- Email: [email protected]
Autoradiographic observations on the innervation of the carotid body of the domestic fowl
Brain Research, 266 (1983) 193-201 Elsevier Biomedical Press
193
Autoradiographic Observations on the Innervation of the Carotid Body of the Domestic Fowl A. A. M. TAHA* and A. S. KING**
Department of Veterinary Anatomy, University of Liverpool, P.O. Box 147, Brownlow Hill, Liverpool, L69 3BX (U.K.) (Accepted October 18th, 1982)
Key words: carotid body innervation - - autoradiography - - synaptic morphology - - domestic fowl
Tritiated leucine was injected into the distal vagal ganglion of 11 domestic fowl, which survived for 12-24 h under general anaesthesia. The cells of this ganglion are known to be exclusively afferent. EM autoradiography showed that in all 11 birds the vast majority of the silver grains fell upon the nervous tissues of the carotid body. In 5 of these birds a quantitative analysis was made, using point-countingmorphometry. The incidence of silver grains per unit area was found to be 26 times greater in axonal endings than in the non-nervous components, and 15 times greater in axons in transit than in non-nervous components. The difference in incidence per unit area between these nervous and non-nervous components was highly significant (P < 0.001). Of all the observed axonal endings 77 ~ were labelled, but there is evidence that this is a substantial underestimate of the total population of afferent endings; in one bird 88 ~ of the endings were labelled. Of the axons in transit, 18 ~ were labelled. This low value is believed to be related to transfer of the label to the axonal endings by the fast component of axonal transport. Afferent and reciprocal synapses occurred in labelled axonal endings, which were therefore considered to have an afferent function. 'Efferent' type synapses also occurred in labelled endings, and therefore belonged to axons which in fact were afferent in function. It is concluded that the innervation of the carotid body of the domestic fowl is almost entirely afferent, the nerve cell bodies being in the distal vagal ganglion. Only very few efferent axonal endings are present. Ultrastructural features, including synaptic morphology, appear to constitute unreliable criteria for distinguishing between afferent and efferent axonal endings in the carotid body. INTRODUCTION It has long been accepted that the carotid b o d y is a chemoreceptor organ, b u t the identity o f the afferent nerve endings within the carotid b o d y has r e m a i n e d controversial despite extensive m o r p h o logical observations a n d degeneration studies i n b o t h m a m m a l s (for review see F i d o n e et al. 7) a n d birds (see A b d e l - M a g i e d a n d King3). It has proved unsatisfactory to rely solely o n general ultrastructure a n d synaptic m o r p h o l o g y to distinguish afferent from efferent axonal elements in the carotid body. I n theory, degeneration experiments can make this distinction, b u t i n practice the interpretation of pathological changes, particularly in their early
stages, is often uncertain. D e g e n e r a t i o n is also n o n selective since it necessarily involves all the axons, whether afferent or efferent, which h a p p e n to pass t h r o u g h the site of the cut. I n a n a t t e m p t to clarify the identity of the afferent nerve endings in the avian carotid b o d y it was decided to apply the technique of a u t o r a d i o g r a p h y to the n o r m a l physiological process of axoplasmic flow. T o the best o f our knowledge this has been attempted in only two investigations of the m a m m a lian carotid bodyT, 16, a n d has n o t previously been reported at all for the avian carotid b o d y except for a p r e l i m i n a r y a c c o u n t of the present experiments a n d a related doctoral thesis18,19.
* Present address: Department of Anatomy, Faculty of Veterinary Science, University of Khartoum, P.O. Box 32, Khartoum North, Sudan. ** To whom reprint requests should be addressed. 0006-8993/83/0000-0000/$03.00 © 1983 Elsevier Science Publishers
194 MATERIALS AND METHODS
Birds Eleven adult domestic hens (Gallus gallus var. domesticus) weighing 1.5-2 kg were used in this study. All were of the Golden Hubbard Comet strain. The carotid body in this species of bird has been shown to receive almost its entire innervation from the carotid body nerve of the distal vagal ganglion2,3. The anatomical nomenclature follows the Nornina Anatomica Avium.
The radioisotope Tritiated [4,5-3H]L-leucine (72 Ci/mmol, 1 mCi/ml; Amersham) was immediately freeze-dried after its arrival, and then redissolved in avian Ringer's solution to give a volume of 0.1 ml. It was used within 24 h.
Surgical exposure of the distal vagal ganglion Each bird was anaesthetized initially by slowly injecting 2-3 ml of Sagatal (M and B) into the left or right brachial vein. The opposite brachial vein was cannulated and a deep level of anaesthesia was continuously maintained with i.v. urethane. An incision, about 8 cm long, was made over the shoulder joint and the adjacent region of the scapula. The neck of the scapula was revealed by blunt dissection of the muscles covering it. The bone was then cut and the two stumps were separated by a retractor. The clavicular air sac was exposed by this procedure and cleared by blunt dissection. The brachial plexus was cut by a pair of fine scissors. This approach gave a clear view of the thyroid gland, jugular vein, and the vagus nerve itself. The distal vagal ganglion was identified by its topographical relations 2 and by being slightly swollen, very vascular, and lacking the cross-striations which are seen in the vagus nerve elsewhere.
Injection of the isotope Injection of either the left or right distal vagal ganglion was carried out under a dissecting microscope with a micropipette. The micropipette was introduced obliquely through the dorsal surface of the ganglion, to a depth of about 1 mm. The isotope was injected very slowly. A pledget of cotton wool soaked in avian Ringer's solution was placed on the
site of the operation to prevent dehydration, and the retractor was then removed. The birds were allowed to survive for 12-24 h under continuous deep anaesthesia before they were killed.
Fixation and processing for LM and EM autoradiography Fixation was carried out by perfusion with halfstrength Karnovsky's~° solution. The left and right distal vagal ganglia and carotid bodies were removed and routinely processed for electron microscopy. For light microscopic autoradiography 1 #m thick resin sections of the distal vagal ganglion were coated with Ilford K2 emulsion (diluted 1:1 with distilled water). They were exposed at 4 °C for periods of 7-21 days, developed in D-19 for 5 min, rinsed in distilled water for 30 s, and fixed in 30 sodium thiosulphate for 5 rain. They were finally rinsed in 3 changes of distilled water for 1 rain each, before being stained with a 1 ~ solution of toluidine blue. For electron microscopic autoradiography ultrathin sections of the carotid body were cut, collected on formvar-coated grids, and stained with lead citrate. They were then coated with Ilford L-4 emulsion (diluted 1:2 with distilled water) using the wire loop method of Caro and van Tubergen 5. The sections were exposed at 4 °C for periods of up to 6 months, developed in D-19 for 3 min at 20 °C, rinsed in distilled water for 20 s, fixed in 30 ~ sodium thiosulphate for 4 rain, and examined electron microscopically. The distal vagal ganglion and the carotid body of the uninjected side were used as controls.
Analysis of EM autoradiographs Qualitative observations were made on EM autoradiographs from the distal vagal ganglia and carotid bodies of all 11 birds. A quantitative analysis by point-counting morphometry15 was then performed on 5 of these birds, which were selected because the injected ganglion was proven by light microscopy to be heavily labelled. For each of these birds a stratified sampling technique was adopted. Three sections were taken randomly, at widely separated intervals, from the carotid body of the same side as that of the
Fig. 1. Light microscope autoradiograph of the distal vagal ganglion at the site of injection of tritiated leucine. A ganglion cell (G) is heavily labelled. Some myelinated axons are labelled (thick arrow) and some are not labelled (thin arrow). The labelled axons probably have their cell bodies in the ganglion itself whereas the unlabelled axons are probably passage fibres. Resin section, 1/~m thick, toluidine blue stain, x 950. Fig. 2. General view of the parenchyma of the carotid body after injecting the ipsilateral distal vagal ganglion with tritiated leucine. Labelling of the various components of the carotid body is non-uniform. Silver grains exclusively overlie the nervous tissue (asterisks). 1 -- type 1 cells; 2 ~ type 2 cell. x 15,000. Figs. 3 and 4. Capsular region of the carotid body, after injecting the ipsilateral distal vagal ganglion with tritiated leucine. A myelinated (M) and an unmyelinated axon (U) are labelled. The axonal profile (A), which is characterized by dense-cored granular vesicles, is unlabelled; this profile is interpreted as being either an amine or peptide containing ending. Fig. 3, x 10,000; Fig. 4, × 21,000.
196 injected ganglion. F r o m each section 5 autoradiographs were obtained from predetermined corners of the grid squares at a primary magnification of × 10,000 and enlarged to give a final magnification of x 25,000. This procedure yielded 75 micrographs from the 5 birds. A quadratic lattice grid with 306 points was superimposed over each print. The tissues of the carotid body were divided into axonal endings (expanded axonal profiles containing relatively numerous vesicles and mitochondria), axons in transit (small diameter axonal profiles characterized by neurotubules with few or no mitochondria and vesicles), and all non-nervous components (type 1 cells, type 2 cells, connective tissue, and blood vessels). The number of grid points, and the number of silver grains, falling on each component was counted. Where a grain overlaid two components the position of its centre was estimated and the grain was allocated to the component under the centre. The total numbers of labelled and unlabelled axons in transit, and axonal endings with and without synapses, were counted in the same 5 birds. RESULTS
Distal vagal ganglion In all 11 birds the light microscopic autoradiographs of the injected ganglion showed heavy concentrations of silver grains selectively overlying the ganglion cells (Fig. 1). Some of the myelinated axons were also labelled and others were not (Fig. 1).
Carotid body: qualitative observations In the EM autoradiographs of the carotid body, the silver grains were non-uniformly distributed
amongst the various structural components. In all 11 birds the vast majority of the grains fell upon the nervous tissues, especially the axonal endings (Fig. 2), only a few falling upon the non-nervous tissues. In the capsule, both myelinated and unmyelinated axons were labelled (Figs. 3, 4). The only type of axonal ending which constantly displayed no silver grains was that characterized by dense-cored granular vesicles (axonal ending A, in Fig. 4); only 7 such axons were observed out of a total of 207 axonal endings. The labelled axonal endings were either large caliciform endings or bouton-like endings. Some of these endings contained great numbers of small clear vesicles (Fig. 5), and others were well-filled with mitochondria (Fig. 2, top right-hand corner). In yet others there were numerous mitochondria in one region, and many clear vesicles in another region of the same ending (Fig. 6). Some of the labelled axonal endings apposed to type 1 cells showed synaptic complexes. O f these, some fulfilled the morphological criteria of afferent synapses transmitting from cell to axon (Figs. 7, 8). Others satisfied the morphological features of efferent synapses (Figs. 9-11). Finally, some of the synapses possessed the characteristics of reciprocal synapses (Fig. 12).
Quantitative observations The areas of the nervous and non-nervous components of the carotid body, and the incidence of silver grains upon each of these components, are summarized in Table I. The incidence of grains per unit area is 26 times greater on axonal endings than on the non-nervous components, and 15 times greater on axons in transit than on non-nervous components. These differences in the number of
Fig. 5. Parenchyma of carotid body after injecting the ipsilateral distal vagal ganglion with tritiated leucine. A labelled, bouton-like, axonal ending (AE) is apposed to a type 1 cell (1). The ending shows numerous small clear vesicles, some dense-cored vesicles, and a few mitochondria. Morphologically this ending could have been interpreted as an efferent ending (cholinergic), but the labelling proves that it is in fact afferent. Fig. 6. Parenchyma of carotid body after injecting the ipsilateral distal vagal ganglion with tritiated leucine. A labelled axonal ending (AE) is extensively apposed to a type 1 cell (1). The upper part of the ending is packed with mitochondria, whereas the lower part contains only a few mitochondria but numerous clear vesicles. A section passing through the plane A-B, would probably have given a profile packed with mitochondria which could have been interpreted (correctly, as the labelling shows) as afferent. A section through C-D would probably have yielded a profile containing many small clear vesicles, and this could have been interpreted (incorrectly) as an efferent ending. Figs. 7 and 8. Parenchyma of carotid body after injecting the ipsilateral distal vagal ganglion with tritiated leucine. In each Fig., a
197
labelled axonal ending (AE) is apposed to a Type 1 cell (1). A synapse (arrows, indicating direction of transmission is formed between each ending and the apposing cell. Small, intermediate, and large sized dense-cored granular vesicles (between the arrows) have accumulated on the presynaptic membrane (here, the cell side). The morphology of these synapses suggests transmission from cell to axon, the axon being afferent, i.e. projecting centrally. The labelling of these endings confirms that they were afferent. Fig. 7, x 39,000; Fig. 8, × 27,000.
Figs. 9-11. Parenchyma of carotid body after injecting the ipsilateral distal vagal ganglion with tritiated leucine. Each of the 3 labelled axonal endings (AE) is apposed to a type 1 cell (1). A synapse (arrows) is formed between each ending and its apposed type 1 cell. Clear vesicles have accumulated on the presynaptic membrane (here, the axon side). The postsynaptic membrane is thicker than the presynaptic membrane, and shows no accumulation of either clear or dense-cored vesicles. On the basis of synaptic morphology these 3 endings could have been interpreted as efferent endings (transmitting in the direction of the arrows, i.e. from axon to cell), but since each ending is labelled it must have been afferent. Fig. 9, × 41,500; Fig. 10, x 38,000; Fig. 11, × 26,500. Fig. 12. Parenchyma of carotid body, after injecting the ipsilateral distal vagal ganglion with tritiated leueine. A labelled axonal ending (AE) is apposed to a type 1 cell (1). A reciprocal synapse (arrows) is formed between the ending and the apposing type 1 cell. The labelling shows this ending to be afferent, x 42,500.
199 TABLE I Areas and grain counts o f nervous and non-nervous components o f the carotM body
The data were obtained from 75 EM autoradiographs from the labelled carotid bodies of 5 birds. Area o f autoradiographs
Structural component
Axons in transit Axonal endings Non-nervous components
%
m~
2.06 19.41 78.53
0.042 0.396 1.602
No. o f grains
No. o f grainsper m 2
57 915 143
1357 2311 89
TABLE II Labelled and unlabelled axonal elements
The data were obtained from75 EM autoradiographs from the labelled carotid bodies of 5 birds. 'Efferent' type synapses possessed accumulations of clear vesicles against the axonal membrane, and thickened type 1 cell membrane (Figs. 9-11). 'Afferent' type synapses showed accumulations of dense-cored granular vesicles against the type 1 cell membrane (Figs. 7, 8). Reciprocal synapses comprised 'afferent' and 'efferent' type structures in the same axonal ending (Fig. 12). The 7 axonal endings (all unlabelled) which contained dense-cored granular vesicles are excluded from the Table. Axons in transit
Labelled Unlabelled Percent labelled
36 175 17
Axonal endings o f all types
Axonalendingswithsynapses 'Efferent' type
'Afferent' type
Reciprocal
154 46 77
13 1
9 2
4 0
grains per unit area between axons in transit or axonal endings on the one hand, and non-nervous components on the other, were highly significant (P < 0.001). The numbers of labelled and unlabelled axons in transit, and axonal endings with and without synapses, are summarized in Table II. In all 5 birds together 77 ~ of the axonal endings were labelled, but in the individual birds the value ranged from 68 ~ to 88 ~ . C o n t r o l tissues
The control ganglia and carotid bodies always showed negligible labelling. DISCUSSION The highly significant difference between the degree of labelling of the nervous and non-nervous components of the carotid body on the injected side, and the negligible labelling of all control tissues,
showed that the radioactive tracer had reached the carotid body on the injected side solely from the injected distal vagal ganglion. In the domestic fowl this ganglion has been shown to contain exclusively afferent nerve cell bodies ~°, although many efferent axons pass through it. It is widely accepted that such axons in passage take up and incorporate into protein very little of an injected amino acid 6,7, and that whatever is so taken up is incorporated mostly into the soluble fraction 13 and is subsequently lost during fixation 8. Consequently axons passing between nerve cell bodies are believed to contribute little or nothing to the subsequent axonal transport of labelled material 6,7. It is therefore concluded that the labelled axonal elements observed in the carotid body in the present investigation belonged to afferent nerve fibres with their cell bodies in the distal vagal ganglion. Since 77 ~ of the axonal endings were labelled, it is clear that the axonal endings in the avian carotid body are predominantly afferent in function. The
200 method of allocating the silver grains was rigorous, in that some grains partly overlaid axonal endings, but were allocated to adjacent non-nervous components according to the position of the centre of the grain. This makes no allowance for the fact that an emission from an axonal ending may be oblique and induce a silver grain somewhat outside the axonal profile. Moreover the distal vagal ganglion in birds is very elongated; therefore a single injection of isotope could have led to its uneven distribution amongst the ganglion cells, as has been suggested 16 for the relatively compact petrosal ganglion of the cat. For all these reasons the 77 ~ of labelled axonal endings is almost certainly a substantial underestimate of the proportion of afferent axonal endings. A further indication of this is shown by one bird in which 88~ of the axonal endings were labelled. Indeed only one type of axonal ending (characterized by dense-cored granular vesicles) was never labelled, and such endings formed only 3.4 of all the observed endings. It is noteworthy that a very small number of axonal endings of this type were the only ones which survived removal of the distal vagal ganglion in the same species of bird by Abdel-Magied and KingL Of the axons in transit, only 17 ~ were labelled in the present study. Since the birds survived only 12-24 h after injecting the isotope, the incorporated amino acid had presumably reached the carotid body by the fast component of axonal transport; this has been reported to be 250-350 mm/day in the chicken (see LaseklZ), and therefore any label travelling in the fast component in our experiments could have been largely cleared from the axons in transit and concentrated in the axonal endings by the time that the animals were killed. On the other hand, the slow component of transport (about 1 ram/day13) would not have had time to reach the carotid body. The low proportion of labelled axons in transit which we observed is therefore not inconsistent with the degeneration experiments by AbdelMagied and Kings in which almost all the axons in transit (as well as axonal endings) degenerated after distal vagal ganglionectomy. We conclude that the innervation of the carotid body of the domestic fowl is almost entirely afferent. This confirms the conclusion reached by AbdelMagied and King3 from degeneration experiments
in the same species of bird. The evidence for the almost exclusively afferent innervation of the carotid body in the two highest vertebrates, i.e. in both birds and mammals, has now been so consistently demonstrated by so many different experimental techniques that there seems no longer any justification for regarding the matter as controversial. Moreover, the concept of transganglionic degeneration offers a reasonable explanation~,9,17 for the controversial degeneration results obtained by Biscoe et al. 4 after intracranial section of the glossopharyngeal nerve of the cat. Since autoradiographic labelling of axonal endings is achieved by the normal physiological process of axoplasmic flow, without disturbing neuronal structure, this technique offers a reliable way of checking the ultrastructural characteristics of afferent axonal endings. It has been suggestedal, 14 that efferent axonal endings in the carotid body are closely packed with small clear vesicles about 35-50 nm in diameter. On this basis the axonal ending in Fig. 5 would have been interpreted as efferent, but nevertheless it must have been afferent because it is labelled. There has been less agreement about the general ultrastructure of afferent endings, but one view is that such endings are packed with small mitochondria12. The axonal profile in Fig. 6 shows the dangers of using these criteria to distinguish between efferent and afferent endings. One end of this profile is filled with mitochondria, so that a transverse section through A-B could have been packed with mitochondria and interpreted (correctly) as afferent. At the other end, a section through C-D would be expected to yield a profile containing many small clear vesicles, and this could have been interpreted (incorrectly) as efferent. Synaptic morphology has been used with relative confidence to identify afferent and efferent axonal endings, utilizing the accumulation of vesicles against the presynaptic membrane and the often asymmetrical pre- and postsynaptic membranes 11. The 'afferent' type of synapse (Figs. 7, 8) and the reciprocal synapses (Fig. 12) observed in this investigation were nearly all labelled (Table II) and therefore belonged to afferent endings. Of the 13 'efferent' type synapses, 12 occurred in labelled endings (Figs. 9-11) and therefore belonged to afferent axons. This observation seems to call into
201 question the reliability of the classical concepts of synaptic morphology. Alternatively, it is j u s t conceivable that these synapses were the efferent
ACKNOWLEDGEMENTS
c o m p o n e n t s of reciprocal synapses, a possibility which was recognized b y K i n g et al. 11.
H e n r y with electron microscopy, J. G e a r y with
O n the basis o f these observations we conclude that purely m o r p h o l o g i c a l criteria are n o t reliable for distinguishing between afferent a n d efferent axonal endings, at least in the carotid body.
REFERENCES 1 Abdel-Magied, E. M., Observations on Chemoreceptor and Baroreceptor Structures in the Carotid Body Region of the Domestic Fowl, P h . D . Thesis, 1979, University of Liverpool. 2 Abdel-Magied, E. M. and King, A. S., The topographical anatomy and blood supply of the carotid body region of the domestic fowl, J. Anat., 126 (1978) 535-546. 3 Abdel-Magied, E. M. and King, A. S., Effects of distal vagal ganglionectomy and midcervical vagotomy on the ultrastructure of axonal elements in the carotid body of the domestic fowl, J. Anat., 134 (1982) 643-652 4 Biscoe, T. J., Lall, A. and Sampson, S. R., Electron microscopic and electrophysiological studies on the carotid body following intracranial section of the glossopharyngeal nerve, J. PhysioL (Lond.), 208 (1970) 133-152. 5 Caro, L. G. and van Tubergen, R. P., High resolution autoradiography, I. Methods, J. Cell Biol., 15 (1962) 173-188. 6 Cowan, W. M., Gottlieb, D. I., Hendrickson, A. E., Price, J. L. and Woolsey, T. A., The autoradiographic demonstration of axonal connections in the central nervous system, Brain Research, 37 (1972) 21-51. 7 Fidone, S. J., Zapata, P. and Stensaas, L. J., Axonal transport of labeled material into sensory nerve endings of cat carotid body, Brain Research, 124 (1977) 9-28. 8 Hendrickson, A. E., Electron microscopic distribution of axoplasmic transport, J. comp. NeuroL, 144 (1972) 381-398. 9 Hess, A. and Zapata, P., Innervation of the cat carotid body: normal and experimental studies, Fed. Proc., 31 (1972) 1365-1382. 10 Karnovsky, M. J., A formaldehyde/glutaraldehyde fixative of high osmolarity for use in electron microscopy,
W e gratefully acknowledge the assistance of Julie photography, a n d C a r o l y n Roberts with typing. Dr. M. A. Williams gave valuable advice o n the interp r e t a t i o n of autoradiographs. We t h a n k The British C o u n c i l a n d University of K h a r t o u m for m a k i n g our c o l l a b o r a t i o n possible.
J. Cell Biol., 27 (1965) 137-138a. 11 King, A. S., King, D. Z., Hodges, R. D. and Henry, J., Synaptic morphology of the carotid body of the domestic fowl, Cell Tiss. Res., 162 (1975) 459-473. 12 King, A. S., McLelland, J., Cook, R. D., King, D. Z. and Walsh, C., The ultrastructure of afferent nerve endings in the avian lung, Respir. Physiol., 22 (1974) 21-40. 13 Lasek, R. J., Protein transport in neurons, Int. Rev. Neurobiol., 13 (1970) 289-324. 14 McDonald, D. M. and Mitchell, R. A., The innervation of glomus cells, ganglion cells and blood vessels in the rat carotid body: a quantitative ultrastructural analysis, J. Neurocytol., 4 (1975) 177-230. 15 Ross, R. and Benditt, E. P., Quantitative electron microscope radioautographic observations of proline-Hz utilization by fibroblasts, J. Cell Biol., 27 (1965) 83-106. 16 Smith, P. G. and Mills, E., Autoradiographic identification of the terminations of petrosal ganglion neurons in the cat carotid body, Brain Research, 113 (1976) 174-178. 17 Smith, P. G. and Mills, E., Time course of Wallerian degeneration in the carotid body after carotid sinus nerve transection. In C. Belmonte, D. J. Pallor, H. Acker and S. Fidone (Eds.), Arterial Chemoreceptors, Proceedings of the 111International Meeting, University Press, Leicester, 1981, pp. 430-439. 18 Taha, A. A. M., Autoradiographic evidence for the afferent innervation of the carotid body in Gallus domesticu~, J. Anat., 133 (1981) 134. 19 Taha, A. A. M., Functional Anatomy of Axons and Granular Cells of the Carotid Body, Aorta, and Pulmonary Arteries in the Domestic Fowl, P h . D . Thesis, 1981, University of Liverpool. 20 Wakley, G. K. and Bower, A. J., The distal vagal ganglion of the hen (Gallus domesticus), a histological and physiological study, J. Anat., 132 (1981) 95-105.
193
Autoradiographic Observations on the Innervation of the Carotid Body of the Domestic Fowl A. A. M. TAHA* and A. S. KING**
Department of Veterinary Anatomy, University of Liverpool, P.O. Box 147, Brownlow Hill, Liverpool, L69 3BX (U.K.) (Accepted October 18th, 1982)
Key words: carotid body innervation - - autoradiography - - synaptic morphology - - domestic fowl
Tritiated leucine was injected into the distal vagal ganglion of 11 domestic fowl, which survived for 12-24 h under general anaesthesia. The cells of this ganglion are known to be exclusively afferent. EM autoradiography showed that in all 11 birds the vast majority of the silver grains fell upon the nervous tissues of the carotid body. In 5 of these birds a quantitative analysis was made, using point-countingmorphometry. The incidence of silver grains per unit area was found to be 26 times greater in axonal endings than in the non-nervous components, and 15 times greater in axons in transit than in non-nervous components. The difference in incidence per unit area between these nervous and non-nervous components was highly significant (P < 0.001). Of all the observed axonal endings 77 ~ were labelled, but there is evidence that this is a substantial underestimate of the total population of afferent endings; in one bird 88 ~ of the endings were labelled. Of the axons in transit, 18 ~ were labelled. This low value is believed to be related to transfer of the label to the axonal endings by the fast component of axonal transport. Afferent and reciprocal synapses occurred in labelled axonal endings, which were therefore considered to have an afferent function. 'Efferent' type synapses also occurred in labelled endings, and therefore belonged to axons which in fact were afferent in function. It is concluded that the innervation of the carotid body of the domestic fowl is almost entirely afferent, the nerve cell bodies being in the distal vagal ganglion. Only very few efferent axonal endings are present. Ultrastructural features, including synaptic morphology, appear to constitute unreliable criteria for distinguishing between afferent and efferent axonal endings in the carotid body. INTRODUCTION It has long been accepted that the carotid b o d y is a chemoreceptor organ, b u t the identity o f the afferent nerve endings within the carotid b o d y has r e m a i n e d controversial despite extensive m o r p h o logical observations a n d degeneration studies i n b o t h m a m m a l s (for review see F i d o n e et al. 7) a n d birds (see A b d e l - M a g i e d a n d King3). It has proved unsatisfactory to rely solely o n general ultrastructure a n d synaptic m o r p h o l o g y to distinguish afferent from efferent axonal elements in the carotid body. I n theory, degeneration experiments can make this distinction, b u t i n practice the interpretation of pathological changes, particularly in their early
stages, is often uncertain. D e g e n e r a t i o n is also n o n selective since it necessarily involves all the axons, whether afferent or efferent, which h a p p e n to pass t h r o u g h the site of the cut. I n a n a t t e m p t to clarify the identity of the afferent nerve endings in the avian carotid b o d y it was decided to apply the technique of a u t o r a d i o g r a p h y to the n o r m a l physiological process of axoplasmic flow. T o the best o f our knowledge this has been attempted in only two investigations of the m a m m a lian carotid bodyT, 16, a n d has n o t previously been reported at all for the avian carotid b o d y except for a p r e l i m i n a r y a c c o u n t of the present experiments a n d a related doctoral thesis18,19.
* Present address: Department of Anatomy, Faculty of Veterinary Science, University of Khartoum, P.O. Box 32, Khartoum North, Sudan. ** To whom reprint requests should be addressed. 0006-8993/83/0000-0000/$03.00 © 1983 Elsevier Science Publishers
194 MATERIALS AND METHODS
Birds Eleven adult domestic hens (Gallus gallus var. domesticus) weighing 1.5-2 kg were used in this study. All were of the Golden Hubbard Comet strain. The carotid body in this species of bird has been shown to receive almost its entire innervation from the carotid body nerve of the distal vagal ganglion2,3. The anatomical nomenclature follows the Nornina Anatomica Avium.
The radioisotope Tritiated [4,5-3H]L-leucine (72 Ci/mmol, 1 mCi/ml; Amersham) was immediately freeze-dried after its arrival, and then redissolved in avian Ringer's solution to give a volume of 0.1 ml. It was used within 24 h.
Surgical exposure of the distal vagal ganglion Each bird was anaesthetized initially by slowly injecting 2-3 ml of Sagatal (M and B) into the left or right brachial vein. The opposite brachial vein was cannulated and a deep level of anaesthesia was continuously maintained with i.v. urethane. An incision, about 8 cm long, was made over the shoulder joint and the adjacent region of the scapula. The neck of the scapula was revealed by blunt dissection of the muscles covering it. The bone was then cut and the two stumps were separated by a retractor. The clavicular air sac was exposed by this procedure and cleared by blunt dissection. The brachial plexus was cut by a pair of fine scissors. This approach gave a clear view of the thyroid gland, jugular vein, and the vagus nerve itself. The distal vagal ganglion was identified by its topographical relations 2 and by being slightly swollen, very vascular, and lacking the cross-striations which are seen in the vagus nerve elsewhere.
Injection of the isotope Injection of either the left or right distal vagal ganglion was carried out under a dissecting microscope with a micropipette. The micropipette was introduced obliquely through the dorsal surface of the ganglion, to a depth of about 1 mm. The isotope was injected very slowly. A pledget of cotton wool soaked in avian Ringer's solution was placed on the
site of the operation to prevent dehydration, and the retractor was then removed. The birds were allowed to survive for 12-24 h under continuous deep anaesthesia before they were killed.
Fixation and processing for LM and EM autoradiography Fixation was carried out by perfusion with halfstrength Karnovsky's~° solution. The left and right distal vagal ganglia and carotid bodies were removed and routinely processed for electron microscopy. For light microscopic autoradiography 1 #m thick resin sections of the distal vagal ganglion were coated with Ilford K2 emulsion (diluted 1:1 with distilled water). They were exposed at 4 °C for periods of 7-21 days, developed in D-19 for 5 min, rinsed in distilled water for 30 s, and fixed in 30 sodium thiosulphate for 5 rain. They were finally rinsed in 3 changes of distilled water for 1 rain each, before being stained with a 1 ~ solution of toluidine blue. For electron microscopic autoradiography ultrathin sections of the carotid body were cut, collected on formvar-coated grids, and stained with lead citrate. They were then coated with Ilford L-4 emulsion (diluted 1:2 with distilled water) using the wire loop method of Caro and van Tubergen 5. The sections were exposed at 4 °C for periods of up to 6 months, developed in D-19 for 3 min at 20 °C, rinsed in distilled water for 20 s, fixed in 30 ~ sodium thiosulphate for 4 rain, and examined electron microscopically. The distal vagal ganglion and the carotid body of the uninjected side were used as controls.
Analysis of EM autoradiographs Qualitative observations were made on EM autoradiographs from the distal vagal ganglia and carotid bodies of all 11 birds. A quantitative analysis by point-counting morphometry15 was then performed on 5 of these birds, which were selected because the injected ganglion was proven by light microscopy to be heavily labelled. For each of these birds a stratified sampling technique was adopted. Three sections were taken randomly, at widely separated intervals, from the carotid body of the same side as that of the
Fig. 1. Light microscope autoradiograph of the distal vagal ganglion at the site of injection of tritiated leucine. A ganglion cell (G) is heavily labelled. Some myelinated axons are labelled (thick arrow) and some are not labelled (thin arrow). The labelled axons probably have their cell bodies in the ganglion itself whereas the unlabelled axons are probably passage fibres. Resin section, 1/~m thick, toluidine blue stain, x 950. Fig. 2. General view of the parenchyma of the carotid body after injecting the ipsilateral distal vagal ganglion with tritiated leucine. Labelling of the various components of the carotid body is non-uniform. Silver grains exclusively overlie the nervous tissue (asterisks). 1 -- type 1 cells; 2 ~ type 2 cell. x 15,000. Figs. 3 and 4. Capsular region of the carotid body, after injecting the ipsilateral distal vagal ganglion with tritiated leucine. A myelinated (M) and an unmyelinated axon (U) are labelled. The axonal profile (A), which is characterized by dense-cored granular vesicles, is unlabelled; this profile is interpreted as being either an amine or peptide containing ending. Fig. 3, x 10,000; Fig. 4, × 21,000.
196 injected ganglion. F r o m each section 5 autoradiographs were obtained from predetermined corners of the grid squares at a primary magnification of × 10,000 and enlarged to give a final magnification of x 25,000. This procedure yielded 75 micrographs from the 5 birds. A quadratic lattice grid with 306 points was superimposed over each print. The tissues of the carotid body were divided into axonal endings (expanded axonal profiles containing relatively numerous vesicles and mitochondria), axons in transit (small diameter axonal profiles characterized by neurotubules with few or no mitochondria and vesicles), and all non-nervous components (type 1 cells, type 2 cells, connective tissue, and blood vessels). The number of grid points, and the number of silver grains, falling on each component was counted. Where a grain overlaid two components the position of its centre was estimated and the grain was allocated to the component under the centre. The total numbers of labelled and unlabelled axons in transit, and axonal endings with and without synapses, were counted in the same 5 birds. RESULTS
Distal vagal ganglion In all 11 birds the light microscopic autoradiographs of the injected ganglion showed heavy concentrations of silver grains selectively overlying the ganglion cells (Fig. 1). Some of the myelinated axons were also labelled and others were not (Fig. 1).
Carotid body: qualitative observations In the EM autoradiographs of the carotid body, the silver grains were non-uniformly distributed
amongst the various structural components. In all 11 birds the vast majority of the grains fell upon the nervous tissues, especially the axonal endings (Fig. 2), only a few falling upon the non-nervous tissues. In the capsule, both myelinated and unmyelinated axons were labelled (Figs. 3, 4). The only type of axonal ending which constantly displayed no silver grains was that characterized by dense-cored granular vesicles (axonal ending A, in Fig. 4); only 7 such axons were observed out of a total of 207 axonal endings. The labelled axonal endings were either large caliciform endings or bouton-like endings. Some of these endings contained great numbers of small clear vesicles (Fig. 5), and others were well-filled with mitochondria (Fig. 2, top right-hand corner). In yet others there were numerous mitochondria in one region, and many clear vesicles in another region of the same ending (Fig. 6). Some of the labelled axonal endings apposed to type 1 cells showed synaptic complexes. O f these, some fulfilled the morphological criteria of afferent synapses transmitting from cell to axon (Figs. 7, 8). Others satisfied the morphological features of efferent synapses (Figs. 9-11). Finally, some of the synapses possessed the characteristics of reciprocal synapses (Fig. 12).
Quantitative observations The areas of the nervous and non-nervous components of the carotid body, and the incidence of silver grains upon each of these components, are summarized in Table I. The incidence of grains per unit area is 26 times greater on axonal endings than on the non-nervous components, and 15 times greater on axons in transit than on non-nervous components. These differences in the number of
Fig. 5. Parenchyma of carotid body after injecting the ipsilateral distal vagal ganglion with tritiated leucine. A labelled, bouton-like, axonal ending (AE) is apposed to a type 1 cell (1). The ending shows numerous small clear vesicles, some dense-cored vesicles, and a few mitochondria. Morphologically this ending could have been interpreted as an efferent ending (cholinergic), but the labelling proves that it is in fact afferent. Fig. 6. Parenchyma of carotid body after injecting the ipsilateral distal vagal ganglion with tritiated leucine. A labelled axonal ending (AE) is extensively apposed to a type 1 cell (1). The upper part of the ending is packed with mitochondria, whereas the lower part contains only a few mitochondria but numerous clear vesicles. A section passing through the plane A-B, would probably have given a profile packed with mitochondria which could have been interpreted (correctly, as the labelling shows) as afferent. A section through C-D would probably have yielded a profile containing many small clear vesicles, and this could have been interpreted (incorrectly) as an efferent ending. Figs. 7 and 8. Parenchyma of carotid body after injecting the ipsilateral distal vagal ganglion with tritiated leucine. In each Fig., a
197
labelled axonal ending (AE) is apposed to a Type 1 cell (1). A synapse (arrows, indicating direction of transmission is formed between each ending and the apposing cell. Small, intermediate, and large sized dense-cored granular vesicles (between the arrows) have accumulated on the presynaptic membrane (here, the cell side). The morphology of these synapses suggests transmission from cell to axon, the axon being afferent, i.e. projecting centrally. The labelling of these endings confirms that they were afferent. Fig. 7, x 39,000; Fig. 8, × 27,000.
Figs. 9-11. Parenchyma of carotid body after injecting the ipsilateral distal vagal ganglion with tritiated leucine. Each of the 3 labelled axonal endings (AE) is apposed to a type 1 cell (1). A synapse (arrows) is formed between each ending and its apposed type 1 cell. Clear vesicles have accumulated on the presynaptic membrane (here, the axon side). The postsynaptic membrane is thicker than the presynaptic membrane, and shows no accumulation of either clear or dense-cored vesicles. On the basis of synaptic morphology these 3 endings could have been interpreted as efferent endings (transmitting in the direction of the arrows, i.e. from axon to cell), but since each ending is labelled it must have been afferent. Fig. 9, × 41,500; Fig. 10, x 38,000; Fig. 11, × 26,500. Fig. 12. Parenchyma of carotid body, after injecting the ipsilateral distal vagal ganglion with tritiated leueine. A labelled axonal ending (AE) is apposed to a type 1 cell (1). A reciprocal synapse (arrows) is formed between the ending and the apposing type 1 cell. The labelling shows this ending to be afferent, x 42,500.
199 TABLE I Areas and grain counts o f nervous and non-nervous components o f the carotM body
The data were obtained from 75 EM autoradiographs from the labelled carotid bodies of 5 birds. Area o f autoradiographs
Structural component
Axons in transit Axonal endings Non-nervous components
%
m~
2.06 19.41 78.53
0.042 0.396 1.602
No. o f grains
No. o f grainsper m 2
57 915 143
1357 2311 89
TABLE II Labelled and unlabelled axonal elements
The data were obtained from75 EM autoradiographs from the labelled carotid bodies of 5 birds. 'Efferent' type synapses possessed accumulations of clear vesicles against the axonal membrane, and thickened type 1 cell membrane (Figs. 9-11). 'Afferent' type synapses showed accumulations of dense-cored granular vesicles against the type 1 cell membrane (Figs. 7, 8). Reciprocal synapses comprised 'afferent' and 'efferent' type structures in the same axonal ending (Fig. 12). The 7 axonal endings (all unlabelled) which contained dense-cored granular vesicles are excluded from the Table. Axons in transit
Labelled Unlabelled Percent labelled
36 175 17
Axonal endings o f all types
Axonalendingswithsynapses 'Efferent' type
'Afferent' type
Reciprocal
154 46 77
13 1
9 2
4 0
grains per unit area between axons in transit or axonal endings on the one hand, and non-nervous components on the other, were highly significant (P < 0.001). The numbers of labelled and unlabelled axons in transit, and axonal endings with and without synapses, are summarized in Table II. In all 5 birds together 77 ~ of the axonal endings were labelled, but in the individual birds the value ranged from 68 ~ to 88 ~ . C o n t r o l tissues
The control ganglia and carotid bodies always showed negligible labelling. DISCUSSION The highly significant difference between the degree of labelling of the nervous and non-nervous components of the carotid body on the injected side, and the negligible labelling of all control tissues,
showed that the radioactive tracer had reached the carotid body on the injected side solely from the injected distal vagal ganglion. In the domestic fowl this ganglion has been shown to contain exclusively afferent nerve cell bodies ~°, although many efferent axons pass through it. It is widely accepted that such axons in passage take up and incorporate into protein very little of an injected amino acid 6,7, and that whatever is so taken up is incorporated mostly into the soluble fraction 13 and is subsequently lost during fixation 8. Consequently axons passing between nerve cell bodies are believed to contribute little or nothing to the subsequent axonal transport of labelled material 6,7. It is therefore concluded that the labelled axonal elements observed in the carotid body in the present investigation belonged to afferent nerve fibres with their cell bodies in the distal vagal ganglion. Since 77 ~ of the axonal endings were labelled, it is clear that the axonal endings in the avian carotid body are predominantly afferent in function. The
200 method of allocating the silver grains was rigorous, in that some grains partly overlaid axonal endings, but were allocated to adjacent non-nervous components according to the position of the centre of the grain. This makes no allowance for the fact that an emission from an axonal ending may be oblique and induce a silver grain somewhat outside the axonal profile. Moreover the distal vagal ganglion in birds is very elongated; therefore a single injection of isotope could have led to its uneven distribution amongst the ganglion cells, as has been suggested 16 for the relatively compact petrosal ganglion of the cat. For all these reasons the 77 ~ of labelled axonal endings is almost certainly a substantial underestimate of the proportion of afferent axonal endings. A further indication of this is shown by one bird in which 88~ of the axonal endings were labelled. Indeed only one type of axonal ending (characterized by dense-cored granular vesicles) was never labelled, and such endings formed only 3.4 of all the observed endings. It is noteworthy that a very small number of axonal endings of this type were the only ones which survived removal of the distal vagal ganglion in the same species of bird by Abdel-Magied and KingL Of the axons in transit, only 17 ~ were labelled in the present study. Since the birds survived only 12-24 h after injecting the isotope, the incorporated amino acid had presumably reached the carotid body by the fast component of axonal transport; this has been reported to be 250-350 mm/day in the chicken (see LaseklZ), and therefore any label travelling in the fast component in our experiments could have been largely cleared from the axons in transit and concentrated in the axonal endings by the time that the animals were killed. On the other hand, the slow component of transport (about 1 ram/day13) would not have had time to reach the carotid body. The low proportion of labelled axons in transit which we observed is therefore not inconsistent with the degeneration experiments by AbdelMagied and Kings in which almost all the axons in transit (as well as axonal endings) degenerated after distal vagal ganglionectomy. We conclude that the innervation of the carotid body of the domestic fowl is almost entirely afferent. This confirms the conclusion reached by AbdelMagied and King3 from degeneration experiments
in the same species of bird. The evidence for the almost exclusively afferent innervation of the carotid body in the two highest vertebrates, i.e. in both birds and mammals, has now been so consistently demonstrated by so many different experimental techniques that there seems no longer any justification for regarding the matter as controversial. Moreover, the concept of transganglionic degeneration offers a reasonable explanation~,9,17 for the controversial degeneration results obtained by Biscoe et al. 4 after intracranial section of the glossopharyngeal nerve of the cat. Since autoradiographic labelling of axonal endings is achieved by the normal physiological process of axoplasmic flow, without disturbing neuronal structure, this technique offers a reliable way of checking the ultrastructural characteristics of afferent axonal endings. It has been suggestedal, 14 that efferent axonal endings in the carotid body are closely packed with small clear vesicles about 35-50 nm in diameter. On this basis the axonal ending in Fig. 5 would have been interpreted as efferent, but nevertheless it must have been afferent because it is labelled. There has been less agreement about the general ultrastructure of afferent endings, but one view is that such endings are packed with small mitochondria12. The axonal profile in Fig. 6 shows the dangers of using these criteria to distinguish between efferent and afferent endings. One end of this profile is filled with mitochondria, so that a transverse section through A-B could have been packed with mitochondria and interpreted (correctly) as afferent. At the other end, a section through C-D would be expected to yield a profile containing many small clear vesicles, and this could have been interpreted (incorrectly) as efferent. Synaptic morphology has been used with relative confidence to identify afferent and efferent axonal endings, utilizing the accumulation of vesicles against the presynaptic membrane and the often asymmetrical pre- and postsynaptic membranes 11. The 'afferent' type of synapse (Figs. 7, 8) and the reciprocal synapses (Fig. 12) observed in this investigation were nearly all labelled (Table II) and therefore belonged to afferent endings. Of the 13 'efferent' type synapses, 12 occurred in labelled endings (Figs. 9-11) and therefore belonged to afferent axons. This observation seems to call into
201 question the reliability of the classical concepts of synaptic morphology. Alternatively, it is j u s t conceivable that these synapses were the efferent
ACKNOWLEDGEMENTS
c o m p o n e n t s of reciprocal synapses, a possibility which was recognized b y K i n g et al. 11.
H e n r y with electron microscopy, J. G e a r y with
O n the basis o f these observations we conclude that purely m o r p h o l o g i c a l criteria are n o t reliable for distinguishing between afferent a n d efferent axonal endings, at least in the carotid body.
REFERENCES 1 Abdel-Magied, E. M., Observations on Chemoreceptor and Baroreceptor Structures in the Carotid Body Region of the Domestic Fowl, P h . D . Thesis, 1979, University of Liverpool. 2 Abdel-Magied, E. M. and King, A. S., The topographical anatomy and blood supply of the carotid body region of the domestic fowl, J. Anat., 126 (1978) 535-546. 3 Abdel-Magied, E. M. and King, A. S., Effects of distal vagal ganglionectomy and midcervical vagotomy on the ultrastructure of axonal elements in the carotid body of the domestic fowl, J. Anat., 134 (1982) 643-652 4 Biscoe, T. J., Lall, A. and Sampson, S. R., Electron microscopic and electrophysiological studies on the carotid body following intracranial section of the glossopharyngeal nerve, J. PhysioL (Lond.), 208 (1970) 133-152. 5 Caro, L. G. and van Tubergen, R. P., High resolution autoradiography, I. Methods, J. Cell Biol., 15 (1962) 173-188. 6 Cowan, W. M., Gottlieb, D. I., Hendrickson, A. E., Price, J. L. and Woolsey, T. A., The autoradiographic demonstration of axonal connections in the central nervous system, Brain Research, 37 (1972) 21-51. 7 Fidone, S. J., Zapata, P. and Stensaas, L. J., Axonal transport of labeled material into sensory nerve endings of cat carotid body, Brain Research, 124 (1977) 9-28. 8 Hendrickson, A. E., Electron microscopic distribution of axoplasmic transport, J. comp. NeuroL, 144 (1972) 381-398. 9 Hess, A. and Zapata, P., Innervation of the cat carotid body: normal and experimental studies, Fed. Proc., 31 (1972) 1365-1382. 10 Karnovsky, M. J., A formaldehyde/glutaraldehyde fixative of high osmolarity for use in electron microscopy,
W e gratefully acknowledge the assistance of Julie photography, a n d C a r o l y n Roberts with typing. Dr. M. A. Williams gave valuable advice o n the interp r e t a t i o n of autoradiographs. We t h a n k The British C o u n c i l a n d University of K h a r t o u m for m a k i n g our c o l l a b o r a t i o n possible.
J. Cell Biol., 27 (1965) 137-138a. 11 King, A. S., King, D. Z., Hodges, R. D. and Henry, J., Synaptic morphology of the carotid body of the domestic fowl, Cell Tiss. Res., 162 (1975) 459-473. 12 King, A. S., McLelland, J., Cook, R. D., King, D. Z. and Walsh, C., The ultrastructure of afferent nerve endings in the avian lung, Respir. Physiol., 22 (1974) 21-40. 13 Lasek, R. J., Protein transport in neurons, Int. Rev. Neurobiol., 13 (1970) 289-324. 14 McDonald, D. M. and Mitchell, R. A., The innervation of glomus cells, ganglion cells and blood vessels in the rat carotid body: a quantitative ultrastructural analysis, J. Neurocytol., 4 (1975) 177-230. 15 Ross, R. and Benditt, E. P., Quantitative electron microscope radioautographic observations of proline-Hz utilization by fibroblasts, J. Cell Biol., 27 (1965) 83-106. 16 Smith, P. G. and Mills, E., Autoradiographic identification of the terminations of petrosal ganglion neurons in the cat carotid body, Brain Research, 113 (1976) 174-178. 17 Smith, P. G. and Mills, E., Time course of Wallerian degeneration in the carotid body after carotid sinus nerve transection. In C. Belmonte, D. J. Pallor, H. Acker and S. Fidone (Eds.), Arterial Chemoreceptors, Proceedings of the 111International Meeting, University Press, Leicester, 1981, pp. 430-439. 18 Taha, A. A. M., Autoradiographic evidence for the afferent innervation of the carotid body in Gallus domesticu~, J. Anat., 133 (1981) 134. 19 Taha, A. A. M., Functional Anatomy of Axons and Granular Cells of the Carotid Body, Aorta, and Pulmonary Arteries in the Domestic Fowl, P h . D . Thesis, 1981, University of Liverpool. 20 Wakley, G. K. and Bower, A. J., The distal vagal ganglion of the hen (Gallus domesticus), a histological and physiological study, J. Anat., 132 (1981) 95-105.