Genesis of neuronal and glial progenitors in the cerebellar cortex of peripuberal and adult rabbits

Abstract

Adult neurogenesis in mammals is restricted to some brain regions, in contrast with other vertebrates in which the genesis of new neurons is more widespread in different areas of the nervous system. In the mammalian cerebellum, neurogenesis is thought to be limited to the early postnatal period, coinciding with end of the granule cell genesis and disappearance of the external granule cell layer (EGL). We recently showed that in the rabbit cerebellum the EGL is replaced by a proliferative layer called 'subpial layer' (SPL) which persists beyond puberty on the cerebellar surface. Here we investigated what happens in the cerebellar cortex of peripuberal rabbits by using endogenous and exogenously-administered cell proliferation antigens in association with a cohort of typical markers for neurogenesis. We show that cortical cell progenitors extensively continue to be generated herein. Surprisingly, this neurogenic process continues to a lesser extent in the adult, even in the absence of a proliferative SPL. We describe two populations of newly generated cells, involving neuronal cells and multipolar, glia-like cells. The genesis of neuronal precursors is restricted to the molecular layer, giving rise to cells immunoreactive for GABA, and for the transcription factor Pax2, a marker for GABAergic cerebellar interneuronal precursors of neuroepithelial origin that ascend through the white matter during early postnatal development. The multipolar cells are Map5+, contain Olig2 and Sox2 transcription factors, and are detectable in all cerebellar layers. Some dividing Sox2+ cells are Bergmann glia cells. All the cortical newly generated cells are independent from the SPL and from granule cell genesis, the latter ending before puberty. This study reveals that adult cerebellar neurogenesis can exist in some mammals. Since rabbits have a longer lifespan than rodents, the protracted neurogenesis within its cerebellar parenchyma could be a suitable model for studying adult nervous tissue permissiveness in mammals.

Figure 1. Newly generated cells in the cerebellar cortex of peripuberal rabbits.

In addition to the chain of neuroblasts of the subpial layer (SPL; [9]), the peripuberal rabbit cerebellar cortex contains PSA-NCAM+ cells in all layers (A). In parallel with the progressive dilution and disappearance of the SPL, numerous PSA-NCAM+ cells remain detectable in the molecular layer (B–D). PSA-NCAM (E–G) and Map5 (I,J) identify two morphologically-distinct cell populations, both immunoreactive for CRMP-4 (H,J,K). PSA-NCAM+ cells are bipolar- and neuronal-shaped (E–H), whereas Map5+ cells are multipolar (I–K). All these cell types are newly generated, as revealed by BrdU staining (L–Q) at different post-injection survival times (5 [L], 10 [M,N], 15 [O–Q]). Quantitative analysis of the two cell populations in different cerebellar layers (R,S). For PSA-NCAM+ cells in the molecular (ML) and granule layer (GL) different morphologies are also indicated. Bars: A–D, 50 µm; E–Q, 10 µm.

Figure 2. Newly generated, neuronal-shaped cells in the cerebellar cortex of peripuberal rabbits.

Bipolar, intermediate, and neuronal-like shaped cells are immunoreactive for the neuronal marker doublecortin (DCX; A–D) and double stained with BrdU (E; 15 days survival post-injection), PSA-NCAM (F), Pax2 (G,H), and GABA (I,J). See also Table in (L); (1), Overlapping populations; red lines, antigens revealing subpopulations of (1); (2), Antigens whose immunoreactivity is present/absent in a subpopulation of bipolar cells (their counterpart being represented in Figure 4L). (K) Model for the newlyborn neuronal cell population: coexpression of PSA-NCAM, DCX, and BrdU (green) in different morphologies reveals a neurogenic sequence within two weeks, and coexpression with the transcription factor Pax2 (red) indicates that these cells are interneuronal progenitors of neuroepithelial origin ascended from the white matter (red arrow). Bars: 10 µm.

Figure 3. Newly generated, multipolar cells in the cerebellar cortex of peripuberal rabbits.

Multipolar cells of different shape are immunoreactive for Map5 in different cerebellar layers (A,D–H,J,K, ML; B,C,I, GL) and CRMP-4 (E), but not DCX (D). These cells are newly generated, as they are marked with BrdU at 15 days survival post-injection (F,G). They also express the transcription factors Olig2 (H,I) and Sox2 (J,K). See also Table in (M). (L) Model for the newlyborn Map5+ cell population: the cells are homogeneously distributed in all cerebellar layers. Bars: 10 µm.

Figure 4. Residual granule cell precursors in the cerebellar cortex of peripuberal rabbits.

Neuroblasts forming chains in the SPL are immunoreactive for PSA-NCAM (A), DCX (A,B,E), and Map5 (E), partially for Tag1 (B,C) and NeuN (F), but not for CRMP-4 (D). NeuN (F,G) and Pax6 (H–K) reveal residual granule cell precursors both in the SPL and in the ML, but they are not present in all DCX+ cells (J). See also Table in (L); (1), Overlapping populations; red lines, antigens revealing subpopulations of (1). (2), Antigens whose immunoreactivity is present/absent in a subpopulation of bipolar cells (their counterpart being represented in Figure 2L). (M), Cell death revealed with Tunel technique is reduced in the shift from the third to the fourth month of age (3 m, 4 m), suggesting that granule cell genesis is coming to an end. Student's T-test revealed a significant decrease of the number of tunel labelled cells in each layer (SPL p = 0,010; ML p = 0,005; GL p = 0,004 WM p = 0,013).This is confirmed by the fact that Pax6 and NeuN are no more detectable after the fourth postnatal month (N). Bars: 10 µm.

Figure 5. Newly generated cells 1–2 months after their genesis in the cerebellar cortex of peripuberal rabbits.

Countings of BrdU+ cells in the cerebellar cortex at 5, 15, and 65 days after injection (A) reveal a marked drop during the first two weeks followed by a relative stabilization during the subsequent months (F = 19,01, p = 0,001). Post-hoc comparison revealed a decrease in the number of BrdU+ cells between 5 and 15 days. (B) The slight decrease observed at long term survival mainly affects cells in the molecular layer (ML). Student's T-test revealed that the decrease of the number of cells between 15 and 65 days of survival in the ML shows a trend (p = 0,09) while in the GL it is not significant (p = 0,45). C–H, BrdU-stained cells visualized 1 month (C–G) and 2 months (H) after the last of 5 injections in association with different neuronal (D–G) and structural plasticity (A,H) markers. (GL) granule cell layer. I–K, Ultrastructural analysis of BrdU+ cells 2 months after systemic injection of the DNA marker. After identification of the marked nuclei in semithin sections of the molecular layer (I,J,K, color insets), the same cells localized at the electron microscope in ultrathin sections (I',J',K') reveal features of a synantocyte (I'), showing a round cell soma and three initial, multipolar processes (asterisks), and two polarized neurons with an oval cell soma (J',K'). K'', inset showing an higher magnification of an axo-somatic synapse (arrow). Bars: A–F, 10 µm; I,I',J,J',K,K', 3 µm; (the cells in ultrastructural photographs seem smaller than in confocal images since the sections are not cut through their main diameter).

Figure 6. Newly generated cells in the cerebellar cortex of the adult rabbit.

All morphological cell types, expressing and coexpressing the same markers as in the peripuberal rabbit cerebellum are detectable in the adult (1 and 3 years; A–G,Q,R). Cell proliferation is detectable in both cortical layers, as showed by Ki67 staining (H–J), and newly generated neuronal (K,L) and multipolar, glial-like cells (S) are visible after BrdU staining at 15 days survival. Also in the adult, the newlyborn neuronal cells do contain GABA (M,N) and Pax2 (O,P), as well as the multipolar cells are immunoreactive for Sox2 (T). See also Table in (X). U, Absolute and relative amounts of PSA-NCAM+ and Map5+ cells in the adult cerebellum. V, Model for the newlyborn PSA-NCAM+ and Map5+ cell populations in the adult. Bars: 10 µm.

Figure 7. Relationships among transcription factors in newly generated neuronal and glial-like cell populations.

Both Olig2 and Sox2 are never coexpressed with Pax2 (A,B), whereas some Sox2+ nuclei also contain Olig2 (C, arrow). Virtually all BLBP+ Bergmann glia cells contain Sox2 (D) but not Olig2 (E). Some large Sox2+ nuclei in the Purkinje cell layer do incorporate BrdU at 1–5 days postinjection survival (F). Incorporation of BrdU is also detectable in some Bergmann glia cells (G; two immuno-negative Bergmann glial cells are indicated by arrows), that retain the marker for at least two months (H). P, Purkinje neuron (immuno-negative). Also Olig2+ cells can incorporate BrdU at 1–5 days postinjection survival (I). (J,K), Proliferating Sox2+ nuclei and BLBP+ Bergmann glia cells in the Purkinje cell layer of the adult rabbit. (L), Schematic summary: Pax2+ neuronal precursors and multipolar glial-like cell precursors generated in the peripuberal and adult rabbit cerebellar cortex belong to two distinct populations. Sox2+, dividing Bergmann glia could generate Sox2+/Olig2+ precursors that then express Map5 in multipolar cells. Bars: 10 µm.

Figure 8. Model for protracted cell genesis in the cerebellum of the peripuberal and adult rabbit.

A, Summary of antigens distribution and morpholgies associated with different populations of newlyborn cells in the rabbit cerebellar cortex. B, Protracted neurogenesis in the postnatal/peripuberal cerebellum and persistent neurogenesis in the adult. All processes represented at the right of the vertical line are typical of rabbits.