Journal of Neiir.ochemi.rrr?. Raven Press, Ltd.. New York Q 1992 International Society for Neurochernistry
Effects of Hypoxia on Choline Exchange Among Organs *T§O. U. Scremin and $OD. J. Jenden *West L.,4. V.A.M.C. Research Service (Wadsworch Division); and Departinenis oftPhysiology and SPhurinacologj, and $Brain Research Inslitirte UCLA School of Medicine. Los Angeles, Califimiu. U.S.A.
Abstract: Cerebral blood flow (CBF) and the artenovenous (A-V) difference for choline (Ch) across brain, lung, splanchnic temtory, liver, kidney, and lower limb were studied in anesthetized, mechanically ventilated rats subjected to 10-20-min periods of hypoxia induced by lowering the inspired O2concentration to 13%. A large, time-dependent increase in arterial blood Ch concentration occurred during hypoxia. This phenomenon coincided with a net rate of uptake of Ch by the brain during hypoxia (0.81 k 0.24 nmol/min, n = 10; p < 0.05),which contrasted with a net rate of loss of Ch by this organ during the control period that preceded hypoxia (-0.20 k 0.08 nmol/min. n = 10;1.1 < 0.05). During hypoxia. lungs and splanchnic territory showed negative A-V differences for Ch levels (net Ch loss), whereas brain. liver. kidney, and lower limb showed posi-
tive A-V differences for Ch levels (net Ch uptake). Ch output from lungs was already detected at 5 min within the period of hypoxia and reversed rapidly after restoration of normal oxygenation. On the other hand. Ch output from the splanchnic territory became evident only 10 min after commencement of hypoxia and outlasted this experimental condition. It is concluded that extracerebral production of Ch during hypocapnic hypoxia raises the arterial concentration of this molecule and. by reversing the gradient across cerebral capillaries. prevents the cerebral loss of Ch in this condition. Key Words: Choline-Acetylcholine-Hypoxia -Cerebrovascular circulation. Scremin 0. U. and Jenden D. J. Effects of hypoxia on choline exchange among organs. J. NelirochPtn. 59, 906-9 14 ( 1992).
Choline (Ch) is a precursor for the synthesis of acetylcholine and phospholipids (Ansell, 1973; Zeisel, 1981 ; Jenden, 199 1 ). Brain tissue lacks the ability for de novo formation of Ch at a rate capable of supplying its own demand of this base and depends on extracerebral sources (Freeman and Jenden, 1976; Blusztajn and Wurtman, 1983; TuEek, 1984). The cerebral concentration of free Ch increases during periods of in vivo ischemia (Scremin and Jenden, 1989a-c, 1991; Beley et al., 199 1: Kumagae and Matsui, 1991) because all enzymatic processes that utilize this molecule require energy, whereas those that produce free Ch by degradation of phospholipids do not (Jenden, 1991). This same phenomenon underlies the wellknown accumulation of Ch by preparations of brain tissue deprived of energy substrates in vitro (Zeisel, 1985). During reperfusion after ischemia, the cerebral free Ch concentration drops gradually to levels below normal, a phenomenon that could lead to impairment of phospholipid or acetylcholine synthesis (Scremin and Jenden, 1991).
Another condition in which cerebral Ch could be lost is hypoxia. As in ischemia, hypoxia may restrict energy conversion in the brain, but it is accompanied by an elevation of cerebral blood flow (CBF), thus creating a situation of enhanced production of free Ch by degradation of Ch-containing compounds and increased removal of free Ch by the circulation that could lead to an even greater loss of cerebral Ch than that observed in ischemia. To test this hypothesis, the cerebral arteriovenous (A-V) difference for Ch was studied in rats subjected to a transient period of hypoxia. Because an initial series of experiments showed large changes in arterial blood Ch concentration in this condition, the A-V difference for Ch across several key organs was also studied to localize the sources responsible for this phenomenon.
MATERIALS AND METHODS Adult Sprague-Dawley male rats (weighing 300-350 g ) were used. Experiments were performed in a laboratory situ.4hbrcitrrtiron\-u w d ANOVA. analysis of variance: A-V. arteriovenous: CBF. cerebral blood flow: Ch. choline: nCBF, normal cerebra1 blood flow: RBC. red blood cell.
Received November 25, 199 I : revised manuscript received February 14. 1992: accepted FebruaQ 25, 1992. Address correspondence and reprint requests to Dr. 0. U. Scremin at V.A.M.C. (Wadsworth Division) Research Service (15 I). Building 1 15, Room 3 13. Wilshire and Sawtelle Boulevards. Los Angeles. CA 90073. U.S.A.
906
HYPOXIA AND CHOLINE ated in Albuquerque, NM, U.S.A., at an altitude of 1,600 m (average barometric pressure, 630 mm Hg). Animals were fasted for 16-18 h before experiments. Anesthesia was induced with halothane (2.5% in air) and maintained with 1.5% halothane in air during performance of surgical procedures. Indwelling catheters were inserted in the aorta and inferior vena cava in all animals. In addition, catheters were inserted in various regional veins as described below. Five groups of animals were used in which the A-V difference for Ch was studied through (a) brain, (b) lung, (c) splanchnic territory and liver, (d) kidney, and (e) hindlimb. The retroglenoid vein was used to obtain samples of cerebral blood and for measurement of CBF. This vessel was approached at its exit from the retroglenoid foramen. A skin incision was performed rostra1 to the ear lobe, and the posterior belly of the masseter muscle was mobilized anteriorly to expose the vein. A Silastic catheter (outer diameter, 0.93 mm) was inserted in this vessel and advanced in the cephalic direction beyond the retroglenoid foramen. A 6-0 silk ligature was tied around the vessel, above the level of entrance of extracranial branches. CBF was measured by timed collection and gravimetric determination of blood volume draining from this vessel under a free-flow condition. The retroglenoid vein outflow method has been extensively validated previously as an estimate ofCBF in the rat (Johansson and Nilsson, 1979; Hardebo and Nilsson, 1980; Nilsson and Siesjo. 1983). To measure the A-V difference of Ch levels across the lungs, a Silastic catheter was inserted in the right external jugular vein and advanced 25 mm beyond the right clavicle so that its tip would reach the right atrium to obtain blood representative of that in the pulmonary artery. Accurate placement of the catheter was confirmed by dissection postmortem. Blood in the aorta was taken as representative of that in the pulmonary veins. The renal vein was cannulated nonocclusively. A small incision was made in the ventral wall of this vessel through which a Silastic catheter (outside diameter, 0.93 mm) was passed. The vessel wall was closed with a purse-string suture (10-0 silk). Portal vein blood was obtained by cannulation of a small tributary vein. A catheter was advanced through this vessel until its tip could be seen emerging into, but not occluding, the portal vein. The portal and renal vein catheters were exteriorized, and the abdominal wall was closed by suturing muscle and skin separately with 3-0 silk. A suprahepatic vein was cannulated by advancing a Silastic catheter from the jugular vein toward the abdomen. At -50 mm from the level of the right clavicle. the catheter wedged into a suprahepatic vein. This was corroborated visually during the procedure by exposing the hepatic hilus through a subcostal laparotomy. For collection of hindlimb venous blood. the superficial circumflex vein was cannulated with a Silastic catheter directed caudally. The tip ofthis catheter was positioned inside the femoral vein, and care was taken not to impede blood flow. To estimate Ch production by blood under hypoxia, 3.5 ml of heparinized blood was equilibrated in microtonometers, at 37"C, with air or N, to which 5% C 0 2 was added for a period of 30 min. Samples were obtained at 2, 5, 10, 20, and 30 rnin after blood was placed in the microtonometers and immediately centrifuged to separate plasma from red blood cells (RBCs). and both fractions were extracted for Ch assay. The trachea was cannulated with a polyethylene tube (outside diameter, 2.5 mm). After all surgical procedures
90 7
were completed, mechanical ventilation was started with 70% N20/30% 0,. Muscle relaxation was achieved with a continuous intravenous infusion of pancuronium bromide, 0.93 mg/kg/h. The ventilation rate was adjusted to obtain P,co2 levels between 30 and 35 mm Hg initially and maintained at a constant rate throughout the experiment. A period of 60 min was allowed between discontinuation of halothane and the first blood sample for Ch content determination.
Measurement of blood Ch concentration Blood samples were obtained in heparinized tubes, and 50 pl was immediately withdrawn and added to tubes containing 2 ml of formic acid/acetone with a deuterated Ch standard for determination ofCh content by gas chromatography-mass spectrometry as described before (Jenden et al.. 1973). Three consecutive samples of arterial and regional venous blood, spaced at 3-min intervals, were initially obtained to establish baseline arterial and venous Ch levels. Then hypoxia was induced by replacing 0, with N, in the inspired gas mixture to a concentration of 13%. This was continuously monitored by a Beckman model OM1 I 0, analyzer. In the case of the brain, hypoxia was maintained for 10 min, with paired samples of retroglenoid and aortic blood obtained at 5 and 10 rnin within the hypoxic period in the first series of four animals. Six additional animals were studied in which hypoxia was sustained for 20 rnin with an additional sample at the end of that period. In all other cases, hypoxia was maintained for 10 min. After the hypoxic bout, the initial 0, concentration (30%)was restored, and paired samples of arterial and regional venous blood were obtained at I , 3. 5, 10, 30, and 60 rnin after termination of hypoxia. Arterial blood pressure was continuously monitored from the aorta with a Hewlett-Packard strain-gauge transducer and polygraph. Samples for measurements of arterial blood gas concentration and pH were obtained during the control period, 5 rnin into the hypoxic episode. and 5 and 60 rnin after hypoxia was terminated. For this purpose, blood was collected into heparinized glass capillary tubes (70 pl) and introduced in a Radiometer model ABL-30 acid-base analyzer.
Data analysis Means and SE values of variables were calculated for every experimental period. The significance of differences was assessed by Student's i test or by analysis of variance (ANOVA) and a multiple-comparisons procedure (ScheE test) where appropriate, as indicated in Results. A mean net exchange rate of Ch (nanomoles per minute) was calculated separately for control. hypoxia, and recovery periods by integrating the product of Ch output [Ch A-V difference (nanomoles per milliliter) CBF (milliliters per minute)] over time and dividing by the interval between the first and last samples of every period. A regression of venous Ch on arterial Ch concentration was calculated for each organ studied. It should be noted that ifCh, = Ch, b a res (Eq. I ) , then Ch, - Ch, = Cham ( 1 - h) - a - res (Eq. 2). where Ch, represents venous Ch concentration, Ch, represents arterial Ch concentration. and h. u, and res are the slope. intercept. and residual. respectively, of the regression equation represented by Eq. I . It is obviously equivalent to test the slope of Eq. 1 for significance against 1 or the slope of Eq. 2 for significance against 0.
-
+ +
J. Neurochem.. V d 59. No. 3, 19Y2
0. U. SCREMIN AND D. J. JENDEN
908
TABLE 1. Blood gmes arid p H under variozis esprririiental conciirions
Control Hypoxia ( 5 min) Recovery ( 5 min) Recovery (60 min)
= =
104.2 f 3.4
34.8 f 0.9
7.401 f 0.009
28
29.6 f 0.9"
26.5 f 0.9"
7.386 f 0.014
24
96.1 f 4.3
40.3 f 1.8*
7.146 f 0.028"
19
92.5 -+ 3.9
33.1 2 1.3
7.329 f 0.029
21
CBF and cerebral A-V Ch difference Under control conditions, retroglenoid vein outflow averaged 0.23 & 0.0 19 ml/min. Hypoxia induced a significant increase in CBF [represented in Fig. 1 as the ratio of the observed value over the first control determination (nCBF)] in spite of the blood pressure drop. Calculated cerebrovascular resistance (mm Hg/ nCBF) decreased from 123.2 -t 4.8 (control) to 52.36 k 5.9 (5-min hypoxia), 49.4 t 6.9 (10-min hypoxia), and 58.2 f 10.7 (20-min hypoxia). The phenomenon reversed rapidly after restoration of normal P,o, levels (Fig. 1). The analysis of pooled data from the 10- and 20rnin hypoxia series revealed that during the initial control period of normoxia, the A-V difference for Ch levels (A-V Ch) across brain was negative (-1.03 f 0.44 nmol/ml, n = 29) and significantly different from 0 ( p < 0.05). The net exchange rate of Ch showed a value of -0.20 f 0.087 nmol/min (n = 10) during this period ( p < 0.05). Thus, during the normoxic period that preceded the hypoxic episode, the brain showed a net loss of Ch. A large increase in arterial blood Ch concentration was observed during hypoxia (Fig. 2). This change increased over time while hypoxia was maintained and did not show evidence of reaching stable levels even after 20 rnin of hypoxia. The A-V difference for Ch levels showed a trend toward an increase during hypoxia (Fig. 2), which, together with the increased CBF, led to a net exchange rate of Ch of 0.80 k 0.239 nmol/min (n = 10, p < 0.05) during the period of hypoxia. Cerebral and venous Ch levels decreased in parallel during the recovery period (Fig. 2). The net exchange rate of Ch during the recovery period was -0.33 f 0.26 nmol/min (n = 10). Regression analysis of pooled data from control,
~
Data are mean f SEM values. The statistical significance of differences against the control value was tested by ANOVA with multiple comparisons (SchefE test): " p < 0.0 I . p < 0.05.
'
RESULTS Changes in arterial blood gases and pH in the three experimental periods are shown in Table 1 for pooled data from all experiments. Significant hypoxia and hypocapnia were observed during ventilation with the 13% O2 gas mixture, followed by rapid recovery of P,02, a rebound of Paco2,and a decrease in blood pH after cessation of hypoxia. The acidosis subsided 60 min after restoration of normoxia. The arterial blood pressure (mean f SE) showed a significant drop during hypoxia: control, 134.1 t 2.3 mm Hg, n = 76 (26) [number of samples (number of rats)]; 5-min hypoxia, 77.4 t 5.6 mm Hg, n = 26 (26); and 10-min hypoxia, 77.0 f 5.5 mm Hg, n = 26 (26). Control levels were restored soon after hypoxia was terminated and remained stable thereafter: 3-min recovery, 1 10.1 f 4.1, n = 26 (26); 5-min recovery, 116.2 t 3.7, n = 26 (26); 10-min recovery, 120.2 & 4.0, n = 26 (26); 30-min recovery, 123.5 f 4.9, n HYPOXIA
2.5
-
26 (26); and 60-min recovery, 115.5 f 4.5, n 26 (26).
150
2.0
120
Y
I
II
-
2
D 1.5
80
y
1.0
80
3 3
c
c 0
-0
I
LL
e.
m 0
=
0.5
0
nCEF MABP
30
0
0 0
10
20
30
40
50
60
TIME (rninl
J . Neurochem.. Val. 59. No. 3. I992
70
80
80
100
FIG. 1. CBF was measured volumetrically from the retroglenoid vein and expressed as the ratio of the observed value to the first control determination (nCBF; 0). In spite of a decrease in mean arterial blood pressure (MABP; O), a large increase in nCBF is observed during hypoxia. The statistical significance of changes was assessed by ANOVA and Scheffe contrasts among pooled data of five conditions (control, 5, 10, and 20 min of hypoxia, and recovery). For nCBF, F4,60 = 5.86, p < 0.0005; significance against the control of the second and third hypoxia periods, p < 0.05. The recovery period was not significantly different from the control. For MABP, F4,60= 5.46, p 4 0.0008; significance against the control of the second and third hypoxia periods, p < 0.05. The recovery period was not significantly different from the control.
HYPOXIA AND CHOLINE BRAIN HYPOXIA
O A-V A cer
dlff
I
10
0
- 10 0
10
20
30
40
50
60
70
80
90
100
Time ( m i d FIG. 2. Cerebral venous blood was obtained from the retroglenoid vein and arterial blood from the aorta. During hypoxia, there was a steady increase in Ch concentration in arterial blood (0) and cerebral venous blood (0).The A-V difference for Ch levels across brain (A) was significantlyless than 0 in the second sample of the control period and the last sample of the recovery period (p i0.05).The statistical significance of changes was assessed by ANOVA and Scheffe contrasts among pooled data of five conditions (control, 5.10, and 20 min of hypoxia. and recovery). For Ch concentration in aortic blood, F4.60= 14.90, p < 0.0001; the second and third hypoxia periods and recovery period were different from the control (p < 0.01). For Ch concentration in retroglenoid vein blood, F,,, = 18.82, p < 0.0001;the second and third hypoxia periods and recovery period were different from the control (p < 0.01).
hypoxic, and recovery periods of 10- and 20-min series showed that the slope relating cerebral A-V Ch to arterial Ch was positive and significantly different from 0 (Fig. 3), supporting the concept that increased arterial Ch concentration is associated with a net gain of Ch by the brain.
Examination of regional A-V differences for Ch levels revealed that lungs released significant amounts of Ch into the circulation during hypoxia, starting early after commencement of hypoxia and reversing rapidly after its termination (Fig. 5). The splanchnic territory also showed an increase of its output of Ch over control levels. This phenomenon was slower in onset than the one observed in lungs, but it lasted beyond termination of hypoxia (Fig. 6).
Extracerebral sinks for Ch The liver showed a significant increase in uptake of Ch after 10 min of hypoxia when compared with the control period ( p < 0.01) but not after 5 min of hypoxia (Fig. 6). However, this phenomenon did not offset the increase in splanchnic output completely because the Ch concentration in suprahepatic vein blood increased significantly over control values after 10 min of hypoxia ( p < 0.05; Fig. 6). Kidney and lower limb also showed significant uptake of Ch during the hypoxic period (Fig. 7). DISCUSSION The present experiments were designed to test the hypothesis that hypoxia would enhance the cerebral loss of Ch through the circulation. On the contrary, the experiments revealed that the negative A-V difference for Ch levels across brain disappeared during hypoxia, a phenomenon most likely related to the dramatic surge in arterial Ch concentration in this condition. This is suggested by the significant dependence
160
1
1.1 Car,ran.Cn
120
Extracerebral sources of Ch The blood itself was initially examined as a source of Ch in hypoxia. Two experiments were performed in which blood extracted from normoxic rats was equilibrated in vitro with air or nitrogen as described in Materials and Methods. Po, was initially 73.6 and 64 mm Hg in experiments 1 and 2, respectively,and it increased in the blood of the microtonometers that contained air to final values of 93.8 (experiment 1) and 94.8 (experiment 2) mm Hg, but dropped continuously in those containing N,, to final values of 28.2 (experiment 1) and 25.5 (experiment 2) mm Hg at the end of 30 min. A steady increase in plasma Ch as a function of time could be demonstrated, with air- and nitrogen-equilibrated samples being represented by the same regression line (Fig. 4). No such trend occurred for RBC Ch. This phenomenon was not affected by the partial pressure of O2because no apparent differences were observed in plasma or RBC Ch for blood exposed to air or nitrogen.
909
80
Y - 0.82 X
+ 4.03
r- 0.97
< 0.0001
P
./
,
-/ IoICer. A - v cn Y - 0.08 X - 4.03 P- 0.002
r- 0.31
40
-
a
0
-
'
.
..'..
e
*D
B
~
2
0
-40
',So
o
0
Mq..0
I
I
I
I
40
80
120
160
200
J. Neirrachem.. Vol. 59. No. 3, 1992
0. U. SCREMIN AND D. J. JENDEN
910
Plasma
RBC
20.0,
-
20.0
Y = 0.16 X t 7.37
16.0
r
=
=
0.88, P
Effects of Hypoxia on Choline Exchange Among Organs *T§O. U. Scremin and $OD. J. Jenden *West L.,4. V.A.M.C. Research Service (Wadsworch Division); and Departinenis oftPhysiology and SPhurinacologj, and $Brain Research Inslitirte UCLA School of Medicine. Los Angeles, Califimiu. U.S.A.
Abstract: Cerebral blood flow (CBF) and the artenovenous (A-V) difference for choline (Ch) across brain, lung, splanchnic temtory, liver, kidney, and lower limb were studied in anesthetized, mechanically ventilated rats subjected to 10-20-min periods of hypoxia induced by lowering the inspired O2concentration to 13%. A large, time-dependent increase in arterial blood Ch concentration occurred during hypoxia. This phenomenon coincided with a net rate of uptake of Ch by the brain during hypoxia (0.81 k 0.24 nmol/min, n = 10; p < 0.05),which contrasted with a net rate of loss of Ch by this organ during the control period that preceded hypoxia (-0.20 k 0.08 nmol/min. n = 10;1.1 < 0.05). During hypoxia. lungs and splanchnic territory showed negative A-V differences for Ch levels (net Ch loss), whereas brain. liver. kidney, and lower limb showed posi-
tive A-V differences for Ch levels (net Ch uptake). Ch output from lungs was already detected at 5 min within the period of hypoxia and reversed rapidly after restoration of normal oxygenation. On the other hand. Ch output from the splanchnic territory became evident only 10 min after commencement of hypoxia and outlasted this experimental condition. It is concluded that extracerebral production of Ch during hypocapnic hypoxia raises the arterial concentration of this molecule and. by reversing the gradient across cerebral capillaries. prevents the cerebral loss of Ch in this condition. Key Words: Choline-Acetylcholine-Hypoxia -Cerebrovascular circulation. Scremin 0. U. and Jenden D. J. Effects of hypoxia on choline exchange among organs. J. NelirochPtn. 59, 906-9 14 ( 1992).
Choline (Ch) is a precursor for the synthesis of acetylcholine and phospholipids (Ansell, 1973; Zeisel, 1981 ; Jenden, 199 1 ). Brain tissue lacks the ability for de novo formation of Ch at a rate capable of supplying its own demand of this base and depends on extracerebral sources (Freeman and Jenden, 1976; Blusztajn and Wurtman, 1983; TuEek, 1984). The cerebral concentration of free Ch increases during periods of in vivo ischemia (Scremin and Jenden, 1989a-c, 1991; Beley et al., 199 1: Kumagae and Matsui, 1991) because all enzymatic processes that utilize this molecule require energy, whereas those that produce free Ch by degradation of phospholipids do not (Jenden, 1991). This same phenomenon underlies the wellknown accumulation of Ch by preparations of brain tissue deprived of energy substrates in vitro (Zeisel, 1985). During reperfusion after ischemia, the cerebral free Ch concentration drops gradually to levels below normal, a phenomenon that could lead to impairment of phospholipid or acetylcholine synthesis (Scremin and Jenden, 1991).
Another condition in which cerebral Ch could be lost is hypoxia. As in ischemia, hypoxia may restrict energy conversion in the brain, but it is accompanied by an elevation of cerebral blood flow (CBF), thus creating a situation of enhanced production of free Ch by degradation of Ch-containing compounds and increased removal of free Ch by the circulation that could lead to an even greater loss of cerebral Ch than that observed in ischemia. To test this hypothesis, the cerebral arteriovenous (A-V) difference for Ch was studied in rats subjected to a transient period of hypoxia. Because an initial series of experiments showed large changes in arterial blood Ch concentration in this condition, the A-V difference for Ch across several key organs was also studied to localize the sources responsible for this phenomenon.
MATERIALS AND METHODS Adult Sprague-Dawley male rats (weighing 300-350 g ) were used. Experiments were performed in a laboratory situ.4hbrcitrrtiron\-u w d ANOVA. analysis of variance: A-V. arteriovenous: CBF. cerebral blood flow: Ch. choline: nCBF, normal cerebra1 blood flow: RBC. red blood cell.
Received November 25, 199 I : revised manuscript received February 14. 1992: accepted FebruaQ 25, 1992. Address correspondence and reprint requests to Dr. 0. U. Scremin at V.A.M.C. (Wadsworth Division) Research Service (15 I). Building 1 15, Room 3 13. Wilshire and Sawtelle Boulevards. Los Angeles. CA 90073. U.S.A.
906
HYPOXIA AND CHOLINE ated in Albuquerque, NM, U.S.A., at an altitude of 1,600 m (average barometric pressure, 630 mm Hg). Animals were fasted for 16-18 h before experiments. Anesthesia was induced with halothane (2.5% in air) and maintained with 1.5% halothane in air during performance of surgical procedures. Indwelling catheters were inserted in the aorta and inferior vena cava in all animals. In addition, catheters were inserted in various regional veins as described below. Five groups of animals were used in which the A-V difference for Ch was studied through (a) brain, (b) lung, (c) splanchnic territory and liver, (d) kidney, and (e) hindlimb. The retroglenoid vein was used to obtain samples of cerebral blood and for measurement of CBF. This vessel was approached at its exit from the retroglenoid foramen. A skin incision was performed rostra1 to the ear lobe, and the posterior belly of the masseter muscle was mobilized anteriorly to expose the vein. A Silastic catheter (outer diameter, 0.93 mm) was inserted in this vessel and advanced in the cephalic direction beyond the retroglenoid foramen. A 6-0 silk ligature was tied around the vessel, above the level of entrance of extracranial branches. CBF was measured by timed collection and gravimetric determination of blood volume draining from this vessel under a free-flow condition. The retroglenoid vein outflow method has been extensively validated previously as an estimate ofCBF in the rat (Johansson and Nilsson, 1979; Hardebo and Nilsson, 1980; Nilsson and Siesjo. 1983). To measure the A-V difference of Ch levels across the lungs, a Silastic catheter was inserted in the right external jugular vein and advanced 25 mm beyond the right clavicle so that its tip would reach the right atrium to obtain blood representative of that in the pulmonary artery. Accurate placement of the catheter was confirmed by dissection postmortem. Blood in the aorta was taken as representative of that in the pulmonary veins. The renal vein was cannulated nonocclusively. A small incision was made in the ventral wall of this vessel through which a Silastic catheter (outside diameter, 0.93 mm) was passed. The vessel wall was closed with a purse-string suture (10-0 silk). Portal vein blood was obtained by cannulation of a small tributary vein. A catheter was advanced through this vessel until its tip could be seen emerging into, but not occluding, the portal vein. The portal and renal vein catheters were exteriorized, and the abdominal wall was closed by suturing muscle and skin separately with 3-0 silk. A suprahepatic vein was cannulated by advancing a Silastic catheter from the jugular vein toward the abdomen. At -50 mm from the level of the right clavicle. the catheter wedged into a suprahepatic vein. This was corroborated visually during the procedure by exposing the hepatic hilus through a subcostal laparotomy. For collection of hindlimb venous blood. the superficial circumflex vein was cannulated with a Silastic catheter directed caudally. The tip ofthis catheter was positioned inside the femoral vein, and care was taken not to impede blood flow. To estimate Ch production by blood under hypoxia, 3.5 ml of heparinized blood was equilibrated in microtonometers, at 37"C, with air or N, to which 5% C 0 2 was added for a period of 30 min. Samples were obtained at 2, 5, 10, 20, and 30 rnin after blood was placed in the microtonometers and immediately centrifuged to separate plasma from red blood cells (RBCs). and both fractions were extracted for Ch assay. The trachea was cannulated with a polyethylene tube (outside diameter, 2.5 mm). After all surgical procedures
90 7
were completed, mechanical ventilation was started with 70% N20/30% 0,. Muscle relaxation was achieved with a continuous intravenous infusion of pancuronium bromide, 0.93 mg/kg/h. The ventilation rate was adjusted to obtain P,co2 levels between 30 and 35 mm Hg initially and maintained at a constant rate throughout the experiment. A period of 60 min was allowed between discontinuation of halothane and the first blood sample for Ch content determination.
Measurement of blood Ch concentration Blood samples were obtained in heparinized tubes, and 50 pl was immediately withdrawn and added to tubes containing 2 ml of formic acid/acetone with a deuterated Ch standard for determination ofCh content by gas chromatography-mass spectrometry as described before (Jenden et al.. 1973). Three consecutive samples of arterial and regional venous blood, spaced at 3-min intervals, were initially obtained to establish baseline arterial and venous Ch levels. Then hypoxia was induced by replacing 0, with N, in the inspired gas mixture to a concentration of 13%. This was continuously monitored by a Beckman model OM1 I 0, analyzer. In the case of the brain, hypoxia was maintained for 10 min, with paired samples of retroglenoid and aortic blood obtained at 5 and 10 rnin within the hypoxic period in the first series of four animals. Six additional animals were studied in which hypoxia was sustained for 20 rnin with an additional sample at the end of that period. In all other cases, hypoxia was maintained for 10 min. After the hypoxic bout, the initial 0, concentration (30%)was restored, and paired samples of arterial and regional venous blood were obtained at I , 3. 5, 10, 30, and 60 rnin after termination of hypoxia. Arterial blood pressure was continuously monitored from the aorta with a Hewlett-Packard strain-gauge transducer and polygraph. Samples for measurements of arterial blood gas concentration and pH were obtained during the control period, 5 rnin into the hypoxic episode. and 5 and 60 rnin after hypoxia was terminated. For this purpose, blood was collected into heparinized glass capillary tubes (70 pl) and introduced in a Radiometer model ABL-30 acid-base analyzer.
Data analysis Means and SE values of variables were calculated for every experimental period. The significance of differences was assessed by Student's i test or by analysis of variance (ANOVA) and a multiple-comparisons procedure (ScheE test) where appropriate, as indicated in Results. A mean net exchange rate of Ch (nanomoles per minute) was calculated separately for control. hypoxia, and recovery periods by integrating the product of Ch output [Ch A-V difference (nanomoles per milliliter) CBF (milliliters per minute)] over time and dividing by the interval between the first and last samples of every period. A regression of venous Ch on arterial Ch concentration was calculated for each organ studied. It should be noted that ifCh, = Ch, b a res (Eq. I ) , then Ch, - Ch, = Cham ( 1 - h) - a - res (Eq. 2). where Ch, represents venous Ch concentration, Ch, represents arterial Ch concentration. and h. u, and res are the slope. intercept. and residual. respectively, of the regression equation represented by Eq. I . It is obviously equivalent to test the slope of Eq. 1 for significance against 1 or the slope of Eq. 2 for significance against 0.
-
+ +
J. Neurochem.. V d 59. No. 3, 19Y2
0. U. SCREMIN AND D. J. JENDEN
908
TABLE 1. Blood gmes arid p H under variozis esprririiental conciirions
Control Hypoxia ( 5 min) Recovery ( 5 min) Recovery (60 min)
= =
104.2 f 3.4
34.8 f 0.9
7.401 f 0.009
28
29.6 f 0.9"
26.5 f 0.9"
7.386 f 0.014
24
96.1 f 4.3
40.3 f 1.8*
7.146 f 0.028"
19
92.5 -+ 3.9
33.1 2 1.3
7.329 f 0.029
21
CBF and cerebral A-V Ch difference Under control conditions, retroglenoid vein outflow averaged 0.23 & 0.0 19 ml/min. Hypoxia induced a significant increase in CBF [represented in Fig. 1 as the ratio of the observed value over the first control determination (nCBF)] in spite of the blood pressure drop. Calculated cerebrovascular resistance (mm Hg/ nCBF) decreased from 123.2 -t 4.8 (control) to 52.36 k 5.9 (5-min hypoxia), 49.4 t 6.9 (10-min hypoxia), and 58.2 f 10.7 (20-min hypoxia). The phenomenon reversed rapidly after restoration of normal P,o, levels (Fig. 1). The analysis of pooled data from the 10- and 20rnin hypoxia series revealed that during the initial control period of normoxia, the A-V difference for Ch levels (A-V Ch) across brain was negative (-1.03 f 0.44 nmol/ml, n = 29) and significantly different from 0 ( p < 0.05). The net exchange rate of Ch showed a value of -0.20 f 0.087 nmol/min (n = 10) during this period ( p < 0.05). Thus, during the normoxic period that preceded the hypoxic episode, the brain showed a net loss of Ch. A large increase in arterial blood Ch concentration was observed during hypoxia (Fig. 2). This change increased over time while hypoxia was maintained and did not show evidence of reaching stable levels even after 20 rnin of hypoxia. The A-V difference for Ch levels showed a trend toward an increase during hypoxia (Fig. 2), which, together with the increased CBF, led to a net exchange rate of Ch of 0.80 k 0.239 nmol/min (n = 10, p < 0.05) during the period of hypoxia. Cerebral and venous Ch levels decreased in parallel during the recovery period (Fig. 2). The net exchange rate of Ch during the recovery period was -0.33 f 0.26 nmol/min (n = 10). Regression analysis of pooled data from control,
~
Data are mean f SEM values. The statistical significance of differences against the control value was tested by ANOVA with multiple comparisons (SchefE test): " p < 0.0 I . p < 0.05.
'
RESULTS Changes in arterial blood gases and pH in the three experimental periods are shown in Table 1 for pooled data from all experiments. Significant hypoxia and hypocapnia were observed during ventilation with the 13% O2 gas mixture, followed by rapid recovery of P,02, a rebound of Paco2,and a decrease in blood pH after cessation of hypoxia. The acidosis subsided 60 min after restoration of normoxia. The arterial blood pressure (mean f SE) showed a significant drop during hypoxia: control, 134.1 t 2.3 mm Hg, n = 76 (26) [number of samples (number of rats)]; 5-min hypoxia, 77.4 t 5.6 mm Hg, n = 26 (26); and 10-min hypoxia, 77.0 f 5.5 mm Hg, n = 26 (26). Control levels were restored soon after hypoxia was terminated and remained stable thereafter: 3-min recovery, 1 10.1 f 4.1, n = 26 (26); 5-min recovery, 116.2 t 3.7, n = 26 (26); 10-min recovery, 120.2 & 4.0, n = 26 (26); 30-min recovery, 123.5 f 4.9, n HYPOXIA
2.5
-
26 (26); and 60-min recovery, 115.5 f 4.5, n 26 (26).
150
2.0
120
Y
I
II
-
2
D 1.5
80
y
1.0
80
3 3
c
c 0
-0
I
LL
e.
m 0
=
0.5
0
nCEF MABP
30
0
0 0
10
20
30
40
50
60
TIME (rninl
J . Neurochem.. Val. 59. No. 3. I992
70
80
80
100
FIG. 1. CBF was measured volumetrically from the retroglenoid vein and expressed as the ratio of the observed value to the first control determination (nCBF; 0). In spite of a decrease in mean arterial blood pressure (MABP; O), a large increase in nCBF is observed during hypoxia. The statistical significance of changes was assessed by ANOVA and Scheffe contrasts among pooled data of five conditions (control, 5, 10, and 20 min of hypoxia, and recovery). For nCBF, F4,60 = 5.86, p < 0.0005; significance against the control of the second and third hypoxia periods, p < 0.05. The recovery period was not significantly different from the control. For MABP, F4,60= 5.46, p 4 0.0008; significance against the control of the second and third hypoxia periods, p < 0.05. The recovery period was not significantly different from the control.
HYPOXIA AND CHOLINE BRAIN HYPOXIA
O A-V A cer
dlff
I
10
0
- 10 0
10
20
30
40
50
60
70
80
90
100
Time ( m i d FIG. 2. Cerebral venous blood was obtained from the retroglenoid vein and arterial blood from the aorta. During hypoxia, there was a steady increase in Ch concentration in arterial blood (0) and cerebral venous blood (0).The A-V difference for Ch levels across brain (A) was significantlyless than 0 in the second sample of the control period and the last sample of the recovery period (p i0.05).The statistical significance of changes was assessed by ANOVA and Scheffe contrasts among pooled data of five conditions (control, 5.10, and 20 min of hypoxia. and recovery). For Ch concentration in aortic blood, F4.60= 14.90, p < 0.0001; the second and third hypoxia periods and recovery period were different from the control (p < 0.01). For Ch concentration in retroglenoid vein blood, F,,, = 18.82, p < 0.0001;the second and third hypoxia periods and recovery period were different from the control (p < 0.01).
hypoxic, and recovery periods of 10- and 20-min series showed that the slope relating cerebral A-V Ch to arterial Ch was positive and significantly different from 0 (Fig. 3), supporting the concept that increased arterial Ch concentration is associated with a net gain of Ch by the brain.
Examination of regional A-V differences for Ch levels revealed that lungs released significant amounts of Ch into the circulation during hypoxia, starting early after commencement of hypoxia and reversing rapidly after its termination (Fig. 5). The splanchnic territory also showed an increase of its output of Ch over control levels. This phenomenon was slower in onset than the one observed in lungs, but it lasted beyond termination of hypoxia (Fig. 6).
Extracerebral sinks for Ch The liver showed a significant increase in uptake of Ch after 10 min of hypoxia when compared with the control period ( p < 0.01) but not after 5 min of hypoxia (Fig. 6). However, this phenomenon did not offset the increase in splanchnic output completely because the Ch concentration in suprahepatic vein blood increased significantly over control values after 10 min of hypoxia ( p < 0.05; Fig. 6). Kidney and lower limb also showed significant uptake of Ch during the hypoxic period (Fig. 7). DISCUSSION The present experiments were designed to test the hypothesis that hypoxia would enhance the cerebral loss of Ch through the circulation. On the contrary, the experiments revealed that the negative A-V difference for Ch levels across brain disappeared during hypoxia, a phenomenon most likely related to the dramatic surge in arterial Ch concentration in this condition. This is suggested by the significant dependence
160
1
1.1 Car,ran.Cn
120
Extracerebral sources of Ch The blood itself was initially examined as a source of Ch in hypoxia. Two experiments were performed in which blood extracted from normoxic rats was equilibrated in vitro with air or nitrogen as described in Materials and Methods. Po, was initially 73.6 and 64 mm Hg in experiments 1 and 2, respectively,and it increased in the blood of the microtonometers that contained air to final values of 93.8 (experiment 1) and 94.8 (experiment 2) mm Hg, but dropped continuously in those containing N,, to final values of 28.2 (experiment 1) and 25.5 (experiment 2) mm Hg at the end of 30 min. A steady increase in plasma Ch as a function of time could be demonstrated, with air- and nitrogen-equilibrated samples being represented by the same regression line (Fig. 4). No such trend occurred for RBC Ch. This phenomenon was not affected by the partial pressure of O2because no apparent differences were observed in plasma or RBC Ch for blood exposed to air or nitrogen.
909
80
Y - 0.82 X
+ 4.03
r- 0.97
< 0.0001
P
./
,
-/ IoICer. A - v cn Y - 0.08 X - 4.03 P- 0.002
r- 0.31
40
-
a
0
-
'
.
..'..
e
*D
B
~
2
0
-40
',So
o
0
Mq..0
I
I
I
I
40
80
120
160
200
J. Neirrachem.. Vol. 59. No. 3, 1992
0. U. SCREMIN AND D. J. JENDEN
910
Plasma
RBC
20.0,
-
20.0
Y = 0.16 X t 7.37
16.0
r
=
=
0.88, P
