Brain Research, 569 (1992) 341-347 Elsevier
341
BRES 24977
Increase in the number of axospinous synapses with segmented postsynaptic densities following hippocampal kindling Yuri Geinisman 1, Frank Morrell 2 and Leyla deToledo-Morrell 2'3 1Department of CMS Biology, Northwestern University Medical School, Chicago, 1L 60611 (U.S.A.) and Departments of 2Neurological Sciences' and 3Psychology, Rush Medical College, Chicago, IL 60612 (U.S.A.) (Accepted 10 September 1991) Key words: Synaptic plasticity; Synapse quantitation; Perforated synapse; Dentate gyrus
Kindling results from intermittent electrical stimulation of a local brain region and leads to a virtually permanent augmentation of synaptic responsiveness in the stimulated circuit. It has been hypothesized that an increase in the number of synapses may represent a structural basis for the enduring expression of synaptic plasticity following kindling, but such an alteration has not been demonstrated unequivocally. The present report provides evidence that hippocampal kindling is indeed accompanied by an increase in synaptic numbers. Young adult rats were kindled via medial perforant path stimulation and sacrificed 4 weeks after reaching a criterion of 5 generalized seizures. Stimulated but not kindled and implanted but not stimulated rats served as controls. Synapses were analyzed in the middle (MML) and inner (IML) molecular layer of the hippocampal dentate gyrus. Using the stereological disector technique, unbiased estimates of the number of synapses per neuron were differentially obtained for 3 morphological subtypes of perforated axospinous synapses characterized by a fenestrated, horseshoe-shaped or segmented postsynaptic density (PSD). A significant increase in synaptic numbers was found to selectively involve only those perforated synapses which are distinguished by a segmented PSD consisting of 2-5 discrete plates. This structural modification was restricted to the terminal synaptic field of stimulated axons (MML), but was not observed in an immediately adjacent synaptic field (IML) which was not directly stimulated during kindling. Since synapses distinguished by a segmented PSD may represent specialized synaptic contacts of an unusually high efficacy, a selective increase in their numbers is likely to provide a structural substrate of the augmented synaptic gain associated with kindling. Kindling may be r e g a r d e d as a dramatic and robust form of neural plasticity2°'23. The kindling effect is the result of intermittent, low-level electrical stimulation of a local brain area. Such stimulation, without any change in p a r a m e t e r s , leads to the progressive intensification of paroxysmal neural activity and gradual alterations in motor behavior which culminate in a generalized seizure 13'15'21'23'27. Once established, the latter response to
tained by examining the total synaptic population, there is a growing b o d y of evidence suggesting that the socalled perforated synapses may be especially important for synaptic plasticity (see ref. 4 for a review). For this reason, our previous work 9-11 was focused on differential quantitative analyses of axospinous p e r f o r a t e d synapses with a discontinuous postsynaptic density (PSD) 3'4'6'8'26'3° and n o n - p e r f o r a t e d ones with a contin-
the originally subconvulsive stimulus becomes virtually p e r m a n e n t , since it can be reliably e v o k e d many months or even years after cessation of stimulation 15'23'27. The
uous PSD. The relative p r o p o r t i o n of p e r f o r a t e d over n o n - p e r f o r a t e d synapses was found to be m a r k e d l y and significantly increased in kindled animals relative to controls 9-11 Since the category of p e r f o r a t e d axospinous
exceptionally enduring augmentation of synaptic responsiveness in the stimulated circuit e n g e n d e r e d by kindling and its d e p e n d e n c e on protein synthesis 2'19'24 as well as axonal transport 2z strongly suggest that the kindling process may entail structural synaptic modifications. Of special importance in this regard may be an increase in synaptic numbers which could account for the durable e n h a n c e m e n t of synaptic efficacy typical of kindling. Such a notion, however, cannot be s u p p o r t e d by earlier morphological studies 14'z8 which have shown that a significant increment in the n u m b e r of synapses does n o t occur following kindling. While these results were ob-
synapses is c o m p o s e d of different morphological subtypes 3'6'8'26'3°, it is possible that our finding could reflect an increase in the n u m b e r of a certain synaptic subtype(s). The present study was designed to test the validity of this supposition. We now report that hippocampal kindling is indeed associated with an increase in synaptic numbers and that this change selectively involves p e r f o r a t e d axospinous synapses distinguished by a segmented PSD. The material o b t a i n e d in our previous work 9'1° was analyzed in this study. E x p e r i m e n t a l procedures used, as
Correspondence: Y. Geinisman, Department of CMS Biology, Northwestern University Medical School, 303 East Chicago Avenue, Chicago, IL 60611, U.S.A.
342 well as the protocols of tissue preparation for electron microscopy and of synapse quantitation were described in detail earlier 9A°. Briefly, young adult male rats of the Fischer 344 strain were implanted chronically with stimulating electrodes into the right medial perforant path and randomly assigned to the following groups consisting of: (1) kindled animals stimulated twice a day (with 1-ms pulses at 60 Hz for 2 s at a current level that initially produced an afterdischarge of 10 s or less) to a criterion of 5 generalized seizures; (2) coulombic controls stimulated with the same total current as corresponding kindled rats, but with parameters (120 pulses at 2 Hz) that do not evoke kindling; or (3) unstimulated controls handled in the same manner as rats from the other two groups. There were 7 animals in each group. Three rats, one from each group, were treated as a cohort. They were coded and sacrificed 4 weeks after an animal belonging to the first group reached criterion. Following an intracardiac perfusion with paraformaldehyde-glutaraldehyde fixatives, the right hippocampal formation was dissected free and divided perpendicular to its long axis into blocks of 1 mm in thickness. Two blocks with their rostral faces 1.5 or 3.5 mm caudal to the septal pole of the hippocampal formation were treated with OsO 4, dehydrated and embedded in Araldite. All blocks were assigned code numbers, and their rostral face was trimmed down to prepare ultrathin sections, each one spanning the entire width of the molecular and granule cell layer (GCL) in the central segment (in the mediolateral direction) of the hidden blade of the hippocampal dentate gyrus. Two series of consecutive sections were prepared from each block, stained with uranyl acetate-lead citrate and mounted on slot grids. For neuronal counts, electron micrographs of the same GCL area were taken from the first, several intermediate and last sections of each series and enlarged photographically to a final magnification of 2500x. For synaptic counts, electron micrographs were obtained from the central zone (in relation to the hippocampal fissure and GCL) of the middle molecular layer (MML) and from the zone of the inner molecular layer (IML) located 20-50 ktm dorsal to the GCL. The same neuropil area of the MML or IML was photographed in all sections of each series and enlarged to a final magnification of 20,O00x. Synapses were examined in the neuropil of the two different subdivisions of the dentate gyrus molecular layer for the following reason. While the MML is predominantly innervated by axons of the medial perforant path (which form virtually all perforated axospinous synapses in the MML25'32), the IML does not receive perforant path fibers and its major source of innervation is
provided by the commissural and associational afferents (see ref. 29 for a review). In rats kindled via medial perforant path stimulation, only the MML is a directly activated synaptic field. Structural synaptic modifications associated with perforant path kindling could be expected to predominantly involve synapses in the MML. but not in the IML. On the other hand, possible generalized or systemic effects of kindling-induced seizures on synaptic morphology are likely to be similar in the MML and IML. Unbiased estimates of the number of synapses per neuron were obtained with the aid of the stereological disector technique m7'3~. For the sampling of neurons, the first and last of k+ I sections of each series were alternately used as reference and look-up sections of disectors. For synapse sampling, 12 reference sections were randomly selected from each series, a section immediately above a reference one being used as a took-up section of disectors. Only those neurons or synapses were sampled which had their nucleus or PSD observed in a reference section micrograph (within an area limited by an unbiased sampling frame~6), but not in a micrograph of a look-up section of a disector. The number of synapses per neuron (n/N) was calculated by the formula: n / N = [(~q-.,Y,A.k)/(~Q-.,~a)].(w/W), In this formula, the summation 2; is over all disectors of a series, Q- and q- indicate numbers of neurons and synapses sampled in an area A or a, respectively, w designates the width of the middle or inner third of the molecular layer, and W represents the width of the GCL. A final n/N estimate per animal was calculated by averaging n/N values derived from 4 section series. The data presented here for each individual animal were obtained by examining series of 25-73 (mean = 42~ consecutive ultrathin sections which were used to obtain samples of 35-90 (mean = 61) or 20-58 (mean = 39~ perforated axospinous synapses per 4509 or 4517 um 2 (average areas) of the MML or IML. respectively. Additionally, 53-96 (mean = 68) neurons were sampled per GCL average area of 24,839 um 2. Examination of electron micrographs of serial sections of the dentate molecular layer showed that perforated axospinous synapses can be subdivided according to the configuration of their PSDs into 3 morphological subtypes characterized by (1) a (enestrated PSD having a hole(s) in its plate, (2) a horseshoe-shaped P S D exhibiting a U-type configuration of the PSD plate or (3) a segmented P S D consisting of 2-5 separate PSD plates (Fig. 1). The shape of the PSD plate extending along the postsynaptic membrane was ascertained in each perforated synapse sampled by means of z planar PSD reconstruction from serial sections (Fig. la,b,c) as described by us earlier 8. The number of synapses per neuron was differentially estimated for perforated synaptic junctions
343
a
b
c
Fig. 1. Electron micrographs of serial sections demonstrating 3 axospinous synapses indicated by solid arrows, open arrows or pointing fingers, respectively. All synaptic profiles exhibiting the PSD are presented. In some profiles of each synapse, PSD discontinuities or perforations (solid arrowheads) are observed indicating that these synaptic contacts belong to the category of perforated junctions. Planar PSD reconstructions from the micrographs of serial sections show that the population of perforated axospinous synapses can be divided according to the PSD configuration into three morphologically distinct subtypes. They are characterized by a fenestrated PSD (a) exhibiting a hole or holes in its single plate, by a horseshoe-shaped PSD (b) having a single plate that assumes a U-type configuration or by a segmented PSD (c) consisting of separate plates. Bar = 0.25 /~m.
344 having a fenestrated, horseshoe-shaped or segmented
TABLE II
PSD (Fig. 1). Q u a n t i t a t i o n of various subtypes of perforated axos-
Number of various subtypes o f perforated axospinous synapses per neuron in the inner molecular layer o f the dentate gyrus of control and kindled rats
pinous synapses in the M M L showed that the n u m b e r of synaptic contacts with segmented PSDs per n e u r o n was significantly increased in kindled rats relative to either
Designations are the same as in Table I.
unstimulated or coulombic controls (Table I). In contrast, synapses with fenestrated and horseshoe-shaped PSDs did not change significantly in n u m b e r s following
Rat
kindling, although a tendency towards a decrease was observed (Table I). In the IML, however, all 3 synaptic subtypes exhibited noticeable trends towards a decrease in their n u m b e r per n e u r o n which did not reach statistical significance (Table II). The n u m b e r of synapses per n e u r o n is a parameter
Number o f various subtypes o f perforated axospinous synapses per neuron in the middle molecular layer o f the dentate gyrus o f control and kindled rats
Designations: n/N, number of synapses per neuron; A, difference between group means. (A) Unstimulated control n/N
(B) Coulombic control
n/N n/N
48
AC-A
AC-B
A B-C
Synapses with fenestrated PSDs 1 43 48 2 44 77 3 50 41 4 16 55 5 50 36 6 53 41 7 38 36 Per group 42
(C) Kindling
-7.1%
67
+8.1%
56
Synapses with segmented PSDs 1 79 121 2 109 86 3 81 71 4 87 69 5 76 100 6 127 132 7 77 61
117 116 145 135 112 101 164
Per group 91
127
91
0.0%
-9.7%
(C) Kindling ................. n/N AC-A
AC-B
3 26 8 30 19 16 32
+1111%19
-29.6%
-36.7%
-23.3%
-29.8%
-19.7%
-25.6%
Synapses with horseshoe-shaped PSDs 1 60 52 35 2 33 28 19 3 33 64 58 4 45 67 25 5 62 27 27 6 22 44 27 7 44 46 39 Per group 43
Per group 76
47
+9.3% 33
82
63 35 86 47 71 56 66
+7.9% 61
-18.8%
Synapses with horseshoe-shaped PSDs 65 1 52 81 41 2 67 87 58 3 56 65 61 4 47 63 59 5 72 46 45 6 84 78 64 7 55 52 Per group 62
30
Synapses with segmented PSDs 1 75 114 2 62 60 3 67 70 4 99 106 5 68 46 6 80 102 7 83 76
45 46 40 35 41 38 28
+14.3% 39
(B) Coulombic control .............. n/N A B-('
Synapses withfenestrated PSDs 1 19 51 2 33 21 3 33 26 4 38 19 5 21 9 6 21 51 7 27 35 Per group 27
ATABLE I
Rat
(A) Unstimulated control n/N
-t6.4%
+39.6%* +39.6%*
* P < 0.05 and less, two-tailed randomization test for two independent samples.
that depends not only on the synaptic numerical density, but also on the numerical density of neurons and on dimensions of n e u r o n a l and synaptic layers in the case of stratified structures such as the dentate gyrus. A direct measure of relative quantities of the synaptic subtypes u n d e r comparison in the neuropil is provided by their ratio which can be assessed from raw synaptic counts in the same disectors. Therefore. the ratio of synapses with segmented PSDs to other perforated axospinous synapses was additionally estimated. It was found to be significantly increased in the M M L of kindled rats as compared with controls (Table III). No significant change in this ratio was observed in the IML (Table III). A major finding of this study is that kindling is associated with an increase in the n u m b e r of synapses per n e u r o n which is limited to only those axospinous synaptic contacts that are distinguished by a segmented PSD (Ta-
345 TABLE III Ratio of synapses with segmented postsynaptic densities to other perforated synaptic junctions in the middle and inner molecular layer of the dentate gyrus of control and kindled rats
Designation: A, difference between group means. (A) Unstimulated control Ratio
Rat
(B) Coulombic control
Ratio A C - A Ratio
0.905
0.820
1.064 1.333 1.480 1.406 1.120 1.217 1.783 -9.4%
Inner molecular layer 1 0.949 1.107 2 0.939 1.224 3 1.015 0.778 4 1.193 1.233 5 0.819 1.278 6 1.860 1.074 7 1.169 0.938 Group mean
1.135
1.090
AC-B
A B-A
Middle molecular layer l 0.832 0.938 2 0.982 0.524 3 0.764 0.670 4 1.381 0.585 5 0.623 1.220 6 0.927 1.109 7 0.828 0.693 Group mean
(C) Kindling
1.343 +48.4%* +63.8%* 1.658 0.778 1.303 0.855 1.543 1.302 0.930
-4.0%
1.196 +5.4%
+9.7%
* P < 0.05 and less, two-tailed randomization test for two independent samples.
-z
6
"~+36
9
J
Fig. 2. Diagram illustrating the process of synapse turnover described in the text. Schematics are shown for a non-perforated axospinous synapse (designated by a value of -420) and for perforated ones with a fenestrated (-6), horseshoe-shaped (-9) or segmented (+36) PSD. Differences in the number of synapses per neuron between control and kindled animals are indicated for each synaptic subtype in the middle molecular layer of the dentate gyrus, the value for non-perforated synaptic contacts being taken from the results reported earlier91°. Since the two groups of control rats studied did not differ significantly from each other with respect to the number of all synaptic subtypes examined, the data obtained from unstimulated and coulombic controls were combined. Kindling-induced alterations in synaptic numbers, which are shown here, were determined relative to the combined control values. Statistically significant changes include the decrease in the number of non-perforated junctions and the increment in perforated synapses exhibiting a segmented PSD.
ble I). This restructuring of connectivity is highly selective, since other subtypes of perforated axospinous synapses characterized by a fenestrated or horseshoe-shaped PSD (Table I), as well as n o n - p e r f o r a t e d axospinous synapses 9,w and synaptic junctions involving dendritic shafts 9, do not exhibit such a change. M o r e o v e r , the observed structural synaptic modification is restricted to the terminal synaptic field of stimulated axons ( M M L ) and does not involve an immediately adjacent synaptic field (IML) which was not directly activated during kindling (Table II). Such anatomical selectivity cannot be accounted for by non-specific effects of generalized seizures and is m o r e consistent with the notion that the observed structural synaptic modification is indeed a direct consequence of the kindling process. The kindling-induced increment in the n u m b e r of axospinous synaptic junctions exhibiting a segmented PSD may be due to a remodelling of some pre-existing axospinous synapses or to synaptogenesis. The former possibility has been p r o p o s e d by the existing models of synapse turnover and splitting4,5,7,es. It has been postulated 5'7'25 that axospinous synapses with a segm e n t e d PSD are constantly formed from non-perforated ones through the stages of p e r f o r a t e d synaptic junctions having first a fenestrated, then a horseshoe-shaped and finally a segmented PSD; eventually, a synapse of the segmented subtype splits into n o n - p e r f o r a t e d synaptic contacts (Fig. 2). The alternative hypothesis, namely, new synapse formation is equally likely and may be supp o r t e d by the discovery that perforant path kindling is accompanied by sprouting of mossy fibers and their terminals in the inner molecular layer of the rat dentate gyrus 34. This finding cannot, of course, account for our data, since the increase in the n u m b e r of segmented synapses was found to be restricted to the middle molecular layer. Unfortunately, the results of synapse quantitation presented here (Fig. 2) are insufficient to resolve the question of whether an intensification of synapse turnover or an induction of synaptogenesis is responsible for the kindling-induced increase in the n u m b e r of synaptic contacts with segmented PSDs. The total synaptic population of the M M L in control, unstimulated animals is c o m p o s e d of axodendritic (6.5%), p e r f o r a t e d axospinous (8.6%) and nonp e r f o r a t e d axospinous (85.0%) synapses 9. Those perforated synaptic contacts that have a segmented PSD consititute only a small p r o p o r t i o n (4.0%) of the M M L synaptic population. Nevertheless, the fact that this particular subtype of p e r f o r a t e d axospinous synapses is selectively increased in numbers following kindling is especially interesting. Such synaptic junctions exhibit two to five discrete PSD plates, rather than only one present in all other synapses. Coexistent with a PSD plate is a
346 specialized presynaptic m e m b r a n e area a p p o s e d by clusters of synaptic vesicles which constitute together a region within the synapse known as the active zone tS. Presynaptic secretion and postsynaptic reception of substances mediating chemical impulse transmission are believed to occur within this region TM. The availability of additional transmission sites within a single synapse would be expected to enhance its strength, especially if each such site is associated with a r e c e p t o r cluster. Therefore, an increase in the n u m b e r of synaptic contacts having s e g m e n t e d PSDs m a y provide a mechanism for the enduring amplification of synaptic responsiveness in the stimulated circuit which characterizes kindling. Of special significance is the fact that the structural synaptic modification r e p o r t e d here is not unique for the kindling p h e n o m e n o n . A selective increase in the number of axospinous synapses exhibiting s e g m e n t e d PSDs has also been d e m o n s t r a t e d by us to occur in the M M L of the d e n t a t e gyrus after the induction of long-term potentiation (LTP) via medial p e r f o r a n t path stimulation 12. LTP is a n o t h e r m o d e l of synaptic plasticity which is closely related to kindling 23. E a r l y stages of kindling e v o k e d by stimulation of the pefforant path are invariably a c c o m p a n i e d by potentiation at perforant p a t h -
The authors are thankful to William Goossens, Nicholas Kriho and Daniel Lanciloti for their skillful technical assistance and to Dr. L.T. Thompson for preparing Fig. 2. This work was supported by grants AG 08794 from NIA and BNS 88-19902 from NSE
1 Br~endgaard, H. and Gundersen, H.J.G., The impact of recent stereological advances on quantitative studies of the nervous system, J. Neurosci. Methods, 18 (1986) 39-78. 2 Cain, D.P., Corcoran, M. and Staines, W., Effects of protein synthesis inhibition on kindling in the mouse, Exp. Neurol., 68 (1980) 409-416. 3 Calverley, P.K.S. and Jones, D.G., A serial section study of perforated synapses in rat neocortex, Cell Tiss. Res., 247 (1987) 565-572. 4 Calverley, P.K.S. and Jones, D.G., Contribution of dendritic spines and perforated synapses to synaptic plasticity, Brain Res. Rev., 15 (1990) 215-249. 5 Carlin, P.K. and Siekevitz, P., Plasticity in the central nervous system: do synapses divide?, Proc. Natl. Acad. Sci. U.S.A., 80 (1983) 3517-3521. 6 Cohen, R.S. and Siekevitz, P., Form of the postsynaptic density. A serial section study, J. Cell Biol., 78 (1978) 36-46. 7 Dyson, S.E. and Jones, D.G., Synaptic remodelling during development and maturation: junction differentiation and splitting as a mechanism of modifying connectivity, Dev. Brain Res., 13 (1984) 125-137. 8 Geinisman, Y., Morrell, E and deToledo-Morrell, L., Axospinous synapses with segmented postsynaptic densities: a morphologically distinct synaptic subtype contributing to the number of profiles of 'perforated' synapses visualized in random sections, Brain Research, 423 (1987) 179-188. 9 Geinisman, Y., Morrell, F. and deToledo-Morrell, L., Remodeling of synaptic architecture during hippocampal 'kindling', Proc. Natl. Acad. Sci. U.S.A., 85 (1988) 3260-3264. 10 Geinisman, Y., Morrell, F. and deToledo-Morrell, L., Increase in the relative proportion of perforated axospinous synapses following hippocampal kindling is specific for the synaptic field of stimulated axons, Brain Research, 507 (1990) 325-331. 11 Geinisman, Y., Morrell, F. and deToledo-MorreU, L., Alterations of synaptic ultrastructure induced by hippoeampal kindling. In J.A. Wada (Ed.), Kindling 4, Plenum, New York,
1990, pp. 75-88. 12 Geinisman, Y., deToledo-Morrell, L. and Morrell, E, Induction of long-term potentiation is associated with an increase in the number of axospinous synapses with segmented postsynaptic densities, Brain Research, in press. 13 Goddard, G.V., The development of epileptic seizures through brain stimulation at low intensity, Nature, 214 (1967) 1020-1021. 14 Goddard, G.V. and Douglas, RM., Does the engram of kindling model the engram of normal long term memory?, Can. J. Neurol. Sci., 2 (1975) 385-394. 15 Goddard, G.V., Mclntyre, D. and Leech, C., A permanent change in brain function resulting from daily electrical stimulation, Exp. Neurol., 25 (1969) 295-;330. 16 Gundersen, H.J.G., Notes on the estimation of the numerical density of arbitrary profiles: the edge effect, J. Microsc., 111 (1977) 219-223. 17 Gundersen, H.J.G., Stereology of the arbitrary particles, J. Microsc., 143 (1986) 3-45. 18 Heuser, J.E. and Reese, T.S., Structure of the synapse. In E.R. Kandel (Ed.), Handbook of Physiology, Sect. 1, Vol. 1, Am. Physiol. Soc. and Williams & Wilkins, Bethesda, MD, 1977, 261-294. 19 Jonek, V. and Wasterlain, C.G, Effect of inhibitors of protein synthesis on the development of kindled seizures, Exp. Neurol, 66 (1979) 524-532. 20 Le Gal La Salle, G., Kindling as a model of long-lasting plasticity: new perspective for the search for molecular rneehanisms. In J.A. Wada (Ed.), Kindling 4, Plenum, New York, 1990, pp. 1-9. 21 McNamara, J.O., Byrne, M.C., Dasheiff, R.M. and Fitz, J.G.; The kindling model of epilepsy: a review, Prog. Neurobiol., 15 (1980) 139-159. 22 Morrell, F., Callosal mechanisms in epileptogenesis. In A. Reeves (Ed.), Epilepsy and the Corpus Callosum, Plenum, New York, 1985, pp. 99-130. 23 Morrell, F. and deToledo-Morrelk t,., Kindling as a model of
dentate granule cell synapses ~3. It is conceivable, therefore, that the remodelling of synaptic architecture observed in kindled animals may actually result from LTP elicited prior to the acquisition of the fully develo p e d kindled state. A n o t h e r possibility is that various models of synaptic plasticity share a similar mechanism of modifying synaptic connectivity. In fact, there are some indications that visual discrimination training may also be accompanied by an increase in the n u m b e r of synapses with segmented PSDs. In the visual cortex of trained animals, clusters of closely a p p r o x i m a t e d synaptic grids are seen more frequently than in controls suggesting that visual training m a y lead to focalization of dense projections and related p a r a m e m b r a n e o u s densities 35 one of which is the PSD. Thus, an increase in the n u m b e r of axospinous synapses distinguished by a segm e n t e d PSD may represent a structural substrate of kindling, LTP and o t h e r forms of synaptic plasticity associated with learning and memor~
347
24
25
26
27 28
29
neuronal plasticity, In J.A. Wada (Ed.), Kindling 3, Raven, New York, 1986, pp. 17-33. Morrell, E, Tsuru, N., Hoeppner, T.J., Morgan, D. and Harrison, W.H., Secondary epileptogenesis in the frog forebrain: effect of inhibition of protein synthesis, Can. J. Neurol. Sci., 2 (1975) 407-416. Nieto-Sampedro, M., Hoff, S.W. and Cotman, C.W., Perforated postsynaptic densities: probable intermediates in synapse turnover, Proc. Natl. Acad. Sci. U.S.A., 79 (1982) 5718-5722. Peters, A. and Kaiserman-Abramof, I.R., The small pyramidal neuron of the rat cerebral cortex. The synapses upon dendritic spines, Z. Zellforsch., 100 (1969) 487-506. Racine, R., Kindling: the first decade, Neurosurgery, 3 (1978) 234-252. Racine, R.J. and Zaide, J., A further investigation into the mechanisms of the kindling phenomenon. In K. Livingston and O. Hornykiewicz (Eds.), Limbic Mechanisms: The Continuining Evolution of the Limbic System Concept, Plenum, New York, 1978, pp. 457-493. Rosene, D.L., and Van Hoesen, G.W., The hippocampal formation of the primate brain. In E.G. Jones and A. Peters (Eds.), Cerebral Cortex, Vol. 6, Further Aspects of Cortical function, Including Hippocampus, Plenum, New York, 1987,
pp. 345-456. 30 Spacek, J., and Hartmann, M., Three-dimensional analysis of dendritic spines. I. Quantitative observations related to dendritic spina and synaptic morphology in cerebral and cerebellar cortices, Anat. Embryol., 167 (1983) 289-310. 31 Sterio, D.C., The unbiased estimation of number and sizes of arbitrary particles using the disector, J. Microsc., 134 (1984) 127-136. 32 Steward, O. and Vinsant, S., The process of reinnervation in the dentate gyrus of the adult rat: a quantitative electron microscopic analysis of terminal proliferation and reactive synaptogenesis, J. Comp. Neurol., 214 (1983) 370-386. 33 Sutula, T. and Steward, O., Quantitative analysis of synaptic potentiation during kindling of the perforant path, J. Neurophysiol., 56 (1986) 732-746. 34 Sutula, T., Xiao-Xian, H., Cavazos, J. and Scott, G., Synaptic reorganization in the hippocampus induced by abnormal functional activity, Science, 239 (1988) 1147-1150. 35 Vrensen, G. and Nunes Cardozo, J., Changes in size and shape of synaptic connections after visual training: an ultrastructural approach to synaptic plasticity, Brain Research, 218 (1981) 7997.
341
BRES 24977
Increase in the number of axospinous synapses with segmented postsynaptic densities following hippocampal kindling Yuri Geinisman 1, Frank Morrell 2 and Leyla deToledo-Morrell 2'3 1Department of CMS Biology, Northwestern University Medical School, Chicago, 1L 60611 (U.S.A.) and Departments of 2Neurological Sciences' and 3Psychology, Rush Medical College, Chicago, IL 60612 (U.S.A.) (Accepted 10 September 1991) Key words: Synaptic plasticity; Synapse quantitation; Perforated synapse; Dentate gyrus
Kindling results from intermittent electrical stimulation of a local brain region and leads to a virtually permanent augmentation of synaptic responsiveness in the stimulated circuit. It has been hypothesized that an increase in the number of synapses may represent a structural basis for the enduring expression of synaptic plasticity following kindling, but such an alteration has not been demonstrated unequivocally. The present report provides evidence that hippocampal kindling is indeed accompanied by an increase in synaptic numbers. Young adult rats were kindled via medial perforant path stimulation and sacrificed 4 weeks after reaching a criterion of 5 generalized seizures. Stimulated but not kindled and implanted but not stimulated rats served as controls. Synapses were analyzed in the middle (MML) and inner (IML) molecular layer of the hippocampal dentate gyrus. Using the stereological disector technique, unbiased estimates of the number of synapses per neuron were differentially obtained for 3 morphological subtypes of perforated axospinous synapses characterized by a fenestrated, horseshoe-shaped or segmented postsynaptic density (PSD). A significant increase in synaptic numbers was found to selectively involve only those perforated synapses which are distinguished by a segmented PSD consisting of 2-5 discrete plates. This structural modification was restricted to the terminal synaptic field of stimulated axons (MML), but was not observed in an immediately adjacent synaptic field (IML) which was not directly stimulated during kindling. Since synapses distinguished by a segmented PSD may represent specialized synaptic contacts of an unusually high efficacy, a selective increase in their numbers is likely to provide a structural substrate of the augmented synaptic gain associated with kindling. Kindling may be r e g a r d e d as a dramatic and robust form of neural plasticity2°'23. The kindling effect is the result of intermittent, low-level electrical stimulation of a local brain area. Such stimulation, without any change in p a r a m e t e r s , leads to the progressive intensification of paroxysmal neural activity and gradual alterations in motor behavior which culminate in a generalized seizure 13'15'21'23'27. Once established, the latter response to
tained by examining the total synaptic population, there is a growing b o d y of evidence suggesting that the socalled perforated synapses may be especially important for synaptic plasticity (see ref. 4 for a review). For this reason, our previous work 9-11 was focused on differential quantitative analyses of axospinous p e r f o r a t e d synapses with a discontinuous postsynaptic density (PSD) 3'4'6'8'26'3° and n o n - p e r f o r a t e d ones with a contin-
the originally subconvulsive stimulus becomes virtually p e r m a n e n t , since it can be reliably e v o k e d many months or even years after cessation of stimulation 15'23'27. The
uous PSD. The relative p r o p o r t i o n of p e r f o r a t e d over n o n - p e r f o r a t e d synapses was found to be m a r k e d l y and significantly increased in kindled animals relative to controls 9-11 Since the category of p e r f o r a t e d axospinous
exceptionally enduring augmentation of synaptic responsiveness in the stimulated circuit e n g e n d e r e d by kindling and its d e p e n d e n c e on protein synthesis 2'19'24 as well as axonal transport 2z strongly suggest that the kindling process may entail structural synaptic modifications. Of special importance in this regard may be an increase in synaptic numbers which could account for the durable e n h a n c e m e n t of synaptic efficacy typical of kindling. Such a notion, however, cannot be s u p p o r t e d by earlier morphological studies 14'z8 which have shown that a significant increment in the n u m b e r of synapses does n o t occur following kindling. While these results were ob-
synapses is c o m p o s e d of different morphological subtypes 3'6'8'26'3°, it is possible that our finding could reflect an increase in the n u m b e r of a certain synaptic subtype(s). The present study was designed to test the validity of this supposition. We now report that hippocampal kindling is indeed associated with an increase in synaptic numbers and that this change selectively involves p e r f o r a t e d axospinous synapses distinguished by a segmented PSD. The material o b t a i n e d in our previous work 9'1° was analyzed in this study. E x p e r i m e n t a l procedures used, as
Correspondence: Y. Geinisman, Department of CMS Biology, Northwestern University Medical School, 303 East Chicago Avenue, Chicago, IL 60611, U.S.A.
342 well as the protocols of tissue preparation for electron microscopy and of synapse quantitation were described in detail earlier 9A°. Briefly, young adult male rats of the Fischer 344 strain were implanted chronically with stimulating electrodes into the right medial perforant path and randomly assigned to the following groups consisting of: (1) kindled animals stimulated twice a day (with 1-ms pulses at 60 Hz for 2 s at a current level that initially produced an afterdischarge of 10 s or less) to a criterion of 5 generalized seizures; (2) coulombic controls stimulated with the same total current as corresponding kindled rats, but with parameters (120 pulses at 2 Hz) that do not evoke kindling; or (3) unstimulated controls handled in the same manner as rats from the other two groups. There were 7 animals in each group. Three rats, one from each group, were treated as a cohort. They were coded and sacrificed 4 weeks after an animal belonging to the first group reached criterion. Following an intracardiac perfusion with paraformaldehyde-glutaraldehyde fixatives, the right hippocampal formation was dissected free and divided perpendicular to its long axis into blocks of 1 mm in thickness. Two blocks with their rostral faces 1.5 or 3.5 mm caudal to the septal pole of the hippocampal formation were treated with OsO 4, dehydrated and embedded in Araldite. All blocks were assigned code numbers, and their rostral face was trimmed down to prepare ultrathin sections, each one spanning the entire width of the molecular and granule cell layer (GCL) in the central segment (in the mediolateral direction) of the hidden blade of the hippocampal dentate gyrus. Two series of consecutive sections were prepared from each block, stained with uranyl acetate-lead citrate and mounted on slot grids. For neuronal counts, electron micrographs of the same GCL area were taken from the first, several intermediate and last sections of each series and enlarged photographically to a final magnification of 2500x. For synaptic counts, electron micrographs were obtained from the central zone (in relation to the hippocampal fissure and GCL) of the middle molecular layer (MML) and from the zone of the inner molecular layer (IML) located 20-50 ktm dorsal to the GCL. The same neuropil area of the MML or IML was photographed in all sections of each series and enlarged to a final magnification of 20,O00x. Synapses were examined in the neuropil of the two different subdivisions of the dentate gyrus molecular layer for the following reason. While the MML is predominantly innervated by axons of the medial perforant path (which form virtually all perforated axospinous synapses in the MML25'32), the IML does not receive perforant path fibers and its major source of innervation is
provided by the commissural and associational afferents (see ref. 29 for a review). In rats kindled via medial perforant path stimulation, only the MML is a directly activated synaptic field. Structural synaptic modifications associated with perforant path kindling could be expected to predominantly involve synapses in the MML. but not in the IML. On the other hand, possible generalized or systemic effects of kindling-induced seizures on synaptic morphology are likely to be similar in the MML and IML. Unbiased estimates of the number of synapses per neuron were obtained with the aid of the stereological disector technique m7'3~. For the sampling of neurons, the first and last of k+ I sections of each series were alternately used as reference and look-up sections of disectors. For synapse sampling, 12 reference sections were randomly selected from each series, a section immediately above a reference one being used as a took-up section of disectors. Only those neurons or synapses were sampled which had their nucleus or PSD observed in a reference section micrograph (within an area limited by an unbiased sampling frame~6), but not in a micrograph of a look-up section of a disector. The number of synapses per neuron (n/N) was calculated by the formula: n / N = [(~q-.,Y,A.k)/(~Q-.,~a)].(w/W), In this formula, the summation 2; is over all disectors of a series, Q- and q- indicate numbers of neurons and synapses sampled in an area A or a, respectively, w designates the width of the middle or inner third of the molecular layer, and W represents the width of the GCL. A final n/N estimate per animal was calculated by averaging n/N values derived from 4 section series. The data presented here for each individual animal were obtained by examining series of 25-73 (mean = 42~ consecutive ultrathin sections which were used to obtain samples of 35-90 (mean = 61) or 20-58 (mean = 39~ perforated axospinous synapses per 4509 or 4517 um 2 (average areas) of the MML or IML. respectively. Additionally, 53-96 (mean = 68) neurons were sampled per GCL average area of 24,839 um 2. Examination of electron micrographs of serial sections of the dentate molecular layer showed that perforated axospinous synapses can be subdivided according to the configuration of their PSDs into 3 morphological subtypes characterized by (1) a (enestrated PSD having a hole(s) in its plate, (2) a horseshoe-shaped P S D exhibiting a U-type configuration of the PSD plate or (3) a segmented P S D consisting of 2-5 separate PSD plates (Fig. 1). The shape of the PSD plate extending along the postsynaptic membrane was ascertained in each perforated synapse sampled by means of z planar PSD reconstruction from serial sections (Fig. la,b,c) as described by us earlier 8. The number of synapses per neuron was differentially estimated for perforated synaptic junctions
343
a
b
c
Fig. 1. Electron micrographs of serial sections demonstrating 3 axospinous synapses indicated by solid arrows, open arrows or pointing fingers, respectively. All synaptic profiles exhibiting the PSD are presented. In some profiles of each synapse, PSD discontinuities or perforations (solid arrowheads) are observed indicating that these synaptic contacts belong to the category of perforated junctions. Planar PSD reconstructions from the micrographs of serial sections show that the population of perforated axospinous synapses can be divided according to the PSD configuration into three morphologically distinct subtypes. They are characterized by a fenestrated PSD (a) exhibiting a hole or holes in its single plate, by a horseshoe-shaped PSD (b) having a single plate that assumes a U-type configuration or by a segmented PSD (c) consisting of separate plates. Bar = 0.25 /~m.
344 having a fenestrated, horseshoe-shaped or segmented
TABLE II
PSD (Fig. 1). Q u a n t i t a t i o n of various subtypes of perforated axos-
Number of various subtypes o f perforated axospinous synapses per neuron in the inner molecular layer o f the dentate gyrus of control and kindled rats
pinous synapses in the M M L showed that the n u m b e r of synaptic contacts with segmented PSDs per n e u r o n was significantly increased in kindled rats relative to either
Designations are the same as in Table I.
unstimulated or coulombic controls (Table I). In contrast, synapses with fenestrated and horseshoe-shaped PSDs did not change significantly in n u m b e r s following
Rat
kindling, although a tendency towards a decrease was observed (Table I). In the IML, however, all 3 synaptic subtypes exhibited noticeable trends towards a decrease in their n u m b e r per n e u r o n which did not reach statistical significance (Table II). The n u m b e r of synapses per n e u r o n is a parameter
Number o f various subtypes o f perforated axospinous synapses per neuron in the middle molecular layer o f the dentate gyrus o f control and kindled rats
Designations: n/N, number of synapses per neuron; A, difference between group means. (A) Unstimulated control n/N
(B) Coulombic control
n/N n/N
48
AC-A
AC-B
A B-C
Synapses with fenestrated PSDs 1 43 48 2 44 77 3 50 41 4 16 55 5 50 36 6 53 41 7 38 36 Per group 42
(C) Kindling
-7.1%
67
+8.1%
56
Synapses with segmented PSDs 1 79 121 2 109 86 3 81 71 4 87 69 5 76 100 6 127 132 7 77 61
117 116 145 135 112 101 164
Per group 91
127
91
0.0%
-9.7%
(C) Kindling ................. n/N AC-A
AC-B
3 26 8 30 19 16 32
+1111%19
-29.6%
-36.7%
-23.3%
-29.8%
-19.7%
-25.6%
Synapses with horseshoe-shaped PSDs 1 60 52 35 2 33 28 19 3 33 64 58 4 45 67 25 5 62 27 27 6 22 44 27 7 44 46 39 Per group 43
Per group 76
47
+9.3% 33
82
63 35 86 47 71 56 66
+7.9% 61
-18.8%
Synapses with horseshoe-shaped PSDs 65 1 52 81 41 2 67 87 58 3 56 65 61 4 47 63 59 5 72 46 45 6 84 78 64 7 55 52 Per group 62
30
Synapses with segmented PSDs 1 75 114 2 62 60 3 67 70 4 99 106 5 68 46 6 80 102 7 83 76
45 46 40 35 41 38 28
+14.3% 39
(B) Coulombic control .............. n/N A B-('
Synapses withfenestrated PSDs 1 19 51 2 33 21 3 33 26 4 38 19 5 21 9 6 21 51 7 27 35 Per group 27
ATABLE I
Rat
(A) Unstimulated control n/N
-t6.4%
+39.6%* +39.6%*
* P < 0.05 and less, two-tailed randomization test for two independent samples.
that depends not only on the synaptic numerical density, but also on the numerical density of neurons and on dimensions of n e u r o n a l and synaptic layers in the case of stratified structures such as the dentate gyrus. A direct measure of relative quantities of the synaptic subtypes u n d e r comparison in the neuropil is provided by their ratio which can be assessed from raw synaptic counts in the same disectors. Therefore. the ratio of synapses with segmented PSDs to other perforated axospinous synapses was additionally estimated. It was found to be significantly increased in the M M L of kindled rats as compared with controls (Table III). No significant change in this ratio was observed in the IML (Table III). A major finding of this study is that kindling is associated with an increase in the n u m b e r of synapses per n e u r o n which is limited to only those axospinous synaptic contacts that are distinguished by a segmented PSD (Ta-
345 TABLE III Ratio of synapses with segmented postsynaptic densities to other perforated synaptic junctions in the middle and inner molecular layer of the dentate gyrus of control and kindled rats
Designation: A, difference between group means. (A) Unstimulated control Ratio
Rat
(B) Coulombic control
Ratio A C - A Ratio
0.905
0.820
1.064 1.333 1.480 1.406 1.120 1.217 1.783 -9.4%
Inner molecular layer 1 0.949 1.107 2 0.939 1.224 3 1.015 0.778 4 1.193 1.233 5 0.819 1.278 6 1.860 1.074 7 1.169 0.938 Group mean
1.135
1.090
AC-B
A B-A
Middle molecular layer l 0.832 0.938 2 0.982 0.524 3 0.764 0.670 4 1.381 0.585 5 0.623 1.220 6 0.927 1.109 7 0.828 0.693 Group mean
(C) Kindling
1.343 +48.4%* +63.8%* 1.658 0.778 1.303 0.855 1.543 1.302 0.930
-4.0%
1.196 +5.4%
+9.7%
* P < 0.05 and less, two-tailed randomization test for two independent samples.
-z
6
"~+36
9
J
Fig. 2. Diagram illustrating the process of synapse turnover described in the text. Schematics are shown for a non-perforated axospinous synapse (designated by a value of -420) and for perforated ones with a fenestrated (-6), horseshoe-shaped (-9) or segmented (+36) PSD. Differences in the number of synapses per neuron between control and kindled animals are indicated for each synaptic subtype in the middle molecular layer of the dentate gyrus, the value for non-perforated synaptic contacts being taken from the results reported earlier91°. Since the two groups of control rats studied did not differ significantly from each other with respect to the number of all synaptic subtypes examined, the data obtained from unstimulated and coulombic controls were combined. Kindling-induced alterations in synaptic numbers, which are shown here, were determined relative to the combined control values. Statistically significant changes include the decrease in the number of non-perforated junctions and the increment in perforated synapses exhibiting a segmented PSD.
ble I). This restructuring of connectivity is highly selective, since other subtypes of perforated axospinous synapses characterized by a fenestrated or horseshoe-shaped PSD (Table I), as well as n o n - p e r f o r a t e d axospinous synapses 9,w and synaptic junctions involving dendritic shafts 9, do not exhibit such a change. M o r e o v e r , the observed structural synaptic modification is restricted to the terminal synaptic field of stimulated axons ( M M L ) and does not involve an immediately adjacent synaptic field (IML) which was not directly activated during kindling (Table II). Such anatomical selectivity cannot be accounted for by non-specific effects of generalized seizures and is m o r e consistent with the notion that the observed structural synaptic modification is indeed a direct consequence of the kindling process. The kindling-induced increment in the n u m b e r of axospinous synaptic junctions exhibiting a segmented PSD may be due to a remodelling of some pre-existing axospinous synapses or to synaptogenesis. The former possibility has been p r o p o s e d by the existing models of synapse turnover and splitting4,5,7,es. It has been postulated 5'7'25 that axospinous synapses with a segm e n t e d PSD are constantly formed from non-perforated ones through the stages of p e r f o r a t e d synaptic junctions having first a fenestrated, then a horseshoe-shaped and finally a segmented PSD; eventually, a synapse of the segmented subtype splits into n o n - p e r f o r a t e d synaptic contacts (Fig. 2). The alternative hypothesis, namely, new synapse formation is equally likely and may be supp o r t e d by the discovery that perforant path kindling is accompanied by sprouting of mossy fibers and their terminals in the inner molecular layer of the rat dentate gyrus 34. This finding cannot, of course, account for our data, since the increase in the n u m b e r of segmented synapses was found to be restricted to the middle molecular layer. Unfortunately, the results of synapse quantitation presented here (Fig. 2) are insufficient to resolve the question of whether an intensification of synapse turnover or an induction of synaptogenesis is responsible for the kindling-induced increase in the n u m b e r of synaptic contacts with segmented PSDs. The total synaptic population of the M M L in control, unstimulated animals is c o m p o s e d of axodendritic (6.5%), p e r f o r a t e d axospinous (8.6%) and nonp e r f o r a t e d axospinous (85.0%) synapses 9. Those perforated synaptic contacts that have a segmented PSD consititute only a small p r o p o r t i o n (4.0%) of the M M L synaptic population. Nevertheless, the fact that this particular subtype of p e r f o r a t e d axospinous synapses is selectively increased in numbers following kindling is especially interesting. Such synaptic junctions exhibit two to five discrete PSD plates, rather than only one present in all other synapses. Coexistent with a PSD plate is a
346 specialized presynaptic m e m b r a n e area a p p o s e d by clusters of synaptic vesicles which constitute together a region within the synapse known as the active zone tS. Presynaptic secretion and postsynaptic reception of substances mediating chemical impulse transmission are believed to occur within this region TM. The availability of additional transmission sites within a single synapse would be expected to enhance its strength, especially if each such site is associated with a r e c e p t o r cluster. Therefore, an increase in the n u m b e r of synaptic contacts having s e g m e n t e d PSDs m a y provide a mechanism for the enduring amplification of synaptic responsiveness in the stimulated circuit which characterizes kindling. Of special significance is the fact that the structural synaptic modification r e p o r t e d here is not unique for the kindling p h e n o m e n o n . A selective increase in the number of axospinous synapses exhibiting s e g m e n t e d PSDs has also been d e m o n s t r a t e d by us to occur in the M M L of the d e n t a t e gyrus after the induction of long-term potentiation (LTP) via medial p e r f o r a n t path stimulation 12. LTP is a n o t h e r m o d e l of synaptic plasticity which is closely related to kindling 23. E a r l y stages of kindling e v o k e d by stimulation of the pefforant path are invariably a c c o m p a n i e d by potentiation at perforant p a t h -
The authors are thankful to William Goossens, Nicholas Kriho and Daniel Lanciloti for their skillful technical assistance and to Dr. L.T. Thompson for preparing Fig. 2. This work was supported by grants AG 08794 from NIA and BNS 88-19902 from NSE
1 Br~endgaard, H. and Gundersen, H.J.G., The impact of recent stereological advances on quantitative studies of the nervous system, J. Neurosci. Methods, 18 (1986) 39-78. 2 Cain, D.P., Corcoran, M. and Staines, W., Effects of protein synthesis inhibition on kindling in the mouse, Exp. Neurol., 68 (1980) 409-416. 3 Calverley, P.K.S. and Jones, D.G., A serial section study of perforated synapses in rat neocortex, Cell Tiss. Res., 247 (1987) 565-572. 4 Calverley, P.K.S. and Jones, D.G., Contribution of dendritic spines and perforated synapses to synaptic plasticity, Brain Res. Rev., 15 (1990) 215-249. 5 Carlin, P.K. and Siekevitz, P., Plasticity in the central nervous system: do synapses divide?, Proc. Natl. Acad. Sci. U.S.A., 80 (1983) 3517-3521. 6 Cohen, R.S. and Siekevitz, P., Form of the postsynaptic density. A serial section study, J. Cell Biol., 78 (1978) 36-46. 7 Dyson, S.E. and Jones, D.G., Synaptic remodelling during development and maturation: junction differentiation and splitting as a mechanism of modifying connectivity, Dev. Brain Res., 13 (1984) 125-137. 8 Geinisman, Y., Morrell, E and deToledo-Morrell, L., Axospinous synapses with segmented postsynaptic densities: a morphologically distinct synaptic subtype contributing to the number of profiles of 'perforated' synapses visualized in random sections, Brain Research, 423 (1987) 179-188. 9 Geinisman, Y., Morrell, F. and deToledo-Morrell, L., Remodeling of synaptic architecture during hippocampal 'kindling', Proc. Natl. Acad. Sci. U.S.A., 85 (1988) 3260-3264. 10 Geinisman, Y., Morrell, F. and deToledo-Morrell, L., Increase in the relative proportion of perforated axospinous synapses following hippocampal kindling is specific for the synaptic field of stimulated axons, Brain Research, 507 (1990) 325-331. 11 Geinisman, Y., Morrell, F. and deToledo-MorreU, L., Alterations of synaptic ultrastructure induced by hippoeampal kindling. In J.A. Wada (Ed.), Kindling 4, Plenum, New York,
1990, pp. 75-88. 12 Geinisman, Y., deToledo-Morrell, L. and Morrell, E, Induction of long-term potentiation is associated with an increase in the number of axospinous synapses with segmented postsynaptic densities, Brain Research, in press. 13 Goddard, G.V., The development of epileptic seizures through brain stimulation at low intensity, Nature, 214 (1967) 1020-1021. 14 Goddard, G.V. and Douglas, RM., Does the engram of kindling model the engram of normal long term memory?, Can. J. Neurol. Sci., 2 (1975) 385-394. 15 Goddard, G.V., Mclntyre, D. and Leech, C., A permanent change in brain function resulting from daily electrical stimulation, Exp. Neurol., 25 (1969) 295-;330. 16 Gundersen, H.J.G., Notes on the estimation of the numerical density of arbitrary profiles: the edge effect, J. Microsc., 111 (1977) 219-223. 17 Gundersen, H.J.G., Stereology of the arbitrary particles, J. Microsc., 143 (1986) 3-45. 18 Heuser, J.E. and Reese, T.S., Structure of the synapse. In E.R. Kandel (Ed.), Handbook of Physiology, Sect. 1, Vol. 1, Am. Physiol. Soc. and Williams & Wilkins, Bethesda, MD, 1977, 261-294. 19 Jonek, V. and Wasterlain, C.G, Effect of inhibitors of protein synthesis on the development of kindled seizures, Exp. Neurol, 66 (1979) 524-532. 20 Le Gal La Salle, G., Kindling as a model of long-lasting plasticity: new perspective for the search for molecular rneehanisms. In J.A. Wada (Ed.), Kindling 4, Plenum, New York, 1990, pp. 1-9. 21 McNamara, J.O., Byrne, M.C., Dasheiff, R.M. and Fitz, J.G.; The kindling model of epilepsy: a review, Prog. Neurobiol., 15 (1980) 139-159. 22 Morrell, F., Callosal mechanisms in epileptogenesis. In A. Reeves (Ed.), Epilepsy and the Corpus Callosum, Plenum, New York, 1985, pp. 99-130. 23 Morrell, F. and deToledo-Morrelk t,., Kindling as a model of
dentate granule cell synapses ~3. It is conceivable, therefore, that the remodelling of synaptic architecture observed in kindled animals may actually result from LTP elicited prior to the acquisition of the fully develo p e d kindled state. A n o t h e r possibility is that various models of synaptic plasticity share a similar mechanism of modifying synaptic connectivity. In fact, there are some indications that visual discrimination training may also be accompanied by an increase in the n u m b e r of synapses with segmented PSDs. In the visual cortex of trained animals, clusters of closely a p p r o x i m a t e d synaptic grids are seen more frequently than in controls suggesting that visual training m a y lead to focalization of dense projections and related p a r a m e m b r a n e o u s densities 35 one of which is the PSD. Thus, an increase in the n u m b e r of axospinous synapses distinguished by a segm e n t e d PSD may represent a structural substrate of kindling, LTP and o t h e r forms of synaptic plasticity associated with learning and memor~
347
24
25
26
27 28
29
neuronal plasticity, In J.A. Wada (Ed.), Kindling 3, Raven, New York, 1986, pp. 17-33. Morrell, E, Tsuru, N., Hoeppner, T.J., Morgan, D. and Harrison, W.H., Secondary epileptogenesis in the frog forebrain: effect of inhibition of protein synthesis, Can. J. Neurol. Sci., 2 (1975) 407-416. Nieto-Sampedro, M., Hoff, S.W. and Cotman, C.W., Perforated postsynaptic densities: probable intermediates in synapse turnover, Proc. Natl. Acad. Sci. U.S.A., 79 (1982) 5718-5722. Peters, A. and Kaiserman-Abramof, I.R., The small pyramidal neuron of the rat cerebral cortex. The synapses upon dendritic spines, Z. Zellforsch., 100 (1969) 487-506. Racine, R., Kindling: the first decade, Neurosurgery, 3 (1978) 234-252. Racine, R.J. and Zaide, J., A further investigation into the mechanisms of the kindling phenomenon. In K. Livingston and O. Hornykiewicz (Eds.), Limbic Mechanisms: The Continuining Evolution of the Limbic System Concept, Plenum, New York, 1978, pp. 457-493. Rosene, D.L., and Van Hoesen, G.W., The hippocampal formation of the primate brain. In E.G. Jones and A. Peters (Eds.), Cerebral Cortex, Vol. 6, Further Aspects of Cortical function, Including Hippocampus, Plenum, New York, 1987,
pp. 345-456. 30 Spacek, J., and Hartmann, M., Three-dimensional analysis of dendritic spines. I. Quantitative observations related to dendritic spina and synaptic morphology in cerebral and cerebellar cortices, Anat. Embryol., 167 (1983) 289-310. 31 Sterio, D.C., The unbiased estimation of number and sizes of arbitrary particles using the disector, J. Microsc., 134 (1984) 127-136. 32 Steward, O. and Vinsant, S., The process of reinnervation in the dentate gyrus of the adult rat: a quantitative electron microscopic analysis of terminal proliferation and reactive synaptogenesis, J. Comp. Neurol., 214 (1983) 370-386. 33 Sutula, T. and Steward, O., Quantitative analysis of synaptic potentiation during kindling of the perforant path, J. Neurophysiol., 56 (1986) 732-746. 34 Sutula, T., Xiao-Xian, H., Cavazos, J. and Scott, G., Synaptic reorganization in the hippocampus induced by abnormal functional activity, Science, 239 (1988) 1147-1150. 35 Vrensen, G. and Nunes Cardozo, J., Changes in size and shape of synaptic connections after visual training: an ultrastructural approach to synaptic plasticity, Brain Research, 218 (1981) 7997.
