LncRNA LYPLAL1-AS1 rejuvenates human adipose-derived mesenchymal stem cell senescence via transcriptional MIRLET7B inactivation

Mesenchymal stem cell (MSC) senescence is a phenotype of aging. Long noncoding RNAs (lncRNAs) are emerging as potential key regulators of senescence. However, the role of lncRNAs in MSC senescence remains largely unknown. We performed transcriptome analysis in senescent human adipose-derived MSCs (hADSCs) and identified that the lncRNA LYPLAL1 antisense RNA1 (LYPLAL1-AS1) was significantly downregulated in senescent hADSCs. LYPLAL1-AS1 expression in peripheral blood was lower in middle-aged healthy donors than in young adult donors, and correlated negatively with age. Knockdown of LYPLAL1-AS1 accelerated hADSC senescence, while LYPLAL1-AS1 overexpression attenuated it. Chromatin isolation by RNA purification (ChIRP) sequencing indicated that LYPLAL1-AS1 bound to the MIRLET7B promoter region and suppressed its transcription activity, as demonstrated by dual-luciferase assay. miR-let-7b, the transcript of MIRLET7B, was upregulated during hADSC senescence and was regulated by LYPLAL1-AS1. Furthermore, miR-let-7b mimics promoted hADSC senescence, while the inhibitors repressed it. Finally, LYPLAL1-AS1 overexpression reversed miR-let-7b-induced hADSC senescence. Our data demonstrate that LYPLAL1-AS1 rejuvenates hADSCs through the transcriptional inhibition of MIRLET7B. Our work provides new insights into the mechanism of MSC senescence and indicates lncRNA LYPLAL1-AS1 and miR-let-7b as potential therapeutic targets in aging.


Introduction
Aging and aging-related chronic diseases such as cardiovascular and cerebrovascular diseases, central nervous diseases, and degenerative osteoarthritis are increasingly burdensome health issues. Stem cell exhaustion, impaired proliferation, and regenerative capacity are important causes of physiological and pathological aging [1,2]. Mesenchymal stem cells (MSCs) originate in the mesoderm and are isolated from diverse tissues including adipose [3], bone marrow [4], the umbilical cord [5], and peripheral blood [6]. MSC populations and pools also decline with age, contributing to human aging and agerelated diseases [7].
Given their self-renewal properties, multilineage differentiation potential, and extensive immunomodulatory effects, MSCs are promising tools for cell-based therapies for various diseases, including hematological diseases, autoimmune diseases, peripheral nerve injuries, cardiovascular diseases, and pulmonary infection [8][9][10][11], with numerous clinical trials currently underway [12], which demands extensive expansion of MSCs in vitro. However, like other cell types, MSCs undergo senescence in culture. Therefore, elucidating the molecular mechanisms of MSC senescence is essential for stem cell-based therapy in translational medicine.
Acquired senescence such as replicative senescence following extensive passaging including cell cycle arrest, impaired function, or loss of the regenerative phenotype limits the use of MSCs in aging-related disease [13][14][15][16]. Cellular senescence is a complex and potentially irreversible process driven by oxidative stress, DNA damage, telomere shortening, and oncogene activation [16]. Although MSCs have significant proliferative potential, they, as with other cells, present replicative senescence after multiple divisions, which is promoted by oxidative stressors such as hydrogen peroxide (H 2 O 2 ) [17]. Senescent MSCs exhibit enlarged and granular morphology, deficient cell proliferation and differentiation capacity, produce the senescence-associated secretory phenotype (SASP), and have increased senescence-associated beta-galactosidase (SA-β-Gal) activity [16,18]. The complex underlying mechanisms and regulatory networks of senescence remain to be fully elucidated. The P53-P21 and PRB-P16 pathways are two complementary senescence regulatory pathways that trigger and maintain cellular senescence [19].
Long noncoding RNAs (lncRNAs) are transcripts longer than 200 bp that lack protein-coding capacity. They are emerging as important and divergent regulators in biological processes through their interaction with chromatin modifiers, DNA, RNA, and RNA-binding proteins (RBPs) [20,21]. Accumulating evidence shows that lncRNAs are altered during aging and senescence stimulation [22,23]. A recent study using transcriptome profiling and loss-of-function screening showed that lncRNA-OIS1 is essential for oncogene-induced human fibroblast senescence by regulating CDKN1A and DPP4 [24]. Dong et al. revealed that lncRNA CYP7A1-1 is upregulated with age in human bone marrow-derived MSCs and that downregulating CYP7A1-1 rejuvenated aged bone marrow-derived MSCs through SYNE1 [25]. However, the role of lncRNAs in MSC senescence remains largely unknown.
In the present study, we profiled the transcriptome in senescent human adipose-derived MSCs (hADSCs). We found that lncRNA LYPLAL1-AS1 was significantly downregulated in senescent hADSCs and was inversely associated with the age of healthy donors. We also demonstrate that LYPLAL1-AS1 was a negative regulator of hADSC senescence, which was potentially mediated by downregulating miR-let-7b levels by suppressing MIRLET7B promoter activity. Our findings provide new insight into the underlying mechanism of hADSC senescence and indicate potential therapeutic targets in anti-aging.

Cell culture and differentiation
Human adipose-derived MSCs (hADSCs) were isolated from adipose tissues of healthy donors undergoing liposuction as described previously [26][27][28]. All experiments and procedures were approved by the Ethics Committee of Peking Union Medical College Hospital. hADSCs were maintained at a density of 1.7 × 10 5 cells/ml in a 75 cm 2 flask at 37 °C, 5% CO2 and were passaged with trypsin/ EDTA on 90% confluence as we previously described in detail [29].

Induction of hADSCs senescence
For replicative senescence, hADSCs were continuously cultured in normal medium in 75 cm 2 flask and were passaged at 1:2 ratio normally following the hADSCs culture steps [29]. For H 2 O 2 -induced senescence, hADSCs at 70% confluence were treated with hydrogen peroxide (H 2 O 2 ) (100 nM, 300 nM, and 500 nM) for 2 h [32], then cells were washed with PBS and were incubated in fresh media for 24-48 h. Cell senescence was evaluated by β-galactosidase assay using senescence β-galactosidase staining kit (YEASEN, Shanghai, China) according to manufacturer's instructions. Briefly, cells were fixed for 15 min at room temperature in 4% paraformaldehyde, washed twice with PBS, then were incubated with SA-βgal staining working solution overnight at 37 °C without CO 2 in dark. The positive cells were stained blue, and images were acquired using an inverted microscope (Olympus, Japan).

Clinical samples
Peripheral blood samples of healthy donors (n = 42) were provided by Peking Union Medical College Hospital. The study protocol was approved by the Ethics Committee of Peking Union Medical College Hospital and written informed consent was provided by all donors. The demographic data of donors was listed in Additional file 6: Table S1.
For overexpression, full-length LYPLAL1-AS1 was inserted into the LV5-EF1-a-EGFP-Puro lentivirus expression vector and packaged by GenePharma (Shanghai, China), and a lentiviral vector that expressed scrambled RNA was used as the negative control. hADSCs were infected with viral precipitates at a multiplicity of infection of 10, and stable cell lines were established by puromycin treatment as described previously [29].

RNA extraction and quantitative real-time polymerase chain reaction (qRT-PCR)
Whole blood RNA was extracted by the MolPure Blood RNA Kit (YEASEN, Shanghai) according to manufacturer's instructions. Total RNA of cultured cells was extracted using TRIzol (Invitrogen) according to manufacturer's instructions. cDNA was synthesized using M-MLV Reverse Transcriptase kit (Takara, Japan). qPCR was performed using SYBR-Green Master mix (YEASEN, Shanghai, China) on an QuantStudio Design & Analysis system (ABI, USA). Relative RNA levels were normalized to GAPDH expression using 2 −ΔΔCt method.
For miRNA detection, primers of miR-let-7b used for reverse transcription and qRT-PCR were designed by Sango Biotech (Shanghai, China), U6 was served as internal control. The primer sequences were listed in Additional file 6: Table S2.
Transcriptome RNA sequencing hADSCs from three donors were maintained in 6-well plates and were harvested at passage 3 and passage 10. RNA was extracted and library preparations and sequencing was processed on a Hiseq 4000 platform by Novogene (Beijing, China). Differential expression analysis was performed using the DESeq2 R package (1.20.0) with a false discover rate (FDR) cutoff of 0.05. The resulting P-values were adjusted using the Benjamini and Hochberg's approach for controlling the false discovery rate. |log2FoldChange|≥ 2 and adjusted P-value < 0.05 were chosen as the cutoff criteria for differentially expressed genes. The RNA-Sequencing data is available at NCBI under SRA accession number PRJNA803433.

Subcellular fractionation
Nuclear and cytosolic fractions was extracted using NE-PER Nuclear and Cytoplasmic Extraction Reagents (Thermo Fisher Scientific, USA) according to manufacturer's instructions [33]. RNA was extracted and qRT-PCR was performed to assess the relative expression in nuclear and cytoplasmic fractions.

Fluorescent in situ hybridization (FISH) assay
FISH was performed using an RNA FISH kit (RiboBio, China) as described previously [29]. Briefly, hADSCs cultured on coverslips were rinsed in PBS and fixed with 4% formaldehyde for 10 min. Cells were permeabilized in PBS containing 0.5% Triton X-100 at 4 °C for 5 min, washed three times, and prehybridized at 37 °C for 30 min. Then, anti-LYPLAL1-AS1, anti-U6, or anti-18S oligodeoxynucleotide probes (RiboBio, China) diluted in hybridization solution were incubated with cells in dark at 37 °C overnight. Cells were stained with DAPI and were imaged using a fluorescence microscope (Carl Zeiss, Germany).

Chromatin isolation by RNA purification (ChIRP)-sequencing
hADSCs were subjected to ChIRP assay as described [29,34]. Briefly, Antisense RNA probes were designed (Aksomics, China) to bind every 100 bp of LYPLAL1-AS1 and U1 (positive control) transcripts, with BiotinTEG biotin label conjugated to 3'end. hADSCs (2 × 10 7 ) were fixed with 1% glutaraldehyde for RNA-chromatin crosslinking, and were snap-frozen in liquid nitrogen and stored at -80 °C. Samples were added to lysis buffer supplemented with protease inhibitor PMSF at room temperature, and cross-linked DNA-RNA complexes were sonicated until cell lysate was clear to break DNA into 100-500 bp fragments. Cell lysate was hybridized with biotinylated RNA probes. The extraction and subsequent analysis of RNA, DNA, and nucleic acid-binding proteins were performed. Sequencing analysis was performed by Aksomics (Shanghai, China). Sequence reads were generated from Illumina HiSeq 4000, image analysis and base calling were performed using Off-Line Basecaller software (OLB V1.8.0). After passing Solexa CHASTITY quality filter, the clean reads were aligned to Human genome (HG19) using BOWTIE software (V2.1.0); The mapped reads were used for peak detection by MACS V1.4.2 (Model-based Analysis of ChIRP-Seq) software. Statistically significant ChIRP-enriched regions (peaks) were identified by comparison of IP vs Input or comparison to a Poisson background model (Cut-off p-value = 10 -5 ); The GO categories are derived from Gene Ontology (www. geneo ntolo gy. org); Pathway analysis are based on the latest KEGG (Kyoto Encyclopedia of Genes and Genomes) database. The ChIRP-sequencing data is available at NCBI under SRA accession number PRJNA788657.

Statistical analysis
Data were expressed as mean ± standard deviation. Student's t-test and One-way analysis of variance were used for comparisons between two groups and multiple groups, respectively. Statistically significant differences were defined as follows: *p < 0.05, ** p < 0.01, *** p < 0.001, and **** p < 0.0001. All statistical analysis was performed using GraphPad Prism7 software (GraphPad Prism, San Diego, CA).

LYPLAL1-AS1-regulated genes identified by ChIRP-seq
LYPLAL1-AS1 was distributed in both the nucleus and cytoplasm (Fig. 5A and Additional file 4: Fig.  S3). To explore the downstream genes regulated by LYPLAL1-AS1, we conducted chromatin isolation by RNA purification sequencing (ChIRP-seq) analysis of LYPLAL1-AS1 in hADSCs to capture its direct binding DNA targets, which showed that LYPLAL1-AS1 bound at the intron, exon, promoter, upstream regions, and intergenic sites (Fig. 5B, C). We performed Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses on all binding genes. KEGG pathway enrichment analysis suggested that LYPLAL1-AS1 potentially regulated the senescence-related pathways, including the JAK-STAT signaling pathway and the apoptosis pathway (Fig. 5D). GO analysis revealed that the target genes of LYPLAL1-AS1 participated in diverse biological processes, including regulation of oxidative stress-induced genes and regulation of ARF protein signal transduction (Fig. 5E).

Discussion
Elucidating the regulation, intervention, and rejuvenation of MSC senescence is important for MSC-based therapy for anti-aging. Repeated subculture and H 2 O 2 exposure induce MSC senescence, which are suitable models of MSC senescence in vitro [32,38]. In the present study, we demonstrate that LYPLAL1-AS1 is a negative regulator of hADSC senescence induced by repeated subculture and H 2 O 2 exposure, which was inversely correlated with aging. We also reveal that LYPLAL1-AS1 rejuvenates hADSC senescence through transcriptional inhibition of the MIRLET7B promoter.
LncRNAs have versatile regulation functions at multiple levels. LncRNA ANCR suppresses epidermal and definitive endoderm differentiation [33] and modulates osteogenic and adipogenic differentiation and tumor progression. LncRNA MEG3 promotes MSC osteogenic differentiation at transcriptional level by acting as a decoy to dissociate SOX2 binding at the BMP4 promoter to activate BMP4 [39], and acts as a competing endogenous RNA (ceRNA) to regulate osteogenic gene expression at post-transcriptional level [40]. LncRNA-p21 is distributed in both the cytoplasm and nuclear, coordinates with RCK RNA helicase to suppress target mRNA translation in the cytoplasm [41], and methylates and maintains H3K9me3 at the promoters of pluripotent genes to repress their transcription by interacting with heterogeneous nuclear ribonucleoprotein (hnRNP) K protein [42]. We have reported that LYPLAL1-AS1 is a key promoter of hADSC adipogenic differentiation by directly targeting and modulating DSP protein stability in the cytoplasm [29]. Herein, we demonstrate that LYPLAL1-AS1 participates in regulating hADSCs senescence, which was mediated by the transcriptional inhibition of MIRLET7B. Therefore, LYPLAL1-AS1 regulates hADSCs at multiple levels, and other potential functions of LYPLAL1-AS1 on hADSCs remained to be explored. In addition, the function of specific lncRNAs is not restricted to a specific protein and they may influence multiple potential targets in a cell type-dependent way. For instance, lncRNA HOTAIR participates in regulating MSC function, which may cause senescence-associated DNA methylation to impact MSC proliferation and differentiation [43]. In senescent human fibroblasts, HOTAIR is upregulated and prevents premature senescence by causing rapid decay of targets Ataxin-1 and Snurportin-1 [44]. In cancer cell lines, HOTAIR depletion may be correlated with cell cycle arrest by decreasing S phase cells, inhibiting cell proliferation rate, and promoting apoptotic levels [45,46]. Our study revealed one function of LYPLAL1-AS1 in preventing human adipose-derived-MSC (hADSC) senescence, and besides adipose tissue, LYPLAL1-AS1 also exhibits high expression in ovary, mammary and subcutaneous according to human RNA sequencing (RNA-Seq) expression data from UCSC (http:// genome. ucsc. edu), therefore, extending our research to other cell types is a promising direction in the future.
MSC senescence is a complex process. Besides the innate molecular dynamic changes in SA-β-Gal staining and P16, P21, and LMNB1 expression examined in the present research, their functional abilities might also change, including the differentiation property which is intricately regulated by multiple factors, including transcriptional factors, growth factors, and epigenetic factors such as ncRNAs (miRNAs, lncRNAs) [47][48][49]. Our previous research indicate that LYPLAL1-AS1 is a key regulator of adipogenic differentiation, promoting the differentiation of normal hADSCs toward adipogenic cells by directly targeting and modulating DSP protein stability [29]. As shown in Additional file 5: Fig. S4A, overexpression of LYPLAL1-AS1 partially reversed the adipogenic differentiation of senescent hADSCs, as indicated by oil red O staining, suggesting that LYPLAL1-AS1 overexpression ameliorates hADSC senescence, resulting in increased cell differentiation potential. The differentiation potential of MSCs changed with senescence [50,51]. Increasing evidence indicates that senescent MSCs exhibit elevated adipogenesis at the expense of osteogenesis which is consistent with the in vivo increased bone marrow adiposity in aging, leading to damaged bone formation ability in humans and animals [51][52][53]. Most studies indicate that osteogenic differentiation potential of MSCs deteriorates with age [38,54,55], and the adipogenic differentiation potential of MSCs is disputable. Some studies indicate that the adipogenesis potential of MSCs tends to decline with senescence [56,57]. In a well-controlled study using rat model, MSCs Renilla luciferase vector (pRL-TK), and LYPLAL1-AS1 siRNAs or control siRNAs. D qRT-PCR of miR-let-7b in LYPLAL1-AS1-silenced or -overexpressed hADSCs. E qRT-PCR of miR-let-7b in hADSCs during replicative senescence. F qRT-PCR of miR-let-7b in H 2 O 2 -treated hADSCs. qPCR data were normalized to GAPDH, n = 3. Data are shown as mean ± SD from three independent experiments; **p < 0.01, ***p < 0.001, ****p < 0.0001 undergoing long-term passaging exhibit impaired in vitro differentiation potential including complete loss of osteogenic potential and attenuated adipogenic potential [58].
Other studies indicate the increasing adipogenic potential of senescent MSCs [59][60][61]. The discrepancy about whether senescent MSCs harbor increased or decreased adipogenic potential may be due to the different in vitro culture conditions and complex in vivo microenvironment including cell-intrinsic cytokines and hormones. Therefore, cellular senescence might not adequately explain the change in MSC differentiation, and the cell-intrinsic mechanism that regulates the age-related Fig. 7 miR-let-7b promotes hADSC senescence, which is reversed by LYPLAL1-AS1. A qRT-PCR of miR-let-7b in hADSCs transfected with miR-let-7b mimic, NC, or miR-let-7b inhibitor. B SA-β-Gal staining in hADSCs transfected with miR-let-7b mimic, NC, or miR-let-7b inhibitor. C qRT-PCR of P16, P21, and LMNB1 in hADSCs transfected with miR-let-7b mimic, NC, or miR-let-7b inhibitor. D Western blot of P16, P21, and LMNB1 in hADSCs transfected with miR-let-7b mimic, NC, or miR-let-7b inhibitor. E qRT-PCR of miR-let-7b, P16, P21, and LMNB1 in hADSCs transfected with miR-let-7b mimic and (or) LYPLAL1-AS1 expression vector. F Western blot of P16, P21, and LMNB1 in hADSCs transfected with miR-let-7b mimic and (or) LYPLAL1-AS1 expression vectors. qPCR data were normalized to GAPDH, n = 3. Data are shown as mean ± SD from three independent experiments; *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001; scale bars: 200 µm change in MSC differentiation deserved comprehensive studies in the future. Besides, deficiency in cell proliferation is also a characteristic of senescent MSCs. We also found knockdown and overexpression of LYPLAL1-AS1 had little effect on the proliferation of young (P3) hAD-SCs (Additional file 5: Fig. S4B, C), while its overexpression significantly reversed the proliferation of senescent hADSCs (Additional file 5: Fig. S4C, D), suggesting that LYPLAL1-AS1 overexpression ameliorates cellular senescence by rejuvenating senescent hADSCs, resulting in increased cell proliferation. Future studies should explore the underlying cellular and molecular defects that may explain the altered cell proliferation in senescent hADSCs.
A common lncRNA-miRNA interaction mechanism is lncRNAs acting as sponges to compete with mRNAs for miRNA binding, also known as the ceRNA mechanism [62][63][64]. However, TargetScan predicted no binding sequence between LYPLAL1-AS1 and miR-let-7b (data not shown). Combined with the ChIRP-seq analysis showing direct binding between LYPLAL1-AS1 and the MIRLET7B promoter, we propose that LYPLAL1-AS1 inhibits MIRLET7B promoter activity, thereby indirectly modulating miR-let-7b levels at the transcriptional level. LncRNA regulation of gene transcription has been extensively studied [65,66]. LncRNAs may act as guides to recruit proteins to DNA through RNA-DNA and RNA-protein interactions [67][68][69]. Guide lncRNAs have both protein-binding functions and genome-interfacing functions. Here, we reveal that LYPLAL1-AS1 interfaced directly with the MIRLET7B promoter and regulate its activity negatively. Further studies are warranted to determine whether other proteins are involved in this process. Therefore, we suggest a new underlying mechanism of LYPLAL1-AS1 inhibition of hADSC senescence: it potentially inhibits the MIRLET7B promoter transcriptionally, thereby downregulating miR-let-7b levels.
Similar to protein-coding genes, miRNA host gene promoters contain CpG islands, TATA box sequences, initiation elements, and certain histone modifications that enable their control by transcription factors, enhancers, silencing elements, and chromatin modifications [70,71]. miRNA processing from the host gene or miRNA precursor transcription may affect mature miRNA levels [72]. Therefore, transcriptional regulation of miRNAs is partially responsible for their specific spatial and temporal expression patterns [73]. For example, p53 promotes miR-34 and miR-107 family expression, which enhances tumor cell cycle arrest and apoptosis [70,74]. PITX3 stimulates miR-133b transcription and contributes to the maturation and function of midbrain dopaminergic neurons [75]. Herein, we found that LYPLAL1-AS1 downregulated miR-let-7b levels by binding to the promoter region of its host gene MIRLET7B and suppressing transcriptional activity, thereby attenuating hADSC senescence. miR-let-7b is a let-7 family miRNA that is upregulated in the plasma of patients with agerelated macular degeneration [76]. A recent study has indicated that miR-let-7b is augmented in choroid and retinal pigment epithelium exposed to oxygen and that increased p53-miR-let-7b activity promotes the inability of the choroid to revascularize in oxygen-induced retinopathy [37]. However, whether miR-let-7b plays a role in hADSC senescence remains unknown. Our data demonstrate that miR-let-7b accelerates hADSC senescence, which is regulated by LYPLAL1-AS1. Therefore, targeting miR-let-7b might be a promising approach for reducing hADSC senescence.

Conclusions
We demonstrate that LYPLAL1-AS1 rejuvenates hADSCs through the transcriptional inhibition of MIRLET7B. Our work provides new insights into the mechanism of stem cell senescence. LYPLAL1-AS1 and miR-let-7b might be potential therapeutic targets in aging-related diseases.