Figure 1.
Lgr5 is expressed in the cerebellum during postnatal development.
Sagittal brain immunofluorescence sections from Lgr5EGFP-CreERT2 mice stained for EGFP to mark Lgr5+ cells at multiple time points. a) Tiled 10x magnified images of a midline sagittal section were stitched together to reveal Lgr5-EGFP expression at the P14 time point. b) Lgr5-EGFP expression at P4, P7, P10, P14, P21, P28 and P56 in the cerebellum demonstrating that Lgr5 expression turns on at P4, ramps up expression until its peak from P10-P14 and then expression is lost permanently over the next 7–14 days. c) Microarray expression data of Lgr5 generated from the Cerebellar Development Transcriptome Database [20] taken from total cerebellum RNA at indicated time points. Data acquired from http://www.cdtdb.neuroinf.jp/CDT/ReferTemporal.do?cdIdCh=CD12762.1. d) Olfactory Bulb (OB) at P7 and P56 reveal Lgr5-EGFP expression throughout development and adulthood. Scale bars, 400 microns.
Figure 2.
Postnatal Lgr5+ cells are exclusively cerebellar granule neurons.
a) Sections from Lgr5EGFP-CreERT2 P14 mouse brains were stained for EGFP to mark Lgr5+ cells, Gabra6 to identify cerebellar granule neuron, NeuN to identify neurons and DAPI to mark cell nuclei. Lgr5-EGFP+ cells are restricted to the internal granule layer (IGL) and all Lgr5+ cells are also positive for GABRA6 and NeuN. Lgr5-EGFP+ processes, presumably from Lgr5+ CGNs in the IGL can be seen in the molecular layer (ML). Top left panel, scale bar, 100 microns. Remaining 5 panels taken from marked area of top left panel. Scale bar, 25 microns. b,c,d) Representative images of Lgr5EGFP-CreERT2 P10 and P14 mouse brains stained for Lgr5-EGFP and b) BLBP to identify Bergmann glia, c) Calb1 to identify Purkinje neurons and d) Olig2 to stain cells of the oligodendrocyte lineage. Top panels, scale bars, 100 microns. Bottom two panels taken from marked area of left panel. Scale bar, 50 microns. e) Cells that stained positive for one of the analyzed markers were counted and examined for Lgr5 expression. Numbers presented are aggregated from all sections analyzed (≥4 per marker). There was no overlap of Lgr5 with Calb1, BLBP or Olig2 in any sections analyzed. EGL –external granule layer; ML – molecular layer; PL – Purkinje layer; IGL – internal granule layer; OC – occipital cortex; WM – white matter tract.
Figure 3.
Lgr5-positive cells in the postnatal cerebellum are lineage-restricted.
a) Schematic for lineage tracing design. Cells from Lgr5EGFP-CreERT2; R26RtdTomato mice express EGFP and CreER2 when the Lgr5 locus is active. Addition of tamoxifen (TAM) activates CreER2, which removes the STOP element in the Rosa26-tdTomato locus permanently marking the cell and its progeny with tdTomato. Cells that maintain Lgr5 expression are double positive (yellow). b) Examples of lineage tracing time points. Cells that were once Lgr5+ (tdTomato+) were invariably Gabra6 positive, indicating granule neuron lineage restriction. Top: Lineage tracing was initiated at the beginning of Lgr5 expression in the IGL (P4) and cerebellums were analyzed at P10. Bottom: Lineage tracing was initiated near the end of Lgr5 expression in the IGL (P14) and analyzed at P28, when the Lgr5 is no longer active in the IGL. Scale bar, 25 microns. c) Lineage tracing initiated at P14 was analyzed at extended time points, P21 and P56, demonstrating Lgr5+ cells at P14 survive and integrate into the adult IGL architecture. Scale bars, top: 400 microns, bottom: 100 microns.
Figure 4.
Cells previously Lgr5+ remain lineage restricted.
In addition to tdTomato+ cells staining positive for Gabra6 (Fig. 3b), neither Lgr5+ cells nor cells previously Lgr5+ are positive for other lineage markers. Sections from Lgr5EGFP-CreERT2; R26RtdTomato mice that were lineage traced from P4 to P10 or P7 to P14 were stained for a) BLBP to identify Bergmann glia, b) Calb1 to identify Purkinje neurons and c) Olig2 to identify cell of the oligodendrocyte lineage. Representative images shown. Sections were also stained for Lgr5-EGFP and DAPI, while tdTomato marks lineage traced cells. d) Cells staining positive for a lineage marker or tdTomato were counted. Numbers shown are aggregated from all sections analyzed (≥4 per marker). There was no overlap of tdTomato and any lineage marker analyzed. Yellow arrows indicate examples of cells that are Lgr5+/tdTomato+ but lineage marker negative, while red arrows indicate examples of cells Lgr5-/tdTomato+ that are lineage marker negative. Scale bars, top 3 panels: 100 microns, bottom 9 panels: 25 microns. ML – molecular layer; PL – Purkinje layer; IGL – internal granule layer; WM – white matter tract.
Figure 5.
Lgr5 CGNs are non-proliferative and post-mitotic.
a, b) Cerebellar sections from Lgr5EGFP-CreERT2 mice were stained for the proliferation marker Ki67 and Lgr5-EGFP at both a) early (P7) and b) late (P14) stages of Lgr5 expression. Lgr5+ cells were uniformly negative for Ki67. Scale bars 50 microns. c) Lineage tracing was initiated at P4 in Lgr5EGFP-CreERT2; R26RtdTomato mice and cerebella analyzed for Ki67 at P10. Many other lineage trace time points were also analyzed (data not shown). Cells post-Lgr5 expression (tdTomato+) were uniformly negative for Ki67 in every time point analyzed. Note that immature granule neurons in the EGL in A-C were Ki67+, as expected. Scale bars for C top: 400 microns, bottom: 100 microns. d) Lgr5EGFP-CreERT2 mice were crossed with Catnbflox(exon3) mice to yield Lgr5EGFP-CreERT2; Catnbflox(exon3) mice. Upon tamoxifen administration, Cre recombinase initiates expression of a constitutively active form of β-catenin in cells with an active Lgr5 locus. e) Tamoxifen was administered at P4 and cerebella were analyzed at P25, when these mice die of other complications related to transgene expression. β-catenin over-activity led to prolonged Lgr5 expression in the IGL in Lgr5EGFP-CreERT2; Catnbflox(exon3 mice (center panel), but not wild type (right panel). The normal architecture of the cerebellum was maintained in both conditions. Scale bars 400 microns. f, g) Lgr5+ CGNs from P25 Lgr5EGFP-CreERT2; Catnbflox(exon3) mice with tamoxifen administration at P4 were f) positive for Gabra6 and g) negative for Ki67. Scale bars, bottom left: 200 microns, other 5 panels: 25 microns.
Figure 6.
Maturing CGNs progress through a transient Lgr5 phase resulting in mature CGNs.
a) Schematic of lineage tracing time course. Tamoxifen (TAM) was injected at P4 into several large litters and brains from pairs of wildtype (WT) and Lgr5EGFP-CreERT2; R26RtdTomato mice from each litter, were analyzed at the indicated time points. Period of predicted Cre activity is indicated. b) Multiple sections from each brain analyzed were stained for Lgr5-EGFP to mark cells currently Lgr5+, while tdTomato expression marked previously Lgr5+ cells that had undergone recombination. Shown is an example from a P10 brain that had lineage tracing initiated at P4. Yellow dots indicate cells marked as Lgr5+/tdTomato+ by image analysis algorithm (yellow arrows indicate examples), while blue dots indicate cells marked as Lgr5-/tdTomato+ (blue arrow with red outline point out examples). Scale bars, 50 microns. c) Quantification from an image analysis algorithm of all sections analyzed.
Figure 7.
The source of the Lgr5 ligand, R-Spondin1, temporally varies with CGN maturation.
Cerebellar sections from Lgr5EGFP-CreERT2 mice at increasing developmental time points a) P4, b) P8, c) P14 and d) P28 were stained for Lgr5-EGFP, Rspo1, Calb1 and DAPI. Top row panels are representative images while boxes indicate magnified regions in bottom 4 rows. Scale bars, top row: 100 microns, bottom 4 rows: 50 microns.
Figure 8.
Allen Developing Mouse Brain Atlas ISH data supports Rspo1 data.
RNA in situ hybridization data from the Allen Developing Brain Atlas was queried for Rspo1 (a, d, g, j)(http://developingmouse.brain-map.org/gene/show/82848), Shh (b, e, h, k) (http://developingmouse.brain-map.org/gene/show/20186) and Calb1 (c, f, i, l) (http://developingmouse.brain-map.org/gene/show/12092) at P4, P14, P28 and P56. Sagitttal sections from indicated time points were used. Representative images of each gene at each time point are presented. Calb1 serves as a consistent positive control marker for Purkinje neurons in the Purkinje layer, while Rspo1 and Shh RNA transcription varies over time within this layer. Feulgen-HP yellow DNA stain was used as a counter stain for all time points except P56 to provide tissue context to the ISH signal (violet). Scale bars, 200 microns.