J. Anat. (1991), 174, pp. 229-238 229 With 3 figures Printed in Great Britain Neural plasticity of the hippocampal (CA1) - quantitative changes in spine density following handling and injection for drug testing

C. H. HORNER, M. O'REGAN* AND E. ARBUTHNOTTt Department of Anatomy, * Department of Statistics and t Department of Physiology, Trinity College, Dublin 2, Ireland (Accepted 10 August 1990)

INTRODUCTION Plasticity of the central nervous system is a well documented phenomenon. Following Ramon y Cajal's (1911) discovery of dendritic spines it became apparent that these structures can change in response to a wide spectrum of conditions, including disease states, administration of chemical substances, non-physiological stimulation and manipulation of factors such as food intake and sleep. Examples of the early findings are found in the published work of Demoor (1898), Monti (1895) and Querton (1898). Although plasticity has been demonstrated in many areas, the neural circuitries of the possess remarkable capacities for both functional and structural changes (Alger & Teyler, 1976; Deadwyler, Dudek, Cotman & Lynch, 1975; Lee, Stanford, Cotman & Lynch, 1977; Lynch, Gall & Dunwiddie, 1978; McWilliams & Lynch, 1978). This study demonstrates the plastic nature of the hippocampus in response to the sensory stimuli of handling and injection. Although the mechanisms of action of drug groups are generally investigated using biochemically- and physiologically-based studies, morphological changes following drug administration have been recorded (Arbuthnott & Folan, 1983; Bigotte & Olsson, 1983; Shikai & Miyakawa, 1981). The tricyclic antidepressant group of drugs is a frequently used treatment for depression, a common psychiatric disorder. The means by which these drugs alleviate depressive symptoms, after approximately three weeks, has been much researched and debated. The biogenic monoamine theory of depression (Coppen, 1967) suggests that a neurotransmitter deficiency at the synaptic junction is responsible. However, the relationship between the acute primary effects of tricyclic antidepressants on the neurotransmitter levels (Ross & Renyi, 1975) and their clinical lag in effectiveness is poorly correlated (Pollock et al. 1986). Therefore the possibility of involvement of postsynaptic receptors in the mediation of antidepressant action has been suggested (Sulser, 1983; Fuxe et al. 1983). A process requiring time for change to occur, such as long-term morphological alterations, may explain the three weeks delay in clinical improvement. Since the synaptic junction is the site of action of these drugs, whether it be on the neurotransmitter levels or the postsynaptic receptor, this study involved a quantitative assessment of the spine density of several dendritic types, the dendritic spines being the site of the majority of synapses (Gray, 1959; Colonnier, 1968). The hippocampus was chosen as the area for study since it is considered part of the 'limbic system" and so has been linked with emotional and behavioural changes (MacLean, 1952; Kluiver & Bucy, 1939). It also has a highly organised structure which is readily identified (Brodal, 230 C. H. HORNER AND OTHERS

1981). In particular the pyramidal cells of area CAl were chosen for this study, these being the site of synapse of the Schaffer-commissural collaterals within the intrinsic circuit of the hippocampal formation.

MATERIALS AND METHODS Young outbred male Wistar albino rats, supplied by the Wellcome Research Animal Laboratory at Trinity College, Dublin, were used in this study. They were eight weeks old at the start of treatment and had a mean weight of 226 g (208-244 g). They were reared in pairs under the same conditions receiving water and rat/mouse diet by Odlum's ad libitum. All animals were free from grossly detectable pathology. The animals were divided into three groups identified as control, drug and saline. The control group (n = 6) received no treatment and were unhandled except for weekly weighing. The drug group (n = 5) had daily intraperitoneal injections of clomipramine (Anafranil), a tricyclic antidepressant, for 22 days. The dose administered was 10 mg per kg body weight. The saline group (n = 5) served as handled controls. They received daily intraperitoneal injections of normal saline of the same volume and were handled in the same way as the drug-treated group. Following 22 days of treatment, the animals were anaesthetised with 6% sodium pentobarbitone (Sagatal) using a dose of 0A44 ml per kg. The animals were artificially ventilated using 95 % 02/5 % CO2 until vascular perfusion was initiated through the left ventricle using a dilute fixative consisting of 1 % paraformaldehyde/ 1 -25 % glutaraldehyde in 0-2M cacodylate buffer initially, followed by a concentrated fixative of 4% paraformaldehyde/5 % glutaraldehyde in 0-2M cacodylate buffer. The crania were removed and stored in concentrated fixative at 4°C overnight. The brains were dissected out, bisected in the midsagittal plane and blocks of hippocampal tissue were taken from each cerebral hemisphere using Paxinos & Watson's stereotaxic atlas (1982) as a guideline. A modified Golgi-Kopsch technique described by Riley (1979) was used for impregnating the neural tissue. The blocks were immersed in fixative for five days, rinsed and further immersed in 0-75 % silver nitrate for two days while being kept in the dark. Approximately 30 ml of each solution were used per block of tissue. Post- impregnation, the blocks were dehydrated, orientated and shelled in paraffin wax for sectioning. Thick coronal sections of 120,um were cut using a sliding microtome. Thick sections are favoured for Golgi-impregnated neural tissue (Feldman, 1976; Fitch, Juraska & Washington, 1989; Uylings, Kuypers, Diamond & Veltman, 1978) since they enable viewing of the wide dendritic field and yet permit focusing at x 100 oil magnification. Spine density on pyramidal cell (CA1) dendrites was estimated at three loci. The first was a 50,sm segment of starting 50,um from the origin of this dendrite at the apex of the cell body. The second was a 25,cm section of basal dendrite beginning 25 ,um from its origin at the perikaryon base and finally a 25,tm segment of oblique dendrite from its point of origin of the apical dendrite (Fig. 1). Visible spines were counted over these sections of dendrites. Twenty estimations for each type of dendrite for each animal were made, choosing clearly visible dendrites which were relatively straight. All protrusions whether pedunculated or stubby, with or without terminal bulbous expansions, were counted as spines if they appeared to be in direct continuity with the dendritic shaft (Feldman & Dowd, 1975). The dendritic diameter and the exact length of dendrite over which the spines were counted was measured Spine density plasticity in hippocampal 231 Terminal tuft

Apical 50 um

.I

Fig. 1. The apical, basal and oblique dendritic loci chosen for estimation of spine density. using semi-automatic image analysis. Using the same system, the spine length and diameter of the spine head of a typical spine on the segment of dendrite being assessed were recorded. These values were applied to the geometrical equation devised by Feldman & Peters (1979) which produces estimates of 'true' spine density which are corrected for dendritic diameter and size of the spines. A nested analysis of variance was used to compare the three group means. In the case of significant group differences a Newman-Keuls test was performed to find the groups between which the difference existed.

RESULTS On examination of the pyramidal cells in the hippocampal sections at low magnification ( x 10) there were no obvious differences between cells within or between groups. At higher magnification ( x 40) there appeared to be differences between the density of spines on cells. Although not the case for all cells, the greater spine numbers appeared to be in the drug-treated and saline-injected animals (Fig. 2). At x 100, oil immersion, spine counts were performed to confirm or negate this impression. Spine density is defined as the number of dendritic spines per micron of dendrite. A total of 900 estimates of spine density was made, twenty estimates per locus per animal were recorded and the mean per group calculated. One control animal was poorly fixed and was discarded, leaving five animals per group. The mean spine density for a particular locus was similar for animals within the same group. However, the cell 232 C. H. HORNER AND OTHERS Basal Apical and oblique

Control,

Saline

Drug

Fig. 2. Camera lucida drawings of 40-50 ,um segments of basal, apical and oblique dendrites in drug- and saline-injected, and unhandled control experimental groups. to cell variability for visible spine counts at any given locus within the same animal were quite large, e.g. 28-75 spines per dendritic segment in different cells of one saline- injected animal. Similar variability was noted in data from all animals in the study. The mean spine densities, corrected for obscuration from varying dendritic diameter and spine size, for apical dendrites were 4-83 for the drug group, 4-33 for the saline- injected group and 4 05 for the unhandled control group. Basal dendrites had mean spine densities of 5-71 in the drug-treated, 5-17 in the saline-injected and 3 97 in the control groups while oblique dendrites had mean values of 6-30 for the drug animals, 5-77 for the saline animals and 3-87 for the control animals. These values and the standard error of the group means are recorded in Table 1 and displayed graphically in Figure 3. These results show basal and oblique spine densities to be higher than apical spine density except in the unhandled controls where the spine densities relate to the diameter of the dendrites, apical dendrites having the highest and oblique dendrites the lowest densities. Basal and oblique spine densities were very similar but oblique spine density was the greater in both drug and saline treatment groups. Drug- injected animals had consistently higher spine densities than saline-injected and unhandled controls while saline-treated animals had greater values than controls. The differences between treatment groups appeared to be quite constant for basal and oblique dendrites, drug- and saline-injected animals having similar spine densities which were statistically greater than unhandled control animals. The apical dendritic Spine density plasticity in hippocampal neurons 233 Table 1. Mean spine densities for apical, basal and oblique dendrites (± S.E.M.) on hippocampal (CA 1) pyramidal cells in drug-treated, saline-injected and unhandled controls. These values are corrected for dendritic diameter and spine size using Feldman & Peters (1979) formula Apical spine density Basal spine density Oblique spine density no./jum no./,am no./jfm Number of cells 300 300 300 sampled Drug 4-83 5-71 6-30 Saline 4-33 5-17 5.77 Control 4 05 3.97 3-87 S.E.M. 0-244 0-12 0-158 'F' 2-68 553 654 no., Estimate of true number of spines. S.E.M., Standard error of the group means. 'F', Statistic derived from analysis of variance for comparison of group means.

6.5. { Ob

iB + Ob 5.5 E 6 C B

.t_ CCA 0 i Ap ~0 C .a 4.5 1 i Ap

Ap B Ob

3.5

Drug Saline Control Fig. 3. Mean spine densities (no./jum) of apical (Ap), basal (B) and oblique (Ob) dendrites in hippocampal (CAl) pyramidal cells in drug-treated (Drug), saline-injected control (Saline) and unhandled control (Control) groups. These values are estimates of'true' spine density using Feldman & Peters (1979) correction formula incorporating dendritic diameter and spine size.

ANA 174 234 C. H. HORNER AND OTHERS Table 2. Comparisons of treatment group mean spine densities of apical, basal and oblique dendrites on hippocampal (CAI) pyramidal cells Drug Drug Saline versus versus versus saline control control Apical spine density N.S N.S N.S Basal spine density N.S ** ** Oblique spine density N.S ** ** N.S., non-significant. ** P < 0-01. spine densities showed the same trend, with drug-related being greater than saline- treated which was greater than control, but the differences between groups were more evenly distributed than those noted between groups for oblique and basal dendrites. Apical spine density showed a 19-3 % increase in drug-treated over unhandled control and a 7-9 % increase in saline-injected over the control group. An 1 -5% increase in the drug group over the saline group was noted. The basal dendritic spine densities showed 57-4 % and 10-4% increases in the drug-treated group over the control and saline groups respectively. The saline-injected group had 43-8 % greater spine density than the unhandled controls. Oblique dendritic spine densities showed the largest differences with drug-treated animals having a 62-8 % increase and saline- injected animals having a 49 1 % greater value than controls. A small difference of 9-2% was noted between drug- and saline-injected groups, the drug-treated being the greater. An analysis of variance gave F values of 2-68, 55-3 and 65 4 for apical, basal and oblique spine densities respectively, each for two degrees of freedom. Significant differences (P < 0O01) were noted between control and drug groups and between control and saline groups for both basal and oblique dendritic spine densities. Apical spine densities did not prove to be significantly different between treatment groups (Table 2).

DISCUSSION In this study visible spine counts recorded in the unhandled control group gave spine densities of 1-07 for apical, 1-29 for basal and 1-25 for oblique dendrites, prior to correction for dendritic diameter and spine size. These compare well with those noted by other researchers, which suggests that they are reasonable estimates. Riley & Walker (1978) recorded 1-36 spines/,sm on basal dendrites in hippocampal (CA1) pyramidal cells. An apical dendritic spine density of one per micrometer was estimated by Rutledge, Duncan & Cant (1972) in the cerebral cortex of kittens, while Ryugo, Ryugo, Globus & Killackey (1975) reported 1 19 spines/,um on apical dendrites in the auditory cortex of rats. Feldman & Dowd (1975) recorded 1-26 spines/,um on oblique dendrites of the rat visual cortex, although their apical and basal estimates correlate poorly with those in this study. The hippocampus is an inwardly folded continuation of the cortex and therefore the pyramidal cells of these areas should be similar in structure. However, age differences and the amount of sensory input into the neurons could produce variations in spine numbers. Initially it would appear that chronic tricyclic antidepressant treatment has caused significantly increased spine densities in both basal and oblique dendrites compared Spine density plasticity in hippocampal neurons 235 with unhandled and untreated controls. Other authors have also noted changes following chronic antidepressant treatment. Alpers & Himwich (1972) recorded changes in amine levels in the hippocampus of rats and Arbuthnott & Folan (1983) noted structural changes in the hippocampus of mice after imipramine treatment. However, both these authors also found some changes in saline-injected animals which were similar to those in the drug-treated, as is the case in this study. It is clear that when all three groups are analysed together, the drug- and saline-injected animals have reacted similarly in regard to both basal and oblique spine densities and although the trend is similar for apical spine density, the results are not significant. Therefore, it would appear that the sensory stimulus of handling or the stressful and painful sensation of injection is responsible for the morphological changes that have occurred. Sensory stimuli whether pleasant or painful have been shown repeatedly to cause neuronal plasticity (Greenough & Volkmar, 1973; Globus, Rosenzweig, Bennett & Diamond, 1973; Schapiro & Vukovich, 1970; Uylings et al. 1978). Several authors have described comparisons between animals reared in different environments, one being socially enriched and the other isolated or impoverished. The former regime involves sensory stimulation of animals, either gently with handling and interesting surroundings (Globus et al. 1973) or by stressful means such as temperature changes, electric shock and shaking (Schapiro & Vukovich, 1970). Differential experience has produced differences in brain enzymes and cortical weight (Bennett, Diamond, Krech & Rosenzweig, 1964; Rosenzweig, Bennett & Diamond, 1970) and in spine density (Globus et al. 1973; Schapiro & Vukovich, 1970). Both authors noted increased spine densities in animals experiencing an enriched environment using similar counting techniques but the extent of the increase differed. Globus et al. (1973) used gentle non-stressful methods and found significant increases of 9-7 % in basal and 3f6 % in oblique dendritic spine densities. Schapiro & Vukovich (1970) employed disturbing and stressful procedures and noted much greater increases of 33 % in both basal and oblique spine density. In both studies apical spine density did not differ significantly. The results of this study compare very well with those recorded by Schapiro & Vukovich (1970). Therefore it would appear that the daily intraperitoneal injection received by the rats was not only a painful sensory stimulus but stressful. The contrary situation, where animals are deprived of sensory stimuli, has produced decreases in spine density, either by modification of the normal environment (Globus & Scheibel, 1967b; Valverde, 1967) or manipulation of the sensory pathways (Colonnier, 1964; Globus & Scheibel, 1967a; Valverde & Esteban, 1968). Globus et al. (1973) suggest that enriched conditions involve variation of stimulation in several sensory modalities such as visual, tactile, kinaesthetic, auditory and olfactory. The demand for integration of this complex multisensory information requires increased intracortical connections, which involves ramification of axon terminals, growth of dendritic spines and formation of new synapses. A hypothesis describing a possible division of pre-existing synapses has been proposed by Carlin & Siekevitz (1983). This model suggests that various types of stimulation result in a sequence leading to dendritic spine division, with each subdivision having the ability to function as a synapse. One response where synapse division may play a role is in learning and memory. This study suggests that the significantly increased spine densities recorded in basal and oblique dendrites of drug- and saline-injected animals is, in fact, due to sensory stimulation and possibly stress. Although drug-treated animals have higher spine densities than saline-injected, these differences are small and not significant. It would appear that if tricyclic antidepressants do act on the postsynaptic receptors as

9-2 236 C. H. HORNER AND OTHERS suggested, they do not involve quantitative changes in the spines bearing these receptors. Increased spine numbers suggest an increase in synapse numbers and a change or upgrading of the neural circuits. This may well be the basis by which the new experience is learned and memorised. Since the simple process of handling and injection can produce such marked quantitative changes it is extremely important to plan experimental procedures carefully. If the unhandled control group had been omitted, the fact that the injection procedure produced obvious morphological changes would have been missed. If the saline-injected controls were omitted, the assumption that the tricyclic antidepressant drug produced significant changes may have been made. A further group of orally drug-treated animals should be investigated in any further work and a handled but non-injected group as a representative of less stressful, gentle stimulation would also clarify the degree to which sensory stimuli produce morphological plasticity.

SUMMARY A quantitative assessment of the spine density of apical, basal and oblique dendrites on pyramidal neurons of area CAl of the hippocampus was made in three experimental groups. The results of a group injected with a tricyclic antidepressant were compared statistically with a saline-injected group and an unhandled control group. A statistical analysis of variance indicated significant (P < -01) differences between drug and control groups and between saline and control groups in two of the loci assessed (basal and oblique dendrites). These findings suggest that the sensory stimulation provided by daily injection and handling is responsible for the increases in spine density in drug- and saline-injected animals. A single control group is insufficient in experiments of this type. Experimental protocol is extremely important and the dramatic morphological changes produced by simple routine processes should not be under-estimated. We are grateful to St Patrick's Hospital for a grant from the Neuropsychiatric Fund and to Ciba Geigy for a supply of clomipramine (Anafranil). We also wish to thank Ms P. Power and Ms S. Ward for their technical assistance.

REFERENCES ALGER, B. E. & TEYLER, T. J. (1976). Long-term and short-term plasticity in the CAl, CA3, and dentate regions of the rat hippocampal slice. Brain Research 110, 463-480. ALPERS, H. S. & HIMWICH, H. E. (1972). The effects of chronic imipramine administration on rat brain levels of , 5-hydroxyindoleacetic acid, norepinephrine and dopamine. Journal of Pharmacology and Experimental Therapeutics 180, 531-538. ARBUTHNOTT, E & FOLAN, J. (1983). A study of the structural changes induced in CNS neurones of the mouse following drug administration and handling. Irish Journal of Medical Science 151, 416. BENNETT, E. L., DLuMoND, M. C., KRECH, D. & ROSENZWEIG, M. R. (1964). Chemical and anatomical plasticity of brain. Science 146, 610-619. BIGoTrE, L. & OLSSON, Y. (1983). Toxic effects of adriamycin on the central nervous system. Ultrastructural changes in some circumventricular organs of the mouse after intravenous administration of the drug. Acta neuropathologica 61, 291-299. BRODAL, A. (1981). The limbic system. In Neurological Anatomy in Relation to Clinical Medicine (ed. A. Brodal), pp. 689-690. Oxford: Oxford University Press. CAJAL, S. RAMON Y (1911). Histologie du Systeme Nerveux de PHomme et des Vertebres, Vol. 2A. Paris: Maloine. CARLIN, R. K. & SIEKEVITZ, P. (1983). Plasticity in the central nervous system: Do synapses divide? Proceedings of the National Academy of Sciences of the USA 80, 3517-3521. COLONNIER, M. (1964). Experimental degeneration in the cerebral cortex. Journal ofAnatomy 98, 47-53. Spine density plasticity in hippocampal neurons 237 COLONNEER, M. (1968). Synaptic patterns on different cell types in the different laminae of the cat visual cortex. An electron microscope study. Brain Research 9, 268-287. COPPEN, A. (1967). The biochemistry of affective disorders. British Journal of Psychiatry 113, 1237-1264. DEADWYLER, S. A., DUDEK, F. E., COTMAN, C. W. & LYNCH, G. (1975). Intracellular responses of rat dentate granule cells in vitro: Post-tetanic potentiation to perforant path stimulation. Brain Research 88, 80-85. DEMOOR, J. (1898). Le mecanisme et la signification de l'etat moniliforme des neurons. Annales de la Societe des sciences midicales et naturelles de Bruxelles 7, 205-250. FELDMAN, M. L. (1976). Aging changes in the morphology of cortical dendrites. In Neurobiology ofAging (ed. R. D. Terry & S. Gershon), pp. 211-227. New York: Raven Press. FELDMAN, M. L. & DowD, C. (1975). Loss of dendritic spines in aging cerebral cortex. Anatomy and Embryology 148, 279-301. FELDMAN, M. L. & PETERs, A. (1979). A technique for estimating total spine numbers on Golgi-impregnated dendrites. Journal of Comparative Neurology 188, 527-542. FITCH, J. M., JURASKA, J. M. & WASHINGTON, L. W. (1989). The dendritic morphology of pyramidal neurons in the rat hippocampus CA3 area. I. Cell types. Brain Research 479, 105-114. FuxE, K., OGREN, S. O., AGNATI, L. F., BENEFATI, F., FREDHOLM, B., ANDERSSON, K., ZINI, I. & ENEROTH, P. (1983). Chronic antidepressant treatment and central 5-HT synapses. Neuropharmacology 22, 389-400. GLOBUS, A., ROSENZWEIG, M. R., BENNETr, E. L. & DIAMOND, M. C. (1973). Effects of differential experience on dendritic spine counts in rat cerebral cortex. Journal of Comparative Physiology and Psychology 82, 175-181. GLOBUS, A. & SCHEIBEL, A. B. (1967a). Synaptic loci on visual cortical neurons of the rabbit: The specific afferent radiation. Experimental Neurology 18, 116-131. GLOBUS, A. & SCHEIBEL, A. B. (1967b). Effect of visual deprivation on cortical neurons: A Golgi study. Experimental Neurology 19, 331-345. GRAY, E. G. (1959). Electron microscopy of synaptic contacts on dendritic spines of the cerebral cortex. Nature 183, 1592-1593. GREENOUGH, W. T. & VOLKMAR, F. R. (1973). Pattern of dendritic branching in occipital cortex of rats reared in complex environments. Experimental Neurology 40, 491-504. KLUVER, H. & BUCY, P. C. (1939). Preliminary analysis of functions of the temporal lobes in monkeys. Archives of Neurology and Psychiatry 42, 979-1000. LFx, K., STANFORD, E. J., COTMAN, C. W. & LYNCH, G. S. (1977). Ultrastructural evidence for bouton proliferation in the partially deafferented dentate gyrus of the rat. Experimental Brain Research 129, 475-485. LYNCH, G., GALL, C. & DUNWIDDIE, T. V. (1978). in the hippocampal formation. In Maturation of the Nervous System (ed. M. A. Corner), pp. 113-128. Amsterdam: Elsevier. MAcLEAN, P. D. (1952). Some psychiatric implications of physiological studies on frontotemporal portion of limbic system (visceral brain). Electroencephalography and Clinical Neurophysiology 4, 407-418. MCWILLLAMS, J. R. & LYNCH, G. (1978). Terminal proliferation and following partial deafferentation: the reinnervation of the inner molecular layer of the dentate gyrus following removal of its commissural afferents. Journal of Comparative Neurology 180, 581-616. MONTI, A. (1895). Sur les alterations du systeme nerveux dans l'inanition. Archives italiennes de biologie 24, 347-360. PAxINOs, G. & WATSON, C. (1982). The Rat Brain in Stereotaxic Coordinates. Sydney: Academic Press. POLLOCK, B. G., PEREL, J. M., SHOSTAK, M., ANTELMAN, S. M., BRANDOM, B. & KUPFER, D. J. (1986). Understanding the response lag to tricyclics. I. Application of pulse loading regimens with intravenous clomipramine. Psychopharmacology Bulletin 22, 214-219. QUERTON, L. (1898). Le sommeil hibernal et les modifications des neurones cerebraux. Annales de la Socieite' des sciences medicales et naturelles de Bruxelles 7, 147-204. RiLEY, J. N. (1979). A reliable Golgi-Kopsch modification. Brain Research Bulletin 4, 127-129. RILEY, J. N. & WALKER, D. W. (1978). Morphological alterations in the hippocampus after long-term alcohol consumption in mice. Science 201, 646-648. ROSENZWEIG, M. R., BENNETT, E. L. & DiAMOND, M. C. (1970). Chemical and anatomical plasticity of brain: replications and extensions. In Macromolecules and Behaviour (ed. J. Gaito). New York: Appleton-Century- Crofts. Ross, S. B. & RENtn, A. L. (1975). Tricyclic antidepressant agents. I. Comparison of the inhibition of the uptake of 3H-noradrenaline and "4C-5-hydroxytryptamine in slices and crude synaptosomal preparations of mid-brain hypothalamus region of the rat brain. Acta pharmacologica and toxicologica 36, 382-394. RUTLEDGE, L. T., DUNCAN, J. & CANT, N. (1972). Long-term status of pyramidal cell axon collaterals and apical dendritic spines in denervated cortex. Brain Research 41, 249-262. RYuGo, D. K., RYuGo, R., GLOBUS, A. & KILLACKEY, H. P. (1975). Increased spine density in auditory cortex following visual or somatic deafferentation. Brain Research 90, 143-146. SCHAPIRO, S. & VUKOVICH, K. R. (1970). Early experience effects upon cortical dendrites: a proposed model for development. Science 167, 292-294. SHIKAI, I. & MIYAKAWA, T. (1981). Morphological changes induced in the central nervous system of rats by steroid drugs. Folia psychiatrica et neurologica japonica 35, 217-224. SULSER, F. (1983). Mode of action of antidepressant drugs. Journal of Clinical Psychiatry 44, 14-20. 238 C. H. HORNER AND OTHERS UYLINGS, H. B. M., KuYPERs, K., DIAMOND, M. C. & VELTMAN, W. A. M. (1978). Effects of differential environments on plasticity of dendrites of cortical pyramidal neurons in adult rats. Experimental Neurology 62, 658-677. VALVERDE, F. (1967). Apical dendritic spines of the visual cortex and light deprivation in the mouse. Experimental Brain Research 3, 337-352. VALVERDE, F. & ESTEBAN, M. E. (1968). Peristriate cortex of mouse: location and the effects of enucleation on the number of dendritic spines. Brain Research 9, 145-148.