Localization of sympathetic postganglionic neurons innervating mesenteric artery and vein in rats

Journal of the Autonomic Nervous System 80 (2000) 1–7 www.elsevier.com / locate / jans

Localization of sympathetic postganglionic neurons innervating mesenteric artery and vein in rats a ,b c a, Nan K. Hsieh , Jiang C. Liu , Hsing I. Chen * a

Institute of Medical Science and Department of Physiology, Tzu Chi College of Medicine and Humanities, Hualien, Taiwan b Graduate Institute of Medical Sciences, Tzu Chi College of Medicine and Humanities, Hualien, Taiwan c Department of Anatomy, National Defense Medical Center, Taipei, Taiwan Received 2 July 1999; received in revised form 22 September 1999; accepted 22 September 1999

Abstract Physiological and histochemical studies have demonstrated the control and innervation of sympathetic nerves to the artery and vein vessels of splanchnic circulation. In our laboratory, we first used the technique of retrograde transport of horseradish peroxidase to identify the origin of sympathetic neurons innervating the mesenteric vein. In this study, double fluorescence staining technique was used for a simultaneous localization of the sympathetic postganglionic neurons supplying the mesenteric artery and vein in rats. First-order branches of mesenteric artery (A) and vein (V) in the vicinity of ileo-cecal junction were isolated for application of fluorescent dyes (Fast Blue, FB and Diamidino Yellow, DY). The application of FB and DY on A and V was alternated in the next animal to minimize the difference in dye uptake. The animal was allowed to recover for 6–7 days assuring a complete uptake of FB and DY into the cytoplasm and nucleus, respectively. The number of FB, DY and double staining neurons in the prevertebral and paravertebral ganglia were counted under a fluorescent microscope after animal fixation and serial frozen section (30 mm) of the sympathetic ganglia. Our study revealed the following findings: (1) Distribution of the fluorescence-staining neurons in the sympathetic ganglia was as follows: right celiac ganglion (39%), superior mesenteric ganglion (30%), left celiac ganglion (26%), inferior mesenteric ganglion (1%) and paravertebral ganglia (4%). (2) Double staining neurons that dually innervate A and V amounted to 54% of total staining neurons. There were 41% neurons singly innervating A and 5% innervating V. (3) The ratio of neurons supplying the A and V ranged from 1.41 to 1.75 (average 1.61). (4) There was no distinct topographical distribution with respect to the neuron location innervating A and V. The distribution of neurons appeared in a scattering pattern.  2000 Elsevier Science B.V. All rights reserved. Keywords: Sympathetic innervation; Double fluorescence; Retrograde transport; Mesenteric vessels; Artery; Vein

1. Introduction Sympathetic control of arterial and venous vessels is physiologically important in the regulation of circulatory dynamics such as pressure, flow, resistance, impedance and capacitance etc. (Mellander, 1960; Brooksby and Donald, 1971; Shepherd and Vanhoutte, 1975; Rothe, 1983; Chen and Wang, 1984; Chen et al., 1991). Anatomically, histochemical studies with fluorometric method have demonstrated the existence of adrenergic nerve endings around artery and vein vessels in various beds (Mayer et al., 1968; Furness, 1973; Furness and Marshall, 1974; Nilsson et al., 1986). Furthermore, retrograde tracing studies using horseradish peroxidase (HRP) or fluores*Corresponding author. Tel.: 1886-3-8560-824; fax: 1886-3-8573053. E-mail address: [email protected] (H.I. Chen)

cence dyes have been carried out for the uptake from nerve endings to localize the sympathetic postganglionic neurons that innervate the mesenteric and cerebral arteries (Tsai et al., 1985; Hill et al., 1987). With respect to the sympathetic innervation to the vein vessels, our laboratory has reported several studies concerning the origins of ganglionic neurons innervating the veins in skin, skeletal muscle and mesentery (Chen and Liu, 1993; Chen and Chi, 1994; Chen et al., 1994). In the autonomic nervous control, a few studies address these questions: Do artery and vein share the same neurons in sympathetic ganglia? Furthermore, do these neurons have specific sites of distribution? Dehal et al. (1992) used double fluorescence (Fast Blue and Fluoro-Gold) technique for a simultaneous localization of the postganglionic neurons innervating artery and vein of the rat’s hindlimb. They found that limb vein and artery were innervated by separate neurons. In addition, the neurons innervating

0165-1838 / 00 / $ – see front matter  2000 Elsevier Science B.V. All rights reserved. PII: S0165-1838( 99 )00070-3

2

N.K. Hsieh et al. / Journal of the Autonomic Nervous System 80 (2000) 1 – 7

artery and vein were not segregated in distinct group; they were intermixed randomly in the ganglia of sympathetic chain. Whether this pattern of sympathetic innervation to artery and vein occurs in other vascular bed is unknown. Sympathetic effects and adrenergic control on the mesenteric bed are more prominent than the cutaneous and muscle vessels (Fuxe and Sedvall, 1965; Furness and Marshall, 1974; Shepherd and Vanhoutte, 1975; Nilsson, 1985; Chen and Chi, 1994; Chen et al., 1994). Localization of the postganglionic neurons innervating the mesenteric vein has recently been accomplished in our laboratory (Chen et al., 1994). However, a simultaneous localization of the postganglionic neurons innervating the mesenteric artery and vein has not been reported. In the present study, we also used the technique of retrograde transport of double fluorescence (Fast Blue and Diamidino Yellow) for tracing the origins of sympathetic neurons that innervate the mesenteric artery and vein in rats.

2. Materials and methods

2.1. Animals Adult rats of Sprague–Dawley (SD) strain, of either sex, weighing 280–390 g were used. The rats were housed in a standard 12 h on / off light cycle with ad libitum access to food and water. Ten days before the surgical preparation, the rats were fed with cabbage instead of chew food in order to reduce the omentum fat. After the surgical preparation, they were allowed to gain access to the original food. Amoxicillin, 500 mg was added to every 100 ml drinking water for 5 days.

2.2. Preparation For an implantation of fluorescence dyes around the mesenteric artery and vein, the rat was anesthetized with an intraperitoneal injection of sodium pentobarbital, 40 mg / kg. The abdomen was opened with a midline incision. The omentum and mesentery were externalized and wrapped with surgical gauze moisturized with warm saline solution. The superior mesenteric artery and vein were identified at their junction with aorta and inferior vena cava. First-order branches of mesenteric artery and vein in the vicinity of ileo-cecal junction were isolated for the application of fluorescence dyes (Fig. 1).

2.3. Double fluorescence technique Double fluorescence technique was used to trace the neurons innervating the mesenteric artery and vein from the prevertebral and paravertebral ganglia. Fast Blue (FB) and Diamidino Yellow (DY) have been shown to be ideal retrograde tracers for double labeling because: (1) FB and DY are uptaken by neuron into the cytoplasm and nucleus, respectively. Under fluorescent microscope at the same

Fig. 1. Diagram showing the location (asteriks) where fluorescence dyes were applied outside the mesenteric artery and vein segments. Fast Blue (FB) was applied on the artery segment and Diamidino Yellow (DY) on the vein segment in four rats. In the other four rats, FB was applied on vein segment and DY on artery segment.

excitation wave length (360 nm), FB produces a blue labeling of cytoplasm, while DY a yellow labeling of nucleus (Kuypers et al., 1980; Keizer et al., 1983; Kuypers and Huisman, 1984). The distinct feature of uptake and staining is an useful guideline to identify single- and double-labeling neurons. (2) Two fluorescent dyes have approximately the same uptake and survival time (Kuypers et al., 1980; Keizer et al., 1983; Kuypers and Huisman, 1984; Hill et al., 1987). Hill et al. (1987) applied FB into rat’s mesenteric artery and found that the maximal uptake of FB into the prevertebral ganglia occurred at 6–7 days after tracer application. In the present experiment, the same period of time (6–7 days) was allowed for the uptake of fluorescence dyes from the mesenteric artery and vein segments to sympathetic ganglia. On the artery and vein branches, a segment (15 mm) was completely isolated from the surrounding tissues. The outer vessel wall was torn off by fine forceps as much as possible to facilitate dye uptake by nerve endings (Yates et al., 1987; Chen and Chi, 1994). The vessel segments were enclosed in a silicone tube (10 mm in length) with a longitudinal slit. A piece of gelform pledget was carefully inserted into the silicon tube and outside the vessel wall. FB or DY solution (5%, 0.4–0.5 ml) was slowly injected into the gelform pledget until the whole piece was impregnated with the dye solution. The fluorescence dyes (FB and DY, Sigma) were dissolved in saline solution. The solution was shaken well immediately before use. Stock solution was stored at 48C in a dark tube and covered with aluminum wrap to keep it from light exposure. The solution was discarded if stocked without use over two weeks. To minimize the possibility of variation in dye uptake due to different dyes, FB was applied to artery and DY to vein in one animal. Alternatively, FB was applied to vein and DY to artery in the next one. After the dye application, a small piece of thin plate was used to separate the artery

N.K. Hsieh et al. / Journal of the Autonomic Nervous System 80 (2000) 1 – 7

and vein segments. The whole vessel segments with silicone tubes were sealed with petroleum jelly to avoid the possibility of dye spreading. The abdomen was closed and the animal allowed to recover.

2.4. Animal fixation, tissue section and neuron counting Six to seven days after the application of fluorescence dyes, the animal was anesthetized again with intraperitoneal injection of sodium pentobarbital, 50 mg / kg. Heparin (1000 I.U.) was given intravenously to prevent coagulation. The animal was then sacrificed by an overdose of pentobarbital. The chest was opened to expose the heart. The right atrium was cut to bleed the animal and the left ventricle was punctured to infuse saline solution 300– 400 ml followed by fixative (300 ml of 1.25% glutaraldehyde and 1% paraformaldehyde in 0.1 M phosphate buffer, pH 7.4). The abdomen and chest were opened to expose the pre- and paravertebral sympathetic ganglia. The prevertebral ganglia including right celiac ganglion (RCG), left celiac ganglion (LCG), superior mesenteric ganglion (SMG) and inferior mesenteric ganglion (IMG) were excised from the ventral aspect of abdominal aorta. The paravertebral ganglia from T 10 to L 3 were removed along both sides of the spine. The tissues were stored in sucrosebuffer solution at 48C, and fixed in place by small needles on a polyvinyl plate to avoid loss of orientation. The whole container was covered by aluminum wrap for light-proofing. A cryostat microtome was used for frozen serial section (30 mm). To avoid fading of fluorescence, the procedures of section, microscope observation and counting were accomplished at a time for only one ganglion. The other tissues were stored in a freezer and remained unexposed before frozen section. The tissue sections were observed under a fluorescent microscope (Nikon, Microphot-FAA). The excitation length for ultraviolet light was 360 nm. The excitation length allowed visualization of FB and DY (Kuypers et al., 1980; Keizer et al., 1983). Photomicrography was selected and taken with ASA 400 film (Kodak Ektachrome). Since FB stains the cytoplasm with blue background and DY stains the nucleus with yellow color (Kuypers et al., 1980; Keizer et al., 1983; Kuypers and Huisman, 1984), the single- and double-staining neurons were easily identified. Some neurons were relatively thick or thin and some were not typically stained, these neurons were not counted. Neuron counts were corrected for double counting using the Abercrombie equation (Abercrombie, 1946).

3

of the postganglionic neurons stained with fluorescence was done after serial section of the sympathetic ganglia. Under the fluorescence microscope, four types of neurons were identified (Fig. 2): (a) FB and DY double-labeling neurons. These neurons dually innervate the mesenteric A and V, (b) FB-labeling neurons, (c) DY-labeling neurons and (d) blank neurons without fluorescence staining. Neurons that were singly stained with FB or DY (b and c) indicate the neurons that singly innervate the A or V. The blank neurons were possibly supporting, connecting interneurons or neurons supplying the other splanchnic arterial and venous vessels.

3.2. Distribution of labeling neurons in sympathetic ganglia Table 1 shows the number of fluorescence-labeling neurons in prevertebral and paravertebral ganglia of each rat. In 8 rats, FB was applied on A and DY on V in rats no. 1, 3, 5 and 7. Alternatively, DY was applied on A and FB on V in rats no. 2, 4, 6 and 8. There was certain variation in the number of labeling neurons among individual rats. However, it was consistent that the number of doublelabeling neurons dually innervating A and V (denoted DIN) was higher than those singly innervating A (denoted SIN-A), and the latter was also higher than those singly innervating V (denoted SIN-V). Abundant labeling neurons could be found in the right celiac ganglion (RCG), left celiac ganglion (LCG) and superior mesenteric ganglion (SMG). Some labeling neurons were observed in the inferior mesenteric ganglion (IMG) and paravertebral ganglia (PVG, T 10 |L 3 ). In each ganglion, the distribution of SIN and DIN appeared in a scattering pattern. There was no specific site for the location of SIN-A, SIN-V or DIN. The results were quite different from those obtained in the hindlimb. We found that the mesenteric beds, artery

3. Results

3.1. Neuron labeling In a total of 8 rats with application of double fluorescence on mesenteric artery (A) and vein (V), observation

Fig. 2. A microphotography showing ganglionic neurons labeling with fluorescence. Double labeling neurons (double arrowhead), FB-labeling neurons (arrowhead) and DY-labeling neurons (small arrow) are intermixed within the ganglion. Staining neurons without a tag are considered atypical and are not counted. Calibration bar equals 100 mm.

N.K. Hsieh et al. / Journal of the Autonomic Nervous System 80 (2000) 1 – 7

4

Table 1 Distribution of fluorescence-labeling neurons in prevertebral and paravertebral ganglia of each rat Right celiac G.

Left celiac G.

Superior mesenteric G.

Inferior mesenteric G.

Paravertebral G. T 10 |L 3

SIN A

SIN A

SIN V

DIN A1V

14 6 12 5 6 4 16 7 8 2 4 1 16 6 12 8

5 2 6 4 4 2 4 2 6 0 2 0 5 1 4 2

34 20 28 18 11 6 38 21 12 4 8 4 42 22 28 19

SIN A

SIN V

DIN A1V

SIN A

SIN V

DIN A1V

SIN A

SIN V

DIN A1V

SIN V

DIN A1V

Rat No. 1

285

41

374

178

26

231

227

33

293

3

0

5

2

386

43

583

295

34

325

277

13

345

6

1

7

3

377

27

391

228

12

256

270

40

275

0

0

0

4

223

56

318

169

44

242

142

8

210

7

3

11

5

237

34

356

155

16

238

191

29

286

5

1

9

6

276

36

365

182

17

212

235

47

317

4

0

1

7

415

59

542

393

42

478

354

26

473

10

4

12

8

369

36

417

239

29

275

305

39

334

14

3

18

Rt Lt Rt Lt Rt Lt Rt Lt Rt Lt Rt Lt Rt Lt Rt Lt

G.5ganglion; SIN5single-innervating neuron; DIN5dual-innervating neuron; A5artery; V5vein.

and vein were supplied in part by the same neurons. There were additional neurons innervating the artery and vein alone. Similar to the findings of previous studies (Hill et al., 1987; Dehal et al., 1992; Chen et al., 1994), the distribution of vasomotor neurons to the artery and vein appeared to be scattered in pattern. The neuron groups were not segregated in a distinct site.

3.3. Analysis of the distribution The average number of neurons (SIN-A, SIN-V and DIN) in RCG, LCG, SMG, IMG and PVG are shown in Table 2. The data revealed that RCG had the highest labeling neurons (39%). SMG and LCG contained about the same percent (27–30%) of total labeling neurons. Small number of labeling neurons (1–2%) were observed in the IMG and PVG. In PVG, most of the labeling Table 2 The average number of neurons in sympathetic ganglia Ganglia

SIN A

SIN V

DIN A1V

RCG LCG SMG IMG PVG Rt Lt

321626 (41%) 224631 (42%) 250623 (42%) 662 (37%)

4264 (5%) 2864 (5%) 2965 (5%) 261 (13%)

418633 (54%) 282630 (53%) 317627 (53%) 862 (50%)

1062 (25%) 561 (24%)

561 (12%) 261 (10%)

2565 (63%) 1463 (66%) Total

Sum 781 [39%] 534 [27%] 596 [30%] 16 [1%] 40 [2%] 21 [1%] 1988

Values are mean6SE (n58). RCG5right celiac ganglion; LCG5left celiac ganglion; SMG5superior mesenteric ganglion; IMG5inferior mesenteric ganglion; PVG5paravertebral ganglia (T 10 |L 3 ). The other abbreviations are the same as Table 1. The figures in () are percent of SIN and DIN in each ganglion; the figures in [] are percent of neuron distribution (sum / total) in each ganglion.

neurons were found in T 13 and L 1 . In addition, the number of labeling neurons in the right sympathetic chain appeared to be slightly higher than those in the left side. This table also shows the percent of SIN-A, SIN-V and DIN in each ganglion. The double-labeling neurons (DIN) were higher than the SIN-A. In RCG, LCG and SMG in which abundant neurons were observed, SIN-V amounted to only 5%. In IMG and PVG, there were approximately 10–12% of neurons that singly innervate the vein (SIN-V). In this experiment, we obtained 15904 fluorescencestaining neurons in 8 rats (average 1988 / rat). To get an idea how much was the proportion of these staining neurons in the total population of ganglionic neurons, photographs of all serial sections were taken in two rats. The total ganglionic neurons were tediously counted and corrected by Abercrombie equation (Abercrombie, 1946). The number of total neurons in the sympathetic ganglia of two rats amounted to 25,683 and 27,894 (an average of 26,789). Although the total neuron counting might not be very precise, the figures gave an estimation that the fluorescence containing neurons approximated 7.4% of the total ganglionic neurons. Fig. 3 schematically illustrates the distribution of double-labeling neurons in prevertebral and paravertebral ganglia. In a total of 1064 DINs (average from 8 rats), the RCG contained 39%; LCG 26% and SMG 30%. Approximately 1% of the total DINs were found in the IMG and 4% in the PVG (Rt plus Lt side). In Table 3, the total number of SIN-A, SIN-V and DIN were counted for each rat to show the percent of neurons that singly innervate the A or V (SIN-A or SIN-V) and that dually innervate the A and V (DIN). It was quite interesting that the percentage for SIN-A, SIN-V and DIN was rather close among these rats. Overall, the artery and vein

N.K. Hsieh et al. / Journal of the Autonomic Nervous System 80 (2000) 1 – 7

5

41%) of the labeling neurons innervate the artery and 59% (54%15%) innervate the vein. The ratio of arterial to venous neurons averaged 1.61 (range 1.41 to 1.75).

4. Discussion

Fig. 3. Distribution of double-labeling neurons in the sympathetic ganglia. Table 3 The number and percent of SIN-A, SIN-V and DIN in each rat Rat No.

SIN A

SIN V

DIN A1V

1 2 3 4 5 6 7 8 Mean6SE

713 (40%) 981 (41%) 885 (46%) 514 (35%) 598 (38%) 702 (41%) 1194 (41%) 941 (44%) 816679 (4161%)

107 (6%) 101 (4%) 85 (5%) 117 (8%) 86 (5%) 102 (6%) 137 (5%) 113 (5%) 10666 (561%)

957 (54%) 1306 (55%) 939 (49%) 840 (57%) 905 (57%) 907 (53%) 1569 (54%) 1091 (51%) 1064689 (5461%)

Abbreviations are the same as Table 1.

share 54% of the same postganglionic neurons. A total of 41% staining neurons supply the artery alone, and 5% singly supply the vein (Fig. 4). Accordingly, 95% (54%1

Fig. 4. A schematic representation of the proportion of total labeling neurons that innervate the mesenteric artery and vein.

Substantial evidence indicates that sympathetic stimulation causes constriction of splanchnic arterial and venous vessels (Brooksby and Donald, 1971; Furness and Marshall, 1974; Karim and Hainsworth, 1976; Greenway, 1983; Ozono et al., 1989). Histochemical studies have demonstrated that the mesenteric arteries and veins are all innervated by adrenergic nerves with higher density in the arterial than venous vessels (Furness, 1973; Furness and Marshall, 1974). Hill et al. (1987) first used microinjection of Fast Blue into the mesenteric arterial wall for a retrograde tracing of the postganglionic neurons supplying the mesenteric artery in rats. Our laboratory (Chen et al., 1994) recently used retrograde transport of horseradish peroxidase (HRP) to localize the postganglionic neurons that innervate the mesenteric vein in cats. In the present study, we further employed the double-fluorescence tracing technique for a simultaneous localization of the sympathetic postganglionic neurons innervating the superior mesenteric artery and vein. The major findings were as follows: (1) In total fluorescence-staining neurons, 54% of these ganglionic neurons coinnervate the artery and vein. Additional neurons that singly innervate the artery and vein amount to 41% and 5%, respectively. (2) The fluorescence-labeling neurons are distributed as follows: right celiac ganglion 39%, superior mesenteric ganglion 30%, left celiac ganglion 26%, inferior mesenteric ganglion 1% and paravertebral ganglia approximately 4%. (3) In each ganglion, the distribution of double- and singlelabeling neurons appears in a scattered pattern. There are no specific sites for the postganglionic neurons innervating the artery and / or vein. Our findings that mesenteric artery and vein were innervated by common and separate neurons from the sympathetic ganglia were different from the results obtained in the rat’s hindlimb (Dehal et al., 1992). They applied Fast Blue to a femoral vein segment and FluoroGold to a femoral artery segment. The dyes were then blotted away 60 min after the application. Observation of the fluorescence-staining neurons in sympathetic ganglia was performed seven days after the dye application. In six rats, a total of 4405 fluorescence containing neurons (average 734 / rat) was counted in paravertebral ganglia from T 13 to L 6 . They were all single-labeling neurons. No double-labeling neurons were found, indicating the femoral artery and vein were separately innervated by different groups of neurons in the sympathetic ganglia. Among the total neurons, 3359 (average 560 / rat) and 1046 (average 174 / rat) neurons supply the femoral artery and vein, respectively. The arterial and venous neurons were ran-

6

N.K. Hsieh et al. / Journal of the Autonomic Nervous System 80 (2000) 1 – 7

domly intermixed; no topographical distribution within the ganglia was observed. In comparison with our findings in the mesenteric bed, we counted a total of 15,904 fluorescence-containing neurons in 8 rats (average 1988 / rat). The number was 2.7 times the neurons number obtained in rat’s limb. The results agree with the findings that adrenergic nerve endings are richer in mesenteric than limb vessels (Fuxe and Sedvall, 1965; Furness and Marshall, 1974; Shepherd and Vanhoutte, 1975). It is also a general consensus that capacity control of the limb veins depends mainly on the venous valves and muscle pump, while sympathetic venoconstriction is the major mechanism for the splanchnic beds (Shepherd and Vanhoutte, 1975; Chen et al., 1979; Rothe, 1983). However, the discrepancy in the postganglionic innervation to the artery and vein between limb and mesenteric vessels is difficult to explain. Whether the artery and vein share the same postganglionic neurons or they use separate neurons in other vascular beds deserves further investigation. The major problem of retrograde tracing using horseradish peroxidase or fluorescence dyes is the possibility of contamination and uptake by other nerve endings. Care was taken in minimizing fluorescence spreading as described in Section 2. In addition, FB and DY which were applied on the artery and vein segment were switched in the next rat. Our results also strongly indicate that crossover contamination between artery and vein and uptake by enteric nerves are not likely. Although there was variation in the number of staining neurons among individual animals, it was consistent that DIN was more numerous than SIN-A and the latter was higher than SIN-V in every rat (Table 1). In addition, the variation in the percentage of DIN, SIN-A and SIN-V was very small among all animals (Table 3). If countamination had occurred, there would be no consistency in the neuron proportion. In this study, we also estimated the total ganglionic neurons in two rats and obtained an average number of 26,789. The total fluorescence-staining neurons were averaged 1988 per rat. The figure was 7.4% of the total ganglionic neurons. It should be noted that the area of fluorescence application on arterial and venous segments might be far less than 1% of the total area of mesenteric bed. At first sight, one may consider that we have overestimated the fluorescence containing neurons. However, the sympathetic innervations to effector organs are usually diffuse and are not operated ¨ in a one-on-one pattern (Lefkowitz et al., 1990; Janig and McLachlan, 1992). Divergence possibly occurs from one neuron to several vessel beds. Although this possibilty deserves further investigation, the diffuse pattern of sympathetic innervation gives an explanation of the relatively large number of fluorescence-labeling neurons obtained in the present experiment. In a previous study (Chen et al., 1994), we used horseradish peroxidase (HRP) to localize the postganglionic neurons innervating the mesenteric vein. The HRP was applied on a vein segment (6–8 mm)

near the duodeno-jejunal junction. We obtained an average number of 928 venous neurons. If we take 59% (54%1 5%) as the percentage of venous neurons over total innervating neurons, it gives rise to a total number of 1573. The figure is only slightly lower than that (1988) obtained in the present study. However, it should be noted that the length of vein segment was shorter than that in the present study (6–8 mm vs. 10 mm). In addition, the location and dye application were also different. Nevertheless, a neuron number over 1500 and approximately 2000 is likely a reasonable value for retrograde tracing by dye application in mesenteric artery and vein segments of about 10 mm in length. The physiological significance of this pattern of postganglionic innervation to mesenteric artery and vein is not clear. There were detailed studies on the effects of sympathetic stimulation and norepinephrine application on the constrictor responses of various portions of mesenteric artery and vein vessels (Furness and Marshall, 1974; Nilsson et al., 1986). These studies revealed that the frequency needed for half-maximal constriction in the mesenteric artery (4.5 Hz) was slightly higher than that in the mesenteric vein (3.4 Hz). The findings do not provide a good explanation why the artery and vein share some neurons in common and use part of the separate neurons in the sympathetic ganglia. Further studies may be carried out to determine whether central and reflex sympathetic drives activate the common neuronal pool for a simultaneous constriction of artery and vein. Additional arterial and / or venous constriction will then recruit the separate neuron population. From the data of this study, the artery and vein shared 54% of the staining neurons; the artery had another 41% and the vein had 5% for separate innervation. Accordingly, the ratio of artery-to-vein prostganglionic innervation averaged 1.61 [(54141) /(5415)]. Furness and Marshall (1974) found that the maximal arterial constriction in response to perivascular nerve stimulation and norepinephrine was about 2.47 times the vasoconstrictor response of the vein. Our ratio of 1.61 for the ganglionic neurons appeared to be lower than the ratio of response. However, the involvement of several factors such as vessel size, neurotransmitter release and receptor binding in the stimulus–response process should be considered for a comparison. Nevertheless, our results provide an anatomical localization of the postganglionic neurons innervating the mesenteric artery and vein, and agree that arterial neurons are more numerous than venous neurons.

Acknowledgements The authors are grateful to Drs. J.Y. Wang and S.T. Wang for their technical assistance. This study is supported by research grant (TCMRC 8507) from the Tzu Chi Charity Foundation and Outstanding Scholarship Foundation.

N.K. Hsieh et al. / Journal of the Autonomic Nervous System 80 (2000) 1 – 7

References Abercrombie, M., 1946. Estimation of nuclear population from microtome sections. Anat. Rec. 94, 239–247. Brooksby, G.A., Donald, D.E., 1971. Dynamic changes in splanchnic blood flow and blood volume in dogs during activation of sympathetic nerves. Circ. Res. 29, 227–228. Chen, H.I., Wang, D.J., 1984. Systemic and pulmonary hemodynamic responses to intracranial hypertension. Am. J. Physiol. 247, H715– H721. Chen, H.I., Liu, J.C., 1993. Localization of sympathetic postganglionic neurons innervating the femoral-saphenous vein in cats. Neurosci. Lett. 155, 140–143. Chen, H.I., Chi, T., 1994. The sympathetic efferent innervation of the cutaneous and muscle veins in cats — a comparative study using retrograde localization with horseradish peroxidase. J. Auton. Nerv. Syst. 46, 189–197. Chen, H.I., Chen, S.J., Chen, H.C., 1979. Brain stem control on the vascular resistance and capacitance in the feline and canine hindquarters. Proc. Natl. Sci. Counc. (ROC) 3, 178–186. Chen, H.I., Chang, K.C., Hsieh, K.S., 1991. Vascular factors in isovolumic systemic and pulmonary circuit. Am. J. Physiol. 260, H201–H209. Chen, H.I., Kang, B.H., Wang, J.Y., Tseng, Y.L., 1994. Origins of sympathetic innervation to the mesenteric vein vessel in cats. Neurosci. Lett. 173, 181–184. Dehal, N.S., Kartseva, A., Weaver, L.C., 1992. Comparison of locations and peptide content of postganglionic neurons innervating veins and arteries of the rat hindlimb. J. Auton. Nerv. Syst. 39, 61–72. Furness, J.B., 1973. Arrangement of blood vessels and their relation with adrenergic nerves in the rat mesentery. J. Anat. 115, 347–364. Furness, J.B., Marshall, J.M., 1974. Correlation of the directly observed responses of mesenteric vessels of the rat to nerve stimulation and noradrenaline with the distribution of adrenergic nerves. J. Physiol. (London) 239, 75–88. Fuxe, K., Sedvall, G., 1965. The distribution of adrenergic nerve fibers to the blood vessels in skeletal muscule. Acta Physiol. Scand. 64, 75–82. Greenway, C.V., 1983. Role of splanchnic venous system in overall cardiovascular homeostasis. Fed. Proc. 42, 1678–1684. Hill, C.E., Hendry, I.A., Sheppard, A., 1987. Use of the fluorescent dye, Fast Blue, to label sympathetic postganglionic neurons of the rat. J. Auton. Nerv. Syst. 18, 73–82. ¨ Janig, W., McLachlan, E.M., 1992. Characteristics of function-specific pathways in the sympathetic nervous system. Trends Neurosci. 15, 475–481. Karim, F., Hainsworth, R., 1976. Responses of abdominal vascular

7

capacitance to stimulation of splanchnic nerves. Am. J. Physiol. 231, 434–440. Keizer, K., Kuypers, H.G.J.M., Huisman, A.M., Dann, O., 1983. Diamidino yellow dihydrochloride (DY ? 2HCl), a new fluorescent retrograde neuronal tracer, which migrates only very slowly out of the cell. Exp. Brain Res. 51, 179–191. Kuypers, H.G.J.M., Huisman, A.M., 1984. Fluorescent neuronal tracers. Advances in cellular neurobiology, Vol 5. In: Federoff, S. (Ed.), Labeling Methods Applicable to the Study of Neuronal Pathways. Academic Press, New York, pp. 307–340. Kuypers, H.G.J.M., Bentivoglio, M., Catsman-Berrevoets, C.E., Bharos, A.T., 1980. Double retrograde neuronal labeling through divergent axon collaterals, using two fluorescent tracers with the same excitation wavelength which label different features of the cell. Exp. Brain Res. 40, 383–392. Lefkowitz, R.J., Hoffman, B.B., Taylor, P., 1990. Neurohumoral transmission: The autonomic and somatic nervous systems. In: Gilman, A.G., et al. (Eds.), The Pharmacological Basis of Therapeutics. Pergamon, New York, pp. 84–121. Mayer, H.E., Abboud, F.M., Ballard, D.R., Eckstein, J.W., 1968. Catecholamines in arteries and veins of the foreleg of the dog. Circ. Res. 23, 653–661. Mellander, S., 1960. Comparative studies on the adrenergic neurohormonal control of resistance and capacitance blood vessels in the cat. Acta Physiol. Scand. 50 (Suppl 176), 1–86. Nilsson, H., 1985. Adrenergic nervous control of resistance and capacitance vessels: studies on isolated blood vessels from the rat. Acta Physiol. Scand. 124 (Suppl 541), 1–34. Nilsson, H., Goldstein, M., Nilsson, O., 1986. Adrenergic innervation and neurogenic response in large and small arteries and veins from the rat. Acta Physiol. Scand. 126, 121–133. Ozono, K., Bornjak, Z.J., Kampine, J.P., 1989. Reflex control of mesenteric vein diameter and pressure in situ in rabbits. Am. J. Physiol. 256, H1066–H1072. Rothe, C.F., 1983. Reflex control of veins and vascular capacitance. Physiol. Rev. 63, 1284–1342. Shepherd, J.T., Vanhoutte, P.M., 1975. Veins and their control. Saunders WB, London. Tsai, S.H., Lin, S.Z., Wang, S.D., Liu, J.C., Shih, C.J., 1985. Retrograde localization of the innervation of the middle cerebral artery with horseradish peroxidase in cats. Neurosurgery 16, 463–467. Yates, B.J., Mickle, J.P., Hedden, W.J., Thompson, T.J., 1987. Tracing of afferent pathways from the femoral-saphenous vein to the dorsal root ganglia using transport of horseradish peroxidase. J. Auton. Nerv. Syst. 20, 1–11.