Lesions of the rostral ventrolateral medulla reduce the cerebrovascular response to hypoxia

BRAIN RESEARCH ELSEVIER

Brain Research 635 (1994) 217-223

Lesions of the rostral ventrolateral medulla reduce the cerebrovascular response to hypoxia Mark D. Underwood *, Costantino Iadecola, Donald J. Reis Department of Neurology, Dwtston of Neuroblology and Neurosctence, Cornell UntL,ersttyMedtcal College, New York, NY 14853, USA (Accepted 21 September 1993)

Abstract Sympathoexcitatory neurons of the rostral ventrolateral medulla are tomcally active and reqmred for maintenance of resting levels of arterial pressure. They are also selectively excited by hypoxia and responsible for the associated sympathoexcitation. Since electrical or chemical stimulation of RVL will increase regional cerebral blood flow (rCBF) independently of changes in regional cerebral glucose utilizaUon (rCGU) we investigated whether the RVL was also required to maintain resting levels of rCBF and also participated in the cerebrovascular vasodilation elicited by hypoxm. Rats were anesthetized (chloralose; 40 m g / k g , s.c.), paralyzed (tubocurarine) and ventilated (100% 02). rCBF was measured in 10 dissected brain regions using [14C]iodoantlpyrine; rCGU was measured by 2-deoxy-D-[lac]glucose. In controls (n = 6) rCBF ranged from 56 + 5 in corpus callosum to 101 + 6 m l / m i n × 100 g in inferior colliculus. Hypoxlc-hypoxia (PaO 2 = 36 + 1 mmHg, n = 6) increased rCBF m all structures maximally, at 204% of control, in occipital cortex. Hypercapnia (PaCO 2 = 63.5 + 0.9, n = 5) also increased rCBF ( P < 0.01) maximally to 299% of control in superior colliculus. Spinal cord transection with maintenance of arterial pressure did not affect resting rCBF and increased the vasodilatlon to hypoxia (PaO 2 = 39 + 1 mmHg, n = 5) from 2- to 3-fold m all structures ( P < 0.01). Bilateral lesions within the RVL had no effect on resting rCBF or rCGU. However, they significantly reduced, in all areas by 50-69% ( P < 0.01, n = 5), the cerebrovascular dilation ehclted by hypoxia but not hypercapnia. Bilateral lesions in the spinal trlgeminal nucleus (PaO 2 = 35 _+ 1; n = 6), or transection of the IXth and Xth cramal nerves did not affect the rCBF response to hypoxia (PaO2 = 41 + 2; n = 6) ( P > 0.05) indicating that the effect of RVL lesions was not attributable to interference with arterial baro- or chemoreceptor reflexes. We conclude that neurons within RVL are not responsible for maintaining tonic levels of rCBF. However they contribute to the cerebrovascular vasodllatlon elicited by hypoxia but not hypercapma. The cerebrovascular response to hypoxla appears reflexive and, in part, due to stimulation of oxygen-sensing neurons in RVL. In contrast, the vasodilation ehclted by hypercapnla reflects local chemical signals in the cerebral microcirculation.

Key words. Cerebral blood flow; Cerebral glucose utilization; Hypoxla; Rostral ventrolateral medulla; Rat

1. Introduction S y m p a t h o e x c i t a t o r y r e t i c u l o s p i n a l n e u r o n s within a s u b z o n e o f v e n t r o l a t e r a l r e t i c u l a r nucleus of t h e m e d u l l a o b l o n g a t a ( R V L ) , t h e C1 a r e a [30], m a i n t a i n resting levels o f a r t e r i a l p r e s s u r e ( A P ) by tonically exciting p r e g a n g l i o n i c s y m p a t h e t i c n e u r o n s o f t h e spinal cord. T h e s e R V L v a s o m o t o r n e u r o n s a r e also inti-

* Corresponding author. Present address Laboratories of Neuropharmacology, Western Psychiatric Institute and Chine, Biomedical Soence Tower, W1643, Pittsburgh, PA 15213, USA. Fax. (1) (412) 624 0233 0006-8993/94/$07.00 © 1994 Elsevier Soence B.V All rights reserved

SSDI 0 0 0 6 - 8 9 9 3 ( 9 3 ) E 1 2 8 8 - E

m a t e l y involved in the r e g u l a t i o n o f s y m p a t h e t i c n e u rons in r e s p o n s e to hypoxia. R V L n e u r o n s n o t only m e d i a t e t h e reflex s y m p a t h o e x c i t a t i o n elicited by hypoxic s t i m u l a t i o n of a r t e r i a l c h e m o r e c e p t o r s [22], b u t a r e also directly a n d selectively excited by hypoxia a n d / o r b r a i n s t e m ischemia [6,12,37,38,42]. T h a t n e u r o n s of t h e C1 a r e a of R V L m a y also influence t h e c e r e b r a l c i r c u l a t i o n has b e e n s u g g e s t e d by o b s e r v a t i o n s that electrical or c h e m i c a l excitation o f R V L n e u r o n s will e l e v a t e r e g i o n a l c e r e b r a l b l o o d flow ( r C B F ) [10,32,45] w i t h o u t c h a n g e in r e g i o n a l c e r e b r a l glucose u t i l i z a t i o n ( r C G U ) [45]. T h e o b s e r v a t i o n s raise two questions. First, d o e s the R V L c o n t r i b u t e to the tonic r e g u l a t i o n o f r C B F ? Like the systemic circula-

218

M D Underwood et al /Brain Research 635 (1994) 217-223

tion, the cerebral c~rculation is tonically neurogenically maintained above basal levels depending on the activity of brain areas since resting rCBF can be reduced by lesions [13], cold blockade [34] or by stimulation of brainstem or cerebellum [5,24]. Second, does the RVL contribute to the cerebrovascular vasodilation ehctted by hypoxia? Hypoxta and hypercapnia, like stimulation of RVL, elevate rCBF but not rCGU [2,4,17] by an action independent of arterial chemoreceptors [27,28, 44]. However, only hypoxia excites RVL sympathoexcitatory neurons [37-40,42]. In the present study we have therefore investigated the effects of bilateral electrolytic lesions of the RVL on resting rCBF and rCGU and on the elevations in rCBF elicited by hypoxia or hypocapnia. We demonstrate that bilateral lesions of the RVL have no effects on basal flow or metabolism. However, they substantially reduce the vasodilation elicited by hypoxia but not hypercapnia suggesting that the two responses are mediated by independent mechanisms and that the RVL participates in this cerebrovascular reflex.

2. Materials and methods The methods used m this study have been extensively documented in previous studies from this Laboratory [15,24,25] and will only be summarized All aspects of the surgical preparation and experimental protocol were approved by the Institutional Animal Care and Use Committee of Cornell Unwerslty Med,cal College

2 1 General surgtcal procedures Studxes were undertaken in male Sprague-Dawley rats weighing 325-425 g After induction with halothane (2-5% m 100% 0 2) a-chloralose (40 m g / k g , s c ) was administered and halothane was continued at a reduced rate (1%) for the next one hour. Catheters were placed in each femoral artery and vein and the trachea was cannulated. The animal was placed m a stereotaxlc frame and a hmlted cranlotomy was performed for insertion of electrodes The animal was then paralyzed with tubocurarlne (0.5 mg/kg, s c ) and ventilated with 100% O 2 A small volume of arterial blood (0.3 ml) was sampled one hour following the onset of artificial ventilation and again just before the measurement of cerebral blood flow for the measurement of PaO 2, PaCO 2 and pH by a blood gas analyzer (Instrumentation Laboratory Mod Mlcrol3) These procedures were typically completed within one hour Animals were then placed in a stereotaxlc frame with the head placed so that the floor of the fourth ventricle was horizontal (bite bar = - 1 1 mm) One of the arterial catheters was connected to a Statham P23Db transducer to continuously monitor systemic AP and HR on channels of a chart recorder The medulla and caudal portion of the cerebellum were exposed by an occipital cranlotomy, marking the completion of surgery, and halothane was discontinued. Body temperature was measured rectally and maintained 37-38°C (YSI Instruments Inc. Mod 73) by an infrared lamp Electrolytic lesions m the RVL or spinal trlgemmal nuclei were produced by passage of anodal current (750/xA for 30-45 s) from a lesion maker (Grass Mod LM5A) Electrodes were fabricated from stainless-steel needles (o d 200 g m ) Insulated with Epoxyhte except for the exposed hp (400/zm) The cathode was a clip applied to the

animal's scalp In experiments m which the effects of bilateral lesions of the RVL were evaluated, the most sens,tlve site m the RVL for elevating AP was located bilaterally by electrical stimulation with the les~omng electrode The electrode was left In place whale spinal cord transection was performed (see below) and lesions made 30-60 mm before the infusion of IAP or 2-DG for rCBF or rCGU measurement, respectively. Interruption of chemoreceptor afferents arising from the aoruc arch and carot,d bod,es was produced by bilateral transection of the vagus (Xth) and glossopharyngeal (IXth) nerves Following measurement of rCBF or rCGU and removal of the brain, the portion of the medulla containing the lesion was placed m freon ( - 2 5 ° C ) , subsequently sectioned at 30 p,m and stained with thlonln and lesions reconstructed

2 2 Measurement of cerebral blood flow and metabohsm rCBF was measured m tissue homogenates using 4-,odo-[N-

methvl-laC]antlpyrlne (lAP, New England Nuclear, 40-60 mC1/ mmol) as indicator [15] Heparln (2000 U l.V, A H Robblns) was administered and, 15 mln later, IAP was infused intravenously over 30-40 s by an infusion pump (Harvard Apparatus, Mod 940). Arterial blood was sampled to obtain the arterial concentration time curve of IAP Ahquots (40 tzl) of blood were transferred to vials, solublhzed (Protosol, New England Nuclear), incubated (55°C for 30 mln) and decolorlzed with hydrogen peroxide (30%) Ethanol (2 ml) and scintillator (15 ml, Beckman H P / b ) was added and radioactivity (dlslntegrahons per minute) was measured by a hquld scmtdlatlon spectrophotometer (Beckman Mod. LS5801) After the animal was killed (1 ml saturated KCI Iv ), the brain was removed and dissected bdaterally Into cerebellum, inferior colhculus, superior colhculus, hypothalamus, frontal, occipital and parietal cortices, thalamus, hlppocampus, caudate nuclei and white matter Samples were transferred to prewelghed scintillation vials and, after tissue solublhzatlon (Protosol, NEN), ethanol (2 ml) and scintillation cocktail (15 ml) was added and the amount of rad,oactwlty determined rCBF (ml/100 g / m l n ) was calculated using a computerized approximation of Kety's integral from the relationship between CBF and tissue IAP concentration rCGU was measured m dissected brain regions using the 2-deoxyD-[14C]glucose (2-DG) technique [36] 2-DG (New England Nuclear, spec act 50-60 m C , / m m o l ) was dissolved m saline after ehmlnatlon of ethanol and refused (12 5/zCl/100 g body weight) intravenously at a constant rate over 40 s Arterml blood was sampled every 10 s during the first minute, at 90 s and at 2, 3, 5, 10, 15, 25, 35 and 45 mln after infusion onset Blood samples were centrifuged and stored in ice following collection Ahquots (20 /zl) of plasma were transferred to vials, had scintillator added and the radloactw,ty measured as described above Plasma glucose concentration was measured with a glucose analyzer (Beckman Mod 2) rCGU ( ~ m o l / 1 0 0 g × m l n ) was calculated from the regional brain radloactlwty and the arterial t~me-course of 2-DG and glucose [33]. The values adopted for the rate and lumped constants were those determined by Sokoloff et al [36] m the albino rat

2 3 Spectfic procedures In experimental animals, after administration of anesthesm, cannulatlon, placement of the ammal in the stereotaxlc frame and exposure of the cerebellum, the electrode was inserted into the cerebellum 2 0 mm rostral to the calamus scrlptorlus and 1 9 mm lateral to the mldhne The electrode was then lowered through the medulla in 0 2 mm increments with stimulation at each step to locate the most active site for an elevation of AP The stimulus consisted of an 8 s tram of pulses dehvered at a frequency of 100 Hz with a pulse duration of 0.5 ms and a st,mulus current of 10-20 /zA Such

M D Underwoodet al. / Bram Research 635 (1994) 217-223 stimulation never elevated A P more than 15 m m H g Such precautions were taken because an abrupt rise m arterial pressure above the autoregulated range might result in paralysis of vascular responsivity and loss of autoregulatlon [19]. Once the most sensltwe site m the R V L was located, the electrode was left m place. Halothane ( 1 - 2 % ) was then administered and electrolytic lesions ( 7 5 0 / z A for 30-45 s) were placed In the R V L bilaterally To prevent the fall in A P resulting from bilateral lesion within the RVL, A P was maintained by continuous intravenous infusion of phenylephrme. Procaine (0.1 ml, 2%) was reJected into the spinal cord between the first and second cervical vertebrae to abort the elevation in A P that results from spinal trauma. A segment of spinal cord (2 ram) was then aspirated with AP maintained by continuous intravenous infusion of phenylephrme. Following completion of the spinal cord transection halothane was d~scontlnued and the animals were allowed to stabihze for 1 h prior to infusion of tracer. Blood gases were carefully adjusted so that the PaCO 2 was maintained within the normocapmc range by adjusting the stroke volume of the ventilator P a O : was maintained by ventilation with 100% 0 2 to offset the atelectasls and hypoxm invariably assocmted with paralysis and artificial ventdatlon [26]. O u r resulting PaO~ values are well above the 80 m m H g normal level for rat However, this ~s of httle consequence since hyperoxm is not a significant variable in controlling rCBF [29,35]. Hypoxlc-hypoxia was established by varying the mspiratory oxygen.mtrogen ratio to an P a O : of approximately 40 m m H g . Hypercapma (PaCO 2 = 6 0 ± 1 m m H g ) was produced by introduction of CO 2 into the ventllatory system Ten to fifteen mins after the start of hypoxla or hypercapnm, l A P was infused The cerebrovascular response to hypoxla was studied m two groups of spinal cord transected animals. Lesions in the R V L were made one hour prior to infusion of l A P with the A P maintained by continuous infusion of phenylephrine ( 1 - 4 t z g / m i n ) In the group of animals in which rCBF was measured following transection of the IXth and Xth cranial nerves, the gradual rise in arterial pressure

219

consequent to the transection was controlled by removal of a small amount of arterial blood untd the desired level of AP was obtained The spinal cord transection was performed approximately one hour following the nerve section with A P similarly maintained by infusion of phenylephrme. Forty-five minutes following completion of the spinal cord transection animals were made hypoxlc by varying the oxygen.nitrogen ratio of inspired air. rCBF was measured 10-15 m m later during which time arterial blood gases were measured The duration of rCBF and r C G U experiments was 3 h and 3.5 h followmg the administration of a-chloralose, respectwely. rCBF was measured in spinal cord transected ammals with (n = 5) and without (n = 5) bilateral electrolytic lesions in the R V L CO 2 was introduced through the ventdator approximately 45 m m followmg completion of spinal cord surgery and placement of bilateral lesions. Ten to fifteen minutes later IAP was infused and rCBF measured Controls consisted of rats with spinal cords transected and AP maintained by phenylephrme.

2 4 Statlsttcal analysts Bdateral differences m rCBF were evaluated by the Student's pa~red t-test Multiple comparisons between experimental and control groups were made using the analysis of variance and N e w m a n Keuls tests

3. R e s u l t s

3.1. Effect o f R V L lesions on r C B F and r C G U

Bilateral lesions of the RVL result in cessation of ongoing sympathetic nerve activity requiring maintenance of AP with continuous infusion of pressor agents.

Table 1 Arterml pressure (AP), blood gases, regional cerebral blood flow (rCBF) and regional cerebral glucose utdlzatlon (rCGU) m controls and ammals with bdateral electrolyhc lesion of the rostral ventrolateral medulla (RVL) with or without spinal cord transection during m e a s u r e m e n t of regional cerebral blood flow (rCBF) or regional cerebral glucose utilization (rCGU) Control

R V L lesion

Intact

Spinal Trans. a

CBF

CGU

CBF

Spinal Trans a CGU

CBF

CGU

Regton Cerebellum Inf. Coil Sup Coll. Hypothalamus Thalamus Caudate Hippocampus FrontalCx. ParietalCx OccipltalCx. Corpus Call

68 101 97 95 79 82 70 83 95 82 56

± 7 + 6 + 7 + 6 ± 9 ± 9 ± 8 + 6 ±10 ± 8 ± 5

21 51 50 42 46 51 44 53 59 64 37

+ 3 + 4 + 3 5:3 ± 3 ± 4 ± 3 _+ 3 ± 3 ± 5 ± 2

68 103 91 84 71 77 57 69 76 66 50

+ + + + ± _+ ± ± ± ± ±

4 7 8 6 5 7 3 4 5 6 6

18 41 34 29 33 35 29 43 41 42 26

_+ + ± + ± ± ± ± _+ ± ±

3 2 2 * 2 * 2 * 3 * 3* 2 4" 3 * 2 *

67 105 86 74 67 72 60 76 83 78 51

± + _+ ± ± ± ± ± _+ ± ±

5 6 5 3 2 3 3 4 5 5 2

15 35 32 25 30 30 23 41 42 40 24

+ + + ± ± _+ ± + + ± +

2 3 1" 1* 1" 1" 1* 4 5" 3" 1

Phystologtcalparameters AP(mmHg) P a C O 2 (mmHg) P a O 2 (mmHg) pH n

134 ± 34.7 ± 413 ± 7.43+ 6

4 04 15 001

131 + 36.3 ± 358 ± 744+ 5

4 0.6 59 0.01

124 ± 36.4 + 402 + 740± 6

2 0.5 18 002

116 ± 7 34.7 ± 0.9 370 ± 60 7 4 4 ± 001 5

123 ± 35.6 ± 457 ± 7.40± 5

4 07 20 001

* P < 0 05 from intact control P < 0.05 from spinal cord transected control (analysis of variance and N e w m a n - K e u l s test) a Spinal cord transected at first cervical segment with A P maintained by phenylephrlne infusion

120 + 33.3 + 428 ± 7.45+ 4

6 0.5 # 38 0.02

M D Underwood et al /Bram Research 635 (1994) 217-223

220

Thus, to determine whether the conditions required to maintain AP influenced rCBF and rCGU, these were measured regionally in groups of anesthetized rats that were otherwise intact, in which the spinal cord was transected and AP maintained by phenylephrine and in a third group in which the spinal cord was transected and the RVL electrolytically lesioned bilaterally (Table 1) maintaining AP in the same way. Blood gases and AP did not appreciably differ between groups (Table 1). As indicated in Table 1 the rCBF and rCGU of spinalized rats with or without RVL lesions did not differ from each other or from intact (control) rats. Lesions of RVL always destroyed the region of RVL containing the rostral cluster of C1 cells [11,31] and also damaged the inferior olive medially, the spinal trigeminal nucleus (STN) laterally, the nucleus ambiguus dorsally and the ventral surface of the medulla (Fig. 1). Such lesions were typical of RVL lesions verified histologically in the remainder of the study.

3.2. Effect of btlateral lestons of RVL, spinal trigeminal nuclet (STN), or transection of IXth and Xth cranial nert,es on the reflex cerebrouascular response to hypoxia rCBF and rCGU were measured during hypoxia produced in groups of spinalized rats without or with lesions of RVL or, as control, with lesions of STN. In spinalized rats, hypoxia (36 + 1 mmHg, n = 6; Table 2) globally increased rCBF maximally, at 204% in occipital cortex ( P < 0.01)(Fig. 2), with values comparable to those obtained by others in anesthetized rat [3,18]. In agreement with others [17], rCGU was unaffected (PaO 2 = 36 + 1 mmHg, n = 6; P > 0.05) (Table 2).

ECN, ICP.

.~ \,

/ \

/ //

RPa _

_

J imm

Fig. 1 Schematic drawing dlustratmg the location of representatwe electrolytic lesions of the CI area of the rostral ventrolateral medulla (RVL) or spinal tngemmal nucleus (STN) The dark stippling represents the location of lesions encompassing the RVL The hght stippled boundaries mark the location of representatwe control lesions Note that bdateral lesions within the RVL, but not adjacent regions, slgmficantly reduced the cerebrovasoddatlon ehoted by hypoxm but not hypercapnm CST, cortlcospmal tract, ECN, external cuneate nucleus, ICP, inferior cerebellar peduncle, IO, inferior ohvary nucleus; IVN, inferior vestibular nucleus; MLF, medial longitudinal fasoculus, MVN, medml vestibular nucleus, NTS, nucleus of the sohtary tract, PP, nucleus preposltus; RF, retrofacml nucleus, RPa, raphe palhdus, RO, raphe obscurus, RVL, rostral ventrolateral medulla, STN, spinal trlgemmal nucleus, STT, spinal trlgemlnal tract

Bilateral lesions of the RVL (Fig. 1) significantly attenuated the magnitude of the increase in rCBF elicited by hypoxia (PaO 2 = 40 __. 1 mmHg) in all brain regions, maximally, by 69%, in the inferior and superior colliculi (P < 0.05, Fig. 2; Table 2). To control for non-specific effects of the lesion, bilateral electrolytic lesions were placed in the STN.

Table 2 Effect of electrolytic lesions of the rostroventrolateral medulla (RVL) on the cerebrovascular response to hypoxm or hypercarbm Hypoxm

Hypercarbia

Intact CBF a

Intact CGU h

RVL lesion d

STN lesion ~

Denervatlon a

Intact d

RVL lesion d

146 124 113 104 108 94 199 136 122 85

278 224 186 168 167 140 189 218 218 114

249 219 174 180 170 140 199 210 210 120

27l 271 202 178 210 146 170 166 159 113

335 310 189 198 198 159 184 175 176 128

Regton Inf Coil Sup Coil Hypothalamus Thalamus Caudate Hlppocampush Frontal Cx Parietal Cx Ocopltal Cx Corpus Call

244 198 173 158 163 137 175 201 197 120

+ 18 _+ 18 _+ 16 _+ 13 _+ 15 _+ 11 _+ 15 + 18 _+ 20 _+ 11

37 30 26 30 30 26 35 38 42 23

-+ 3 _+ 3 + 2 _+ 3 _+ 2 _+ 2 +_ 2 _+ 3 + 3 _+ 1

_+ 6 + 7 _+ 6 _+ 5 _+ 10 _+ 6 _+ 6 _+ 6 _+ 7 _+ 10

* * * * * * * * * *

+_ 25 ± 18 _+ 16 _+ 15 _+ 14 _+ 10 -+ 21 -+ 18 _+ 25 _+ 7

_+ 40 _+ 36 + 22 _+ 28 _+ 21 _+ 18 _+ 30 + 29 + 29 _+ 16

_+ 19 + 15 _+ 8 -+ 14 + 41 + 9 _+ 10 _+ 9 -+ 7 _+ 3

+ 37 _+ 39 -+ 13 -+ 16 _+ 17 _+ 10 _+ 15 _+ 14 _+ 8 +_ 10

Phystologtcal parameters AP(mmHg) PaCO z PaO 2 pH n

115 _+ 342 _+ 39 _+ 739_+ 7

5 06 1 002

102 _+8 359 +_08 36 _+ 1 740_+002 5

112 _+ 35.8 _+ 41 _+ 740_+ 5

6 0.3 1 001

112 _+ 36.1 -+ 35 _+ 739_+ 6

6 0.7 1 002

119 _+ 336 _+ 41 _+ 728_+ 6

8 16 1 004

123 _+ 3 635 -+ 0 9 462 -+35 722_+ 001 5

119 _+ 5 644 + 05 386 +31 720_+ 001 5

* P < 0 05 from intact controls, spinal tngemmal nucleus (STN) lesioned and denervated groups (analysm of variance and Neuman-Keuls tests) d Values are mean -+ S.E.M m l / m m × 100 g b Values are mean _+ S E M ~zmol/mm × 100 g.

M.D Underwood et al ~Brain Research 635 (1994) 217-223 400 ~"

o

300 l -

"6

200

[]

Intact, *p < 0 01 from control (n = 6)

*

•

"

•

ooij 0

400 vl~ "5

bilaterally in spinalized rats. Such lesions interrupt all arterial chemoreceptor afferents reaching the brainstem. In agreement with others [27,28,44], deafferentation of arterial chemo- (and baro-) receptors had no effect on resting rCBF or the magnitude of the response to hypoxia (Table 2).

[ ] Lesion, • p < 0 01 from control & intact (n = 5) , *

_

lnfC SupC Hyp Thai

T ]" Ill~ T /

CN

Htpp FCx PCx OCx

CC

[ ] CO2 (n = 5) [ ] Les'°n + 002 (n = 5) AIIstructuresabovecontrolp<005

300

"~ 200

3.3. Effect of RVL lesions on the reflex cerebroz'ascular response to hypercapnia Elevation of PaCO: to 63.5 + 1 mmHg in 5 spinalized rats globally and significantly (P < 0.01) increased rCBF (Fig. 2), maximally to 299% in brainstem, values similar to those previously reported from this laboratory in intact animals [1,14,16,25]. Lesions of RVL did not modify the magnitude of the vasodilation (PaCO 2 = 64.4 + 1 mmHg; n = 5).

4. Discussion

100

0

221

InfC SupC Hyp Thai

CN

Hipp FCx PCx OCx

CC

Fig. 2 U p p e r effect of bilateral electrolytic lesions In the rostral ventrolateral medulla (RVL) on the increase m regional cerebral blood flow (rCBF) ehcJted by hypoxia Note that bilateral lesions within the R V L significantly reduced the m a g m t u d e of the cerebrovascular response to hypoxna m all brain regions and abolished nt m all brain stem regions and the caudate nucleus The residual increases in rCBF to hypoxm were significantly elevated above control across the cerebral cortex, m the hlppocampus and the corpus callosum ( P < 0.05 from intact controls). Crb, cerebellum, Inf C, inferior colhculus; Sup C, superior colllculus; Hyp, hypothalamus; CN, caudate nucleus, Hlpp, hlppocampus; FCx, frontal cortex; PCx, parietal cortex, OCx, occipital cortex; CC, corpus callosum. Lower effect of bilateral electrolytic lesions m the rostral ventrolateral medulla (RVL) on the cerebrovascular response to hypercapma in spmal cord transected rats. Note that m unlesioned controls hypercapnla ehcited the well estabhshed global increase m rCBF ( P < 0 05) and that bilateral lesions m the R V L did not affect the response m any brain region.

Typical lesions of STN are depicted in Fig. 1 and extended to the lateral portion of the parvocellular reticular nucleus medially, the spinal trigeminal tract laterally, the inferior cerebellar peduncle and external cuneate nucleus dorsally and to the ventral border of the STN but entirely spared the C1 area of RVL. Lesions of STN had no effect on the vasodilation elicited by hypoxia (PaO 2 = 35 + 1 mmHg; P > 0.05; Table 2). To determine if the attenuation of the hypoxic vasodilation by RVL lesions resulted from impairment of arterial chemoreceptor reflexes which are mediated by RVL, the IXth and Xth cranial nerves were transected

In the present study we have investigated whether bilateral electrolytic lesions of the RVL contribute to the tonic regulation of rCBF a n d / o r rCGU and whether they participate in the cerebrovascular vasodilation elicited by hypoxia or hypercapnia. To avert the changes in rCBF consequent to the fall of AP below the autoregulated range which follow lesions of the area (e.g. [31]) we studied rats in which AP and blood gases were stabilized by transecting the spinal cord and maintaining AP by phenylephrine. As we have demonstrated elsewhere [15,25] spinal cord transection in anesthetized rat does not modify restmg rCBF or rCGU when AP and blood gases are maintained. Moreover, as demonstrated here, it does not alter the global cerebrovascular vasodilation elicited by hypoxia or hypercarbia. We observed that unlike the effects on sympathetic activity, bilateral lesions of the RVL had no effect on resting rCBF or rCGU. The absence of an effect cannot be attributed to rCBF a n d / o r rCGU being at a 'floor' since under comparable experimental condition they can be reduced, by up to 50%, by electrical stimulation of the parabrachial nucleus [24] or by chemical stimulation of the cerebellar fastigial nucleus [5]. Moreover, that lesions of the brainstem [13] or anesthesia [7] will substantially reduce resting r C B F / rCGU indicates that cerebral mechanisms contribute to set rCBF within midrange. While not influencing resting levels of rCBF, lesions of the RVL reduced, by over 50%, the cerebrovascular vasodilation elicited by hypoxia. The attenuation of the response by RVL lesions appears anatomically and functionally selective. Thus lesions of adjacent STN were without effect indtcating that the attenuation was not the result of nonspecific damage. Moreover since

222

M D Underwood et al /Bram Research 635 (1994) 217-223

restmg r C G U was not changed indicates that diminished response was not secondary to a reduction of cerebral metabolism. Interruption of all afferent information from arterial, pulmonary and cardiac chemoreceptors by bilateral transection of the IXth and Xth cranial nerves also failed to modify the responses to hypoxia indicating that the effect of R V L lesions was not due to interrupting these reflexes whose cardiovascular c o m p o n e n t s are integrated in RVL. It also confirms the long-standing view that arterial chemo- or baroreceptors do not contribute to the reflex control of r C B F by hypoxia [28,41,43]. Finally, the observation that R V L lesions failed to impair the cerebral vasodtlation ehcited by hypercapnia indicates that cerebrovascular reactivity was not impaired by the lesions. The findings therefore strongly suggest that a substantial c o m p o n e n t of the cerebrovascular vasodilatlon elicited by hypoxia is reflexive and d e p e n d e n t on the integrity of the RVL. In contrast, the response to hypercapnla appears to be mediated in large measure by a distinct mechanism conceivably, as widely believed, the result of a direct action of H ÷ generated by CO2 on vascular smooth muscles in the cerebral microcirculation [20]. That neurons within the R V L may selectively mediate the cerebrovascular effects of hypoxia is supported by recent findings that sympathoexcitatory neurons but not adjacent respiratory or noncardiovascular non-respiratory neurons of R V L are directly excited by hypoxia and cyanide in vivo and in vitro but not hypercapnia or H ÷ [37-40,42]. While electrolytic lesions of the R V L would destroy elements other than intrinsic neurons in the region, the facts that chemical a n d / o r electrical stimulation of the R V L evoke elevations of r C B F of comparable magnitude [10,32,45] which are also unassociated with changes in r C G U [45] make it most likely that the responses can be attributed to neurons in the region. M o r e o v e r the fact that these neurons are richly e n d o w e d with mitochondria and are in extremely close apposition to blood vessels, even to the point of being penetrated by capillaries [23], is consistent with the view that they may be central oxygen sensors. T h e effector pathway mediating the R V L - d e p e n d ent response to hypoxia is not known. It is not likely mediated by extracerebral neural pathways innervating cerebral vessels. Sympathetic nerves can be excluded since they are vasoconstrictor, will not be centrally activated in animals with the spinal cord transected, and the vasodilation elicited by electrical stimulation of the R V L persists after sympathectomy [45]. While parasympathetic fibers traveling with the V I I t h nerve when sttmulated will elicit a primary vasodilation [8], the magnitude of vasodilation evoked by electrical stimulation of the nerves is substantially less than that elicited by hypoxia. Finally, while the sensory fibers of the trigeminal nerve will, when antidromically excited,

elicit a primary vasodilation [9], lesions of the Vth cranial nerve have no effect on the response. More likely the effect is mediated by excitation of an intrinsic pathway originating in, or projecting to or through, the R V L and acting through connections yet to be determined that widely ramify throughout brain. Since projections from R V L do not innervate the cerebral cortex [21], the increase in r C B F in that target must be mediated t h r o u g h a relay in the brain stem whose identity has yet to be established. In conclusion our results suggest that neural systems in part related to R V L participate in the cerebrovascular vasodilation elicited by hypoxia. Such a system may be important in providing amplification of the vasodilator response to hypoxia generated by local mechanisms thereby serving to offer protection to neuronal viability.

References [1] Arnerlc, S.P, Iadecola, C, Underwood, M.D. and Rexs, D.J., Local cholinerglc mechamsms participate m the increase m cortical cerebral blood flow ehclted by electrical stimulation of the fasttgial nucleus in rat, Bram Res, 411 (1987) 212-225 [2] Berntman, L. and Slesjo, B K, Cerebral metabohc and orculatory changes induced by hypoxla m starved rats, Z Neurochem, 31 (1978) 1254-1276 [3] Borgstrom, L, Johannsson, H and Slesjo, B K, The relationship between arterial pO2 and cerebral blood flow m hypoxic hypoxla, Acta Phystol Stand, 93 (1975) 423-432 [4] Borgstrom, L, Norberg, K and Slesjo, B.K, Glucose consumption in rat cerebral cortex m normoxla, hypoxia and hypercapma, Acta Phystol Scand, 96 (1976) 569-574 [5] Chlda, K, Iadecola, C and Rels, D J, Global reduction m cerebral blood flow and metabolism ehclted from intrinsic neurons of fastiglal nucleus, Brain Res, 500 (1-2) (1989) 177-192 [6] Dampney, R A L and Moon, E A, Role ofventrolateral medulla in vasomotor response to cerebral lschemla, Am J Phystol, 239 (1980) H349-H358 [7] Dudley, R.E, Nelson, S R and Samson, F, Influence of chloralose on brain regional glucose utlhzatton, Bram Res, 233 (1982) 173-180 [8] Goadsby, P.J, Effect of stimulation of faoal nerve on regional cerebral blood flow and glucose utlhzatlon m cats, Am J Phystol Regul Integr Comp Phystol, 257 (1989)R517-R521 [9] Goadsby, P J and Duckworth, J W., Effect of stimulation of tngemmal ganghon on regional cerebral blood flow in cats, Am J Phystol, 253 (1987) R270-R274 [10] Golanov, E.V and Rels, D.J, Nztrlc oxide and prostanolds participate m cerebral vasodllatlon ehcited by electrical stimulation of rostral ventrolataeral medulla, J Cereb Blood Flow Metab, m press [11] Granata, A R, Rugglero, D A, Park, D H., Joh, T H and Rels, D J, Brain stem area wzth C1 epinephrine neurons medmtes baroreflex vasodepressor responses, Am J Physlol, 248 (1985) H547-H567. [12] Haselton, J R., Haselton, C L, Vera, P L., Ellenberger, H.H, LeBlanc, W.G, Scbnelderman, N and McCabe, P M., Nucleus reticulans laterahs involvement m the pressor component of the cerebral Ischemlc response, Bram Res, 335 (1985) 315-320 [13] Hass, W K, Hawkins, R A and Ransohoff, J, Reduction of cerebral blood flow, glucose utlhzatlon, and oxldatwe metabolism

M D Underwood et al /Bram Research 635 (1994) 217-223 after bilateral reticular formation lesions, Trans Am Neurol Assoc, 102 (1977) 19-20. [14] ladecola, C , Mraowtch, S., Meeley, M.P and Rels, D J , Lesions of the basal forebraln in rat selectively impair the cortical vasodllat~on ehclted from cerebellar fast~gml nucleus, Bram Res, 279 (1983) 41-52. [15] Iadecola, C, Nakal, M , Arblt, E and Reis, D.J., Global cerebral vasodflatatlon ehclted by focal electrical stimulation within the dorsal medullary reticular formation m anesthetized rat, J Cereb Blood Flow Metab , 3 (1983) 270-279. [16] Ishltsuka, T , Iadecola, C , Underwood, M.D and Reis, D.J., Lesions of nucleus tractus sohtarn globally impair cerebrovascular autoregulatlon, Am. J Phystol, 251 (1986) H269-H281 [17] Johannsson, H. and Slesjo, B K., Cerebral blood flow and oxygen consumption m the rat m hypoxlc hypoxm, Acta Phystol Scand, 93 (1975) 269-276 [18] Kagstrom, E., Smith, M.L. and S~esjo, B K, Cerebral c~rculatory responses to bypercapma and hypoxla In the recovery period following complete and incomplete cerebral ischemia m thr rat, Acta Physlol Scand, 118 (1983) 281-291 [19] Kontos, H.A., Wei, E.P., Navarl, R.M, Levasseur, J E , Rosenblum, W I. and Patterson, J.L ,Jr, Responses of cerebral arteries and arterioles to acute hypotenslon and hypertension, Am J Phystol, 234 (1978) H371-H383 [20] Kuschinsky, W. and Wahl, M., Local chemical and neurogemc regulation of cerebral vascular resistance, Phystol Ret,, 58.3 (1978) 656-689 [21] Loewy, A D., Wallach, J.H. and McKellar, S, Efferent connections of the ventral medulla oblongata m the rat, Bratn Res Ret,, 3 (1981) 63-80 [22] Mlllhorn, D.E and Eldndge, F.L., Role of ventrolateral medulla m regulation of respiratory and cardiovascular systems, J Appl Phystol, 61 (1986) 1249-1263 [23] Mdner, T.A., Plckel, V.M, Park, D.H., Joh, T.H. and Rels, D J., Phenylethanolamme N-methyltransferase-contammg neurons m the rostral ventrolateral medulla of the rat' I Normal ultrastructure, Bram Res, 411 (1987) 28-45 [24] Mraovltch, S, Iadecola, C, Rugglero, D.A. and Rels, D.J, Widespread reductions in cerebral blood flow and metabohsm ehclted by electrical stimulation of the parabrachial nucleus in rat, Bram Res, 341 (1985) 283-296 [25] Nakal, M., Iadecola, C and Rels, D J., Global cerebral vasodllation by stimulation of rat fastlgial cerebellar nucleus, A m J Phystol, 243 (1982) H226-H235 [26] Nathan, M.A. and Rels, D.J, Hypoxemla, atelectasls, and the elevation of arterial pressure and heart rate in paralyzed artlficmlly ventilated rat, Ltfe Sct, 16 (1975) 1103-1120 [27] Park, D H., Baetge, E E , Kaplan, B B, Albert, V R , Reis, D J and Joh, T H , Different forms of adrenal phenylethanolamme N-methyltransferase' Species-specific posttranslatlonal modification, J Neurochem, 38 (1982) 410-414. [28] Rapela, C.E, Green, H D and Demson, A.B, Baroreceptor reflexes and autoregulatlon of cerebral blood flow m the dog, Ctrc Res, 21 (1967) 559-508 [29] Rewlch, M , Cerebral circulatory responses to respiratory influences In F J Toole, R G Slekert and J P Whlsnant (Eds) Cerebral Vascular Diseases, Grune and Stratton, New York, 1968, pp 91-100

223

[30] Ross, C A , Rugglero, D A , Joh, T H , Park, D H. and Rels, D.J, Rostral ventrolateral medulla selectwe projections to the thoracic autonomic cell column from the region containing C1 adrenahne neurons, J Comp Neurol, 228 (1984) 168-185 [31] Ross, C A , Rugglero, D.A, Park, D H , Joh, T H , Sved, A F , Fernandez-Pardal, J , Saavedra, J.M and Rels, D J , Tonic vasomotor control by the rostral ventrolateral medulla Effect of electrical or chemical stimulation of the area containing C1 adrenahne neurons on arterial pressure, heart rate, and plasma catecholamlnes and vasopressm, J Neurosct, 4 (1984) 474-494 [32] Saekl, Y , Sato, A., Sato, Y. and Trzebski, A , Stimulation of the rostral ventrolateral medullary neurons increases cortical cerebral blood flow via actwatlon of the lntracerebral neural pathway, Neurosct Lett, 107 (1989) 26-32 [33] Savakl, H E., Davldsen, L, Smith, C. and Sokoloff, L, Measurement of free glucose turnover m brain, J Neurochem, 35 2 (1980) 495-502 [34] Shaht, M.N., Relnmuth, O , Shlmojyo, S. and Schemberg, P , Carbon dioxide and cerebral circulatory control III. The effects of brain stem lesions, Arch Neurol, 17 (1967) 342-353 [35] Shmozuka, T., Nemoto, E M and Winter, P M , Mechanisms of cerebrovascular 0 2 sensitw~ty from hyperoxia to moderate hypoxla m the rat, J Cereb Blood Flow Metab, 9 (1989) 187-195 [36] Sokoloff, L., Rewich, M, Kennedy, C, Des Roslers, M H , Patlak, C S, Pettigrew, K.D, Sakurada, D and Shmohara, M , The ~4C-deoxyglucose method for the measurement of local cerebral glucose utilization theory, procedure, and normal values in the conscious albino rat, J Neurochem, 28 (1977) 897-916 [37] Sun, M -K and Rels, D J , Differential responses of barosensitive neurons of rostral ventrolateral medulla to hypoxia m rats, Bram Res, 609 (1993) 333-337 [38] Sun, M -K. and Rels, D J , Extracellular H lontophoresls modities responses to y-amlnobutync acid and cyamde of retJculospmal vasomotor meurons in rats, Eur J Pharmacol, 236 (1993) 305-313. [39] Sun, M - K and Reis, D J , Hypoxia selectively and directly excites vasomotor neurons of rostral ventrolateral medulla m rats, A m J Phystol, in press [40] Sun, M.-K and Reis, D J., Hypoxic excitation of medullary vasomotor neurons m rats are not mediated by glutamate or mtric oxide, Neurosct Lett, 157 (1993) 219-222 [41] Sun, M - K and Spyer, K M , Responses of rostroventrolateral medulla spinal vasomotor neurones to chemoreceptor stlmulat~on in rats, J Auton Nert' Syst, 33 (1991) 79-84 [42] Sun, M K, Jeske, 1 T and Rels, D.J, Cyanide excites medullary sympathoexcltatory neurons m rats, Am J Physlol, 262 (1992) R182-R189 [43] Traystman, R J and Fitzgerald, R.S, Cerebrovascular response to hypoxla an baroreceptor and chemoreceptor-denervated dogs, Am J Phystol, 241 (1981) H724-H731 [44] Traystman, R J , Fitzgerald, R S and Loscutoff, S C, Cerebral c~rculatory responses to arterial hypoxm m normal and chemodenervated dogs, Clrc Res, 42 (1978) 649-657 [45] Underwood, M.D, ladecola, C, Sved, A and ReJs, D.J, Stimulation of C1 area neurons globally increases regional cerebral blood flow but not metabolism, J Cereb Blood Flow Metab, 12 (1992) 844-855