Activation of neural stem cells from quiescence drives reactive hippocampal neurogenesis after alcohol dependence

Abstract

Neural stem cell-driven adult neurogenesis contributes to the integrity of the hippocampus. Excessive alcohol consumption in alcoholism results in hippocampal degeneration that may recover with abstinence. Reactive, increased adult neurogenesis during abstinence following alcohol dependence may contribute to recovery, but the mechanism driving reactive neurogenesis is not known. Therefore, adult, male rats were exposed to alcohol for four days and various markers were used to examine cell cycle dynamics, the percentage and number of neural progenitor cell subtypes, and the percentage of quiescent versus activated progenitors. Using a screen for cell cycle perturbation, we showed that the cell cycle is not likely altered at 7 days in abstinence. As the vast majority of Bromodeoxyuridine-positive (+) cells were co-labeled with progenitor cell marker, Sox2, we then developed a quadruple fluorescent labeling scheme to examine Type-1, -2a, -2b and -3 progenitor cells simultaneously. Prior alcohol dependence indiscriminately increased all subtypes at 7 days, the peak of the reactive proliferation. An evaluation of the time course of reactive cell proliferation revealed that cells begin proliferating at 5 days post alcohol, where only actively dividing Type 2 progenitors were increased by alcohol. Furthermore, prior alcohol increased the percentage of actively dividing Sox2+ progenitors, which supported that reactive neurogenesis is likely due to the activation of progenitors out of quiescence. These observations were associated with granule cell number returning to normal at 28 days. Therefore, activating stem and progenitor cells out of quiescence may be the mechanism underlying hippocampal recovery in abstinence following alcohol dependence.

Keywords: Adult neurogenesis; Alcoholism; Ethanol; Hippocampus; Neurodegeneration; Progenitor cell.

Figure 1

NeuroD1+ immunoreactivity is increased following 14, but not 7, days of abstinence in ethanol-treated rats. Representative photomicrographs of control (A,C) and ethanol-exposed animals (B,D) at 7 days (A,B) and 14 days (C,D) post ethanol exposure. The number of NeuroD1+ cells in the subgranular zone (SGZ) of the hippocampal dentate gyrus was determined using unbiased stereology (E) for ethanol (n=8–9) and controls (n=6–8) and shown as total estimated NeuroD1+ cells ± SEM. GCL, granule cell layer; scale bar = 100 µm; inset scale bar = 20 µm; *p<0.001.

Figure 2

Alcohol-induced reactive proliferation begins five days post-binge, which is two days earlier than originally reported (Nixon and Crews, 2004). The number of BrdU+ profiles (A) were counted only in the SGZ in order to avoid proliferating microglia (Nixon et al., 2008) and are presented as mean BrdU+ cells/section ± SEM. Two to four controls were counted at each time point, and upon one-way ANOVA to confirm a lack of difference between time points, were collapsed into a common control group. Representative images are shown for BrdU immunoreactivity in control (B) and ethanol-exposed rats at 4 (C), 5 (D), and 7 (E) days after alcohol exposure. Scale bar = 100 µm; 20 µm inset; *p<0.05 versus control.

Figure 3

Seven days of abstinence following binge ethanol exposure increased the number of actively proliferating cells (Ki67+) without altering cell cycle dynamics. (A) A diagram indicating each stage of the cell cycle and the antibody used for identification: BrdU for cells recently in S-phase (within 2 h), pHisH₃ for cells in the G₂ and M phases, and Ki67 for labelling cells in all phases of active division (not G₀). The G₁ population was estimated by subtracting the sum of BrdU+ and pHisH₃+ cells from number of Ki67+ cells number. Representative photomicrographs from control (B-D, n=7) and ethanol exposed (E-G, n=5) rats are shown for Ki67 (B, E), BrdU (C, F) and pHisH₃ (D, G). (H) Profile counts were used to quantify the number of cells/section due to nonhomogeneous distribution of these markers (Crews et al., 2004; Noori and Fornal, 2011). Abstinence following ethanol exposure yielded a significant increase in the number of proliferating cells in all phases of the cell cycle (H), but did not alter the proportion of cells in any phase of the cell cycle (I). Data are mean ± SEM. BrdU, bromo-deoxy-uridine; pHisH₃, serine 10 phosphorylated Histone H₃; GCL, granule cell layer; scale bar = 100 µm; inset scale bar = 20 µm; *p<0.05.

Figure 4

Reactive hippocampal neurogenesis following ethanol exposure and abstinence is due to an increase in the number of proliferating neural progenitor cells. (A,B) Representative photomicrographs of Sox2+ cells in the hippocampus of control (A, n=6) and ethanol-exposed (B, n=8) rats after 7 days of abstinence, quantification of which is shown in C. (D, E) Representative confocal images are shown for Sox2 (D), BrdU (E), and merged (F). Every BrdU+ cell was counted (G, left) and evaluated for Sox2 colabeling (G, right). The original finding of an increase in BrdU was replicated and extended to show that the majority of proliferating cells are Sox2+ NPCs. Data are mean ± SEM. BrdU, Bromo-deoxy-Uridine; GCL, granule cell layer; scale bar = 100 µm (A); scale bar = 20 µm (A inset and D). *p<0.05

Figure 5

Four-day binge ethanol exposure results in recruitment of hippocampal NPCs out of quiescence and into active proliferation at T7 of abstinence. Representative confocal images of Sox2+ (A), Ki67+ (B), and dual-labeled (C) cells were obtained from rats in abstinence at 7 days after binge ethanol exposure (n=8) or control diet (n=6). All Sox2+ cells were evaluated for the presence or absence of Ki67 and presented as the mean percentage of NPCs in active portion of the cell cycle ± SEM. (D) Progenitor cell activation is evidenced by a shift in the proportion of Sox2+/Ki67− (quiescent) to Sox2+/Ki67+ (proliferating) cells in rats with prior ethanol exposure at this time point. Scale bar = 20 µm. *p<0.01.

Figure 6

Adult neurogenesis is a multi-step process that begins with Type 1 stem-like radial glial cells that proliferate and divide asymmetrically to generate a progenitor cell (Type 2a) which can then proliferate and/or differentiate into a Type 2b NPC (circular arrows indicate that the cell is able to proliferate (Kempermann et al., 2004). Type 2b NPCs can proliferate and/or differentiate into a Type 3 NPC, which can then proliferate and differentiate into an immature neuron. Fluorescent quadruple immunohistochemistry was designed to identify each of the various progenitor cell subtypes as described in A. For example, NeuroD1 is expressed in Type 2b and Type 3 NPCs and is considered to be a Type 2b cell when it co-expresses with Sox2, but a Type 3 cell when Sox2 is no longer expressed. (B) Representative confocal images are shown: type 1 cell is positive for Ki67, GFAP, and Sox2 but is negative for NeuroD1 (row 1). Type 2a proliferating cells are positive for Ki67 and Sox2 but negative for GFAP and NeuroD1 (row 2). A combination of Ki67+, Sox2+, and NeuroD1+ yet GFAP- signifies a type 2b cell (row 3). Finally, a cell that is only Ki67+ and NeuroD1+ is a type 3 cell (row 4). ML, molecular layer; GCL, granule cell layer; SGZ, subgranular zone; scale bar = 20µm.

Figure 7

Characterization of the proliferating neural progenitor cell pool in the hippocampus of adult rats 7 days after 4 days of binge alcohol exposure. Each Ki67+ cell was investigated for the presence or absence of the other markers. A) NPC subtypes are shown as a percent of total Ki67+ cells counted ± SEM. (B) In order to approximate the number of cells in each progenitor phenotype, the percentage was multiplied by the total number of Ki67+ cells and these data are presented as mean cells ± SEM. Control n=6, ethanol n=8. *p<0.05.

Figure 8

The proliferating pool of NPCs in the hippocampus of adult rats was characterized at 5 days after 4 days of binge alcohol exposure, the point at which reactive proliferation appears to emerge. Each Ki67+ cell was investigated for the presence or absence of the other markers. (A) Progenitor subtypes are shown as a percent of total Ki67+ cells counted ± SEM. (B) In order to approximate the number of cells in each progenitor phenotype, the percentage in A was multiplied by the total number of Ki67+ cells and these data are presented as mean cells ± SEM. Control n=6, ethanol n=8. *p<0.05.