Brain Research, 136 (1977) 345-350 © Elsevier/North-Holland Biomedical Press
345
A specific 'axo-axonal' interneuron in the visual cortex of the rat
P. SOMOGYI 1st Departmentof Anatomy, Faculty of Medicine, Semmelweis University, H-1450 Budapest (Hungary)
(Accepted July 7th, 1977)
Synaptic contacts located on the axon hillock and the initial axon segment of pyramidal neurons have been observed already by Palay et alp and have been described in more detail by Peters et al. 1°. Nothing, so far, is known about the origin and possible significance of these axo-axonic contacts. More recently we have come across a specific type of interneuron in the visual cortex of the rat that, due to the abundance in which it appears to occur in materials stained with our Golgi procedure, and due to its peculiar axon terminals, was deemed to be an interesting subject for a combined Golgi and EM study. The technique used will be reported in detail in a paper dealing with another major problem of cortical synaptology that can be solved with such a combined approach. It might suffice, therefore, to mention only very briefly that adult albino rats of our local breed were perfused through the heart with a solution of 1.25 % glutaraldehyde and 2 ~ paraformaldehyde in 0.1 M sodium phosphate buffer (pH 7.2-7.4) at room temperature for 30 min. The entire primary visual cortex was removed according to the stereotaxic coordinates given by Schober and Winkelmann 11. The tissue blocks were further immersed in the same fixation fluid and postosmicated in 2 % osmium tetroxide solution after 2 × 1 h washing in the buffer. The blocks were then immersed in 4 % potassium dichromate solution for one to two days and were transferred into a 0.75 % solution of silver nitrate. Sections of 100/~m thickness were cut with the aid of a Sorvall TC-2 tissue chopper and were mounted after dehydration in Durcupan. Suitably stained areas were selected and re-embedded for electron microscopy. Serial thin sections were mounted on formvar coated single-slot grids. The contrast was enhanced by 'en-bloc' staining with uranyl acetate and of the sections with lead citrate. A continuous light microscopic control enabled us to follow the serial sectioning, so that not only Golgi-stained cells and their dendrites but also various axon branches and individual terminal boutons could be easily identified at the EM level. The specific interneuron that caught our interest is a cell that resembles slightly the so-called 'chandelier cell' of Szentfigothai la,14 but even more the interneuron type 4 of Jones 6 in the monkey somatosensory cortex. Such cells have been found in our material mainly in layers I[ and III, they are of medium size and ovoid shape with a
347 vertically oriented ascending and descending dendritic arborization. The ascending arbor reaches layer I, where its branches turn into tangential direction, the descending reaches lamina IV. The dendrites are covered with a moderate number of delicate drumstick-shaped spines. The axon arises from the lower part of the body or from one of the major descending dendrites, and branches profusely after a short descending course. What is remarkable, are the somewhat coarser terminal portions of the axons, consisting of a single row of vertically oriented beads of 1-2/~m size. The length of these terminal portions is about 10-20/~m, or even more and consists of 5-7 beads on the average (2-12). The Golgi picture of Figs. 1A and 3 shows the principal feature of this cell type. Eleven such vertical beaded terminal portions, belonging to two different cells have been traced, so far, in EM series, all of them (including each terminal enlargement) exactly identified with light microscope photographs of the Golgi stain (Fig. 1A, C and D). Without a single exception these beads established contacts with the initial segments of pyramid cell axons, and none of the beads showed any evidence of a synaptic contact with any other neural structure (dendrite or cell body). The pyramid cell nature of the recipient neurons could be established partly on the basis of the apical dendrite and partly on the basis of the characteristic site of origin of the axon (Fig. 1C). The initial axon segments could be identified by tracing them from their origins in the uninterrupted EM section series, and on the basis of the generally accepted criteria of (i) an undercoat of the axon membrane of specific structure 2,9, (ii) the characteristic bundling of microtubules s, (iii) stacked flattened endoplasmic cisterns, described by Peters et al. 10 as being characteristic for pyramid axon initial segments. The dilatation of each beaded terminal portion established synapses with the same initial axon segment, however, in addition, several non-stained boutons established synapses with the same initial segment. This offered a unique opportunity to study the presynaptic terminals of this type, which are otherwise obscured by the Golgi precipitate (Fig. 2). Although it is possible that other cell types form occasional synapses with pyramid axon initial segments, it seems likely that the impregnated and at least the majority, if not all of the unstained boutons, belong to the same cell type, since their size is similar, non-impregnated boutons are interconnected along the initial segment and the characteristics of the synaptic loci are the same. The synaptic vesicles of the unstained terminals were usually flattened or at least pleomorphic (which is
Fig. 1. A: photomontage of Golgi-stained specific interneuron. Large curved arrow shows the site where the axon originates. Characteristic vertically oriented terminal axon portions are indicated by full arrows. Outline arrow indicates terminal portion shown in electron micrographs C, D and E. B: vertical intertwinement of several vertical terminal axon portions of simultaneously stained cells of similar character. C and D: low power electron micrograph of pyramid cell, the origin of the apical dendrite is indicated by thick arrows. Curved arrow showsaxon hillock, and the small arrows indicate the initial segment of the axon that could be traced in an uninterrupted section series (see D, the upper margin of which is continuous with the lower margin of C). Area indicated in D is shown at larger magnification in E. The synaptic contact is indicated by outline arrow, and below this a stacked membrane system becomes visible. This presynaptic bouton is the uppermost enlargement seen in the light microscope picture. Other parts of this terminal portion appear as black patches in Fig. D. Scale in A and B = 10#m; in C and D = 1/~m, in E = 0.25/~m.
348
Fig. 2. Two synaptic boutons on pyramid cell initial axon segment, one stained with Golgi method and was also identified as part of a beaded axon terminal portion. The synaptic contact of the impregnated bouton is shown by outline arrow. Series of arrows at lower right point to characteristic undercoat of initial axon segment. Non-stained bouton (asterisk) contains pleomorphic flattened synaptJc vesicles and has two small patches of synaptic membrane specializations (large arrows). Scale -- 0.25 pro. slightly at variance with the description o f Peters et al. 10, however, the shape o f the vesicles varies considerably with the histotechnical procedure); the presynaptic densities were restricted to small patches, while the postsynaptic membrane specializations are barely visible. The synaptic cleft was slightly widened and showed an increased opacity as compared to non-synaptic intercellular clefts. There could be little d o u b t that these axo-axonic synapses more closely resembled G r a y ' s type il 4 and Uchizono's F-type t6, i.e. the structural characteristics very generally encountered in inhibitory synapses. The remarkable feature in this interneuron type is that it is specifically establishing axo-axonic synapses with pyramid cell axon initial segments. If inhibitory, this neuron type might secure an elegant mechanism o f output control, exercized by the convergence o f several neurons o f the same type upon each pyramid neuron, in the two hitherto identified cases o f synapses located on initial axon segments: the M a u t h n e r neuron 7 and the Purkinje cei15,8,12, these were shown on physiological basis to be inhibitory 1,3. The tissue space within which any given cell o f this type could exercise its function would be around 200/~m in the tangential direction. Similar cells have turned up, so far, only in laminae II and Ill, while Jones' type 4 interneuron is more widely distributed 6,15. The observation of the axo-axonic nature o f the synapses is at variance with the assumption o f Szentfigothai a3,14 on his so called 'chandelier' cell. On the basis o f the arrangement of the vertically oriented b o u t o n groups a r o u n d a cylinder of about 2 # m diameter, Szentfigothai 14 thought it likely that the terminals o f the 'chandelier' cell are arranged a r o u n d apical dendrite shafts o f pyramidal cells, where rows o f F-type boutons can readily be observed. The Golgi pictures o f terminal axon profiles studied in this work do indeed differ significantly from those considered characteristic for the 'chandelier' cell. The cell type 4 of
349
Fig. 3. Light micrograph of another 'axo-axonic' cell, in layer III of area 17. The pial surface is at the top of the picture. Note the vertically oriented spiny dendrites. Curved arrow indicates the origin of the descending axon. Beaded terminal axon segments (small arrows). One of them (arrowhead) is seen to originate from the main axon. Scale - 100 ~m.
Jones 6, however, seems to be identical with the cell that we are dealing with here. It remains, therefore, in question, w h e t h e r the v a r i o u s cell types having such vertically oriented b e a d e d t e r m i n a l synaptic p o r t i o n s are all having a x o - a x o n a l contacts or, conversely, whether different types o f cells with these characteristics d o exist. M o r e i m p o r t a n t still are questions concerning the afferent connections o f these cells, whether they receive specific sensory afferents on their descending dendrites or o t h e r d i s t a n t
350 afferents or, w h e t h e r their c o n t a c t s are m a i n l y o f local s o u r c e - - for e x a m p l e , p y r a m i d a x o n collaterals - - w h i c h l a t t e r case w o u l d be p a r t i c u l a r l y i n t e r e s t i n g as a s h o r t n e g a t i v e f e e d b a c k loop. All o f these q u e s t i o n s r e q u i r e f u r t h e r studies, w h i c h are p a r t l y in p r o g r e s s o r still in the stage o f p l a n n i n g . T h e a u t h o r is g r a t e f u l to D r . J. Szenfftgothai f o r his a d v i c e a n d e n c o u r a g e m e n t a n d to M i s s C. G i p p e r t for h e r e x c e l l e n t t e c h n i c a l a n d p h o t o g r a p h i c w o r k .
1 Andersen, P., Eccles, J. C. and Voorhoeve, P. E., Inhibitory synapses on somas of Purkinje cells in the cerebellum, Nature (Lond.), 199 (1963) 655 656. 2 Chan-Palay, V., The tripartite structure of the undercoat in initial segments of Purkinje cell axons, Z. Anat. Entwickl.-Gesch., 139 (1972) 1-10. 3 Furukawa, T. and Furshpan, E. J., Two inhibitory mechanisms in the Mauthner neurons of goldfish, J. NeurophysioL, 26 (1963) 140-176. 4 Gray, E. G., Electron microscopy of presynaptic organelles of the spinal cord, J. Anat.(Lond.), 97 (1963) 101-106. 5 HS.mori, J. and Szent~igothai, J., The Purkinje cell basket: ultrastructure of an inhibitory synapse, Acta BioL Acad. Sci. hung., 15 (1965) 465-479. 6 Jones, E. G., Varieties and distribution of non-pyramidal ceils in the somatic sensory cortex of the squirrel monkey, J. comp. NeuroL, 160 (1975) 205-268. 7 Nakajima, Y., Fine structure of the synaptic endings on the Mauthner cell of the goldfish, J. comp. Neurol., 156 (1974)375-402. 8 Palay, S. L., The structural basis for neural action. In M. A. B. Brazier (Ed.), Brain Function, Vol. H. RNA and Brain Function: Memory and Learning, UCLA Forum Med. Sci., Univ. Calif. Press, Los Angeles, 1964, pp. 69-108. 9 Palay, S. L., Sotelo, C., Peters, A. and Orkand, P. M., The axon hillock and the initial segment, J. Cell BioL, 38 (1968) 193-201. 10 Peters, A., Proskauer, C. C. and Kaiserman-Abramof, 1. R., The small pyramidal neuron of the rat cerebral cortex: the axon hillock and initial segment, J. Cell BioL, 39 (1968) 604-619. 11 Schober, W. und Winkelmann, E., Die visuelle Kortex der Ratte Cytoarchitektonik und stereotaktische Parameter, Z. mikrosk.-anat. Forseh., 89 (1975) 431-446. 12 Somogyi, P. and H~imori, J., A quantitative electron microscopic study of the Purkinje cell axon initial segment, Neuroscience, I (1976) 361-365. 13 Szent~.gothai, J. and Arbib, M.A., Conceptual models of neural organization, Neurosci. Res. Progr. Bull., 12 (1974) 307-510. 14 SzentS_gothai, J., The module-concept in cerebral cortex architecture, Brain Research, 95 (1975) 475-796. 15 T6mbSl, T., Golgi analysis of the internal layers (V-VI) of the cat visual cortex, Exp. Brain Res., Suppl. 1 (1976)292-295. 16 Uchizono, K., Characteristics of excitatory and inhibitory synapses in the central nervous system of the cat, Nature (Lgnd.), 207 (1965) 642-643.
345
A specific 'axo-axonal' interneuron in the visual cortex of the rat
P. SOMOGYI 1st Departmentof Anatomy, Faculty of Medicine, Semmelweis University, H-1450 Budapest (Hungary)
(Accepted July 7th, 1977)
Synaptic contacts located on the axon hillock and the initial axon segment of pyramidal neurons have been observed already by Palay et alp and have been described in more detail by Peters et al. 1°. Nothing, so far, is known about the origin and possible significance of these axo-axonic contacts. More recently we have come across a specific type of interneuron in the visual cortex of the rat that, due to the abundance in which it appears to occur in materials stained with our Golgi procedure, and due to its peculiar axon terminals, was deemed to be an interesting subject for a combined Golgi and EM study. The technique used will be reported in detail in a paper dealing with another major problem of cortical synaptology that can be solved with such a combined approach. It might suffice, therefore, to mention only very briefly that adult albino rats of our local breed were perfused through the heart with a solution of 1.25 % glutaraldehyde and 2 ~ paraformaldehyde in 0.1 M sodium phosphate buffer (pH 7.2-7.4) at room temperature for 30 min. The entire primary visual cortex was removed according to the stereotaxic coordinates given by Schober and Winkelmann 11. The tissue blocks were further immersed in the same fixation fluid and postosmicated in 2 % osmium tetroxide solution after 2 × 1 h washing in the buffer. The blocks were then immersed in 4 % potassium dichromate solution for one to two days and were transferred into a 0.75 % solution of silver nitrate. Sections of 100/~m thickness were cut with the aid of a Sorvall TC-2 tissue chopper and were mounted after dehydration in Durcupan. Suitably stained areas were selected and re-embedded for electron microscopy. Serial thin sections were mounted on formvar coated single-slot grids. The contrast was enhanced by 'en-bloc' staining with uranyl acetate and of the sections with lead citrate. A continuous light microscopic control enabled us to follow the serial sectioning, so that not only Golgi-stained cells and their dendrites but also various axon branches and individual terminal boutons could be easily identified at the EM level. The specific interneuron that caught our interest is a cell that resembles slightly the so-called 'chandelier cell' of Szentfigothai la,14 but even more the interneuron type 4 of Jones 6 in the monkey somatosensory cortex. Such cells have been found in our material mainly in layers I[ and III, they are of medium size and ovoid shape with a
347 vertically oriented ascending and descending dendritic arborization. The ascending arbor reaches layer I, where its branches turn into tangential direction, the descending reaches lamina IV. The dendrites are covered with a moderate number of delicate drumstick-shaped spines. The axon arises from the lower part of the body or from one of the major descending dendrites, and branches profusely after a short descending course. What is remarkable, are the somewhat coarser terminal portions of the axons, consisting of a single row of vertically oriented beads of 1-2/~m size. The length of these terminal portions is about 10-20/~m, or even more and consists of 5-7 beads on the average (2-12). The Golgi picture of Figs. 1A and 3 shows the principal feature of this cell type. Eleven such vertical beaded terminal portions, belonging to two different cells have been traced, so far, in EM series, all of them (including each terminal enlargement) exactly identified with light microscope photographs of the Golgi stain (Fig. 1A, C and D). Without a single exception these beads established contacts with the initial segments of pyramid cell axons, and none of the beads showed any evidence of a synaptic contact with any other neural structure (dendrite or cell body). The pyramid cell nature of the recipient neurons could be established partly on the basis of the apical dendrite and partly on the basis of the characteristic site of origin of the axon (Fig. 1C). The initial axon segments could be identified by tracing them from their origins in the uninterrupted EM section series, and on the basis of the generally accepted criteria of (i) an undercoat of the axon membrane of specific structure 2,9, (ii) the characteristic bundling of microtubules s, (iii) stacked flattened endoplasmic cisterns, described by Peters et al. 10 as being characteristic for pyramid axon initial segments. The dilatation of each beaded terminal portion established synapses with the same initial axon segment, however, in addition, several non-stained boutons established synapses with the same initial segment. This offered a unique opportunity to study the presynaptic terminals of this type, which are otherwise obscured by the Golgi precipitate (Fig. 2). Although it is possible that other cell types form occasional synapses with pyramid axon initial segments, it seems likely that the impregnated and at least the majority, if not all of the unstained boutons, belong to the same cell type, since their size is similar, non-impregnated boutons are interconnected along the initial segment and the characteristics of the synaptic loci are the same. The synaptic vesicles of the unstained terminals were usually flattened or at least pleomorphic (which is
Fig. 1. A: photomontage of Golgi-stained specific interneuron. Large curved arrow shows the site where the axon originates. Characteristic vertically oriented terminal axon portions are indicated by full arrows. Outline arrow indicates terminal portion shown in electron micrographs C, D and E. B: vertical intertwinement of several vertical terminal axon portions of simultaneously stained cells of similar character. C and D: low power electron micrograph of pyramid cell, the origin of the apical dendrite is indicated by thick arrows. Curved arrow showsaxon hillock, and the small arrows indicate the initial segment of the axon that could be traced in an uninterrupted section series (see D, the upper margin of which is continuous with the lower margin of C). Area indicated in D is shown at larger magnification in E. The synaptic contact is indicated by outline arrow, and below this a stacked membrane system becomes visible. This presynaptic bouton is the uppermost enlargement seen in the light microscope picture. Other parts of this terminal portion appear as black patches in Fig. D. Scale in A and B = 10#m; in C and D = 1/~m, in E = 0.25/~m.
348
Fig. 2. Two synaptic boutons on pyramid cell initial axon segment, one stained with Golgi method and was also identified as part of a beaded axon terminal portion. The synaptic contact of the impregnated bouton is shown by outline arrow. Series of arrows at lower right point to characteristic undercoat of initial axon segment. Non-stained bouton (asterisk) contains pleomorphic flattened synaptJc vesicles and has two small patches of synaptic membrane specializations (large arrows). Scale -- 0.25 pro. slightly at variance with the description o f Peters et al. 10, however, the shape o f the vesicles varies considerably with the histotechnical procedure); the presynaptic densities were restricted to small patches, while the postsynaptic membrane specializations are barely visible. The synaptic cleft was slightly widened and showed an increased opacity as compared to non-synaptic intercellular clefts. There could be little d o u b t that these axo-axonic synapses more closely resembled G r a y ' s type il 4 and Uchizono's F-type t6, i.e. the structural characteristics very generally encountered in inhibitory synapses. The remarkable feature in this interneuron type is that it is specifically establishing axo-axonic synapses with pyramid cell axon initial segments. If inhibitory, this neuron type might secure an elegant mechanism o f output control, exercized by the convergence o f several neurons o f the same type upon each pyramid neuron, in the two hitherto identified cases o f synapses located on initial axon segments: the M a u t h n e r neuron 7 and the Purkinje cei15,8,12, these were shown on physiological basis to be inhibitory 1,3. The tissue space within which any given cell o f this type could exercise its function would be around 200/~m in the tangential direction. Similar cells have turned up, so far, only in laminae II and Ill, while Jones' type 4 interneuron is more widely distributed 6,15. The observation of the axo-axonic nature o f the synapses is at variance with the assumption o f Szentfigothai a3,14 on his so called 'chandelier' cell. On the basis o f the arrangement of the vertically oriented b o u t o n groups a r o u n d a cylinder of about 2 # m diameter, Szentfigothai 14 thought it likely that the terminals o f the 'chandelier' cell are arranged a r o u n d apical dendrite shafts o f pyramidal cells, where rows o f F-type boutons can readily be observed. The Golgi pictures o f terminal axon profiles studied in this work do indeed differ significantly from those considered characteristic for the 'chandelier' cell. The cell type 4 of
349
Fig. 3. Light micrograph of another 'axo-axonic' cell, in layer III of area 17. The pial surface is at the top of the picture. Note the vertically oriented spiny dendrites. Curved arrow indicates the origin of the descending axon. Beaded terminal axon segments (small arrows). One of them (arrowhead) is seen to originate from the main axon. Scale - 100 ~m.
Jones 6, however, seems to be identical with the cell that we are dealing with here. It remains, therefore, in question, w h e t h e r the v a r i o u s cell types having such vertically oriented b e a d e d t e r m i n a l synaptic p o r t i o n s are all having a x o - a x o n a l contacts or, conversely, whether different types o f cells with these characteristics d o exist. M o r e i m p o r t a n t still are questions concerning the afferent connections o f these cells, whether they receive specific sensory afferents on their descending dendrites or o t h e r d i s t a n t
350 afferents or, w h e t h e r their c o n t a c t s are m a i n l y o f local s o u r c e - - for e x a m p l e , p y r a m i d a x o n collaterals - - w h i c h l a t t e r case w o u l d be p a r t i c u l a r l y i n t e r e s t i n g as a s h o r t n e g a t i v e f e e d b a c k loop. All o f these q u e s t i o n s r e q u i r e f u r t h e r studies, w h i c h are p a r t l y in p r o g r e s s o r still in the stage o f p l a n n i n g . T h e a u t h o r is g r a t e f u l to D r . J. Szenfftgothai f o r his a d v i c e a n d e n c o u r a g e m e n t a n d to M i s s C. G i p p e r t for h e r e x c e l l e n t t e c h n i c a l a n d p h o t o g r a p h i c w o r k .
1 Andersen, P., Eccles, J. C. and Voorhoeve, P. E., Inhibitory synapses on somas of Purkinje cells in the cerebellum, Nature (Lond.), 199 (1963) 655 656. 2 Chan-Palay, V., The tripartite structure of the undercoat in initial segments of Purkinje cell axons, Z. Anat. Entwickl.-Gesch., 139 (1972) 1-10. 3 Furukawa, T. and Furshpan, E. J., Two inhibitory mechanisms in the Mauthner neurons of goldfish, J. NeurophysioL, 26 (1963) 140-176. 4 Gray, E. G., Electron microscopy of presynaptic organelles of the spinal cord, J. Anat.(Lond.), 97 (1963) 101-106. 5 HS.mori, J. and Szent~igothai, J., The Purkinje cell basket: ultrastructure of an inhibitory synapse, Acta BioL Acad. Sci. hung., 15 (1965) 465-479. 6 Jones, E. G., Varieties and distribution of non-pyramidal ceils in the somatic sensory cortex of the squirrel monkey, J. comp. NeuroL, 160 (1975) 205-268. 7 Nakajima, Y., Fine structure of the synaptic endings on the Mauthner cell of the goldfish, J. comp. Neurol., 156 (1974)375-402. 8 Palay, S. L., The structural basis for neural action. In M. A. B. Brazier (Ed.), Brain Function, Vol. H. RNA and Brain Function: Memory and Learning, UCLA Forum Med. Sci., Univ. Calif. Press, Los Angeles, 1964, pp. 69-108. 9 Palay, S. L., Sotelo, C., Peters, A. and Orkand, P. M., The axon hillock and the initial segment, J. Cell BioL, 38 (1968) 193-201. 10 Peters, A., Proskauer, C. C. and Kaiserman-Abramof, 1. R., The small pyramidal neuron of the rat cerebral cortex: the axon hillock and initial segment, J. Cell BioL, 39 (1968) 604-619. 11 Schober, W. und Winkelmann, E., Die visuelle Kortex der Ratte Cytoarchitektonik und stereotaktische Parameter, Z. mikrosk.-anat. Forseh., 89 (1975) 431-446. 12 Somogyi, P. and H~imori, J., A quantitative electron microscopic study of the Purkinje cell axon initial segment, Neuroscience, I (1976) 361-365. 13 Szent~.gothai, J. and Arbib, M.A., Conceptual models of neural organization, Neurosci. Res. Progr. Bull., 12 (1974) 307-510. 14 SzentS_gothai, J., The module-concept in cerebral cortex architecture, Brain Research, 95 (1975) 475-796. 15 T6mbSl, T., Golgi analysis of the internal layers (V-VI) of the cat visual cortex, Exp. Brain Res., Suppl. 1 (1976)292-295. 16 Uchizono, K., Characteristics of excitatory and inhibitory synapses in the central nervous system of the cat, Nature (Lgnd.), 207 (1965) 642-643.
