Received 2001-05-09Accepted 2001-07-18

Corresponding author. Tel: 86-311-6044121 ext 5566; Fax: 86-311-6048177; E-mail: syho@hebmu.edu.cn

Acta Physiologica Sinica47

Feb. 2002, 54 (1), 47～54

Research Paper

Effects of intracarotid injection of 17β -estradiol on electrical activity of rostral ventrolateral medullary neurons in male rats

WANG Sheng, HE Rui-Rong*

Department of Physiology, Hebei Medical University, Shijiazhuang 050017

Abstract: The purpose of this study was to determine the effects of 17β-estradiol (E2) on electrical activity of the rostral ventrolateral medulla (RVLM) neurons in rats. Male Sprague-Dawley rats were anesthetized with urethane (1.0 g/kg) and subjected to sino-aortic denervation. Blood pressure, heart rate and spontaneous discharge of RVLM neurons were recorded simultaneously. Intracarotid injection of E2 (10 ng/kg) decreased the discharge rate from 14.46±0.47 to 9.73±0.33 spikes/s (P<0.001) in 25 out of 30 RVLM neurons, while blood pressure and heart rate showed no significant change. The inhibitory effect of E2 on RVLM neuronal activity was rapid at the onset (within 1 min) and long-lasting (>5 min). Prior administration of antiestrogen tamoxifen (TAM) did not affect the effect of E2. However, pretreatment with Nω-nitro-L-arginine-methyl ester (L-NAME), an inhibitor of nitric oxide (NO) synthase, significantly attenuated the inhibitory effect of E2. In addition, NO donor 3-morpholinosydnonimine (SIN-1) potentiated the effect of E2. These results suggest that E2 may inhibit spontaneous electrical activity of RVLM neurons, an effect which is mediated by the activation of NOS with the resultant of NO release via nongenomic actions.

Key words: estrogen; rostral ventrolateral medulla; spontaneous discharge; nitric oxide

颈动脉注射17β -雌二醇对去缓冲神经大鼠延髓腹外侧头端区神经元放电活动的影响

王升, 何瑞荣*

河北医科大学基础医学研究所生理室, 石家庄 050017

摘要: 本研究旨在观察17β-雌二醇(E2)对雄性大鼠延髓腹外侧头端区(RVLM)神经元自发放电活动的影响。在切断双侧缓冲神经的麻醉雄性Sprague-Dawley大鼠上, 同步记录血压、 心率和RVLM神经元的自发放电活动。颈动脉内注射E2 (10 ng/kg), 30个RVLM神经元自发放电单位中有25个单位的放电频率由14.46±0.47降至9.73±0.33 spikes/s (P<0.05), 与此同时血压和心率无明显改变。E2的抑制效应在1 min内起效, 持续时间长于5 min。 雌激素受体拮抗剂tamoxifen (5 mg/kg)不能阻断E2 的抑制效应。预先给予一氧化氮(NO)合酶阻断剂L-NAME (2.7 μg/kg)能明显阻断E2的抑制效应。应用NO供体SIN-1 (0.5 μg/kg)可增强E2的抑制效应。以上结果提示, E2可通过非基因组效应激活RVLM神经元的NOS而引发NO释放, 进而抑制其自发放电活动。

关键词: 雌激素; 延髓腹外侧头端区; 自发放电; 一氧化氮

学科分类号: Q463; R331.3　

The protective effect of estrogen on cardiovascular system received a great deal of attention, particularly, estrogen has been used as a replacement hormone in postmenopausal women to achieve a wide range of health benefits[1]. However, little is known about the effect of estrogen on central autonomic nuclei. There is increasing evidence supporting a role for estrogen as a central modulator of autonomic tone. Previous study[2] showed that intracerebroventricular injection of estrogen increased blood pressure (BP). In anesthetized female rats, intravenous injection of 17β-estradiol inhibited area postrema (AP) neuronal activity without chan-ges in BP and heart rate (HR)[3]. Saleh et al. observed that in anesthetized rats injection of estrogen into the nucleus tractus solitarius (NTS) decreased mean blood pressure, HR and renal sympathetic nerve activity. Injection of estrogen into the rostral ventrolateral medulla (RVLM) also decreased BP and renal sympathetic nerve activity. However, no significant changes in BP and HR were observed following the injection of estrogen into the parabrachial nucleus, central nucleus of the amygdala and insular cortex[4,5]. It is generally accepted that RVLM plays an important role in the regulation of vascular tone and the maintenance of blood pressure. RVLM neurons have been shown to be involved in various sympathetic reflex and integration of the inputs from a variety of visceral, somatic and supra-medullary structures[6,7]. Estrogen receptors have been identified in the rat RVLM and suggested to modulate the activity of RVLM neurons[8,9]. However, the direct effects of estrogen on the electrical activity of RVLM neurons have not yet been well-documented. The purpose of this study was to ascertain the effect of estrogen on the spontaneous electrical activity of RVLM neurons and to elucidate the possible underlying mechanism.

1MATERIALS AND METHODS

1.1 General surgical preparation and procedures Experiments were performed on a total of 35 male SD rats weighing 300～350 g. Rats were anesthetized with urethane (1.0 g/kg i.p.). The experimental procedures and protocols were approved by the Animal Care and Use Committee of Hebei Medical University. The trachea was cannulated for ventilation. The left femoral artery was cannulated, and BP was measured with a pressure transducer (MPU-0.5A, Nihon Kohden) and carrier amplifier (AP-621G, Nihon Kohden). HR was monitored by a HR counter (AT-601G, Nihon Kohden) triggered by differential signals of arterial pressure pulse. The left carotid artery and the femoral vein were cannulated for infusion of drugs. PO2, PCO2 and pH in arterial blood monitored by a blood gas analyzer (ABL-1, Radiometer) were 13.20±0.53 kPa, 4.67±0.27 kPa and 7.38±0.03, respectively.

A midline incision was made on the ventral surface of the neck. After turned rostrally the trachea and esophagus, the bilateral longus capitis muscles were retracted. The basal portion of the occipital bone was carefully cut away to expose the ventral surface of the rostral medulla. After incising the dura and arachnoid, the exposed ventral surface of the medulla was covered with warm (37℃) liquid paraffin. Cerebrospinal fluid was constantly drained with a strip of thinly twisted cotton. Body temperature was maintained in the range of 37～38℃ with a heating lamp. Animals were killed at the end of the experiments by intravenous injection of an overdose of urethane.

1.2 Sino-aortic denervation Carotid sinus areas were fully exposed. Sternohyoideus muscles and superior laryngeal nerves were cut. The aortic depressor nerves and the carotid sinus nerves were sectioned under a dissecting microscope. The superior cervical sympathetic trunk and recurrent laryngeal nerves were also cut. Completeness of barodenervation was ensured by the absence of decrease in HR when hypertension was initiated by i.v. injection of phenylephrine (2～3 μg).

1.3 Extracellular single-unit recordingsThe rat was fixed on a stereotaxic frame (Model 1C Jiangwan, China) in a supine position. Extracellular single-unit recordings were made with a glass microelectrode (impedance: 5～10 MΩ; tip diameter: <1 μm). The microelectrode was filled with 2% pontamine sky and 0.5 mol/L acetate sodium and advanced at a speed of 1 μm/s by a micromanipulator (SM-21, Narishige) into the RVLM (coordinates: 1.5～2.0 mm lateral to the midline, 2.6～3.3 mm caudal to the interaural line, and 0.3～0.5 mm from the ventral surface). The electrical signals were amplified first by a microelectrode amplifier (MEZ-7200, Nihon Kohden) and then by a biophysical amplifier (AB-621G, Nihon Kohden). The amplified bioelectrical signals were recorded along with BP and HR on a polygraph system (RM-6000, Nihon Kohden) with a thermal array recorder (WS-682G, Nihon Kohden; band-pass width: 0～2.8 kHz).

1.4 Experimental protocolsA period of 30 min was allowed for stabilization after the operation, and the microelectrode was inserted into the RVLM to record the spontaneous electrical activity. After a stable recording was obtained for more than 5 min, the drugs were administered. Before the application of drugs, the vehicle (normal saline) was used as control. The experimental animals were divided into the following groups. Group 1: after a stable recording was obtained, in a total of 20 male rats intracarotid injection of 17β-estradiol (10 ng/kg, Sigma) was performed, and the changes in BP, HR and electroneurogram (ENG) of RVLM neurons were examined. An interval of 15 min was allowed for the successive application of E2 in one rat. Group 2: BP, HR and ENG were examined following intracarotid injection of E2 before and after administration of an antiestrogen tamoxifen (5 mg/kg, Sigma). Group 3: BP, HR and ENG were examined following intracarotid injection of E2 before and after administration of Nω-nitro-L-arginine-methyl ester (L-NAME, 2.7 μg/kg, Sigma). Group 4: BP, HR and ENG were examined following intracarotid injection of E2 before and after administration of 3-morpholinosydnonimine (SIN-1, 0.5 μg/kg, Cassella).

1.5 Histological localization of recording sitesThe localization of recorded RVLM neurons was marked by passing direct current (50 μA for 30 s) at the site of the electrode tip. At the end of the experiment, the rat was deeply anesthetized, and the brain stem was quickly removed and fixed in 10% buffered formalin. The brain stem was cut into 40 μm coronal sections on a freezing microtome, mounted on slides, and then stained with cresyl violet. The lesion sites were identified and plotted on the standardized sections from Paxinos and Watson′s atlas[10]. Data were excluded if the recorded neurons were not located in the RVLM area.

1.6 Data analysisAll values are presented as means±SE. Statistical differences were evaluated by paired Student′s t test. Differences between groups were assessed using unpaired t test. Statistical significance was set at P<0.05.

2RESULTS

　Sixty spontaneously active single-units included in the desired area were recorded from 35 male rats. Mean firing rate of neurons was 14.82±0.48 spikes/s. Figure 1 is a composite picture of the locations where the spontaneously active single-units were recorded.

Fig.1. Localization of microelectrode tips in RVLM.Numbers indicate caudal to interaural line in mm. Coronal sections of medulla are modified from Paxinos and Watson. IO, inferior olive; NA, nucleus ambiguus; PY, pyramidal tract; RM, nucleus raphe magnus; RO, nucleus raphe obscurus; RP, nucleus raphe pallidus; closed circle, units responsive to moxonidine; triangle, units not in the nucleus paragigantocellularis lateralis (PGL).

2.1 Effects of E2 on ENG, BP and HR

Intracarotid injection of E2 evoked an immediate decrease in firing rate from 14.46±0.47 to 9.73±0.33 spikes/s (P<0.001) in 25 out of 30 single-units in a total of 20 rats, 3 single-units showed no marked changes and the other 2 were increased, while BP and HR were unaffected. The inhibitory effect of E2 on RVLM neuronal activity was rapid at the onset (within 1 min) and long-lasting (>5 min)(Table 1). When E2 was administered many times in one rat, the same effect was observed. Figure 2 shows the time course of the change in firing rate of RVLM neurons after administration of E2. Intracarotid injection of solvent induced no significant changes in baseline values of these parameters.

Table　１. Changes in MAP, HR and ENG of RVLM neurons in anesthetized male rats following intracarotid injection of E2 (10 ng/kg) (n=25)

(beats/min)〖〗ENG

(spikes/s)Control[]13.17±0.39[]386.60±7.32[]14.46±0.47E2[]12.90±0.45〖〗384.33±8.72〖〗 9.73±0.33***All values are means±SE. MAP, mean arterial pressure; HR, heart rate; ENG, electroneurogram. RVLM, rostral ventrolateral medulla. ***P<0.001, compared with control group.

Fig.2.Time course of the change in firing rate of RVLM neurons after administration of E2. n=25. *P<0.05, **P<0.01, ***P<0.001, vs control.

2.2 Effects of TAM on the actions of E2

　Intracarotid injection of E2 reduced firing rate from 14.82±0.46 to 10.11±0.50 spikes/s (P<0.001) in 10 units in 5 rats. After ENG returned to the control level, intracarotid injection of an antiestrogen TAM (5 mg/kg) did not produce any effects on BP, HR and ENG (data not shown). After 30 min, intracarotid administration of E2 still significantly changed the discharge rate of RVLM neurons from 14.82±0.46 to 9.91±0.39 spikes/s (P<0.001) (Table 2).

2.3 Effects of L-NAME on the actions of E2

　Intracarotid administration of E2 (10 ng/kg) reduced firing rate from 15.13±0.38 to 9.26±0.36 spikes/s (P<0.001) in 10 units in 5 rats. After ENG returned to the control level, intracarotid injection of L-NAME (2.7 μg/kg) per se had no significant effects on ENG, BP and HR within 3 min,while 3 min later intracarotid administration of L-NAME markedly attenuated the inhibitory effect of E2 on RVLM neurons (Fig.3, Table 2).

2.4 Effects of SIN-1 on the actions of E2

　Intracarotid injection of E2 (10 ng/kg) noticeably reduced firing rate from 15.26±0.65 to 10.87±0.73 spikes/s in 10 units in 5 rats (P<0.001). After ENG restored to the control level, intracarotid

Table　2. Effects of pretreatment with TAM (5 mg/kg), L-NAME (2.7 μg/kg) and SIN-1

(0.5 μg/kg ) on respones of MAP, HR and ENG of RVLM to intracarotid injec-

tion of E2 (10 ng/kg)

[]n[]MAP (kPa)[]HR (beats/min)[]ENG (spikes/s)Control[]25[]14.63±0.53[]395.32±8.42[]14.82±0.46E2[]25[]14.23±0.56[]393.11±7.64[]10.11±0.50***TAM+E2[]25[]14.76±0.78[]395.89±9.01[]9.91±0.39***Control[]25[]14.90±0.30[]390.41±7.50[]15.13±0.38E2[]25[]15.16±0.42[]392.21±6.34[]9.26±0.36***L-NAME+E2[]25[]15.03±0.57[]388.33±8.24[]12.53±0.51**## Control[]25[]13.57±0.39[]386.25±7.78[]15.26±0.65E2[]25[] 13.70±0.52[]389.97±7.34[]10.87±0.73***SIN-1+E2[]25[]13.64±0.60[]389.60±8.04[]8.13±0.41***## **P<0.01, ***P<0.001, compared with control groups. ##P<0.01, compared with E2 groups.

Fig.3.Effects of intracarotid injection of E2 (10 ng/kg) on BP, HR and ENG of RVLM neurons before and after L-NAME (2.7 μg/kg) administration. ↓, injection of 17β-estradiol; -〖〗↓, injection of L-NAME.

Fig.4.Changes in BP, HR and ENG of RVLM neurons following intracarotid injection of E2 (10 ng/kg) before and after SIN-1 (0.5 μg/kg) administration. ↓, injection of 17β-estradiol; -〖〗↓, injection of SIN-1.

injection of SIN-1 (0.5 μg/kg) only resulted in a slight decrease in firing rate (data not shown). After 20 s, administration of E2 significantly reduced the discharge rate of RVLM neurons from 15.26±0.65 to 8.13±0.41 spikes/s (P<0.001) (Fig.4, Table 2).

3DISCUSSION

　The results from this study demonstrated that intracarotid injection of E2 inhibited spontaneous electrical activity of RVLM neurons. The inhibitory effect may be ascribed to the direct action of E2 on RVLM neurons. This is based on the following reasons: (1) E2 is a liposoluble substance which can easily cross the blood-brain barrier; (2) the experiments were performed in sinoaortic denervated rats, thereby excluding the indirect effects from baroreflex on RVLM neurons; and (3) blood pressure had no significant change in our experiments, so the effect secondary to the alteration in cerebral blood flow might be ruled out.

　There was evidence suggesting that the overall effect of estrogen within autonomic nuclei is to enhance parasympathetic tone by direct activation of parasympathetic preganglionic neurons located in NTS and nucleus ambiguous or indirectly by an attenuation of sympathetic output at the level of RVLM or spinal cord[4,11]. Saleh et al. have found that estrogen acting within NTS increases parasympathetic output and decreases sympathetic output as well, resulting in a decrease of mean blood pressure, HR and renal sympathetic nerve activities in anesthetized rats[4]. They also reported that direct microinjection of estrogen into RVLM exerted an inhibitory influence on RVLM neurons, which exhibited a decrease in mean arterial pressure and renal sympathetic nerve activity. Also, a decrease in sympathetic tone following estrogen injection into RVLM may be responsible for it[4]. The results from our study indicated that intracarotid injection of E2 reduced the discharge rate of RVLM neurons, while BP and HR showed no significant change. Similarly, Li et al.[3] also observed that intravenous injection of E2 inhibited AP neuronal activity without changes in BP and HR. That the inhibitory effect of E2 on RVLM neurons was not accompanied by a change in BP may be ascribed to the possibility that E2 increased cardiac output in this experiment, counterbalancing the effect on BP of the reduced peripheral resistance. This possibility is supported by the reports that acute treatment with E2 increases cardiac output but decreases peripheral resistance[12]. However, both direct microinjection of estrogen into RVLM and intracarotid injection of estrogen exert an inhibitory influence on RVLM neuronal activity.

　The α and β forms of estrogen receptors have been cloned from human and rats[13,14], which have been identified in a wide range of tissues including the cardiovascular system[15,16] and central nervous system[17]. Multiple brain regions concerning central cardiovascular regulation, including NTS, ventrolateral medulla and AP, contained both forms of estrogen receptor[17]. It is generally accepted that estrogen exerts genomic and nongenomic actions on target tissues. Since the genomic actions of estrogen require a complex cascade of events including hormone-receptor binding, targeted gene expression, and protein synthesis, it may take hours for hormone signal to be translated into membrane excitability changes. In contrast to genomic mechanism, the nongenomic actions are rapid at the onset and do need membrane receptor and/or other cellular second-messenger system rather than nuclear receptor to change neuronal activity. The time course of the response in the present study indicates that the genomic actions are unlikely and nongenomic actions are the reasonable alternative mechanism underlying E2 action in our study. In addition, the antiestrogen tamoxifen, which acts by binding to specific cytosol/nuclear E2 receptors and has been shown to prevent classic E2 effects[18], did not influence the inhibitory effect of E2 on RVLM neurons. Thus, the observed effect of E2 in our experiments was probably mediated through nongenomic mechanism.

　It is well-documented that estrogen may produce nongenomic actions via NO release[19]. Investigation into the effects of estrogen on endothelial, neuronal and inducible nitric oxide synthase (NOS) has shown that E2 significantly enhanced the activity of endothelial NOS (NOS-3) and neuronal NOS (NOS-1), but inhibited that of inducible NOS (NOS-2)[20]. As well, Hayashi et al. observed that estrogen had a biphasic effect on the activity of NOS-1 through the Ca2+-calmodulin. A low concentration of estrogen enhanced the activity of partially purified NOS-1, while a high dose attenuated it[21]. Our previous study also showed that E2 decreased the discharge rate of subfornical organ neurons via the activation of NOS with the resultant production of NO[22]. Therefore, it was tempting to speculate that E2 might inhibit spontaneous electrical activity of RVLM neurons via activation of NOS. In our experiments, the application of L-NAME, an inhibitor of NO synthase, markedly blocked the effect of E2. Furthermore, NO donor SIN-1 could potentiate the inhibitory effect of E2. Such results were accordant with our previous study which demonstrated that NO donor markedly suppressed the spontaneous electrical activity of RVLM neurons[23]. Accordingly, the generation of NO through the result of NOS-1 activation may be responsible for the inhibitory effect of E2 on RVLM neurons. However, it remains unclear whether estrogen activation of NOS depends on binding to the precise location of putative membrane receptor. Recent years have been characterized by a booming interest in research on caveolae, which are cholesterol/sphingolipid-rich microdomains of the plasma membrane and well-known to be enriched in many signaling molecules, including NOS-3[24]. It has been reported that the caveolae were found in neuronal cells[25]. Based on the above observations, it is eligible to speculate that the rapid activation of NOS-1 by E2 might be mediated by ERα localized in caveolae in RVLM neuronal cells.

　In summary, intracarotid injection of E2 decreases spontaneous discharge rate of RVLM neurons, which is due to NO release via nongenomic mechanism.

REFERENCES

[1]Prelevic GM, Beljic T. The effect of estrogen and pro-gestogen replacement therapy on systolic flow velocity in healthy postmenopausal women. Maturitas, 1994,20(1):37～44.

[２]Takahashi H, Ashizawa H, Takeda KU, Yoshimura M, Ijichi H. Central vasopressor responses to conjugated estrogens in rats may be mediated via a rennin-angiotensin system in the brain. J Pharmacol Exp Ther, 1982,222:726～730.

[３]Li Z, Hay M. 17β-estradiol modulation of area postrema potassium currents. J Neurophysiol, 2000,84:1385～1391.

[４]Saleh MC, Connell BJ, Saleh TM. Medullary and intrathecal injection of 17β-estradiol in male rats. Brain Res, 2000,867(1-2):200～209.

[５]Saleh MC, Connell BJ, Saleh TM. Autonomic and cardiovascular reflex responses to central estrogen injection in ovariectomized female rats. Brain Res, 2000,879:105～114.

[６]Terui N, Saeki Y, Kumada M. Barosensory neurons in the ventrolateral medulla in rabbits and their responses to various afferent inputs from peripheral and central sources. Jpn J Physiol, 1986,36:1141～1164.

[７]Verberne AJ, Srornetta RL, Guyenet PG. Properties of C1 and other ventrolateral medullary neurons with hypothalamic projections in the rat. J Physiol (Lond), 1999,517:477～494.

[８]Pelletier G, Liao N, Follea N, Govindan MV. Mapping of estrogen receptor-producing cells in the rat brain by in situ hybridization. Neurosci Lett, 1988,94:23～28.

[９]Simonian SX, Herbison AE. Differential expression of estrogen receptor and neuropeptide Y by brainstem A1 and A2 noradrenaline neurons. Neuroscience, 1997,76:517～529.

[1０]Paxinos G, Watson C. The Rat Brain in Stereotaxic Coordinates. 4th ed. San Diego: Academic Press, 1999.

[１1]Saleh TM, Connell BJ. Centrally mediated effect of 17β-estradiol on parasympathetic tone in male rats. Am J Physiol, 1999,276:R474～R481.

[1２]Magness RR, Rosenfeld CR. Local and systemic estradiol-17β: effects on uterine and systemic vasodilation. Am J Physiol, 1989,256:E536～E542.

[1３]Korke S, Sakai M, Muramatsu M. Molecular cloning and characterization of rat estrogen receptor. Nucleic Acids Res, 1987,15:2499～2513.

[1４]Kuiper GG, Enmark E, Peltor-Huikko M, Nilsson S, Gustafsson JA. Cloning of a novel receptor expressed in rat prostate and ovary. Proc Natl Acad Sci, 1997,138:863～870.

[1５]Grohe C, Kahlert S, Lobbert K, Vetter H. Expression of oestrogen receptor alpha and beta in rat heart: role of local oestrogen synthesis. J Endocrinol, 1998,156:R1～R7.

[1６]Karas RH, Patterson BL, Mendelsohn ME. Human vascular smooth muscle cells contain functional estrogen receptor. Circulation, 1994,89:1943～1950.

[1７]Shughrue PJ, Lane MV, Merchenthaler I. Comparative distribution of estrogen receptor-alpha and -beta mRNA in the rat central nervous system. J Comp Neurol, 1997,388:507～525.

[1８]Katzenellenbogen BS, Miller MA, Eckert RL, Sugo K. Antiestrogen pharmacology and mechanism of action. J Steroid Biochem, 1983,19:59～68.

[1９]Kim HP, Lee JY, Jeong JK, Bae SW, Lee HK, Jo I. Nongenomic stimulation of nitric oxide release by estrogen is mediated by estrogen receptor α localized in caveolae. Biochem Biophys Res Commun, 1999,263:257～262.

[２０]Hayashi T, Yamada K, Esaki T, Mutoh E, Iguchi A. Effect of estrogen on isoforms of nitric oxide synthase: possible mechanism of antiatherosclerotic effect of estrogen. Gerontology, 1997,43:24～34.

[２1]Hayashi T, Ishikawa T, Yamada K, Kuzuya M, Naito M, Hidaka H, Iguchi A. Biphasic effect of estrogen on neuronal constitutive nitric oxide synthase via Ca2+-calmodulin dependent mechanism. Biochem Biophys Res Commun, 1994,203:1013～1019.

[２２]Wang H (王 红), Wang ZA (王志安), He RR (何瑞荣). Modulatory effects of 17β-estradiol on electrical activity of subfornical organ neurons. Acta Physiol Sin (生理学报), 2000,52(6):515～518.

[２３]Yang LL (杨丽明), He RR (何瑞荣). Effects of intravenous injection of N-nitro-L-arginine and NO donors on spontaneous activity of rostral ventrolateral medulla neurons. Acta Physiol Sin (生理学报), 1996,48(4):320～328.

[２４]Shaul PW, Anderson RG. Role of plasmalemmal caveolae in signal transduction. Am J Physiol, 1998,275:L843～L851.

[２５]Wu C, Butz S, Ying Y, Anderson RG. Tyrosine kinase receptors concentrated in caveolae-like domains from neuronal plasma membrane. J Biol Chem, 1997,272:3554～3559.