SEARCH

SEARCH BY CITATION

Keywords:

  • cerebellum;
  • development;
  • proliferation;
  • migration;
  • cytostatic drug

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. LITERATURE CITED

We evaluated the relationship among proliferation, death and migration of granule cells in lobules VI–VIII of vermis, in comparison with lobule III, during cerebellar development. To this aim, a single injection of cisplatin, i.e., a cytostatic agent that is known to induce death of proliferating granule cells, was given to 10-day-old rats. Histochemical markers of proliferating (PCNA immunoreaction) and apoptotic (TUNEL staining) cells were used; the variations of the external granular layer (EGL) thickness were evaluated in parallel. After PCNA and TUNEL reactions, evident changes of the whole EGL were found on PD11 (1 day after treatment), when a reduction of the thickness of this layer was found in treated rats, mainly in consequence of the high number of apoptotic cells in all the cerebellar lobules. On PD17 (7 days after treatment), a thick layer of proliferating cells was observed in lobules VI–VIII of treated rats, while the peculiar pattern of the normal development showed a thin EGL. At the same time, in treated rats, the number of apoptotic cells in EGL was low. In all developmental stages of treated rats, after GFAP immunoreaction, glial fibers appeared twisted, thickened, and with an irregular course; intensely labeled end-feet were present. The damage of radial glia suggests an alteration of migratory processes of granule cells, which is also evidenced by the decreased thickness of the premigratory zone of the EGL. Injured radial glia fibers were restricted to lobules VI–VIII and they persisted at PD30, leading to the presence of ectopic granule cells in the molecular layer, as we previously described. © 2005 Wiley-Liss, Inc.

During postnatal development of rat cerebellum, the external granule cell layer (EGL) represents the matrix area in which two distinct zones can be clearly identified, the proliferative and the premigratory zones (Mareš et al.,1970; Altman,1972). The postmitotic cells of the EGL migrate to their final destination in the internal granule cell layer (IGL) (Rakic,1971; Altman,1972; Hatten,1990); the correct positioning of the cells is important for the final cytoarchitecture and in particular for the pattern of synaptic connections in laminated brain regions such as the cerebellum (Rakic et al.,1994).

Proper development of the nervous system requires apoptosis, a form of programmed cell death, that systematically removes a lot of cells not only in the proliferative zones, but also among the postmitotic cells (Blaschke et al.,1996,1998). In the cerebellum, a few pycnotic cells, presumed to be degenerating, are found during all stages of development of the external granular layer (Miale and Sidman,1961). However, cerebellar granule cells undergo extensive nuclear DNA fragmentation between postnatal days 5–9 and it was deduced that the deaths prior to synaptogenesis might help regulate cell number (Wood et al.,1993). Considering the CNS cytoarchitecture, the synaptic targets of the developing neurons have been the major focus of interest for understanding the regulation of naturally occurring neuronal death, though variations in neuron/target ratios may be compensated by other changes, including the efficacy, number, and distribution of synapses, and the branching of axons and dendrites (Oppenheim,1991). Recently, Lossi et al. (2002) have proposed that both synapse-independent and synapse-dependent programmed cell death affect the granule cells during postnatal development; the former occurs when premigratory granule cell precursors are still populating the EGL, while the latter concerns the postmigratory neurons in the IGL.

It is worthwhile to note that when proliferation is disrupted, migration is often affected. This can be also demonstrated with X-ray irradiation (Doughty et al.,1998), methyl-azoxy-methanol (MAM) (Doughty et al.,1998; Lafarga et al.,1998), or methylmercury (Rodier et al.,1984; Choi,1986,1989; Ponce et al.,1994).

In this work, we have chosen to treat the rats with cisplatin, a cytostatic drug that interferes with mitotic cells in the developing rat cerebellum (Mareš et al.,1986; Scherini and Bernocchi,1994). Our goal was to assess the relationship of fundamental mechanisms regulating cerebellum morphogenesis, examining proliferation and death, and migration of granule cells in the early and late phases following drug injury. To the aim of the present research, we have considered the histochemical markers of proliferating cells [proliferating cell nuclear antigen (PCNA) immunoreaction], apoptotic cells [terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling (TUNEL) staining], and the thickness of the EGL. A marker of radial glia [glial fibrillary acidic protein (GFAP)] has been analyzed to describe cisplatin-induced changes in the morphology and reactivity of Bergmann radial glial fibers that are the guide substrate of granule cells; in fact, the feature of these fibers gives information on the proper granule cell migration. Finally, we will discuss the relationship, if any, between the early developmental phases of granule cells (before their positioning in the IGL) and the altered maturation of Purkinje cells we have recently reported (Pisu et al.,2004)

In our experimental model, we have taken into account the vulnerable periods of cerebellum development, during which the basic histogenetic steps are sensitive to environmental insults (Barone et al.,2000; Rice and Barone,2000). In particular, we thought important to consider the different timing of the morphogenesis of cerebellar lobulation (Altman,1982; Doughty et al.,1998), and therefore we compared the neocerebellar lobules VI–VIII and a paleocerebellar lobule, i.e., lobule III; the cell lining of fissure I and fissure II was also analyzed.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. LITERATURE CITED

Ten-day-old Wistar rats were given a single s.c. injection of cisplatin (0.5 mg/ml; Teva Pharma, Italy) on the nape of the neck at a dose of 5 μg/g body wt [corresponding to the therapeutic dose suggested by Bodenner et al. (1986)].

Six hours (PD10), 1 day (PD11), 7 days (PD17), and 20 days (PD30) after drug administration, treated (three per stage) and untreated control rats (three per stage) of the same age were deeply anesthetized with an intraperitoneal injection of 35% chloral hydrate (100 μl/100 g body wt) and perfused intracardially with saline followed by 4% paraformaldehyde in a 0.1 M phosphate buffer, pH 7.3. The brains were immediately excised, postfixed in the same fixative medium at 4°C for 1.5 hr, kept in 30% sucrose in a 0.1 M phosphate buffer at 4°C for 48 hr, and frozen in liquid nitrogen. Twelve micrometer thick cryostatic sections of cerebellar vermis were cut on the midsagittal plane and collected on silan-coated slides. The sections from PD10, PD11, and PD17 stages and from all the stages were then processed for TUNEL staining and GFAP immunocytochemistry, respectively.

Moreover, treated (three per stage) and untreated rats (three per stage) at PD10, PD11, and PD17 stages were deeply anesthetized with an intraperitoneal injection of 35% chloral hydrate (100 μl/100 g body wt), the brains were immediately excised, fixed in Carnoy fixative for 18 hr at 4°C, kept in absolute ethanol and in aceton, then embedded in Paraplast. Eight μm thick sections of cerebellar vermis were cut in the midsagittal plane and collected on silan-coated slides. The sections were then processed for the PCNA immunocytochemistry.

To avoid possible staining differences due to small changes in the procedure, the reactions were carried out simultaneously on slides of control and treated animals at all stages. All experiments were conducted in accordance with the guidelines of the Italian law 116/92 regarding the care and use of laboratory animals.

PCNA Immunocytochemistry

Immunoreaction was performed overnight at room temperature using a primary mouse monoclonal anti-PCNA antibody (Oncogene, Boston, MA) diluted 1:600 in PBS. A biotinylated anti-mouse secondary antibody and an avidin-biotynilated horseradish peroxidase complex (Vector Laboratories, Burlingame, CA) were used to reveal the sites of antigen/antibody reaction. 3,3′-diaminobenzidine tetrahydrochloride (DAB; Sigma, St. Louis, MO) was used as a chromogen. Sections were dehydrated in ethanol, cleared in xylene, and mounted in Eukitt (Kindler, Freiburg, Germany).

For the control staining, some sections were incubated with normal mouse serum, or PBS, instead of the primary antibody. No immunoreactivity was present in these sections.

TUNEL Staining

The reaction was performed using a TdT–mediated dUTP nick end-labeling kit (Roche, Penzberg, Germany); the sections were incubated with 50 μl of TUNEL mixture conjugated with FITC (5 μl enzyme solution and 45μl label solution) according to the manufacturer's instructions for 3 hr at room temperature. After washing, the nuclei were counterstained with 0.1 μg/ml Hoechst 33258 for 10 min; coverslips were finally mounted in a drop of Mowiol (Calbiochem, Darmstadt, Germany). For the control staining, some sections were incubated without enzyme solution. No reactivity was present in these sections.

GFAP Immunocytochemistry

Immunoreaction was performed overnight at room temperature using a primary goat polyclonal anti-GFAP (C-19) antibody (Santa Cruz Biotechnology, Santa Cruz, CA) diluted 1:200 in PBS; after washing, the sections were incubated for 1 hr with the secondary antibody, Alexa-Fluor 488-conjugated anti-goat, at a dilution of 1:50 (Molecular Probes, Space, Milano, Italy) and then the nuclei were counterstained with 0.1 μg/ml Hoechst 33258 for 10 min; coverslips were finally mounted in a drop of Mowiol (Calbiochem) and the slices were viewed in fluorescence microscopy with an Olympus BX50 microscope equipped with a 100 W mercury lamp. The following equipment was used: 330–385 nm excitation filter (excf), 400 nm dichroic mirror (dm), and 420 nm barrier filter (bf) for Hoechst 33258; 450–480 nm excf, 500 nm dm, and 515 nm bf for Alexa 488. Images were recorded with an Olympus Camedia C-2000 Z digital camera and stored on a PC running Olympus software for processing and printing. For control staining, some sections were incubated with PBS instead of the primary antibodies. No immunoreactivity was present in these sections.

Quantitative Evaluation

The thickness of the EGL was directly measured on PCNA-stained slides by using a 10× Zeiss ocular micrometer with a 40× objective; the depth of the outer proliferating and premigratory zones was also separately considered. For each section (five sections for each control and five sections for each treated rat), measurements were taken in randomly selected points of the surface in neocerebellar lobules VI–VIII (150 measurements) and in the paleocerebellar lobule III (60 measurements); moreover, the evaluations concerned the thickness of EGL along fissure I (90 measurements) and fissure II (90 measurements). The mean average (μm) was calculated and the Student's t-test was provided to obtain the significance of differences between the experimental groups.

Sagittal sections of the cerebellum were reproduced by means of a camera lucida attached to a Leitz Laborlux K microscope at 40× objective magnification. On the drawings, the length (mm) of the EGL of lobules VI–VIII and III was measured with a curvimeter. Then, the counting of TUNEL-labeled cells was performed with an Olympus BX50 microscope using a 40× objective. We considered 5 sections for each control (15 sections in total) and 5 sections for each treated rat (15 sections in total). The mean average (cell number per mm) was calculated and the Student's t-test was provided to obtain the significance of differences between the experimental groups.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. LITERATURE CITED

Histological Observations

PCNA immunoreactivity.

Labeling with the anti-PCNA antibody showed the prominent response in the EGL, where the outer proliferating (labeled) zones and the inner (not labeled) zones can be detected in all the lobules and along fissures. Labeled nuclei, likely glial cells or apoptotic cells, were also observed in the molecular layer (ML), internal granular layer (IGL), and in the white matter.

From PD10 to PD17, the EGL of control rats progressively decreased in thickness in all lobules and along fissures (Fig. 1a, c, e, and g). In particular, zones with proliferating and nonproliferating cells were clearly quantifiable on PD10 (not shown) and PD11 (Fig. 1c).

thumbnail image

Figure 1. Immunoreaction for PCNA in lobules VI–VIII at PD11, PD17. Control (a, c, e, and g) and treated rats (b, d, f, and h). On PD11, in controls (a and c), zones with PCNA-positive (asterisk) and nonproliferating cells are evident along fissure I. In the age-matched treated rats 1 day after treatment (b and d), the EGL appears thinner in comparison to the control; moreover, the anti-PCNA antibody also labels cells with apoptotic features (arrows, inset; d). On PD17, in controls the zone of proliferating cells is thin (e and g). After treatment at the top of the lobules (f and h), a thickened layer of proliferating cells is observed (asterisk). Note the presence of hemorragic foci (arrowheads). Scale bar = 250 μm (a and b); 54 μm (c, d, g, and h); 137 μm (e and f).

Download figure to PowerPoint

The treatment with cisplatin showed on PD10 (not shown) the persistence of the EGL proliferating zone, but the premigratory zone appeared to be reduced. More evident changes of the whole EGL were found on PD11 (Fig. 1b and d); in fact, the EGL appeared thinner in treated rats (Fig. 1d) than in controls (Fig. 1c). In treated rats, the anti-PCNA antibody also labeled cells with apoptotic features (Fig. 1d, inset); the premigratory zone often appeared less identifiable.

The zone of proliferating cells was thin in controls on PD17 (Fig. 1e and g). After treatment, a peculiar pattern was found in neocerebellar lobules VI–VIII. In fact, only at the top of these lobules (Fig. 1f and h) was a thick layer of proliferating cells was observed, while along fissure I, the EGL was often absent (not shown). At this stage, hemorragic foci were found in the brain parenchyma (Fig. 1f).

TUNEL staining.

After in situ TUNEL staining, labeled nuclei were present in the EGL. Positive nuclei, however, were observed in all the layers of the cerebellar cortex, though Purkinje cells themselves were never labeled.

The apoptotic nuclei were few in the EGL of control rats at all stages (not shown), and this was lobules-independent. On PD11, TUNEL-positive (Tp) cells greatly increased in lobule III and lobules VI–VIII and along fissures in treated rats (Fig. 2a). TUNEL staining labeled few cells in the thick EGL of 17-day-old treated rats (Fig. 2b).

thumbnail image

Figure 2. Double staining for TUNEL reaction and Hoechst in lobules VI–VIII at PD11, PD17. Treated rats. The green fluorescence is related to TUNEL reaction, the blue fluorescence is due to the nuclear counterstaining with Hoechst 33258. On PD11 (a), several TUNEL-positive (apoptotic) cells are present in the EGL. On PD17 (b), few TUNEL-labeled cells are found in the EGL (arrows); in the molecular layer (ML), some TUNEL-labeled cells are detectable (arrowheads). Scale bar = 62 μm.

Download figure to PowerPoint

GFAP immunoreactivity.

In a comparison between controls and treated rats, changes were observed in lobules VI–VIII only; along fissures, no apparent differences emerged. No changes in GFAP immunoreactivity were detected at PD10.

Thin glial GFAP-positive fibers had a regular feature running in parallel in the ML and in the EGL in lobules VI–VIII of controls at PD11 (Fig. 3a). Few GFAP-positive fibers were observed at the top of lobules VI–VIII in treated rats (Fig. 3b); sometimes, the fibers appeared interrupted.

thumbnail image

Figure 3. Double staining GFAP immunoreaction and Hoechst in lobules VI–VIII at PD11, PD17, PD30. Control (a, c, and e). Treated rats (b, d, and f). The green fluorescence is related to GFAP immunoreactivity, the blue fluorescence is due to the nuclear counterstaining with Hoechst 33258. On PD11, in controls (a), thin glial GFAP-positive fibers are present in the EGL and ML. In treated rats (b), few fibers are present at the top of the lobules; sometimes they are interrupted. On PD17, in control rats (c), GFAP-positive fibers appear thin and regular. In treated rats (d), glial-positive fibers show a twisted and irregular feature (arrows); often these fibers have intensely stained thickenings (end-feet) at the pial surface of the EGL (arrowheads). On PD30, in comparison with controls (e), in treated rats (f), GFAP-positive fibers appear twisted, irregular, and interrupted; sometimes they form extended thickenings (end-feet; arrowhead) in the superficial ML. In the lower part of the ML, ectopic cells, identifiable as granule cells, are found (arrows). Scale bar = 62 μm (a and b); 54 μm (c–f).

Download figure to PowerPoint

GFAP-positive fibers were thin and regular, reaching the pial surface in the controls at PD17 (Fig. 3c). In lobules VI–VIII of treated rats, glial fibers appeared twisted with an irregular course (Fig. 3d) and sometimes they traversed the ML at various angles. Then, GFAP-positive fibers often had thickened and intensely stained end-feet in the pial surface of the EGL (Fig. 3d).

Lack of GFAP positivity was found in lobules VI–VIII of treated rats on PD30 (Fig. 3f) in comparison with controls (Fig. 3e). GFAP immunoreactive fibers were twisted and irregular (Fig. 3f); extended end-feet were also observed in the pial surface of the ML. Hoechst labeled ectopic cells in the proximal part of the ML of the same lobules (Fig. 3f).

Quantitative Evaluations

Total EGL thickness and its components, i.e., proliferating and premigratory zones, showed differences existing between lobules and along fissures at some stages. By contrast, the number of Tp cells did not differ among lobules.

Thickness of EGL in lobules III, VI–VIII, and along fissures I and II.

Generally, the trend of EGL thickness (Fig. 4) showed a decrease between PD10 and PD17, with lower values in treated rats in comparison with controls.

thumbnail image

Figure 4. Thickness of the external granular layer, proliferating and premigratory zones in lobules VI–VIII, lobule III, and along fissures I and II. On PD10, the differences of EGL thickness between controls and treated rats are significant (Student's t-test) only in lobules VI–VIII (P < 0.001). The decreased depth of EGL in 11-day-old rats after cisplatin is significant in all the lobules, though the highest reduction is observed in lobules VI–VIII. On PD17, increased thickness of EGL is visible in lobules VI–VIII. Both in proliferating and premigratory zones on PD10, the differences are significant only in lobules VI–VIII (P < 0.001). On PD11, while in the proliferating zone there are no differences between controls and treated rats, in the premigratory zone a significant decrease in thickness is observed in all the lobules and along fissures I and II (P < 0.001). On PD17, very significant differences are visible in both proliferating and premigratory zones; in particular, a drastic decrease of the thickness of premigratory zone along fissures I and II of treated rats was observed. In lobules VI–VIII, an increase in depth is evident in treated rats versus controls in the two EGL zones (P < 0.001), while in lobule III no differences are observed. Double asterisk, Student's t-test, P < 0.001. White columns, controls; gray columns, treated rats

Download figure to PowerPoint

In 10-day-old control rats, the depth of the EGL had the highest values at the top of lobules VI–VIII and along fissure I when compared to lobule III. After cisplatin treatment, a significant decrease was found in lobules VI–VIII (P < 0.001). The cisplatin injection induced a general significant decrease of EGL thickness at PD11; the highest decrease was found in lobules VI–VIII and in fissure I (P < 0.001).

A peculiar result concerned lobules VI–VIII, where a significant increase of EGL thickness was found after the cisplatin treatment at PD17 (P < 0.001); instead, a significant decrease was observed in lobule III and along fissure I (P < 0.001).

Proliferating zone of EGL.

The PCNA immunoreaction labeled the proliferating zone that was clearly distinguishable from the premigratory one. On PD10 and PD11 (Fig. 4), no significant differences were revealed in lobules III, VI–VIII, and along fissures I and II after the cisplatin treatment in comparison with controls. In lobules VI–VIII and along fissure II of treated rats, the thickness of the proliferating zone showed a significant increase (P < 0.001) compared to controls at PD17 (Fig. 4).

Premigratory zone of EGL.

Premigratory cells constituted a PCNA immunonegative zone below the proliferating layer. On PD10 (Fig. 4), the cisplatin caused a significant decrease (P < 0.001) of the thickness of this zone in lobules VI–VIII in comparison with controls. The premigratory zone was still present in controls but it had disappeared in treated rats at PD11 (Fig. 4); the differences were significant in lobules III, VI–VIII, and along fissures I and II (P < 0.001).

On PD17, the premigratory zone of lobules III and VI–VIII (Fig. 4) was not identifiable in control rats, while it was present along fissures I and II. After treatment with cisplatin, a very thin premigratory zone was present in lobules VI–VIII (P < 0.001), but not in lobule III and along fissures I and II (P < 0.001).

TUNEL labeling.

The number of Tp cells was low in the EGL of 10-day-old rats (Fig. 5); there were no significant differences between control and treated rats in lobules III and VI–VIII. On PD11 (Fig. 5), treated rats, in comparison with controls, showed a significant increase in the number of Tp cells in both lobules III and VI–VIII (P < 0.001).

thumbnail image

Figure 5. Number of TUNEL-positive cells per mm in lobules VI–VIII and lobule III. On PD10 and PD17, no significant differences are observed in treated rats versus controls. On PD11, a drastic and significant increase (Student's t-test) of TUNEL cells number is observed in all the lobules of treated rats (P < 0.001). Double asterisk, Student's t-test, P < 0.001. White columns, controls; gray columns, treated rats

Download figure to PowerPoint

Tp cells were rare in the EGL of 17-day-old control rats (Fig. 5). In treated rats, the incidence of Tp cells was low in lobule III as well as in the recovered EGL of lobules VI–VIII.

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. LITERATURE CITED

A previous study on the influence of several doses of cisplatin on the morphology of immature cerebellum revealed that the acute necrotizing effect of the cytostatic is more evident in the population of immature and dividing cells of the EGL than in the postmitotic one, such as in the IGL or among Purkinje cells (Mareš et al.,1986). However, we have not dealt with the interaction of the basic mechanisms of CNS development, i.e., proliferation, cell death, and migration.

Therefore, the present research has considered the immunocytochemical markers of dividing cells and programmed cell death to demonstrate the cisplatin-induced trend and distribution of cell proliferation/apoptotic cell death in crucial stages of postnatal cerebellar cortex development.

In the EGL, a relevant result has been obtained with regard to the premigratory inner zone, suggesting an evident interference of cisplatin with cell migration in the neocerebellar lobules. Alterations of radial glia fibers were observed at all developmental stages and they were maintained also after recovery of EGL proliferation. All the data contribute to the knowledge of developmental mechanisms in the critical and vulnerable phases of cerebellar lobule morphogenesis and indicate how and where the immature cerebellar cytoarchitecture reacts to injury.

Effects of Cisplatin on Proliferation and Death in EGL

The proliferating cell nuclear antigen is a 36 kDa protein that is required for DNA replication (Prelich et al.,1987a,1987b; Tan et al.,1987; Bauer and Burges,1988) and repair (Shivji et al.,1992; Zeng et al.,1994). The PCNA is synthesized during the late G1/early S-phase of the cell cycle, immediately preceding the onset of the DNA synthesis, and it is abundant during the S-phase and declines during the G2/M-phase (Kurki et al.,1986,1988) so that its expression declines as cells become quiescent.

PCNA-dividing cells were conspicuously present in the outer proliferating zone of the EGL in the developing cerebellum of untreated rats. The labeled zone of the EGL permits the distinguishing of the nonlabeled premigratory zone, i.e., the site from which newly postmitotic cells migrate into the IGL (Komuro and Rakic,1995; Zhang and Goldman,1996). Labeled cells were normally present in other layers of the cerebellar cortex and in the white matter (Tanaka and Marunouchi,1988; Lossi et al.,1998; Migheli et al.,1999; Ino and Chiba,2000). The PCNA-positive cells within the white matter could be immature glial cells. Alternatively, proliferating cells in the ML, the IGL, and in the white matter may correspond to cerebellar interneurons generated after interstitial proliferation (Zhang and Goldman,1996).

On the other hand, death of granule cells occurs naturally during developmental stages (Koppel et al.,1983; Smeyne and Goldowitz,1989; Ashwell,1990). Programmed cell death by apoptosis is currently visualized by the TUNEL procedure, which labels in situ fragmented DNA (Wood et al.,1993; Whiteside et al.,1998). In the developing cerebellum, apoptotic nuclei, as demonstrated after TUNEL staining, occur mainly within the EGL. Apoptotic nuclei are present in other cerebellar layers and in the white matter, although they are sparse in this location. Our findings showed that the number of apoptotic cells in the EGL of control rats was low on PD10 and PD11, in agreement with the decrease of cell death after the first postnatal week (Wood et al.,1993).

The balance of proliferating and apoptotic cells, and also premigratory cells, determines the thickness of the EGL, which represents a key factor in the cerebellum morphogenesis, since the variation of the EGL influences the density of granule cells in the IGL (Mareš et al.,1986; Doughty et al.,1998; Lossi et al.,1998) and therefore the subsequent synaptogenesis by parallel fibers in the cerebellar cortex.

The link between genesis and death of cells, and cell migration in the development of cerebellar lobules was largely supported by findings obtained after the cisplatin treatment. Cisplatin is a cytostatic drug introduced into clinical use, which activates a cascade of degenerative events in the developing cerebellum, both in mitotic and postmitotic cells (Scherini and Bernocchi,1994).

As previously reported (Mareš et al.,1986), a massive degenerating peak was found in the EGL on PD11, 1 day after treatment. Present findings indicate that the degeneration was due to apoptosis as evidenced by both PCNA immunoreaction and TUNEL staining; PCNA was also strongly expressed in apoptotic cells as a likely reflection of its other activity in nucleotide excision repair of damaged DNA (Shivji et al.,1992).

It is worthwhile to emphasize that in the present research, we have found that after treatment with cisplatin, apoptotic events were not present differently in the cerebellar lobules or along fissures, but they occurred in the entire EGL of the cerebellar cortex.

Activation of apoptosis occurred in the cerebellum within the first postnatal weeks in the weaver mouse (Gillardon et al.,1995; Wüllner et al.,1995; Migheli et al.,1997); after treatment with neurotoxic substances, such as MAM (Ciaroni et al.,1995,1998; Lafarga et al.,1997; Doughty et al.,1998) and ethanol (Liesi,1997; Light et al.,2002), cell death has been described in the cerebellum and other brain areas. In particular, the MAM treatment causes massive cell death of granule cell precursors, which has the biochemical and ultrastructural characteristics of apoptosis. Ultrastructural analysis showed fine structural changes of the cell nucleus in cells undergoing apoptosis, which clearly reflects a disruption of the nuclear compartments involved in the transcription, the processing, and the transport of RNA and are related to the patterns of DNA and RNA degradation (Lafarga et al.,1997). Hypertrophy and an increased thickness of EGL in MAM cerebella were observed, but the density of cells within the IGL decreased during the postnatal histogenesis (Doughty et al.,1998).

Finally, the recovered genesis of cells we found in the EGL of lobules VI–VIII on PD17, 7 days after the cisplatin treatment, was not affected by a massive apoptotic event, but it was accompanied by normal cell death, as it was shown by the number of TUNEL-positive cells. This finding will assume great value in the interpretation of the remodeled cytoarchitecture of cisplatin-treated rats at the end of cerebellum histogenesis.

Effects of Cisplatin on Cell Migration

The EGL thickness normally decreases during postnatal histogenesis (Altman,1972,1982; Doughty et al.,1998; Lossi et al.,1998) but, after treatment with cisplatin, a more prominent decrease was found on PD11, especially in lobules VI–VIII. The analysis of the two zones of the EGL evidenced how the premigratory component had a drastic decrease in all lobules and along fissures.

Notwithstanding the general loss of premigratory cells, GFAP immunofluorescence showed that cisplatin altered the pattern of radial glia in the neocerebellar lobules VI–VIII; in these lobules fibers were lacking or thicker and intensely labeled, also in the end-feet.

Acute disturbance of Bergmann glial processes was described 2 days after a single injection of the cytotoxic agent MAM (Lafarga et al.,1998), which produces extensive cell death in the EGL and overexpression of GFAP in the distal half of Bergmann fibers, including the end-feet at the pial surface. The last finding has been interpreted as a reactive response, in which changes in the interactions between Bergmann glia and granule cell precursors have a role (Lafarga et al.,1998).

In the glial reactions and alterations, signal molecules and receptors could be involved. GluR1 receptors had drastic temporal and spatial changes in the cerebellar histogenesis and they decreased in Lurcher rats, showing a loss of granule cells (Ryo et al.,1993); the findings have suggested that the decreased activity of the receptors affected the migration of cells from the EGL to the IGL. However, after treatment with cisplatin, we previously observed an increased immunoreactivity for the ionotropic glutamate receptors GluR1 and GluR4, mainly in the distal half of the radial glia fibers (Pisu et al.,2002). The different results we found could be interpreted as due to the fact that glial alterations were restricted only to confined zones, i.e., lobules VI–VIII of the cerebellum.

At the latest stage following the cisplatin treatment, i.e., on PD17 and PD30, distortions of radial glia remained. In particular, on PD17, the GFAP-immunostained fibers traverse the molecular layer at various angles. Following treatments with MAM and X-rays on E16-21 (Doughty et al.,1998), similar features of Bergmann glia were observed on PD14.

The cytoarchitecture of lobules VI–VIII on PD30, 20 days after the cisplatin treatment, induces us to consider greatly the involvement of the altered radial glia in the abnormalities of the cerebellar cortex, such as the ectopia of granule cells that occupied the inferior half of the molecular layer. On the other hand, the morphological pattern of radial glia was still damaged 20 days after treatment with cisplatin. Taken together, the analysis of GFAP reactivity clearly stressed the early action of cisplatin on migration effectors, with the possible involvement of signal molecules.

As to cell migration, particular emphasis must be placed on GFAP changes that, as other markers we have studied (Pisu et al.,2004), concerned lobules VI–VIII. In our opinion, what deserves particular importance is the vulnerable state of the development of lobules VI–VIII when treatment was performed, e.g., as regards critical events such as the regression of perisomatic spines/elimination of climbing fiber synapses on the somata of Purkinje cells (for a review, see Altman and Bayer,1997; Lopez-Bendito et al.,2001; Pisu et al.,2003).

Effects of Cisplatin-Injured Cell Proliferation and Migration on Cerebellar Cytoarchitecture

In our model of injury, we have reported that early morphological and molecular alteration of cell proliferation in the EGL is due to a high incidence of apoptotic cells. The consequent decrease of thickness in the premigratory zone of the EGL, the altered expression of GFAP, and the distorted morphology in the radial glia system could early affect the differentiation of Purkinje cell arbor and the formation of proper synaptic contacts by parallel and climbing fibers, as we previously described (Pisu et al.,2003).

Twenty days after the cisplatin injury, when the recovery of the neurogenetic processes occurred in the EGL, alterations of Bergmann radial glia persisted, probably leading to the ectopia of granule cells (Pisu et al.,2004). This event has been largely described in the morphological alterations induced by X-ray irradiation during cerebellar histogenesis (Altman,1973,1976); ectopia of granule cells is considered a consequence of late neurogenetic inputs and it is produced when the late descending granule cells stopped in the molecular layer (Altman,1973).

Besides irradiation, the MAM treatment induces ectopia of granule cells (de Barry et al.,1987; Garcia-Ladona et al.,1991; Fonnum and Lock,2000). This phenomenon is accompanied by overexpression of glutamate receptors (Garcia-Ladona et al.,1993).

The ectopia of granule cells has deep consequences on the proper development of cerebellar cortex cytoarchitecture. In the cerebellar cortex of the apical zones of lobules VI–VIII, remodeling of Purkinje cell dendrites, as shown by both morphological and molecular markers, was found after cisplatin (Pisu et al.,2004). Changed cytoarchitecture of the cerebellar cortex, from PD17 to PD30, includes reorientation of the main dendritic branches and branchlets of Purkinje cells and changes in the inhibitory control of the compartments of Purkinje cell dendrites.

The most intriguing result, which deserves particular attention to further development of the research in this field, is that ectopia, remodeled Purkinje cells, and altered radial glia are restricted to lobules VI–VIII and persisted after the end of development. In this light, we have to consider the crucial events in the cerebellar cortex at the end of the second postnatal week. The development of neocerebellar lobules VI–VIII is largely occupied as in other lobules by EGL proliferation, but probably their peculiar critical event is the regulation of its molecular environment affecting Purkinje dendrite growth, elimination of perisomatic spines and of transient climbing fiber contacts, and also granule cell migration. In fact signaling molecules have a prominent role in the cerebellum development. It is shown by the findings we have recently published (Pisu et al.,2003,2004) and is in progress.

LITERATURE CITED

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. LITERATURE CITED