Purkinje Cell Dendritic Tree Research Articles

Cellular and molecular control of dendritic growth and development of cerebellar Purkinje cells

Purkinje cells are the principal neurons of the cerebellar cortex and are characterized by a large and highly branched dendritic tree. For this reason, they have for a long time been an attractive model system to study the regulation of dendritic growth and differentiation. In this article, I will first review studies on different aspects of Purkinje cell dendritic development and then go on to present studies which have aimed at experimentally altering Purkinje cell dendritic development. Some of the cellular and molecular mechanisms which have been shown by these studies to be important determinants of Purkinje cell dendritic development will be discussed, in particular the role of the parallel fiber input, of hormones, and of neuronal growth factors. The organotypic slice culture method will be introduced as an important experimental tool to study Purkinje cell dendritic development under controlled conditions. Using cerebellar slice cultures, protein kinase C (PKC) has been identified as a major determinant of Purkinje cell dendritic development and the contribution of specific isoforms of PKC will be discussed. Finally, it will be shown that Purkinje cell dendritic development in slice cultures does not depend on the activation of glutamate receptors and appears to be independent of the presence of the neurotrophin BDNF. These studies indicate that the initial outgrowth of the Purkinje cell dendritic tree can occur in the absence of signals derived from afferent fibers, but is under control of PKC signaling.

Purkinje cell dendritic tree development in the absence of excitatory neurotransmission and of brain-derived neurotrophic factor in organotypic slice cultures

Purkinje cell dendritic tree development in the absence of excitatory neurotransmission and of brain-derived neurotrophic factor in organotypic slice cultures

Purkinje cell dendritic tree development in the absence of excitatory neurotransmission and of brain-derived neurotrophic factor in organotypic slice cultures

The development of the dendritic tree of a neuron is a complex process which is thought to be regulated strongly by signals from afferent fibers. In particular the synaptic activity of afferent fibers and activity-dependent signaling by neurotrophic factors are thought to affect dendritic growth. We have studied Purkinje cell dendritic arbor development in organotypic cultures under suppression of glutamate-mediated excitatory neurotransmission, achieved with multiple combinations of blockers of glutamate receptors. Despite the presence of either single receptor blockers or combinations of blockers predicted to fully suppress glutamate-mediated excitatory neurotransmission Purkinje cell dendritic arbors developed similar to those of control cultures. Furthermore, Purkinje cell dendritic arbors in organotypic cultures from brain-derived neurotrophic factor (BDNF) knockout mice or after pharmacological blockade of trk-receptors also developed in a way similar to control cultures. Our results demonstrate that during the stage of rapid dendritic arbor growth signals from afferent fibers are of minor importance for Purkinje cell dendritic development because a seemingly normal Purkinje cell dendritic tree developed in the absence of excitatory neurotransmission and BDNF signaling. Our results suggest that many aspects of Purkinje cell dendritic development can be achieved by an intrinsic growth program.

Protein kinase C: its role in activity-dependent Purkinje cell dendritic development and plasticity.

The cerebellum is a central organ in the control of motor learning and performance. In this respect, the cellular plasticity model systems of multiple climbing fiber elimination and long-term depression have been intensively studied. The signalling pathways involved in these plastic changes are now well understood on a molecular level and protein kinase C (PKC) activity appears to be crucially involved in both processes. Furthermore, as shown in recent studies, Purkinje cell dendritic development also critically depends on the activity of PKC. Thereby, the Ca(2+)-dependent PKC subtypes, activated by synaptic inputs through metabotropic glutamate receptors, trigger functional changes as well as long-term anatomical maturation of the Purkinje cell dendritic tree during cerebellar development at different time levels. This review summarizes these findings and forwards the hypothesis of a link between the functional mechanisms underlying LTD and the differentiation of Purkinje cell dendrites.

The Ducky Mutation in Cacna2d2 Results in Altered Purkinje Cell Morphology and Is Associated with the Expression of a Truncated α2δ-2 Protein with Abnormal Function

The mouse mutant ducky, a model for absence epilepsy, is characterized by spike-wave seizures and cerebellar ataxia. A mutation in Cacna2d2, the gene encoding the alpha 2 delta-2 voltage-dependent calcium channel accessory subunit, has been found to underlie the ducky phenotype. The alpha 2 delta-2 mRNA is strongly expressed in cerebellar Purkinje cells. We show that du/du mice have abnormalities in their Purkinje cell dendritic tree. The mutation in alpha 2 delta-2 results in the introduction of a premature stop codon and predicts the expression of a truncated protein encoded by the first three exons of Cacna2d2, followed by 8 novel amino acids. We show that both mRNA and protein corresponding to this predicted transcript are expressed in du/du cerebellum and present in Purkinje cells. Whereas the alpha 2 delta-2 subunit increased the peak current density of the Ca(V)2.1/beta(4) channel combination when co-expressed in vitro, co-expression with the truncated mutant alpha 2 delta-2 protein reduced current density, indicating that it may contribute to the du phenotype.

Cerebellar diffuse amyloid plaques are derived from dendritic Abeta42 accumulations in Purkinje cells.

beta-amyloid(1-42) (Abeta42)-rich amyloid plaques (APs) may be derived from destroyed neurons that were burdened with extensive intracellular Abeta42 accumulations. Since most cells that accumulate Abeta42 express the alpha7 nicotinic acetylcholine receptor (alpha7nAChR), we examined the relationship between the intracellular accumulation of Abeta42 and the expression of the alpha7nAChR in cells from the cerebellum of sporadic Alzheimer's disease (AD) patients. Abeta42, but not Abeta40 or Abeta43, accumulates intracellularly in Purkinje, Golgi II, stellate and basket cells in the AD cerebellum, all of which express the alpha7nAChR. Abeta42 deposits were also prominent within dendrites of Purkinje cells, especially at points of their bifurcation that were often occluded with this material. Diffuse APs appeared to represent the remnants of destroyed Abeta42-laden segments of Purkinje cell dendritic trees. Similarly, the accumulation of Abeta42 and early loss of Golgi II cells in AD cerebella correlated directly to their high level of alpha7nAChR expression. Furthermore, the presence and relative abundance of neuron-derived Abeta42/alpha7nAChR-positive materials within Bergman glia may be indicative of the stage of AD. These data are consistent with a role for the alpha7nAChR in mediating intracellular Abeta42 accumulation and also support the notion that the intracellular and intradendritic accumulation of Abeta42 may eventually result in cell lysis and the formation of APs.

Altered dendritic development of cerebellar Purkinje cells in slice cultures from protein kinase Cγ-deficient mice

Protein kinase C (PKC) is a key molecule for the expression of long-term depression at the parallel fiber–Purkinje cell synapse in the cerebellum, a well known model for synaptic plasticity. We have recently shown that activity of PKC also profoundly affects the dendritic morphology of Purkinje cells in rat cerebellar slice cultures suggesting that synaptic efficacy and dendritic development may be controlled by similar intracellular signalling pathways. Here we have analyzed the role of the γ-isoform of protein kinase C (PKCγ), which is strongly and specifically expressed in Purkinje cells, during dendritic development. After pharmacological treatment with PKC modulators, phosphorylation of PKCγ at serine 660 was altered in cerebellar slices suggesting that a change of PKCγ activity by these treatments was taking place within the Purkinje cells. In PKCγ-deficient mice, Purkinje cell dendritic trees were enlarged and had an increased number of branching points compared to wild-type mice indicating a role for the PKCγ isoform as a negative regulator of dendritic growth and branching. Furthermore, the branching-stimulating effects of the PKC inhibitors 2-[1-(3-dimethylaminopropyl)indol-3-yl]-3-(indol-3-yl)maleimide and Gö6976 found in wild-type cultures were abolished in the absence of PKCγ. In contrast, the strong inhibitory effect on dendritic growth by the PKC activator phorbol-12-myristate-13-acetate (PMA) did not require the presence of the PKCγ isoform since it was still present in the cultures of PKCγ-deficient mice. Our results clearly demonstrate an involvement of PKCγ in Purkinje cell dendritic differentiation in cerebellar slice cultures.

Hypoxemia near mid-gestation has long-term effects on fetal brain development.

We tested the hypotheses that an episode of hypoxemia near mid-gestation in fetal sheep has long-term effects on brain development and that the extent and type of damage is related to the stage of development within a particular brain structure at the time of the hypoxemia. Fetal sheep (n = 8) were made hypoxemic at 90 +/- 2 days (term approximately 147 days) by restricting the maternal blood supply to the placenta for 12 hours (h) using a vascular clamp so as to reduce fetal arterial O2 saturation by 50%-60%. Fetuses were killed 35 days later and the brains analysed histologically and immunohistochemically. Age-matched fetuses (n = 8) were used as controls. Gross brain damage was observed in only 1 fetus, the most acidemic during the period of hypoxemia. There was a reduction of 12% (p < 0.05) in the cross-sectional area of the cerebral cortex in hypoxemic fetuses compared with controls. In lobule 6 of the cerebella of hypoxemic fetuses, significant reductions were seen in (a) the volume density of Purkinje cells (33%), (b) the width of the molecular layer (13%), (c) the area of the inner granule cell layer (13%), (d) the area of the white matter (18%), and (e) the total cross-sectional area (15%). There were also significant reductions in the area of arborization of Purkinje cell dendritic trees (50%), in the branching density (25%), and in the number of dendritic spines (31%). In the ventral hippocampi of hypoxemic fetuses, there was a 36% reduction (p < 0.05) in the volume density of CA1 pyramidal cells and a 50% increase (p < 0.05%) in the number of astrocytes. We conclude that an episode of hypoxemia near mid-gestation reduces neuronal numbers in the hippocampus and cerebellum and probably also in the cerebral cortex. The growth of neural processes in a particular region will be significantly retarded if the hypoxemia occurs at an early stage of the growth of neural processes (e.g. cerebellum) but not if development is well advanced at the time of the insult (e.g. hippocampus). Damage is sustained in the white matter of the cerebral hemispheres if the insult is particularly severe. Together, these deficits could affect neural connectivity and impair postnatal brain function.

Neonatal Borna disease virus infection in the rat causes a loss of Purkinje cells in the cerebellum.

Viral insults that occur during early postnatal periods, can affect neuronal systems which exhibit significant postnatal development, such as the cerebral cortex and cerebellum. Borna disease virus (BDV) is a single-strand RNA virus which replicates in the nervous system of many species after experimental inoculation and causes acute neurological disease. Neonatal rats infected with BDV do not mount an aggressive response to the virus like their adult counterparts, but instead develop a persistent BDV infection with less overt clinical sequelae. Recently, the cerebellum, a neural structure associated with regulation of motor behavior, and perhaps with higher cognitive functions, has been demonstrated to be a target of neonatal BDV infections in rats (Bautista et al, 1995). In the present study neonatal rats were infected with BDV and their cerebella were analyzed histologically and immunohistochemically at 7 months of age. The cerebella of infected animals were reduced in size but normal foliation and laminar organization was present. However, as visualized with immunohistochemistry for the Purkinje cell-specific antigen calbindin, there were numerous gaps within the Purkinje cell layer and in the molecular layer which contains the Purkinje cell dendritic trees. We estimated the number of Purkinje cells and found there was an approximately 75% loss of PC in adult rats neonatally infected with BDV. These results suggest that neonatal BDV infection may either (1) target the PC and cause the death of these cells directly or (2) acts indirectly by triggering an immune response which is then responsible for the loss of these cells.

Natural variability in the number of dendritic segments: Model-based inferences about branching during neurite outgrowth

A study was made of the possible basis for naturally occurring variations in the number of segments in individual dendritic trees. Distributions of the number of terminal segments have been studied in dendrites from rat, cat, and frog motoneurons, basal dendrites from rat visual cortex pyramidal and non-pyramidal neurons, in rat cerebellar Purkinje cell dendritic trees, and in human hippocampal dentate granule cells. By means of a mathematical model for dendritic branching, it was shown that the variation in the number of dendritic segments can be accounted for by assuming that new branches during neurite outgrowth are formed randomly at terminal segments. The observed terminal segment number distributions could be closely approximated by additionally assuming that branching probabilities decline with increasing number of terminal segments in growing dendrites. The pyramidal neuron group differed significantly from the other neuron groups in such a way as to suggest that this decline is stronger than in the dendrites of other types of neurons. By using literature data on the mean number of terminal segments in rat cerebellar Purkinje cells, measured at different times during early development, an estimate could be obtained of the time-course of the branching probabilities. The branching probability of a terminal segment was found to be in the order of 0.002 per hour in the first 4 weeks postnatal with a 5-fold transient increase in the second week.

Abnormal Purkinje cell dendrites in lurcher chimeric mice result from a deafferentation-induced atrophy.

Previous studies of Purkinje cell dendrites in lurcher<-->wild-type mouse chimeras (lurcher chimeras) have documented the surprising occurrence of unusual atrophic dendritic morphologies among the wild-type cells of the mosaic cerebella. We have hypothesized that these aberrant morphologies arise from a process of developmental deafferentation that is due to the unique loss of mutant Purkinje cells in these chimeras. These earlier studies left unanswered the question of whether the abnormal dendrites were the result of a blocked developmental process (agenesis) or regressive events that deform a previously well-developed dendritic arbor (atrophy). Using a set of simple morphometric measures, we now examine wild-type Purkinje cells in young lurcher chimeras. At postnatal day 20, normal Purkinje cell development is nearly but not fully complete. In lurcher chimeras, the morphologies of the wild-type Purkinje cell dendrites are similar to those in wild-type controls of the same age. This means that they are larger in height, width, and cross-section than their counterparts in adult lurcher chimeras. The younger cells exhibit almost none of the atrophic morphologies described in mature animals. We conclude that the aberrant morphologies found in adult lurcher chimeras arise from atrophy rather than through a failure in development. Furthermore, consideration of the details of the wild-type dendrites in the lurcher chimeras leads to the proposal that the height and width of the Purkinje cell dendritic tree are controlled by two independent mechanisms.

Purkinje cell dendritic arbors in chick embryos following chronic treatment with an N‐methyl‐D‐aspartate receptor antagonist

The normal development of Purkinje cell dendrites is dependent on afferent innervation. To investigate the role of neuronal activity in Purkinje cell dendritic development, chick embryos were chronically treated with a potent, selective, and systemically active competitive N-methyl-D-aspartate (NMDA) receptor antagonist, NPC 12626. The NMDA receptor was chosen as a target for pharmacological blockade because of the importance of the NMDA receptor in synaptic plasticity and stabilization in development. Chick embryos were given daily injections of NPC 12626 (25 to 100 mg/kg) from embryonic day 14 (E14) to E17. The initial injections of NPC 12626 dramatically blocked embryo movements, but activity levels partially recovered following subsequent injections. Embryo movements were reduced by 24% at the end of the experiment. Embryos were killed on E18, and their brains processed for Golgi-Cox staining. The morphology of Golgi-stained Purkinje cells in drug-treated embryos was similar to control embryos. Morphometric analysis showed, however, that chronic treatment with NPC 12626 resulted in a 19% reduction in Purkinje cell dendritic tree area and a 13% reduction in the number of dendritic branch points. The overall width and height of the drug-treated dendritic trees were not significantly different from controls, suggesting that NPC 12626 reduced Purkinje cell dendritic area by interfering with branch formation. The volume of the granule cell layer and the heights of the molecular and external granule cell layers was not reduced, suggesting that NPC 12626 treatment did not simply delay development. These results suggest that activation of the NMDA receptor may mediate the afferent-target interactions in the cerebellum that regulate the elaboration of Purkinje cell dendrites.

Physiology of Bergmann Gllal Cells

While Bergmann glial cells play an important role in the development of the cerebellum they were thought to serve as passive insulators of the Purkinje cell dendritic tree and its synaptic connections. New results challenge this view and demonstrate that Bergmann glial cells are equipped with a large repertoire of receptors allowing them to sense the activity of synapses. These receptors have distinct biophysical and pharmacological features activating second-messenger pathways in the Bergmann glial cells. It is evident that the synapse has to be viewed as consisting of three elements, the presynaptic and postsynaptic region and the glial ensheathment. All three elements of this synaptic complex may undergo plastic changes as a prerequisite for central nervous system plasticity. Glial cells could interfere with synaptic transmission by communicating with neurons via the extracellular space, e.g., by modulating ion concentrations or transmitter levels in the cleft (Fig. 6).

Association of BEN glycoprotein expression with climbing fiber axonogenesis in the avian cerebellum

In a previous study, we have identified an avian 100 kDa membrane glycoprotein that we called BEN and demonstrated that it is transiently present in the CNS and PNS on the cell somas and axons of neurons that establish the peripheral neuronal circuitry. We report here that in the developing chick cerebellar system BEN is selectively expressed on fibers whose ingrowth and synaptogenesis pattern corresponds to that described for climbing fibers. We have constructed quail-chick chimeras in which the chick mesencephalon and anterior metencephalon were replaced by their quail counterparts, thus generating a cerebellum and mesencephalon exclusively composed of quail cells whereas the main nuclei emitting afferent fibers to the cerebellar cortex were of chick origin. Then, using species-specific monoclonal antibodies we were able to show in double staining experiments that BEN protein is specifically expressed on fibers arising from the inferior olivary nucleus. The spatiotemporal pattern of BEN expression on the climbing fibers leads us to propose that this molecule is associated with the growth of these fibers and with the establishment of synapses between them and the Purkinje cell dendritic tree.

Nerve growth factor receptor immunoreactivity in the new world monkey (Cebus apella) and human cerebellum.

The present study used the NGFR-5 monoclonal antibody raised against human nerve growth factor receptor (NGFR) to determine the extent of NGFR immunoreactivity within the embryonic and young adult Cebus apella cerebellum as well as the human cerebellum. Immunohistochemically processed tissue revealed NGFR expressing Purkinje cell somata, axons, and dendrites, the latter being observed within the molecular layer of both adult species. Within all regions of the cerebellum we observed both darkly and lightly immunostained Purkinje cells. The proximal axons of these cells, which were visualized for short distances within the granular cell layer, appeared to contain bulbous aggregates of reaction product. In sagittal sections, the full extent of the Purkinje cell dendritic tree was observed in the more lightly stained portions of the cerebellum. In situ hybridization experiments revealed NGFR mRNA within Purkinje cells in a pattern similar to that seen with immunohistochemistry. The distribution of NGFR immunoreactivity within the cerebellum exhibits a general topographic organization with the heaviest and most consistent staining occurring within the archi- and neocerebellum and weaker staining within the paleocerebellum. In fetal Cebus monkey cerebellum obtained at gestational day 50 and 70, NGFR immunoreactivity was observed as a band composed of developing Purkinje cell neurites. These profiles were seen in the paleo- and neocerebellum, but not the archicerebellum. The present investigation is the first demonstration of NGFR immunoreactive profiles in the adult monkey and human cerebellum. These findings suggest that nerve growth factor may influence locomotor and vestibular behaviors that are mediated by cerebellar circuity. The precise mode of action for the NGF/NGFR system within the cerebellum remains to be determined.

Permanent alterations of the dendritic tree of cerebellar Purkinje neurons in the rat following postnatal exposure to cis-dichlorodiammineplatinum

The aim of this study was to characterize further the effect of cis-dichlorodiammineplatinum (cisDDP) on cerebellar Purkinje neurons of the immature rat. Ten-day-old rats were treated with cisDDP subcutaneously and killed after 1, 7, 20 or 65 days. The cerebellar vermis was impregnated by the Golgi-Cox method to evaluate the extent of morphological maturation of the Purkinje cell dendritic tree. One day after treatment, the dendritic network of Purkinje cells of treated animals was poorly developed and the cell somata still showed numerous perisomatic processes. This indicates that cisDDP interferes with the organization of microtubules and microfilaments by the cell. Later, several abnormal shapes of the Purkinje cell dendritic tree were observed. These included: (1) elongated primary dendrites; (2) asymmetrical dendrites; (3) sprouting of secondary and spiny branches in two planes of the molecular layer; and (4) damming of spiny branchlets at the pial surface. Moreover, all the dendritic networks of Purkinje cells in treated animals were of a lower Strahler order than in controls. All these data suggest that the late anomalies of the dendritic trees are secondary to the general cisDDP-induced damage of the cerebellar cortex, rather than being a primary effect of the drug on the dendritic tree growth.

Topological Link-Vertex Analysis of the growth of Purkinje cell dendritic trees in normal, reeler, and weaver mice.

The growth of Purkinje cell dendritic trees in normal, reeler, and weaver mice has been defined by using Link-Vertex Analysis. Growth probably occurs in three phases in normal and cortically located reeler trees. Phase I, completed by 7 days postnatum (dpn), establishes a rudimentary tree of 100 segments by random terminal branching. Phase II lasts from 7 to 20 dpn when some 690 and 450 segments are generated in normal and cortical reeler trees respectively. Phase II is initiated by an inductive stimulus mediated by a finite number of parallel fibres. Thereafter, dendritic trees develop a similar topology in normal and cortical reeler Purkinje cells through random interactions between parallel fibres and dendritic growth cones. We have defined this process with the aid of computer simulation techniques. Interactions appear to be restricted to a narrow growth front occupied by the highest centrifugally ordered terminals behind which adhesions occur at a greatly reduced frequency. The cortical reeler tree thus has fewer segments than normal because fewer parallel fibres are available, but it is surprisingly normal in most other respects. Phase III is a period of remodelling and extends from 20 dpn into adulthood when high-ordered terminals are eroded and middle-ordered terminals are added, with no change in total segment number in both normal and cortical reeler trees. Weaver and deeply placed reeler Purkinje cell dendritic trees are not influenced by parallel fibres. Accordingly, their growth is arrested at the end of Phase I, when both types of mutant tree have generated 100 segments by unrestrained random terminal branching. In the absence of parallel fibres, Phase II is not induced and remodelling, characteristic of Phase III, does not occur.

Identification of a Ca 2+-ATPase in cerebellar Purkinje cells

The expression of a sarcoplasmic reticulum (SR)-like Ca 2+-ATPase was studied in the adult chicken cerebellum. A monoclonal antibody, CaS/Cl-IgG, specific for the cardiac/slow-twitch skeletal muscle SR Ca 2+-ATPase, was used as a probe of protein expression. An immunoblot analysis showed that CaS/Cl-IgG recognized similar size polypeptides in adult chicken heart and cerebellum. CaS/Cl-IgG recognized fragments of similar size after limited tryptic digestion of cardiac and cerebellar membranes. A two-dimensional α-chymotryptic peptide map analysis demonstrated that the cardiac and cerebellar Ca 2+-ATPases were structurally very similar. Immunofluorescence microscopy localized the cerebellar Ca 2+-ATPase to Purkinje cell bodies and dendritic trees. These results suggest that the well-known Ca 2+ uptake system of skeletal and cardiac muscle SR has a remarkably similar counterpart in some neurons.

Link-Vertex Analysis of Purkinje cell dendritic trees from the murine cerebellum

A method of analysis of dendritic trees is described, called Link-Vertex Analysis, which defines the branching patterns (topology) by means of a discontinuous function, the terminal/link (TL) vertex ratio, computed at all successive vertex orders within a network, and represented as a graphical plot for a given tree. The T/L vertex ratio is defined as the number ( n) of terminal (pendant) vertices (Vp) divided by the total number ( n) of all link vertices (both dichotomous, Vd, and trichotomous, Vt) multiplied by the number of segments arising from each vertex—i.e., nVp(2nVd+ 3nVt)per vertex order. Hypotheses of growth are simulated by computer modelling and the distributions of computer generated T/L vertex ratios per order compared with those of actual dendritic tress. The differential distribution of T/L vertex ratios per order in trees grown by either random terminal or random segmental branching discriminates between these two forms of growth. Significant advances of this method include the detection of regional variations in growth and remodelling in a tree and the incorporation of Vt with little peturbation of the Vd data. The analytical powers of this method have been tested using the dendritic fields of 20-day-old mouse cerebellar Purkinje cells defining possible temporal sequences and relative contributions of random and non-random terminal growth, and also the frequency of branching over a growth front relative to that of the rest of the tree.

Computer-assisted analysis of the developing Purkinje neuron. II. A comparative study of the dendritic development in differently aged kittens and in organotypic cultures explanted at the same ages

Golgi-prepared cerebella from 10–13- (model K) and 22-day-old (model C) kittens were analyzed and compared with 30–40 days in vitro (DIV) horseradish peroxidase (HRP)-stained organotypic cultures from kitten cerebella explanted at comparable ages (12 (model E) and 21 (model T) postnatal days). In addition, HRP-stained cerebella explanted from 1-day-old kittens were used as a reference (model N). Computer reconstructions and morphometric parameters (14 variables) allowed a comparative quantitative analysis of the Purkinje cell dendritic trees and axonal processes. The biplanar arrangement of the dendritic sheets observed in vivo at 10–13 days (K) was replaced by a mostly bipolar one in the in vitro Purkinje cells (E) explanted at the same age with their axon following the direction in the number of dendritic. The dendritic expansion observed in intact animals between 12 and 22 days was mostly due to an increase in the number of dendritic segments havinig shorter lengths rather than an increase in the number of rows. A discriminant analysis permitting the recognition of 3 populations of Purkinje cells in both groups reinforced the previous hypothesis that the dendritic shapes are modelled by specific afferents pre-existent in vivo.