Oleamide,a Sleep-Inducing Supplement,Upregulates Doublecortin in Hippocampal Progenitor Cells via PPAR α

Reagents and antibodies

Rabbit anti-Ki67 antibody (Cat #PA1-21520; dilution 1 : 500 for IF), Mouse anti-BrdU antibody (Cat #MS-1058-P0; dilution 1 : 500 for IF), Rabbit anti-NeuN antibody (Cat #711054; dilution 1 : 500 for IF), Rabbit anti-Nestin antibody (Cat #PA5-11887; dilution 1 : 500 for IF), anti-DCX antibody (Cat #48-1200; dilution 1 : 500 for IF), and anti-BMP4 antibody (Cat #PA5-27288; dilution 1 : 500 for IF) were purchased from ThermoFisher. Rabbit anti-MAP2 (Cat #AB5622; dilution 1 : 500 for IF) was obtained from EMD Millipore. Mouse anti-PPARα antibody (Cat #sc-398394; dilution 1 : 500 for IF) was bought from Santa Cruz Biotechnology. BrdU, chemically known as 5-Bromo-2^′-deoxyuridine (Cat #B5002-1G) was purchased from Sigma. Recombinant EGF (Cat #236-EG) and FGF (Cat #233-FB) were purchased from R&D system.

Isolation and culture of hippocampal progenitor cells

Hippocampi of E18 mouse embryos were dissected as discussed before [14, 18]. Briefly, after peeling off the meninges, a fine incision was made in the middle line around the circle of Willis followed by making the medial temporal lobe exposed. Hippocampus was isolated as a thin slice of tissue located near the cortical edge of medial temporal lobe. After that the tissues were disintegrated by mild trypsinization, neutralized with DMEM media, washed, and plated in a 75 mm flask with complete DMEM media. EGF (20 ng/mL) and FGF (20 ng/mL) were added in every alternate day for seven days until the formation of discrete neural rosettes. Next, rosettes were collected, combined, disintegrated by passing through glass pipette, and then plated in poly-D-lysine coated plate for further differentiation.

Culture of human neural progenitor cells

Human fetal brain tissues (11–17 weeks old) were obtained from the Human Embryology Laboratory (University of Washington, Seattle, WA). All of the experimental protocols were reviewed and approved by the Institutional Review Board of the Rush University Medical Center. Brains were dissociated by trituration and trypsinization (0.25% trypsin in PBS at 37°C for 15 min). The trypsin was inactivated with 10% heat-inactivated fetal bovine serum (Mediatech). The dissociated cells were filtered through 380- and 140-μm meshes (Sigma) and pelleted by centrifugation followed by plating with complete DMEM supplemented with 20 ng/mL EGF and FGF for alternate days.

Cloning of the Dcx promoter and site-directed mutagenesis

It was performed as described before with some modifications [17, 19–21]. Briefly, mouse Dcx promoter was cloned in lenti plasmid (pEZX-LvPF02). The promoter is 1760 bp long starting from 0 to –1760 bp upstream of start point. A GFP reporter gene was cloned under the guidance of that promoter. Primers were designed from NEB base changer of New England Bio labs Inc. website (nebasechanger.neb.com). The sequences of primers are as follows. Dcx: sense: 5^′ CAG CCT GGG CTA CAC AGC GAG TT 3^′; antisense: 5^′ CCA TCG CAT TTG CTG AAA ATT TTC AGC CTC. With the help of these primers, Dcx promoter was amplified, transformed into E. coli DH-5α competent cells and then enriched with Maxiprep method. After that, the plasmid DNA was mutated with the help of Q5 site directed mutagenesis kit (NEB, Cat# E0554S). PCR product was generated using Q5 hot start high fidelity master mix. The reaction was designed as instructed by manufacturer with annealing temperature set at 63°C. Amplified PCR product was then digested with enzyme mix containing Kinase, Ligase and DpnI. The mutated plasmid was transformed in E. coli DH5-α competent cells, enriched with Maxiprep method.

Lentiviral packaging

It was performed as described before [14, 22]. Briefly, HEK 293FT cells were cultured with 95% confluency in Opti-MEM media without antibiotics as described in manufacturer’s protocol (Invitrogen). Next day, ViraPowertrademark Packaging Mix (9μg/reaction) and pLenti expression plasmid DNA containing either pDcx (WT)-gfp or pDcx (mut)-gfp (3μg/reaction) (12μg total) were mixed in serum-free Opti-MEM® I Medium. In another tube, 36μL of Lipofectamine® 2000 was added in serum-free Opti-MEM® I Medium. After 5 min of incubation at room temperature, both the reactions were mixed and incubated for another 20 min. After that, the mixture was applied to HEK-293FT cells and incubated overnight at 37°C in a humidified 5% CO₂ incubator. Next day, the media was replaced with serum-free Opti-MEM media and further incubated for 48–72 h at 37°C in a humidified 5% CO₂ incubator and then sup containing viral particles was collected. Viral particles were concentrated with lenti-concentrator solution and MOI was calculated before application.

GFP reporter assay

Mouse HSCs were transduced with pDcx (WT)-gfp or pDcx (mut)-gfp lentiviral particles. Particles were added at a MOI of 10 on 70–80% confluent HSCs for 48 h. After that, cells were treated with different doses of OCT and gemfibrozil for 4 h. A reporter lysis buffer was added to end the reaction, and then kept at –80°C refrigerator for overnight. Next day, cell lysate was loaded in 96 well plates and GFP signal was measured in VictorX2 fluorimeter at a wavelength of 480 nm. Fluorescence was adjusted with the background value.

Embryo-specific calcium assay

A pregnant female cam ^cre Ppara ^{fl ox} (Ppara^ΔHip) mouse carries heterogeneous population of embryos with cam ^cre , Ppara ^flox , and pure cam ^cre Ppara ^flox genotypes. Therefore, combining all embryos for achieving sufficient numbers of hippocampal neurons for calcium assay irrespective of their genotype would be erroneous. Accordingly, we adopted embryo-specific calcium assay to nullify that error. Briefly, each embryo was separated, washed with sterile PBS, carefully dissected with sterile scissors to isolate hippocampus, disintegrated with sterile glass pipette to achieve single cell suspension, and then plated in a single row of 96 well plates. At the same time, the tails were cut to genotype the fetus. After one week of culture, the cells were analyzed with Fluo-4 based calcium influx assay as described elsewhere [14, 24].

Immunoblot and immunofluorescence assay

Immunoblot and immunofluorescence analyses were performed as described elsewhere [14, 25–27].

Software and statistical analyses

Counting of DCX, PPARα, nestin, and BrdU-ir cells were performed in touch counting module of Olympus Microsuite software. Quantitative data was presented as the mean±SEM for animal studies. Statistical significance was assessed via an unpaired two-tailed Student’s t test, one-way ANOVA for single effector, and two-way ANOVA for dual factor with Student-Newman-Keuls post hoc analyses.

Study approval

Mice were maintained and experiments conducted in accordance with National Institute of Health guidelines and were approved (protocol # 18-045) by the Rush University Medical Center Institutional Animal Care and Use Committee.

The role of PPARα in neuronal commitment of hippocampal progenitor cells (HPCs)

Upregulation of DCX in HPCs has been frequently observed in the synthesis of new neurons in DG region of the hippocampus. However, the mechanism for the upregulation of DCX has been poorly studied. Recently we demonstrated that PPARα plays a central role in the maintenance of hippocampal function [14, 28], primarily learning and memory. Since, upregulation of DCX in HPCs and subsequent neurogenesis are inherent mechanisms of hippocampal function, in our present study, we wanted to explore the role of PPARα in the upregulation of DCX. First, we determined whether HPCs expressed PPARα. HPCs were prepared from the dorsomedial telencephalon of E18 fetal brains. After 7 days in vitro (DIV), HPCs were immunostained with PPARα and Ki67, a pluripotent marker of progenitor cells (Fig. 1A, B). Our immunofluorescence analyses indicated that similar to hippocampal neurons (Fig. 1C), HPCs strongly expressed PPARα. However, measuring co-localization of PPARα with DAPI (Fig. 1D) followed by analyzing MFI of PPARα in nucleus and cytosol (Fig. 1E) revealed that the distribution of PPARα was mostly cytosolic in undifferentiated, partly nuclear in differentiating and fully nuclear in differentiated cells (Fig. 1D-E).

OPEN IN VIEWER

Fig. 1

Expression of PPARα in hippocampal progenitor cells (HPCs). HPCs were prepared from dorsomedial telencephalon of E18 embryos of both wild-type and Ppara-null mouse, cultured in complete DMEM media for 7 days, supplemented with EGF (20 ng/mL) and FGF (20 ng/mL) on alternate days, and then single progenitor cells were generated after disintegrating the neurosphere. Differentiation of wild-type progenitor cells were analyzed after treatment with 20μM of 9-octadecenamide (OCT), an endogenous hippocampal ligand of PPARα, for 2 days. Similarly, the differentiation of Ppara-null progenitor cells was monitored after overexpression of full-length Ppara gene followed by the treatment with 20μM OCT. A) Undifferentiated mouse stem cells and B) differentiating progenitor cells with fewer protrusions were labeled with Ki67 (red) and PPARα (green). C) Terminally differentiated neurons were stained with dendritic marker MAP2 (red) and PPARα (green). Nuclei were stained with DAPI. D) A comparative analysis of intracellular distribution of PPARα in three different stages of stem cells including undifferentiated, differentiating and fully differentiated neurons. E) A histogram analysis represents mean fluorescence intensity (MFI) of cytosolic and nuclear PPARα for all three types of stem cells. Results are mean±SD of 9 different cells per group. Significance of mean was calculated by paired t-test with F_1,16 = 35.07654 (> F _c  = 4.49); ^***p < 0.001 (= 0.0000215) between nuclear and cytosolic PPARα in undifferentiated cells. Similarly, significance was tested between nuclear and cytosolic PPARα in differentiated cells with F_1,16 = 177.4408 (> F _c  = 4.49); ^***p < 0.0001 (= 4.47^*10^–10), whereas no significance was observed in case of differentiating cells. NS, no significance.

Next, we were interested to study the effect of PPARα on neuronal commitment in these progenitor cells. HPCs prepared from both wild-type and Ppara-null E18 embryos were cultured for 7 DIVs followed by co-immunostaining with nestin and the dendritic marker MAP2 (Fig. 2A), which is also an early marker of neuronal maturation. Interestingly, we observed that the absence of PPARα significantly impaired the expression of dendritic protein MAP2 in nestin-stained Ppara-null progenitor cells, whereas no reduction of MAP2 immunoreactivity was observed in wild-type HPCs, suggesting that ablation of PPARα impairs the neuronal commitment of HPCs. The result was further validated by quantifying the number of MAP2-ir cells per every 20 nestin-ir cells (Fig. 2B) and with a Pearson correlation analysis, which revealed a strong negative correlation (r = –0.085; p < 0.0001) between MFIs of MAP2 and nestin (Fig. 2C). Next, we wanted to study if OCT-mediated direct activation of PPARα stimulated the numbers of MAP2-ir neuronal cells. Accordingly, 48 h of treatment with 20μM OCT strongly stimulated MAP2-ir cells in wild-type (Fig. 2D, E) HPCs. Consistent with this result, our MAP2 immunoblot (Fig. 2F) followed by relative densitometric analysis (Fig. 2G) clearly indicated that OCT strongly upregulated the expression of MAP2 in wild-type HPCs. Interestingly, treatment with OCT was unable to stimulate the expression of MAP2 (Fig. 2H–J) in Ppara-null HPCs, suggesting that PPARα is essential for the differentiation of HPCs into neurons. Moreover, ectopic overexpression of PPARα significantly restored the neurogenic ability of Ppara-null HPCs as we confirmed with immunofluorescence (Fig. 2K, L) and immunoblot analyses (Fig. 2M, N), suggesting that PPARα is indeed involved in the generation of new neurons from progenitor cells. The neurogenic niche of DG has been reported to be heterogeneous in terms of morphology and marker protein expression [29, 30]. Accordingly, our immunofluorescence analyses revealed that along with nestin, both wild-type and Ppara-null progenitor cells express the same levels of astroglial marker glial fibrillary acidic protein or GFAP (Fig. 3A, B) and oligodendroglial marker galactocerebrosidase or GalC (Fig. 3C, D). The result was further confirmed with a quantification study by counting number of GFAP-ir cells (Fig. 3E) and GalC-ir cells (Fig. 3F) per 20 nestin-ir HPCs. Although, 20μM OCT apparently stimulated the numbers of GalC-ir cells in WT HPCs, but similar upregulation was also observed in Ppara-null HPCs indicating that OCT might employ a PPARα-independent pathway for oligodendroglial differentiation. Interestingly, agonizing PPARα with OCT did not upregulate the expression of these glial markers further indicating that activation of PPARα does not stimulate glial differentiation of HPCs.

OPEN IN VIEWER

Fig. 2

The role of PPARα in the differentiation of hippocampal progenitor cells (HPCs) into neurons. A) Wild-type and Ppara-null HPCs were immunostained with nestin (green) and MAP2 (red). B) Based on MAP2 signal, number of neurons per 20 nestin-immunoreactive (ir) WT and Ppara-null cells were counted and plotted as a histogram. A paired t-test was applied to test the significance of mean between two groups with ^*F_1,38 = 11.79245 (> F _c  = 4.35); ^*p < 0.005 (= 0.00262). C) Mean Fluorescence Intensities (MFIs) of MAP2 and nestin were calculated from randomly chosen 34 cells of WT and Ppara-null groups and then analyzed for the correlation as shown in a scattered plot with r = –0.085 and p < 0.0001. D) WT HPCs were treated with 20μM of OCT followed by staining with nestin (green) and MAP2 (red). E) Number of neurons were calculated per 20 nestin-ir cells from each group and then plotted as a histogram. Paired t-test was adopted to test the significance of mean between groups resulting F_1,38 = 6.054 (> F _c  = 4.50); ^*p < 0.05 (= 0.023). F) HPCs were treated with 20μM OCT for 24 h followed by immunoblot analyses with MAP2 (Millipore: Cat #AB5622-I dilution: 1 : 500). G) Relative density was measured with respect to control after normalizing with respective actin bands. ^**p < 0.001 (= 0.00074). Results are mean±SD of three different experiments. H) Similarly, Ppara-null HPCs were treated with 20μM of OCT followed by immunostaining with nestin (green) and MAP2 (red). I) WT and Ppara-null HPCs were treated with 20μM OCT for 24 h followed by MAP2 immunoblot analyses. J) Densitometric analysis was performed after normalizing with respective actin. A paired t-test was adopted to test the significance of mean that results ^***p < 0.001 (= 0.0006) versus WT control. K) Dual-immunostaining of nestin (green) and MAP2 (red) in Ppara-null HPCs transduced with GFP and full-length (FL)-Ppara lentivirus. L) Histogram analysis represents number of MAP2-ir neurons per 20 nestin-ir neurons. Significance of mean between GFP and FL-Ppara groups was tested by t-test with ^*F_1,22 = 11.06128 (> F _c  = 4.550); p < 0.01 (= 0.00306). (M) Immunoblot followed by (N) densitometric analyses were performed in gfp- and FL-Ppara-transduced HPCs. Densitometric values were normalized with actin followed by normalizing with maximum expression of OCT-treated FL-Ppara-transduced HPCs. ^*p < 0.05 (= 0.038) versus GFP-control and ^***p < 0.001 (= 0.00054). Results are mean±SD of three different experiments.

OPEN IN VIEWER

Fig. 3

The role of PPARα in the expression of glial markers in HPCs. A) Ppara-null and B) wild type HPCs were split from neurosphere and then cultured as single cells for 7 days in vitro (DIV) followed by stimulation with 20μM OCT. After that, progenitor cells were immunostained with astroglial marker GFAP (red) and nestin (green). Similarly, both C) Ppara-null and D) wild-type HPCs were treated with or without 20μM of OCT followed by double-labeling with oligodendroglial marker GaLC (red) and nestin (green). E) Number of GFAP-ir cells were counted in 20 nestin-ir cells of wild-type (control and Oct 20μM-treated; paired t-test, ns, no significance) and Ppara-null HPCs (control and Oct 20μM-treated; paired t-test, ns, no significance) and then plotted as histogram. F) Similar analyses were done while counting number of GalC-ir cells in HPCs. One-way ANOVA was performed to test the significance of mean between groups that results F_3,36 = 27.41 (> F_c = 2.86). A paired t-test between no-treated and treated (20μM) wild-type GalC-ir cells resulted ^**p < 0.01(= 0.00073). Similarly, t-test between untreated and treated Ppara-null cells generated ^**p < 0.01 (= 0.00497).

Next, we examined the role of PPARα in human neural progenitor cells (hNPCs). Our immunocytochemical analyses revealed that 48 h of treatment with OCT strongly upregulated the expression of MAP2 in human neural progenitor cells (Fig. 4A, B, G). This upregulation of MAP2 was also accompanied with neuron-like morphology and very little nestin immunoreactivity in the dendritic growth-cone (arrowhead) as shown in Fig. 4C. To confirm the role of PPARα in OCT-mediated upregulation of MAP2, hNPCs were infected with a dominant-negative mutant of PPARα or lenti-Y464D-Ppara [22], followed by stimulation with OCT. Interestingly, lenti-Y464D-Ppara strongly inhibited the commitment of these progenitor cells into neurons as evident by negligible MAP2 immunoreactivity in these cells either before (Fig. 4D, G) or after (Fig. 4E, G) OCT treatment. Moreover, careful observation revealed that these cells acquired astroglial morphology (inset) with GFAP immunoreactivity (Fig. 4F), suggesting that inactivation of PPARα in human progenitor cells might induce a gliogenic response. Together, these results suggest that the OCT-mediated activation of PPARα is necessary for determining the neuronal fate of hNPCs.

OPEN IN VIEWER

Fig. 4

PPARα is required for the differentiation of human progenitor cells to neurons. Human progenitor cells were cultured in DMEM for 7 days with the supplementation of EGF (20 ng/mL) and FGF (20 ng/mL) on alternate days. After that, neurospheres were split as single cell and then cultured. After two days, cells were treated with 20μM OCT for 48 h and then analyzed for immunocytochemistry. A) Control and (B) OCT-treated cells were immunostained with nestin (green) and MAP2 (red). Representative images were low magnification images with broader field. Higher magnification images were provided underneath after splitting into green and red channels. C) OCT-treated progenitor cell displayed residual nestin only at its dendritic growth cone (arrowhead). Human progenitor cells were transduced with D) Y464D, a ligand-dead mutant of PPARα for 48 h and then treated with E) 20μM OCT for another two days. After that, cells were immunostained with nestin (green) and MAP2 (red). Higher magnification images were split into separate channels and then provided under respective low magnification images. F) (Y464DPpara+OCT)-treated cells acquired astroglial morphology with strong GFAP (red) signal. G) MFI of MAP2-ir progenitor cells were counted form 12 randomly selected cells per group and then presented as a histogram. A Two-way ANOVA analysis (effectors are genotype and treatment) was performed to justify the significance of mean ^**p < 0.01 (= 0.001232) between control and OCT-treated group and ^***p < 0.001 (= 1.83^*10^–7) between control and Y464D group.

PPARα upregulates the expression of DCX

Since DCX has recently emerged as a selective marker of cells committed to the neuronal lineage, next, we were interested to study if PPARα regulated the expression of DCX in HPCs. Calcium/calmodulin-dependent protein kinase type II subunit alpha (CaMKIIα) is predominantly expressed in the hippocampal neurons [31–33] and neuronal progenitor cells of hippocampal niche [34]. Therefore, we generated conditional knock-out mice with PPARα floxed under the influence of CamKIIα cre gene. The resultant Ppara^ΔHip mice (Fig. 5A) were used to study the effect of hippocampus specific knock-down of PPARα on the expression of neuronal lineage makers. First, we analyzed the expression of 84 such genes in the dissected hippocampal tissue of E18 Ppara^ΔHip fetus and then compared with the age-matched wild-type littermates. The cDNA-based microarray analysis, summarized in a heat-map image (Fig. 5B), revealed the downregulation of 15 genes in Ppara^ΔHip embryos. These genes were midkine or neurotic growth promoting factor-2 gene (Mdk), neuritic growth supportive tyrosine kinase gene Alk, vascular endothelial growth factor a gene (Vegfa), PSD95 protein encoding discs large homolog4 gene (Dlg4), cAMP response element binding protein encoding gene (Creb1), microtubule associated protein encoding gene Map2, brain-derived neurotrophic factor gene Bdnf, oligodendrocyte transcription factor 2 gene Olig2, adenosine a1 receptor gene adora1, class IV POU domain containing transcription factor gene Pou4f1, epidermal growth factor gene Egf, neurogenin 2 or Neurog2, roundabout guidance reptor1 gene robo1, and doublecortin or Dcx. Interestingly, while analyzing the relationship among all these genes, our STRING network analyses revealed that Dcx was located at the center nodes of the gene cluster (Fig. 5C), indicating that the downregulation of Dcx could play pivotal role in the impairment of neuronal fate of the HPCs of Ppara^ΔHip animals.

OPEN IN VIEWER

Fig. 5

PPARα is required for the transcription of dcx in the DG. A) Cam ^cre Ppara ^flox (Ppara^ΔHip) mice were generated and genotyped for the transgenes of Cam ^cre (324 bp; top) and Ppara ^flox (700 bp; bottom) in our laboratory. B) A cDNA-based microarray analysis of 84 genes related to hippocampal neurogenesis was carried out in hippocampal tissue of E18 WT and Ppara^ΔHip fetuses. C) STRING network analyses of downregulated genes in Ppara^ΔHip hippocampus resulted Dcx in center node. D) A schema displaying PPRE in the Dcx promoter with detailed bit score map. Core bases marked in block letters. Primers were designed 133 bp around the PPRE for ChIP assay. E) ChIP assay followed by PCR amplification of Dcx promoter was performed in 7 DIV wild-type and Ppara-null HPCs treated with different doses of OCT. F) The result was further confirmed by real-time PCR. ^***p < 0.001 (= 0.000813) versus control (OCT 10μM) and ^***p < 0.001 (= 1.2^*10^–5) versus control (OCT 20μM). G) In situ ChIP of Dcx promoter was performed in hippocampal lysate of wild-type and Ppara^ΔHip mice after pulling-down with PPARα antibody. PCR amplification of RNA polymerase-precipitated DNA with primers of dcx promoter confirmed that PPARα was involved in the transcription of dcx gene. H) Real-time PCR analysis of dcx promoter was performed in PPARα-pulled down fraction. Results are mean±SD of three independent experiments with ^***p < 0.0001 versus PPARα-pulled down fraction of WT hippocampus. I) Core pluripotency network consists of three genes including nanog, oct4 and sox2. Hippocampal extracts of 6–8 weeks old wild-type and Ppara^ΔHip animals were analyzed for mRNA expression of (J) nanog [^***p < 0.001 (= 0.00034)], (K) oct4 [^***p < 0.001 (= 0.000026)] and (L) sox2 [^**p < 0.01 (= 0.00347)] genes by real-time PCR.

While exploring promoters of these genes, we observed the presence of two conserved PPAR responsive elements (PPRE) in the promoters of the Dcx gene with more than 90% core homology, one at –537 to –547 bp upstream and another between –544 and –566 bp upstream of start site. Between the two PPREs, the core homology is maximum in the proximal one (–537 to –547 bp) (Fig. 5D). Therefore, to investigate the recruitment of PPARα to the proximal PPRE, we performed ChIP analyses in both WT and Ppara-null HPCs upon OCT treatment. Interestingly, increasing doses of OCT stimulated the recruitment of PPARα to the Dcx promoter in WT, but not Ppara-null, HPCs (Fig. 5E, F). Similarly, in order to validate the recruitment of PPARα to the Dcx promoter in vivo in the hippocampus, we performed in-situ ChIP. We observed the recruitment of PPARα to the promoter of Dcx gene and this recruitment was impaired in the hippocampus of Ppara^ΔHip animals (Fig. 5G, H), indicating that the expression of Dcx in cultured HPCs and in vivo in the hippocampus is directly controlled by the recruitment of PPARα to its promoter. Activation of core pluripotency genes including Nanog, Oct4, and Sox2 was reported to inhibit neuronal commitment [35, 36] of HPCs (Fig. 5I). Therefore, we analyzed whether genetic ablation of PPARα in neuronal niche of the hippocampus stimulated the expression of these genes. Accordingly, we observed that the expression of all these genes significantly increased (Fig. 5J–L) in the hippocampus of E18 Ppara^ΔHip fetus. These results suggest that PPARα determines neuronal fate of HPCs not only by stimulating the transcription of Dcx gene, but also via downregulation of core pluripotency network comprising Nanog, Oct4, and Sox2 genes in the hippocampus.

To further confirm this finding, we performed a Dcx promoter reporter assay using Lenti-pdcx(WT)-Gfp construct, in which the transcription of Gfp is under the control of WT Dcx promoter (Fig. 6A). Human neural progenitor cells (hNPCs) were transduced with Lenti-pdcx(WT)-Gfp followed by treatment with OCT and gemfibrozil. Accordingly, increasing doses of OCT and gemfibrozil upregulated the expression of Dcx promoter-driven GFP as indicated with a steady increase in GFP fluorescence (Fig. 6B) and resultant MFI (Fig. 6C). Moreover, OCT-mediated stimulation of Dcx promoter activity was strongly inhibited with 2μM GW6471, a selective antagonist of PPARα [34], demonstrating that PPARα is indeed involved in the transcription of Dcx gene (Fig. 6B, C).

OPEN IN VIEWER

Fig. 6

Analysis of Dcx promoter activity in neural progenitor cells. A) Human neural progenitor cells were transduced with pdcx-gfp (GFP reporter under the control of dcx promoter) lentivirions. Two days after transduction, cells were treated with increasing doses of OCT, gemfibrozil and PPARα antagonist GW6471. B) GFP immunofluorescence was monitored after treatment with 5, 10, and 20μM of OCT, 25μM of gemfibrozil and 20μM OCT+2μM of GW6471. Cells without transduction were included as negative control. C) MFI analyses of GFP were performed in 6 cells from each group and summarized in a histogram. ^*F_1,10 = 19.3315 (> F _c  = 4.50); ^*p < 0.05 (= 0.0312) between control and 10μM OCT, ^***F_1,10 = 55.709 (> F _c  = 4.50); ^***p < 0.001 (= 0.000018) between 25μM OCT and OCT+GW6471, and ^***F_1,10 = 55.728 (> F _c  = 4.50); ^***p < 0.0001 (= 0.000013) between control and 25μM Gem. D) Dcx promoter-driven Gfp construct [pdcx (WT)-Gfp] was cloned in lentivirus with topo vector-cloning strategy. Dcx promoter was mutated [pdcx(mut)-gfp] by site-directed mutagenesis. Mutated bases were marked with red color. E) WT HPCs were transduced with Lenti-pdcx(WT)-Gfp and after 48 h, cells were treated with different doses of OCT and Gemfibrozil (Gem) followed by GFP reporter assay measured in 480 nm filter at Victor X2 fluorimeter. ^*p < 0.005 (= 0.0015) versus control ^**p < 0.01(= 0.000917) versus control. F) Similarly, Dcx promoter activity was also monitored in Ppara-null HPCs infected with Lenti-pdcx(WT)-Gfp. G) WT HPCs were transduced with Lenti-pdcx(mut)-Gfp and after 48 h, cells were treated with different doses of OCT and Gemfibrozil (Gem) followed by GFP reporter assay. H) WT HPCs were transduced with Lenti-pdcx(WT)-Gfp and Lenti-pdcx(mut)-Gfp. After 48 h, cells were treated with different doses of OCT for 4 h followed by monitoring GFP in fluorescence microscope. I) MFI of GFP-ir cells were calculated from 20 random images from each group. After background correction, results are displayed in a histogram analysis. ^*p < 0.01 (= 0.00729).

To confirm the involvement of PPARα, we infected Ppara-null HPCs with Lenti-pdcx(WT)-Gfp (Fig. 6D) followed by stimulation with OCT and gemfibrozil (Fig. 6F). In contrast to WT HPCs (Fig. 6E), OCT and gemfibrozil remained unable to induce Dcx promoter-driven GFP activity (Fig. 6F), suggesting that PPARα plays an important role in the transcription of Dcx gene. To further demonstrate a direct role of PPARα in the transcription of Dcx gene, we performed site-directed mutagenesis to mutate PPRE of the Dcx promoter (Fig. 6D; lower panel). Then the resultant pdcx(mut)-Gfp was packaged in lentivirus to transduce 7 DIV WT HPCs. In contrast to Lenti-pdcx(WT)-Gfp, OCT and gemfibrozil did not induce GFP reporter activity in WT HPCs that were transduced with Lenti-pdcx(mut)-Gfp (Fig. 6G). Additionally, monitoring GFP fluorescence (Fig. 6H) followed by MFI quantification (Fig. 6I) further revealed that OCT was only able to upregulate Dcx promoter-driven GFP activity when HPCs were infected with Lenti-pdcx(WT)-Gfp, but not Lenti-pdcx(mut)-Gfp.

The direct role of PPARα in the upregulation of DCX is further confirmed when we observed that PPARα agonist OCT stimulated the expression of DCX in wild-type, but not Ppara-null HPCs (Fig. 7A, B). Moreover, our immunoblot results further demonstrated that reconstitution of PPARα restored neurogenic ability in Ppara-null HPCs and that property was further enhanced with the stimulation of OCT (Fig. 7C, D). Collectively, our results indicate that PPARα stimulates neurogenic events in the hippocampus via upregulation of DCX.

OPEN IN VIEWER

Fig. 7

The essential role of PPARα in the upregulation of DCX in HPCs. A) Wild-type and Ppara-null HPCs were treated 20μM OCT followed by immunoblot analyses of DCX. Actin was run as housekeeping protein. B) Densitometric analysis was performed after normalizing with respective actin. ^*p < 0.05 (= 0.032) and ^**p < 0.05 (= 0.038) versus wild-type control. C) Ppara-null HPCs were transduced with gfp and FL-Ppara lentivirus for 48 h followed by treatment with 20μM OCT for 24 h. After that, cells were analyzed for DCX and β-actin immunoblot expression analyses. ^*p < 0.05 (= 0.0327) versus GFP control and ^*p < 0.05 (= 0.0161) versus FL-Ppara. D) Relative densitometric analyses after normalizing with respective beta-actin bands. Results are mean±SD of three independent experiments.

PPARα-mediated upregulation of DCX preserves hippocampal function in mouse brain

Next, we wanted to study whether the downregulation of DCX and subsequent impairment of neuronal fate of HPCs affected the hippocampal function. Hippocampal function can be best assessed with measuring the influx of calcium ion via post-synaptic ionotropic membrane receptors including NMDA and AMPA receptors [23, 24]. Therefore, first we wanted to analyze if the genetic ablation of PPARα in the neurogenic niche of DG impaired the expressions of NMDA and AMPA receptor subunits in hippocampus. Interestingly, the levels of NMDA receptor subunit NR2A, AMPA receptor subunit GluR1, and post-synaptic protein PSD95 were significantly less in the DG of Ppara^ΔHip mice as compared to wild-type mice, suggesting that PPARα stimulates the expression of calcium regulatory genes the in neurogenic niche in hippocampus (Fig. 8A, B). To evaluate the functions of NMDA and AMPA receptors in Ppara^ΔHip animals, we performed a unique embryo-specific calcium assay. Briefly, we pooled a batch of E18 fetus, generated a purified culture of neuronal lineaged cells from hippocampal tissue of each of the fetus and plated separately in each row of a 96 well plate. Each fetus was genotyped from their tail tissue. After one week, these cells were analyzed for NMDA and AMPA-dependent calcium influx and the result from each row was matched with the respective genotype data. Interestingly, these cells of Ppara^ΔHip, but not wild-type mice, displayed significantly reduced AMPA- (Fig. 6C) and NMDA- (Fig. 6D) stimulated calcium influx, suggesting that genetic ablation of PPARα in the neurogenic niche of hippocampus significantly impaired their excitatory calcium conduction properties. However, when compared with wild-type mice, both Ppara ^flox and cam ^cre mice displayed partial but non-significant impairment. In order to achieve unbiased analyses of our experiment, results of each individual experiment was randomized and then plotted in a scattered plot as a function of maximum fluorescence. Interestingly, most of the points distributed over the median line were derived from wild-type embryos, whereas values derived from Ppara^ΔHip fetus remained dispersed under the median line, suggesting that embryonic disruption of PPARα in hippocampus significantly impaired the calcium buffering capacity both via AMPA (Fig. 8E) and NMDA (Fig. 8F) receptors. Together, our results suggest that PPARα plays an important role in the regulation of DCX and ionotropic transmission in the neurogenic niche of HPCs.

OPEN IN VIEWER

Fig. 8

Downregulation of plasticity-associated hippocampal proteins and impairment of NMDA- and AMPA-induced calcium oscillation in Ppara^ΔHip hippocampal neurons. A) Immunoblot analyses of NR2A, GluR1 and PSD95 in hippocampal neurons of wild-type and Ppara^ΔHip animals followed by B) densitometric analyses normalized with β-actin. ^**p < 0.01 (= 0.0026) versus wild-type NR2A, ^**p < 0.01 (= 0.00894) versus wild-type GluR1, ^**p < 0.01(= 0.00204) versus wild-type PSD95. C) AMPA- and (D) NMDA-sensitive calcium oscillation in cultured hippocampal neurons of wild-type (Green), Ppara ^flox (Red), Cam ^cre (Yellow) and Cam ^cre /Ppara ^flox (Grey) embryos (n = 4). Max fluorescence of (E) AMPA and (F) NMDA-sensitive calcium assay collected from each assay was combined and then plotted. The solid line represents the median line.