Cerebral Acid Buffering Capacity at Different Ages Measured In Vivo by 31 P and 1 H Nuclear Magnetic...

Journal oJNeurochemistry Raven Press, Ltd., New York 0 I992 International Society for Neurochemistry

Cerebral Acid Buffering Capacity at Different Ages Measured In Vivo by 31P and 'H Nuclear

Magnetic Resonance Spectroscopy

Ronald J. T. Corbett, Abbot R. Laptook, Damian Garcia, and Joan I. Ruley

Ralph Rogers and Mary Nell Magnetic Resonance Center, Department of Radiology; and Department of Pediatrics, University of Texas Southwestern Medical Center at Dallas, Dallas, Texas, U.S.A.

Abstract: Cerebral acidosis occumng during ischemia has been proposed as one determinant of tissue damage. New- born animals appear to be less susceptible to ischemic tissue damage than adults. One possible component of ischemic tolerance could derive from maturational differences in the extent of acid production and buffering in newborns compared to adults. The purpose of this study was to measure the dependency of acid production on the blood plasma glucose concentrations and acid buffering capacity of piglets at different stages of development. Complete ischemia was induced in 29 piglets ranging in postconceptual age from 1 1 1 to 156 days (normal term conception, 1 15 days). Brain buffering capacity during the first 30 min of ischemia was quantified in vivo, via and 'H nuclear magnetic resonance (NMR) spectroscopy, by measuring the change in intracellular brain pH for a given change in the concentration of compounds that contribute to the production of hydrogen ions. Animals from all four age groups showed a similar linear correlation between preischemia blood glucose concentration and intracellular pH after 30 min of ischemia. For each animal the slope of the plot of intracellular pH versus cerebral buffer base deficit was used to calculate the buffer capacity. Using data obtained over the entire 30 rnin of ischemia, there was no difference in the mean buffer capacity of the different age groups, nor was there a signifi-

cant correlation between buffer capacity and age. However, there was a significant increase in buffer capacity for the intracellular pH range 6.6-6.0, compared to 7.0-6.6, for all age groups. No significant differences in buffer capacity for these two pH ranges were observed between any of the age groups. Acid buffering capacity was also measured by performing pH titrations on brain tissue homogenized in the presence of inhibitors of glycolysis and creatine kinase. Plots of homogenate pH versus buffer base deficit showed a nonlinear trend similar to that seen in vivo, indicating an increase in buffer capacity as intracellular pH decreases. A comparison of newborn and I-month-old brain tissue frozen under control conditions or after 45 min of ischemia revealed no differences that could be attributed to age and a slight decrease in buffer capacity of ischemic brain compared to control brain tissue homogenates. There was no difference between the brain buffering capacity measured in vivo using "P and 'H NMR and that measured in vitro using brain homogenates. Key Words: Neonatal brain pH-Ischemia-Acid buffering-Nuclear magnetic resonance spectroscopy-Brain development. Corbett R. J. T. et al. Cerebral acid buffering capacity at different ages measured in vivo by "P and 'H nuclear magnetic resonance spectroscopy. J. Neurochem. 59,2 16-226 ( 1992).

The phenomenon of greater tolerance toward hypoxic ischemia-induced brain injury in newborns compared to adults is not fully understood (Volpe, 1987). One aspect that remains unclear is whether or not an elevated blood glucose concentration plays a deleteri-

ous or beneficial role in neonates exposed to hypoxic ischemia. In adults, elevated glucose prior to ischemia has been shown to be detrimental, most likely because of enhanced anaerobic glycolysis of glucose to lactate, with parallel increased acidosis (Helgason, I 988).

Received July 24, 199 I ; revised manuscript received December I 6, I99 I ; accepted December 19, I99 1.

Address correspondence and reprint requests to Dr. R. Corbett at Department of Radiology, University of Texas Southwestern Medi- cal Center at Dallas, 5801 Forest Park Road, Dallas, TX 75235- 9085, U.S.A.

Abbreviations used: ANOVA, analysis of variance; @, buffer capacity (millimolar base deficit per pH unit); ABB, change in buffer base (or buffer base deficit); FID, free induction decay; [glu- C O S ~ ] ~ ~ ~ ~ ~ ~ ~ arterial plasma glucose concentration; HCT, hemato-

crit; [La], brain lactic acid concentration; [lactate],,, final brain lactate concentration; [la~tate],~,,,,, arterial plasma lactate concentration; MANOVA, multivariant analysis of variance; MAP, mean arterial blood pressure; NMR, nuclear magnetic resonance; NTP, nucleoside triphosphate; P,, inorganic phosphate; PCA, perchloric acid; Pco,, arterial blood CO, tension; PCr, phosphocreatine; pH.. arterial blood pH; pH,, pH measured from an electrode inserted into the brain homogenate; pHi, intracellular brain pH (or pH measured from the chemical shift of Pi in the "P NMR spectrum after ischemia); Po,, arterial blood 0, tension; Ptco,, tissue PCo,.

216

CEREBRAL ACID BUFFERING CAPACITY AND AGE 217

Based on studies of adult animals, exaggeration of ischemic damage by lactic acidosis is thought to occur at tissue lactate concentrations of -20 pmol/g and intracellular pH values of -6 (Siesjo et al., 1990). It has been speculated that slower rates of glycolysis and/or transport of glucose across the blood-brain bamer will limit lactic acidosis in neonates compared to adults (Vannucci, 1990). This hypothesis presup- poses the validity of a second unmentioned and un- tested hypothesis, namely, that the extent of brain acidosis for a given degree of brain lactate formation is the same in neonates and adults. For this to be true, cerebral acid buffering capacity must not change as a function of age. However, we have hypothesized that during maturation, changes in acid buffering capacity could occur due to changes in the concentration of intrinsic physiochemical buffers (proteins or low mo- lecular weight compounds) or changes in the ability of the brain to extrude acid via active transport mechanisms (Corbett et al., 1988; Corbett and Laptook, 1990). To our knowledge there have been no previous attempts to characterize age-related changes in cerebral acid buffering capacity in animals of any species.

In the present study we report our measurements of the cerebral acid buffering capacity of piglets, ranging in age from newborn to 1 month. We used two com- plementary approaches. First, combined 'H and "P nuclear magnetic resonance (NMR) spectroscopy was used to measure changes in buffering capacity in vivo during the first 30 min following complete ischemia induced via cardiac arrest. Second, the pH titration curves of brain tissue homogenates were compared for newborn and 1-month-old piglets. The specific questions addressed here are (1) What effect does blood plasma glucose concentration have on intracellular brain pH during ischemia in animals of different ages? (2) Does cerebral acid buffering capacity change as a function of age? and (3) Do changes in acid buffering capacity occur as a result of ischemia?

MATERIALS AND METHODS

Surgical preparation and experimental procedures The surgical procedures and experimental protocol were

approved by the U.T. Southwestern Medical Center Institu- tional Review Board for Animal Research. A total of 39 miniature swine (Sinclair strain) was studied, ranging in postconceptual age from 1 11 to 153 days (normal term gestation is 1 15 days). Twenty-nine animals were used to determine age-related changes in acid buffering capacity during the 30 min following complete cerebral ischemia, measured in vivo, via combined 'H and "P NMR. Ten of these animals were also used to evaluate the acid buffering capacity of ischemic brain by performing pH titrations in vitro on brain tissue homogenates. In vitro pH titrations were also performed on a separate group of 10 animals whose brains were frozen prior to ischemia, i.e., under control physiologic conditions.

In two litters, birth was initiated after 1 10 and I 14 days of gestation, respectively, via prostaglandin-induced labor (Linderkamp et al., 1981). Two injections of F2 a-prosta-

glandin (10 mg/kg Lutalyse i.m.; Upjohn, Kalamazoo, MI, U.S.A.), given I2 h apart, resulted in labor within 24 h ofthe first injection. Piglets from other litters were studied after spontaneous term delivery ( I 14- 1 I6 days); all piglets were kept with their sow until the day of the study. Surgery was preceded by anesthesia with ketamine (10 mg/kg i.m.), plus infiltration of surgical sites with 1% xylocaine for local anesthesia. The piglet was then rapidly tracheotomized and ven- tilated with N20 /02 70:30 (Harvard small animal respira- tor). Intravascular catheters were placed in the carotid artery and the superior vena cava via the internal jugular vein, respectively. D-Tubocurarine C1 (0.10 mg/kg) and nalbuphine (0. I 5 mg/kg) were then administered intravenously for muscle relaxation and analgesia, respectively. Ketamine is a short-acting anesthetic; by the time experimental data collection began (-2 h later), its potential effect on cerebral metabolism was negligible. The use of nalbuphine for analgesia during the NMR measurements is not expected to affect cerebral metabolism, as narcotic opiates can relieve pain but have no or only mild depressive effects on CNS activity (Benedetti and Butler, 1990).

For experiments involving in vivo NMR, skin overlying the skull was retracted so that the NMR radiofrequency coil rested directly on the skull. After surgery the animal was transported to the magnet and stabilized for I h before in vivo NMR measurements began. Control data consisted of a pair of-"P and 'H NMR spectra and arterial blood gas and substrate measurements, to assess baseline brain metabolite levels and systemic physiologic status, respectively. Next, a continuous intravenous infusion of glucose ( 10-30 mg/kg/ min) or a bolus intravenous injection ofinsulin (1-2 IU/kg) was given to adjust the arterial plasma glucose concentration of individual piglets to a predesignated value in the range of I to 30 mM. After approximately 30 min, when the desired was reached, an additional pair of "P and 'H NMR spectra and arterial blood samples were collected. Complete cerebral ischemia was then rapidly initiated via cardiac arrest, induced by an intravenous injection of 4 ml of 4 M KCI. A second injection of nalbuphine was given 15 min prior to initiating cardiac arrest. Changes in brain phosphorylated metabolites and lactate were monitored by collecting "P and 'H NMR spectra during the period following cardiac arrest. After 30 min, NMR data collection ended and the animal was removed from the magnet. The portion ofthe cerebral cortex immediately below the NMR coil was removed and stored at -70°C for subsequent analysis of the final brain lactate concentration ([lactate],,,,) and for in vitro pH titrations.

For experiments involving in vitro pH titrations of control brain homogenates, skin overlying the skull was retracted and a craniotomy was performed. After a 30-min stabilization period. a blood sample was obtained to measure arterial blood gas tension, pH, [ g l u c o ~ e ] ~ , , ~ ~ ~ , and arterial plasma lactate concentration ([lactate],,,,,,). Next. the brain was funnel frozen in situ for 10-20 min, using the method of Ponten et al. (1973). Throughout the freezing period, heart rate and blood pressure were monitored continuously. Finally, the animal was killed (pentobarbital, 150 mg/kg), and the frozen brain cortex (10-20 g) was removed and stored at -70°C until brain tissue homogenates were prepared.

Analytical techniques NMR spectroscopy. In vivo NMR measurements were

gathered on two systems, both of which have been described

J. Neirroclrem.. I ' d . 59. N o . 1. 1993

218 R. J. T. CORBETT ET AL.

in detail previously (Corbett and Laptook, 1990; Corbett et al., 1991).

Initially, six animals were studied on a Nicolet NT-80 spectroscopy system with a 30-cm diameter-bore TMR 321 200 superconducting magnet (Oxford Instruments) operating at 1.9 T. The NMR signal from brain was measured using two concentric rectangular radiofrequency coils (outer dimensions, 5 X 4 cm) tuned to the 'H and 31P resonance frequencies via two separate balanced matched cir- cuits (Murphy-Boesch and Koretsky, 1983). Piglets laid su- pine, with their skulls resting directly on top of the coil. The 'H tuned coil was used to observe the NMR signal from H20 while adjusting the homogeneity of the magnetic field until the linewidth in the corresponding 'H NMR spectrum was -0.25 ppm (or -0.1 ppm on the 4.7-T system). The "P tuned coil was used to acquire 31P NMR spectra at control and 25-30 min after cardiac arrest. The pulse sequence for "P NMR data collection consisted of a 90" excitation pulse (50 ps) and a 409-ms acquisition time; the sweep width was 5,000 Hz and there were 2,048 data points per free induction decay (FID). Each 3'P NMR spectrum was the sum of 128 FID transients; the delay time between excitation pulses equaled 2.8 s. To reduce spurious curvature in the baseline, the first three data points in the FID were set to zero, by left shift removal, followed by right shifting to rees- tablish the phase of the FID; this has the effect of setting the acquisition delay equal to three dwell times (i.e., 3 x 200 ps). This was followed by applying exponential apodization ( 10-Hz linebroadening), Fourier transformation, and baseline straightening via an interpolation routine supplied with the NMR spectrometer software.

Subsequently, 23 additional animals were studied on a General Electric CSI system and 40-cm diameter-bore Ox- ford superconducting magnet operating at a magnetic field strength of 4.7 T. The "P and 'H NMR spectra were measured using a two-turn radiofrequency coil (outer coil dimensions: 5 x 3 cm), double-tuned to the "P and 'H resonance frequencies (Gonnella and Silverman, 1989). The piglets rested prone, with the coil lying directly above the skull. Besides the advantage of a better signal-to-noise ratio due to the higher magnetic field strength, this spectrometer system also enabled rapid switching of data collection between 'H and "P resonance frequencies. This feature allowed the simultaneous comparison of cerebral NMR results at both 'H and 31P frequencies over the first 30 rnin after initiating ischemia. The protocol for data collection consisted of acquiring I6 FIDs at the 'H frequency for 25 s, followed by the collection of four FIDs at the "P frequency for 5 s. A 3-s interval was required for the computer to switch data collection parameters, making the total time to collect a pair of 'H and "P NMR spectra equal to 36 s. To establish a steady state of magnetization before the record- ing of data, we applied a 90" pulse I s prior to collecting the four FIDs at the 31P frequency. 31P NMR data collection consisted of a 90" excitation pulse (50-100 ps), a 256-ms acquisition time, and a 700-ms delay time prior to the next excitation pulse; the sweep width was 5,000 Hz and there were 2,048 data points per FID. The accumulated FID was processed in the same manner as described above for the 1.9-T system, with the exception that an apodization function corresponding to 20-Hz linebroadening was applied. To measure the @lactate 'H NMR spectruq, signalsfr_om water and lipids were suppressed using a I33 1-7-2662- 7-AC spin echo pulse sequence, where "1" corresponds to the first portion of a binomial pulse equal to one-eighth of

the 90" pulse width; the other pulses are multiples of this and the over bar indicates phase inversion (Hore, 1983). The prepulse delay time was I s, the delay time for echo formation ( 7 ) equaled 150 ms, the sweep width was 5,000 Hz, and there were 2,048 data points per FID. The accumulated 'H FID (16 transients collected every 36 s) was processed by applying baseline correction for DC offset, followed by multiplication by an exponential apodization function corresponding to 10-Hz linebroadening, Fourier transformation, and zero-order phasing. The phase angle used was determined from the setting that gave the maximal positive peak for the final postmortem &lactate NMR spectrum. To suppress the nonlactate 'H NMR signal, a control spectrum was subtracted from all subsequent spectra collected during ischemia, to yield a series of difference spectra showing the time course of cerebral lactate accumulation following complete ischemia. To reduce the amount of noise accumulated by performing difference spectroscopy, the control spectrum was obtained as the sum of 160 FID transients. The validity of the difference spectroscopy ap- proach for measuring changes in brain lactate concentration was established through tests on phantoms and in vivo brain tissue in this laboratory (Corbett et al., 1988) and elsewhere (Gyulai et al., 1987; Petroff et al., 1988).

Physiological monitoring and blood biochemistry. Mean arterial blood pressure (MAP) and heart rate were measured via the catheter in the carotid artery using a Gould pressure transducer and a Coulbourn polygraph. The catheter in the carotid artery was also used to obtain samples for monitoring arterial blood: pH (pH,), CO, tension (Pco,), 0, tension (Po2), lactate, glucose, and hematocrit (HCT). The catheter in the superior vena cava was used for the adminis- tration of nalbuphine, glucose, insulin, and 4 M KCI. pH,, Pco,, and Po, were measured from a 200-pl sample using a Corning 165 blood gas analyzer. [Lactate],,,,,, was measured using the lactate dehydrogenase assay after deproteinization of 200 p1 of plasma with 8% perchloric acid (PCA) and centrifugation for I min at 1,500 g. [ G ~ U C O S ~ ] , ~ , ~ ~ ~ was determined by the glucose OxidaseJperoxidase assay after deproteinization of 50 pl of plasma with 2% PCA and centrifugation. Polygraph traces of heart rate and MAP were recorded continuously during the experiments, whereas blood samples were drawn at the midpoint of the two preischemia NMR data collection periods.

Brain tissue biochemistry and pH titrations. Frozen postmortem cortical brain samples were ground to a powder and mixed thoroughly to ensure a homogeneous sample representative of the entire region of brain measured by the NMR radiofrequency coil. Brain tissue lactate concentrations were determined using a lactate dehydrogenase enzy- matic assay of the neutralized aqueous portion from 0.5 g of brain powder, homogenized and extracted in 5% PCA (Cor- bett et al., 1987).

Three in vitro pH titrations were performed per animal, using homogenates prepared from 0.7-, 0 3 , and 0.3-g samples of brain powder, The frozen brain powder was rapidly transferred to a 50-ml test tube, to which was added 3 ml of a solution containing 5 mM iodoacetic acid and 0.2 mM fluorodinitrobenzene, to inhibit 3-phosphoglyceraldehyde dehydrogenase and creatine kinase, respectively (Rose et al., 1988). The mixture was homogenized for 1 min with a glass pestle attached to a rotary homogenizer. Next, 3 ml of min- eral oil was layered on top, and a semimicro combination pH electrode (Markson Scientific) was inserted into the homogenate. The entire assembly was then placed into a con-

J. Neurochem.. Val. 59. No. 1. 1992


stant water bath and allowed to reach 38°C over a 2-min period. The homogenate was titrated over the pH range 7. I to 5.5 using 50- to 300-p1 aliquots of either 0.02 M HCI or KOH. Typically 15-10 aliquots were added for the acid titration curve and a 1-min interval was allowed for the pH reading to stabilize. The acid pH titration was always com- pleted within 20 rnin after the initial weighing and homogenization of the tissue. The concentration of stock HCI was standardized by titration against solutions of sodium car- bonate; the concentration of stock KOH was established by titration against stock HCI (Peters et al., 1974).

To examine the effect of animal age, time of measurement, and pH range on buffer capacity, pH titrations were performed on four classes of brain tissue (five animals in each class): (1) newborn-frozen at control; (2) newborn- frozen after 45 rnin of ischemia; (3) 30-day-old-frozen at control; and (4) 30-day-old-frozen after 45 rnin of ischemia. Animals from classes 2 and 4 were selected from the previously described in vivo NMR study, as their brains had been removed and stored for tissue lactate determinations, as part of the procedure to calibrate the in vivo lactate 'H NMR signal.

Preliminary experiments indicated that the amount of acid or base required to titrate the inhibitors ( 5 mM iodoacetic acid and 0.2 mM fluorodinitrobenzene) was less than 0.5 mMof base or acid. Therefore we did not need to correct for acid buffering due to the presence of these compounds. We also examined the stability of the pH reading for selected brain homogenates in the presence of the inhibitor solution over a 30-min period at 37"C, without any titration being performed. The initial pH for control brain tissue (classes I and 3) 2 rnin after homogenization ranged from 6.93 to 6.79; this value decreased slowly by less than 0.1 pH unit over 30 min, suggesting that the samples were suffi- ciently stable to perform pH titrations within this interval.

Calculations Intracellular pH, lactate, and energy metabolites. Mean

intracellular brain pH (pH,) was calculated from the chemical shift of inorganic phosphate (Pi) using the following equation:

pHi = 6.683 + log [ ( p - 3.153)/(5.728 - p)] ( 1 )

where p equals the chemical shift difference (parts per mil- lion) between the Pi resonance frequency at different stages in the protocol and the phosphocreatine (PCr) resonance frequency at control. Since the animal is placed in the magnet and its position is never adjusted, it is possible to use the resonance frequency of PCr measured in a control spectrum as the chemical shift reference standard for subsequent spectra acquired during ischemia, when the concentration of PCr rapidly decreases to zero. We have experimentally veri- fied that magnetic field drift on our magnet is less than 0.01 ppm over the interval required to carry out the present study protocol (30 min). The accuracy and limitation of making pHi measurements via 31P NMR have been dis- cussed at length elsewhere (Nuccitelli, 1982; Petroff et al., 1985; Corbett and Laptook, 1990).

Alterations in phosphorylated metabolite levels were calculated as the percentage of change in PCr and @-nucleoside triphosphate (NTP) peak height relative to the mean control spectra.

The postmortem brain lactate content measured from the tissue extract was used to calibrate the intensity of the @-lac-

tate signal measured in the final 'H NMR spectrum obtained in the protocol summarized above. The procedure to derive brain lactate concentrations from the in vivo 'H NMR spectra collected during ischemia assumes that the spin-spin and spin-lattice relaxation times of the 0-lactate 'H do not change. Past studies indicate that there is not a significant change in the spin-lattice relaxation time constant between control and ischemia, since control brain lactate concentrations derived from the postmortem calibra- tion of the @-lactate 'H NMR signal are in good agreement with the brain lactate concentration measured from PCA extracts (Corbett et al., 1988). For 10 animals studied in the present protocol, we also examined the linewidth of the lactate peak from the time-course 'H NMR difference spectra. There were no significant changes in linewidth, suggesting that the apparent spin-spin relaxation time for 0-lactate re- mained constant. Therefore, changes in the height of the @-lactate 'H NMR peak were used to quantify the change in brain lactate concentration over time. The postmortem brain lactate content measured from the tissue extract was used to calibrate the height of the @-lactate signal measured in the final 'H NMR spectrum obtained in the protocol summarized above.

Buffer base deficit (ABB) and buffering capacity. During ischemia the ABB equals the amount of acid generated from the production of lactic acid, the hydrolysis of endogenous NTP, and the net hydrolysis of PCr to PI (Siesjo, 1985). As in our previous report (Corbett et al., 1988), we estimated the ABB at different times ( t ) during ischemia, using the following formula:

4BBI = [Lalin,t,al - [La], + 0.7"TPlt - "TPIuutm,

Brain lactic acid concentration ([La]) was calculated from the calibrated 'H NMR signal, as described above. The changes in concentration of NTP and PCr were estimated from the percentage change in their corresponding "P NMR signals, assuming that initial concentrations for both equal 3 pmol/g (Corbett et al., 1988). It should be noted that the concentration of cerebral PCr in neonates is lower than in adults (Tofts and Wray, 1985), whereas ATP remains constant with maturation (Miller and Shamban, 1977). The coefficients 0.7 and 0.4 reflect fractional protons produced per mole ofNTP hydrolyzed to ADP plus P, and the hydrolysis of PCr to Cr plus P,, respectively (Wilkie, 1979; Siesjo, 1985). The volume fraction ofintra- and extracellular water in brain (V) was assumed to equal 0.76 ml/g (Siesjo, 1985) and is assumed to remain constant during ischemia.

The buffer capacity (@) during ischemia was calculated using the expression (Siesjo, 1985)

@I = (ABB)J(APH,)I (3) where ApH, = (pHl),n,tlal - (pH,),. To calculate @, it is desir- able to compare 31P and 'H NMR data collected over the same interval. This was achieved by averaging data from two "P NMR spectra acquired immediately before and after each 'H NMR spectrum obtained during ischemia.

It should be noted that Eq. 2 differs from the phenomeno- logical formula developed by Siesjo (1985) to calculate a theoretical ABB (defined as [BB],, by Corbett et al., 1988). The latter formula requires knowledge of the brain tissue COz tension, and assumptions must be made about the total concentrations and pK, values of weak buffer acids. In comparison, the calculation of A[BB] and @ using Eqs. 2 and 3 is

J. Neurochem., Vol. 59. No. 1. 1992


based on experimentallv determined changes in cerebral lactate, NTP, PCr, and pH,. As detailed in the previous paragraph, Eq. 2 also requires assumptions about initial concentrations of PCr and ATP and the volume fraction of water in the brain. However, we believe that even if these assumptions only approximate true values, the results from Eqs. 2 and 3 are still useful, especially when relative compari- sons are made, as in the present study.

Statistical anal-vsb ofdata. All statistical tests and regression fits were performed using SPSS/PC (Version 3.1, Sta- tistical Package for the Social Sciences Inc., Chicago, IL, U.S.A.), on a AST premium/286 personal computer. Com- bined data are presented as the mean t SD, tests for multiple-group differences were made by a one-way analysis of variance (ANOVA) using Duncan's (1955) multiple comparison test to identify statistically significant differences ( p < 0.05) between specific pairs of means. Multiple measurements of fl were also calculated and compared at different times or over different pH ranges for the various age groups of animals studied. Therefore, it was necessary to perform a multivariant ANOVA (MANOVA) with consideration of within-subject repeated measures. Multivariant tests for the significance of F were performed using the Pillais trace test (Olsen, 1976; Norusis, 1988).

RESULTS

Physiologic status The mean control physiologic values for the four

age groups of piglets studied by in vivo NMR are summarized in Table 1. As expected, significant increases in body weight, HCT, and MAP occurred with increasing age. The two youngest groups of piglets had slightly higher blood lactate concentrations than the oldest two groups. Newborn piglets (group 1) also had a slightly lower than the older three groups of animals. The lower mean pHa for group 1 compared to groups 3 and 4 can be totally accounted for by the slightly elevated Pco2 observed for some animals in this group. Overall we judged the difference in blood gases between group 1 and the other three groups to be small and not likely to cause a differential group effect on cerebral metabolism during ischemia.

Following the 30-min interval during which [glu- C O S ~ ] ~ ~ ~ ~ ~ ~ was adjusted to a value in the range 1 to 30 M, a second series of physiologic measurements and a pair of "P and IH NMR spectra were collected for each animal. With the exception of there were no significant changes in group means in any of the physiologic variables, compared to the first control measurements (Table 1). After the injection of KC1, the heart rate dropped to zero within one or two beats, with a concurrent drop in blood pressure, indicating a well-defined time point at which complete cerebral ischemia was initiated. For five piglets in which the postmortem rectal temperature was monitored, the temperature measured after 30 min differed by less than 1°C compared to the control (37.8 to 38.3"C).

Ten additional animals were used in the experiments involving pH titrations of brain homogenates (see below). These animals were subdivided into two groups with mean postconceptual ages of 1 15 * 2 and 144 f 6 days, respectively. The mean physiologic values for these two groups did not differ significantly from the means shown in Table 1 for groups 1 (newborn) and 4 (1 month), respectively, with the exception of a slightly lower mean Pco2 (38 f 4 mm Hg) for the newborn group.

In vivo NMR A series of representative brain 31P and difference

'H NMR spectra measured in vivo for a 2-day old piglet is shown in Fig. 1. The spin-echo pulse sequence using binomial pulses produced excellent suppression of the 'H NMR signal from water, as evi- denced by the absence of a peak at 4.8 ppm. The most prominent peaks in the 'H spectra were from 2.0 to 1 .O ppm, corresponding to N-acetylaspartate, lipid, and lactate. In the postarrest 'H NMR difference spectra, a progressively larger signal appeared at 1.4 ppm, corresponding to the /3 protons of lactate. No peaks other than the one corresponding to lactate were usually present in the difference spectrum except for a small peak at 2 ppm. The 31P NMR data revealed

TABLE 1. Physiologic data for diferent age groups of piglets studied using in vivo NMR (mean * SO)

Group I (n = 10)

Group 2 (n = 5 )

Group 3 (n = 7)

Group 4 (n = 7)

Age (days) 114 f 3," I 2 I rt I '-3.4 128 + 31,2.4 146 f 41.2,4 Weight (kg) 0.9 f 0.23.4 1.8 +_ 0.14 2.7 f 0.41.4 4.4 5 1.7'-' [GluCOSelplasma (M) 4.1 f 2.8'" 6.9 f 0.5' 7.5 f 2.1' 6.8 f 1.7' [Lactatel,,,,, (mw 2.0 f 0.93 2.3 * 0.93 1. I ? 0.4'32 1.4 f 0.6 PH 7.31 + 0.13.4 1.37 f 0.07 7.40 rt 0.05' 7.44 f 0.06' Pco, (mm Hg) 47 * 102" 35 + 3' 37 * 3' 36 f 3' Po2 (mm Hg) 99 f 193,4 91 f 463,4 146 f 37'*' 160 + 24'.'

Heart rate (beat/min) 198 f 34,s' 276 rt 3 6 ' ~ ~ 255 -+ 40' 21 1 f 60' MAP (mm Hg) 6 9 + 174 71 rt 2i4 87 -c 8 95 f 18'.'

The superscripts 1-4 refer to the group means that are significantly different from that of the group listed in the column heading, p < 0.05

HCT (volume W ) 28 f 4'" 34 f 2' 33 f 5' 37 f 3'

using the Duncan multiple-comparison test. For example, group 1 has a significantly lower HCT than groups 2-4.

J . Neurochem., Vol. 59, No. I , 1992

CEREBRAL ACID BUFFERING CAPACITY AND AGE 22 I

1 I 41

A. I B. I

+cardiac arrest

T - 10 0 -10 -20ppm 5 -5 PPm

rapid postarrest declines in the peaks corresponding to PCr and ATP, with a parallel increase in Pi. pH, decreased during ischemia, as indicated by a shift in the resonance frequency of Pi. There was no significant difference between pH, measured after 12 to 30 min of ischemia and lactate measured after 20 to 30 min. For all four age groups, there were significant linear correlations (rZ ranging from 0.66 to 0.93) between measured immediately prior to cardiac arrest and final pH, (Fig. 2). The correlation between and pH, was lower for earlier intervals.

For the subset of animals studied on the 4.7-T NMR system, each pair of 3LP and 'H spectra was used to calculate pH, and ABB throughout the first 1 1 min of ischemia (Fig. 3). The p for each animal was determined from the slope ofpH, versus ABB; individual plots had a r2 of 0.9 or higher. The mean p for the full pH range measured for each age group is shown in Table 2 as pfUn. A one-way ANOVA, using Duncan's multiple-comparison test, revealed no significant differences between age groups. Consistent with this, linear regression analysis of pfull, calculated for each animal, versus postconceptual age, gave an r2 of 0.082 (not significant) and a slope close to zero (0.1 mM BB/pH unit per day). Despite the high r2 value for plots of pH, versus ABB, closer inspection of individual plots suggested a slight curvature, suggestive of an

FIG. 1. Representative in vivo NMR spectra from a newborn piglet: A, 31P; B, lactate-edited 'H. The data shown correspond to pre-cardiac arrest (first spectra) and the first 10 min after inducing brain ischemia via cardiac arrest. Each 31P NMR spectrum is the sum of eight FID transients collected within 0.6 min; the hori- zontal dimension shows the chemical shift (ppm) with PCr at 0 ppm, and the vertical dimension corresponds to relative signal intensity. The peaks are labeled as (1) P, (2) PCr, (3) /3 peak of NTP, and (4) 0 peak of lactate. Typi- cally peak height signaknoise ratios at control were -5:l for either PCr or 0-NTP. The lactate-edited 'H NMR spectra are the sum of 16 transients collected within 0.6 min; typical postmortem signaknoise ratios were 210:i.

increase in p as pH, decreased. To test for this we recalculated p for two pH ranges: p7.1-6.6 for pH, 7.1- 6.6, and p6.6-6.0 for pH, 6.6-6.0. There was a high degree of interanimal variation for the lower pH range of p6.6-6.0 (pHi 6.5 to 5.9), because the final pHi de- pended on the at the time of cardiac arrest and was adjusted to a range of values for each age group (see above). Mean values of p 7 . '-6.6 and p6.6-6,0 for each age group are summarized in Table 2. A paired t test suggested significant differences between p7.1-6.6 and p6.6-6.0 for all four age groups 0, = 0.027,0.087,0.012, and 0.062 for groups 1 to 4, respectively).

In our initial analysis to examine whether or not significant age-group differences in buffering capacity occur for either pH range, we assumed that the measurements of p were all statistically independent. Us- ing a one-way ANOVA, there were no significant differences in p7.1-6.6 between any of the four age groups (Duncan's multiple-comparison test at p = 0.05). However, p6.6-6.0 for the oldest age group (group 4) was significantly greater than ,86.6_6.0 for the first three groups. Because the individual pairs p7. 1-6.6 and 66.6-6.0

were measured from the same animal, the above as- sumption about independence may not be correct, so we also performed a repeated-measures MANOVA. For one animal in group 4 there were insufficient data to calculate p7.1-6.6. leaving a total of 22 subjects. The

.I. Nerrrochhcm., V d 59. N o . 1. 1992


FINAL pH 7.2

t I

5.8

5 10 15 20 25 30 35 5.6

0

PLASMA GLUCOSE CONCENTRATION ( mM

FIG. 2. Relationship between [ g l ~ c o s e ] ~ , ~ ~ ~ ~ and the final brain pH,. The symbols shown in the legend represent the final pH, value (measured after 12 to 30 min of ischemia) for 29 animals in each of the four age groups as defined in Table 1. The solid lines show the best fit by linear regression analysis for each of the four age groups. The slopes (+SD) for groups 1 to 4 are -0.020 f 0.005(n = lo), -0.0162 0.003(n = 5), -0.012+ 0.002(n = 7), and -0.018 f 0.002 (n = 7) pH unit/mM glucose, respectively. Intercepts for groups 1 to 4 are 6.42 f 0.07, 6.50 f 0.04, 6.48 f 0.03, and 6.48 -+ 0.03, respectively.

within-subject variables, P7.1-6.6 and P6.6-6.0, were transformed into a single within-subject variable, BUFF, represented as the orthonormalized difference (or contrast) between 66.6-6,0 and 67.1-6.6 for each animal. Age group was the between-subject factor. The MANOVA of between-subject effects indicated that there was no difference between the overall mean P for animals in the four age groups (p = 0.076). The MANOVA of within-subject effects involving BUFF indicated that the interaction between BUFF and age group was not significant (p = 0.077). However, the within-subject contrast between 67.1-6.6 and 66.6-6.0 was significantly different from zero (p < 0.0005; mean value of the difference = 18.0 f 3). In vitro pH titrations

The pH of ischemic brain samples (designated classes 2 and 4 in Materials and Methods) prior to titration ranged from 6.7 to 6.0. As these animals had also been studied using in vivo NMR, we were able to compare the pH measured from the chemical shift of Pi in the "P NMR spectrum after 28-30 min of ischemia (pHi) to the pH measured from an electrode inserted into the brain homogenate (pH,,) from the same animal (n = 9). Linear regression analysis of the plot of pH, versus pH,, gave a slope of I . 13 f 0.08 and an r2 value of 0.97, suggesting an excellent agreement between the two pH measurements.

Figure 4 shows the acid titration curves obtained for two newborn animals frozen at control and after 45 min of ischemia. Prior to acid titration, all the homogenates were adjusted to pH 7.1-7.0 using the ap- propriate amount of base. After correcting for differ-

ences in the weight of brain powder used, the three titration curves performed for each animal were similar. Like the in vivo NMR results (Fig. 3), the titration plots for both control and ischemia brain homogenates had a nonlinear trend, suggesting that P increases as pH decreases. To quantify this, we calculated 67.1-6.6 to 66.0-5.5 (slope of pH versus ABB) for three distinct pH ranges: 7.1-6.6, 6.6-6.0, and 6.0- 5.5. The means and standard deviations of ,87,!-6.6, p6.6-6.0, and P6.0_5.5 for the four classes of in vitro titration data are summarized in Table 2.

To identify significant differences, we performed a repeated-measures MANOVA. The within-subject dependent variables, P7.L-6.6, P6.6-6.07 and P6.0-5.5 were transformed into orthonormalized linear combinations of their differences; a variety of contrasts was examined in separate MANOVA tests. The combination of all transformed within-subject variables was designated the pH range. Age group (i.e., newborn or 1 month old) and time of measurement (i.e., brains frozen at control or after 45 min of ischemia) were designated between-subject variables. The MANOVA for between-subject effects indicated that age group was not significant (p = 0.58), whereas time of measurement was highly significant (p < 0.0005). The multivariant tests of significance of within-subject effects indicated a significant interaction between time of measurement and pH range (Pillais test, p = 0.002), but the interaction between age group and pH range was not significant (p = 0.13). Finally, the within-subject variable, pH range, was highly significant (p < 0.0005), indicating a significant difference between one or more of 67.1-6.6, 66.&6.0, and 66.&5.5. Univariate F tests for different combinations of the transformed within-subject variables indicated that the interaction between time of measurement and pH range was due mainly to significant differences between P6.6-6.0 and

lntracellulor pH 75

5.5 1 I I J -5 0 5 $3 15 20 25 M

- A [ Buffer Base ]

FIG. 3. Relationship between pH, and A[68] based on measurements made in vivo using combined3'P and 'H NMR. The pH, was determined from the chemical shift of Pi; the A[BB] was determined from the changes in PCr, NTP, and lactate occurring during the first 10 min of ischemia (see Calculations). The data from seven newborn animals are presented; solid lines connect data points collected at different time points for individual animals.

J. Nertrochem., Vol. 59. No. I, 1992

CEREBRAL ACID BUFFERING CAPACITY AND AGE

6.5

6.0

223

-

-

TABLE 2. Comparison of cerebral p (mean & SD) for different pH ranges and age groups measured in vivo by NMR and in vitro via the uH titration ofbrain homoaenates

Group 1 Group 2 Group 3 Group 4 In vivo (n = 7) (n = 5 ) (n = 6) (n = 5)

Prua -23.9 t 4.7 -2 1.7 f 4.4 -25.7 f 5.6 -26.6 f 3.4 07.1-6.6 -19.9 t 3.7 -17.0 f 3.5 -17.4 f 1.7 - 18.2 % 6.0 06.64.0 -31.5 % 9.2 -27.3 f 10.0 -34.6 f 11.2 -52.8 f 19.3

In vitro (n = 5) (n = 5)

Control 07.1-6.6 -17.9 k 2.6 NM b6.6-6.0 -31.3 f 1.0 NM 06.0-5.5 -36.0 t 1.2 NM

07.1-6.6 -18.0 f 0.9 NM 86 .66 .0 -26.3 k 0.8 NM 06.0-5.5 -32.0 t 1.0 NM

Ischemic

NM -19.1 k 2.2 NM -29.6 k 2.2 NM -36.0 f 1.8

NM -18.9 f 1.8 NM -25.5 f 1.4 NM -30.6 f 1.9

Age groups I to 4 correspond to piglets whose mean postconceptual ages are listed in the first row in Table 1. The subscript "full" refers to 0 measured over the entire range ofpHi observed for each animal; the other subscripts refer to different subranges of pH, or brain homogenate pH. Values under the heading'% vivo" refer to 0 measured for 23 piglets studied on the 4.7-T NMR system (see Materials and Methods). For one animal in group 4 there were insufficient data to calculate 06.6-6.0. The values under the heading "In vitro" refer to 0 measured by performing pH titrations on 10 brain homogenates. NM, not measured; control, brain tissue funnel frozen under control physiologic conditions; ischemic, brain tissue frozen after 45 min of ischemia. An analysis of significant group differences is presented in the Results.

P7.1-6.6 and between P6.0-5.5 and 07.1-6.6. The estimate of the coefficient of difference between P6.6-6.0 and P7. 1-6.6

due to the time of measurement and pH range interaction was 2.2 & 0.5; the analogous estimate for the P6,&5,5 and p7.1-6.6 difference was 2.3 k 0.5. Similar univariate F tests indicated significant differences among all three values. The parameter estimates of the coefficients of the intrinsic differences between values (i.e., exclusive of between-subject effects) were 15.2 k 0.6 for P6.&5.5 and P7.1-6.6, 9.7 k 0.5 for P6.6-6.0

and 07.1-6.6, and 5.4 0.5 for P6.0-5.5 and 06.6-6.W

DISCUSSION

All four age groups had similar correlations between prior to ischemia and final pHi at the end of ischemia (Fig. 2). The negative slopes for all four plots are consistent with previous observations that elevated blood glucose levels accentuate brain acidosis during complete ischemia in adult brain (von Hanwehr et al., 1986; Chopp et al., 1988). We inter- pret the similarity in slopes to imply that for a given increase in blood glucose concentration, the same increase in intracellular brain glucose occurs for all four age groups and therefore the same increase in acidosis is observed. We conclude that for piglets exposed to complete ischemia, the potential for hyperglycemia to increase brain acidosis remains the same throughout the first month of life.

Another noteworthy finding illustrated in Fig. 2 is that the intercept of these plots ranged from 6.5 to 6.4 for all four age groups. The implication is that -0.5

pH unit of the observed decline in pH, during total ischemia is not influenced by the preischemic [glu- C O S ~ ] ~ ~ ~ ~ ~ ~ . Assuming an acid buffering capacity of -25 M / p H unit (Results), we predict that a pHi decrease of 0.5 unit would be generated by a ABB of - 12 mM. About 2.5 mM of the ABB would derive from ATP and PCr hydrolysis [see Eq. 21, which we expect to be the same for all four age groups. We pos- tulate that the remaining - 10 mM is from lactate generated from endogenous brain glycogen reserves. The similarity of intercepts for all four age groups suggests that the concentration of the endogenous glycogen reserve is the same for all four age groups.

pH of Brain Homogenate

7s I m frozen at ronlrol 7.0 [ I I

m 5.5' " ' ' ' ' ' ' -5 0 5 XI 15 20 25 10 35 40

- A [ Buffer Bose ]

FIG. 4. Relationship between pH and A[BB] based on pH titrations of brain tissue homogenates. Three titration curves, performed at 37"C, on one control and one ischemic brain homogenate are presented (0.7-, 0.5, and 0.33 samples were used). The samples were adjusted to pH 7.1-7.0 with base prior to performing an acid titration with 0.02 M HCI.

J. Neurochem., Vol. 59, No. 1. 1992


The results from the present study are consistent with our earlier finding (Corbett et al., 1988) that the relationship between pH, and ABB is approximately linear from pH, 7 to 6. However, the present experiments (in vivo NMR data collected on the 4.7-T system) allowed a closer examination of this relationship for individual animals. The results indicate a significant increase in 0 as pH, decreases. This trend occurred consistently for all four age groups of animals studied (Table 2). Thus, it is increasingly difficult to reach profound levels of acidosis (i.e., pH, -6) during ischemia because the brain is more strongly buffered as the pH, decreases. This result suggests that caution should be exercised when using brain lactate concentrations to estimate brain pH, or when using previously determined pH,-versus-lactate plots to extrapo- late a pH, based on an assumed continued linear trend outside of the observed range. A more detailed charac- terization of brain buffering in other species at the extremes of acidosis is warranted, especially in light of numerous reports of the occurrence of profound intracellular acidosis during and immediately following ischemia (Smith et al., 1986; Chopp et al., 1988; Kraig and Chesler, 1990; Laptook et al.. 1990).

The calculation of buffering capacity using Eq. 3 assumes that changes in pH, can be totally accounted for by the change in ABB as defined in Eq. 2. How- ever, it is likely that H+ generated during complete ischemia will react with HCO;, potentially causing a greater rise in PCO, in hyperglycemic, compared to normoglycemic or hypoglycemic, animals. Thus the possibility exists that elevated tissue Pco2 (Ptco,) levels could account in part for the drop in pH, and that this effect becomes more pronounced with increasing

To evaluate the relative importance that different changes in Ptco, have on the reduction in pH,, we used the theoretical formula developed by Siesjo (1985) that relates A[BB], pH,, and Ptco,. Ek- holm et al. (199 1) reported a linear relationship between Ptco2 and brain lactate in adult rats; increases in brain lactate concentrations of 10 and 20 pmollg result in a corresponding increase in Ptco2 tension to - 160 and -210 mm Hg, respectively. To use Siesjo’s model, we assumed that the above Rco2 values are valid for the present animal model, and we used the same formula and constants reported in our previous publication [Eq. 2 of Corbett et al. (1988)l. Specifi- cally, we assume that the preischemic control pH, and PtCo, equal 7.0 and 44 mm Hg, respectively, and that piglet brain can be represented by six buffer acids: three buffers, each at a concentration of 10 mrM, with pK, values of 6.7,6.4, and 6.1, and three buffers, each at a concentration of 20 mM, with pKa values of 5.8, 5.5, and 5.2, respectively (see Fig. 5 of Corbett et al. (1988)l. Thus for A[BB] equal to -15 and -25 mM with Ptco, fixed at 44 mm Hg, we calculate pH, values of -6.58 and -6.3 1, respectively. For the same pair of A[BB] values, but allowing Ptco2 to increase to 160 and 210 mm Hg, we calculate pH, values of

-6.38 and -6.16, respectively. Like von Hanwehr et al. (1986), we conclude that a minor fraction of the total decline in pH, (i.e., 0.2 to 0.15 pH unit) is due to the rise in Ptco2. Furthermore, the expected differences in elevated Ptco, during ischemia would account for a negligibly small differential decrease in pH, (i.e., less than 0.05 pH unit) in hyperglycemic versus normoglycemic animals.

It has been proposed that brain tissue pHi regulation occurs via three mechanisms: physicochemical buffering, consumption of metabolic acids, and transmembrane transport of H+ or HCOj (Siesjo, 1985). Acid regulation due to the presence of intrinsic intracellular buffers is the most important source of buffering for the present study, as transmembrane transport of hydrogen ions (or their equivalents) or the consumption of metabolic acids would require energy derived from ATP. As illustrated in Fig. 1, complete ischemia resulted in a rapid depletion of ATP. Fur- thermore, as the cerebral blood flow was zero, no means exist by which actively extruded H+ could be removed. We therefore considered the observed increase in at decreased pHi to originate by two means: ( 1) time- and ischemia-independent changes in p that are intrinsic to the pH range under observa- tion and (2) time- and ischemia-dependent changes in 0 during the course of the in vivo NMR experiment, occurring by virtue of the fact that brain tissue buffering is changing during ischemia. In other words, does p increase because there is simply a higher concentration of intrinsic buffers with a pK, of 6.5 to 6.0 compared to buffers with a pKa of 7.0 to 6.5, or do ischemia-induced changes in buffering capacity occur during the time course of the experiment? For example, it is conceivable that intracellular proteolysis during ischemia could increase p. Furthermore, acid-induced changes in brain buffering due to protein dena- turation are a recognized phenomenon, although extreme acidosis (i.e., pH, < 4.5) was required before a change in buffering capacity occurred in adult rats (Kraig and Wagner, 1987); analogous experiments on neonates have not been reported.

Initially we attempted to use data from the in vivo NMR experiment to determine the relative importance that age group, time of measurement, and pH range have on buffering capacity. However, for the data collected in these experiments pHi decreases continuously during the time course of ischemia, and therefore it is difficult to separate the effects of time of measurement and pH range. Nevertheless, we tried an ANOVA using the data for all four age groups to assess the independent effects of age group, time of measurement, or pH range. However, several empty cells existed in the matrix of means for different combinations of these three variables, and consequently, the interaction between variables could not be calculated. Furthermore, this type of analysis ignores the fact that multiple p values are estimated from the same animal and therefore are not statistically independent obser-

J. Ncurochem.. Vol. 59. N o . 1. 1992


vations. A repeated-measures MANOVA is more ap- propriate but would require a measurement of p in each animal for all combinations of time of measurement and pH range. An examination of our data revealed that, even with a small number of time and pH intervals, there would still be some empty cells in the data matrix for the majority of animals.

Therefore to investigate further the relative importance that age, pH range, and time of measurement have on 6, we performed pH titrations on brain homogenates obtained from newborn and 1 -month-old piglets. Homogenates prepared in the presence of the metabolic inhibitors, iodoacetic acid and fluorodinitrobenzene, had a stable pH, indicating that lactate formation and PCr degradation had been prevented. As a result it was possible to measure the intrinsic @ over a broad pH range and to compare brain tissue funnel frozen at control to brains exposed to 45 min of ischemia prior to freezing. Both control and ischemic tissue were examined for newborn and 1- month-old piglets, giving four independent classes of data. Similar to the in vivo NMR results, the relationship between pH and ABB (Fig. 4) was indicative of an increase in for progressively lower pH ranges (Ta- ble 2). A repeated-measures MANOVA revealed that there were no significant changes in @ that could be attributed to animal age. However, significant differences in @ existed for different pH ranges. The largest increase in p (AD = 9.7 f 0.5) occurred between pH range 7.1-6.6 and pH range 6.6-6.0. From pH 6.6- 6.0 to pH 6.0-5.5, there was a further increase in @ (Ap = 5.5 f 0.5 mM BB/pH unit). In addition, there was a small but significant decrease in p that could be attributed to time of measurement. Brain exposed to 45 min of ischemia prior to the pH titration had a lower buffering capacity than controls (A@ = -2.3 mM BB/pH unit). Thus the consequence of ischemia is to reduce the p, although this effect is relatively small. We conclude that the increase in @ observed in vivo is not caused by ischemia or age-induced changes but, rather, by the presence of higher concentrations of intrinsic intracellular buffers from pH range 6.6 to 6.0 compared to pH range 7.0 to 6.6.

In summary, the results from this study indicate that p during ischemia does not change as a function of age in piglets from newborn to 1 month old. This statement holds true for p calculated for pH ranges studied in vivo or in vitro using brain homogenates. Overall there is excellent agreement between p values measured in vivo and those measured in vitro (Table 2). For a given base deficit, brain pHi changes to the same extent regardless of animal age. We therefore conclude that piglets from newborn to I month in age have the same potential for brain lactic acidosis during ischemia, provided that similar brain lactate levels were generated during complete ischemia. However, this does not address the broader issue of whether or not there are age-related differences in irreversible damage enhanced by lactic acidosis. For example,

newborns could be exposed to similar levels of acidosis as adults but be intrinsically less sensitive to its putative harmful effects. Furthermore, we cannot ex- clude the possibility that the rate of formation of acidosis during ischemia is lower and the rate of clear- ance of acidosis following ischemia is higher in newborns compared to older animals. The former may occur because of slower glycolytic rates and the latter may occur due to a more permeable blood-brain barrier in newborns compared to adults. Thus, newborns could be exposed to shorter durations and de- grees of acidosis than adults, even if the interval of ischemia is identical for either age. Finally, the present study deals only with complete ischemia; during partial ischemia, maturational differences in pHi regulation via the consumption of metabolic acids and transmembrane transport of H" or HCO; could be- come important factors affecting the degree and dura- tion of lactic acidosis. We believe that the complex issue of acidosis-enhanced injury in immature brain cannot be fully resolved until these issues are addressed by further studies.

Acknowledgment: Funding was provided by Grant-in- Aid 906-075 from the American Heart Association Texas Affiliates, BRSG 2 SO7 RR 07 175- 13 awarded by the Bio- medical Research Support Grant Program, Division of Re- search Resources, NIH, National Institutes of Health South- western in Vivo Resource Facility (5P4 I-PR02584), and the Departments of Pediatrics and Radiology. We thank Dr. Loren Bertocci for his advice about performing pH titrations on tissue homogenates.

REFERENCES

Benedetti C. and Butler S. H. (1990) Systemic analgesia, in The Management ((Pain (Benica J. J., eds), Vol. 11, pp. 1640- 1675. Lea & Febiger, Philadelphia.

Chopp M., Welch K. M. A., Tidwell C. D., and Helpern J. A. (1988) Global cerebral ischemia and intracellular pH during hyperglycemia and hypoglycemia in cats. Stroke 19, 1383-1387.

Corbett R. J. T. and Laptook A. R. (1 990) Acid homeostasis following partial ischemia in neonatal brain measured in vivo by "P and 'H nuclear magnetic resonance spectroscopy. J. Neuro- chem. 54, 1208- I2 17.

Corbett R. J. T., Laptook A. R., and Nunnally R. L. ( 1987) The use of the chemical shift of the phosphomonoester P-3 I magnetic resonance peak for the determination of intracellular pH in the brains of neonates. Neurology 37, I77 1 - 1779.

Corbett R. J. T., Laptook A. R., Nunnally R. L., Hassan A., and Jackson J. (1 988) Intracellular pH, lactate, and energy metabolism in neonatal brain during partial ischemia measured in vivo by "P and 'H nuclear magnetic resonance spectroscopy. J . Neurochem. 51, 1501-1509.

Corbett R. J. T., Laptook A. R., and Olivares E. (1991) Simulta- neous measurement of cerebral blood flow and energy metabolites in piglets using deuterium and phosphorus nuclear magnetic resonance. J. Cereb. Blood Flow Metab. 11, 55-65.

Duncan D. B. (1955) Multiple range and multiple F tests. Biomet- rics 11, 1-42.

Ekholm A., Katsura K., and Siesjo B. K. (199 I) Tissue lactate content and tissue PC02 in complete brain ischaemia: implica- tions for compartmentation of H+. Neurol. Res. 13,74-76.

Gonnella N. C. and Silverman R. F. (1989) Design and construc- tion of a simple double tuned single input surface coil probe

J. Neurochem.. Vol. >9. No. 1, 1992

226 R. J. T. CORBETT ET A L

with multinuclear capability. Soc. Magnet. Res. Med. Abstr. 2, 948.

Gyulai L.. Schnall M., McLaughlin A. C., Leigh J. S., and Chance 9. (1987) Simultaneous "P- and ' H nuclear magnetic resonance studies of hypoxia and ischemia in the cat brain. J . C'ereb. Blood Flow Metab. 7, 543-55 1.

Helgason C. M. (1988) Blood glucose and stroke. Stroke 19, 1049- 1053.

Hore P. J. (1983) Solvent suppression of Fourier transform nuclear magnetic resonance. J . Magnet Res. 55,283-300.

Kraig R. P. and Chesler M. ( 1990) Astrocytic acidosis in hyperglycemic and complete ischemia. J . Cereb. Blood Flow Metab. 10, 104-1 14.

Kraig R. P. and Wagner R. J. (1987) Acid-induced changes in brain protein buffering. Brain Res. 410, 390-394.

Laptook A. R., Corbett R. J. T., and Nunnally R. L. (1990) Effect of plasma glucose concentration on neonatal cerebral metabolism during partial ischemia. Stroke 21,435-440.

Linderkamp 0.. Betke K., Guntner M., Jap G. H.. Riegel K. P., and Walser K. ( 198 1) Blood volume in newborn piglets: effects of time of natural cord rupture, intra-uterine growth retardation, asphyxia, and prostaglandin-induced prematurity. Pediarr.

Miller A. L. and Shamban A. (1977) A comparison of methods for stopping intermediary metabolism of developing rat brain. J. Neurochem. 28, 1327- 1334.

Murphy-Boesch J . and Koretsky A. P. (1983) An in vivo NMR probe circuit for improved sensitivity. J. Magnet. Res. 54,

Norusis M. J . (1988) Storing memories: Repeated measures analysis of variance, in SPSS/PC+ Advanced Statistics V2.0, Chap. 6. pp. B153-9181. SPSS Inc., Chicago, Illinois.

Nuccitelli R. (1982) lntracellular pH measurement techniques: their advantages and limitations, in Inrracellular pH: Its Mea- surement. Regulation. and Utilization in Cellular Funciions (Nuccitelli R.. ed), pp. 161-169. Alan Liss, New York.

Olsen C. L. ( 1976) On choosing a test statistic in multivariate analysis of variance. Psychol. Bull. 83, 579-586.

Peters D. G., Hayes J. M.. and Hieftje G. M. (1974) ChemicalSepa- rations and Measurements: Theory and Practice of Analytical Chemistry. W. B. Saunders, Philadelphia.

RCS. 15, 53-57.

526-532.

Petroff 0. A. C., Prichard J. W., Behar K. L., Rothman D. L., Alger J. R., and Shulman R. G. (1985) Cerebral intracellular pH by 3' P nuclear magnetic resonance spectroscopy. Neurology 35,

Petroff 0. A. C., Prichard J. W.. Ogino T., and Shulman R. G. ( 1 988) Proton magnetic resonance spectroscopy studies of agonal carbohydrate metabolism in rabbit brain. Neurology

Ponten U., Ratcheson R. A,, Salford L. G.. and Siesjo 9. K. (1973) Optimal freezing conditions for cerebral metabolites in rats. J. Neurochem. 21, 1127-1 138.

Rose R. J., Hodgson D. R., Kelso T. B., McCutcheon L. J., Reid T. A,, Bayly W. M., and Gollnick P. D. (1988) Maximum 0, uptake, O2 debt and deficit, and muscle metabolites in thor- oughbred horses. J. Appl. Physiol. 64,781-788.

Siesjo B. K. (1985) Acid-base homeostasis in the brain: physiology, chemistry, and neurochemical pathology, in MolecularMecha- nisms of Ischemic Brain Damage (Kogure K., Hossmann K.-A,, Siesjo B. K., and Welsh F. A., eds), pp. 121-154. Else- vier, Amsterdam/New York/Oxford.

Siesjo 9 . K., Ekholm A,, Katsura K., and Theander S. (1990) Acid- base changes during complete brain ischemia. Stroke 21

Smith M.-L., von Hanwehr R., and Siesjo B. K. (1986) Extra- and intracellular pH during near-complete forebrain ischemia in the rat. J. Neurochem. 46, 331-339.

Tofts P. S. and Wray S. ( 1985) Changes in brain phosphorus metabolites during the post-natal development of the rat. J. Physiol.

Vannucci R. C. (1990) Experimental biology of cerebral hypoxic- ischemia: relation to perinatal brain damage. Pediatr. Res. 27,

Volpe J. J. ( 1987) Hypoxic-ischemic encephalopathy: basic aspects and fetal assessment. in Neurology ofthe Newborn (Volpe J. J., ed), pp. 160-195. W. B. Saunders, Philadelphia.

von Hanwehr R., Smith M. L., and Siesjo B. K. (1986) Extra- and intracellular pH during near-complete forebrain ischemia in the rat. J. Neurochem. 46, 33 1-339.

Wilkie D. R. (1979) Generation of protons by metabolic processes other than glycolysis in muscle cells: a critical view. J . Mol. Cell. Cardiol. 11, 325-330.

781-788.

38, 1569-1574.

(SUPPI. Ill), 111-194-111-199.

359,4 17-429.

3 17-326.

J N~~itroclietn.. 1'01. 59, No I . 1992

Cerebral Acid Buffering Capacity at Different Ages Measured In Vivo by 31 P and 1 H Nuclear Magnetic...

Documents

Transcript of Cerebral Acid Buffering Capacity at Different Ages Measured In Vivo by 31 P and 1 H Nuclear Magnetic...