The mouse NSCs used in this study were originally generated in our previous study (Lee et al., 2015). NSCs were cultured on Matrigel (Corning)-coated 24-well plates and under expansion conditions, with the culture medium consisting of DMEM/F12 (Corning) with N2 supplement (Gibco, United States), B27 supplement without vitamin (Gibco), penicillin/streptomycin (Gibco), non-essential amino acids (Gibco), β-mercaptoethanol (Gibco), fibroblast growth factor-basic (PeproTech, Korea, 20 ng/mL), and platelet-derived growth factor-AA (Prospec, East United States, 20 ng/mL).

For differentiation into astrocytes, NSCs were seeded at 2.5 × 10⁵ cells on Matrigel-coated 24-well plates. The following day, the cells were fed with astrocyte differentiation medium, consisting of DMEM/F12 with 15% fetal bovine serum (FBS; Corning, United States), N2 supplement (Gibco, United States), penicillin/streptomycin (Gibco), non-essential amino acids (Gibco), and β-mercaptoethanol (Gibco) for 7 days. For the differentiation of NSCs into neurons, NSCs were seeded at 1.5 × 10⁵ cells on poly-l-ornithine/laminin–coated 24-well plates. During the first 4 days, the cells were cultured in the neuronal differentiation medium containing DMFM/F12 (Corning), B27 supplement (Gibco), N2 supplement (Gibco), penicillin/streptomycin (Gibco), non-essential amino acids (Gibco), β-mercaptoethanol (Gibco), and fibroblast growth factor-basic (20 ng/mL; PeproTech). On day 4 of neuronal differentiation, the medium was replaced without fibroblast growth factor-basic, and the cells were cultured until day 8 of neuronal differentiation.

The quantitative PCR (qRT-PCR) array was used for parallel quantitative analysis of microRNA expression. MicroRNA profiling was performed using the Mouse miRNome Sanger miRBase microRNA Profiler Set (#RA670A-1, System Biosciences, United States). The expression of mouse microRNAs (based on Sanger miRBase version 14) was analyzed as per the manufacturer’s instructions. https://www.systembio.com/wp/wp-content/uploads/2020/10/miRNome_Profilers_Manual_WEB-1.pdf. The microRNA array data have been deposited in the Gene Expression Omnibus (GEO) under the accession number GSE241920 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE241920).

MicroRNA microarray data were normalized by logarithmic transformation, and values not measured in the microarray were removed. MicroRNA expression difference values between MEF, ESC, and NSC were visualized using the pheatmap R package.

Lentiviral green fluorescent protein (GFP)-microRNA sponge directed against the microRNA-325-3p vector was designed based on the pLVTHM (#12247, Addgene, United States) plasmid. First, the microRNA sponge was inserted by using T4 ligase into the Bspd1 (NEB, United States)- and Mlu1 (NEB)-digested pLVTHM plasmid and transformed into DH5α by using the heat shock method. Furthermore, selected colonies were incubated in Luria–Bertani broth with ampicillin for minipreparation. For self-ligation, the plasmid was digested with the Bspd1 and Mlu1 restriction enzyme and electroporation was performed to confirm the size.

To insert the plasmid inserted with the microRNA sponge into the inducible system vector pLVET-tTR-KRAB (#11644, Addgene), the backbone and insert were digested with the restriction enzymes Fsp1 and Msc1, and the cut insert was ligated into the backbone plasmid to generate an inducible microRNA sponge (Cosmogenetech, South Korea).

Lentiviruses were produced in HEK293T (293T) packaging cells in culture medium with DMEM containing 10% FBS by effect vector, packaging vector ps-PAX2 (#12260, Addgene), and envelop vector VSV-G (#12259, Addgene) for 48 h. The supernatant was transferred to a 50-mL conical tube and mixed with a Lenti-X Concentrator (Clontech, United States). After overnight incubation at 4°C, viral particles were collected and suspended in a conditioned medium.

Viral particles suspended in the NSC culture medium were added to 1 × 10⁴ NSCs on a Matrigel-coated 12-well plate with polybrene (5 μm/mL; Sigma Aldrich) for 48 h. After 2 days, the medium was carefully removed by using a pipette, and fresh NSC medium was added.

After lentivirus infection to NSCs, the infected cells were harvested and filtered using a 70-μm cell strainer. The filtered cells were washed with Dulbecco’s phosphate-buffered saline (DPBS) and transferred into a new FACS tube. An FACS Aria I flow cytometer (BD Biosciences, United States) was used by an operator to separate and sort GFP-positive cells.

Cell counting was conducted for every culture using an automated cell counter (Bio-Rad, United States) and counter slides (Bio-Rad) with trypan blue (Sigma Aldrich, United States). The numbers in the graphs were conceived by carefully checking them three times.

Fresh NSC medium 900 μL and MTT solution 100 μL (Sigma) were added into each well of a 96-well plate, and the plate was incubated at 37°C for 3 h. After incubation, the MTT solution was replaced with DMSO (Sigma, United States) and incubated on an orbital shaker for 15 min. The absorbance of the plate was read at OD = 540 nm.

The cells were collected by centrifugation and resuspended in 500 μL of binding buffer and 5 μL of annexin v-Cy3 (BioVision, United States) and incubated at room temperature (RT) for 10 min in the dark. The stained cells were analyzed by flow cytometry using an FL2 channel for annexin v-cy3.

The cells were stained with propidium iodide (PI) (Sigma-Aldrich) in binding buffer and incubated at RT for 15 min in the dark. After incubation, the stained cells were analyzed by flow cytometry.

To extract total RNA or microRNA-enriched RNA, we used an RNeasy and miRNeasy Mini Kit (Qiagen, Germany), following the supplier’s instructions.

Total RNA (1 μg) was reverse-transcribed into cDNA using a high-capacity cDNA reverse transcription kit (Applied Biosystems, United States) and miScript RT (Qiagen) following the manufacturer’s protocol.

All reverse transcription polymerase chain reactions (RT-PCRs) used Ex Taq polymerase (TaKaRa, Japan) and were performed for 25–28 cycles for all markers. The primers used to amplify cDNA samples are listed in Supplementary Table S1. Gene expression levels were evaluated by qRT-PCR using a Roche real-time PCR system (Roche, Switzerland) and SYBR Green (Thermo Scientific, United States). Primer sequences are listed in Supplementary Table S1.

The cells were fixed in 4% paraformaldehyde for 10 min at RT and washed with Dulbecco’s phosphate-buffered saline (DPBS; HyClone, United States). For cell permeabilization, 0.5% Triton X-100 (Sigma-Aldrich) in DPBS was added to the fixed cells for 10 min at RT, and the cells were blocked with 2% of bovine serum albumin (Gibco) in DPBS for 1 h at RT. Then, the cells were rinsed and incubated in the primary antibody solution overnight at 4°C and washed with 0.2% Tween 20 in DPBS. After washing, the cells were incubated in the secondary antibody solution for 1 h at RT and washed with 0.2% Tween 20 in DPBS. For nuclear acid staining, the cells were incubated in the 4′-6-diamidino-2-phenylindole (DAPI) antibody for 5 min at RT and washed with 0.2% Tween 20 in DPBS. Antibody information is shown in Supplementary Table S2.

The cells were lysed in RIPA lysis buffer (Thermo Scientific) with a protease inhibitor cocktail (Thermo Scientific). The samples were run on 8% SDS-PAGE gel for 20 min at 80 V and 40 min at 120 V and then transferred to a transfer membrane (Pall Life Sciences, United States) for 90 min at 110 V. The membrane was blocked in 5% skimmed milk in Tris-buffered saline with Tween 20 (TBST) for 1 h at RT. The membrane was incubated with the primary antibody in blot buffer overnight at 4°C and washed three times for 5 min each with TBST. Then, the blots were incubated with a secondary antibody conjugate to HRP for 1 h at RT. The membrane was washed three times for 5 min each with TBST and incubated with an appropriate luminol-based chemiluminescent substrate. After incubation, the membrane was visualized on an ImageQuant LAS 4000. Images were quantified using ImageStation and ImageJ software. Antibody information is shown in Supplementary Table S3.

To search for mRNA that might regulate mmu-miR-325-3p expression, we used the following public prediction algorithms and databases (miRDB and MGI):

https://www.informatics.jax.org/interaction/explorer?markerIDs=MGI:3619336 and https://mirdb.org/cgi-bin/search.cgi?searchType=miRNA&full=mirbase&searchBox=MIMAT0000558.

All data are presented as means ± SEM. All statistics were calculated using independent t-tests or analysis of variance (ANOVA) with the least significant difference test for post hoc analysis. A value of p < 0.05 was considered significant.

The samples consisting of MEFs, ESCs, and NSCs were used to identify NSC-specific microRNA. Small non-coding RNA was extracted from these samples. Then, we screened out NSC-specific microRNAs by miRNome-wide screening and confirmed by using real-time PCR. Heatmap analysis data indicated candidates that had a relatively high number of NSCs compared to MEFs and ESCs (Figure 1A). After normalization to the MEF group, microRNAs highly expressed in NSCs were identified (Figure 1A; left panel). Similarly, after normalization to the ESC group, microRNAs highly expressed in NSCs were identified (Figure 1A; right panel).

Based on these results, it was confirmed that microRNA-325 was highly expressed specifically in NSCs. Following the heatmap analysis data, we profiled microRNAs that were expressed highly in NSCs (Figure 1B and Supplementary Figure S1) by using a volcano plot. Among these microRNAs, we found microRNA-325-3p, which had the highest expression in NSCs, followed by brain- and neuronal-specific microRNA-9, which is known to play an important role in neuronal development (Figure 1C). Furthermore, we confirmed through qRT-PCR analysis that NSC-specific microRNA is highly expressed in NSCs (Figure 1D). In addition, we observed an increase in the expression of microRNA-325-3P in NSC-derived neurons and astrocytes after induced differentiation, suggesting that microRNA-325-3p is not only NSC-specific but also could play a critical role in the differentiation into the neuronal or glial lineage (Figure 1E).

To investigate the role of microRNA-325-3p in NSCs and NSC-derived differentiated cells, we designed a microRNA-325-3p-sponge carrying anti-sequence for binding to the microRNA-325-3p (Figure 2A). For generating an inducible system of sponge expression, we inserted the microRNA-325-3p-sponge into the EF1α/EGFP/TetR plasmid. To enhance the binding efficiency, three sites were designed on the sponge for binding to the target microRNA (Figure 2B). After generating lentiviral particles, the viral supernatant was concentrated with a concentrator and added to NSCs for 48 h to establish lentiviral-infected cells. For accurate separation of lentiviral-infected cells, only GFP-positive cells were cultured after fluorescence-activated cell sorting (Figure 2C). The GFP expression of the cultured cells was confirmed by flow cytometry (Figure 2D), and viral integration of the infected cells was confirmed by genomic DNA sequencing (Supplementary Figure S2).

To confirm the properties of NSCs in the infected group, we examined the mRNA and protein expression of NSC-specific genes in the infected NSCs (Supplementary Figures S3A, B). Additionally, we validated the differentiation ability of the infected NSCs by demonstrating their capacity to differentiate into neurons and astrocytes (Supplementary Figure S3C). These data showed that the infected cells had the characteristics of NSCs. Therefore, we successfully established a cell line capable of modulating the expression of microRNA-325-3p, a regulator of cell proliferation and differentiation in NSCs, by using an inducible system with the microRNA sponge.

MicroRNAs are known to be important to cell viability and proliferation. Here, we investigated a downregulation of microRNA-325-3p that affects the proliferation of NSCs by carrying out trypan blue and MTT assays. To induce downregulation of microRNA-325-3p with an inducible doxycycline system, doxycycline (final concentration: 100 ng/mL; Sigma-Aldrich) was added into the NSC culture medium for 72 h. The data showed that there was no difference between mock NSCs and scramble NSCs, but there was a small but significant decrease in the number of cells in sponge NSCs when downregulation of microRNA-325-3p was induced by the sponge (Figure 3A). Consistently, in the MTT assay, cell viability tended to decrease in sponge NSCs (Supplementary Figure S4). In addition, an annexin-v assay was performed with flow cytometry to measure the extent of apoptosis in cells (Aubry et al., 1999). The results show that the difference between control and staining treatment in mock NSCs and scramble NSCs was approximately 8%, while in sponge NSCs, labeled cells were increased to approximately 25% (Figures 3B, D). These results show that microRNA-325-3p downregulation can inhibit the proliferative capacity of NSCs and induce apoptosis.

To confirm cell death, an assay for the observation of the sub-G1 population was performed using propidium iodide (PI) staining in combination with flow cytometry. As shown in the results, in mock NSCs and scramble NSCs, there was no difference in the ratio of sub-G1 population of cells (Figures 3C, E). On the other hand, when the downregulation of microRNA-325-3p was induced by the sponge in sponge NSCs, the sub-G1 population of cells increased by approximately 15% compared with mock NSCs and scramble NSCs (Figures 3C, E). These results indicate that the downregulation of microRNA-325-3p induces accumulation in the sub-G1 phase population and that microRNA-325-3p regulates cell proliferation and apoptosis by targeting proliferation and apoptosis-related genes directly.

By conducting qRT-PCR analysis, we identified that the expression of microRNA-325-3p increases after differentiation into astrocytes from NSCs (Figures 4A, B). These results suggest the potential of microRNA-325-3p as a regulatory factor involved in the generation and function of astrocytes. To analyze whether the downregulation of microRNA-325-3p by the sponge affects astrocyte cell death upon differentiation, we used annexin-v and PI staining to analyze the population of cell death. Annexin-v analysis shows an approximate 16% increase in the population undergoing apoptosis in sponge NSC-derived astrocytes compared to mock NSC-derived astrocytes and scramble NSC-derived astrocytes (Figures 4C, D). Consistently, the populations of PI-positive cells were 6.6% ± 0.8% in mock NSCs, 6.5% ± 0.5% in scramble NSCs, and 17.3% ± 2.4% in sponge NSCs (Figure 4E). The graph shows a 10.5% increase in PI-positive cells during astrocyte differentiation in the group that induced downregulation with sponge expression (Figure 4F). These results suggest that microRNA-325-3p is involved in the differentiation from NSCs to astrocytes and targets genes related to cell differentiation into astrocytes and cell death in astrocytes.

Before the examination of differentiation and apoptosis-related regulation, we used prediction programs (miRDB and MGI) to search for the target of microRNA-325-3p, and candidate genes with the highest scores were screened out. Moreover, by using a database, we found that microRNA325-3p has binding sites in the 3’-untranslated region (UTR) of brain-specific angiogenesis inhibitor (Bai1) (Figure 5A). Bai1 has a conversed stretch of seven to nine nucleotides that can base-pair with the corresponding microRNA-325-3p. Furthermore, the qRT-PCR analysis revealed a significant increase in the mRNA expression pattern of Bai1 in NSC-derived astrocytes and a small increase in NSCs by microRNA-325-3p downregulation (Figure 5B). Furthermore, Western blot showed that microRNA-325-3p was a negative regulator of BAI1 in NSCs and NSC-derived astrocytes (Figure 5C). These results suggest that microRNA-325-3p targets and regulates BAI1 expression in NSCs and NSC-derived astrocytes.