- Email: [email protected]
Regulation of pH and HCO3 in brain and CSF of the developing mammalian central nervous system
Developmental Brain Research, 38 (1988) 255-264 Elsevier
255
BRD 50684
Regulation of pH and H C O 3 in brain and CSF of the developing mammalian central nervous system Conrad E. Johanson, James Allen* and C. Dean Withrow Department of Neurosurgery, Brown University and Rhode Island Hospital, Providence, R102902 (U.S.A. ) and Department of Pharmacology, University of Utah School of Medicine, Salt Lake City. UT84132 (U.S.A.) (Accepted 11 August 1987) Key words." Cerebrospinal fluid pH; Cerebral cortex pH; Central nervous system acid-base regulation; Dimethadione method; Bicarbonate and pCO2; Metabolic acidosis and alkalosis; Ontogeny of cell pH; Cell H ion homeostasis
Acute (2-h) metabolic acidosis or alkalosis was induced in immature rats to ascertain the ability of their incompletely-developed CNS to regulate pH when challenged with perturbations in blood [H] and [HCO3]. Brain and cisternal CSF pH were determined from steady-state distribution of [14C]dimethadione, a weak organic acid. By 1 week post partum, there was a remarkable stability of pH in the cerebral cortex of animals subjected to arterial pH extremes of 7.1 and 7.5. However, CSF pH in 1-week-old animals rendered alkalotic remained 0.07-0.08 units above control due to lack of a compensatory increase in pCO 2, and to a blood-CSF barrier apparently more permeabie to HCO 3. As arterial HCO3, i.e. [HCO3]art, was varied from about 10 to 30 mmol/l, the infants maintained [HCO3J,f only half as effectively as adults, i.e. A[HCO3]csf/A[HfO3]art was 0.4 and 0.2 at l and >4 weeks, respectively. Throughout postnatal ontogenesis, [HCO3]csf was more resistant to alteration by metabolic acidosis than by alkalosis. Overall, the results indicate that immature rats challenged with systemic acid-base loads are less capable than adults in regulating CSF pH, but they are able to maintain brain pH.
oped CNS at birth. The rat is atypical among several laboratory mammals because it is born with a brain less mature than that of many other species. Several studies of neonatal rats l'3,7,~t'tS'28 have revealed
of the CSF secretory process 32, that may influence acid-base regulation. Previous acid-base investigations of infant rat CNS have been carried out primarily in 1-week-old animals t2'31. In this investigation we systematically analyzed acid-base balance in 1-, 2-, 3- and > 4 - w e e k - o l d rats. By measuring p H and total CO2 in C S F and the p H of cerebral cortex in response to acute metabolic acidosis (and alkalosis), we have assessed the regulatory capacity of the CNS at various ontogenetic
greater permeability of the b l o o d - b r a i n (BBB) and b l o o d - c e r e b r o s p i n a l fluid (CSF) barriers, larger content of total CO2 in brain tissue, and greater extraceilular space and water content in the i m m a t u r e CNS than in the mature brain. There are o t h e r characteristics of the immature rat brain, such as the postnatal proliferation of astroglia and d e v e l o p m e n t
stages. The response of [HCO3]cs f has also been analyzed in the context of the maturation of neural barrier systems to H C O 3. Since the time course of maturation of CSF secretion and astroglial function is known, it is possible to evaluate deductively the role of these factors in CNS acid-base regulation 32. The results for young rats indicate that although H C O 3 in
INTRODUCTION Few studies have addressed acid-base regulation in the immature central nervous system (CNS); most of these investigations have e m p l o y e d the neonatal lamb 4'5 or dog t7 which have a relatively well-devel-
,p
resent address: Searle Pharmaceuucals Inc., Searle Safety Assessment, 4901 Searle Parkway, Skokie, IL 60076, U.S.A. Correspondence." C.E. Johanson Department of Neurosurgery, Brown University and Rhode Island Hospital, 593 Eddy Street, Providence, RI 02902, U.S.A. 0165-3806/88/$03.50 © 1988 Elsevier Science Publishers B.V. (Biomedical Division)
25~ plasma gains access to CSF, there is remarkable regulation of brain pH after 2 h of metabolic distortions. CSF pH is less well maintained, especially in infant rats challenged with HCO~ loads. MATERIALS AND METHODS
Experimentalprotocol. The effect of 2-h metabolic acid-base distortions on the pH of arterial blood, CSF and brain was analyzed in immature Sprague-Dawley rats (1, 2, and 3 weeks) and compared to young adults (4-5 weeks). Ten or more animals were used for each treatment and age. Anesthesia was induced by ketamine (75-125 mg/kg), with the larger doses used in the older animals. Bristol Laboratories supplied ketamine as the HC1 salt, 100 mg/ml. The procedure for the injection of [14C]DMO (dimethadione), to indicate pH, is described elsewhere s. The radio-DMO (5,5-dimethyloxazoladine-2,4-dione, New England Nuclear; 0.4 mCi/mg) was injected intraperitoneally, (t.02 gCi/g. Salt solutions were injected intraperitoneally (i.p.) (0.01 ml/g) to induce the metabolic distortions: acidosis (0.3 mg NH4CI/g body wt.), control (0.1 mg NaCl/g), and alkalosis (0.8 mg NaHCO)/g). All salts were analytical grade from Sigma. A time-course preliminary study revealed that, following the injection of each salt, there was rapid induction (by 5 rain) of the acid-base disturbance which remained stable in blood for up to 4 h in the nephrectomized animals. Sampling of fluids and brain. At the end of each 2h experiment, blood was sampled (0.3-0.5 ml) from the abdominal aorta with a small heparinized syringe. Immediately thereafter the animal was positioned for CSF sampling 12. The dural membrane covering the cisterna magna was punctured with a 100-/al glass micropipet attached by tubing to a 1-ml syringe. CSF was withdrawn by gentle suction; if blood contamination occurred, the sample was discarded. Analysis of the CSF total CO 2 content was performed immediately after collection to minimize CO2 changes in the fluid, A 0.05- to 0.15-g sample of cortical tissue was sliced from the parietal lobe of both cerebral hemispheres. Meninges and residual blood were blotted to remove extraparenchymal elements from the brain sample. Tissues were weighed to 0.1 mg on a Sartorius analytical balance. Analyses of samples. The handling of arterial sam-
pies for blood gas and radioactivity analyses has been describe& s. Total CO 2 content in CSF was determined by gas chromatography in a Fisher Clinical Gas Partitioner. Between 10 and 20 f~l of CSF was used for each analysis. The CO: content of CSF was quantitated by comparing peak heights obtained with unknown samples to those obtained with equal volumes of a standard solution of NaHCOs (24.6 raM). CSF not analyzed for CO 2 was used for measurement of [14C]DMO activity. Prior to counting the l~C-activity by liquid scintillation, all of the samples of CSF, plasma and brain were solubilized r'< The DMO technique for pH. DMO as well as C(), distribution can be used to calculate pit and [HCO3]. Infant rat CSF samples are poorly buffered and small (10-15 hal), and thus are not suitable for reliable direct measurement of pH by electrode, ]'here is agreement among adult pH values for CSF, and brain, whether determined by DMO, CO2 distribution or pH electrodes s'2~'22"2a. Reed et al. 2~ showed that DMO distribution between arterial blood and CSF yielded a value agreeable with directly measured CSF pH, Similarly, Roos >'24 reported that un-ionized DMO distributes at equilibrium between CSF and arterial blood. Such findings indicate feasibility of DMO for pH measurement of CNS regions 2e, Calculations and statistical analyses. CSF pH was calculated from the steady-state distribution of [14C]DMO29, and alternatively from the distribution of HCO 3 and CO225'26. CNF pH was determined from steady-state distribution of [14ClDMOS'm, the percent volume of distribution (Vdt of which was calculated as: 100 × dpm/mg CSF water/dpm/~l plasma water Values for arterial pH and total activity of [14C]DMO were used to calculate concentrations of ionized and un-ionized [14C]DMO in plasma. The formulae and assumptions for calculating the ratio of ionized to unionized [~4C]DMO in brain tissue and hence cell pH, have been previously presented s. Due to the limiting effect of the BBB on ion distribution, the concentration of ionized [14C]DMO in extracellular fluid (ECF) of brain was taken as that in CSF, rather than in plasma s. The H : O contents of plasma and brain for various ages were determined as the difference be-
257 tween wet and dry weights of samples. Values for rat cerebral cortical ECF volume (% of wet tissue wt.) of 22.5, 16, 14.5 and 13.5%, respectively, were utilized in cortical pH calculations for 1-, 2-, 3- and >4-weekold animals 28. Neither metabolic acidosis (NH4CI) nor alkalosis (NaHCO3) alters ECF volume 3°. Severe respiratory acidosis, e.g. 20% CO2 to bring about arterialpCO 2 > 100 torr, does not change ECF volume ~5. Arterial [HCO3] was calculated from electrodemeasured pH and pCO2 values, by use of the Henderson-Hasselbalch equation (pKa = 6.10, aCO 2 = 0.030). C S F [ H C O 3 ] w a s determined as the difference between total CO 2 and carbonic acid (i.e. apCO2) where a for CSF at 37 °C is 0.03216. CSF pCO 2 was derived by the method of Brzezinski et al. 2, who systematically described the pCO2 of cisternal CSF as a function of arterial pCO 2. Electrode measurements have confirmed that brain pCO 2 can be reliably estimated as arterial pCO2 + 6 torr 2. For brevity, [HCO3] in mmol/kg H20 is expressed as mM. One-way analysis of variance and the multiplerange test (Tukey version) were used to test for significant differences (P < 0.05) among age groups. The paired Student's t-test was also utilized in CSFplasma analysis.
stantially different in respect to pH, p C O 2 and [HCO3]; see Fig. 1. The respiratory acidosis nature of blood in immature animals converted between 1 and 2 weeks to adult-like composition. The utilized doses of NH4C1 and NaHCO3 caused comparable pH deflections of 0.10-0.15 pH units in arterial blood (Fig. 2). Except for 1-week-old animals, all stages of development displayed the typical adult-like response of CO 2 elimination ($pCO2) in metabolic acidosis and CO 2 retention ( I' pCO2) in aikalosis. CSFand brainpH values. In controls, CSF pH (derived from DMO distribution) did not change significantly over the developmental period studied (1-5 weeks); pH values were usually in the range of 7.30 + 0.04 (Table I). CSF pH was also calculated from the distribution ratio of HCO 3 to CO2, by using data for total CO 2 (Table III) and CSF pCO2. In l-, 2-, and 3-week-old rats, there were generally no statistically significant differences between DMO- and CO2-derived values for CSF pH (in either control, acid- or base-loaded animals). At >4 weeks, the CO: method tended to yield CSF pH values greater than those by DMO. Table II contains an ontogenetic analysis of acid-base distortion-induced alterations in CSF pH relative to those in arterial pH. In metabolic acidosis (2 h) the ratio ofACSF pH to d arterial
ARTERIAL BLOOD
RESULTS
E~ METABOLIC ACIDOSIS r--] METABOLIC ALKALOSIS
Baseline and experimental values for arterial parameters. The composition of arterial blood in infant
U u. U
rats (1 week), compared to young adults, was sub-
pH
pCO2(torr)
[HCO~ (mM) & pCO2 (From control )
25
"'7.5
15
20t
~71 7.0
I
I
I0
I
2
5 >4
2
5
>4
POSTNATAL AGE (weeks)
0 I
2
5 >4
0j
I
2
3>4
POSTNATAL AGE (weeks) Fig. 1. Ontogenetic changes in the baseline composition of arterial blood acid-base parameters. Each bar is the mean + S.E.M. for at least 11 rats. All controls received NaCI, i.p., 0.1 mg/g body weight, 2 h before sacrifice. *P < 0.05, 1 or 2 weeks vs >4 weeks, by multiple-range test.
Fig. 2. Developmental analysis of the magnitude of distortioninduced changes (from corresponding values for NaCl-injected controls) in pH and pCO 2 (torr) of arterial blood in rats subjected for 2 h either to metabolic acidosis (0.3 mg/g) or to alkalosis (0.8 mg/g). Each bar is the mean + S.E.M. for 11 or more animals. *P < 0.05, by multiple-range test, that the experimental alteration is significantly different from corresponding control.
25S ]ABLE 1 p H ()f cerebrospinal fluid and cerebral corte_r in developing rats challenged with metabolic acidosis or alkalosis Each mean _+ S.E.M. for CSF pH is for 10-12 animals; for cerebral cortex data, n = 12. CSF and brain samples were taken from the cisterna magna and frontoparietal cortex, respectively. Cerebral cortex pH
(erebrospinal fluid pH 1 week
2 weeks
3 weeks
>4 weeks
I week
Metabolic acidosis Control
7.25 ± (I.(12 7.26 _+ 0.02
7.21 ± 0.02* 7.29 _+ 0.02
7.31 +- 0.02 7.34 + (I.04
7.26 +- 0.03 7.26 ± 0.03
Metabolic alkalosis
7.33 +__0.03*
7.27 +__0.01
7.34 + 0.(13
7.30 + 0.03
2 weeks
3 week,~
>4 weeks
7.14 +- 0.02 ~ 7.03 + 0.02 7.12 +__0.02 7.08 +- 0.03
7.(19L 0.03 7.12 -+_1!.02
7.02 ± 0.03 7.07 +__0.02
7.13 __+0.04
7.10 ± 0.t13
7,06 +__0,03
7.08 +__0.01
* P < 0.05, control vs acidosis or alkalosis. * P < 0.05, 1 week vs >4 weeks, by multiple-range test. pH values were determined from the steady-state distribution of [laC]dimethadione, Treatments for 2 h are described in Materials and Methods.
p H was only slightly g r e a t e r at 1 and 2 w e e k s than at 3
than that in animals at 2 w e e k s (21.4), 3 w e e k s (20.3)
and > 4 weeks. A C S F p H / A arterial p H in m e t a b o l i c
or > 4 w e e k s (22.7); see T a b l e l l I . T h e r e s p o n s e of
alkalosis was substantially g r e a t e r at 1 w e e k ((1.68)
the C S F system to a c i d - b a s e c h a l l e n g e v a r i e d accord-
than in adults (close to z e r o ) : w d u e s at 2 and 3 w e e k s
ing to age as well as distortion. T h e l - w e e k - o l d rat
were i n t e r m e d i a t e ,
e x h i b i t e d a significant d e c r e a s e (by 3.4 m m o l ) and in-
but closer to the adult
level
(Table II).
crease (by 3.6 retool) in total CO2 d u r i n g m e t a b o l i c
T h e p H v a l u e s for c e r e b r a l c o r t e x t e n d e d to be
acidosis and alkalosis, r e s p e c t i v e l y , f r o m a c o n t r o l
m o r e alkaline in the less m a t u r e a n i m a l s ( T a b l e I).
value of 28 m m o l / l . M e t a b o l i c alkalosis c a u s e d a sta-
N e i t h e r 2-h m e t a b o l i c acidosis n o r alkalosis c a u s e d a
tistically significant a u g m e n t a t i o n in C S F total CO~
significant a l t e r a t i o n in cortical p H at any age.
at 2, 3, and > 4 w e e k s ; in c o n t r a s t , C S F in animals at
Cerebrospinal fluid content o f total C O 2 and bicar-
these s a m e 3 stages of d e v e l o p m e n t was p r o t e c t e d
bonate. In c o n t r o l rats at 1 w e e k the C S F had a
from a substantial fall in total C O z d u r i n g m e t a b o l i c
m a r k e d l y h i g h e r c o n t e n t of total CO2 (28 mmol/I)
acidosis for 2 h ( T a b l e III).
TABLE 11 Ontogenetic analysis of acid-base distortion-induced alterations in CSF pH relative to those in arterial pH Average change refers to difference between mean control and mean experimental values. Since there were generally no statistically significant differences between CSF pH values, DMO vs CO: methods, the overall mean was determined by averaging pooled CSF values. Postnatal age 1 week
2 weeks
3 weeks
;>4 weeks
Metabolic acidosis Average change in CSF pH units Average change in arterial pH units A CSF pH/A arterial pH
-0.015 -0.13 0.12
-0.030 -0.17 0.18
+0.010" -0.10 0.10
+0.005* "0.17 0.(/3
Metabolic alkalosis Aver age change in CSF pH units Average change in arterial pH units A CSF pH/A arterial pH
+ 0.075 +0.11 0.68
+ 0.0 l 0 +0.09 0.11
+ 0.015 +0.13 0.12
it +0.10 0
* CSF pH increase was in the opposite direction to the change in arterial pH, but it was not significantly elevated above the control value for CSF pH.
259 T A B L E III
Concentration of total carbon dioxide in the cisternal cerebrospinal fiuid of rats at various stages of development Each m e a n + S.E.M. is for 10 or more animals. Control rats were injected with NaCI solution; metabolic acidosis and alkalosis, respectively, were induced with NH4CI and N a H C O 3 for 2 h. CO 2 concentration was determined by gas chromatography.
Treatment
Concentration of CO e (mmol/l CSF)
Metabolic acidosis (NH4CI) Control (NaCI) Metabolic alkalosis (NaHCO3)
1 week
2 weeks
3 weeks
>4 weeks
24.6 _+ 0.8* 28.0 + 1.0 31.6 _+ 0.7*
20.0 __+0.6 21.4 _+ 0.7 26.7 + 1.4"
19.3 + 0.6 20.3 _+ 0.4 25.8 + 1.5"
22.0 _+ 0.9 22.7 _+ 0.4 25.9 _ 1.3'
* A statistically significant difference (P < 0.05, multiple-range test) between control and treated animals within an age group.
The mean concentration of HCO 3 in CSF relative to plasma is presented according to age and treatment in Fig. 3. In controls, CSF [HCO3] was generally not significantly different from plasma [HCO3], except at 1 week. At all ages investigated, CSF [HCO3] was significantly greater than plasma [HCO3] after 2 h of acidosis. In metabolic alkalosis, the [HCO3] in CSF remained significantly lower than that in plasma only in the older animals (3 and >4 weeks).
30
[ ] CSF [ ] PLASMA
***
2O
The relationship between CSF [HCO3] and plasma [HCO3] is given on an individual basis in Fig. 4; for each age, there is a positive linear relationship between the [HCO3] in the two fluids. The slope value at 1 week was 0.40 + 0.08 (S.E.M.) and it progressively decreased with age to values of 0.38 + 0.07, 0.23 _+ 0.05, and 0.20 + 0.06 at 2, 3 and >4 weeks, respectively; the least squares regression for 1-week data is significantly different from that for >4 weeks. Brain cell [HCO3] was calculated to be 17 mM at 1 weele, and it declined to 11-12 mM at 2, 3 and >4 weeks. The cellular HCO 3 was not significantly affected by the acid-base disturbances at any stage of development.
I0
}
DISCUSSION
Developmental changes" in baseline values. In adult
~o-t 401 C. ALKALOSIS
I
2
3
>4
POSTNATAL AGE (WEEKS)
Fig. 3. Relationship between the concentration of H C O 3 in plasma and that in CSF during metabolic acidosis (Part A) and alkalosis (C). Data for control animals (B) are included for comparison. Rats of 4 postnatal ages are represented. CSF was sampled from cisterna m a g n a and blood was collected from abdominal aorta. Each bar is the m e a n + S.E.M. for at least 10 animals. Statistical analysis was performed with paired t-test, CSF vs plasma. ***P < 0.001, **P < 0.01, *P < 0.05.
mammals, CSF [HCO3] is approximately equal to plasma [HCO3] s. During early development in the rat (1 week), CSF [HCO3] and arterial [HCO3] are both elevated several mM above corresponding adult control levels. The substantial attenuation of CSF [HCO3] by 5 mM between the 1st and 2nd weeks in controls is accompanied by a comparable decrease in [HCO3] in plasma and brain cells. Thus, the HCO3 concentration gradients among the 3 compartments appear not to be substantially different in maturation vs early rapid growth. The 1-week-old rat cortical pH values tended to be more alkaline than adult values 8 which are close to 7.0. Thus, the higher pHs in infants for brain, and for non-neural tissues 3°, likely reflect greater intraceilular [HCO3]. In controls, the lack of significant change in CSF pH with age indicates a preferred value at which CSF
26i~
40] I WEEK
2 WEEKS I
35l
o
To
25
•
.j
4
.2
° . L
]Z
20
sE 15 E
n- 5 ~
5
1
IO 1% 20
---
-T---'-T--"
I
is
4s s
20
sb
3~5
4O
45
] ~4 WEEKS
d 40- 3 WEEKS Z
h-
~ 30 Ill m
~
•
40 45 5
10
~25
] .-
......
20 •
A•
15Z t0 55
©
15 20
25
30
35
15
20
25
30
35
40
45
ARTERIAL PLASMA [HCO~](mmol/I.H20) Fig. 4. Correlation between the concentration of H C O 3 in cisternal CSF and that in arterial plasma. Each datum point corresponds to the concentrations of H C O 3 in the plasma and CSF of a single rat. The symbols represent the 3 treatments: NH4CI (liD), NaCI (A), NaHCO 3 (O). Each panel depicts the relationship for a given age. Slopes, determined by least-squares best fit, of 0.40 _+ 0.08 (S.E.M.), 0.38 +_0.07, 0.23 + 0.05 and 0.20 + 0.06 were calculated for the I-, 2-, 3- and >4-week-old animals, respectively. By analysis of covariance, there is a significant difference (P < 0.05) between the slope value at 1 week compared to that at >4 weeks.
p H is regulated. Consistent with this view, the CSF system throughout development continues to regulate p H at about 7.25 when challenged with arterial acidosis. While CSF p H remains stable after birth, the pH of arterial blood continually rises thereby gradually setting up the p H gradient of 0.1 unit (CSF more acidic) across the adult mammalian barrier (Fig. 5). That there is no appreciable p H gradient at 1 week is consistent with the more permeable barriers 7J2'~s'2s and perhaps immaturity of systems translocating ions in the incompletely developed choroid plexus 28. It seems reasonable to hypothesize that zl/~ H ÷ (i.e. electrochemical potential difference between blood and CSF) is close to zero in neonatal rats. Since Ap H ÷ is normally about +5 to 10 mV in adults, the postnatal rat is a promising developmen-
tal model for pursuing the nature of H-ion distribution across neural barrier systems.
Ontogenetic differences in stability of CSF pH in acute distortions. CSF p H is not well-regulated in immature rats subjected to metabolic alkalosis. In the present experiment, CSF pH in 1-week-old animals loaded with HCOs (and anesthetized with ketamine) increased significantly by 0.07 and 0.08 units, from [MC]DMO and CO 2 data, respectively; one cause for this alkaline shift was insufficiently retained CO23° to counteract the substantially elevated CSF [HCO3]. In a similar acute study, of infa,it rats 12. but with a different anesthetic (pentobarbital) and experimental duration (1 h), CSF p H in metabolic alkalosis was also increased from 7.27 (control) to 7.40: the greater CSF pH augmentation in the latter case Iz occurred
261
*0•18-
*0.15-
0
*0.12-
13
(/3
o *0.09I 0
+0.06*0.03-
o • 0
,,=::1
O-i
0 •
8
• 0
0
-0.03adult POSTNATAL AGE (weeks)
Fig. 5. Gradient of pH across the blood-CSF barrier in control rats as a function of postnatal age. Blood and CSF were sampled from abdominal aorta and cisterna magna, respectively. Symbols refer to investigations in which CSF pH was calculated from the distribution of CO_, (filled) and DMO (unfilled), respectively. Each circle represents a mean value (n = 10) from the present investigation for ketamine-anesthetized rats; A, ref. 12, n = 14 pentobarbital-anesthetized rats; V, V, ref. 21. n = 10. ether anesthesia; ~ , ref. 20, n = 6 unanesthetized rats; []• II, ref. 21. n = 10 pentobarbital-anesthetized rats: ,o . ref. IlL pentobarbital and ether, respectively. because the 24% rise in p C O 2 secondary to alkalosis had not counteracted the 50%-elevated [HCO3] in the CSF 1 h after the beginning of the distortion. Thus, two investigations have shown that immature rat (1-week) CSF does not defend pH effectively in acute metabolic alkalosis. However, by 2-3 weeks, CSF displayed pH homeostasis when challenged with non-respiratory alkalosis, presumably since: (1) barrier systems at blood interfaces are more impermeable 7, (2) central and/or peripheral chemoreceptors are capable of adjusting pCO2, and (3) choroid plexus secretion TM is more developed (hence greater buffering ability). With respect to CSF pH regulation, metabolic acidosis appears less problematic to immature rats (1 week) than metabolic alkalosis 12. We confirmed that CSF pH is unaltered in animals injected with NH4C1 at 1 week, such protection presumably related to relatively high normal baseline concentration of HCO3 in brain cells, cerebral interstitial fluid, CSF and choroid plexus t°'31. Thus, especially at 1 week, the concentration gradient for H C O 3 (barrier cell, or brain cell, to CSF) in metabolic acidosis favors net diffusion of HCO 3 into CSF replacing H C O 3 depleted by
NH4CI treatment. The substantially reduced baseline [HCO3] in these CNS compartments at 2 weeks (vs 1 week) may explain why CSF pH was not maintainable even in acidotic animals• CSF pH maintenance at 3 weeks and later, in both distortions, coincides with nearly complete development of BBB restrictiveness, astrocyte function, and choroid plexus barrier and transport systems for ion homeostasis and fluid secretion. Intracellular p H o f cerebral cortex• Cortex pH is more alkaline in immature than in mature animals s. Similarly, cell pH in extra-CNS tissues (liver, muscle, etc. 3°) is also more basic in infants than adults• Alkaline pH i is apparently conducive to growth processes. Interestingly, the relative alkalinity of cells in immature animals occurs in the face of elevated levels of tissue pCO2; in this regard, arterial p C O 2 is 20 torr higher at 1 week vs >4 weeks. The brain pH i in young rats was not altered by metabolic acidosis or alkalosis (Table I). Regulation of pH i in adult mammalian brain has been delineated TM 24,26, but the ability of immature brain to maintain pH in face of variations in [H] and [HCO3] in blood and CSF has apparently not been investigated. Siesjo and Messeter 25 concluded that adult brain cells regulate pH i by organic acid production or consumption, physicochemical buffering and active transport. There may be maturational differences in ability to buffer by virtue of cell protein content or capacity to metabolize acids like lactate. Because the untreated week-old rat has relatively high [HCO3] in brain cells 3°'3t, this likely contributes to H-buffering during acid challenges to neurons. That infant rats actively translocate H and/or H C O 3 can be argued from control pHs. With a CSF (i.e. interstitial fluid) pH of 7.26 and cerebral cortex pH i of 7.12 at 1 week (Table I), the Nernst equation predicts that E H is - 9 mV at 37 °C. By deduction, H-ions must distribute against an electrochemical gradient, i,e. be actively extruded, because neurons in infants would not function adequately with such a low membrane potential. Awaiting elucidation is the role of secondary active transport ( N a - H and C1-HCO3 exchange) during perturbations of the milieu of the perinatal CNS. Assumptions and limitations in calculations. D M O usage to measure cell pH in brain (pHi) was scrutinized by Roos 23'24for effects of theoretical variations in membrane potential and ECF volume. His find-
ings of reliability are pertinent to this study in which potential difference (PD) and ECF volume in brain vary significantly with age. The absolute value of pH i is not particularly sensitive to error in brain ECF volume measurements: even so, our emphasis has been on experimental changes in pH i, so any error in ECF volume (for a given age) should not seriously affect relative wtlues for control vs treated groups. Accurate calculation of [HCO3] from CO: data depends on a sound value for pCO2. CO, distribution among plasma, CSF and brain tissue in rats was studied by Siesjo and co-workers 2'>, who established with electrodes that arterial pC02 data can be used to estimate accurately the p C O 2 of CSF and cerebral cortex. Changes in CSF and brain [HCO ff with maturation and treatment. Infant rats (1 week) maintained CSF [HCO3] half as effectively as young adults in metabolic distortions; compare slopes in Fig. 4. One explanation is that barrier systems (components of transport and permeability) are less developed in the youngest rats 7'~s'2s and consequently transmit changes in plasma [HCO3] more extensively to CSF. However, it should be emphasized that in 1-week rats the A[HCO3]c,f/A[HCOs]~r t at 2 h was only 0.4; thus, even with incompletely-developed barriers, the distortion in [HCO3]art was not fully transmitted to the CSF. At all ages, CSF [HCO3] was more resistant to alteration in acute metabolic acidosis than in alkalosis. This supports the hypothesis that brain cell HCO 3 is a buffer reservoir partially to replace CSF [HCO3] titrated downward in acidosis. Thus, brain [ H C Q l i (highest in neonates) may help to maintain CSF [HCO3] in the face of metabolic acidosis, thereby counteracting potentially deleterious effects of leaky barriers to HCO3 and H early in development. Even if brain cells supply HCO3 to buffer CSF, they maintain their own [HCO3]. Of several tissues in infants s°, brain cell [HCQ] was generally the most stable in distortions. This is interesting since, ontogenetically, glial (and thus carbonic anhydrase) proliferation has not yet occurred 1 week after birth 32. By deduction this intimates that neurons (deficient in carbonic anhydrase) in infants have effective pH homeostatic ability. Another possible factor affecting stability of [HCO3],is fluctuation in interstitial fluid H C O 3, i.e. [HCO3]ist, determined by [HCOs]cs f (already discussed) and by H C O 3 flUX across the BBB.
Plasma HCO 3 significantly penetrates adult BBB ~:~. and it is likely that movement of HCO 3 and other solutes 7"js2~ is even less restricted across infant cerebral capillaries 7'~, Yet, there must be some impediment to HCO3 movement across infant BBB, as is the case for non-electrolytes 7'~'~s'~s. Thus, even after 2 h of distortions in infants, one expects that Zl[HCO3]isf/Zl[HCO3]ar > similar to the case for ,d[HCO3]csf/,d[HCO3lar t, would be less for brain than other tissues. Non-neural parenchyma >, unprotected by barrier systems, are subjected extracellularly to the full brunt of alterations in [HCO3]~,. Thus, the BBB and choroid plexus, though more permeable in infants than in adults, help by some barrier effect to confer stability to brain phi and [HCO31i in 1-week animals challenged with H / H C O 3 loads. Blood-CSF barrier vs BBB. In adult rats, the blood-CSF barrier epithelium is leakier than the BBB endothelium. Ions and small non-electrolytes, like 22Na, 42K and mannitol (tool. wl. = 181), have permeability coefficients in choroid plexus at least 10-100 times greater than those in cerebral capillaries 27. Urea and HCO3, with comparable molecular weights of 60 and 61, respectively, permeate the choroidal membrane in infants more readily than m adults 12a8. Neonatal rats rendered hyperkalemic have stable [K]isf, but a marked rise in [K]cs6 this demonstrates that developmental barrier tightening to K in choroid plexus lags the BBW 4. Collectively, the information indicates that the parenchyma in postnatal rat brain is protected against acidosis and hyperkalemia, in spite of the leaky blood-CSF barrier. Perspective. The CNS in the rat one week after birth, with its somewhat leaky neural barrier systems 79'1s, incompletely developed astroglial and choroid plexus transport systems ~'~3:, and minimal CSF sink action ~s, can protect the pH of its cellular elements against acute systemic metabolic acid-base insults. That even the neonate regulates brain cell pH so vigorously suggests that stability of this physiological parameter is of prime importance to neurons during development. The lesser ability of CSF in infants (1 week) to defend pH, ~specialty against alkaIosis, is attributed to substantial penetration of HCO 3 across the incompletely-developed choroid plexus ~> ~8,2s. On the other hand, the effective impermeability of the external limiting membranes of brain paren-
263 c h y m a l e l e m e n t s ( n e u r o n s ) , e v e n in infants, assures
G r e e n w o o d for c o n t r i b u t i o n s to the p r o j e c t . This
that p H i and [HCO3] i are not a l t e r e d in the face of
w o r k was s u p p o r t e d by U S P H S G r a n t 1 R01 N S
moderate
displacements
of
cortical
extracellular
[HCO3] , i.e. 2 0 - 4 0 % of arterial fluctuations.
13988,
NINCDS
and
a Research
Career
Devel-
o p m e n t A w a r d to C . E . J . This p r o j e c t was also partially s u p p o r t e d by N I H g r a n t G M 07579 and by
ACKNOWLEDGEMENTS
funds f r o m the U n i v e r s i t y of U t a h R e s e a r c h C o m m i t tee.
W e w o u l d like to t h a n k W a y n e B a m o s s y and J e a n
REFERENCES 1 Amtorp, O. and Sorensen, S.C., The ontogenetic development of concentration differences for protein and ions between plasma and cerebrospinal fluid in rabbits and rats, J. Physiol. (London), 243 (1974) 387-400. 2 Brzezinski, J., Kjallquist, A. and Siesjo, B.K., Mean carbon dioxide tension in the brain after carbonic anhydrase inhibition, J. Physiol. (London), 188 (1967) 13-23. 3 Ferguson, R.K. and Woodbury, D.M., Penetration of [~4C]inulin and [14C]sucrose into brain, cerebrospinal fluid, and skeletal muscle of developing rats, Exp. Brain Res., 7 (1969) 181-194. 4 Herrington, R.T., Harned, H.S., Jr., Ferreiro, J.I. and Griffin, C.S., The role of the central nervous system in perinatal respiration: studies of chemoregulatory mechanisms in the term lamb, Pediatrics, 47 (1971) 857-864. 5 Hodson, W.A., Fenner, A., Brumley, G., Chernick, V. and Avery, M.E., Cerebrospinal fluid and blood acid-base relationships in fetal and neonatal lambs and pregnant ewes, Resp. Physiol., 4 (1968) 322-332. 6 Johanson, C.E., Choroid epithelial cell pH, Life Sci., 23 (1978) 861-868. 7 Johanson, C.E., Permeability and vascularity of the developing brain: cerebellum vs cerebral cortex, Brain Res., 190 (1980) 3-16. 8 Johanson, C.E., Differential effects of acetazolamide, benzolamide and systemic acidosis on hydrogen and bicarbonate gradients across the apical and basolateral membranes of the choroid plexus, J. Pharmacol. Exp. Ther., 231 (1984) 502-511. 9 Johanson, C.E., Ontogeny of the blood-brain barrier. In E.A. Neuwelt (Ed.), Clinical Implications of the BloodBrain Barrier and its Manipulation, Plenum, New York, in press. 10 Johanson, C.E. and Parandoosh, Z., Ontogenetic and methodological differences in the response of choroid plexus pH to acetazolamide, Soc. Neurosci. Abstr., 7 (1981) 86. 11 Johanson, C.E., Reed, D.J. and Woodbury, D.M., Developmental studies of the compartmentalization of water and electrolytes in the choroid plexus of the neonatal rat brain, Brain Res., 116 (1976) 35-48. 12 Johanson, C.E., Woodbury, D.M. and Withrow, C.D., Distribution of bicarbonate between blood and cerebrospinal fluid in the neonatal rat in metabolic acidosis and alkalosis, Life Sci., 19 (1976) 35-48. 13 Johnson, D.C., Hoop, B. and Kazemi, H., Movement of CO 2 and HCO 3 from blood to brain in dogs, Am. J. Physiol., 54 (1983) 989-996. 14 Jones, H.C. and Keep, R.F., The control of potassium concentration in the cerebrospinal fluid and brain interstitial
fluid of developing rats, J. Physiol. (London), 383 (1987) 441-453. 15 Madsen, J., Johanson, C.E. and Withrow, C.D., Extracellular fluid volume of several tissues in severely acidotic rats, Fed. Proc., 40 (1981) 470. 16 Mitchell, R.A., Herbert, D.A. and Carman, C.T., Acidbase constants and temperature coefficients for cerebrospinal fluid, J. Appl. Physiol., 20 (1965) 27-30. 17 Nattie, E.E. and Edwards, W.H.. CSF acid-base regulation and ventilation during acute hypercapnia in the newborn dog, J. Appl. Physiol. Resp. Environ. Exercise Physiol., 50 (1981) 566-574. 18 Parandoosh, Z. and Johanson, C.E., Ontogeny of the blood-brain barrier to and cerebrospinal fluid sink action on, [l~C]urea, Am. J. Physiol., Reg. lnteg., Comp. Sect.. 243 (1982) R400- R407. 19 Pershing, L.K. and Johanson, C.E., Acidosis-induced enhanced activity of the Na-K exchange pump in the in vivo choroid plexus: an ontogenetic analysis of possible role in cerebrospinal fluid pH homeostasis, J. Neurochem., 38 (1982) 322-332. 20 Ponten, O. and Siesjo, B.K., Acid-base relations in arterial blood and cerebrospinal fluid of the unanesthetized rat, Acta Physiol. Scand., 71 (1967) 89-95. 21 Reed, D.J., Withrow, C.D. and Woodbury, D.M., Electrolyte and acid-base parameters of rat cerebrospinal fluid, Exp. Brain Res., 3 (1967) 212-219. 22 Rollins, D.E. and Reed, D.J., Transport of DMO out of cerebrospinal fluid of rats, Am. J. Physiol., 219 (1970) 1200-1204. 23 Roos, A., Intraeellular pH and intracellular buffering power of the cat brain, Am. J. Physiol., 209 (1965) 1233-1246. 24 Roos, A., Intracellular pH and buffering power of rat brain, Am. J. Physiol., 221 (1971) 176-181. 25 Siesjo, B.K. and Messeter, K., Factors determining intracellular pH. In B.K. Siesjo and S.C. Sorensen (Eds.), Ion Homeostasis of the Brain, Munksgaard, Copenhagen, 1971, pp. 244-262. 26 Siesj6, B.K. and Ponten, U., Acid-base changes in the brain in non-respiratory acidosis and alkalosis. Exp. Brain Res., 2 (1966) 176-19(I. 27 Smith, Q.R. and Rapoport, S.I., Cerebrovascular permeability coefficients to sodium, potassium, and chloride, J. Neurochem., (1986) 1732-1742. 28 Smith, Q.R., Woodbury, D.M. and Johanson, C.E., Kinetic analysis of 36C1,22Na and [3H]mannitol uptake into the in vivo choroid plexus-cerebrospinal fluid system: ontogeny of the blood-brain and blood-CSF barriers, Dev. Brain Res., 3 (1981) 181-198. 29 Waddell, W.J. and Butler, T.C., Calculation of intracellular pH from the distribution of 5,5-dimethyl-2,4-oxazolidi-
2fl4 nedione (DMO). Application to skeletal muscle of dog, J. Clin. lnvest., 38 (1959) 720-729. 30 Withrow, C.D., Allen, J. and Johanson, C.E., Effects of metabolic acidosis or alkalosis on pH of blood, epithelia and muscle in immature rats, submitted. 31 Withrow, C.D., Woodbury, D.M. and Wilcox, W.D., Acid-base changes in brain and skeletal muscle of matur-
ing rats, Am. J. Physiol., 206 (1964) 521 524. 32 Woodbury, D.M., Johanson, C.E. and Brondsted, H., Maturation of the blood-brain and blood- cerebrospinal fluid barriers and transport systems. In E. Zimmermann and R. George (Eds.), Narcotics and the ft~7~othalamus, Raven. New York. 1974, pp. 225--247.
255
BRD 50684
Regulation of pH and H C O 3 in brain and CSF of the developing mammalian central nervous system Conrad E. Johanson, James Allen* and C. Dean Withrow Department of Neurosurgery, Brown University and Rhode Island Hospital, Providence, R102902 (U.S.A. ) and Department of Pharmacology, University of Utah School of Medicine, Salt Lake City. UT84132 (U.S.A.) (Accepted 11 August 1987) Key words." Cerebrospinal fluid pH; Cerebral cortex pH; Central nervous system acid-base regulation; Dimethadione method; Bicarbonate and pCO2; Metabolic acidosis and alkalosis; Ontogeny of cell pH; Cell H ion homeostasis
Acute (2-h) metabolic acidosis or alkalosis was induced in immature rats to ascertain the ability of their incompletely-developed CNS to regulate pH when challenged with perturbations in blood [H] and [HCO3]. Brain and cisternal CSF pH were determined from steady-state distribution of [14C]dimethadione, a weak organic acid. By 1 week post partum, there was a remarkable stability of pH in the cerebral cortex of animals subjected to arterial pH extremes of 7.1 and 7.5. However, CSF pH in 1-week-old animals rendered alkalotic remained 0.07-0.08 units above control due to lack of a compensatory increase in pCO 2, and to a blood-CSF barrier apparently more permeabie to HCO 3. As arterial HCO3, i.e. [HCO3]art, was varied from about 10 to 30 mmol/l, the infants maintained [HCO3J,f only half as effectively as adults, i.e. A[HCO3]csf/A[HfO3]art was 0.4 and 0.2 at l and >4 weeks, respectively. Throughout postnatal ontogenesis, [HCO3]csf was more resistant to alteration by metabolic acidosis than by alkalosis. Overall, the results indicate that immature rats challenged with systemic acid-base loads are less capable than adults in regulating CSF pH, but they are able to maintain brain pH.
oped CNS at birth. The rat is atypical among several laboratory mammals because it is born with a brain less mature than that of many other species. Several studies of neonatal rats l'3,7,~t'tS'28 have revealed
of the CSF secretory process 32, that may influence acid-base regulation. Previous acid-base investigations of infant rat CNS have been carried out primarily in 1-week-old animals t2'31. In this investigation we systematically analyzed acid-base balance in 1-, 2-, 3- and > 4 - w e e k - o l d rats. By measuring p H and total CO2 in C S F and the p H of cerebral cortex in response to acute metabolic acidosis (and alkalosis), we have assessed the regulatory capacity of the CNS at various ontogenetic
greater permeability of the b l o o d - b r a i n (BBB) and b l o o d - c e r e b r o s p i n a l fluid (CSF) barriers, larger content of total CO2 in brain tissue, and greater extraceilular space and water content in the i m m a t u r e CNS than in the mature brain. There are o t h e r characteristics of the immature rat brain, such as the postnatal proliferation of astroglia and d e v e l o p m e n t
stages. The response of [HCO3]cs f has also been analyzed in the context of the maturation of neural barrier systems to H C O 3. Since the time course of maturation of CSF secretion and astroglial function is known, it is possible to evaluate deductively the role of these factors in CNS acid-base regulation 32. The results for young rats indicate that although H C O 3 in
INTRODUCTION Few studies have addressed acid-base regulation in the immature central nervous system (CNS); most of these investigations have e m p l o y e d the neonatal lamb 4'5 or dog t7 which have a relatively well-devel-
,p
resent address: Searle Pharmaceuucals Inc., Searle Safety Assessment, 4901 Searle Parkway, Skokie, IL 60076, U.S.A. Correspondence." C.E. Johanson Department of Neurosurgery, Brown University and Rhode Island Hospital, 593 Eddy Street, Providence, RI 02902, U.S.A. 0165-3806/88/$03.50 © 1988 Elsevier Science Publishers B.V. (Biomedical Division)
25~ plasma gains access to CSF, there is remarkable regulation of brain pH after 2 h of metabolic distortions. CSF pH is less well maintained, especially in infant rats challenged with HCO~ loads. MATERIALS AND METHODS
Experimentalprotocol. The effect of 2-h metabolic acid-base distortions on the pH of arterial blood, CSF and brain was analyzed in immature Sprague-Dawley rats (1, 2, and 3 weeks) and compared to young adults (4-5 weeks). Ten or more animals were used for each treatment and age. Anesthesia was induced by ketamine (75-125 mg/kg), with the larger doses used in the older animals. Bristol Laboratories supplied ketamine as the HC1 salt, 100 mg/ml. The procedure for the injection of [14C]DMO (dimethadione), to indicate pH, is described elsewhere s. The radio-DMO (5,5-dimethyloxazoladine-2,4-dione, New England Nuclear; 0.4 mCi/mg) was injected intraperitoneally, (t.02 gCi/g. Salt solutions were injected intraperitoneally (i.p.) (0.01 ml/g) to induce the metabolic distortions: acidosis (0.3 mg NH4CI/g body wt.), control (0.1 mg NaCl/g), and alkalosis (0.8 mg NaHCO)/g). All salts were analytical grade from Sigma. A time-course preliminary study revealed that, following the injection of each salt, there was rapid induction (by 5 rain) of the acid-base disturbance which remained stable in blood for up to 4 h in the nephrectomized animals. Sampling of fluids and brain. At the end of each 2h experiment, blood was sampled (0.3-0.5 ml) from the abdominal aorta with a small heparinized syringe. Immediately thereafter the animal was positioned for CSF sampling 12. The dural membrane covering the cisterna magna was punctured with a 100-/al glass micropipet attached by tubing to a 1-ml syringe. CSF was withdrawn by gentle suction; if blood contamination occurred, the sample was discarded. Analysis of the CSF total CO 2 content was performed immediately after collection to minimize CO2 changes in the fluid, A 0.05- to 0.15-g sample of cortical tissue was sliced from the parietal lobe of both cerebral hemispheres. Meninges and residual blood were blotted to remove extraparenchymal elements from the brain sample. Tissues were weighed to 0.1 mg on a Sartorius analytical balance. Analyses of samples. The handling of arterial sam-
pies for blood gas and radioactivity analyses has been describe& s. Total CO 2 content in CSF was determined by gas chromatography in a Fisher Clinical Gas Partitioner. Between 10 and 20 f~l of CSF was used for each analysis. The CO: content of CSF was quantitated by comparing peak heights obtained with unknown samples to those obtained with equal volumes of a standard solution of NaHCOs (24.6 raM). CSF not analyzed for CO 2 was used for measurement of [14C]DMO activity. Prior to counting the l~C-activity by liquid scintillation, all of the samples of CSF, plasma and brain were solubilized r'< The DMO technique for pH. DMO as well as C(), distribution can be used to calculate pit and [HCO3]. Infant rat CSF samples are poorly buffered and small (10-15 hal), and thus are not suitable for reliable direct measurement of pH by electrode, ]'here is agreement among adult pH values for CSF, and brain, whether determined by DMO, CO2 distribution or pH electrodes s'2~'22"2a. Reed et al. 2~ showed that DMO distribution between arterial blood and CSF yielded a value agreeable with directly measured CSF pH, Similarly, Roos >'24 reported that un-ionized DMO distributes at equilibrium between CSF and arterial blood. Such findings indicate feasibility of DMO for pH measurement of CNS regions 2e, Calculations and statistical analyses. CSF pH was calculated from the steady-state distribution of [14C]DMO29, and alternatively from the distribution of HCO 3 and CO225'26. CNF pH was determined from steady-state distribution of [14ClDMOS'm, the percent volume of distribution (Vdt of which was calculated as: 100 × dpm/mg CSF water/dpm/~l plasma water Values for arterial pH and total activity of [14C]DMO were used to calculate concentrations of ionized and un-ionized [14C]DMO in plasma. The formulae and assumptions for calculating the ratio of ionized to unionized [~4C]DMO in brain tissue and hence cell pH, have been previously presented s. Due to the limiting effect of the BBB on ion distribution, the concentration of ionized [14C]DMO in extracellular fluid (ECF) of brain was taken as that in CSF, rather than in plasma s. The H : O contents of plasma and brain for various ages were determined as the difference be-
257 tween wet and dry weights of samples. Values for rat cerebral cortical ECF volume (% of wet tissue wt.) of 22.5, 16, 14.5 and 13.5%, respectively, were utilized in cortical pH calculations for 1-, 2-, 3- and >4-weekold animals 28. Neither metabolic acidosis (NH4CI) nor alkalosis (NaHCO3) alters ECF volume 3°. Severe respiratory acidosis, e.g. 20% CO2 to bring about arterialpCO 2 > 100 torr, does not change ECF volume ~5. Arterial [HCO3] was calculated from electrodemeasured pH and pCO2 values, by use of the Henderson-Hasselbalch equation (pKa = 6.10, aCO 2 = 0.030). C S F [ H C O 3 ] w a s determined as the difference between total CO 2 and carbonic acid (i.e. apCO2) where a for CSF at 37 °C is 0.03216. CSF pCO 2 was derived by the method of Brzezinski et al. 2, who systematically described the pCO2 of cisternal CSF as a function of arterial pCO 2. Electrode measurements have confirmed that brain pCO 2 can be reliably estimated as arterial pCO2 + 6 torr 2. For brevity, [HCO3] in mmol/kg H20 is expressed as mM. One-way analysis of variance and the multiplerange test (Tukey version) were used to test for significant differences (P < 0.05) among age groups. The paired Student's t-test was also utilized in CSFplasma analysis.
stantially different in respect to pH, p C O 2 and [HCO3]; see Fig. 1. The respiratory acidosis nature of blood in immature animals converted between 1 and 2 weeks to adult-like composition. The utilized doses of NH4C1 and NaHCO3 caused comparable pH deflections of 0.10-0.15 pH units in arterial blood (Fig. 2). Except for 1-week-old animals, all stages of development displayed the typical adult-like response of CO 2 elimination ($pCO2) in metabolic acidosis and CO 2 retention ( I' pCO2) in aikalosis. CSFand brainpH values. In controls, CSF pH (derived from DMO distribution) did not change significantly over the developmental period studied (1-5 weeks); pH values were usually in the range of 7.30 + 0.04 (Table I). CSF pH was also calculated from the distribution ratio of HCO 3 to CO2, by using data for total CO 2 (Table III) and CSF pCO2. In l-, 2-, and 3-week-old rats, there were generally no statistically significant differences between DMO- and CO2-derived values for CSF pH (in either control, acid- or base-loaded animals). At >4 weeks, the CO: method tended to yield CSF pH values greater than those by DMO. Table II contains an ontogenetic analysis of acid-base distortion-induced alterations in CSF pH relative to those in arterial pH. In metabolic acidosis (2 h) the ratio ofACSF pH to d arterial
ARTERIAL BLOOD
RESULTS
E~ METABOLIC ACIDOSIS r--] METABOLIC ALKALOSIS
Baseline and experimental values for arterial parameters. The composition of arterial blood in infant
U u. U
rats (1 week), compared to young adults, was sub-
pH
pCO2(torr)
[HCO~ (mM) & pCO2 (From control )
25
"'7.5
15
20t
~71 7.0
I
I
I0
I
2
5 >4
2
5
>4
POSTNATAL AGE (weeks)
0 I
2
5 >4
0j
I
2
3>4
POSTNATAL AGE (weeks) Fig. 1. Ontogenetic changes in the baseline composition of arterial blood acid-base parameters. Each bar is the mean + S.E.M. for at least 11 rats. All controls received NaCI, i.p., 0.1 mg/g body weight, 2 h before sacrifice. *P < 0.05, 1 or 2 weeks vs >4 weeks, by multiple-range test.
Fig. 2. Developmental analysis of the magnitude of distortioninduced changes (from corresponding values for NaCl-injected controls) in pH and pCO 2 (torr) of arterial blood in rats subjected for 2 h either to metabolic acidosis (0.3 mg/g) or to alkalosis (0.8 mg/g). Each bar is the mean + S.E.M. for 11 or more animals. *P < 0.05, by multiple-range test, that the experimental alteration is significantly different from corresponding control.
25S ]ABLE 1 p H ()f cerebrospinal fluid and cerebral corte_r in developing rats challenged with metabolic acidosis or alkalosis Each mean _+ S.E.M. for CSF pH is for 10-12 animals; for cerebral cortex data, n = 12. CSF and brain samples were taken from the cisterna magna and frontoparietal cortex, respectively. Cerebral cortex pH
(erebrospinal fluid pH 1 week
2 weeks
3 weeks
>4 weeks
I week
Metabolic acidosis Control
7.25 ± (I.(12 7.26 _+ 0.02
7.21 ± 0.02* 7.29 _+ 0.02
7.31 +- 0.02 7.34 + (I.04
7.26 +- 0.03 7.26 ± 0.03
Metabolic alkalosis
7.33 +__0.03*
7.27 +__0.01
7.34 + 0.(13
7.30 + 0.03
2 weeks
3 week,~
>4 weeks
7.14 +- 0.02 ~ 7.03 + 0.02 7.12 +__0.02 7.08 +- 0.03
7.(19L 0.03 7.12 -+_1!.02
7.02 ± 0.03 7.07 +__0.02
7.13 __+0.04
7.10 ± 0.t13
7,06 +__0,03
7.08 +__0.01
* P < 0.05, control vs acidosis or alkalosis. * P < 0.05, 1 week vs >4 weeks, by multiple-range test. pH values were determined from the steady-state distribution of [laC]dimethadione, Treatments for 2 h are described in Materials and Methods.
p H was only slightly g r e a t e r at 1 and 2 w e e k s than at 3
than that in animals at 2 w e e k s (21.4), 3 w e e k s (20.3)
and > 4 weeks. A C S F p H / A arterial p H in m e t a b o l i c
or > 4 w e e k s (22.7); see T a b l e l l I . T h e r e s p o n s e of
alkalosis was substantially g r e a t e r at 1 w e e k ((1.68)
the C S F system to a c i d - b a s e c h a l l e n g e v a r i e d accord-
than in adults (close to z e r o ) : w d u e s at 2 and 3 w e e k s
ing to age as well as distortion. T h e l - w e e k - o l d rat
were i n t e r m e d i a t e ,
e x h i b i t e d a significant d e c r e a s e (by 3.4 m m o l ) and in-
but closer to the adult
level
(Table II).
crease (by 3.6 retool) in total CO2 d u r i n g m e t a b o l i c
T h e p H v a l u e s for c e r e b r a l c o r t e x t e n d e d to be
acidosis and alkalosis, r e s p e c t i v e l y , f r o m a c o n t r o l
m o r e alkaline in the less m a t u r e a n i m a l s ( T a b l e I).
value of 28 m m o l / l . M e t a b o l i c alkalosis c a u s e d a sta-
N e i t h e r 2-h m e t a b o l i c acidosis n o r alkalosis c a u s e d a
tistically significant a u g m e n t a t i o n in C S F total CO~
significant a l t e r a t i o n in cortical p H at any age.
at 2, 3, and > 4 w e e k s ; in c o n t r a s t , C S F in animals at
Cerebrospinal fluid content o f total C O 2 and bicar-
these s a m e 3 stages of d e v e l o p m e n t was p r o t e c t e d
bonate. In c o n t r o l rats at 1 w e e k the C S F had a
from a substantial fall in total C O z d u r i n g m e t a b o l i c
m a r k e d l y h i g h e r c o n t e n t of total CO2 (28 mmol/I)
acidosis for 2 h ( T a b l e III).
TABLE 11 Ontogenetic analysis of acid-base distortion-induced alterations in CSF pH relative to those in arterial pH Average change refers to difference between mean control and mean experimental values. Since there were generally no statistically significant differences between CSF pH values, DMO vs CO: methods, the overall mean was determined by averaging pooled CSF values. Postnatal age 1 week
2 weeks
3 weeks
;>4 weeks
Metabolic acidosis Average change in CSF pH units Average change in arterial pH units A CSF pH/A arterial pH
-0.015 -0.13 0.12
-0.030 -0.17 0.18
+0.010" -0.10 0.10
+0.005* "0.17 0.(/3
Metabolic alkalosis Aver age change in CSF pH units Average change in arterial pH units A CSF pH/A arterial pH
+ 0.075 +0.11 0.68
+ 0.0 l 0 +0.09 0.11
+ 0.015 +0.13 0.12
it +0.10 0
* CSF pH increase was in the opposite direction to the change in arterial pH, but it was not significantly elevated above the control value for CSF pH.
259 T A B L E III
Concentration of total carbon dioxide in the cisternal cerebrospinal fiuid of rats at various stages of development Each m e a n + S.E.M. is for 10 or more animals. Control rats were injected with NaCI solution; metabolic acidosis and alkalosis, respectively, were induced with NH4CI and N a H C O 3 for 2 h. CO 2 concentration was determined by gas chromatography.
Treatment
Concentration of CO e (mmol/l CSF)
Metabolic acidosis (NH4CI) Control (NaCI) Metabolic alkalosis (NaHCO3)
1 week
2 weeks
3 weeks
>4 weeks
24.6 _+ 0.8* 28.0 + 1.0 31.6 _+ 0.7*
20.0 __+0.6 21.4 _+ 0.7 26.7 + 1.4"
19.3 + 0.6 20.3 _+ 0.4 25.8 + 1.5"
22.0 _+ 0.9 22.7 _+ 0.4 25.9 _ 1.3'
* A statistically significant difference (P < 0.05, multiple-range test) between control and treated animals within an age group.
The mean concentration of HCO 3 in CSF relative to plasma is presented according to age and treatment in Fig. 3. In controls, CSF [HCO3] was generally not significantly different from plasma [HCO3], except at 1 week. At all ages investigated, CSF [HCO3] was significantly greater than plasma [HCO3] after 2 h of acidosis. In metabolic alkalosis, the [HCO3] in CSF remained significantly lower than that in plasma only in the older animals (3 and >4 weeks).
30
[ ] CSF [ ] PLASMA
***
2O
The relationship between CSF [HCO3] and plasma [HCO3] is given on an individual basis in Fig. 4; for each age, there is a positive linear relationship between the [HCO3] in the two fluids. The slope value at 1 week was 0.40 + 0.08 (S.E.M.) and it progressively decreased with age to values of 0.38 + 0.07, 0.23 _+ 0.05, and 0.20 + 0.06 at 2, 3 and >4 weeks, respectively; the least squares regression for 1-week data is significantly different from that for >4 weeks. Brain cell [HCO3] was calculated to be 17 mM at 1 weele, and it declined to 11-12 mM at 2, 3 and >4 weeks. The cellular HCO 3 was not significantly affected by the acid-base disturbances at any stage of development.
I0
}
DISCUSSION
Developmental changes" in baseline values. In adult
~o-t 401 C. ALKALOSIS
I
2
3
>4
POSTNATAL AGE (WEEKS)
Fig. 3. Relationship between the concentration of H C O 3 in plasma and that in CSF during metabolic acidosis (Part A) and alkalosis (C). Data for control animals (B) are included for comparison. Rats of 4 postnatal ages are represented. CSF was sampled from cisterna m a g n a and blood was collected from abdominal aorta. Each bar is the m e a n + S.E.M. for at least 10 animals. Statistical analysis was performed with paired t-test, CSF vs plasma. ***P < 0.001, **P < 0.01, *P < 0.05.
mammals, CSF [HCO3] is approximately equal to plasma [HCO3] s. During early development in the rat (1 week), CSF [HCO3] and arterial [HCO3] are both elevated several mM above corresponding adult control levels. The substantial attenuation of CSF [HCO3] by 5 mM between the 1st and 2nd weeks in controls is accompanied by a comparable decrease in [HCO3] in plasma and brain cells. Thus, the HCO3 concentration gradients among the 3 compartments appear not to be substantially different in maturation vs early rapid growth. The 1-week-old rat cortical pH values tended to be more alkaline than adult values 8 which are close to 7.0. Thus, the higher pHs in infants for brain, and for non-neural tissues 3°, likely reflect greater intraceilular [HCO3]. In controls, the lack of significant change in CSF pH with age indicates a preferred value at which CSF
26i~
40] I WEEK
2 WEEKS I
35l
o
To
25
•
.j
4
.2
° . L
]Z
20
sE 15 E
n- 5 ~
5
1
IO 1% 20
---
-T---'-T--"
I
is
4s s
20
sb
3~5
4O
45
] ~4 WEEKS
d 40- 3 WEEKS Z
h-
~ 30 Ill m
~
•
40 45 5
10
~25
] .-
......
20 •
A•
15Z t0 55
©
15 20
25
30
35
15
20
25
30
35
40
45
ARTERIAL PLASMA [HCO~](mmol/I.H20) Fig. 4. Correlation between the concentration of H C O 3 in cisternal CSF and that in arterial plasma. Each datum point corresponds to the concentrations of H C O 3 in the plasma and CSF of a single rat. The symbols represent the 3 treatments: NH4CI (liD), NaCI (A), NaHCO 3 (O). Each panel depicts the relationship for a given age. Slopes, determined by least-squares best fit, of 0.40 _+ 0.08 (S.E.M.), 0.38 +_0.07, 0.23 + 0.05 and 0.20 + 0.06 were calculated for the I-, 2-, 3- and >4-week-old animals, respectively. By analysis of covariance, there is a significant difference (P < 0.05) between the slope value at 1 week compared to that at >4 weeks.
p H is regulated. Consistent with this view, the CSF system throughout development continues to regulate p H at about 7.25 when challenged with arterial acidosis. While CSF p H remains stable after birth, the pH of arterial blood continually rises thereby gradually setting up the p H gradient of 0.1 unit (CSF more acidic) across the adult mammalian barrier (Fig. 5). That there is no appreciable p H gradient at 1 week is consistent with the more permeable barriers 7J2'~s'2s and perhaps immaturity of systems translocating ions in the incompletely developed choroid plexus 28. It seems reasonable to hypothesize that zl/~ H ÷ (i.e. electrochemical potential difference between blood and CSF) is close to zero in neonatal rats. Since Ap H ÷ is normally about +5 to 10 mV in adults, the postnatal rat is a promising developmen-
tal model for pursuing the nature of H-ion distribution across neural barrier systems.
Ontogenetic differences in stability of CSF pH in acute distortions. CSF p H is not well-regulated in immature rats subjected to metabolic alkalosis. In the present experiment, CSF pH in 1-week-old animals loaded with HCOs (and anesthetized with ketamine) increased significantly by 0.07 and 0.08 units, from [MC]DMO and CO 2 data, respectively; one cause for this alkaline shift was insufficiently retained CO23° to counteract the substantially elevated CSF [HCO3]. In a similar acute study, of infa,it rats 12. but with a different anesthetic (pentobarbital) and experimental duration (1 h), CSF p H in metabolic alkalosis was also increased from 7.27 (control) to 7.40: the greater CSF pH augmentation in the latter case Iz occurred
261
*0•18-
*0.15-
0
*0.12-
13
(/3
o *0.09I 0
+0.06*0.03-
o • 0
,,=::1
O-i
0 •
8
• 0
0
-0.03adult POSTNATAL AGE (weeks)
Fig. 5. Gradient of pH across the blood-CSF barrier in control rats as a function of postnatal age. Blood and CSF were sampled from abdominal aorta and cisterna magna, respectively. Symbols refer to investigations in which CSF pH was calculated from the distribution of CO_, (filled) and DMO (unfilled), respectively. Each circle represents a mean value (n = 10) from the present investigation for ketamine-anesthetized rats; A, ref. 12, n = 14 pentobarbital-anesthetized rats; V, V, ref. 21. n = 10. ether anesthesia; ~ , ref. 20, n = 6 unanesthetized rats; []• II, ref. 21. n = 10 pentobarbital-anesthetized rats: ,o . ref. IlL pentobarbital and ether, respectively. because the 24% rise in p C O 2 secondary to alkalosis had not counteracted the 50%-elevated [HCO3] in the CSF 1 h after the beginning of the distortion. Thus, two investigations have shown that immature rat (1-week) CSF does not defend pH effectively in acute metabolic alkalosis. However, by 2-3 weeks, CSF displayed pH homeostasis when challenged with non-respiratory alkalosis, presumably since: (1) barrier systems at blood interfaces are more impermeable 7, (2) central and/or peripheral chemoreceptors are capable of adjusting pCO2, and (3) choroid plexus secretion TM is more developed (hence greater buffering ability). With respect to CSF pH regulation, metabolic acidosis appears less problematic to immature rats (1 week) than metabolic alkalosis 12. We confirmed that CSF pH is unaltered in animals injected with NH4C1 at 1 week, such protection presumably related to relatively high normal baseline concentration of HCO3 in brain cells, cerebral interstitial fluid, CSF and choroid plexus t°'31. Thus, especially at 1 week, the concentration gradient for H C O 3 (barrier cell, or brain cell, to CSF) in metabolic acidosis favors net diffusion of HCO 3 into CSF replacing H C O 3 depleted by
NH4CI treatment. The substantially reduced baseline [HCO3] in these CNS compartments at 2 weeks (vs 1 week) may explain why CSF pH was not maintainable even in acidotic animals• CSF pH maintenance at 3 weeks and later, in both distortions, coincides with nearly complete development of BBB restrictiveness, astrocyte function, and choroid plexus barrier and transport systems for ion homeostasis and fluid secretion. Intracellular p H o f cerebral cortex• Cortex pH is more alkaline in immature than in mature animals s. Similarly, cell pH in extra-CNS tissues (liver, muscle, etc. 3°) is also more basic in infants than adults• Alkaline pH i is apparently conducive to growth processes. Interestingly, the relative alkalinity of cells in immature animals occurs in the face of elevated levels of tissue pCO2; in this regard, arterial p C O 2 is 20 torr higher at 1 week vs >4 weeks. The brain pH i in young rats was not altered by metabolic acidosis or alkalosis (Table I). Regulation of pH i in adult mammalian brain has been delineated TM 24,26, but the ability of immature brain to maintain pH in face of variations in [H] and [HCO3] in blood and CSF has apparently not been investigated. Siesjo and Messeter 25 concluded that adult brain cells regulate pH i by organic acid production or consumption, physicochemical buffering and active transport. There may be maturational differences in ability to buffer by virtue of cell protein content or capacity to metabolize acids like lactate. Because the untreated week-old rat has relatively high [HCO3] in brain cells 3°'3t, this likely contributes to H-buffering during acid challenges to neurons. That infant rats actively translocate H and/or H C O 3 can be argued from control pHs. With a CSF (i.e. interstitial fluid) pH of 7.26 and cerebral cortex pH i of 7.12 at 1 week (Table I), the Nernst equation predicts that E H is - 9 mV at 37 °C. By deduction, H-ions must distribute against an electrochemical gradient, i,e. be actively extruded, because neurons in infants would not function adequately with such a low membrane potential. Awaiting elucidation is the role of secondary active transport ( N a - H and C1-HCO3 exchange) during perturbations of the milieu of the perinatal CNS. Assumptions and limitations in calculations. D M O usage to measure cell pH in brain (pHi) was scrutinized by Roos 23'24for effects of theoretical variations in membrane potential and ECF volume. His find-
ings of reliability are pertinent to this study in which potential difference (PD) and ECF volume in brain vary significantly with age. The absolute value of pH i is not particularly sensitive to error in brain ECF volume measurements: even so, our emphasis has been on experimental changes in pH i, so any error in ECF volume (for a given age) should not seriously affect relative wtlues for control vs treated groups. Accurate calculation of [HCO3] from CO: data depends on a sound value for pCO2. CO, distribution among plasma, CSF and brain tissue in rats was studied by Siesjo and co-workers 2'>, who established with electrodes that arterial pC02 data can be used to estimate accurately the p C O 2 of CSF and cerebral cortex. Changes in CSF and brain [HCO ff with maturation and treatment. Infant rats (1 week) maintained CSF [HCO3] half as effectively as young adults in metabolic distortions; compare slopes in Fig. 4. One explanation is that barrier systems (components of transport and permeability) are less developed in the youngest rats 7'~s'2s and consequently transmit changes in plasma [HCO3] more extensively to CSF. However, it should be emphasized that in 1-week rats the A[HCO3]c,f/A[HCOs]~r t at 2 h was only 0.4; thus, even with incompletely-developed barriers, the distortion in [HCO3]art was not fully transmitted to the CSF. At all ages, CSF [HCO3] was more resistant to alteration in acute metabolic acidosis than in alkalosis. This supports the hypothesis that brain cell HCO 3 is a buffer reservoir partially to replace CSF [HCO3] titrated downward in acidosis. Thus, brain [ H C Q l i (highest in neonates) may help to maintain CSF [HCO3] in the face of metabolic acidosis, thereby counteracting potentially deleterious effects of leaky barriers to HCO3 and H early in development. Even if brain cells supply HCO3 to buffer CSF, they maintain their own [HCO3]. Of several tissues in infants s°, brain cell [HCQ] was generally the most stable in distortions. This is interesting since, ontogenetically, glial (and thus carbonic anhydrase) proliferation has not yet occurred 1 week after birth 32. By deduction this intimates that neurons (deficient in carbonic anhydrase) in infants have effective pH homeostatic ability. Another possible factor affecting stability of [HCO3],is fluctuation in interstitial fluid H C O 3, i.e. [HCO3]ist, determined by [HCOs]cs f (already discussed) and by H C O 3 flUX across the BBB.
Plasma HCO 3 significantly penetrates adult BBB ~:~. and it is likely that movement of HCO 3 and other solutes 7"js2~ is even less restricted across infant cerebral capillaries 7'~, Yet, there must be some impediment to HCO3 movement across infant BBB, as is the case for non-electrolytes 7'~'~s'~s. Thus, even after 2 h of distortions in infants, one expects that Zl[HCO3]isf/Zl[HCO3]ar > similar to the case for ,d[HCO3]csf/,d[HCO3lar t, would be less for brain than other tissues. Non-neural parenchyma >, unprotected by barrier systems, are subjected extracellularly to the full brunt of alterations in [HCO3]~,. Thus, the BBB and choroid plexus, though more permeable in infants than in adults, help by some barrier effect to confer stability to brain phi and [HCO31i in 1-week animals challenged with H / H C O 3 loads. Blood-CSF barrier vs BBB. In adult rats, the blood-CSF barrier epithelium is leakier than the BBB endothelium. Ions and small non-electrolytes, like 22Na, 42K and mannitol (tool. wl. = 181), have permeability coefficients in choroid plexus at least 10-100 times greater than those in cerebral capillaries 27. Urea and HCO3, with comparable molecular weights of 60 and 61, respectively, permeate the choroidal membrane in infants more readily than m adults 12a8. Neonatal rats rendered hyperkalemic have stable [K]isf, but a marked rise in [K]cs6 this demonstrates that developmental barrier tightening to K in choroid plexus lags the BBW 4. Collectively, the information indicates that the parenchyma in postnatal rat brain is protected against acidosis and hyperkalemia, in spite of the leaky blood-CSF barrier. Perspective. The CNS in the rat one week after birth, with its somewhat leaky neural barrier systems 79'1s, incompletely developed astroglial and choroid plexus transport systems ~'~3:, and minimal CSF sink action ~s, can protect the pH of its cellular elements against acute systemic metabolic acid-base insults. That even the neonate regulates brain cell pH so vigorously suggests that stability of this physiological parameter is of prime importance to neurons during development. The lesser ability of CSF in infants (1 week) to defend pH, ~specialty against alkaIosis, is attributed to substantial penetration of HCO 3 across the incompletely-developed choroid plexus ~> ~8,2s. On the other hand, the effective impermeability of the external limiting membranes of brain paren-
263 c h y m a l e l e m e n t s ( n e u r o n s ) , e v e n in infants, assures
G r e e n w o o d for c o n t r i b u t i o n s to the p r o j e c t . This
that p H i and [HCO3] i are not a l t e r e d in the face of
w o r k was s u p p o r t e d by U S P H S G r a n t 1 R01 N S
moderate
displacements
of
cortical
extracellular
[HCO3] , i.e. 2 0 - 4 0 % of arterial fluctuations.
13988,
NINCDS
and
a Research
Career
Devel-
o p m e n t A w a r d to C . E . J . This p r o j e c t was also partially s u p p o r t e d by N I H g r a n t G M 07579 and by
ACKNOWLEDGEMENTS
funds f r o m the U n i v e r s i t y of U t a h R e s e a r c h C o m m i t tee.
W e w o u l d like to t h a n k W a y n e B a m o s s y and J e a n
REFERENCES 1 Amtorp, O. and Sorensen, S.C., The ontogenetic development of concentration differences for protein and ions between plasma and cerebrospinal fluid in rabbits and rats, J. Physiol. (London), 243 (1974) 387-400. 2 Brzezinski, J., Kjallquist, A. and Siesjo, B.K., Mean carbon dioxide tension in the brain after carbonic anhydrase inhibition, J. Physiol. (London), 188 (1967) 13-23. 3 Ferguson, R.K. and Woodbury, D.M., Penetration of [~4C]inulin and [14C]sucrose into brain, cerebrospinal fluid, and skeletal muscle of developing rats, Exp. Brain Res., 7 (1969) 181-194. 4 Herrington, R.T., Harned, H.S., Jr., Ferreiro, J.I. and Griffin, C.S., The role of the central nervous system in perinatal respiration: studies of chemoregulatory mechanisms in the term lamb, Pediatrics, 47 (1971) 857-864. 5 Hodson, W.A., Fenner, A., Brumley, G., Chernick, V. and Avery, M.E., Cerebrospinal fluid and blood acid-base relationships in fetal and neonatal lambs and pregnant ewes, Resp. Physiol., 4 (1968) 322-332. 6 Johanson, C.E., Choroid epithelial cell pH, Life Sci., 23 (1978) 861-868. 7 Johanson, C.E., Permeability and vascularity of the developing brain: cerebellum vs cerebral cortex, Brain Res., 190 (1980) 3-16. 8 Johanson, C.E., Differential effects of acetazolamide, benzolamide and systemic acidosis on hydrogen and bicarbonate gradients across the apical and basolateral membranes of the choroid plexus, J. Pharmacol. Exp. Ther., 231 (1984) 502-511. 9 Johanson, C.E., Ontogeny of the blood-brain barrier. In E.A. Neuwelt (Ed.), Clinical Implications of the BloodBrain Barrier and its Manipulation, Plenum, New York, in press. 10 Johanson, C.E. and Parandoosh, Z., Ontogenetic and methodological differences in the response of choroid plexus pH to acetazolamide, Soc. Neurosci. Abstr., 7 (1981) 86. 11 Johanson, C.E., Reed, D.J. and Woodbury, D.M., Developmental studies of the compartmentalization of water and electrolytes in the choroid plexus of the neonatal rat brain, Brain Res., 116 (1976) 35-48. 12 Johanson, C.E., Woodbury, D.M. and Withrow, C.D., Distribution of bicarbonate between blood and cerebrospinal fluid in the neonatal rat in metabolic acidosis and alkalosis, Life Sci., 19 (1976) 35-48. 13 Johnson, D.C., Hoop, B. and Kazemi, H., Movement of CO 2 and HCO 3 from blood to brain in dogs, Am. J. Physiol., 54 (1983) 989-996. 14 Jones, H.C. and Keep, R.F., The control of potassium concentration in the cerebrospinal fluid and brain interstitial
fluid of developing rats, J. Physiol. (London), 383 (1987) 441-453. 15 Madsen, J., Johanson, C.E. and Withrow, C.D., Extracellular fluid volume of several tissues in severely acidotic rats, Fed. Proc., 40 (1981) 470. 16 Mitchell, R.A., Herbert, D.A. and Carman, C.T., Acidbase constants and temperature coefficients for cerebrospinal fluid, J. Appl. Physiol., 20 (1965) 27-30. 17 Nattie, E.E. and Edwards, W.H.. CSF acid-base regulation and ventilation during acute hypercapnia in the newborn dog, J. Appl. Physiol. Resp. Environ. Exercise Physiol., 50 (1981) 566-574. 18 Parandoosh, Z. and Johanson, C.E., Ontogeny of the blood-brain barrier to and cerebrospinal fluid sink action on, [l~C]urea, Am. J. Physiol., Reg. lnteg., Comp. Sect.. 243 (1982) R400- R407. 19 Pershing, L.K. and Johanson, C.E., Acidosis-induced enhanced activity of the Na-K exchange pump in the in vivo choroid plexus: an ontogenetic analysis of possible role in cerebrospinal fluid pH homeostasis, J. Neurochem., 38 (1982) 322-332. 20 Ponten, O. and Siesjo, B.K., Acid-base relations in arterial blood and cerebrospinal fluid of the unanesthetized rat, Acta Physiol. Scand., 71 (1967) 89-95. 21 Reed, D.J., Withrow, C.D. and Woodbury, D.M., Electrolyte and acid-base parameters of rat cerebrospinal fluid, Exp. Brain Res., 3 (1967) 212-219. 22 Rollins, D.E. and Reed, D.J., Transport of DMO out of cerebrospinal fluid of rats, Am. J. Physiol., 219 (1970) 1200-1204. 23 Roos, A., Intraeellular pH and intracellular buffering power of the cat brain, Am. J. Physiol., 209 (1965) 1233-1246. 24 Roos, A., Intracellular pH and buffering power of rat brain, Am. J. Physiol., 221 (1971) 176-181. 25 Siesjo, B.K. and Messeter, K., Factors determining intracellular pH. In B.K. Siesjo and S.C. Sorensen (Eds.), Ion Homeostasis of the Brain, Munksgaard, Copenhagen, 1971, pp. 244-262. 26 Siesj6, B.K. and Ponten, U., Acid-base changes in the brain in non-respiratory acidosis and alkalosis. Exp. Brain Res., 2 (1966) 176-19(I. 27 Smith, Q.R. and Rapoport, S.I., Cerebrovascular permeability coefficients to sodium, potassium, and chloride, J. Neurochem., (1986) 1732-1742. 28 Smith, Q.R., Woodbury, D.M. and Johanson, C.E., Kinetic analysis of 36C1,22Na and [3H]mannitol uptake into the in vivo choroid plexus-cerebrospinal fluid system: ontogeny of the blood-brain and blood-CSF barriers, Dev. Brain Res., 3 (1981) 181-198. 29 Waddell, W.J. and Butler, T.C., Calculation of intracellular pH from the distribution of 5,5-dimethyl-2,4-oxazolidi-
2fl4 nedione (DMO). Application to skeletal muscle of dog, J. Clin. lnvest., 38 (1959) 720-729. 30 Withrow, C.D., Allen, J. and Johanson, C.E., Effects of metabolic acidosis or alkalosis on pH of blood, epithelia and muscle in immature rats, submitted. 31 Withrow, C.D., Woodbury, D.M. and Wilcox, W.D., Acid-base changes in brain and skeletal muscle of matur-
ing rats, Am. J. Physiol., 206 (1964) 521 524. 32 Woodbury, D.M., Johanson, C.E. and Brondsted, H., Maturation of the blood-brain and blood- cerebrospinal fluid barriers and transport systems. In E. Zimmermann and R. George (Eds.), Narcotics and the ft~7~othalamus, Raven. New York. 1974, pp. 225--247.