In vivo phosphocreatine and ATP in piglet cerebral gray and white matter during seizures

Brain Research 783 Ž1998. 19–27

Research report

In vivo phosphocreatine and ATP in piglet cerebral gray and white matter during seizures D. Holtzman

a, )

, R. Mulkern b , R. Meyers a , C. Cook c , E. Allred d , I. Khait a , F. Jensen a , M. Tsuji e , P. Laussen f

a

Department of Neurology, Children’s Hospital and HarÕard Medical School, Boston, MA 02115, USA Department of Radiology, Children’s Hospital and HarÕard Medical School, Boston, MA 02115, USA c Department of Radiology, Deaconess Hospital, Boston, MA 02115, USA Department of Neurology (Neuroepidemiology), Children’s Hospital and HarÕard Medical School, Boston, MA 02115, USA e Department of Pediatrics (Neonatology), Children’s Hospital and HarÕard Medical School, Boston, MA 02115, USA f Department of Anesthesia, Children’s Hospital and HarÕard Medical School, Boston, MA 02115, USA b

d

Accepted 7 October 1997

Abstract The creatine kinase ŽCK. reaction is thought to be important in coupling ATP metabolism and regulating ADP concentration in tissues with high and variable ATP turnover, including cerebral gray matter ŽGM.. There is low phosphocreatine ŽPCr., low CK reaction rates, and high mitochondrial CK ŽMiCK. isoenzyme activity in GM compared to white matter ŽWM.. To compare the CK reaction in GM and WM when ATP metabolism is high, CK reactants and reaction rates were measured in predominantly GM and WM slices in vivo in 2 and 14-day old piglets during pentylenetetrazole ŽPTZ. seizures using 31 P nuclear magnetic resonance ŽNMR. 1-dimensional chemical shift imaging ŽCSI.. Arterial pressure, temperature, and blood gasses were stable at both ages. Before seizures, the PCrrnucleoside triphosphate ŽNTP. ratio was higher in WM than GM at both ages with a developmental increase seen in WM. The CK reaction rate constant increased in both regions between 2 and 14 days. During seizures, PCrrNTP increased in GM at 14 days due to increased PCr while the ratio and PCr decreased in WM. The NTP was more stable in WM and GM at both ages. The CK reaction rate decreased in both regions more at 2 than at 14 days. Thus, brain ATP, deduced from NTP, is stable during seizures in the piglet. In GM stable ATP is associated with a unique increase in PCr concentration. q 1998 Elsevier Science B.V. Keywords: Brain; Creatine kinase; Seizure; White matter; Gray matter; Nuclear magnetic resonance spectroscopy

1. Introduction Study of brain ATP metabolism, including synthesis and utilization, is complicated by regional and cellular heterogeneity. Although rates of ATP turnover are not known, ATP synthesis and ATPase activities are higher and vary more in gray matter ŽGM. than white matter ŽWM. w11,34x. In some species, including altricial rodents, piglets, and the human, brain ATP metabolism increases post-natally w3,9,13–16x. In rats and piglets, this metabolic maturation is later in WM than GM w7,19x.

)

Corresponding author. Department of Radiology, Brigham and Women’s Hospital, 221 Longwood Ave., Boston, MA 02115, USA. Fax: q1-617-278-0610; E-mail: [email protected] 0006-8993r98r$19.00 q 1998 Elsevier Science B.V. All rights reserved. PII S 0 0 0 6 - 8 9 9 3 Ž 9 7 . 0 1 2 6 8 - 7

The mature brain, like other tissues with high and variable rates of ATP metabolism, has high phosphocreatine ŽPCr. concentration and creatine kinase ŽCK. activities, including cytosolic ŽBCK. and mitochondrial ŽMiCK. isoenzymes w11,18,22,30,39x. The CK catalyzed reaction k

PCryq MgADPyq Hq l Cr q MgATPy generally is thought to be important in providing an energy source for rapid resynthesis of ATP and for regulating ATP and ADP concentrations w8,18,39x. The GM has higher MiCK, lower PCrrnucleoside triphosphate ŽNTP. and PCrrCr ratios, and lower CK reaction rates than does WM w16,18,19,23,37,38x. These differences suggest that the physiology of the CK reaction differs in the high and variable ATP metabolism of GM from that in the slower ATP metabolism of WM.

20

D. Holtzman et al.r Brain Research 783 (1998) 19–27

Cerebral hemisphere ATP metabolic rates, including cerebral blood flow ŽCBF. and anaerobic and aerobic glycolysis, increase 2 – 3 fold during seizures w4,6,10,21,24,25,31x. Thus, the transition from baseline to a seizure state is appropriate for in vivo study of the CK reaction in high ATP turnover states. The CK reactants and reaction have been studied in vivo during seizures using 31 P nuclear magnetic resonance ŽNMR. spectroscopy, localized to the cerebral hemispheres by a surface coil on the skull or scalp w20,28,32,40x. These studies show little or no change in ATP while PCr drops about 20–40% indicating close coupling of ATP synthesis and utilization in the high energy turnover seizure state. This close regulation of ATP in vivo during seizures also is true in the immature rat brain in which the CK reaction rates are slower and PCr concentration is less than in the adult brain w16,20x. In this report, pentylenetetrazole ŽPTZ.-induced seizures were used to study the physiology of the brain CK reaction. Regulation of PCr and ATP concentrations and the CK catalyzed reaction rates were measured in vivo in predominantly GM and WM regions of the piglet brain. These regional measurements were made using the 31 P NMR one dimensional chemical shift imaging experiment ŽCSI.. To determine the effects of maturation, metabolic changes were compared in piglets 2 days and 2 weeks of age. The results of these studies provide the first in vivo measures of the CrrCKrPCr system in developing cerebral GM and WM under the condition of high ATP turnover. 2. Methods 2.1. Animals Yorkshire piglets ŽParsons Breeding Farm, Hadley, MA. were studied at 2–3 days or 2 weeks of age as described previously w37x. Piglets arrived the day before an experiment and were fed formula or feed provided by the breeder. The last feeding was about 6 h before the experiment. The piglet was anesthetized with urethane Ž1400 mg kgy1 i.p.., endotracheally intubated, paralyzed with pancuronium Ž0.1 mg kgy1 each h., and mechanically ventilated. Rectal temperature was maintained between 34–368C with a warming blanket. Catheters were inserted through the surgically exposed right femoral artery for arterial blood gas sampling and continuous blood pressure and heart rate monitoring and the vein for injection of PTZ Ž80 mg kgy1 . to initiate seizures. No animals required additional anesthesia. Catheters were continuously flushed at 2 cc hy1 with heparinized normal saline. Arterial blood gasses were analyzed in heparinized samples ŽCiba Corning Diagnostics, Medfield, MA.. Mechanical ventilation was adjusted to maintain arterial pH 7.35–7.45, PaCO 2 35–45 mmHg, and PaO 2 80–120 mmHg. Subcutaneous platinum electrodes were placed biparietally for electroencephalogram ŽEEG. recording. The cop-

per wires were wrapped around each other and fixed to the lucite carrier to assure no movement in the magnetic field. The leads were connected through a low pass filter Ž80 dB at 1 MHz. with shielded cables. A single turn copper surface coil Ž3 cm diameter for 2–3-day old piglets, 4 cm for 2-week old piglets. was sown to the intact scalp centered with the anterior edge just posterior to the supraorbital ridges. A capillary tube containing saturated phenylphosphoric acid in water was positioned in the center of the surface coil to mark the scalp surface for the CSI. The piglet then was positioned in the magnet lying supine with the head centered in the volume coil. Following data collection, the anesthetized piglet was euthanized with intravenous KCl. 2.2. NMR experiments All studies were performed on a 4.7 T, 30 cm horizontal bore Bruker Bio-spec NMR System ŽBruker Instruments, Billerica, MA.. A 17 cm diameter volume coil ŽUS ASIA Instruments, Highland Heights, OH. was used in the transmit mode and the surface coil was used in the receive mode. Surface coil localized 1-pulse brain spectra were acquired after the surface coil was tuned and matched to the 31 P frequency and the magnetic field was optimized using the water 1 H signal. The 908 pulse width usually was 100 m s Spectra were averages of 32 acquisitions with 2 s Žpartially saturated. or 10 s Žalmost fully relaxed. recycle times ŽRD.. The CSI experiments were acquired as previously described w19,37x. A 100 m s volume coil transmit pulse for spin excitation was followed by a phase encoding step applied parallel to the surface coil axis. Signal readout was acquired for 256 ms. A total of 32 phase encoding steps were applied. The RD was 2 s with 32 signal averages per phase encoding step. The field of view was 4 or 6 cm. The CSI data was zero filled to 64 in the spatial domain and 1024 in the frequency domain. Previous studies using these experimental parameters resulted in no systematic differences in the PCrrNTP ratio with distance from the surface coil and no signal contamination from facial or tongue muscles w37x. The surface coil localized ST experiments were carried out as shown in previous publications w12,16,17x. In order to measure transfer of phosphoryl groups from PCr to ATP, selective saturation of the g-phosphorus NTP signal Žg-NTP. was achieved using a 1 s low power pulse centered on this frequency. After the saturating pulse, the usual 908 high power non-selective acquisition pulse was applied and the free induction decay was acquired to record the net z-magnetization. Phosphorus flux in the CK catalyzed reaction was analyzed with a first order exchange simplification w2x. k

PCr l ATP

D. Holtzman et al.r Brain Research 783 (1998) 19–27

The unidirectional pseudo-first order rate constant for the PCr to ATP exchange is equal to the ratio of the chemical flux, J, to the PCr concentration, as measured directly in the ST experiment, k f s Jr w PCr x s 1r Ž T 1 . PCr =

Mo y Ms Ms

where T 1 is the longitudinal relaxation time of PCr in the absence of chemical exchange, Ms is the PCr signal in the presence of g-ATP saturation, and Mo is the PCr signal with the saturating pulse an equal distance on the opposite side of the PCr peak. The T 1 value of 3 s for PCr was taken from a modified progressive saturation experiment in mouse brain w16x. Use of this model does not imply that the reaction is first order. Only the forward rate constant was measured in these experiments but the CK reaction rates in brain are equal in the two directions when reactant concentrations are stable w16x. The ST experiment and CSI were combined ŽST–CSI. in order to compare the CK-catalyzed reaction rates in predominantly GM and WM slices w19x. The two CSI experiments comprising the ST–CSI were run sequentially. In the first the short acquisition pulse was preceded by a 1 s low power pulse at the g-NTP frequency. The acquisition pulse was followed by the usual phase encoded gradient and acquisition periods. After acquiring the full set of saturation spectra, control spectra were acquired with the saturating pulse placed on the opposite side of the PCr peak. 2.3. Experimental design A sequence of 31 P NMR studies and a 1 H image were acquired in five piglets aged 2–3 days and six piglets aged 2 weeks. First, surface coil localized spectra were acquired with 2 and 10 s RD. An initial CSI then was acquired, followed by a surface coil localized ST experiment and the ST–CSI. These experiments required about 1.5 h. The PTZ then was given intravenously. A series of one-pulse spectra Ž2 s RD. were acquired in order to follow changes in PCr and NTP concentrations. When these spectra were stable Žusually 5 min., the CSI, ST, and ST–CSI experiments were repeated. A final surface coil localized spectrum was acquired to be sure no ischemic changes in PCr

21

had occurred during the ST–CSI. Intra-arterial pressure, heart rate, and rectal temperature were acquired continuously. The EEG was acquired between each NMR experiment. Arterial blood gas measurements were obtained before seizures, after the seizure CSI, and after the ST–CSI. At the end of each study, a T 2 weighted 1 H image was acquired in the usual manner w37x. 2.4. Data analysis The surface coil localized spectra were individually phased and peaks were integrated after baseline flattening using the convolution difference technique w1x. The baseline ratios of PCr and b-NTP peak areas were taken from the almost fully relaxed Ž10 s. spectra. The changes during seizures were determined by comparing the partially saturated 2 s spectra before and during seizures. The pH values were taken from the chemical shift of the Pi peak relative to that of PCr w29x. The pseudo-first order forward reaction rate constants Ž k . for the surface coil localized ST experiments were calculated from the integrated PCr peaks of control and saturated spectra. All CSI data sets were reconstructed and analyzed as power spectra in order to avoid individual phasing of spectra from each slice. Spectral peaks were fit with Lorentzian functions and areas were integrated ŽMacFid Software, Tecmag, Houston, TX.. The square roots of peak areas were used to calculate PCrrb-NTP ratios which are proportional to relaxation weighted concentration ratios for the two metabolites w5x. The PCrrNTP ratios were calculated from integrals of peaks from each 1.9 mm slice. All the analytic steps were performed by one person ŽIK.. Printed spectra and peak integrals were reviewed by another observer ŽDH.. Projection depths for the surface ŽGM. and supraventricular ŽWM. slices were chosen from T 1 weighted 1 H images acquired for each animal by one observer ŽRM. who was unaware of the 31 P spectral results. The EEGs were reviewed by one observer ŽFJ. who also was unaware of the NMR results. 2.5. Statistics Results are expressed as the median and range of experimental values. Non-parametric analyses were chosen

Table 1 Mean arterial pressures ŽMAP., arterial blood gas concentrations, and arterial pH before and during penetylenetetrazole-induced seizures in 2-day old Ž n s 5. and 14-day old piglets Ž n s 6. Age

Condition

MAP ŽmmHg.

PO 2 ŽmmHg.

PCO 2 ŽmmHg.

pH Žunits.

2 Days

Baseline Seizure Baseline Seizure

45 Ž35–65. 45 Ž28–75. p s 0.50 75 ) Ž55–95. 65 Ž25–75. p s 0.04

107 Ž81–96. 110 Ž95–103. p s 0.07 79 ) Ž66–94. 65 Ž64–111. p s 0.35

30 Ž28–45. 31 Ž20–32. p s 0.27 35 Ž34–42. 33 Ž29–43. p s 0.35

7.39 Ž7.22–7.43. 7.26 Ž7.19–7.27. p s 0.14 7.37 Ž7.33–7.42. 7.24 Ž7.21–7.28. p s 0.04

14 Days

Results are shown as the median and range of values measured under each condition. Baseline values are compared at the two ages as unpaired samples and are compared to values during seizures as paired samples. Significance values F 0.05 are indicated Ž ) . for baseline values at the two ages.

D. Holtzman et al.r Brain Research 783 (1998) 19–27

22

Table 2 Baseline and seizure PCrrNTP ratios, pH and creatine kinase pseudo-first order reaction rate constants Ž k . measured by surface coil localized spectroscopy

PCrrNTP pH k Žsy1 .

Baseline Seizures Baseline Seizures Baseline Seizures

2 Days

14 Days

1.06 Ž0.56–1.34. 0.69 Ž0.40–0.85. p s 0.08 7.06 Ž7.00–7.17. 7.06 Ž7.03–7.11. p s 0.92 0.37 Ž0.25–1.13. 0.69 Ž0.40–0.85. p s 0.50

0.76 Ž0.60–1.51. 0.73 Ž0.60–1.34. p s 0.60 7.07 Ž7.03–7.11. 7.06 Ž7.01–7.10. p s 1.0 0.26 Ž0.01–0.70. 0.94 Ž0.60–1.66. p s 0.03

31

P NMR

p s 0.85 p s 1.0 p s 0.20

The Yorkshire piglets were 2 days old Ž n s 5. or 14 days old Ž n s 6.. All values are medians with ranges shown in parentheses.

because of the relatively small number of animals studied. Significance of differences between pre-seizure and seizure measurements at the two ages were evaluated using Wilcoxon’s ranked sum test. Significance for differences between GM and WM slices in each group before or during seizures were calculated using analysis of covariance on ranked data Žnon-parametric ANCOVA.. Significance of changes in regional reactant concentrations or rate constants at either age also were calculated using analysis of covariance on ranked data.

3. Results

spectra are shown in Table 2. Brain PCrrNTP ratios and pH values were the same at 2 and 14 days. The CK reaction rate constants also were the same at the two ages. The PCrrNTP ratios and pseudo-first order rate constants from predominantly GM and WM slices of 2 and 14-day old piglets are shown in Table 3. The brain images showed supraventricular slices to be 0.8–1.1 cm below the paramedian cortical surface at both ages. The baseline PCrrNTP ratios were greater in WM than in GM at both ages but the difference was significant only at 14 days. The rate constants were the same in the two regions at 2 days and were about 20% higher Žnot significant. in WM than in GM at 14 days. The rate constants increased with age in both WM and GM, reaching significance in WM.

3.1. Baseline

3.2. Seizures

Baseline arterial blood gas concentrations, pH and mean arterial pressures in 2-day old and 14-day old piglets are shown in Table 1. Mean arterial pressures were consistently higher in older than in younger piglets. Arterial O 2 concentrations were higher and CO 2 concentrations were lower in the younger animals, suggesting that older piglets were not as effectively ventilated. Arterial pH values were the same at the two ages. Baseline measurements of CK reactant concentrations and CK reaction rate constants from surface coil localized

In 2-day old piglets, mean arterial pressures did not change with onset of seizures ŽTable 1.. Arterial PO 2 showed a small increase in all the animals and pH became more acidic in most. In contrast, the older animals showed significant decreases in arterial pressure and pH. Ictal spike discharges appeared within 30 s of the intravenous PTZ injection in animals of both ages Žfour pigs at 2 days, six pigs at 14 days.. Regular high amplitude spikes continued to be seen between NMR studies after seizure onset.

Table 3 Creatine kinase reactant ratios ŽPCrrNTP. and pseudo-first order reaction rate constants Ž k . in predominantly GM and predominantly WM slices from 2 and 14-day old piglets

PCrrNTP k Žsy1 .

Region

2 Days

14 Days

Gray White Gray White

0.38 Ž0.14–1.51. 0.67 Ž0.25–3.08. p s 0.17 0.32 Ž0.23–1.40. 0.33 Ž0.17–0.56. p s 0.75

0.41 Ž0.20–0.84. 1.08 Ž1.03–1.78. p s 0.004 0.87 Ž0.46–4.0. 1.03 Ž0.0–1.04. p s 0.87

p s 1.0 p s 0.47 p s 0.14 p s 0.05

Each value is the median and range for five or six experiments. Values were taken from localized 31 P NMR spectra acquired by the CSI technique Žsee Section 2.. The p-values were calculated by analysis of covariance on ranked data for the GM and WM comparisons and as unranked data for maturational changes in each region.

D. Holtzman et al.r Brain Research 783 (1998) 19–27

23

Fig. 1. Changes in PCr and NTP concentrations during pentylenetetrazole-induced seizures were calculated from surface coil localized 31 P spectra Žsee Section 2.. The concentration changes during seizures are expressed as a ratio of values during seizuresrpre-seizure values for 2-day old and 14-day old piglets. The PCr concentration decreased significantly during seizures at 2 days Ž p s 0.04. and at 14 days Ž p s 0.03.. Decreases in NTP concentration approached significance at 2 days Ž p s 0.08. and at 14 days Ž p s 0.17..

Fig. 2. Changes in GM and WM PCrrNTP ratios during seizures in 2-day old and 14-day old piglets. All results are ratios of seizurerpre-seizure values measured in vivo in superficial predominantly GM and supra-ventricular predominantly WM slices using the CSI NMR experiment as described in Section 2. The PCrrNTP ratio was significantly increased in the GM slices at 14 days Ž p s 0.03., but not at 2 days Ž p s 0.22.. The ratio did not change in WM.

24

D. Holtzman et al.r Brain Research 783 (1998) 19–27

The surface coil localized PCrrNTP ratios were not different during seizures compared to pre-seizure values at 14 days ŽTable 2.. This reactant ratio decreased in the 2-day old piglets. Brain pH values did not change at either age. The CK reaction rate constants increased during seizures only at 14 days. In contrast to the PCrrNTP ratio,

the PCr and NTP concentrations decreased during seizures compared to the pre-seizure values in both the 2 and 14-day old piglets ŽFig. 1.. Only the decreases in PCr reached significance. The PCrrNTP ratio increased during seizures in GM slices in 14-day old but not in 2-day old piglets ŽFig. 2..

Fig. 3. Changes in PCr and NTP concentrations during seizures are shown in brain slices from superficial predominantly GM and supra-ventricular predominantly WM regions in 2-day old Župper figure. and 14-day old Žlower figure. piglets. Each result is expressed as the ratio of seizurerpre-seizure values measured by the CSI experiment described in Section 2. The PCr concentration was significantly increased in GM in the 14-day old piglets Ž p s 0.05.. The PCr increase in the GM was not significant at 2 days. The PCr did not change in WM at either age. The NTP concentration was unchanged at both ages in each region.

D. Holtzman et al.r Brain Research 783 (1998) 19–27

25

Fig. 4. Changes in brain creatine kinase pseudo-first reaction rate constants Ž k . are shown for predominantly GM and WM slices before and during pentylenetetrazole-induced seizures in 2-day old Župper figure. and 14-day old Žlower figure. piglets. The values were calculated from the CSI saturation transfer experiments described in Section 2. At 2 days, the rate constants decreased during seizures in WM Ž p s 0.04.. The decrease in GM approached significance Ž p s 0.17.. At 14 days, there were no changes in the rate constants.

There were no changes in PCrrNTP ratios during seizures compared to baseline values in WM in 2 or 14-day old piglets. There were no changes in NTP during seizures in either region in the 2 or 14-day old piglets ŽFig. 3.. The PCr concentration in GM was unchanged in 2-day old piglets and increased in 14-day old piglets. The PCr did not change in WM with the onset of seizures at either age. The CK catalyzed reaction rate constants decreased

during seizures in both GM and WM at 2 days ŽFig. 4.. At 14 days, there were no changes in the k values in either region. 4. Discussion Baseline surface coil localized brain spectra, comprised of signals from undetermined quantities of GM and WM,

26

D. Holtzman et al.r Brain Research 783 (1998) 19–27

are similar to those previously observed in the mature piglet w19,36,37x. Brain PCrrNTP ratio does not change between 2–14 days unlike the postnatal increases previously described in miniature piglets and the large increase in both altricial rodents and human newborns w3,9,16,19,35x. The stable CK reaction rate constant in the post-natal Yorkshire piglet also is in contrast to the large increases seen in miniature piglets and the rat and human w16,19x. Thus, the increases in brain PCrrNTP and in CK catalyzed reaction rates either are much smaller or occur earlier in the piglet than in the altricial rodent. The PCr decrease and stable NTP concentration during seizures at both ages are similar to reactant concentration changes seen in the rat w20,40x. The increase in reaction rate constant in the more mature piglets during seizures is larger than in the rat w20x. The reaction rate constants do not increase in the immature rat or piglet. Results of CSI studies show that regulation of brain CK reactants during a high ATP turnover state is different in GM and WM. As seen previously, the PCrrNTP ratio is larger in WM than GM in the mature brain due to a postnatal increase only in WM w19x. During seizures the PCrrNTP ratio and the two reactant concentrations decrease a small amount in the predominantly WM slices. In contrast, the PCrrNTP ratio increases in GM during seizures more in older than in younger piglets. This increase is due to increased PCr. As in the predominantly WM slice, ATP concentration, deduced from the b-NTP signal, does not change during seizures in GM at either age. The stable NTP signal from GM and WM slices is consistent with the stable NTP seen in surface coil localized spectra. The PCr decrease in the surface coil localized spectra may reflect the metabolic change in WM. The small decrease in PCr concentration in the deeper WM slice may be due to averaging the PCr increase in GM and decrease in WM, since the latter makes up only about 60% of this slice w37x. The pseudo-first order CK rate constants during seizures were different when measured in surface coil localized and CSI saturation transfer experiments. In the surface coil localized experiment, reaction rate constants increase in the more mature piglets. In the CSI experiment, the more mature brain shows no changes in rate constants while the immature brain shows a decrease in both the WM and GM slices. These differences are not readily explained by differences in the NMR techniques. The experimental differences in CK rate constants can be interpreted cautiously using previous observations in other species. The piglets were paralyzed and ventilated with close monitoring of blood gasses and arterial pressure. Lower arterial O 2 and mild hypercarbia in the 14-day old pigs suggest that ventilation was less effective than in younger piglets. However, there were no indications of ischemic changes in brain 31 P spectra in these animals. During seizures, the arterial pH decreased without changes in brain pH at both ages w28,29x. Importantly, arterial

pressure decreased during seizures in the older piglets. This decrease suggests the animals had passed the transition point, described in prolonged seizures in rats, after which the brain may be relatively hypoperfused w25x. It is unlikely that decreased arterial pressure contributes to increased PCr and stable ATP in GM in the seizing 14-day old piglets. A relatively decreased CBF could contribute to the decreased CK reaction rates during seizures in both GM and WM after the long CSI saturation transfer experiment w17x. These results provide a comparison of the PCrrCKrCr system and energy regulation in cerebral GM and WM during seizures. Seizures produce a high ATP turnover state in GM, while in WM the ATP turnover is uncertain w10,21x. The PCrrCKrCr system differs in these regions, with lower PCr and higher Cr concentrations, high MiCK activity, and lower CK catalyzed reaction rates in GM w18,19,37x. During seizures the PCr concentration decreases in WM and increases in GM. With the different CK systems, the ATP concentration is closely maintained in both regions. Similarly, ATP does not change during seizures in either region in the younger animals, where PCr is relatively stable. The observation that PCr concentration increases only in the mature GM supports the proposal that MiCK, which is present in large concentrations only in mature GM, is important in this metabolic response to increased ATP demand w19,23,39x. A second proposal is that the increase in PCr concentration is due to Cr and PCr moving closer to their equilibrium concentrations w33x. Understanding regulation of ATP metabolism during high ATP turnover states is central in understanding the physiology of energy requiring processes in brain. This study demonstrates the importance of metabolic heterogeneity in understanding brain energy regulation during seizures and, probably, other states of increased ATP turnover. The coupling of PCr and ATP concentrations differs markedly in cerebral GM and WM in the piglet. The small postnatal changes in brain ATP metabolism suggest that the piglet matures early and may be of limited value in studying development of brain energy metabolism as seen in the pre-term and term human newborn w3x. The small decrease in PCr with stable ATP in mature WM is consistent with PCr as a temporal energy reserve as suggested for skeletal muscle w26x. The PCr increase in GM during seizures has not been described previously and suggests a novel physiological mechanism in ATP regulation. Faster localized phosphorus NMR acquisitions will allow further study of these critical physiological questions in brain w27x.

Acknowledgements This work was supported by research grants to D.H. from the Cerebral Palsy Foundation ŽR-4010. and the National Institutes of Health ŽNS 26371.. M.T. was sup-

D. Holtzman et al.r Brain Research 783 (1998) 19–27

ported in part by a NIH post-doctoral training grant ŽNS 07264. and a Physician Scientist Award ŽHD 01010..

References w1x J. Ackerman, T. Grove, G. Wong, D. Gadian, G. Radda, Mapping of metabolites in whole animals by 31 P NMR using surface coils, Nature 283 Ž1980. 167–179. w2x J. Alger, R. Shulman, NMR studies of enzymatic rates in vitro by magnetization transfer, Q. Biophys. Rev. 17 Ž1984. 83–124. w3x D. Azzopardi, J. Wyatt, P. Hamilton, E. Cady, D. Delpy, P. Hope, E. Reynolds, Phosphorus metabolites and intracellular pH in the brains of normal and small for gestational age infants investigated by magnetic resonance spectroscopy, Pediatr. Res. 25 Ž1989. 440– 444. w4x G. Blennow, J. Folbergrova, B. Nilsson, B. Siesjo, Cerebral metabolic and circulatory changes in the rat during sustained seizures induced by DL-homocysteine, Brain Res. 179 Ž1979. 129–146. w5x E. Cady, Clinical Magnetic Resonance Spectroscopy, Plenum, New York, 1990, 278 pp. w6x A. Chapman, E. Westerberg, B. Siesjo, The metabolism of purine and pyrimidine nucleotides during insulin-induced hypoglycemia and recovery, J. Neurochem. 36 Ž1981. 179–189. w7x A. Chesler, H. Himwich, Comparative studies of the rates of oxidation and glycolysis in the cerebral cortex and brain stem of the rat, Am. J. Physiol. 141 Ž1944. 513–517. w8x M. Conway, J. Clark, Creatine and Creatine Phosphate, Academic Press, New York, 1996, 242 pp. w9x R. Corbett, A. Laptook, Age-related changes in swine brain creatine kinase-catalyzed 31 P exchange measured in vivo using 31 P NMR magnetization transfer, J. Cereb. Blood Flow Metab. 14 Ž1994. 1070–1077. w10x J. Engle, D. Kuhl, M. Phelps, Patterns of human local cerebral glucose metabolism during epileptic seizures, Science 218 Ž1982. 2–19. w11x M. Erecinska, I. Silver, ATP and brain function, J. Cereb. Blood Flow Metab. 9 Ž1989. 2–19. w12x S. Forsen, R. Hoffmann, Study of moderately rapid chemical exchange reactions by means of nuclear double resonance, J. Chem. Phys. 39 Ž1963. 2892–2901. w13x P. Greengard, H. McIlwain, Metabolic responses to electric pulses in mammalian cerebral tissues during development, in: H. Waelsch ŽEd.., Biochemistry of the Developing Nervous System, Academic Press, New York, 1955, pp. 251–160. w14x H. Himwich, Historical review, in: W. Himwich ŽEd.., Developmental Neurobiology, Charles C. Thomas, Springfield, 1979, pp. 22–44. w15x D. Holtzman, J. Olson, Developmental changes in brain cellular energy metabolism in relation to seizures and sequelae, in: H. Jaspers, N. van Gelder ŽEds.., Basic Mechanisms of Neuronal Hyperexcitability, Liss, New York, 1983, pp. 423–429. w16x D. Holtzman, E. McFarland, D. Jacobs, M. Offutt, L. Neuringer, Maturational increase in mouse brain creatine kinase reaction rates shown by phosphorus magnetic resonance, Dev. Brain Res. 58 Ž1991. 181–188. w17x D. Holtzman, M. Tsuji, M. Offutt, L. Neuringer, D. Jacobs, Creatine kinase reaction rates in the cyanide-poisoned mouse brain, J. Cereb. Blood Flow Metab. 13 Ž1993. 153–161. w18x D. Holtzman, M. Tsuji, T. Wallimann, W. Hemmer, Functional maturation of creatine kinase in rat brain, Dev. Neurosci. 15 Ž1993. 261–270.

27

w19x D. Holtzman, R. Mulkern, M. Tsuji, C. Cook, R. Meyers, Phosphocreatine and creatine kinase systems in gray and white matter of the developing brain, Dev. Neurosci. 18 Ž1996. 535–542. w20x D. Holtzman, R. Meyers, I. Khait, F. Jensen, Brain creatine kinase reaction rates and reactant concentrations during seizures in developing rats, Epilepsy Res. 27 Ž1997. 7–11. w21x D. Howse, J. Caronna, T. Duffy, F. Plum, Cerebral energy metabolism, pH, and blood flow during seizures in the cat, Am. J. Physiol. 277 Ž1974. 1444–1450. w22x W. Jacobus, A. Lehninger, Creatine kinase of rat heart mitochondria, J. Biol. Chem. 248 Ž1973. 4803–4810. w23x P. Kaldis, W. Hemmer, E. Zanolla, D. Holtzman, T. Wallimann, ‘Hot spots’ of creatine kinase localization in the brain: cerebellum, hippocampus and choroid plexus, Dev. Neurosci. 18 Ž1996. 542–554. w24x J. Klein, N. Olson, Effects of convulsive activity upon the concentration of brain glucose, glycogen, lactate, and phosphates, J. Biol. Chem. 167 Ž1947. 747–756. w25x N. Kreisman, J. LaManna, M. Rosenthal, T. Sick, Oxidative metabolic responses with recurrent seizures in rat cerebral cortex: role of systemic factors, Brain Res. 218 Ž1981. 175–188. w26x E. McFarland, M. Kushmerick, T. Moerland, Activity of creatine kinase in a contracting mammalian muscle of uniform fiber type, Biophys. J. 67 Ž1994. 1912–1924. w27x R. Mulkern, H. Chao, J. Bowers, D. Holtzman, Multi-echo approaches to spectroscopic imaging of brain, in: D. Lester ŽEd.., New Developments in Metabolic Brain Imaging, Ann. NY Acad. Sci., 820 Ž1997. 97–122. w28x O. Petroff, J. Prichard, K. Behar, J. Alger, R. Shulman, In vivo phosphorus nuclear magnetic resonance spectroscopy in status epilepticus, Ann. Neurol. 16 Ž1984. 169–177. w29x O. Petroff, J. Prichard, K. Behar, J. Alger, J. den Hollander, R. Shulman, Cerebral intracellular pH by 31 P nuclear magnetic resonance spectroscopy, Neurology 35 Ž1985. 781–788. w30x J. Prichard, J. Alger, K. Behar, O. Petroff, R. Shulman, Cerebral metabolic studies in vivo by 31 P NMR, Proc. Natl. Acad. Sci. USA 80 Ž1983. 2748–2751. w31x B. Saktor, J. Wilson, C. Tickert, Regulation of glycolysis in brain in situ during convulsions, J. Biol. Chem. 241 Ž1966. 5071–5075. w32x A. Sauter, M. Rudin, Determination of creatine kinase kinetic parameters in rat brain by NMR magnetization transfer, J. Biol. Chem. 268 Ž1993. 13166–13171. w33x J. Schlegel, M. Wyss, H. Eppenberger, T. Wallimann, Functional studies with the octameric and dimeric forms of mitochondrial creatine kinase, J. Biol. Chem. 265 Ž1990. 9221–9227. w34x B. Siesjo, Brain Energy Metabolism, Wiley, Chichester, 1978, 607 pp. w35x P. Tofts, S. Wray, Changes in brain phosphorus metabolites during the postnatal development of the rat, J. Physiol. 359 Ž1985. 417–429. w36x M. Tsuji, H. Naruse, J. Volpe, D. Holtzman, Cerebral oxygenation and energy state in hypoxic piglets, Pediatr. Res. 37 Ž1995. 253–259. w37x M. Tsuji, R. Mulkern, C. Cook, R. Myers, D. Holtzman, Relative phosphocreatine and nucleoside triphosphate concentrations in cerebral gray and white matter measured in vivo by 31 P nuclear magnetic resonance, Brain Res. 707 Ž1996. 146–154. w38x T. Whittingham, A. Douglas, D. Holtzman, Creatine and nucleoside triphosphates in rat cerebral gray and white matter, Cereb. Metab. Dis., in press. w39x M. Wyss, J. Smeitink, R. Wevers, T. Wallimann, Mitochondrial creatine kinase: a key enzyme of aerobic energy metabolism, Biochim. Biophys. Acta 1102 Ž1992. 119–166. w40x R. Young, M. Osbakken, R. Briggs, S. Yagel, D. Rice, S. Goldberg, 31 P NMR study of cerebral metabolism during prolonged seizures in the neonatal dog, Ann. Neurol. 18 Ž1985. 14–20.