Effect of insulin hypoglycemia upon cerebral energy metabolism and EEG activity in the rat

Brain Research, 126 (1976) 263-280

263

'~5 Elsevier/North-Holland Biomedical Press, Amsterdam Printed in The Netherlands

E F F E C T OF I N S U L I N H Y P O G L Y C E M I A U P O N C E R E B R A L E N E R G Y M E T A B O L I S M A N D E E G A C T I V I T Y IN T H E RAT

G. FEISE*, K. KOGURE, R. BUSTO, P. SCHEINBERG and O. M. REINMUTH Cerebral Vascular Disease Research Center, Department of Neurol¢L~y, University of Miami School of Medicine, Miami, Fla. (U.S.A.)

(Accepted September 3rd, 1976)

SUMMARY Anesthetized ventilated rats were subjected to insulin-induced hypoglycemia (50 units/kg i.v.) while EEG, ECG, mean arterial pressure, blood gases, arterial pH and rectal temperature were controlled. Animals were sacrificed by rapid transcalvarial freezing of the brain in situ. Glucose, pyruvate and lactate were measured in blood, CSF and cortical tissue, in which additionally glycogen, phosphocreatine, ATP, ADP, AMP, aketoglutarate (aKG), glutamate, oxalacetate, aspartate, ammonia and water content were estimated. A T P / A D P ratio, energy charge (ECh) energy reserve, N A D H / N A D ~- quotient and intracellular pH were calculated. ECh does not correlate with either dysfunction or carbohydrate depletion, but declines in a threshold fashion when tissue glucose has fallen by over 97% and glycogen by over 60%. The E E G correlates with the degree and duration of carbohydrate depletion in cortical tissue. An isoelectric EEG occurs pari passu with the fall of the ECh. Increase in ammonia and decrease in a K G and Glut are supportive evidence of intrinsic substrate. Lactate decrease during hypoglymecia is not reversed by superimposed hypoxia.

INTRODUCTION Insulin hypoglycemia seems to be a suitable model to study the effects of acute substrate deficiency on cerebral energy metabolism isolated from other factors operative in cerebral ischemia. Cerebral blood flow (CBF) has been demonstrated to remain normal in severe hypoglycemia ~8 and systemically adminsitered insulin is generally

* Dammweg 13, 6900 Heidelberg, G.F.R.

264 accepted to have no direct effect on brain metabolism when hypoglycemia is allowed to occur aS. Since glucose is thought to be the only energy yielding substrate of quantitative importance and the ultimate source of chemical energy for the brain 's. it appears reasonable to expect a correlation between severity and duration of hypoglycemia and cerebral levels of high-energy phosphates ( ~ P). Several in vivo studies, however, have failed to demonstrate any significant changes in cerebral ~ P during hypoglycemia10.~s. zT,ag, whereas others have reported decreased cerebral ~ P during hypoglycemiata,z4,4o. These conflicting results are probably due to differences in the experimental methods. Metabolic parameters measured in brain tissue have been variously correlated with onset of typical clinical signs such as seizures or coma, onset of typical EEG patterns, or with the degree of hypoglycemia, commonly expressed as blood glucose level. The time course of hypoglycemia has often been neglected. Most previous studies employed physiologically uncontrolled or only partially controlled animals. Several factors can influence brain metabolism in a physiologically uncontrolled animal during hypoglycemia in unpredictable ways. Hypo- and hypercapnia have been demonstrated to change the pattern of cerebral carbohydrate metabolism in normoglycemia~l,zL Both conditions can influence the availability of glucose to the brain through changes in CBF. Major motor convulsions in non-paralyzed, spontaneously breathing animals lower blood glucose, increase blood lactate, and superimpose asphyxia upon the hypoglycemic condition. Arterial hypotension in advanced hypoglycemia could result in cerebral ischemia, while the hypothermia which occurs in this condition might be protective by decreasing cerebral energy requirements. The method used to stop metabolism is important in using biologically labile compounds. The transcalvarial freezing technique has been found to preserve the metabolic stability of cortical tissue better than some other rapid freezing techniques when applied to small animals during maintenance of respiration and arterial blood pressure aa. The present experiments were designed to study the effects of hypoglycemia on cerebral energy metabolism and EEG under controlled conditions using the magnitude and duration of carbohydrate depletion in cortical tissue as the basic parameter. METHODS The experiments were performed on mate Wistar rats, weighing 250-350 g, fasted for 18-24 h. Anesthesia was induced with vinyl ether in a closed jar, and the animal was paralyzed by intraperitoneal administration of 5 mg/kg of tubocurarine chloride. Tracheostomy established respiration was maintained with a Harvard rodent respirator using 70% nitrous oxide and 30% oxygen. Tidal volume of the respirator and oxygen flow were adjusted to keep arterial CO2 pressure (PaCOD between 35 and 40 tort and arterial oxygen pressure (PaO2) above 100 torr. The femorat artery was cannulated to monitor arterial blood pressure continuously via a Statham pressure transducer and to sample blood anaerobically for measurement of blood gases and pH

265 on a Radiometer-Eschweiler combination unit. The readings were corrected for deviations of the body temperature from 37 °C. The femoral vein was cannulated for intravenous injection of insulin. The animal's scalp was reflected after midline incision, the periosteum removed, and three shallow depressions were drilled into the calvarium to secure a tripod containing two brass screw electrodes of a fronto-occipital bipolar E E G lead. EEG, ECG and mean arterial blood pressure (MAP) were continuously monitored on a oscilloscope. Recordings on a Gould Brush 440 recorder were taken every 15 min or whenever the pattern changed on the oscilloscope. The rectal temperature was kept between 36.5 and 37.5 °C with a heating lamp. After a steady state of at least 30 min on the respirator, 50 units of regular insulin (lletin, Lilly)/kg body weight were injected intravenously in a single dose of approximately 0.3-0.4 ml. At the end of an experiment the cisterna magna was punctured with the sharpened tip of a glass capillary through the previously exposed atlantooccipital membrane. Approximately 150 #1 arterial blood and 100/A cerebrospinal fluid (CSF) were allowed to drip directly into liquid nitrogen and a final blood sample was taken for blood gases and pH. The animal was then sacrificed by transcalvarial freezing of the brain in situ 33. The frozen brain was chiseled out under intermittent irrigation with liquid nitrogen using cooled instruments. The supratentorial hemispheres were taken as samples after scraping off the basal parts and were stored in a deep freezer at - 90 °C until they were analyzed. A fasting control group was treated in the same way as described above apart from the insulin injection and was sacrificed after 120 rain in a normocapnic steady state. In preliminary experiments in all of the 12 unanesthetized, spontaneously breathing animals, the chosen insulin dose provoked stupor within 90-100 rain, generalized convulsions within 100-140 min, and death within 110-180 rain. The onset of coma with loss of reactive movements and corneal reflexes could not be clearly defined due to repeated major motor convulsions. The observed seizure patterns indicated rostrocaudally progressing functional disturbance. Initial tonic clonic convulsions turned later into slower dystonic movements and finally into severe stretch spasms with marked opisthotonus. The rectal temperature declined to 34-36 °C within 135 min after insulin injections at a room temperature of 24 °C. Pilot studies showed that PaCO2 increased constantly and markedly during the first 45 rain of insulin hypoglycemia. The first experimental groups were therefore sacrificed 45 min after insulin injection. PaCOz was allowed to increase spontaneously from a normocapnic steady state in one group (45 min, normocapnic). Further experimental groups were sacrificed 90 and 135 min, respectively, after insulin injection. The 135 rain group was subdivided into animals with persistent E E G activity (135 min group with persistent EEG) and those with flat EEG recordings (135 min group with flat EEG). In an additional experimental group, PaO2 was lowered below 40 torr during the last I0 min of a 135 min observation period of hypoglycemia in an attempt to test the concept that normal brain levels of lactate during hypoglycemia exclude occur-

266 rence of hypoxia and that cerebral lactate levels can increase in severe hypoglycemia when complicated by ischemic hypoxia40. In these animals CO2 had to be added to the respired gas mixture to keep PaCO2 above 35 torr. 02 flow was decreased to an empirical value and was compensated for by nitrogen, Before and 2 rain after the induction of hypoxia, blood gases were checked and minor adjustments were made, MAP dropped below 100 mm Hg in all instances, but remained above 70 mm Hg in all reported experiments.

Analytical techniques Frozen brain, CSF and blood samples were quantitatively assayed for glucose (Glu), pyruvate (Py), lactate (La), adenosine monophosphate (AMP)--diphosphate (ADP)--triphosphate (ATP), and phosphocreatine (PCr) following the methods of Hohorst et al. 14. The initial tissue extraction was with 3 M perchloric acid a t - - 1 0 to -- 15 °C. Glycogen (Gly) was measured as glucose equivalent from the perchloric acid precipitate after hydrolysis with 1 N HCI for 2 h at 100 °C. Total carbon dioxide content of brain and CSF was assayed by crushing the frozen samples in liquid nitrogen in a COz-free atmosphere and analyzing an aliquot of the crushed tissue powder by the Conway microdiffusion technique with the modification of Pont6n and Siesjb 34. The ammonia content of the brain tissue was estimated by Conway's microdiffusion analysis7, using 0.002 N Na~COa as the standard instead of Ba(OH)z, The water content of brain tissue was determined as the difference between the wet and dry weight of a tissue sample of approximately 100 mg dried for 72 h at 105 °C, Alpha-ketoglutarate3, glutamate4, oxalacetate15 and aspartate az were assayed by enzymatic spectrophotometric techniques using minor modifications of the original techniques. All enzymes and co-enzymes for the assays were commercially supplied (Sigma Chemical Co., St. Louis, Mo.). All measurements were made in duplicate on a Zeiss PMQI spectrophotometer with an attached Sargent SRL linear-log recorder. Data for derived parameters were calculated as previously described2% Significant differences between mean values were tested by Student's t,test and P < 0.05 was considered to be significant. RESULTS The physiological parameters are given in Table I. The rectal temperature tended to increase during the first observation period of 45 rain, so that the heating lamp could be switched off intermittently, whereas during the third 45 min period the lamp had to be kept very close to the animal to maintain it's body temperature. The PaCO2 tended to increase spontaneously during the first hour after insulin injection. In the 45 min hypercapnic group, PaCOz was allowed to increase :from a normocapnic steady state and reached 46.5 ~ 3.5 torr compared to 37:5 =~: 1:0 torr in the 45 rain normocapnic group, tn the normocapnic experiment, PaCO2 was controlled by continuous increases in the tidal volume of the respirator. In later stages of hypoglycemia, particularly during the last 30 min of the 135 rain observation period, tidal volume had to be decreased rapidly to below that of the initial steady state to

37.0 115 115 37.8 7.388 22.06 --1.91

-k ± ± ± ± ± ±

0.4 7 11 1.3 0.023 0.99 0.89 37.1 112 110 46.5 7.284 21.10 --5.47

4± ± ± ± ± ±

0.2 10 15 1.0 0.017" 1.05" 1.35"

45 rain normocapnia (12)

±0.2 37.0 ± 7 110 ± 8 117 ± 3.5*,** 37.5 ± 0.030*.** 7.334 ±2.70 19.39 ± 1.69" --5.11

45 rain hypercapnia (10) 37.1 118 107 36.4 7.346 19.65 --4.48

± 0.2 i 11 ± 7 ± 1.3 ± 0.023* i 1.49" ± 1.78'

90 min normocapnia (10)

* Statistically significant at P < 0.05 compared with controls; ** compared with corresponding group.

Rectal temp. (°C) MAP (ram Hg) PaO2 (torr) PaCO2 (tort) Arterial pH HCO~(mEq./l) BE(mEq./l)

Fasting controls (10)

Number of animals in brackets. Mean values ± standard deviation are given.

37.1 138 113 37.1 7.340 19.88 --4.35

± 0.1 ± 15 ± 15 ± 1.3 ± 0.030* ± 1.75" ± 1.97"

135 min EEG present (12)

37.1 137 112 36.7 7.356 19.48 --4.79

± ± ± ~z ± ± ~

135 min flat EEG (12) 0.2 30 2 1.5 0.034* 1.62" 1.88"

Body temperature, mean arterial pressure ( M A P ) , arterial oxygen tension (Pa02), carbon dioxide tension (PaCOz) and arterial p H measured in the experimental groups and the derived standard bicarbonate and base excess (BE) values

TABLE I

tO

268

~ ~t~,~J~..~,~'#~..,~/~D,'-.-~v.~q

~ ,,.o,+w-~-+'~~ " ~ ' , - , , ~ '

~~4/~+'\'v'~'~,"---s~'~J~,,~,--v

, r

,'~-'-~.~,~,D,~+,<~,v-'~'+~,~

II~O

2.67

15 ndm

2.00

|Omin

0.097

0

1111

•

J

j~, /,

I•

°-

.

',?

mmia

/ ? j

"~'.

120Hie

.

.

.

135inin

0.N7 ,..._

135rain

1,15ada

/~ ~,'~L '~

• 7j '. •

, ,, ~l

r ,~.r, ~ "v ~ %'.,~, "~

"" * .....

~:

v~

!~ ,'.~,

,

I', I,/"

0.0N

"~ "

..J 'j ~/;

7,5 Ig|ll

10B Fig. 1. E E G changes during progressive glucose depletion of cortical brain tissue; at right, mean tissue concentrations of glucose, 75-90 min, after insulin injection, onset of paroxysmal activity. Bottom two traces, after 135 rain, various degrees of intermittent slow wave activity; in about 25 % flat EEG.

prevent hypocapnia. Those animals which became hypocapnic were discarded, The mean arterial pH of 7.388 in fasting controls was lower than the pooled control value of this laboratory for fed animals of 7,40429 • Standard bicarbonate (HCOs) and base excess (BE) in arterial blood decreased parallel to arterial pH significantly but not progressively in all normocapnic animals. Heart rate remained fairly constant; The EEG (Fig. 1) underwent no changes during the first 30 min alter insulin

7.19 0.114 0.713 6.25 2.67 3.45 0.097 1.24 12.8 5.38 0.184 2.08 11.3

i 1.27 q- 0.033 i 0.296 i 0.98 ± 0.59 _ 0.23 ± 0.012 ± 0.12 ± 2.0 ± 0.86 ± 0.023 ± 0.31 ± 1.8

2.01 0.225 1.57 6.97 0.225 2.59 0.086 0.922 10.7 1.79 0.172 1.97 11.5

-c 0.62* ± 0.033* ± 0.37* ± 1.13 _ 0.226* ± 0.54* ± 0.021 ± 0.20 ± 2.3 ± 0.72* ± 0.033 ± 0.37 ± 3.6

45 rain normocapnia (12)

± 0.48",** 2.57 ± 0.045* 0.237 5_ 0.41" 1.55 ~ 1.28 6.54 ± 0.125" 0.360 ± 0.35* 2.59 ± 0.015 0.097 ± 0.159",** 1.20 ± 1.0",** 12.4 ± 0.48* 1.87 ± 0.043 0.170 ± 0.30 2.18 ± 3.6 12.8

45 min hypercapnia (10) 1.33 0.155 0.991 6.39 0.097 1.91 0.069 0.687 10.0 0.962 0.164 2.23 13.6

± 0.40* ± 0.041 ± 0.366 ± 1.89 5_ 0.068* ~ 0.52* ± 0.013" ± 0.156" ± 2.9* ± 0.293* ± 0.026 ± 0.27 ± 2.6

90 min normocapnia (10)

* Statistically significant at P < 0.05 c o m p a r e d with controls ; ** c o m p a r e d with c o r r e s p o n d i n g group.

Blood glucose pyruvate lactate La/Py Brain glucose glycogen pyruvate lactate La/Py CSFglucose pyruvate lactate La/Py

Fasting controls (SO)

M e a n values ± s t a n d a r d deviation are given. N u m b e r o f a n i m a l s in brackets.

0.907 0.110 0.601 5.46 0.057 1.31 0.058 0.486 8.4 0.537 0.147 1.86 12.7

± ± ± ± ± ± ± ± ± ± ± ± ±

± 0.155" ± 0.050 $ 0.620 ± 1.93 ± 0.026* ± 0.109",** ± 0.052 5_ 0.140",** + 2.1",** ± 0.062",** ± 0.032 ± 0.55 ± 3.8

135 min flat EEG (12)

0.212",** 0.418 0.028 0.144 0.239 0.926 1.15 6.43 0.034* 0.038 0.30* 0.690 0.015" 0.065 0.156" 0.184 2.1" 2.8 0.154" 0.294 0.022 0.159 0.31 1.84 1.8 11.6

135 rain EEG present (12)

Concentrations q f ,elucose pyruvate, lactate and the derived lactate/pyruvate ratio (La/Py) in arterial blood, brain tissue and CSF, and glycogen concentration in brain tissue in mmole/kg o f wet brain tissue weight

T A B L E 11

~D

270

mUol/ Glucose

Kgwwt

]

~...........,....~...

lie e d CSF

. . . . 0 lS 30 n - 10

S

6

46

, ~ ~ |0

12

10

~rain ; min I3S 12

Fig. 2. Blood, CSF and brain glucose concentrations versus time after i.v. injection of 50 units insulin per kg; mean values from Table II; n, number of experiments.

injection, the period of most rapid decrease in blood and brain glucose. After about 45 min, the EEG slowed and increased in amplitude. At 90-105 min, paroxysmal activity appeared along with intermittent runs of slow, high voltage waves and more continuous atypical spike-wave and sharp wave activity as shown in the two bottom tracings of Fig. 1. During later stages of hypoglycemia slowing and increased amplitude developed progressively while longer isoelectric intervals occurred. In about 25%, the EEG became flat for 5-15 rain at the end of the 135 rain observation period. These animals are reported as a separate experimental group along with their significantly different metabolic date (135 rain group with flat EEG). Table II shows the glycolytic metabolites and the lactate/pyruvate ratios in blood, brain and CSF. The arterial glucose of 7.19 in moles/kg in the rats fasted 24 h

±

5.46 2.83 0.335 0.012 8.45 94.3 24.7 1.68 7.03

± 0.28 ± 0.07 -5_ 0.021 ± 0.011 ± 0.56 -- 0.3 £ 1.7 x 10 3 ± 0.04

Fasting controls (10) 5.34 2.84 0,331 0.013 8.58 94.3 17.3 1.33 7.03

± 0.14 -4- 0.17 ± 0.035 ± 0.011 ± 0.79 ± 0.6 ± 1.3" x 10 -3 5_ 0.03

45 min hypercapnia (10) 5.52 2.76 0.365 0.021 7.56 93.7 17.6 1.71 7.08

i 0.35 ± 0.16 ± 0.083 ± 0.028 ± 1.58 -+- 1.9 ± 1.8" x 10 .3 ± 0.04",**

45 rain normacapnia (12)

standard deviation. N u m b e r of animals in brackets.

5.27 2.72 0.350 0.032 7.77 93.3 15.0 1.09 7.09

~ 0.30 ± 0.11 -~ 0.059 ± 0.027 ± 1.52 _+ 1.5 i 1.7" x 10:3 i 0.03*

90 min normacapnia (10) 4.71 2.62 0.413 0.087 6.34 90.5 12.5 0.85 7.09

± 0.87* ± 0.38 ~+ 0.099* ± 0.143 ~ 1.49" ± 5.7 ~ 1.8" x 10 -3 ± 0.02*

135 rain EEG present (12)

6.95

1.22 0.965 0.701 0.492 1.36 60.4 4.6

± 0.51",** +_ 0.349",** ~_ 0.167',** ± 0.260*,** ± 0.34",** ± 10.8 ± 1.2",** -± 0.09",**

135 rain flat EEG (12)

* Statistically significant at P < 0.05 c o m p a r e d with controls, ** c o m p a r e d with corresponding group. Statistical values were not given to the calculated N A D H / N A D + ratio.

PCr ATP ADP AMP ATP/ADP ECh ERes NADH/NAD ~ pH(

Values are mean

Concentrations o f phosphocreatine, adenosine triphosphate (ATP), adenosine diphosphate ( ADP), adenosine monophosphate ( A M P ) in brain tissue in mmole/kg o f wet brain tissue weight, the derived A TP/A DP ratio, energy charge o f the adenylate pool (ECh) in per cent, energy reserve (ERes) in mmole/kg o f wet brain tissue weight, calculated O, toplasmic N A D H / N A D + ratio and the intracellular p H (pH'a)

TABLE II1

272 is about 3 mmoles/kg lower than the pooled control value for fed animals and fell rapidly and continuously to levels below 1 mmole/kg 22. Arterial lactate and pyruvate showed an intermittent increase at 45 min in the presence of a normal La/Py ratio and normal PaO2. The brain glucose content fell rapidly during insulin hypoglycemia almost to nothing. Glycogen decreased more slowly and less severely. The brain lactate level was lowered in excess of the pyruvate level after 90 rain and further after 135 min, resulting in a decreased La/Py ratio. Animals that developed a flat E E G at the end of the 135 rain observation period showed significantly lower glycogen and lactate levels than those with some slow wave EEG activity (135 min group with persistent EEG), although the glucose levels were not significantly different. The CSF glucose level followed the arterial pattern. Animals of the 135 rain group with flat E E G showed significantly lower values than those of the 135 rain group with persistent EEG. The intermittent increase in arterial lactate was not reflected in the CSF. The La/Py ratio did not change significantly. The time response curve of glucose levels in blood, brain and CSF (Fig. 2) reveals that CSF glucose followed the arterial level after an initial delay of about 45 min with the normal ratio of 0.75. The brain glucose concentration fell by over 85% during the first 45 min and was thereafter maintained at low levels, which decreased further to 1-2% of control values. The initial ratios of brain glucose to blood glucose of 0.38 and of brain glucose to CSF glucose of 0.5 fell during the first and second 45 rain observation periods by 50 ~o each and later leveled off. Table III shows ~ P and derived parameters in brain tissue. Lack of substrate was first reflected by significant decrease in the calculated energy reserve (ERes), which represents the early decrease of glucose and glycogen (45 min groups). After 135 min, when brain glucose was decreased by over 95%, glycogen by 60)~, ERes by 50 ~o and the E E G showed marked depression, a small but significant decrease in PCr occurred together with a shift from ATP to ADP as reflected in a decreased A T P ADP ratio but not yet reflected in a significant change of the ECh (135 min group with flat EEG). A highly significant decrease in ~ P was observed when brain glucose was decreased by over 98 ~ , glycogen and ERes by 80 ~ , and the EEG had become flat (135 min group with flat EEG). ATP/ADP ratio and ECh showed extensive changes. but cannot be compared with other experimental groups, since the adenylate pool. e.g. the sum of ATP, ADP. and AMP concentrations, decreased from 3.1 to 2.1 mmoles/kg as calculated from mean values. A further decline of about 50~o in glycogen contributed to the pronounced decrease in ERes in the 135 mm group with flat EEG apart from the substantial decrease in u p . The rapid and extensive fall in ~ P becomes more obvious when plotted against brain glucose levels (Fig. 3). The curve exhibits a threshold pattern. The calculated N A D H / N A D + ratio from intracellular lactate and pyruvate concentrations approached unity during the later stages of hypoglycemia. The values are distorted, however, in severe hypoglycemia since lactate, and to a lesser degree pyruvate, decrease severely in the brain. The calculation involves correction for 2"., blood and 1 5 ~ CSF and this assumption probably accounts for the fact that the

273

• o

: s

•

0o

~

IP •

ee

•

--

•

o

PCr ,//

AA

A

Wj ~

:s

flu

-

es

•.

-:

~w a'.~

ATP .ll

r

1.0

5.8

O6 O.4

iio .. °

ADP o

--¢

o.~

---

-

o

0.2 ,----

0,i

ss

e-

~

gO

.;-

,,

c>

,4 /

0.8

AMP

0.4 0.2

o

o

e6

0 25

015

0[75

t:

1.0

~

2]0

310

i

. mM011k.

Fig. 3. Cerebral tissue concentration of phosphocreatine, ATP, ADP and AMP versus decreasing glucose concentrations of cortical tissue. 0, fasting controls; ©, 45 rain normocapnic group; L 90 rnin group; L~, 135 rain group with persistent EEG activity; A, 135 rain group with flat EEG.

values for the 135 min group with flat EEG became negative and have been omitted from the table. The calculated intracellular pH increased in all normocapnic hypoglycemic groups to a similar degree except in the 135 rain group with flat EEG, which showed a decreased pH. In the 45 rain hypercapnic group, the pH remained normal. The arterial pH was shifted towards more acid values, the opposite of the calculated i,ltracellular pH of the brain. Table IV shows the Krebs cycle intermediates cc-ketoglutarate ((t-KG), oxalacetate (OAA), and the related amino acids glutamate (Glut) and aspartate (Asp), and also the water and ammonia content of the brain. Glut and ct-KG showed a tendency to decrease during advanced hypoglycemia. OAA and Asp remained normal. The ammonia content increased progressively. The water content remained essentially normal. Small but significant increases exhibited no trend and are of minor magnitude.

IV

..~..

+ 0.28 0.111 j, 0.023 77.4 I 0.7

1.10

0.136 &I 0.026 13.87 2 0.50 0,041 tir 0.013

0.098 :i 0.037****

Il.69 i 0.88* 0.041 f 0.012 1.39 & 0.31* 78.5 & 1.2* 1.21 i: 0.30 0.081 .c 0.025* 77.6 _F 0.9

0.141 ::: 0.025 12.1 I * 0.58* 0.038 & 0.012

45 min normocapnia iIZi

. ..-._--.

i $ i i

0.016 0.291 0.032* 1.2

1;. 1.67

values.

0.048 0.904 0.177 71.6

13.16

0.102 IL 0.031*

90 rnk normocapnia iroi

_... ..----..-...- _

0.905 r 0.364 0.207 I 0.062* 77.8 5 1.5

0.060 & 0.022* 12.81 -;- 1.93* 0.047 f 0.019

!12)

135 min EEG present

_..._ -.._. .~_ -.

~. .__.

0.904 2: 0.289 0.449 i 0.137*,** 78.5 f 1.1*

0.052 + 0.026* 6.56 ‘2 1.73*,** 0.040 i 0.025

135 min flat EEG iLj

.___.._

aspartate iA.sp) nnd ammonia Ih?Ha’) content in mmofelkg of‘ wet brain tissue weight

* Statistically significant at P _=0.05 compared with controls, ** compared with corresponding

.-.

Hz0

OAA ASP NH4

_. - ---... -.

U-KG

Glut

!lO)

(IO)

. . -_-..-_.. 4.5 min hypercapnia

Fasting controls

Mean values rt standard deviation. __... ._.____~_.__ ____..- . .--.-.

and water content of the brain inper cent

Alpha ketoglutarate ((I-KG), glutamate (Gluti. oxalacetate (OAA),

TABLE

275 TABLE V Hypoxemia (10 rain, Pa02 = 40 torr ) superimposed on hypoglycemia

Blood glucose pyruvate lactate La/Py Brain glucose glycogen pyruvate lactate La/Py CSF glucose pyruvate lactate La/Py

Fasting controls' ( lO)

135 min euroxemia (12) 135 min hypoxemia (8)

7.19 ± 1.27 0.114 ± 0.033 0.713 ± 0.296 6,25 ±0.98 2.67 ± 0.59 3.45 ± 0.23 0.097 ± 0.012 1.24 ± 0.21 12.78 ± 2.00 5.38 ± 0.86 0.184 ± 0.023 2.08 _~ 0.3l 11.30 ± 1.17

0.418 ± 0.155 0.144 ± 0.050 0.926 ± 0.620 6.43 ± 1.93 0.038 ± 0.026 0.690 ± 0.109 0.065 ~k 0.051 0.184 ± 0.140 2.83 -5_-2.13 0.249 ± 0.062 0.159 ± 0.062 1.84 % 0.55 11.57 :L3.83

0.431 ± 0.057*** 0.204 ± 0.050 2.975 :+ 0.871"** 14.64 ± 2.23*** 0.014 :k: 0.037* 0.607 L 0.052* 0.059 ± 0.037 0.442 ± 0.177" 8.77 ~:~4.60** 0.164 :£ 0.131" 0.156 L 0.032 2.81 :£ 0.42 18.40 :~- 3.45"**

* Statistically significant at P < 0.05 compared with controls, ** compared with 135 min flat EEG group. All changes were most p r o n o u n c e d in the 135 min group with flat EEG. Only the hypercapnic group already exhibited an increase in aspartate and a decrease in ~t-ketoglutarate 45 min after insulin injection. When hypoxia (PaO~ < 40 torr) was superimposed on hypoglycemia during the last 10 min o f a 135 min observation period, the E E G became isoelectric within 2-3 min. Lactate in blood and C S F increased significantly c o m p a r e d with fasting controls and with animals that developed a flat E E G after 135 rain of hypoglycemia, but brain lactate levels remained lower than in fasting controls and did not differ significantly from those o f the eupoxic 135 rain group with flat E E G (Table V). Brain glucose became unmeasurable and ~-~P decreased further, whereas brain glycogen remained unchanged. DISCUSSION The transient increase in arterial lactate in the presence o f a n o r m a l La/Py ratio and normal PaO2 is a k n o w n effect o f insulin-hypoglycemia resulting from neuronally mediated epinephrine release from the adrenal medulla which in turn stimulates muscle phosphorylase and thereby glycogen breakdown in skeletal muscle 29. Arterial p H decreased in all stages of hypoglycemia but not progressively. This can be explained during the initial stage by the increase in lactate. During later stages, however, other acid metabolites were probably present since lactate returned to normal. These metabolites may have been beta-hydroxybutyric and acetoacetic acid produced by enhanced lipolysis. Glucose deficiency o f the brain has been shown to cause sympathetic stimulation of adipose tissue resulting in release o f free fatty acids into the blood 5. During the first 15 min after insulin injection, brain glucose decreased more

276

mMol/k 1g 4

,t,

" "45

n: 10 6 6 12

"La

135

]0

rain

12(,10)

°La IPy'NADH~NAO *

1.2'

"12

1.0'

'10

0.8'

"8

0.6'

'8

0.4,

"4 a,

0.2'

,2

-1.0 xio "3

1.6 1.4

"1.2 1.0 0.8

Fig. 4. Top: glycogen and glucose; bottom: La/Py ratio, lactate and NADHINAD * quotient in cortical brain tissue after insulin injection. Dotted lines, 135 rain group with fiat EEG (,t); n. number of experiments.

slowly than blood glucose, but more rapidly in the following 30 min (Fig. 2). Since an initial delay in the decline of brain glucose would be expected during the time required for a significant deurease in blood glucose, it is rather remarkable how rapidly hypoglycemia was reflected in decreasing brain glucose levels. The initial delay in the decrease of brain glucose does not support the idea that insulin enhances cerebral glucose uptake. The CSF glucose level lags blood glucose by about 45 min (Fig. 2) as previously shown by Myers and Netsky ~°. The progressively decreasing brain/blood and brain CSF/glucose ratios indicate that the carbohydrate breakdown in brain tissue exceeded the rate of net transfer of glucose from the other two compartments. Glucose uptake by the brain is thought to be mediated by a carrier and to a lesser extent (4-25 ~ ) by diffusi on 8,11. Decreasing blood glucose levels result in desaturation of the carrier and declining diffusion rates. Therefore the net uptake of glucose decreases despite an increasing percentage of blood glucose being extracted by the carrier 8.

277 The glycogen level in cortical tissue declined more slowly and less extensively during hypoglycemia than did glucose (Fig. 4). This delayed pattern of cerebral glycogenolysis during hypoglycemia has previously been reported by others 13,17,27, but the reason for the delay is not yet clear. During the observation period of 135 min a minority of animals (135 min group with flat EEG) developed an isoelectric EEG together with decrease in ~ P and a significantly more advanced carbohydrate depletion than the rest (Table Ill). This difference might be due to random selection of animals with low fasting glucose levels. Another possibility is that there was a prolonged and more severe paroxysmal neuronal discharge in these animals. Electroshock has been found to increase the use of ~ P and thereby, of carbohydrates in the brains of paralyzed, well-oxygenated animals 6. The possibility therefore exists that the animals whose EEG's became flat at 135 rain maintained a higher cerebral metabolic rate due to sustained paroxysmal neuronal discharge and, therefore, reached a more advanced stage of carbohydrate depletion. The visual evaluation of the EEG recordings allows no quantitative estimation of the paroxysmal activity, the onset of which did not occur significantly earlier in these animals than in others. The decrease in PCr is associated with an acid shift in the calculated intracellular pH. A shift in the pH dependent phosphocreatine kinase (CPK) equilibrium could therefore have contributed to the decrease in PCr. The concomitant decrease in ATP, however, indicates that a true drop in ~ P and not only a shift in the equilibrium of the CPK reaction accounts for this finding. Together with the fall in PCr and ATP and the increase in ADP and AMP, the size of the adenylate pool decreases. The increase in AMP probably causes a shift in the reaction equilibria of those metabolic steps that catabolize AMP further. Breakdown of AMP in brain tissue proceeds via adenosine and inosine to hypoxanthine 19 involving dephosphorylation and deamination, which may partly contribute to the accumulation of ammonia and inorganic phosphate 41 observed during advanced hypoglycemia. Contrary to our expectation, the ECh neither correlated with the severity of dysfunction nor with the degree of carbohydrate depletion of brain tissue. Atkinson 2 has pointed out that tight coupling between generation and utilization of chemical energy is an essential feature of metabolism in all living systems. He elaborated that 'most substrates require some type of ATP consuming activation before they can be metabolized to yield ATP; hence, even a cell starving for energy should conserve its ATP supply (maintain a high energy charge) to allow for such activation, when substrate becomes available'. Following this concept, the normal ECh in hypoglycemia must be the result of metabolic adjustments that keep up ATP production and reduce ATP utilization. The normal ECh then could mean an effective metabolic regulation to maintain the balance between ATP production and utilization at a given level of brain function, or the level of brain function might be the result of such regulation. The decrease in glutamic acid and related amino acids and the concomitant increase in ammonia during hypoglycemia has been interpreted to indicate that free amino acids are used as a substrate for oxidative phosphorylation 12. Since deamination of free amino acids can only partially account for the accumulation of ammonia,

278 which occurs before the deaminated nucleotides increase, additional protein brea kdown is likely to occur during hypoglycemia1,20,4z. The fall in ~ P followed a threshold pattern, suggesting failure of tissue homeostatic processes at that point. Since free fatty acids have been found to inhibit oxidative phosphorylationX6,a6,aT,42, it might be speculated that lipid breakdown results in accumulation of endogenous inhibitors of oxidative phosporylation. Hinzen et al. lz have demonstrated that the total phospholipid content of the cerebral cortex decreases in a threshold fashion at critical blood glucose levels of 1.5-1.0 mmolesl kg. Cerebral lactate level, La/Py ratio, underwent no significant changes during the first 45 rain after insulin injection despite a rapid 80 ~o decrease of brain glucose. We assume an essentially normal pattern of carbohydrate stores during this phase. When glucose availability remained restricted, the normal metabolic pattern deteriorated rapidly. (Fig. 4) Ammonia increased and a - K G decreased suggesting that during this period (45-90 min after insulin injection) catabolism of endogenous substrate and decrease of metabolic rate came into effect. When cerebral carbohydrate stores approached total depletion, despite these mechanisms, ~ P levels declined concomitantly with accelerated glycogen breakdown and further deterioration of the cytoplasmic redox state. Our E E G recordings confirm the observation of Olsen and Klein zl and of Lewis et a124, that rapid decline of brain glucose levels occurs without remarkable E E G changes whereas prolonged hypoglycemia results in paroxysmal neuronal discharge and progressive depression until the EEG finally becomes isoelectric. Severe depression of the EEG with only intermittent delta waves was associated with normal or only slightly decreased ~-~P levels, while an isoelectric EEG of more than 5 min duration was always accompanied by markedly reduced ~ P levels. It is tempting, therefore, to correlate cortical electrical activity and ~-~P concentration, but several investigators have found a significant decrease in ~-~P levels only after E E G activity had become severely depressed or completely losO. It has been concluded therefore, that only a fraction of the total ~ P content of cerebral cortex might be readily available for the metabolic processes underlying the EEG. The data obtained in the 135 min hypoxic group confirm the tindings of Ljunggren et al. z5 that lactate increase due to hypoxia depends upon the carbohydrate stores of brain tissue and may be absent in advanced hypoglycemia; therefore normal or even decreased tissue lactate levels do not allow one to exclude hypoxic influences during hypoglycemia. In a previous communication from this laboratory ~3 the observation was reported that hypoglycemia abolishes the hypoxia induced increase in cortical blood flow. It has been previously concluded that lactate is the mediator of the blood flow response to hypoxia and that the increase m brain lactate is prevented by hypoglycemia. The present data are supportive of this conclusion.

279 ACKNOWLEDGEMENTS T h i s i n v e s t i g a t i o n was s u p p o r t e d

by the N a t i o n a l I n s t i t u t e o f N e u r o l o g i c a l

Diseases a n d S t r o k e R e s e a r c h G r a n t NS-05820-08 a n d the M e y e r G o l d R e s e a r c h Fund. W e are i n d e b t e d to Mrs. M. Santiso, Mrs. E. M a r t i n e z , M r . A. M o n t e i l , Mr. F. A l v a r e z , Mrs. O. A l o n s o , a n d Miss H. D i a z for their e x p e r t t e c h n i c a l assistance.

REFERENCES 1 Abood, L. G. and Geiger, A., Breakdown of proteins and lipids during glucose-free perfusion of the cat's brain, Amer. J. PhysioL, 182 (1955) 557-560. 2 Atkinson, D. E., Enzymes as control elements in metabolic regulation. In P. E. Boyer (Ed.), The Enzymes (Student Edition), Vol. 1, Academic Press, New York, 1970, pp. 461M-89. 3 Bergmeyer, H. U. and Bernt, E., ct-Oxog[utarate. In H. U. Bergmeyer (Ed.), Methods' of Enzymatic Analysis, Academic Press, New York, 1963, pp. 324-327. 4 Bernt, E. and Bergmeyer, H. U., L-Glutamate, determination with glutamic dehydrogenase. In H. U. Bergmeyer (Ed.), Methods' of Enzymatic Analysis', Academic Press, New York, 1963, pp. 384-388. 5 Cantu, R. C., Correll, J. W. and Manger, W. M., Reassessment of central neural pathways necessary for adrenal catecholamine output in response to hypoglycemia, Proc. Sac. exp. Biol. (N.Y.), 129 (1968) 155-16l. 6 Collins, R. C., Posner, J. B, and Plum, F., Cerebral energy metabolism during electroshock seizures in mice, Amer. J. Physiol., 218 (1970) 943-950. 7 Conway, E. J., Microdiffusion Analysis and Volumetric Erlvr, Crosby Lockwood, London, 1957, pp. 98-104. 8 Crone, C., Facilitated transfer of glucose from blood into brain tissue, J. Physiol. (Lond.), 181 (1965) 103-113. 9 Duffy, T. E., Nelson, S. R. and Lowry, O. H., Cerebral carbohydrate metabolism during acute hypoxia and recovery, J. Neurochem., 19 (1972) 959 977. 10 Ferendelli, J. A, and Chang, M. M., Brain metabolism during hypoglycemia effect of insulin on regional central nervous system glucose and energy reserve in mice, Areh. Neurol. (Chit'.), 28 (1973) 173-177. 11 Gilboe, D. D. and Betz, A. L., Kinetics of glucose transport in the isolated dog brain, Amer. J. PhysioL, 219 (1970) 774-778. 12 Himwich, W. A. and Agrawal, H. C., Amino acids. In A. Lajtha (Ed.), Handbook of Neurochemistry, Vol. l, Plenum Press, New York, 1969, pp. 33-52. 13 Hinzen, D. H. und Muller, U., Energiestooffwechsel und funktion des kaninchengehirns wahrend insulinhypoglykamie, Pflugers Arch. ges. Physiol., 322 (1971) 47-59. 14 Hohorst, H. J., Kruez, F. H. und Bucher, T., Uber metabolitgehalte und metabolit-konzentrationen in der leber der ratte, Biochem Z., 332 (1959) 18-46. 15 Hohorst, H. J. and Rein, M., Oxalacetate. In H. U. Bergmeyer (Ed.), Methods" of Enzymatic Analysis', Academic Press, New York, 1963, pp. 335-339. 16 Hulsman, W. C,, Elliott, W. B. and Slater, E. C., The nature and mechanism of action of uncoupling agents present in mitochrome preparations, Bioehem. biophys. Acta (Amst.), 39 (1970) 267-276. 17 Kerr, S. E. and Ghantus, M., The carbohydrate metabolism of brain. II. The effect of varying the carbohydrate and insulin supply on the glycogen, free sugar and lactic acid in mammalian brain, J. biol. Chem., 116 (1936) 9-19. 18 King, k. J., Lowry, O. H., Passonneau, J. V. and Venson, V., Effects of convulsants on energy reserves in the cerebral cortex, J. Neurochem., 14 (1967) 599-611. 19 Kleihues, P., Kobayashi, K. and Hossmann, K. A., Purine nucleotide metabolism in the cat brain after one hour of complete ischemia, J. Neurochem., 23 (1974) 417-425. 20 Knauff, H. G., Marx, D. und Mayer, G., Das verhalten der proteine und der serin- und colamin-

280

21 22 23 24 25 26 27

28 29 30 31 32

33 34 35 36 37 38 39 40 41 42

43

hattigen phosphatide des zentratnervesustems wahrend der insulinhypoglykamie, Hoppe-Seylers Z.physiol. Chem. 326 (1961) 227-234. Kogure, K., Busto, R., Matsumoto, A.. Scheinberg, P. and Reinmuth, O. M.. Effect of hyperventilation upon dynamics of cerebral energy metabolism, Amer. J. Physiol., 228 (1975) 1862-1867. Kogure, K.. Busto, R., Scheinberg, P. and Reinmuth. O. M., Dynamics of cerebral energy metabolism during moderate hypercapnia, J. Neurochem.. 24 (1975) 471-478. Kogure, K., Scheinberg, P., Reinmuth, O. M., Fujishima M. and Busto, R., Mechanisms of cerebral vasodilatation in hypoxia, J. appl. Physiol., 29 (1970) 223-229. Lewis, L. D., Ljunggren, B., Ratcheson. R. A. and Siesjo, B. K., Cerebral energy state in insulininduced hypoglycemia, related to blood glucose and EEG, J. Neurochem, 23 (1974) 673--679. Ljunggren, B.. Norberg, K. and Siesjo, B. K., Influence of tissue acidosis upon restitution of brain energy metabolism following total ischemia, Brain Research, 77 (1974) 173-186. Lowry, O. H.. Passonneau, J. V.. Hasselberg, F. X. and Schulz. D. W.. Effect of ischemia on known substrates and cofactors of the glycolytic pathway in brain, J. biol. Chem., 239 ~1964) 18-30. Mayman, C. J. and Tijerina, M. L., The effect of hypoglycemia on energy reserves in adult and newborn brain. In J. B. Brierly and B. S. Meldrum (Eds.), Brain Hypoxia, Heineman. London, 1971. pp. 242-249. Mcllwain, H. and Bachetard, H. S., Biochemistry o/'the Central Nervous System, 4th ed.. Churchill. London. 1971, Chap, 2, pp. 17-19. Mulrow, P. J.. The adrenals. In T. C. Ruch and H. D. Patton (Eds.), Physiology and Biophysics. Vol. 1II, 12th ed., Saunders, Philadelphia, Pa., 1973. pp. 224-247. Myers, G. G. and Netsky, M. G.. Relation of blood and cerebrospinal fluid glucose. Arch. Neurol. (Chic.), 6 (1962) 32-40. Olsen, N. S. and Klein, J. R., Effect of insulin hypoglycemia on brain glucose, glycogen, lactate and phosphates, Arch. Biochem., 13 (1947) 343-347. Pfleiderer, G.. L-Aspartic acid and L-Asparagine, determination with glutamate-oxalacetate transminase and malid dehydrogenase. In H. U. Bergmeyer (Ed.), Methods of Enzymatic Analysis. Academic Press. New York, 1963, pp. 381-383. Pontdn, U., Ratcheson, R. A., Salford, L. G. and Siesjo, B. K.. Optimal freezing conditions for cerebral metabolites in rats. J. Neurochem., 21 (1973) 1127-1138. Pontdn, U. and Siesjo, B. K.. A method for the determination of the total carbon dioxide content of frozen tissues, Acta physiol, scand., 60 (1964) 297-308. Rafaelsen, O. and Mellerup, E.. Insulin action. In A. Lajtha (Ed.), Handbook of Neurochemiso3,, Vol. IV., Plenum Press, New York, 1970, pp. 361-372. Sacks, W., Cerebral Metabolism in vivo. In A. Lajtha (Ed.), Handbook ofNeurochemistry, Vol. I. Plenum Press, New York, 1969, pp. 301-324. Sato, K.. Yamaguchi, M. and Mullan, S., Brain edema. A study of biochemical and structural alterations, Arch. Neurol. (Chic.). 21 (1969) 413-424. Sokoloff, L., The action of drugs on the cerebral circulation, Pharmacol. Rev., I 1 (1959) 1-85. Sokoloff, L., Metabolism of ketone bodies by the brain, Ann. Rev. Med., 24 (1973) 271--28 0. Tarr. M., Brada, D. and Samson, F. E.. Cerebral high energy phosphates during insuling hypoglycemia. Amer. J. Physiol., 203 (1962) 690-692. Tews, J. K., Carter. S. H and Stone, W. E., Chemical changes in the brain during insulin hypoglycemia and recovery, J. Neurochem., 12 (1965) 679-693. Wojtczak, L. and Wojtcsak, A. B., Uncouplic of oxidative phosphorylation and inhibition of ATPpi exchange by a substance from insect mitochondria. Biochem. biophys Acta (Amst.), 39 (1960) 277-286. Yoshino. Y. and Elliot K. A C.. Effects of various conditions o n the movement of carbon atoms derived from the glucose into and out of protein in rat brain. Canad. J. Biochem. 48 (1970) 236243.