory control. Delayed development of these neurons, or toxicity due to exogenous influences such as tobacco smoke, could lead to inappropriate responses to an acid/base challenge in early life. Defining the cellular mechanisms of raphe neurons may help elucidate the
1010
W. Wang and G. B. Richerson
mechanisms of central cardiorespiratory control in adults, developmental changes in central chemoreception that occur over the neonatal period, and potential mechanisms of dysfunction of homeostatic reflexes induced by hypercapnia.
of chemosensitivity of these neurons over the first three postnatal weeks. This increase in chemosensitivity at the cellular level may underlie the increase in respiratory chemoreception in vivo that occurs over the same time period.
CONCLUSIONS
Neurons of the medullary raphe are candidates for central respiratory chemoreceptors. We have found that there is an increase in the incidence and degree
Acknowledgements—We wish to thank Gabriel Haddad for critical review of the manuscript. This work was supported by NIH R01HL52539 and the VAMC.
REFERENCES
1. Bamford O. S., Schuen J. N. and Carroll J. L. (1996) Effect of nicotine exposure on postnatal ventilatory responses to hypoxia and hypercapnia. Resp. Physiol. 106, 1–11. 2. Bernard D. G., Li A. and Nattie E. E. (1996) Evidence for central chemoreception in the midline raphe. J. appl. Physiol. 80, 108–115. 3. Bureau M. A., Lamarche J., Foulon P. and Dalle D. (1985) Postnatal maturation of respiration in intact and carotid bodychemodenervated lambs. J. appl. Physiol. 59, 869–874. 4. Carroll J. L., Bamford O. S. and Fitzgerald R. S. (1993) Postnatal maturation of carotid chemoreceptor responses to O2 and CO2 in the cat. J. appl. Physiol. 75, 2383–2391. 5. Coates E. L., Li A. and Nattie E. E. (1993) Widespread sites of brain stem ventilatory chemoreceptors. J. appl. Physiol. 75, 5– 14. 6. Dean J. B., Bayliss D. A., Erickson J. T., Lawing W. L. and Millhorn D. E. (1990) Depolarization and stimulation of neurons in nucleus tractus solitarii by carbon dioxide does not require chemical synaptic input. Neuroscience 36, 207–216. 7. Dillon G. H. and Waldrop T. G. (1992) In vitro responses of caudal hypothalamic neurons to hypoxia and hypercapnia. Neuroscience 51, 941–950. 8. Donnelly D. F. and Haddad G. G. (1990) Prolonged apnea and impaired survival in piglets after sinus and aortic nerve section. J. appl. Physiol. 68, 1048–1052. 9. Dreshaj I. A., Haxhiu M. A. and Martin R. J. (1998) Role of the medullary raphe nuclei in the respiratory response to CO2. Resp. Physiol. 111, 15–23. 10. Filiano J. J., Choi J. C. and Kinney H. C. (1990) Candidate cell populations for respiratory chemosensitive fields in the human infant medulla. J. comp. Neurol. 293, 448–465. 11. Frantz I. D., Adler S. M., Thach B. T. and Taeusch H. W. Jr (1976) Maturational effects on respiratory responses to carbon dioxide in premature infants. J. appl. Physiol. 41, 41–45. 12. Fukuda Y., See W. R. and Honda Y. (1980) H ⫹-sensitivity and pattern of discharge of neurons in the chemosensitive areas of the ventral medulla oblongata of rats in vitro. Pflu¨gers Arch. 388, 53–61. 13. Guthrie R. D., Standaert T. A., Hodson W. A. and Woodrum D. E. (1980) Sleep and maturation of eucapnic ventilation and CO2 sensitivity in the premature primate. J. appl. Physiol. 48, 347–354. 14. Haddad G. G., Mazza N. M., Defendini R., Blanc W. A., Driscoll J. M., Epstein M. A., Epstein R. A. and Mellins R. B. (1978) Congenital failure of automatic control of ventilation, gastrointestinal motility and heart rate. Medicine 57, 517–526. 15. Hamill O. P., Marty A., Neher E., Sakmann B. and Sigworth F. J. (1981) Improved patch-clamp techniques for high-resolution current recording from cells and cell-free membrane patches. Pflu¨gers Arch. 391, 85–100. 16. Haxhiu M. A., Erokwu B., Prabhakar N. R., Cherniack N. S. and Strohl K. P. (1992) Locus coeruleus neurons express c-fos immunoreactivity upon stimulation of central chemosensory system. Soc. Neurosci. Abstr. 18, 828. 17. Herrington R. T., Harned H. S. Jr, Ferreiro J. I. and Griffin C. A. (1971) The role of the central nervous system in perinatal respiration: studies of chemoregulatory mechanisms in the term lamb. Pediatrics 47, 857–864. 18. Hofer M. A. (1984) Lethal respiratory disturbance in neonatal rats after arterial chemoreceptor denervation. Life Sci. 34, 489– 496. 19. Hohimer A. R., Bissonnette J. M., Richardson B. S. and Machida C. M. (1983) Central chemical regulation of breathing movements in fetal lambs. Resp. Physiol. 52, 99–111. 20. Jacobs B. L. and Azmitia E. C. (1992) Structure and function of the brain serotonin system. Physiol. Rev. 72, 165–229. 21. Jacobs B. L. and Fornal C. A. (1991) Activity of brain serotonergic neurons in the behaving animal. Pharmac. Rev. 43, 563– 578. 22. Jansen A. H. and Chernick V. (1983) Development of respiratory control. Physiol. Rev. 63, 437–483. 23. Jansen A. H. and Chernick V. (1991) Fetal breathing and development of control of breathing. J. appl. Physiol. 70, 1431–1446. 24. Jansen A. H., Ioffe S., Russell B. J. and Chernick V. (1981) Effect of carotid chemoreceptor denervation on breathing in utero and after birth. J. appl. Physiol. 51, 630–633. 25. Jarolimek W., Misgeld U. and Lux H. D. (1990) Neurons sensitive to pH in slices of the rat ventral medulla oblongata. Pflu¨gers Arch. 416, 247–253. 26. Kinney H. C., Filiano J. J., Sleeper L. A., Mandell F., Valdes-Dapena M. and White W. F. (1995) Decreased muscarinic receptor binding in the arcuate nucleus in sudden infant death syndrome. Science 269, 1446–1450. 27. Krauss A. N., Klain D. B., Waldman S. and Auld P. A. M. (1975) Ventilatory response to carbon dioxide in newborn infants. Pediatric Res. 9, 46–50. 28. Lahiri S. (1994) Carotid body chemoreception: mechanisms and dynamic protection against apnea. Biol. Neonate 65, 134– 139. 29. Lalley P. M. (1986) Responses of phrenic motoneurones of the cat to stimulation of medullary raphe nuclei. J. Physiol., Lond. 380, 349–371. 30. Larnicol N., Wallois F., Berquin P., Gros F. and Rose D. (1994) c-fos-like immunoreactivity in the cat’s neuraxis following moderate hypoxia or hypercapnia. J. Physiol., Paris 88, 81–88.
Development of chemosensitivity in the raphe
1011
31. Lindholm D., Hartikka J., Berzaghi M. D., Castren E., Tzimagiorgis G., Hughes R. and Thoenen H. (1994) Fibroblast growth factor-5 promotes differentiation of cultured rat septal cholinergic and raphe serotonergic neurons—comparison with the effects of neurotrophins. Eur. J. Neurosci. 6, 244–252. 32. Mason P. (1997) Physiological identification of pontomedullary serotonergic neurons in the rat. J. Neurophysiol. 77, 1087– 1098. 33. Mitchell R. A., Loeschcke H. H., Massion W. H. and Severinghaus J. W. (1963) Respiratory responses mediated through superficial chemosensitive areas on the medulla. J. appl. Physiol. 18, 523–533. 34. Mueller R. A., Towle A. C. and Breese G. R. (1984) Supersensitivity to the respiratory stimulatory effect of TRH in 5,7dihydroxytryptamine-treated rats. Brain Res. 298, 370–373. 35. Nattie E. E. and Edwards W. H. (1981) CSF acid–base regulation and ventilation during acute hypercapnia in the newborn dog. J. appl. Physiol. 50, 566–574. 36. Nattie E. E. and Li A. H. (1996) Central chemoreception in the region of the ventral respiratory group in the rat. J. appl. Physiol. 81, 1987–1995. 37. Neubauer J. A., Gonsalves S. F., Chou W., Geller H. M. and Edelman N. H. (1991) Chemosensitivity of medullary neurons in explant tissue cultures. Neuroscience 45, 701–708. 38. Paxinos G. and Watson C. (1986) The Rat Brain in Stereotaxic Coordinates. Academic, New York. 39. Pineda J. and Aghajanian G. K. (1997) Carbon dioxide regulates the tonic activity of locus coeruleus neurons by modulating a proton- and polyamine-sensitive inward rectifier potassium current. Neuroscience 77, 723–743. 40. Rae J., Cooper K., Gates P. and Watsky M. (1991) Low access resistance perforated patch recordings using amphotericin B. J. Neurosci. Meth. 37, 15–26. 41. Richerson G. B. (1995) Response to CO2 of neurons in the rostral ventral medulla in vitro. J. Neurophysiol. 73, 933–944. 42. Richerson G. B. (1997) Sudden infant death syndrome: the role of central chemosensitivity. Neuroscientist 3, 3–7. 43. Richerson G. B. and Messer C. (1995) Effect of composition of experimental solutions on neuronal survival during rat brain slicing. Expl Neurol. 131, 133–143. 44. Rigatto H., Brady J. P. and de la Torre Verduzco R. (1975) Chemoreceptor reflexes in preterm infants: II. The effect of gestational and postnatal age on the ventilatory response to inhaled carbon dioxide. Pediatrics 55, 614–620. 45. Sato M., Severinghaus J. W. and Basbaum A. I. (1992) Medullary CO2 chemoreceptor neuron identification by c-fos immunocytochemistry. J. appl. Physiol. 73, 96–100. 46. Schlaefke M. E., See W. R. and Loeschcke H. H. (1970) Ventilatory response to alterations of H ⫹ ion concentration in small areas of the ventral medullary surface. Resp. Physiol. 10, 198–212. 47. Vandermaelen C. P. and Aghajanian G. K. (1983) Electrophysiological and pharmacological characterization of serotonergic dorsal raphe neurons recorded extracellularly and intracellularly in rat brain slices. Brain Res. 289, 109–119. 48. Veasey S. C., Fornal C. A., Metzler C. W. and Jacobs B. L. (1995) Response of serotonergic caudal raphe neurons in relation to specific motor activities in freely moving cats. J. Neurosci. 15, 5346–5359. 49. Wang W., Pizzonia J. H. and Richerson G. B. (1998) Chemosensitivity of rat medullary raphe neurones in primary tissue culture. J. Physiol., Lond. 511, 433–450. 50. Watanabe T., Kumar P. and Hanson M. A. (1996) Development of respiratory chemoreflexes to hypoxia and CO2 in unanaesthetized kittens. Resp. Physiol. 106, 247–254. (Accepted 26 August 1998)
Neuroscience Vol. 90, No.3, pp. 1001–1011, 1999 Copyright 䉷 1999 IBRO. Published by Elsevier Science Ltd Printed in Great Britain. All rights reserved 0306-4522/99 $20.00+0.00 S0306-4522(98)00505-3
DEVELOPMENT OF CHEMOSENSITIVITY OF RAT MEDULLARY RAPHE NEURONS W. WANG* and G. B. RICHERSON†‡ *Yale University, New Haven, CT 06510, U.S.A. †Department of Neurology, Veteran’s Affairs Medical Center, West Haven, CT 06516, U.S.A.
Abstract—In many neonatal mammals, including humans and rats, there is a developmental increase in the ventilatory response to elevated pCO2. This maturation of central respiratory chemoreception may result from maturation of intrinsic chemosensitivity of brainstem neurons. We have examined age-related changes in chemosensitivity of neurons from the rat medullary raphe, a putative site for central chemoreception, using perforated patch-clamp recordings in vitro. In brain slices from rats younger than 12 days old, firing rate increased in 3% of neurons and decreased in 17% of neurons in response to respiratory acidosis (n 36). In contrast, in slices from rats 12 days and older, firing rate increased in 18% of neurons and decreased in 15% of neurons in response to the same stimulus (n 40). A tissue culture preparation of medullary raphe neurons was used to examine changes in chemosensitivity with age from three to 74 days in vitro. In cultured neurons younger than 12 days in vitro, firing rate increased in 4% of neurons and decreased in 44% of neurons in response to respiratory acidosis (n 54). In contrast, in neurons 12 days in vitro and older, firing rate increased in 30% of neurons and decreased in 24% of neurons in response to respiratory acidosis (n 105). In both types of chemosensitive neuron (“stimulated” and “inhibited”), the magnitudes of the changes in firing rate were greater in older neurons than in young neurons. These results indicate that the incidence and the degree of chemosensitivity of medullary raphe neurons increase with age in brain slices and in culture. This age-related increase in cellular chemosensitivity may underlie the development of respiratory chemoreception in vivo. Delays in this maturation process may contribute to developmental abnormalities of breathing, such as sudden infant death syndrome. 䉷 1999 IBRO. Published by Elsevier Science Ltd. Key words: respiration, medulla, chemoreception, tissue culture, brain slice, sudden infant death syndrome.
At the time of birth, control of breathing is sufficiently well developed in mammals to maintain homeostasis. However, in some species the ventilatory response to a CO2 challenge increases in the postnatal period. For example, sensitivity of the respiratory system to CO2 continues to increase after birth in premature and in term human infants. 11,27,44 A postnatal increase in chemosensitivity has also been described in premature monkeys, 13 rats 1 and dogs. 35 The postnatal development of chemoreception appears to be, in part, a result of an increase in central chemosensitivity. This is suggested by the fact that homeostasis is highly dependent on peripheral chemoreceptors in some neonatal mammals, 3,8,18 and peripheral chemoreceptors are generally regarded as essential for the neonate.28 This dependence on peripheral chemoreceptors indicates ‡To whom correspondence should be addressed. Abbreviations: AHP, afterhyperpolarization; AP-5, ( ^ )-2amino-5-phosphonopentanoic acid; CI, chemosensitivity index; CNQX, 6-cyano-7-nitroquinoxaline-2,3-dione; DIV, days in vitro; EGTA, ethyleneglycolbis(b-aminoethyl ether)-N,N,N 0 ,N 0 -tetra-acetate; FBS, fetal bovine serum; HEPES, N-2-hydroxyethylpiperazine-N 0 -2-ethanesulfonic acid; ISI, interspike interval; MEM, modified Eagle’s medium; P, postnatal day.
that central chemoreception is not fully developed in the fetus and undergoes maturation during the neonatal period. 22,23 This appears to be particularly true in altricial species such as rats and humans, which are relatively immature at birth. In those species in which postnatal maturation of the hypercarbic ventilatory response occurs, virtually nothing is known about the cellular mechanisms of development of central chemoreception. The most likely explanation for the increase in CO2 sensitivity of the system as a whole is that the individual neurons responsible for chemotransduction undergo maturation of their cellular properties. Thus, examining development of the response of individual neurons to respiratory acidosis is likely to be a fruitful approach in defining these mechanisms. The medullary raphe is a putative site for central chemoreceptors. We have shown previously that rat medullary raphe neurons are chemosensitive in brain slices 41 and in tissue culture. 49 During these studies, we were interested in using brain slices from young [postnatal day (P) 1–8] rats, because: (i) they are resistant to hypoxic/ischemic and traumatic damage during slice preparation; (ii) it is easier to visualize neurons prior to development of myelin; and (iii) recordings are generally more stable than in adult
1001
1002
W. Wang and G. B. Richerson
tissue. However, we observed that, in the medullary raphe of young slices, most neurons were not chemosensitive. This observation led us to examine the relationship between age and chemosensitivity in rat medullary raphe neurons. Here, we used recordings from brain slices and in tissue culture, 49 and investigated the chemosensitivity of raphe neurons with increasing age during the first few postnatal weeks. EXPERIMENTAL PROCEDURES
Materials Picrotoxin, EGTA, HEPES, cytosine b-d-arabinofuranoside hydrochloride, papain, cysteine, trypsin inhibitor, bovine serum albumin, poly-l-ornithine, laminin, streptomycin, penicillin, fibroblast growth factor-5, and all salts and chemicals not otherwise listed were purchased from Sigma Chemical Co. (St Louis, MO). (^)-2-Amino-5-phosphonopentanoic acid (AP-5) and 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX) were purchased from Research Biochemicals Incorporated (Natick, MA). Culture media, including fetal bovine serum (FBS), basic fibroblast growth factor, Neurobasal medium and B27 supplement, were purchased from Gibco BRL Products (Gaithersburg, MD). Modified Eagle’s medium (MEM; No. 56419) and F12 supplement were purchased from JRH Biosciences (Lenexa, KS). Brain slices Brain slices were prepared using methods described previously. 41,43 Briefly, Sprague–Dawley rats (Charles River Labs, Wilmington, MA) were anesthetized with chloral hydrate (1 g/kg, i.p.), then perfused through the aorta with 50 ml of ice-cold neuroprotective solution (in mM: choline chloride 135, KCl 1, CaCl2 0.5, MgCl2 20, choline bicarbonate 20, NaH2PO4 1.2, dextrose 10, kynurenic acid 1; pH 7.4 at 5% CO2) to cool the brain and decrease extracellular NaCl and CaCl2 concentrations. The medulla was rapidly removed, and thin (100 mm) transverse slices were cut on a Vibratome in the same neuroprotective solution. Slices were placed in a recording chamber on a fixedstage upright microscope (Zeiss Axioskop FS). Recordings from neurons in brain slices were included in analysis only if they were located within the medullary raphe. Localization was determined using visualization at high power ( × 400) with Nomarski DIC optics, and at low power ( × 25). The location of the medullary raphe (raphe pallidus, raphe magnus and raphe obscurus) was determined using the rat brain atlas of Paxinos and Watson. 38 Cell culture Neurons of the ventromedial medulla were dissociated and maintained in primary tissue culture using a protocol described previously. 49 Briefly, neonatal Sprague–Dawley rats (P1–P3) were decapitated and the medulla was removed using an aseptic technique. All efforts were made to minimize animal suffering and to reduce the number of animals used. The medullary raphe and immediately surrounding tissue was isolated under a dissecting microscope. Dissected tissue was digested, triturated and plated on poly-lornithine- and laminin-coated 12-mm round coverslips at a density of 0.5–1 × 10 5 cells per ml. Neurons were fed with glial-conditioned medium (10% FBS in MEM, or 10% FBS in 60% MEM ⫹ 40% Neurobasal/B27; with penicillin/ streptomycin; conditioned for one day by glial cultures obtained from the ventromedial medulla). In most cases, basic fibroblast growth factor (0.1–1 ng/ml) and/or
fibroblast growth factor-5 (1–10 ng/ml) were added to the culture medium to enhance survival. 31 Cells were first fed on days 4–7 with culture medium to which cytosine b-darabinofuranoside (3 mM) was added to inhibit glial growth, and then fed approximately once per week. Recordings were made after cells were grown in culture for three to 74 days. Electrophysiological recordings Neurons were maintained at room temperature and continuously superfused at a rate of 3–4 ml/min with oxygenated Ringer solution (in mM: NaCl 124, KCl 3, CaCl2 2, MgCl2 2, NaHCO3 26, NaH2PO4 1.3, dextrose 10; pH 7.4 at 5% CO2). For recordings in tissue culture, antagonists of ionotropic GABAergic and glutamatergic receptors were included in the Ringer solution (100 mM picrotoxin, 50 mM AP-5 and 10 mM CNQX). Acid/base changes were made by altering pCO2, while maintaining constant NaHCO3 concentration (i.e. “respiratory” acid/base change). 41,49 Control conditions were always at 5% CO2 and pH 7.4. Bath pH was continuously measured with a pH electrode (MI-414; Microelectrodes, Inc., Londonderry, NH) in the inflow to the recording chamber. All recordings in both brain slices and cell culture were made using the amphotericin perforated-patch technique, 40 except as noted. Electrodes (4–10 MV; borosilicate glass, Corning 7052) were filled with intracellular solution containing (in mM): potassium methanesulfonate 135, KCl 10, HEPES 5 and EGTA 1 (pH 7.2; osmolarity 275 ^ 10 mOsm). Neurons were considered healthy with resting potential ⱕ⫺45 mV and action potential height ⱖ 60 mV. Recordings were amplified (Axopatch 1D, Axon Instruments, Foster City, CA), filtered (10 kHz low-pass) and acquired at 10 kilosamples per second with a computerized data acquisition system (AT-MIO-16F-5, National Instruments, Austin, TX) using custom-written software. Data were stored simultaneously on tape (Neurocorder DR-484, Neuro Data Inst. Corp., New York, NY) and on magnetic storage media. All recordings were made in current-clamp mode. In brain slices, all recordings were made without current injection. In culture, if a cell did not have spontaneous firing, it was depolarized to induce firing to near 1 Hz, and changes in firing rate were measured in response to changes in CO2/pH. In culture, the mean firing rate of neurons that fired spontaneously was 1.67 ^ 1.52 (n 72), while the mean firing rate of neurons depolarized to induce firing was 1.46 ^ 1.23 (n 43) after depolarization. Since these values were comparable, we have reported the results together in subsequent analyses. If a neuron required continuous hyperpolarization to decrease firing rate or stabilize membrane potential, that neuron was not used to study chemosensitivity. Membrane capacitance and input resistance were calculated from the current induced by a ⫺10 mV voltage-clamp pulse from a holding potential of ⫺70 mV. Mean membrane potential, spike threshold, afterhyperpolarization (AHP) level and firing rate were measured during the first 5 min of stable recording. The relative interspike interval (relative ISI) was used as a measure of the regularity of firing. 49 For the entire period of stable recording for each neuron, the relative ISI was calculated as the ratio of the ISI before each spike to the ISI after that spike. The standard deviation of the relative ISI (S.D. of relative ISI) was then calculated from all spikes for each neuron. Neurons with a monotonously regular firing pattern are typical of serotonergic neurons, 32 and have a low S.D. of relative ISI. Definition and analysis of chemosensitivity Recordings were analysed using methods described previously. 41,49 Neurons were defined as chemosensitive if they demonstrated changes in firing rate in response to
Development of chemosensitivity in the raphe
1003
Table 1. Electrophysiological properties of medullary raphe neurons in brain slices: changes with age of rat Age (days)
Capacitance (pF)
Input resistance (MV)
Membrane potential (mV)
Firing rate (Hz)
Age ⬍ 12 days in vitro Mean 9.4 S.D. 1.6 n 41
69.9 54.6 11
1246 841 11
⫺ 55.5 6.3 33
2.2 1.6 30
Age ⱖ 12 days in vitro Mean 14 S.D. 2.3 n 45
96.4* 58.2 15
2097 1911 19
⫺ 58.2 8.6 33
2.3 1.9 31
*P ⬍ 0.05.
changes in CO2 that were: (i) reproducible and reversible on two or more exposures to acid/base changes (i.e. four transitions between pH levels); (ii) consistent in time-course for each stimulus; (iii) if exposed to both acidosis and alkalosis, the response was opposite in sign; and (iv) there was a statistically significant change in firing rate of at least 20% per change in pH of 0.2 for at least four transitions. To compare the degree of chemosensitivity of different neurons, the chemosensitivity index (CI) was calculated as the average firing rate (as percentage of control) induced by a decrease in pH of 0.2 units: CI 100% × 10logFR9 =FR3 0:2= pH3 ⫺pH9 ; where FR9 mean firing rate (percentage of control) at 9% CO2, FR3 mean firing rate (percentage of control) at 3% CO2, pH9 mean pH at 9% CO2 and pH3 mean pH at 3% CO2. The CI, which has been described previously, 49 provides a normalized value for the degree of chemosensitivity. For example, a neuron that increased its firing rate to 200% of control when pH decreased from 7.4 to 7.2 would have a CI of 200%, while one that increased its firing rate to 200% of control when pH increased from 7.4 to 7.6 would have a CI of 50%. The CI was calculated for each transition, then the mean CI for each neuron was calculated as the average of all transitions. All data presented in the form x ^ y are mean ^ S.D. unless noted otherwise. Statistical significance was tested using the non-parametric Mann–Whitney–Wilcoxon test, unless specified otherwise.
RESULTS
Development of electrophysiological properties of medullary raphe neurons Recordings from brain slices were divided into two age groups: ⬍12 days old and ⱖ12 days old. Within these two groups, the basic cellular electrophysiology of medullary raphe neurons was similar (Table 1). There was a significant increase in whole-cell membrane capacitance in older neurons (P ⬍ 0.05), consistent with an increase in cell size. There was no difference in input resistance (P ⬎ 0.1), mean resting potential (P ⬎ 0.1) or mean firing rate (P ⬎ 0.1). The mean age of the younger group of neurons was 9.4 days, and none of the recordings were from slices less than five days old. Thus, it is possible that neurons in brain slices
from animals younger than five days old would have shown developmental differences. Development of the basic electrophysiological properties of neurons was also studied in tissue culture. Using the whole-cell patch-clamp recording mode, 15 the cell capacitance and input resistance of neurons grown for 3 ^ 1 days in vitro (DIV; n 18) were compared to those grown for 13 ^ 2 DIV (n 18). There was an increase in cell capacitance from 54 ^ 20 to 100 ^ 35 pF (P ⬍ 0.025) and a decrease in input resistance from 745 ^ 375 to 485 ^ 241 MV (P ⬍ 0.025) in older neurons. This was consistent with an increase in size and neurite extension with maturation. Basic electrophysiological properties were then compared in neurons grown in culture for ⬇6.5 DIV with those grown in culture for ⬇21.5 DIV, and there were no differences in membrane potential, AHP level or spike threshold (Table 2). Older neurons tended to have a slightly lower firing rate and a more regular firing pattern, but these differences were not statistically significant. Older neurons were also slightly more likely to not spontaneously fire (45%) compared to younger neurons (27%). When neurons were separated into those inhibited by CO2 and those with no response to CO2, there were also no age-related differences in any of the parameters except capacitance and input resistance within each class. When each of these parameters was plotted vs age, rather than dividing neurons into two large age groups, there were still no differences seen with age in culture. Most recordings were made from neurons grown in culture for more than four days, and no recordings were made prior to 3 DIV. Thus, there may have been changes that occurred prior to this age. These results indicate that the basic electrophysiological characteristics of medullary raphe neurons were established by at least 3–5 DIV. Examples of two neurons grown in culture for four days are shown in Fig. 1, which display the two basic types of firing pattern described previously in neurons cultured from the medullary raphe. 49 Based solely on firing pattern, firing rate, membrane
1004
W. Wang and G. B. Richerson Table 2. Electrophysiological properties of rat medullary raphe neurons in tissue culture: changes with age in culture Membrane potential (mV)
Spike threshold (mV)
AHP level (mV)
Firing rate (Hz)
S.D. of relative ISI
Age ⬍ 12 days in vitro Mean ⫺ 52.1 S.D. 6.5 n 48
⫺ 42.3 5.0 48
⫺ 57.5 6.4 48
1.18 1.33 45
1.15 0.5 21
Age ⱖ 12 days in vitro Mean ⫺ 51.1 S.D. 7.5 n 101
⫺ 42.3 4.7 101
⫺ 56.6 6.6 101
0.74 1.46 77
0.99 0.58 37
AHP level, minimum afterhyperpolarization potential; S.D. of relative ISI, standard deviation of interspike interval. Firing rate was mean firing rate without current injection, in Ringer solution with AP-5, CNQX and picrotoxin. Mean ages: 6.9 ^ 2.4 DIV (n 49) vs 21.5 ^ 9.8 DIV (n 101) for membrane potential, spike threshold and AHP; 6.5 ^ 2.2 DIV (n 46) vs 22.2 ^ 10.9 DIV (n 78) for firing rate; 6.6 ^ 2.6 DIV (n 21) vs 21.4 ^ 11.8 DIV (n 37) for S.D. of relative ISI.
Fig. 1. Baseline activity of two types of cultured neurons. The basic electrophysiological properties of these neonatal neurons were indistinguishable from those of more mature neurons in culture. 49 (A) Young neuron (4 DIV) with a highly regular firing pattern. The S.D. of the relative ISI in this neuron was 0.45. (B) Young neuron (4 DIV) with an irregular firing pattern. The S.D. of the relative ISI in this neuron was 1.89.
potential or other basic electrophysiological properties, it was not possible to distinguish these recordings from those of medullary raphe neurons
grown in culture for more than two weeks, 49 in brain slices prepared from two-week-old animals 41 or from adult animals in vivo. 32
Development of chemosensitivity in the raphe
1005
Fig. 2. The percentage of raphe neurons that were stimulated by respiratory acidosis increased with age in brain slices and in tissue culture. (A) Brain slices: percentage of neurons stimulated, inhibited or with no response to respiratory acidosis (n 36 for cells ⬍12 days old; n 40 for cells ⭓12 days old). (B) Tissue culture: percentage of neurons with each type of response to respiratory acidosis (n 54 for cells ⬍12 DIV; n 105 for cells ⭓12 DIV). *P ⬍ 0.05, Chi-square test.
Development of chemosensitivity in brain slices In the medullary raphe of brain slices, during the first two weeks of age there was an increase in the percentage of neurons that were stimulated by respiratory acidosis. In brain slices prepared from animals younger than 12 days old, only 3% of neurons (1/36) were stimulated by respiratory acidosis. In contrast, in slices from animals 12 days and older, 18% (7/40) of neurons were stimulated (P ⬍ 0.05, Chi-square test; Fig. 2A). There were no differences in the percentage of neurons inhibited by respiratory acidosis in the two age groups.
The paucity of chemosensitive neurons in the medullary raphe of brain slices from young animals did not make it feasible to use animals less than 10 days old in our earlier study. 41 It is also difficult to use medullary brain slices from animals older than about 18 days for patch-clamp recordings from visualized neurons (see Introduction). Thus, recordings from brain slices were limited to a relatively narrow time window, making it difficult to completely define developmental changes in the degree of chemosensitivity of medullary raphe neurons in slices. We therefore used tissue culture to further examine maturation of chemosensitivity.
1006
W. Wang and G. B. Richerson
Fig. 3. Maturation of the response to CO2-induced acid/base changes in stimulated medullary raphe neurons in tissue culture. Firing rate vs time as CO2 was altered to produce acidosis or alkalosis is shown in both panels. (A) Young stimulated neuron (7 DIV). Note that the firing rate increased each time CO2 increased, but it was not obvious that the changes in firing rate were caused by the CO2 changes, rather than random variations. (B) Older stimulated neuron (13 DIV). The response in this neuron was more robust, and was clearly due to the stimulus. A simultaneous recording of pH is also shown in B.
Development of chemosensitivity in tissue culture In tissue culture, as in brain slices, older neurons were more likely to be stimulated by respiratory acidosis (Fig. 2B). Only 4% (2/54) of neurons cultured for less than 12 days were stimulated, whereas 30% (31/105) of neurons cultured for 12 or more days were stimulated (P ⬍ 0.005, Chisquare test). There was a lower percentage of inhibited neurons in the older age group, but this difference was not statistically significant. In brain slices, the percentage of neurons that were unresponsive to respiratory acidosis was greater than in tissue culture. This may have been due, in part, to the difficulty in recording from neurons
older than 18 days in brain slices. Alternatively, regrowth of dendrites in culture may have resulted in an increase in chemosensitivity, or ischemic damage 43 may have blunted the response in brain slices. Although two of 54 young neurons in culture met the criteria for a stimulated neuron, their responses were not as obvious as most older neurons. An example of one of these young stimulated neurons is shown in Fig. 3A. In response to three periods of exposure to an increase in CO2 to 9%, the firing rate of this neuron (7 DIV) increased each time. An example of an older stimulated neuron is shown for comparison (Fig. 3B). The response of the older neuron was more robust and clearly evident.
Development of chemosensitivity in the raphe
1007
Fig. 4. Relationship between chemosensitivity and age in culture. (A) Histograms of CI for neurons separated into two age groups. Black bars: neurons younger than 12 DIV. Shaded bars: neurons 12 DIV and older. Note that most neurons ⬍ 12 DIV had a CI between 60% and 120%. In contrast, a relatively large percentage of neurons ⱖ 12 DIV had a CI that was below 40% or above 120%, indicating that there was a greater incidence of neurons with either a stimulated or a strong inhibited response to acidosis. (B) Mean CI vs age for inhibited neurons with a CI ⬍ 80%. Note that, with increasing age, the mean CI became smaller, indicating a greater degree of inhibition by acidosis. (C) Mean CI vs age for stimulated neurons with a CI ⬎ 120%. Note that, with increasing age, the mean CI became larger, indicating a greater degree of stimulation by acidosis. Numbers above points in B and C are numbers of neurons for each data point. Error bars indicate ^ S.E.M.
Such a clear and reproducible response is common in older cultured medullary raphe neurons. 49 One of the advantages of using tissue culture for studying chemosensitive neurons is the enhanced ability to quantify responses. We previously developed an approach to quantify the degree of chemosensitivity of neurons in tissue culture. 49 In the current study, we applied this approach to study developmental changes in chemosensitivity. The CI was calculated for 122 neurons. With increasing age, there was an increase in the number of cells
with a CI ⬎ 120%, consistent with the observed increase in the percentage of neurons classified as stimulated. Interestingly, there was also an increase in the number of neurons with a CI ⬍ 30%. Five percent (1/19) of young (⬍12 DIV) inhibited neurons had a CI less than 30%, whereas 50% (11/22) of older (ⱖ12 DIV) inhibited neurons had a CI less than 30% (P ⬍ 0.005, Chi-square test). The mean CI for all young (⬍12 DIV) inhibited neurons was 61 ^ 20%, whereas that for all older (ⱖ12 DIV) inhibited neurons was 33 ^ 21%.
1008
W. Wang and G. B. Richerson
To further compare the distribution of the CI for neurons of different ages, histograms were plotted of the CI for neurons within two age groups (Fig. 4A). Among neurons 12 DIV and older, there was a large percentage of neurons with a CI greater than 200% or less than 40%. The difference in distribution of CI in young neurons and older neurons was statistically significant (P ⬍ 0.05, Chi-square test). Thus, although our classification of neurons revealed no difference in the percentage of inhibited neurons in older neurons, there was an increase in the degree of chemosensitivity of those neurons that were inhibited. Development of chemosensitivity did not occur in a stepwise fashion at the age of 12 DIV, but instead the degree of chemosensitivity of neurons increased progressively over the age range studied. The relationship between the CI and age was determined for inhibited neurons (Fig. 4B) and stimulated neurons (Fig. 4C). With an increase in age above 12 DIV, there was a decrease in the mean CI for inhibited neurons (P ⬍ 0.01, ANOVA) and a trend toward an increase in the mean CI for stimulated neurons (not significant, ANOVA), consistent with an increase in the degree of chemosensitivity of both types of neuron.
DISCUSSION
Development of basic electrophysiology in vivo Neurons of the medullary raphe have a stereotypical firing behavior that has been observed in tissue culture, 49 in brain slices 41,47 and in vivo. 20,32 The developmental time-course for this firing behavior has not been defined, but the adult electrophysiological phenotype is apparently present at a very young age. In the present study, we found that there were no major age-related changes in brain slices or in culture. Thus, it appeared that the basic firing properties which distinguish raphe neurons, a highly regular firing pattern, prominent AHP and ramp depolarization between spikes, were nearly fully developed by P3–P5. The current study does not provide evidence for when raphe neurons first begin to produce their characteristic firing behavior. It is interesting that these properties were present shortly after birth, suggesting that they develop in utero. This may indicate that normal activity of raphe neurons is important for homeostasis at the time of birth, consistent with the proposed role of raphe neurons in regulation of a variety of autonomic functions in vivo. 21 The anatomical development of the medullary raphe has been studied previously. 20 However, little is currently known about prenatal development of the basic electrophysiological properties of raphe neurons. Thus, it will be important in the future to define the developmental sequence of the electrophysiology of raphe neurons.
Development of respiratory chemoreception in vivo It is not known when the central chemoreceptors first begin to function during development. Early work suggested that there was no central response to CO2 in the fetus, 17 but this work was performed in anesthetized and/or exteriorized animals. In later experiments performed in unanesthetized, chronically instrumented fetal lambs between 126 and 136 days gestation, varying HCO⫺ 3 concentration of the cerebrospinal fluid via ventriculocisternal perfusion altered the incidence of fetal breathing movements. 19 Thus, central chemoreceptors are functional even in the fetal lamb, apparently developing in utero. Although central chemoreception exists early in development, the sensitivity of the respiratory system to CO2 continues to increase after birth in some species, 1,13,35 including man. 11,27,44 It is not clear how much of this postnatal development of chemosensitivity is a result of an increase in the central vs peripheral chemosensitivity. However, in general, the central chemoreceptors are more important for the response to changes in CO2, while the response to hypoxia is dominated by the peripheral chemoreceptors. The peripheral chemoreceptors are thought to be inactive in the fetus, but undergo rapid resetting immediately after delivery. 22 In kittens, robust peripheral chemoreceptor responses to CO2 occur shortly after birth, at a time that responses to hypoxia are less than those of the adult. 50 However, in this species, development of the peripheral response to CO2 during hypoxia continues for more than eight weeks, which is longer than development of the response to hypoxia itself. 4 After denervation of peripheral chemoreceptors in three- to nine-day-old piglets, ventilation decreases, arterial pCO2 rises and mortality rate is increased. 8 Similar results are seen in two-day-old lambs. 3 In rats younger than 21 days old, peripheral chemoreceptor denervation results in an abnormal respiratory pattern and mortality rates near 50%. 18 Not all species show this dependence on peripheral chemoreceptors in the neonatal period. 24 However, the dependence of homeostasis on peripheral chemoreceptors in some neonatal mammals, and the postnatal increase in the ventilatory response to hypercapnia, suggest that the central chemoreceptor response undergoes maturation during the neonatal period in some mammals, including rats and humans. If raphe neurons are respiratory chemoreceptors, and they do not develop the cellular property of chemosensitivity until they are older than 12 days postnatal, then it would be expected that there would be a developmental increase in the respiratory response to CO2 over the same age range in the same strain of rats. This has been found to be the case. There is little response in ventilation to inspiration of 5% CO2 in Sprague–Dawley rats of three and
Development of chemosensitivity in the raphe
eight days of age, but ventilation approximately doubles at 19 and 34 days. 1 This similar dependence on age does not prove a causal link, but it is consistent with the hypothesis that development of chemosensitivity of raphe neurons plays a key role in development of the systems level response to CO2. Maturation of cellular chemosensitivity in vitro The cellular mechanisms of chemoreception have not been defined in adult mammals, making it difficult to directly study the mechanisms of maturation of the systems level response to hypercapnia. However, it is generally assumed that central chemoreception is due to intrinsic chemosensitivity of specialized neurons in brainstem nuclei linked to respiratory control. Neurons from a variety of brainstem regions are chemosensitive in adult brain slices, including the ventrolateral medulla, 12,25,37 nucleus of the solitary tract, 6 locus coeruleus, 39 hypothalamus 7 and medullary raphe. 41 It is unclear whether any of these neurons are involved in respiratory control, although each of them have been proposed as sites for central respiratory chemoreception. 2,5,16,30,33,36,39,45,46 Changes in chemosensitivity with age have not been described for neurons in any of these brainstem regions. When neurons from the medullary raphe are isolated in tissue culture, they display a high degree of chemosensitivity due to intrinsic cellular properties. 49 Here, we have used tissue culture to provide the first description at the cellular level of agerelated changes in chemosensitivity for any CNS region. These results indicated that, although chemosensitivity was present in neurons of the medullary raphe postnatally, the percentage of chemosensitive neurons and the degree of chemosensitivity both increased with age. The degree of chemosensitivity of stimulated and inhibited neurons both increased with age, with the largest change during the second postnatal week. These changes in chemosensitivity occurred over a time period during which the basic cellular properties of raphe neurons did not undergo significant agerelated changes other than an increase in capacitance and decrease in input resistance. This suggests that chemosensitivity is a cellular property that develops relatively late in these neurons, compared to other basic neuronal properties. Our results confirm our previous impression that brain slices from rats younger than 10 days of age, although technically superior for patch-clamp recording, are not a suitable model system for studying cellular mechanisms of chemosensitivity in adult rats. By analogy, other in vitro preparations using tissue from neonatal rats also may not demonstrate responses to CO2 that are directly comparable to those seen in adult tissue. Maturation of chemosensitivity was very similar in tissue culture and in brain slices, lending important
1009
validation of the dissociated cell culture preparation for studying chemosensitivity. Since recordings in slices are made acutely from neurons that have developed in situ, the developmental sequence of neurons in slices is likely to be an accurate reflection of what occurs in vivo. Thus, we believe that development of raphe neurons in culture is probably very similar to what occurs in vivo, and tissue culture will be an important and useful method for studying chemosensitivity of these neurons. However, despite the similarity in development of chemosensitivity in slices and culture, we have not yet established a oneto-one correlation between age in slices and days in vitro in culture, making it impossible to make a direct quantitative comparison between data obtained using the two preparations. Relationship between maturation of cellular chemosensitivity and development of respiratory chemoreception The relationship between maturation of cellular chemosensitivity in vitro and development of respiratory chemoreception in vivo is unclear. Within the medullary raphe, chemosensitivity at the cellular level in vitro has not been definitively linked with respiratory chemoreception in vivo. However, there is strong in vivo evidence to suggest that the medullary raphe is an important site for central chemoreception. Raphe neurons in cats increase their firing rate with hypercapnia. 48 Electrical stimulation of the raphe pallidus in cats 29 and acetazolamide microinjections in the raphe in rats 2 stimulate respiration. Chemical lesions of serotonergic neurons of the raphe of rats with 5,7dihydroxytryptamine result in an increase in baseline CO2 and a decrease in the response to hypercapnia, 34 and injections of lidocaine or ibotenic acid into the raphe lead to blunting of respiratory chemoreception. 9 Thus, it is likely that the increase in chemosensitivity of medullary raphe neurons with age contributes to the increase in central respiratory chemoreception that occurs postnatally. The similarity of the time-course of maturation of chemosensitivity in brain slices and in culture indicates that defining age-related changes of medullary raphe neurons in tissue culture is relevant to development of these neurons in vivo. The human homolog of the medullary raphe has been reported to be abnormal in a subset of infants who have died of sudden infant death syndrome, 26 and may also be involved in other developmental breathing disorders such as Ondine’s curse and Haddad syndrome. 10,14,42 Neurons of the medullary raphe are involved in cardiorespiratory control. Delayed development of these neurons, or toxicity due to exogenous influences such as tobacco smoke, could lead to inappropriate responses to an acid/base challenge in early life. Defining the cellular mechanisms of raphe neurons may help elucidate the
1010
W. Wang and G. B. Richerson
mechanisms of central cardiorespiratory control in adults, developmental changes in central chemoreception that occur over the neonatal period, and potential mechanisms of dysfunction of homeostatic reflexes induced by hypercapnia.
of chemosensitivity of these neurons over the first three postnatal weeks. This increase in chemosensitivity at the cellular level may underlie the increase in respiratory chemoreception in vivo that occurs over the same time period.
CONCLUSIONS
Neurons of the medullary raphe are candidates for central respiratory chemoreceptors. We have found that there is an increase in the incidence and degree
Acknowledgements—We wish to thank Gabriel Haddad for critical review of the manuscript. This work was supported by NIH R01HL52539 and the VAMC.
REFERENCES
1. Bamford O. S., Schuen J. N. and Carroll J. L. (1996) Effect of nicotine exposure on postnatal ventilatory responses to hypoxia and hypercapnia. Resp. Physiol. 106, 1–11. 2. Bernard D. G., Li A. and Nattie E. E. (1996) Evidence for central chemoreception in the midline raphe. J. appl. Physiol. 80, 108–115. 3. Bureau M. A., Lamarche J., Foulon P. and Dalle D. (1985) Postnatal maturation of respiration in intact and carotid bodychemodenervated lambs. J. appl. Physiol. 59, 869–874. 4. Carroll J. L., Bamford O. S. and Fitzgerald R. S. (1993) Postnatal maturation of carotid chemoreceptor responses to O2 and CO2 in the cat. J. appl. Physiol. 75, 2383–2391. 5. Coates E. L., Li A. and Nattie E. E. (1993) Widespread sites of brain stem ventilatory chemoreceptors. J. appl. Physiol. 75, 5– 14. 6. Dean J. B., Bayliss D. A., Erickson J. T., Lawing W. L. and Millhorn D. E. (1990) Depolarization and stimulation of neurons in nucleus tractus solitarii by carbon dioxide does not require chemical synaptic input. Neuroscience 36, 207–216. 7. Dillon G. H. and Waldrop T. G. (1992) In vitro responses of caudal hypothalamic neurons to hypoxia and hypercapnia. Neuroscience 51, 941–950. 8. Donnelly D. F. and Haddad G. G. (1990) Prolonged apnea and impaired survival in piglets after sinus and aortic nerve section. J. appl. Physiol. 68, 1048–1052. 9. Dreshaj I. A., Haxhiu M. A. and Martin R. J. (1998) Role of the medullary raphe nuclei in the respiratory response to CO2. Resp. Physiol. 111, 15–23. 10. Filiano J. J., Choi J. C. and Kinney H. C. (1990) Candidate cell populations for respiratory chemosensitive fields in the human infant medulla. J. comp. Neurol. 293, 448–465. 11. Frantz I. D., Adler S. M., Thach B. T. and Taeusch H. W. Jr (1976) Maturational effects on respiratory responses to carbon dioxide in premature infants. J. appl. Physiol. 41, 41–45. 12. Fukuda Y., See W. R. and Honda Y. (1980) H ⫹-sensitivity and pattern of discharge of neurons in the chemosensitive areas of the ventral medulla oblongata of rats in vitro. Pflu¨gers Arch. 388, 53–61. 13. Guthrie R. D., Standaert T. A., Hodson W. A. and Woodrum D. E. (1980) Sleep and maturation of eucapnic ventilation and CO2 sensitivity in the premature primate. J. appl. Physiol. 48, 347–354. 14. Haddad G. G., Mazza N. M., Defendini R., Blanc W. A., Driscoll J. M., Epstein M. A., Epstein R. A. and Mellins R. B. (1978) Congenital failure of automatic control of ventilation, gastrointestinal motility and heart rate. Medicine 57, 517–526. 15. Hamill O. P., Marty A., Neher E., Sakmann B. and Sigworth F. J. (1981) Improved patch-clamp techniques for high-resolution current recording from cells and cell-free membrane patches. Pflu¨gers Arch. 391, 85–100. 16. Haxhiu M. A., Erokwu B., Prabhakar N. R., Cherniack N. S. and Strohl K. P. (1992) Locus coeruleus neurons express c-fos immunoreactivity upon stimulation of central chemosensory system. Soc. Neurosci. Abstr. 18, 828. 17. Herrington R. T., Harned H. S. Jr, Ferreiro J. I. and Griffin C. A. (1971) The role of the central nervous system in perinatal respiration: studies of chemoregulatory mechanisms in the term lamb. Pediatrics 47, 857–864. 18. Hofer M. A. (1984) Lethal respiratory disturbance in neonatal rats after arterial chemoreceptor denervation. Life Sci. 34, 489– 496. 19. Hohimer A. R., Bissonnette J. M., Richardson B. S. and Machida C. M. (1983) Central chemical regulation of breathing movements in fetal lambs. Resp. Physiol. 52, 99–111. 20. Jacobs B. L. and Azmitia E. C. (1992) Structure and function of the brain serotonin system. Physiol. Rev. 72, 165–229. 21. Jacobs B. L. and Fornal C. A. (1991) Activity of brain serotonergic neurons in the behaving animal. Pharmac. Rev. 43, 563– 578. 22. Jansen A. H. and Chernick V. (1983) Development of respiratory control. Physiol. Rev. 63, 437–483. 23. Jansen A. H. and Chernick V. (1991) Fetal breathing and development of control of breathing. J. appl. Physiol. 70, 1431–1446. 24. Jansen A. H., Ioffe S., Russell B. J. and Chernick V. (1981) Effect of carotid chemoreceptor denervation on breathing in utero and after birth. J. appl. Physiol. 51, 630–633. 25. Jarolimek W., Misgeld U. and Lux H. D. (1990) Neurons sensitive to pH in slices of the rat ventral medulla oblongata. Pflu¨gers Arch. 416, 247–253. 26. Kinney H. C., Filiano J. J., Sleeper L. A., Mandell F., Valdes-Dapena M. and White W. F. (1995) Decreased muscarinic receptor binding in the arcuate nucleus in sudden infant death syndrome. Science 269, 1446–1450. 27. Krauss A. N., Klain D. B., Waldman S. and Auld P. A. M. (1975) Ventilatory response to carbon dioxide in newborn infants. Pediatric Res. 9, 46–50. 28. Lahiri S. (1994) Carotid body chemoreception: mechanisms and dynamic protection against apnea. Biol. Neonate 65, 134– 139. 29. Lalley P. M. (1986) Responses of phrenic motoneurones of the cat to stimulation of medullary raphe nuclei. J. Physiol., Lond. 380, 349–371. 30. Larnicol N., Wallois F., Berquin P., Gros F. and Rose D. (1994) c-fos-like immunoreactivity in the cat’s neuraxis following moderate hypoxia or hypercapnia. J. Physiol., Paris 88, 81–88.
Development of chemosensitivity in the raphe
1011
31. Lindholm D., Hartikka J., Berzaghi M. D., Castren E., Tzimagiorgis G., Hughes R. and Thoenen H. (1994) Fibroblast growth factor-5 promotes differentiation of cultured rat septal cholinergic and raphe serotonergic neurons—comparison with the effects of neurotrophins. Eur. J. Neurosci. 6, 244–252. 32. Mason P. (1997) Physiological identification of pontomedullary serotonergic neurons in the rat. J. Neurophysiol. 77, 1087– 1098. 33. Mitchell R. A., Loeschcke H. H., Massion W. H. and Severinghaus J. W. (1963) Respiratory responses mediated through superficial chemosensitive areas on the medulla. J. appl. Physiol. 18, 523–533. 34. Mueller R. A., Towle A. C. and Breese G. R. (1984) Supersensitivity to the respiratory stimulatory effect of TRH in 5,7dihydroxytryptamine-treated rats. Brain Res. 298, 370–373. 35. Nattie E. E. and Edwards W. H. (1981) CSF acid–base regulation and ventilation during acute hypercapnia in the newborn dog. J. appl. Physiol. 50, 566–574. 36. Nattie E. E. and Li A. H. (1996) Central chemoreception in the region of the ventral respiratory group in the rat. J. appl. Physiol. 81, 1987–1995. 37. Neubauer J. A., Gonsalves S. F., Chou W., Geller H. M. and Edelman N. H. (1991) Chemosensitivity of medullary neurons in explant tissue cultures. Neuroscience 45, 701–708. 38. Paxinos G. and Watson C. (1986) The Rat Brain in Stereotaxic Coordinates. Academic, New York. 39. Pineda J. and Aghajanian G. K. (1997) Carbon dioxide regulates the tonic activity of locus coeruleus neurons by modulating a proton- and polyamine-sensitive inward rectifier potassium current. Neuroscience 77, 723–743. 40. Rae J., Cooper K., Gates P. and Watsky M. (1991) Low access resistance perforated patch recordings using amphotericin B. J. Neurosci. Meth. 37, 15–26. 41. Richerson G. B. (1995) Response to CO2 of neurons in the rostral ventral medulla in vitro. J. Neurophysiol. 73, 933–944. 42. Richerson G. B. (1997) Sudden infant death syndrome: the role of central chemosensitivity. Neuroscientist 3, 3–7. 43. Richerson G. B. and Messer C. (1995) Effect of composition of experimental solutions on neuronal survival during rat brain slicing. Expl Neurol. 131, 133–143. 44. Rigatto H., Brady J. P. and de la Torre Verduzco R. (1975) Chemoreceptor reflexes in preterm infants: II. The effect of gestational and postnatal age on the ventilatory response to inhaled carbon dioxide. Pediatrics 55, 614–620. 45. Sato M., Severinghaus J. W. and Basbaum A. I. (1992) Medullary CO2 chemoreceptor neuron identification by c-fos immunocytochemistry. J. appl. Physiol. 73, 96–100. 46. Schlaefke M. E., See W. R. and Loeschcke H. H. (1970) Ventilatory response to alterations of H ⫹ ion concentration in small areas of the ventral medullary surface. Resp. Physiol. 10, 198–212. 47. Vandermaelen C. P. and Aghajanian G. K. (1983) Electrophysiological and pharmacological characterization of serotonergic dorsal raphe neurons recorded extracellularly and intracellularly in rat brain slices. Brain Res. 289, 109–119. 48. Veasey S. C., Fornal C. A., Metzler C. W. and Jacobs B. L. (1995) Response of serotonergic caudal raphe neurons in relation to specific motor activities in freely moving cats. J. Neurosci. 15, 5346–5359. 49. Wang W., Pizzonia J. H. and Richerson G. B. (1998) Chemosensitivity of rat medullary raphe neurones in primary tissue culture. J. Physiol., Lond. 511, 433–450. 50. Watanabe T., Kumar P. and Hanson M. A. (1996) Development of respiratory chemoreflexes to hypoxia and CO2 in unanaesthetized kittens. Resp. Physiol. 106, 247–254. (Accepted 26 August 1998)