Nuclear receptor TLX stimulates hippocampal neurogenesis and enhances learning and memory in a transgenic mouse model

Transgenic TLX Expression Leads to Enlarged Brains and Hippocampal DG.

In this study, we took a gain-of-function approach to investigate the role of TLX in neural stem cell self-renewal, neurogenesis, and the behavioral output of the brain. We generated a TLX Tg mouse model that expresses a GFP-TLX fusion protein under the control of the nestin promoter (Fig. 1A), and established two founder lines that exhibit indistinguable phenotypes. We found that the GFP-TLX transgene was expressed in brains of the Tg mice, but not those of their WT littermates (Fig. 1B). We also detected the GFP-TLX protein using a GFP antibody along the SGZ of the hippocampal DG and the SVZ, the two active adult neurogenic areas where endogenous nestin is expressed (Fig. 1 C and D). The Tg mice had enlarged forebrains with greater brain weight and size compared with their WT littermates (Fig. 1 E–G), although the two groups were of similar total body weight and length (Fig. S1). Histologically, the Tg mice exhibited an enlarged DG and slightly smaller lateral ventricles (Fig. 1 H–J).

TLX Transgene Expression Stimulates Neural Stem Cell Self-Renewal.

The enlarged brains in the Tg mice prompted us to ask whether increased TLX expression led to enhanced neural stem cell self-renewal. To address this question, we isolated neural stem cells from the brains of 6-wk-old WT and Tg mice and cultured them in DMEM F12 medium supplemented with N2, epithelial growth factor (EGF), and fibroblast growth factor 2 (FGF-2). The self-renewal capacity of the WT and Tg neural stem cells was determined by clonal analysis. Increased TLX expression led to enhanced self-renewal of neural stem cells, as demonstrated by increases in both clonal size (sphere size and sphere cell numbers) and clonal rate (sphere formation rate) of primary (1st) and secondary (2nd) neurospheres (Fig. 2 A–H). Both WT and Tg neural stem cells were able to differentiate into Tuj1⁺ neurons, GFAP⁺ astrocytes, and O4⁺ oligodendrocytes (Fig. S2), suggesting that both are multipotent. These results further support the idea that TLX plays an important role in neural stem cell self-renewal.

In previous work, we showed that TLX represses the expression of miRNAs miR-9 and miR-137 in neural stem cells, and that both miRNAs negatively regulate neural stem cell proliferation (11, 12). To determine whether these TLX downstream miRNAs play a role in regulating cell proliferation in Tg neural stem cells, we first examined the expression levels of miR-9 and miR-137 in WT and Tg neural stem cells. We found dramatically lower expression of miR-9 and miR-137 in Tg neural stem cells compared with WT cells (Fig. 2 I and J). Consistent with enhanced self-renewal in Tg neural stem cells (Fig. 2 A–H), we observed elevated cell proliferation in Tg neural stem cells compared with WT cells (Fig. 2K). Moreover, overexpression of miR-9 and miR-137 together reversed the elevated cell proliferation in Tg neural stem cells (Fig. 2K), suggesting that repression of miR-9 and miR-137 expression by TLX contributes to the phenotype of enhanced cell proliferation observed in Tg neural stem cells.

Adult Tg Mice Exhibit Increased Hippocampal Neurogenesis.

To determine whether transgenic TLX expression regulates neural progenitor cell proliferation in adult brains, we performed BrdU labeling of adult mice for 5 d. Because the more prominent anatomic changes in Tg mice were seen in the hippocampal DG (Fig. 1 H–J), we focused on this brain region. Increased BrdU labeling was observed along the SGZ of the DG in Tg brains compared with WT brains (Fig. 3 A and B). In addition, increased staining of Ki67, a marker of proliferating progenitors, was observed in the DG of the Tg mice (Fig. 3 A and C). These results indicate enhanced neural progenitor cell proliferation in the Tg hippocampus.

We next examined whether TLX expression potentiates hippocampal neurogenesis in adult Tg brains. For this, 6-wk-old WT and Tg mice were treated with BrdU for 5 d, followed by a 3-wk survival period. Neurogenesis was assessed by double-labeling with BrdU and a neuronal marker, NeuN, in the SGZ of the DG. Quantification of the BrdU⁺NeuN⁺ cells revealed substantially increased hippocampal neurogenesis in the Tg mice compared with their WT littermates (Fig. 3 D and E). Together, these results indicate that the neural precursor-specific expression of the TLX transgene stimulates neural progenitor proliferation and neurogenesis.

To determine whether TLX directly regulates adult hippocampal neurogenesis, we overexpressed TLX in the DG of adult mouse brains by intracranial lentiviral delivery. A TLX-expressing lentivirus was injected into the DG of adult WT mouse brains by stereotaxic transduction. Viral expression of TLX was monitored by a coexpressed GFP reporter, and cell proliferation was monitored by BrdU labeling. Specific expression of TLX at the adult hippocampal DG led to a substantial increase in cell proliferation, as demonstrated by the increased percentage of BrdU⁺GFP⁺ cells in total GFP⁺ cells (Fig. 4 A and B). When we treated the viral-transduced mice with BrdU for 5 d, followed by a 3-wk survival period, adult DG-specific expression of the TLX transgene also was able to enhance neurogenesis, as demonstrated by the increased percentage of BrdU⁺NeuN⁺GFP⁺ cells (Fig. 4 C and D).

Transgenic TLX Expression Rescues Defects in Hippocampal Neurogenesis in TLX-Null Brains.

To test whether the neural precursor-specific expression of the TLX transgene can rescue the neurogenic defects induced by loss of TLX expression, we bred the Tg mice with the TLX^−/− mice (6) to obtain compound mice with the TLX transgene in a TLX^−/− background (TgTLX^−/−) (Fig. S3A). The TLX transgene was specifically expressed in the TgTLX^−/− mice, but not in the TLX^−/− mice, and the expression level of the transgene was similar to that of endogenous TLX in WT mice (Fig. S3B). The nestin promoter-driven expression of TLX was able to substantially rescue the deficiencies in brain weight and size observed in TLX^−/− mice (Fig. 5 A–C). Histologically, the TgTLX^−/− mice exhibited considerably enlarged hippocampal DG and slightly smaller lateral ventricles compared with their TLX^−/− littermates (Fig. 5D).

We next isolated neural stem cells from the forebrains of TgTLX^−/− and TLX^−/− littermate mice. The TLX^−/− cells did not form neurospheres in DMEM F12 medium supplemented with N2, EGF, and FGF-2, but instead underwent spontaneous differentiation (Fig. 5E). In contrast, cells from the TgTLX^−/− mouse brains cultured in the same medium were able to form neurospheres, similar to WT cells (Fig. S3C).

To determine whether the nestin promoter-driven expression of TLX could rescue neural progenitor cell proliferation in the hippocampal DG of TLX^−/− mice, we treated adult TgTLX^−/− and TLX^−/− littermate mice with BrdU for 5 d. We found a substantial increase in BrdU⁺ cells was observed in the DG of TgTLX^−/− mice compared with their TLX^−/− littermates (Fig. 5 E and F). We also noted increased Ki67 staining in the DG of TgTLX^−/− mice (Fig. 5 E and G). These results indicate that the neural precursor-specific expression of the TLX transgene plays an important role in establishing the proliferative neural progenitor cell populations in adult brains. Expression of the TLX transgene also was able to restore hippocampal neurogenesis in adult TLX^−/− brains to a certain extent, as demonstrated by increased numbers of BrdU⁺NeuN⁺ neurons in the DG of adult TgTLX^−/− mice compared with TLX^−/− mice (Fig. 5 H and I).

Learning and Memory Are Enhanced in TLX Transgenic Mice.

In adult brains, the hippocampus is crucial for spatial learning and memory (21). The Morris water maze has been used extensively to investigate spatial learning and memory. To compare the performance of the WT and Tg mice in the Morris water maze test, we first subjected the mice to the visible platform test. The WT and Tg mice exhibited similar swimming velocity, distance, and latency to the platform in this test (Fig. S4), suggesting comparable vision and motor stamina in the two groups of mice. We next subjected the mice to the hidden platform test, which determines spatial memory acquisition and retention. Specifically, WT and Tg mice were subjected to 10 trials daily for 3 d to evaluate spatial memory acquisition. Probe trials were performed after the end of the last training session (on day 3) to assess short-term memory retention and again at 1 wk after training (on day 10) to evaluate long-term memory retention (22) (Fig. 6A).

The Tg mice exhibited enhanced spatial memory acquisition, as indicated by their shorter latency to reach the platform throughout the trials (Fig. 6B) and shorter latency to reach the platform in each of the 3 d of training (Fig. 6C) compared with their WT littermates. Furthermore, the Tg mice demonstrated superior short-term memory retention by spending more time in the target quadrant, which had previously held the hidden platform (Fig. 6D), and entering the target quadrant at higher frequency (Fig. 6E). At 1 wk after training, the Tg mice also showed evidence of improved long-term memory retention by spending more time in the target quadrant and entering the target quadrant at higher frequency (Fig. 6 F and G). Together, these results suggest that expression of the TLX transgene in neural precursors made a significant contribution to spatial learning and memory.

The enhanced cognitive ability of Tg mice was further demonstrated using startle habituation and prepulse inhibition analyses (23), which provide a tool for studying fundamental properties of nervous system function, including cognition. Startle habituation represents the simplest form of learning. In a 36-trial session, startle habituation was assessed as the percent decrease in startle reactivity between the first block (trials 2–6) and the last block of trials (trials 32–36) (Fig. S5A). The Tg mice exhibited significantly greater startle habituation compared with their WT littermates (Fig. S5B).

Per-pulse inhibition (PPI) provides an operational measure of sensorimotor gating (24), and reduced PPI is a manifestation of cognitive deficits (25, 26). Tg mice subjected to three prepulses at 3, 6, and 12 dB above the 65 dB background noise, followed by a pulse stimulus at 120 dB, exhibited considerably enhanced PPI at all three prepulses compared with their WT littermates (Fig. S5C). These results, combined with the water maze data, strongly suggest that the Tg mice exhibit enhanced cognitive ability.

Viral Production and Intracranial Lentivirual Infection.

The CSC control vector and the CSC-TLX–expressing lentiviruses were generated as described previously (6). For intracranial lentiviral infection, 1 µL of 1 × 10⁸ transduction unit/mL concentrated virus was injected into the hippocampal DG of 6-wk-old WT ICR mice by stereotaxic injection as described previously (15). At 2 wk after viral injection, the transduced mice were injected with BrdU for 5 d, followed by short-term BrdU labeling, or 3 wk of survival for long-term BrdU labeling. The coordinates for the DG of WT mice were anterior-posterior, −2.0 mm, mediolateral, ±1.7 mm; dorsoventral, −1.5; and −1.9 mm from the skull surface.

Morris Water Maze Test.

The water maze test was conducted in a 1.5-m-diameter circular tank filled with opaque water with visible cues outside. The test consisted of 1 d of pretraining, 2 d of visible platform training, and 3 d of hidden platform training. For visible platform, the platform location was indicated by a colorful rod attached to the platform above the water. Mice were trained with 10 trials per day for two days. For the hidden platform test, the platform was placed in a constant position in quadrant 1, hidden 1 cm below the water surface. Mice were trained with 10 trials per day for three days. In each training trial, mouse was released, facing the wall of the tank, from quadrants 2–4. Each mouse was allowed to locate the platform within 2 min. If a mouse failed to find the platform within 2 min, it was guided onto the platform. The mouse was allowed to stay on the platform for 15 s after finding it.

After 10 trials on the last day of training, the platform was removed, and a probe trial was conducted as a test of short-term memory. For testing long-term memory, the probe trial was repeated at 1 wk after the final training trial. For both probe trials, each mouse was allowed to swim for 90 s. The procedure was recorded using a digital video camera system (San Diego Instruments). Additional experimental procedures are provided in SI Materials and Methods.