Evidence of Progenitor Cell Lineage Rerouting in the Adult Mouse Hippocampus After Status Epilepticus

Abstract

Cell lineage in the adult hippocampus comprises multipotent and neuron-committed progenitors. In the present work, we fate-mapped neuronal progenitors using Dcx-CreERT2 and CAG-CAT-EGFP double-transgenic mice (cDCX/EGFP). We show that 3 days after tamoxifen-mediated recombination in cDCX/EGFP adult mice, GFP+ cells in the dentate gyrus (DG) co-expresses DCX and about 6% of these cells are proliferative neuronal progenitors. After 30 days, 20% of GFP+ generated from these progenitors differentiate into GFAP+ astrocytes. Unilateral intrahippocampal administration of the chemoconvulsants kainic acid (KA) or pilocarpine (PL) triggered epileptiform discharges and led to a significant increase in the number of GFP+ cells in both ipsi and contralateral DG. However, while PL favored the differentiation of neurons in both ipsi- and contralateral sides, KA stimulated neurogenesis only in the contralateral side. In the ipsilateral side, KA injection led to an unexpected increase of astrogliogenesis in the Dcx-lineage. We also observed a small number of GFP+/GFAP+ cells displaying radial-glia morphology ipsilaterally 3 days after KA administration, suggesting that some Dcx-progenitors could regress to a multipotent stage. The boosted neurogenesis and astrogliogenesis observed in the Dcx-lineage following chemoconvulsants administration correlated, respectively, with preservation or degeneration of the parvalbuminergic plexus in the DG. Increased inflammatory response, by contrast, was observed both in the DG showing increased neurogenesis or astrogliogenesis. Altogether, our data support the view that cell lineage progression in the adult hippocampus is not unidirectional and could be modulated by local network activity and GABA-mediated signaling.

Keywords: GABAergic interneurons; adult hippocampus; astrogliogenesis; fate-specification; kainic acid; neurogenesis; pilocarpine; status epilepticus.

FIGURE 1

Dcx-lineage comprises mostly neurons and a small fraction of astrocytes. (A) Timeline presenting timepoints (d.a.r. – days after recombination) in which animals were fixed (crosses) for immunohistochemistry and quantification of GFP+ cells after TAM administration. (B–J) Coronal sections of the hippocampus of cDCX/EGFP animals immunolabeled for GFP (green), DCX, GFAP, or CTIP2 (red) 3, 7, or 30 d.a.r. Observe that GFP expression is restricted to small DCX+ cells in the SGZ 3 d.a.r. (B–D). (E–G) Example of a GFP+/GFAP+ cell showing morphologies of astrocytes 30 d.a.r. (arrowhead). (H–J) Examples of GFP+/CTIP2+ cells showing mature granule cell morphology 30 d.a.r. (arrowheads indicate some of these cells). (K) Graphic showing the quantification of GFP cells expressing makers of different cell populations (n = 4–5 animals per timepoint). (L) Schematic representation of adult hippocampal neurogenesis showing markers in each phase of cell development used in this study. NSCs – Neural stem cells (type 1); IPs – Intermediate progenitors (types 2a, 2b, and 3). (M) Timeline presenting experimental protocol from group receiving TAM injection followed by 3 days of continuous BrdU treatment in the drinking water. Animals were perfused 3 or 30 days after recombination (d.a.r.). (N–Q) Expression of DCX in GFP+/BrdU+ cells 3 d.a.r. (R–U) 30 days after Dcx-mediated recombination, GFP+/BrdU+ are rare possibly due to cell death of newly generated cells. (V) Graphic showing the quantification of GFP+/BrdU+ cells expressing DCX or GFAP. (n = 5 animals in “3 days” group; n = 3 in “30 days” group). Scale bars: 20 μm.

FIGURE 2

Kainic acid and pilocarpine have opposite effects in Dcx-lineage progression. (A) Timeline of the experimental protocol used to fate map the Dcx-lineage recombined 7 days after kainic acid (KA), pilocarpine (PL) or vehicle (VEH) intrahippocampal injection. (B–F) Coronal sections of the hippocampus of cDCX/EGFP animals immunolabeled for GFP (green), GFAP or CTIP2 (red). Mosaic composition of VEH (B) and PL (C) ipsilateral side and KA contralateral hippocampus (D) showing GFP+ granule cells with typical morphology. (E) Confocal image of contralateral hippocampus in KA group showing the expression of CTIP2 in GFP+ granule cells in single plane z-projection images. (F) Mosaic composition of ipsilateral hippocampus of KA injected animals showing severe GCD and GFP+ cells with morphologies of reactive astrocytes. (G) Confocal image of ipsilateral DG showing the expression of GFAP in GFP+ cells in single plane z-projection images. ML, molecular layer; GCL, granular cell layer. (H,I) Quantification of the total number of GFP+ cells differentiating into granule neurons (red bars) or astrocytes (blue bars) in the contralateral and ipsilateral sides. (n VEH = 3; n KA = 5; n PL = 3; two-way ANOVA followed by Tukey’s multiple comparisons: **Padj < 0.01; ***Padj < 0.001; ****Padj < 0.0001). Graphic shows mean ± SEM of animals with same number of sections analyzed. Scale bars: 100 μm (B–D,F) and 20 μm (E,G).

FIGURE 3

Kainic acid injection enhances astrogliogenesis from previously recombined DCX-cells. (A) Timeline representing experimental protocols in which one tamoxifen injection was administered 1 day before SE induction with KA. Animals received BrdU in drinking water during 3-days and were perfused right after this period, or 30 days after recombination (d.a.r.). (B,C) Quantification of proliferative GFP+ cells after 3 days (B), and expression of DCX or GFAP among these cells (C) on ipsilateral and contralateral side of vehicle and KA injected animals. (D,E) Quantification of proliferative GFP+ cells after 3 and 30 days in ipsilateral side (D) and contralateral (E) expressing DCX or GFAP. (F–H) Example of GFP+/GFAP+ cell (RGL) in the SGZ of KA injected cDCX/EGFP animals 3 days after recombination. (I) Schematic representation of the fate switches (dotted red arrow) observed in the Dcx-lineage after KA injection.

FIGURE 4

DCX-cells retain the capacity to proliferate and generate astrocytes after 1 month. (A) Timeline of the experimental protocol used to fate map the Dcx-lineage recombined 35 days before kainic acid (KA), pilocarpine (PL) or vehicle (VEH) intrahippocampal injection. Animals were treated with BrdU in the drinking water during the whole period after KA and were perfused 2 weeks later. (B,C) Coronal sections of the hippocampus of cDCX/EGFP animals immunolabeled for GFP (green). Mosaic image of the KA contralateral DG shows the majority of GFP+ cells with typical neuronal morphologies (B). Mosaic image of the KA ipsilateral DG shows GCD and many GFP+ cells with morphologies of astrocytes (C). (D) Coronal sections of the hippocampus of cDCX/EGFP animals immunolabeled for GFP (green) and GFAP (red) showing GFP+/GFAP+ astrocytes (arrowheads) in the ipsilateral DG. (E,F) Quantification of GFP+ cells differentiating into neurons (red bars) or astrocytes (blue bars) in the contra (E) and ipsilateral (F) DG of controls and KA-treated animals. (G) Coronal sections of the hippocampus of cDCX/EGFP animals immunolabeled for GFP (green), GFAP (red), and BrdU (blue). Dashed box indicate a GFP+/GFAP+/BrdU+ astrocyte. Individual channels are shown in the right bottom corner. (H) Quantification of BrdU+ and BrdU− GFP+/GFAP+ cells the contra and ipsilateral DG of KA injected animals. ML, molecular layer; GCL, granular cell layer. Scale bars: 50 μm (B,C) and 20 μm (D,G). [n vehicle = 4; n KA = 4; n PL = 3; two-way ANOVA followed by Tukey’s multiple comparisons: **Adjusted p-value (KA vs. PL) = 0.0018; ***Adjusted p-value (KA vs. VEH) = 0.0008)].

FIGURE 5

Parvalbuminergic plexus degenerates in the ipsilateral side of KA injection. (A) Timeline of the experiment. Animals received unilateral intrahippocampal injection of KA, PL or VEH followed by continuous BrdU administration in the drinking water for 3 days before perfusion. (B) Coronal section of the hippocampus of VEH animals immunolabeled for DCX (green), PV (red), and BrdU (blue). Note the presence of PV+ interneurons in the granule cell layer (GCL). Observe process of PV+ cell in close contact with proliferating BrdU+/DCX+ cell (arrowhead) in the subgranular zone (SGZ). BrdU administration was performed for three consecutive days in the drinking water before perfusion. Scale bar: 50 μm. (C) Ratio of PV+ cells in the ipsi and contralateral sides of the hippocampus in different treatment groups. (D) Quantification of the number of PV+ interneurons per mm² of the dentate gyrus 3 days after injection of saline, KA or PL. [n vehicle = 4; n KA = 4; n PL = 4; Statistics: (B) ANOVA F₍₂,₉₎ = 4.529; p = 0.0436; Tukey’s multiple comparisons test DF = 9; Adjusted p-value (control vs. KA) = 0.0363; (C) ANOVA F₍₂,₉₎ = 8.347; p = 0.0089; Tukey’s multiple comparisons test DF = 9; Adjusted p-value (KA vs. PL) = 0.0077].

FIGURE 6

Microglia activation is stronger in the ipsilateral side of KA injection. (A) Timeline of experimental procedure to quantify Iba1+ cells. (B,C) Microglial cells in control group labeled with anti-Iba1 antibody and showing resting morphologies. (D,E) Increased signal for anti-IBA1 antibody in the KA group, indicating bilateral activation of microglia in the CA1/CA3 regions and in the ipsilateral DG hilus and molecular layer (not quantified). (F,G) Increased signal for anti-IBA1 antibody in the PL group, indicating bilateral activation of microglia in the CA1/CA3 regions (not quantified). (H–K) Higher magnification of the DG in vehicle (H,I) and KA (J,K) animals showing the morphological changes of microglial cells in the ipsilateral KA group. (L,M) Quantification of activated microglia in the ipsi or contralateral DG of vehicle-, KA- and PL-treated animals. [n control = 3; n KA = 2; n PL = 3; Statistics: (L) ANOVA F₍₂,₅₎ = 5.819; p = 0.0495; Tukey’s multiple comparisons test DF = 5; Adjusted p-value (control vs. KA) = 0.0486; (M) ANOVA F₍₂,₆₎ = 13.24; p = 0.0063; Tukey’s multiple comparisons test DF = 6; Adjusted p-value (vehicle vs. KA) = 0.0072; (PL vs. KA) = 0.0166].

FIGURE 7

Cell lineage progression model in the adult hippocampus. (A) Gray arrows indicate previously described transitions during cell lineage progression from RGC to neurons and from RGC to astrocytes in the adult hippocampus. Red arrows indicate the novel transitions described in our work. (B) Enhanced electrical activity accompanied by loss of GABAergic interneurons inhibits the progression of intermediate progenitors toward neurogenesis and stimulate proliferation, generation of RGCs and astrogliogenesis in an environment surrounded by aMGCs. (C) Enhanced electrical activity and sustained GABAergic activity stimulate proliferation of intermediate progenitors and neuronal differentiation/survival both in the presence (KA contralateral DG) or absence (PL ipsi and contralateral DG) of aMGCs.