- Email: [email protected]
Tentative identification of arterial baroreceptors associated with arteriovenous anastomoses
Brain Research, 444 (1988) 273-283
273
Elsevier BRE 13387
Tentative identification of arterial baroreceptors associated with arteriovenous anastomoses Satoru Onodera and Chizuka Ide Department of Anatomy, School of Medicine, lwate Medical University, Morioka (Japan) (Accepted 8 September 1987)
Key words: Wheatgerm-agglutinated horseradish perc.xidase (WGA-HRP); Transganglionic; Trigeminal nerve; Baroreceptor; Arteriovenous anastomosis; Cat
Nerve endings of the trigeminal nerve were examined in the lip, nostril, upper jaw and supraorbitai skin of cat by the horseradish peroxidase (HRP) tracing method. Wheatgerm agglutinin-HRP (WGA-HRP) at injected into the spinal trigeminal nucleus was transported transganglionicallyto nerve endings. The present study deals with presumptive baroreceptive nerve endings found on arteriovenous anastomoses and on some other large arteries. By light microscopy HRP-labeled, complexly arborized nerve endings were found on the walls of arteries located deep in the subcutaneous tissue of the lip and nostril. These arteries were identified in serial sections to be an arterial segment of an arteriovenous anastomosis. By electron microscopy characteristic nerve endings were located in the adventitia and partly extended into the adjacent connective tissue of the arterial wall. These terminals had no connective tissue capsule. Axon terminals were somewhat enlarged ( 1-2 ~m in diameter) and extended through bundles of collagen fibrils. The terminals contained an abundance of mitochondria and some vesicles. Axon terminals were typically covered by thin Schwann cell processes, but parts of the axolemma were sometimes devoid of such Schwann cell coverings, being invested only by basal laminae. Cell bodies of Schwann cells were located apart from axons. These light and electron microscopic features of the endings resembled those of other well-defined baroreceptors reported in the carotid sinus, aortic arch and endocardium, as well as of Ruffini terminals and Golgi tendon organs, suggesting that they would be baroreceptors of arteriovenous anastomosis. In addition HRP-labeled single nerve fibers with varicosities were found in the walls of some large arteries in the facial skin. By electron microscopy, such a HRP-positive nerve fiber contained some mltochondria and vesicles in varicosities and coursed with a bundle of HRP-negative fine fibers in the adventitia of arteries. These HRP-labeled single fibers were considered to be sensory derived from trigeminal nerves.
INTRODUCTION Arterial b a r a r e c e p t o r s have so far been described in the carotid sinus 1'3'5'7,13'!4'2°'24, c o m m o n carotid artery t, subclavian artery I, aortic arch I and endocardium 23'25 of various animals including cat, Baroreceptors in these areas consist of arborized axon terminals containing an a b u n d a n c e of m i t o c h o n d r i a and some small clear vesicles. Schwann cells, which are somewhat separated from axon terminals, send thin cellular processes to invest the axon terminals. Part of the a x o l e m m a is devoid of such Schwann cell covering and directly exposed to the connective tissue. These features are the same as those of nerve terminals of the arteriovenous a n a s t o m o s e s as described in
the present study. Most of these studies on b a r o r e c e p t o r s fail to recognize the striking similarity of these terminals to other m e c h a n o r e c e p t o r s described to date in joints and ligaments, specifically Ruffini terminals 2's-10'16'!7, Golgi t e n d o n organs 2t'22'26 and free nerve endings 16.17. During the e x a m i n a t i o n of sensory nerve endings in the cat facial skin following injection of wheatgerm a g g l u t i n i n - h o r s e r a d i s h peroxidase ( W G A - H R P ) into the spinal trigeminal nucleus, well developed nerve endings were identified on the wall of arteriovenous anastomoses in the deep subcutaneous tissues of the u p p e r lip and nostril skin. In addition H R P labeled single nerve fibers with varicosities were also
Correspondence: S. Onodera, Department of Anatomy, School of Medicine, twate Medical University, Morioka 020, Japan. 0006-8993/88/$03.50 © 1988 Elsevier Science Publishers B.V. (Biomedical Division)
274 found in the wall of some large arteries in these and other parts of the face including upper jaw and supraorbital skin. These presumptive baroreceptive axon terminals on arteries of the facial skin were documented in the present study. MATERIALSAND METHODS Five adult cats of both sexes were used. Animals were anesthetized by Nembutal (sodium pentobarbiturate, 30 mg/kg b. wt.), and a total of 2.7-17 ~l of 5-10% W G A - H R P (Toyobo or Sigma) solution was injected by more than 10 punctures into the extent part of the spinal trigeminal nucleus. The injection was done stereotaxically using a glass micropipette connected to a 50-~1 Hamilton microsyringe. Animals were allowed to survive for 3-5 days after the injection, and they again were deeply anesthetized and peffused through the heart with 800 ml of warm (37 °C) physiological saline followed by 3000 ml of fixative containing 0.5% paraformaldehyde and 2.5% glutaraldehyde in 0.1 M phosphate buffer (pH 7.4), and then by 1000 ml of 0.1 M phosphate buffer
(pH 7.4) containing 10% sucrose. Facial skins were excised from lip, nostril, upper jaw and supraorbital parts and they were stored overnight at 4 °C in 0.1 M phosphate buffer containing 30% sucrose. Frozen sections about 50~m thick were cut from these specimens on a freezing microtome. Sections were incubated in 0.005% tetramethylbenzidine (TMB) solution TM. Some sections were counterstained with a 1% Neutral red for light microscopy. Sections in which characteristic nerve endings with a flower spray pattern were identified by light microscopy were processed for electron microscopy. Sections were osmicared a t. 45 °C for 60 min in 2% osmium tetroxide solution adjusted to pH 6.0 with 0.1 M phosphate buffer, dehydrated through a graded series of ethanol for 7 rain at each concentration of ethanol and embedded in Epon 812. HRP-positive nerve endings were identified by light microscopy and such areas were trimmed for thin sectioning. Serial sections of one typical complexly arborized ending on an arteriovenous anastomosis were prepared by cutting 3 or 4 thick sections (1 ~tm thick) for light microscopy followed by 2 or 3 thin sections for electron microscopy.
Fig. 1. Light micrograph showing an HRP-labeled, complexlyarborized axon terminal on the small artery (large arrow) in the hairy skin of the upper lip. A single large axon (small arrows) gives fine branches, which formed a dense flower spray-like network around the artery, x750.
275
Fig. 2. This electron micrograph shows a part of the HRP-labeled arborized nerve ending of Fig. 1. The reaction products ot !-!!
276 Thin sections were cut with a diamond knife on an LKB Ultrotome and observed in a JEM-100B electron microscope. RESULTS
By light microscopy HRP-labeled nerve fibers could be followed along the wall of arteries in the deep subcutaneous tissues. These fibers gave rise to fine branches which appeared to end as terminals on arterial walls. These HRP-labeled nerve endings were varied in form consisting of apparently single to complexly arborized axon terminals. Complexly arborized nerve endings were found on the walls of small arteries in the hairy skin of the upper lip (Fig. 1) and nostril (Fig. 3). The arborized axons formed a dense flower spray-like network on the wall of the artery extending approximately 100 #m in diameter and covering about half the circumference of the artery. A similar nerve ending found in the nostril skin was also located adjacent to an artery and partly extended into neighboring connective tissue around the arterial wall (Fig. 3). Many small veins were usually observed ~n the area around the artery associated with th~,ze complexly arborized HRP-positive nerve endings. This anatomical relationship suggests that these arteries and veins could be part of an arteriovenous anastomosis, and that these characteristic nerve endings might be part of the arteriovenous anastomosis. Therefore, we carried out serial sections by light and electron microscopy on the HRP-positive nerve ending in the nostril skin of Fig. 3. These serial sections confi~ ,~ . a the fact that the artery adjacent to arborized nerve ending was directly continuous with the vein, identifying this artery as an arterial segment of arteriovenous anastomosis (Fig. 4). Electron microscopic observations of the nerve
Fig. 3. Light micrograph of unstained thick Epon section showing HRP-labeled axon terminals in the adventitia of a small artery (A) in the nostril skin. A small vein (V) is found in the vicinity of the artery, x622.
endings of Figs. 1 and 3 demonstrated the networks of axon terminals located in the adventitia of the artery (Figs. 2, 5 and 6). The ultrastructural features of axon terminals were the same in these two nerve endings. The axon terminals contained crystalloid inclusions of HRP reaction products and thus were derived from the trigeminal nerve. Axon terminals, measuring approximately 1-2/~m in diameter, contained an abundance of mitochondria and small clear vesicles (Fig. 6a). Each axon terminal was covered by thin cellular processes of Schwann cells. Schwann cell bodies were somewhat separated from axon terminals and extended their cell processes toward axon terminals. Axon terminals were partly devoid of Schwann cell coverings, being only invested by basal laminae and in direct contact with the connective tissues. One axon terminal was surrounded by several layers of Schwann cell processes resembling a lamellar corpuscle (Fig. 6b). Tightly packed bundles of collagen fibrils were separated by the cytoplasmic processes of Schwann cells and fibroblasts (Figs. 5 and
Fig. 4. These low-powered electron micrographs show parts of serial sections obtained from the region presented in Fig. 3. It is evident that the artery and vein in Fig. 3 formed an arteriovenous anastomosis, a: the artery (A) has a thick smooth muscle layer and tall endothelial cells. Two groups of axon terminals which correspot~d in location to those in Fig. 3 are found in the wall of the artery: the smaller one is confined in the adventitia, while the larger one extends partly into the connective tissue compartment. These two groups of axon terminals are enlarged in Fig. 5a,b, respectively, b: this electron micrograph was taken at the region about 8g~mdeep from Fig. 4a. The smaller group of axon terminals has disappeared in this level, but larger ones (arrows) are still evident. Some clear smooth muscle cells with large light nuclei (arrowheads) are obserw-d on the artery, c: this micrograph was taken at the region about 12 j~m deep from Fig. 4b. The artery (A) is directly continuous with the vein (V), showing that these vessels are components of an arteriovenous anastomosis, x 1200.
277
278
Fig. 5. These two electron micrographs show the detail of HRP-labeled terminals presented in Fig. 4a, a: HRP-labeled axon terminals (large arrows) are located in the adventitia. Many HRP-negative unmyelinated axons (small arrows) were founc~ closer to the smooth muscle layer (M). b: irregularly shaped HRP-labeled axon terminals are grouped in a small zone which has no connective tissue capsule, Colagen fibrils arc dense around axon terminals. A cell body of Schwann cell (S) is located apart from these terminals, x5500.
279
Fig. 6. Highly magnified electron micrographs of the HRP-labeled axon terminals of Fig. 1. a: the axon terminals contain an accumulation of mitochondria and some small clear vesicles (small arrow). The crystalloid reaction products of HRP (large arrows) are seen within the axon terminals. Schwann cell coverings are very thin on the axon terminals, being partly lacking (arrowhead). Collagen fibrils were dense around axon terminals. S, Schwann cell. x20,000, b: in rare cases an axon terminal was surrounded by layers of Schwann cell processes, x I 1,000.
6). Schwann cells were characterized by the presence of a prominent basal lamina, and fibroblasts lacked any sign of basal lamina. Small finger-like protrusions of Schwann cells sometimes wrapped a small bundle of collagen fibrils. No capsule could be recognized in the connective tissue compartment. Besides these labeled large endings, many unmyelinated axons without HRP labeling were present. Unmyelinated axons were thin, measuring 0.3-0.4 /zm in diameter, and were usually located closer to the smooth muscle layer than HRP-iabeled axon terminals in the adventitia (Figs. 2 and 5a). These unmyelinated fibers were considered to be autonomic nerves. Based on observations of semi-serial sections
by light and electron microscopy of the nerve ending of Fig. 3, and taking into consideration the ultrastructural findings of the nerve ending of Fig. 1, the 3-dimensional structure of part of the complexly arborized nerve ending on the arteriovenous anastomosis was schematically drawn in Fig. 7. The arterial segment had a thick smooth muscle layer and a relatively narrow lumen lined by high endothelial cells. At lower magnifications it is evident that some s.~ooth muscle cells of the arteriovenous anastomosis were round, translucent cells with large light nuclei (Fig. 4). These muscle cells contained scant myofilaments and a few mitochondria, and these are characteristic features of arterial smooth
280
x'b.:o.
~b
t
f j .
Fig. 7. This schematic drawing shows the 3-dimensional structure of part of a complexly arborized nerve ending on the arterial wall of arteriovenous anastomosis based on findings in Figs. 1-5. A myelinated nerve fiber (N) gives fine branches which form groups of flower spray-like complex termipals around the arterial segment (A). In addition well-developed autonomic nerve networks (n) surround the arterial segment. A few autonomic nerve fibers are also found along the vein (V). The arterial segment has tall endothelial cells (E) and a relatively thick smooth muscle layer (M).
281 muscle cells in arteriovenous anastomoses s.Ta9. The wall of an efferent vein was thin, lacking in smooth muscles. Occasionally unmyelinated axons were found on such a thin vein. HRP-positive unmyelinated axons were also contained in a bundle of HRP-negative unmyelinated fibers located in the adventitia of large arteries (Fig. 8). These HRP-positive unmyelinated axons had focal varicosities containing some mitochondria and vesicles (Fig. 9b). In addition an HRP-positive myelinated axon was present in a large unmyelinated fiber bundle close to the arterial wall and was surrounded by the distinct perineural sheath (Fig. 9a). DISCUSSION
The present study has described presumptive baroreceptive nerve endings on the wall of arteriovenous anastomoses and other large arteries in cat facial skin. HRP-labeled complexly arborized nerve endings were found on the wall of the arterial segment of arteriovenous anastomoses in the lip and nostril skin. These endings consisted of arbotized axon terminals containing an accumulation of mitochondria and some vesicles. Part of axolemma wa.~ devoid of Schwann cell covering and directly exposed to the connective tissue consisting of tightly packed bundles of collagen fibrils. These features are the same as those of the baroreceptors in the carotid sinus 1'2'5'7'13'14'20'24in the aortic arch I and in the endocardium 23'25. The morphology of complexly arbo-
Fig. 8. Light micrograph showing HRP-labeled fine fibers with
varicosities (small arrows) on the wall (W) of a large artery in tile upper jaw. x585.
rized nerve endings also resembles that of mechanoreceptors in specialized compartments of dense connective tissue, i.e. Ruffini terminals 2.a-l°a6,17 aLd Golgi tendon organs 21.22,26. Similar but less developed nerve endings containing a large number of mitochondtia and small clear vesicles have been reported to be present within smooth muscle layers of the bandicoot arteriovene:~ anastomoses 19. Since baroreceptors are defined functionally, future physiologic studies are needed to confirm the functional characteristics of these complexly arborized endings. Although some reports conclude that no sensory endings are found in rat arteriovenous anastomoses ~s, the distinction between sensory and autonomic nerves is impossible by ordinary electron microscopic observations. It is inconceivable that there are always baroreceotors as described in the present study in every arteriovenous anastomosis. There have been reports which suggest the presence of various-typed sensory endings in the arteriovenous anastomoses, which include simple unmyelinated axons with agranular vesicles 4, axon varicosities with mitochondria 7 and substance P-immunoreactive nerve fibers n. Therefore it can be considered that arteriovenous anastomoses have generally some kinds of sensory nerve endings, albeit they present wide varieties in the structural complexity. Such sensory endings may play a role of baroreceptive function in the control of the blood flow. In addition the present study has shown that there are sensory nerves derived from trigeminal nerves in a bundle of unmyelinated fibers associated with arteties. Although in the present study some reservation must be made on the diffusion of HRP injected into the brainstem, the finding has been obtained that fine nerve fibers with vaiicosities were labeled in the arterial wall in the preparations in which HRP was injected into trigeminal ganglions (unpublished data). Therefore, it can be considered that the HRP-labeled fine fibers in the present study were net autonomic but sensory. It can be said from the findings of the present study that while a sensory myelinated fiber which is included in a bundle of autonomic unmyclinated fibers is surrounded by its own Schwann cells, its peripheral unmyelinated parts were enveloped by common Schwann cells with other autonomic axons in the walls of the arteries. Substance P-immunoreactive nerve fibers with cell
282
Fig. 9. Electron micrographs showing HRP-labeled fibers in the adventitia of the large artery in the upper lip. a: an HRP-labeled (arrows) myelinated fiber is included in a large bundle of unmyelinated fibers surrounded by perineural sheath (P). × 12,000. b: a small unmyelinated fiber bundle located closer to smooth muscles (M) include fibers containing HRP-reaction products (arrows), which are regionally enlarged in varicosities. × 18,000.
283 bodies in the trigeminal ganglion were o b s e r v e d in the arterial wall in the rat lip skin by light microscopy6. T h e p a t t e r n of distribution of substance P-positive fibers r e p o r t e d in that p a p e r is similar to t h a t of H R P - l a b e l e d fine nerve fibers as described in the p r e s e n t study. In our e l e c t r o n microscopic observations substance P-positive axons shared c o m m o n Schwann cells with bundles of substance P-negative (i.e. p r o b a b l e a u t o n o m i c ) u n m y e l i n a t e d axons (unpublished data). A l t h o u g h sensory nerve endings on p e r i p h e r a l arteries may h a v e b e e n considered to be fine u n m y e l i n a t e d axons c o n t a i n i n g an a b u n d a n c e of m i t o c h o n d r i a t2, such axon terminals were r a t h e r few.
U n m y e l i n a t e d sensory, axons may be p r e s e n t that resemble a u t o n o m i c fibers. T h e H R P tracing m e t h o d is n e e d e d to distinguish u n m y e l i n a t e d sensory endings from the a u t o n o m i c nerves.
REFERENCES
635-646. 13 Knoche, H. and Addicks, K., Electron microscopic studies of the pressoreceptor fields of the carotid sinus of the dog, Cell Tissue Res., 173 (1976) 77-94. 14 Knoche, H., Wiesner-Menzel, L. and Addicks, K., Ultrastructure of baroreceptors in the carotid sinus of the rabbit, Acta. Anat., 106 (1980) 63-83. 15 Kondo, H., An electron microscopic study on the caudal glomerulus of the rat, J. Anat., 113 (1972) 341-358. 16 Munger, B.L. and Halata, Z., The sensory innervation of primate facial skin. I. Hairy skin, Brain Res. Rev., 5 (1983) 45-80. 17 Munger, B.L. and Halata, Z., The sensorineural apparatus of the human eyelid, Am. J. Anat., 170 (1984) 181-204. 18 Mesulam, M.-M., Tetramethyl benzidine for horseradish peroxidase neurohistochemistry: a non-carcinogenic blue reaction-product with superior sensitivity for visualizing neural afferents and efferents, J. Histochem. Cytochem., 26 (1978) 106-117. 19 Molyneux, G.S., Neural control of cutaneous arteriovenous anastomoses In D. Garlick (Ed.), Progress in Microcirculation Research, University of New South Wales, Sydney, 1981, pp. 296-315. 20 Rees, P.H., Observations on the fine structure and distribution of presumptive baroreceptor nerves at the carotid sinus, J. Comp. Neurol., 131 (1967) 517-543. 21 Schoultz, T.W. and Swett, J.M., The fine structure of Golgi tendon organs, J. Neurocytol., 1 (1972) 1-26. 22 Schoultz, T.W. and Swett, J.M., Ultrastructural organization of the sensory fibers innervating the Golgi tendon organ, Anat. Rec.. 179 (i974) i47-i62. 23 Tranum-Jensen, J., The ultrastructure of the sensory endorgans (baroreceptors) in the atrial endocardium of young mini-pig, J. Anat., 119 (1975) 255-275. 24 Yokota, R., lde, C., Nitatori. T. and Onodera, S., Cholinesterase activity in the carotid sinus baroreceptor, Acta Histochem. Cytochem., 15 (1982) 537-542. 25 Yokota. R., lde. C., Nitatori, T. and Onodera, S., Electron microscopic histochemistry of cholinesterase activity in the baroreceptor of the cat arterial endocardium. Acta Histochem. Cx'tochem., 16 (1983) 129-137. 26 Zelemi, J. and Soukup, T., The development of Goigi tendon organs, J. Neurocytol., 6 (1977) 171-194.
1 Aumonier, F.J., Histological observations on the distribution of baroreceptors in the carotid and aortic regions of the rabbit, cat and dog, Acta Anat., 82 (1972) 1-16. 2 Biemesderfer, D., Munger, B.L., Binck, J. and Dubner, R., The pilo-Ruffini complex: a non-sinus hair and associated slowly-adapting mechanoreceptor in primate facial skin, Brain Research, 142 (1978) 197-222. 3 B/Sck, P. and Gorgas, K., Fine structure of baroreceptor terminals in the carotid sinus of guinea pigs and mice, Cell Tissue Res., 170 (1976) 95-112. 4 Cauna, N., The fine structure of the arteriovenous anastomosis and its nerve supply in the human nasal respiratory mucosa, Anat. Rec., 168 (1970) 9-22. 5 Chiba, T., Fine structure of the baroreceptt.~r aerve te.rminals in the carotid sinus of the dog, J. Electron Microsc., 21 (1972) 139-148. 6 Cuello, A.C., Fiacco, M.D. and Paxinos, G., The central and peripheral ends of the substance P-containing sensory neurons in the rat trigeminal system, Brain Research, 152 (1978) 499-509. 7 Gorgas, K., B6ck, P., Tishendorf, F. and Curri, S.B., The fine structure of human digital arterio-venous anastomoses (Hoyer-Grosser's organs), Anat. Embryol., 150 (1977) 269-289. 8 Halata, Z., The ultrastructure of the sensory nerve endings in the articular capsule of the knee joint of the domestic cat (Ruffini corpuscles and Pacinian corpuscles), J. Anat., 124 (1977) 717-729. 9 Halata, Z. and Munger, B.L., The uitrastructure of Ruffini and Herbst corpuscles in the articular capsule of domestic pigeon, Anat. Rec., 198 (1980) 681-692. 10 Halata, Z. and Munger, B.L., Identification of the Ruffini corpuscle in human hairy skin, Cell Tissue Res., 219 (1981) 437-440. 11 Hartschuh, W., Weihe, E. and Reincke, M., Peptidergic (neurotensin, VIP, substance P) nerve fibres in the skin. lmmunohistochemical evidence of an involvement of neuropeptides in nociception, pruritus and inflammation, Br. J. Dermatol., Suppl. 25,109 (1983) 14-17. 12 lwayama, T., Furness, J.B. and Burnstock, G., Dual adrenergic and cholinergic innervation of the cerebral arteries of the rat: an ultrastructural study, Circ. Res., 26 (1970)
ACKNOWLEDGEMENTS The authors wish to t h a n k Professor B.L. Munger, D e p a r t m e n t of A n a t o m y , Faculty of Medicine, The Pennsylvania State University, H e r s h e y , U . S . A . , for his invaluable advice and discussions in preparing the manuscript.
273
Elsevier BRE 13387
Tentative identification of arterial baroreceptors associated with arteriovenous anastomoses Satoru Onodera and Chizuka Ide Department of Anatomy, School of Medicine, lwate Medical University, Morioka (Japan) (Accepted 8 September 1987)
Key words: Wheatgerm-agglutinated horseradish perc.xidase (WGA-HRP); Transganglionic; Trigeminal nerve; Baroreceptor; Arteriovenous anastomosis; Cat
Nerve endings of the trigeminal nerve were examined in the lip, nostril, upper jaw and supraorbitai skin of cat by the horseradish peroxidase (HRP) tracing method. Wheatgerm agglutinin-HRP (WGA-HRP) at injected into the spinal trigeminal nucleus was transported transganglionicallyto nerve endings. The present study deals with presumptive baroreceptive nerve endings found on arteriovenous anastomoses and on some other large arteries. By light microscopy HRP-labeled, complexly arborized nerve endings were found on the walls of arteries located deep in the subcutaneous tissue of the lip and nostril. These arteries were identified in serial sections to be an arterial segment of an arteriovenous anastomosis. By electron microscopy characteristic nerve endings were located in the adventitia and partly extended into the adjacent connective tissue of the arterial wall. These terminals had no connective tissue capsule. Axon terminals were somewhat enlarged ( 1-2 ~m in diameter) and extended through bundles of collagen fibrils. The terminals contained an abundance of mitochondria and some vesicles. Axon terminals were typically covered by thin Schwann cell processes, but parts of the axolemma were sometimes devoid of such Schwann cell coverings, being invested only by basal laminae. Cell bodies of Schwann cells were located apart from axons. These light and electron microscopic features of the endings resembled those of other well-defined baroreceptors reported in the carotid sinus, aortic arch and endocardium, as well as of Ruffini terminals and Golgi tendon organs, suggesting that they would be baroreceptors of arteriovenous anastomosis. In addition HRP-labeled single nerve fibers with varicosities were found in the walls of some large arteries in the facial skin. By electron microscopy, such a HRP-positive nerve fiber contained some mltochondria and vesicles in varicosities and coursed with a bundle of HRP-negative fine fibers in the adventitia of arteries. These HRP-labeled single fibers were considered to be sensory derived from trigeminal nerves.
INTRODUCTION Arterial b a r a r e c e p t o r s have so far been described in the carotid sinus 1'3'5'7,13'!4'2°'24, c o m m o n carotid artery t, subclavian artery I, aortic arch I and endocardium 23'25 of various animals including cat, Baroreceptors in these areas consist of arborized axon terminals containing an a b u n d a n c e of m i t o c h o n d r i a and some small clear vesicles. Schwann cells, which are somewhat separated from axon terminals, send thin cellular processes to invest the axon terminals. Part of the a x o l e m m a is devoid of such Schwann cell covering and directly exposed to the connective tissue. These features are the same as those of nerve terminals of the arteriovenous a n a s t o m o s e s as described in
the present study. Most of these studies on b a r o r e c e p t o r s fail to recognize the striking similarity of these terminals to other m e c h a n o r e c e p t o r s described to date in joints and ligaments, specifically Ruffini terminals 2's-10'16'!7, Golgi t e n d o n organs 2t'22'26 and free nerve endings 16.17. During the e x a m i n a t i o n of sensory nerve endings in the cat facial skin following injection of wheatgerm a g g l u t i n i n - h o r s e r a d i s h peroxidase ( W G A - H R P ) into the spinal trigeminal nucleus, well developed nerve endings were identified on the wall of arteriovenous anastomoses in the deep subcutaneous tissues of the u p p e r lip and nostril skin. In addition H R P labeled single nerve fibers with varicosities were also
Correspondence: S. Onodera, Department of Anatomy, School of Medicine, twate Medical University, Morioka 020, Japan. 0006-8993/88/$03.50 © 1988 Elsevier Science Publishers B.V. (Biomedical Division)
274 found in the wall of some large arteries in these and other parts of the face including upper jaw and supraorbital skin. These presumptive baroreceptive axon terminals on arteries of the facial skin were documented in the present study. MATERIALSAND METHODS Five adult cats of both sexes were used. Animals were anesthetized by Nembutal (sodium pentobarbiturate, 30 mg/kg b. wt.), and a total of 2.7-17 ~l of 5-10% W G A - H R P (Toyobo or Sigma) solution was injected by more than 10 punctures into the extent part of the spinal trigeminal nucleus. The injection was done stereotaxically using a glass micropipette connected to a 50-~1 Hamilton microsyringe. Animals were allowed to survive for 3-5 days after the injection, and they again were deeply anesthetized and peffused through the heart with 800 ml of warm (37 °C) physiological saline followed by 3000 ml of fixative containing 0.5% paraformaldehyde and 2.5% glutaraldehyde in 0.1 M phosphate buffer (pH 7.4), and then by 1000 ml of 0.1 M phosphate buffer
(pH 7.4) containing 10% sucrose. Facial skins were excised from lip, nostril, upper jaw and supraorbital parts and they were stored overnight at 4 °C in 0.1 M phosphate buffer containing 30% sucrose. Frozen sections about 50~m thick were cut from these specimens on a freezing microtome. Sections were incubated in 0.005% tetramethylbenzidine (TMB) solution TM. Some sections were counterstained with a 1% Neutral red for light microscopy. Sections in which characteristic nerve endings with a flower spray pattern were identified by light microscopy were processed for electron microscopy. Sections were osmicared a t. 45 °C for 60 min in 2% osmium tetroxide solution adjusted to pH 6.0 with 0.1 M phosphate buffer, dehydrated through a graded series of ethanol for 7 rain at each concentration of ethanol and embedded in Epon 812. HRP-positive nerve endings were identified by light microscopy and such areas were trimmed for thin sectioning. Serial sections of one typical complexly arborized ending on an arteriovenous anastomosis were prepared by cutting 3 or 4 thick sections (1 ~tm thick) for light microscopy followed by 2 or 3 thin sections for electron microscopy.
Fig. 1. Light micrograph showing an HRP-labeled, complexlyarborized axon terminal on the small artery (large arrow) in the hairy skin of the upper lip. A single large axon (small arrows) gives fine branches, which formed a dense flower spray-like network around the artery, x750.
275
Fig. 2. This electron micrograph shows a part of the HRP-labeled arborized nerve ending of Fig. 1. The reaction products ot !-!!
276 Thin sections were cut with a diamond knife on an LKB Ultrotome and observed in a JEM-100B electron microscope. RESULTS
By light microscopy HRP-labeled nerve fibers could be followed along the wall of arteries in the deep subcutaneous tissues. These fibers gave rise to fine branches which appeared to end as terminals on arterial walls. These HRP-labeled nerve endings were varied in form consisting of apparently single to complexly arborized axon terminals. Complexly arborized nerve endings were found on the walls of small arteries in the hairy skin of the upper lip (Fig. 1) and nostril (Fig. 3). The arborized axons formed a dense flower spray-like network on the wall of the artery extending approximately 100 #m in diameter and covering about half the circumference of the artery. A similar nerve ending found in the nostril skin was also located adjacent to an artery and partly extended into neighboring connective tissue around the arterial wall (Fig. 3). Many small veins were usually observed ~n the area around the artery associated with th~,ze complexly arborized HRP-positive nerve endings. This anatomical relationship suggests that these arteries and veins could be part of an arteriovenous anastomosis, and that these characteristic nerve endings might be part of the arteriovenous anastomosis. Therefore, we carried out serial sections by light and electron microscopy on the HRP-positive nerve ending in the nostril skin of Fig. 3. These serial sections confi~ ,~ . a the fact that the artery adjacent to arborized nerve ending was directly continuous with the vein, identifying this artery as an arterial segment of arteriovenous anastomosis (Fig. 4). Electron microscopic observations of the nerve
Fig. 3. Light micrograph of unstained thick Epon section showing HRP-labeled axon terminals in the adventitia of a small artery (A) in the nostril skin. A small vein (V) is found in the vicinity of the artery, x622.
endings of Figs. 1 and 3 demonstrated the networks of axon terminals located in the adventitia of the artery (Figs. 2, 5 and 6). The ultrastructural features of axon terminals were the same in these two nerve endings. The axon terminals contained crystalloid inclusions of HRP reaction products and thus were derived from the trigeminal nerve. Axon terminals, measuring approximately 1-2/~m in diameter, contained an abundance of mitochondria and small clear vesicles (Fig. 6a). Each axon terminal was covered by thin cellular processes of Schwann cells. Schwann cell bodies were somewhat separated from axon terminals and extended their cell processes toward axon terminals. Axon terminals were partly devoid of Schwann cell coverings, being only invested by basal laminae and in direct contact with the connective tissues. One axon terminal was surrounded by several layers of Schwann cell processes resembling a lamellar corpuscle (Fig. 6b). Tightly packed bundles of collagen fibrils were separated by the cytoplasmic processes of Schwann cells and fibroblasts (Figs. 5 and
Fig. 4. These low-powered electron micrographs show parts of serial sections obtained from the region presented in Fig. 3. It is evident that the artery and vein in Fig. 3 formed an arteriovenous anastomosis, a: the artery (A) has a thick smooth muscle layer and tall endothelial cells. Two groups of axon terminals which correspot~d in location to those in Fig. 3 are found in the wall of the artery: the smaller one is confined in the adventitia, while the larger one extends partly into the connective tissue compartment. These two groups of axon terminals are enlarged in Fig. 5a,b, respectively, b: this electron micrograph was taken at the region about 8g~mdeep from Fig. 4a. The smaller group of axon terminals has disappeared in this level, but larger ones (arrows) are still evident. Some clear smooth muscle cells with large light nuclei (arrowheads) are obserw-d on the artery, c: this micrograph was taken at the region about 12 j~m deep from Fig. 4b. The artery (A) is directly continuous with the vein (V), showing that these vessels are components of an arteriovenous anastomosis, x 1200.
277
278
Fig. 5. These two electron micrographs show the detail of HRP-labeled terminals presented in Fig. 4a, a: HRP-labeled axon terminals (large arrows) are located in the adventitia. Many HRP-negative unmyelinated axons (small arrows) were founc~ closer to the smooth muscle layer (M). b: irregularly shaped HRP-labeled axon terminals are grouped in a small zone which has no connective tissue capsule, Colagen fibrils arc dense around axon terminals. A cell body of Schwann cell (S) is located apart from these terminals, x5500.
279
Fig. 6. Highly magnified electron micrographs of the HRP-labeled axon terminals of Fig. 1. a: the axon terminals contain an accumulation of mitochondria and some small clear vesicles (small arrow). The crystalloid reaction products of HRP (large arrows) are seen within the axon terminals. Schwann cell coverings are very thin on the axon terminals, being partly lacking (arrowhead). Collagen fibrils were dense around axon terminals. S, Schwann cell. x20,000, b: in rare cases an axon terminal was surrounded by layers of Schwann cell processes, x I 1,000.
6). Schwann cells were characterized by the presence of a prominent basal lamina, and fibroblasts lacked any sign of basal lamina. Small finger-like protrusions of Schwann cells sometimes wrapped a small bundle of collagen fibrils. No capsule could be recognized in the connective tissue compartment. Besides these labeled large endings, many unmyelinated axons without HRP labeling were present. Unmyelinated axons were thin, measuring 0.3-0.4 /zm in diameter, and were usually located closer to the smooth muscle layer than HRP-iabeled axon terminals in the adventitia (Figs. 2 and 5a). These unmyelinated fibers were considered to be autonomic nerves. Based on observations of semi-serial sections
by light and electron microscopy of the nerve ending of Fig. 3, and taking into consideration the ultrastructural findings of the nerve ending of Fig. 1, the 3-dimensional structure of part of the complexly arborized nerve ending on the arteriovenous anastomosis was schematically drawn in Fig. 7. The arterial segment had a thick smooth muscle layer and a relatively narrow lumen lined by high endothelial cells. At lower magnifications it is evident that some s.~ooth muscle cells of the arteriovenous anastomosis were round, translucent cells with large light nuclei (Fig. 4). These muscle cells contained scant myofilaments and a few mitochondria, and these are characteristic features of arterial smooth
280
x'b.:o.
~b
t
f j .
Fig. 7. This schematic drawing shows the 3-dimensional structure of part of a complexly arborized nerve ending on the arterial wall of arteriovenous anastomosis based on findings in Figs. 1-5. A myelinated nerve fiber (N) gives fine branches which form groups of flower spray-like complex termipals around the arterial segment (A). In addition well-developed autonomic nerve networks (n) surround the arterial segment. A few autonomic nerve fibers are also found along the vein (V). The arterial segment has tall endothelial cells (E) and a relatively thick smooth muscle layer (M).
281 muscle cells in arteriovenous anastomoses s.Ta9. The wall of an efferent vein was thin, lacking in smooth muscles. Occasionally unmyelinated axons were found on such a thin vein. HRP-positive unmyelinated axons were also contained in a bundle of HRP-negative unmyelinated fibers located in the adventitia of large arteries (Fig. 8). These HRP-positive unmyelinated axons had focal varicosities containing some mitochondria and vesicles (Fig. 9b). In addition an HRP-positive myelinated axon was present in a large unmyelinated fiber bundle close to the arterial wall and was surrounded by the distinct perineural sheath (Fig. 9a). DISCUSSION
The present study has described presumptive baroreceptive nerve endings on the wall of arteriovenous anastomoses and other large arteries in cat facial skin. HRP-labeled complexly arborized nerve endings were found on the wall of the arterial segment of arteriovenous anastomoses in the lip and nostril skin. These endings consisted of arbotized axon terminals containing an accumulation of mitochondria and some vesicles. Part of axolemma wa.~ devoid of Schwann cell covering and directly exposed to the connective tissue consisting of tightly packed bundles of collagen fibrils. These features are the same as those of the baroreceptors in the carotid sinus 1'2'5'7'13'14'20'24in the aortic arch I and in the endocardium 23'25. The morphology of complexly arbo-
Fig. 8. Light micrograph showing HRP-labeled fine fibers with
varicosities (small arrows) on the wall (W) of a large artery in tile upper jaw. x585.
rized nerve endings also resembles that of mechanoreceptors in specialized compartments of dense connective tissue, i.e. Ruffini terminals 2.a-l°a6,17 aLd Golgi tendon organs 21.22,26. Similar but less developed nerve endings containing a large number of mitochondtia and small clear vesicles have been reported to be present within smooth muscle layers of the bandicoot arteriovene:~ anastomoses 19. Since baroreceptors are defined functionally, future physiologic studies are needed to confirm the functional characteristics of these complexly arborized endings. Although some reports conclude that no sensory endings are found in rat arteriovenous anastomoses ~s, the distinction between sensory and autonomic nerves is impossible by ordinary electron microscopic observations. It is inconceivable that there are always baroreceotors as described in the present study in every arteriovenous anastomosis. There have been reports which suggest the presence of various-typed sensory endings in the arteriovenous anastomoses, which include simple unmyelinated axons with agranular vesicles 4, axon varicosities with mitochondria 7 and substance P-immunoreactive nerve fibers n. Therefore it can be considered that arteriovenous anastomoses have generally some kinds of sensory nerve endings, albeit they present wide varieties in the structural complexity. Such sensory endings may play a role of baroreceptive function in the control of the blood flow. In addition the present study has shown that there are sensory nerves derived from trigeminal nerves in a bundle of unmyelinated fibers associated with arteties. Although in the present study some reservation must be made on the diffusion of HRP injected into the brainstem, the finding has been obtained that fine nerve fibers with vaiicosities were labeled in the arterial wall in the preparations in which HRP was injected into trigeminal ganglions (unpublished data). Therefore, it can be considered that the HRP-labeled fine fibers in the present study were net autonomic but sensory. It can be said from the findings of the present study that while a sensory myelinated fiber which is included in a bundle of autonomic unmyclinated fibers is surrounded by its own Schwann cells, its peripheral unmyelinated parts were enveloped by common Schwann cells with other autonomic axons in the walls of the arteries. Substance P-immunoreactive nerve fibers with cell
282
Fig. 9. Electron micrographs showing HRP-labeled fibers in the adventitia of the large artery in the upper lip. a: an HRP-labeled (arrows) myelinated fiber is included in a large bundle of unmyelinated fibers surrounded by perineural sheath (P). × 12,000. b: a small unmyelinated fiber bundle located closer to smooth muscles (M) include fibers containing HRP-reaction products (arrows), which are regionally enlarged in varicosities. × 18,000.
283 bodies in the trigeminal ganglion were o b s e r v e d in the arterial wall in the rat lip skin by light microscopy6. T h e p a t t e r n of distribution of substance P-positive fibers r e p o r t e d in that p a p e r is similar to t h a t of H R P - l a b e l e d fine nerve fibers as described in the p r e s e n t study. In our e l e c t r o n microscopic observations substance P-positive axons shared c o m m o n Schwann cells with bundles of substance P-negative (i.e. p r o b a b l e a u t o n o m i c ) u n m y e l i n a t e d axons (unpublished data). A l t h o u g h sensory nerve endings on p e r i p h e r a l arteries may h a v e b e e n considered to be fine u n m y e l i n a t e d axons c o n t a i n i n g an a b u n d a n c e of m i t o c h o n d r i a t2, such axon terminals were r a t h e r few.
U n m y e l i n a t e d sensory, axons may be p r e s e n t that resemble a u t o n o m i c fibers. T h e H R P tracing m e t h o d is n e e d e d to distinguish u n m y e l i n a t e d sensory endings from the a u t o n o m i c nerves.
REFERENCES
635-646. 13 Knoche, H. and Addicks, K., Electron microscopic studies of the pressoreceptor fields of the carotid sinus of the dog, Cell Tissue Res., 173 (1976) 77-94. 14 Knoche, H., Wiesner-Menzel, L. and Addicks, K., Ultrastructure of baroreceptors in the carotid sinus of the rabbit, Acta. Anat., 106 (1980) 63-83. 15 Kondo, H., An electron microscopic study on the caudal glomerulus of the rat, J. Anat., 113 (1972) 341-358. 16 Munger, B.L. and Halata, Z., The sensory innervation of primate facial skin. I. Hairy skin, Brain Res. Rev., 5 (1983) 45-80. 17 Munger, B.L. and Halata, Z., The sensorineural apparatus of the human eyelid, Am. J. Anat., 170 (1984) 181-204. 18 Mesulam, M.-M., Tetramethyl benzidine for horseradish peroxidase neurohistochemistry: a non-carcinogenic blue reaction-product with superior sensitivity for visualizing neural afferents and efferents, J. Histochem. Cytochem., 26 (1978) 106-117. 19 Molyneux, G.S., Neural control of cutaneous arteriovenous anastomoses In D. Garlick (Ed.), Progress in Microcirculation Research, University of New South Wales, Sydney, 1981, pp. 296-315. 20 Rees, P.H., Observations on the fine structure and distribution of presumptive baroreceptor nerves at the carotid sinus, J. Comp. Neurol., 131 (1967) 517-543. 21 Schoultz, T.W. and Swett, J.M., The fine structure of Golgi tendon organs, J. Neurocytol., 1 (1972) 1-26. 22 Schoultz, T.W. and Swett, J.M., Ultrastructural organization of the sensory fibers innervating the Golgi tendon organ, Anat. Rec.. 179 (i974) i47-i62. 23 Tranum-Jensen, J., The ultrastructure of the sensory endorgans (baroreceptors) in the atrial endocardium of young mini-pig, J. Anat., 119 (1975) 255-275. 24 Yokota, R., lde, C., Nitatori. T. and Onodera, S., Cholinesterase activity in the carotid sinus baroreceptor, Acta Histochem. Cytochem., 15 (1982) 537-542. 25 Yokota. R., lde. C., Nitatori, T. and Onodera, S., Electron microscopic histochemistry of cholinesterase activity in the baroreceptor of the cat arterial endocardium. Acta Histochem. Cx'tochem., 16 (1983) 129-137. 26 Zelemi, J. and Soukup, T., The development of Goigi tendon organs, J. Neurocytol., 6 (1977) 171-194.
1 Aumonier, F.J., Histological observations on the distribution of baroreceptors in the carotid and aortic regions of the rabbit, cat and dog, Acta Anat., 82 (1972) 1-16. 2 Biemesderfer, D., Munger, B.L., Binck, J. and Dubner, R., The pilo-Ruffini complex: a non-sinus hair and associated slowly-adapting mechanoreceptor in primate facial skin, Brain Research, 142 (1978) 197-222. 3 B/Sck, P. and Gorgas, K., Fine structure of baroreceptor terminals in the carotid sinus of guinea pigs and mice, Cell Tissue Res., 170 (1976) 95-112. 4 Cauna, N., The fine structure of the arteriovenous anastomosis and its nerve supply in the human nasal respiratory mucosa, Anat. Rec., 168 (1970) 9-22. 5 Chiba, T., Fine structure of the baroreceptt.~r aerve te.rminals in the carotid sinus of the dog, J. Electron Microsc., 21 (1972) 139-148. 6 Cuello, A.C., Fiacco, M.D. and Paxinos, G., The central and peripheral ends of the substance P-containing sensory neurons in the rat trigeminal system, Brain Research, 152 (1978) 499-509. 7 Gorgas, K., B6ck, P., Tishendorf, F. and Curri, S.B., The fine structure of human digital arterio-venous anastomoses (Hoyer-Grosser's organs), Anat. Embryol., 150 (1977) 269-289. 8 Halata, Z., The ultrastructure of the sensory nerve endings in the articular capsule of the knee joint of the domestic cat (Ruffini corpuscles and Pacinian corpuscles), J. Anat., 124 (1977) 717-729. 9 Halata, Z. and Munger, B.L., The uitrastructure of Ruffini and Herbst corpuscles in the articular capsule of domestic pigeon, Anat. Rec., 198 (1980) 681-692. 10 Halata, Z. and Munger, B.L., Identification of the Ruffini corpuscle in human hairy skin, Cell Tissue Res., 219 (1981) 437-440. 11 Hartschuh, W., Weihe, E. and Reincke, M., Peptidergic (neurotensin, VIP, substance P) nerve fibres in the skin. lmmunohistochemical evidence of an involvement of neuropeptides in nociception, pruritus and inflammation, Br. J. Dermatol., Suppl. 25,109 (1983) 14-17. 12 lwayama, T., Furness, J.B. and Burnstock, G., Dual adrenergic and cholinergic innervation of the cerebral arteries of the rat: an ultrastructural study, Circ. Res., 26 (1970)
ACKNOWLEDGEMENTS The authors wish to t h a n k Professor B.L. Munger, D e p a r t m e n t of A n a t o m y , Faculty of Medicine, The Pennsylvania State University, H e r s h e y , U . S . A . , for his invaluable advice and discussions in preparing the manuscript.