Human immortalized keratinocyte HaCaT cells and human embryonic dermal fibroblast WS1 cells were purchased from the American Type Culture Collection. They were both cultured in DMEM (cat. no. D0822; Sigma-Aldrich; Merck KGaA) supplemented with 10% fetal bovine serum (FBS; cat. no. 04-001-1A; Biological Industries), 100 U/ml streptomycin and 100 U/ml penicillin (cat. no. 15140122; Gibco; Thermo Fisher Scientific, Inc.) at 37˚C in a 5% CO₂ atmosphere.

To explore the function of AUF1 in skin cells, overexpression and RNA interference experiments were carried out. AUF1 complementary (c)DNA was obtained from OriGene Technologies, Inc. (cat. no. SC107836). AUF1 cDNA was subcloned into the pHBAD-EF1-mcs-CMV vector (Hanbio Biotechnology Co., Ltd.) for adenovirus construction. Packaging and purification of adenovirus overexpressing AUF1 (Ad-AUF1) and negative control adenovirus (Ad-NC) were carried out by Hanbio Biotechnology Co., Ltd. AUF1 cDNA was also subcloned into the pHBLV-U6-ZsGreen-Puro vector for lentivirus construction. Lentiviruses targeting AUF1 [named short hairpin (sh)AUF1-1 and shAUF1-2, respectively] and a scrambled negative control lentivirus (sh-NC) were purchased from Hanbio Biotechnology Co., Ltd. The target sequences against AUF1 are listed in Table SI. A mock-infected control group (without vector) was also constructed to demonstrate the background expression value.

HaCaT and WS1 cells were cultured in a six-well culture plate. After reaching 70% confluence, the cells were both infected with adenoviruses and lentiviruses at a multiplicity of infection of 20 with DMEM free-serum for 6 h at 37˚C in 5% CO₂. After 48 h, the cells were harvested for further experiments, and the efficiency of infection was determined by reverse transcription-quantitative PCR (RT-qPCR) and western blotting.

Total RNA of HaCaT and WS1 cells was extracted using TRIzol^® (cat. no. 15596018; Thermo Fisher Scientific, Inc.), and the absorbance values at 230, 260 and 280 nm were measured to determine the purity and density of the RNA. Subsequently, 1 µg of total RNA from each sample was used to synthesize single-stranded cDNA with a PrimeScript™ RT reagent kit (cat. no. RR047A; Takara Bio, Inc.). The reaction was performed at 37°C for 15 min followed by 85˚C for 5 sec. cDNA was amplified via qPCR with TB Green^® dye (cat. no. RR420S Takara Bio, Inc.) on a Prism 7500 real-time PCR machine (Applied Biosystems; Thermo Fisher Scientific, Inc.). The thermocycling conditions included an initial denaturation step of 95˚C for 60 sec, followed by 40 cycles of amplification at 95˚C for 15 sec and then annealing at 60˚C for 30 sec. The β-actin gene was used to normalize the relative fold changes of the target genes (15). The details of the primer sequences are listed in Table SII.

HaCaT and WS1 cells were harvested and lysed in RIPA lysis buffer (containing 50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1% Triton X-100, 1% sodium deoxycholate and 0.1% SDS) containing protease inhibitor cocktail (Sigma-Aldrich; Merck KGaA) for 30 min at 4˚C. Protein concentration was subsequently measured using a BCA Protein Assay kit (cat. no. P0012; Beyotime Institute of Biotechnology). Protein (50 µg/lane) from each lysate was fractionated by 10% SDS-PAGE and transferred to polyvinylidene difluoride membranes (MilliporeSigma). After blocking with 5% non-fat milk in PBS and 0.1% Tween-20 for 1 h at room temperature, the membranes were blotted with AUF1 (cat. no. ab282018; 1:1,000) and GAPDH (cat. no. ab181602; 1:50,000) antibodies overnight at 4˚C. After washing four times with PBST, the membranes were incubated with horseradish peroxidase-conjugated goat anti-rabbit IgG (cat. no. A21020; 1:10,000; Abbkine Scientific, Co., Ltd.) secondary antibodies for 2 h. Following washing in PBST, the protein-bound antibodies were detected using an enhanced chemiluminescence stable peroxide solution (cat. no. DC10100; PointBio). All protein bands were visualized by a FluroChem MI imaging system (Alpha Innotech Corporation) at room temperature. Densitometry was performed using Image J Java 1.8.0-172 software (National Institutes of Health).

After infection, cell viability was assayed using a Cell Counting Kit-8 assay (CCK-8; cat. no. CK04; Dojindo Laboratories, Inc.). Cells were seeded in a 96-well plate (1x10⁴ cells/well) and cultured for 24 and 48 h at 37˚C in a CO₂ incubator. After 10 µl of CCK-8 solution was added to each well, the cells were incubated for another 2 h. Absorbance values were determined at 450 nm using a microplate reader (Bio-Rad Laboratories, Inc.).

To measure the proliferation of infected skin cells, an EdU incorporation assay was performed by an EdU-based cell proliferation kit (cat. no. C0071; Beyotime Institute of Biotechnology) according to the manufacturer's instructions after 48 h of infection. In brief, HaCaT and WS1 cells at 50-70% confluence were incubated with 50 µM EdU for 2 h at 37˚C. Subsequently, the cells were fixed with 4% paraformaldehyde for 15 min at room temperature, permeabilized with 0.3% Triton X-100 and washed with phosphate-buffered saline (PBS). The cell nuclei were stained with 5 µg/ml Hoechst 33342 for 30 min at room temperature. A florescence microscope (Olympus Corporation; magnification, x20) was used to examine the cells.

HaCaT and WS1 cells infected with the indicated virus were seeded in 60-mm petri dishes at low density (1,000 cells/dish). After 10 days, the cells were fixed with 3.7% methanol for 30 min at room temperature and stained with 0.1% crystal violet for 30 min at room temperature. Colonies consisting of >50 cells were counted manually under a light microscope (Olympus Corporation; magnification, x40).

Infected HaCaT and WS1 cells were seeded onto 6-well plates and cultured for 24 h at 37˚C to form a 50% confluent monolayer. A straight scratch was gently and slowly created using a p200 pipette tip in the cell monolayer. Images of the wound were captured using a light microscope (Olympus Corporation; magnification, x10). Following two gentle washes with PBS, the plates were replenished with fresh DMEM containing 2% FBS for regular culture. Images of the width of the scratched regions after 24 and 48 h were captured and wound closure was assessed using Image J Java 1.8.0-172 software (National Institute of Health). The relative wound width rate was calculated by dividing the changed distance in the scratched region by the initial distance.

Early and late stage of apoptosis was evaluated using an Annexin V-FITC apoptosis detection kit (cat. no. C1062; Beyotime Institute of Biotechnology). HaCaT and WS1 cells were harvested 48 h post-infection and then stained with Annexin V/PI for 30 min at room temperature. To compute the percentage of apoptotic cells, a flow cytometer (FACSCalibur; BD Biosciences) with ModFit's LT v.3.0 software (BD Diagnostics) was used for data analysis.

HaCaT and WS1 cells were fixed in 2% formaldehyde/0.2% glutaraldehyde for 5 min at room temperature. β-Galactosidase staining solution containing X-gal (cat. no. C0602; Beyotime Institute of Biotechnology) was added after rinsing with PBS. The cells were then incubated for 6-10 h in a 37˚C incubator without CO₂. Senescent cells (stained blue) were observed and images were captured using light microscopy (Olympus Corporation; magnification, x10), and positive staining areas were calculated by determining the percentage of SA-β-gal⁺ cells in five random fields in each of the three wells.

HaCaT and WS1 cells were seeded in 100-mm petri dishes. At 24 h, the cells were collected, washed with PBS and fixed with 70% precooled ethanol overnight at 4˚C. After the cells were stained with propidium iodide staining solution, including RNase A (50X; cat. no. C1052; Beyotime Institute of Biotechnology) in the dark for 30 min at 37˚C to detect S phase, G₁ phase and G₂/M phase arrest. The cell cycle distribution was analyzed using a FACSCalibur system with ModFit's LT software.

Total RNA was extracted from infected WS1 cells with TRIzol^® reagent (cat. no. 15596018; Invitrogen; Thermo Fisher Scientific, Inc.) and quantified by a NanoDrop ND-2000C spectrophotometer (Thermo Fisher Scientific, Inc.). RNA integrity was assessed using the Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA). The libraries were subsequently constructed using a TruSeq Stranded mRNA LT Sample Prep kit (cat. no. RS-122-2302, Illumina, Inc.) according to the manufacturer's protocol. Transcriptome sequencing and analysis were conducted by Shanghai OE Biotech Co., Ltd. (Shanghai, China) http://oebiotech.bioon.com.cn/. The libraries were sequenced on an Illumina Novaseq platform (Illumina, Inc.) where 150 bp paired-end reads were generated. Raw data (raw reads) of fastq format were firstly processed using in-house perl scripts and the low-quality reads were removed to obtain the clean reads. The clean reads were subsequently mapped to the human genome (GRCh38) using Hierarchical Indexing for Spliced Alignment of Transcripts (Hisat2 v2.0.5.). Fragments per kilobase per million reads sequenced of each gene was calculated using StringTie, after which the read counts of each gene were obtained using featureCounts v1.5.0-p3. Differential expression analysis was performed using the DESeq2 R package (1.16.1). P<0.05 and a fold-change >2 or <0.5 were set as the thresholds for significant differential expression. Hierarchical cluster analysis of differentially expressed genes (DEGs) was performed to demonstrate the expression pattern of genes in different groups and samples. Kyoto Encyclopedia of Genes and Genomes pathway enrichment analysis of DEGs were performed respectively using the clusterProfiler R package sed on the hypergeometric distribution. The Gene Expression Omnibus (GEO) number was GSE138621.

All the data are presented as the mean ± SEM of at least three independent experiments and were analyzed using SPSS 19.0 software (SPSS Inc.). Differences between two groups were determined using a paired Student's t-test, and differences among >2 groups were analyzed by one-way analysis of variance and Tukey post hoc tests. P<0.05 was considered to indicate a statistically significant difference.

To explore the role of AUF1 in human skin cells, AUF1 was overexpressed or silenced using an adenovirus or a lentivirus, respectively. The efficiency of infection in HaCaT and WS1 cells was verified by RT-qPCR (Fig. 1A and B) and western blotting analysis (Fig. 1C and D). Compared with the sh-NC group, sh-AUF1-1- and sh-AUF1-2-infected cells demonstrated reduced levels of AUF1 protein. As expected, AUF1 protein expression in the Ad-AUF1-infected group was increased markedly As presented in Fig. 1A-D, AUF1 expression was successfully modulated by virus infection. Subsequently, cell viability was assayed using CCK-8-based analysis. As presented in Fig. 1E-H, AUF1 overexpression significantly increased cell viability compared with Ad-NC at 24 and 48 h; whereas AUF1 silencing (sh-AUF1-1 and sh-AUF1-2) significantly suppressed cell viability in HaCaT and WS1 cells compared with sh-NC at 24 and 48 h.

To explore the role of AUF1 in cell proliferation of skin cells, EdU staining and colony formation assays were performed. As presented in Fig. 2A and C, WS1 cells in the AUF1 overexpression group revealed a significantly increased EdU staining percentage compared with that in the control groups, while no significant difference was observed in HaCaT cells. Knockdown of AUF1 did not show a significant effect on the EdU-positive cell percentage in the two types of skin cells (Fig. 2B and D).

Next, the role of AUF1 in skin cell colony formation was investigated. Overexpression of AUF1 resulted in a significant increase of HaCaT and WS1 colony foci compared with Ad-NC (Fig. 3A and C). Conversely, infection with the two AUF1-silencing lentiviruses caused a significant decrease in the number of colonies in both HaCaT cells and WS1 cells compared with sh-NC (Fig. 3B and D). Colony formation analysis consistently demonstrated that AUF1 expression promoted cell proliferation.

A scratch-wound assay was used to evaluate cell migration, which is important for wound healing. As presented in Fig. 4A and C, when compared with mock or Ad-NC group, the migration of HaCaT and WS1 cells was significantly increased with AUF1 overexpression at 24 and 48 h. By contrast, when compared with the sh-NC group, fewer AUF1-silenced HaCaT and WS1 cells migrated toward the center of the gap of the cells, leading to a decreased rate of wound healing at 24 and 48 h (Fig. 4B and D). Migration inhibition induced by AUF1 depletion was observed even after incubation for 48 h. Overall, these results demonstrated that AUF1 expression can affect the migration of skin cells.

Cell senescence was analyzed using β-galactosidase staining, and the relative numbers of β-galactosidase-positive skin cells were identified. In the AUF1 overexpression group of skin cells, there was a significant decrease in the number of β-galactosidase-positive cells following infection with AUF1-overexpressing adenovirus compared with the Ad-NC groups (Fig. 3E and G). The positive areas were significantly reduced by Ad-AUF1 infection in HaCaT and WS1 cells (Fig. 3E and G). Conversely, as presented in Fig. 3F and H, knocking down AUF1 nearly tripled the β-galactosidase staining-positive areas in the two shRNA lentivirus-infected HaCaT cells and WS1 cells compared with the sh-NC. These results reflected the positive function of AUF1 in cell senescence.

As cell proliferation was affected by AUF1 expression, the function of AUF1 expression on apoptosis was investigated using an Annexin V/PI double-staining assay. Due to the pretreatment time before cytometry analysis, there were higher basal levels of apoptosis in HaCaT cells. The Ad-AUF1 groups in both HaCaT and WS1 cells demonstrated significantly reduced apoptosis rates compared with the control groups; by contrast, the apoptosis rates of AUF1-downregulated skin cells were significantly increased compared with the controls (Fig. 5A-F). These results indicated the antiapoptotic role of AUF1 in skin cells.

AUF1 may play a role in cell proliferation by disturbing the cell cycle; therefore, a cell cycle analysis was conducted using flow cytometry. However, no statistical significance for the percentage of cells in different phases was observed between any infected group (either the AUF1 overexpression group or AUF1 knockdown groups) and the corresponding control group in the two types of cells (Fig. 5G-J). As presented in Fig. 5K-L, AUF1 did not modulate cell cycle progression in HaCaT and WS1 cells.

As AUF1 is an RBP, multiple transcripts and pathways may be affected by the action of AUF1. To explore AUF1-affected mRNAs, WS1 cells with AUF1 overexpression and silencing were used to perform RNA-Seq and KEGG pathway analysis (Fig. 6A). The raw data are accessible at GSE138621. The results indicated that knockdown of AUF1 influenced >10,000 mRNAs, while overexpression of AUF1 caused a significant expression change in 6,771 mRNAs (Fig. 6B-D). A total of 6,771 mRNAs differentially expressed in AUF1-overexpressing cells included 3,422 up- and 3,349 downregulated mRNAs. AUF1 silencing involved 5,966 up- and 6,023 downregulated transcripts (Fig. 6D). A total of 18 mRNAs (eight mRNAs with positive associations and 10 mRNAs with negative associations) demonstrated consistent associations with both AUF1 overexpression and silencing (Fig. 6D): Mitochondrially encoded cytochrome B, anti-muellerian hormone (AMH), secreted protein acidic and cysteine rich (SPARC), forkhead box G1 (FOXG1), G protein-coupled receptor class c group 5 member A, cysteine rich angiogenic inducer 61 (CYR61), NADH-ubiquinone oxidoreductase chain 3, C-X-C motif chemokine ligand 6, oligosaccharyltransferase complex subunit 4, non-catalytic, mitochondrially encoded ATP synthase membrane subunit 6, phosphoserine aminotransferase 1, solute carrier family 7 member 5, methionyl-TRNA synthetase 1 (MARS), glycyl-TRNA synthetase 1, phospholipid scramblase 1 (PLSCR1), 2'-5'-oligoadenylate synthetase 1, nuclear autoantigen Sp-100 (SP100), chromosome 17 open reading frame 51 (Fig. 6E). KEGG pathway analysis revealed AUF1-affected networks involved multiple biological processes, including cell adhesion molecules, the proteasome and the spliceosome (Fig. 7A and B). mRNAs downregulated by AUF1 were associated with cell adhesion molecules, pathways in cancer, cellular senescence and the TGF-β signaling pathway, among others (Fig. 7C and D). These results suggested a complex regulatory network of AUF1 in skin cells.