Aging Cell

Proliferating and senescent WI‐38 fibroblasts express different levels of antisense lncRNAs

Compared with proliferating (P), early‐passage (PDL 22), senescent (S, PDL 52), WI‐38 HDFs displayed a flattened and enlarged morphology and increased senescence‐associated β‐galactosidase (SA‐βgal) activity, a widely used senescence marker (Debacq‐Chainiaux et al., 2009; Fig. 1A). The flattened and enlarged morphology was also visualized using rhodamine phalloidin, which stains filamentous (F‐)actin (Fig. 1B). Western blot analysis revealed that S WI‐38 cells expressed lower levels of Sirt1 and HuR, while they expressed higher levels of p21 and p53 (Fig. 1C), as previously reported (Marasa et al., 2010). RNA‐Seq analysis was performed using these cell populations, and the subsequent characterization was focused on differentially expressed, SAL‐RNAs.

Among the differentially expressed lncRNAs were a number of naturally occurring antisense (NA) lncRNAs. Listed in Fig. 1D (and Table S1) are WI‐38 NA‐SAL‐RNAs showing >5‐fold higher in S compared with P (top) and those showing >4‐fold higher in P relative to S (bottom). These differences were validated for a handful of NA‐SAL‐RNAs by reverse transcription (RT) using random hexamers and real‐time, quantitative (q)PCR amplification using transcript‐specific primer pairs. Validated NA‐SAL‐RNAs showing higher expression levels in senescent cells (Fig. 1E) included RP11‐346A9.1, targeting the metallopeptidase ADAMTS19, and OSTN‐AS1‐001, complementary to the mRNA encoding the secreted protein osteocrin. Validated NA‐SAL‐RNAs exhibiting lower abundance in senescent cells (Fig. 1F) included VCAN‐AS1‐001, complementary to versican mRNA (encoding an extracellular matrix proteoglycan) and LSAMP‐AS1‐001, antisense to the mRNA encoding a limbic system‐associated membrane protein (LSAMP). In several instances, the levels of the NA‐SAL‐RNA and the complementary mRNA showed positive correlation (not shown), whether this finding reflects a positive influence of one transcript upon the other or simply joint transcriptional control awaits further investigation. In a different model of senescence, achieved by exposure of WI‐38 fibroblasts to ionizing radiation (IR, 10 Gray) followed by culture for an additional 10 days (Fig. S1A), expression of NA‐SAL‐RNAs followed a similar trend, showing elevated levels of RP11‐346A9.1 and OSTN‐AS1‐001 and reduced levels of VCAN‐AS1‐001 and LSAMP‐AS1‐001 in senescent (IR‐treated) cells (Fig. S1B,C). These results indicate that the changes in lncRNA expression are not limited to the replicative exhaustion model (P and S fibroblasts).

Pseudogene‐encoded and additional lncRNAs are differentially expressed with senescence

A number of lncRNAs encoded by pseudogenes were also found to be differentially expressed in senescent compared with proliferating WI‐38 cells (Fig. 2A, Table S2). Validation of a subset of pseudogene‐encoded (PE) transcripts using RT‐qPCR analysis is shown in Fig. 2B (higher in senescent cells) and in Fig. 2C (higher in proliferating cells). Validated upregulated PE‐SAL‐RNAs include the carcinoembryonic antigen‐related cell adhesion molecules CEACAMP10 and CEACAMP11, which are highly expressed in solid tumors, down in senescent cells (Fig. 2B). Among the downregulated PE‐SAL‐RNAs are transcripts expressed from the ribosomal protein L21 pseudogene and from the heterogeneous nuclear ribonucleoprotein A1 (HNRNPA1) pseudogene (high in cancer cells, low in senescent cells). In IR‐induced fibroblast senescence, the relative expression levels of PE‐SAL‐RNAs in IR‐treated relative to untreated fibroblasts were the same as that seen in S relative to P fibroblasts (Fig. S1D,E); in a different cell model of replicative senescence (early‐passage IMR‐90 and late‐passage IMR‐90), PE‐SAL‐RNAs showed a similar trend (Fig. S2A,B).

Other well‐annotated lncRNAs are listed in Fig. 3A (and Table S3). Validation of the differential expression in S cells for a subset of them is shown. BX004987.5, CTD‐2021J15.2, and RP11‐314P12.2 were elevated with senescence (Fig. 3B), and MALAT1, XIST, RP11‐394O4.4, RP11‐255A11.21, and MIAT (also known as GOMAFU) were lower with senescence (Fig. 3C). As the magnitude of changes in RNA levels differed between the RNA‐Seq and RT‐qPCR analyses, we assayed lncRNA changes by Northern blot analysis. As shown, MALAT1 and MIAT were indeed reduced in S relative to P cells (Fig. 3D). As above, in WI‐38 fibroblasts rendered senescent by DNA damage (IR) and in S and P IMR‐90 fibroblasts, the relative expression levels of several previously reported SAL‐RNAs largely recapitulated the trend seen in P and S fibroblasts (Fig. S1F,G and Fig. S2C,D).

The impact of these two lncRNAs, MALAT1 and MIAT, was tested by lowering their abundance in early‐passage (~PDL 22) HDFs; 5 days after transfection, cells were split and re‐transfected to ensure the downregulation of the noncoding RNA. As shown in Fig. 3E, by 10 days of silencing MALAT1 or MIAT using small interfering (si)RNAs directed to the respective RNAs, which reduced their levels by 70% or 50%, respectively (Fig. 3E), cultures displayed an increase in the numbers of senescent, SA‐βgal‐positive cells (Fig. 3F, left), and changes in protein markers of senescence [p53 and p21 were upregulated, HuR and Sirt1 were downregulated (Fig. 3F, right)]. These results suggest that the reduction in some of these SAL‐RNAs could directly contribute to implementing the senescent phenotype.

Identification of novel SAL‐RNAs, both higher and lower in senescent cells

A number of novel SAL‐RNAs were also identified in the RNA‐Seq dataset as being upregulated in proliferating fibroblasts (Fig. 4A; Table S4); among them, an example of the RNA‐Seq coverage plot (XLOC_023166; ‘SAL‐RNA1’) and several validated transcripts of novel senescence‐reduced lncRNAs are shown (Fig. 4B,C). Analysis of the subcellular distribution of these senescence‐downregulated SAL‐RNAs by RT‐qPCR analysis indicates that several of the transcripts were predominantly nuclear (e.g., XXLOC_023166, XLOC_018371, XLOC_007522, XLOC_045655, and XLOC_064216), while others (XLOC_068343 and XLOC_041539) showed extensive presence in both the nucleus and the cytoplasm, with modest change in their subcellular distribution in senescence (Fig. 4D); the predominantly cytoplasmic lncRNA 7SL and nuclear lncRNA 7SK were included as controls in the fractionation experiment.

A comparably long list of SAL‐RNAs was found to be upregulated in senescent fibroblasts (Fig. 5A, Table S5). A coverage plot for one such SAL‐RNA showing higher levels in senescent cells (XLOC_025931; ‘SAL‐RNA2’) and a number of validated SAL‐RNAs are shown (Fig. 5B,C). Many validated lncRNAs in this group showed a primarily nuclear presence (e.g., XLOC_025932, XLOC_024912, XLOC_025922, XLOC_025931), but most of them are found in both the nucleus and the cytoplasm, with modest changes in the subcellular distribution during senescence. The nuclear lncRNA XIST was included as a control. Analysis in other models of senescence, the aforementioned IR‐treated (senescent) versus untreated (proliferating) WI‐38 fibroblasts and IMR‐90 fibroblasts at early (proliferating, PDL 25) and late (senescent, PDL 49) passages, indicated that the trends of expression of novel upregulated and downregulated SAL‐RNAs (Figs 4 and 5) were also conserved among senescence models (Fig. S3). Limited SAL‐RNA analysis by Northern blotting is shown in Fig. S4 (Supporting information). In sum, our results suggest that the RNA‐Seq approach effectively identifies lncRNAs differentially expressed during senescence (Fig. 5D).

Modulation of cell fate by SAL‐RNAs

In order to begin to study whether SAL‐RNAs affect any aspect of cellular senescence (e.g., proliferation, survival, gene expression, SA‐βgal activity), we designed siRNAs directed at SAL‐RNA1 (XLOC_023166), SAL‐RNA2 (XLOC_025931), or SAL‐RNA3 (XLOC_ 025918, all three sequences in supplemental Fig. S5), which achieved reductions to ~50% of the original transcript concentration by 5 days after transfection (Fig. 6A). Other lncRNAs tested (XLOC_ 025932 and XLOC_ 018371) did not achieve sufficient silencing (not shown). As shown in Fig. 6B, by 5 days after transfecting senescent fibroblasts with the siRNAs, silencing SAL‐RNA2 or SAL‐RNA3 caused changes in cell morphology (Fig. 6B) and lowered cell numbers significantly (Fig. 6C). These changes were consistent with increased apoptosis, which was confirmed by the enhanced levels of cleaved PARP [poly (ADP‐ribose) polymerase], a marker of apoptosis (Fig. 6D) and by the increased expression of p53 (Fig. 6E). By contrast, 5 days after silencing SAL‐RNA1, cells still displayed a senescent phenotype and cell numbers did not decline compared with the control group (Fig. 6B,C); no PARP cleavage was seen in these cells, underscoring the absence of apoptosis in this transfection group (Fig. 6D). However, the levels of p53 protein and mRNA also rose in early‐passage fibroblasts 10 days after transfection with SAL‐RNA1‐directed siRNA (Fig. 6F); interestingly, in these fibroblasts, silencing SAL‐RNA1 elicited a senescent phenotype, as revealed by the enhanced expression of senescence protein markers p21 and p16 and the accumulation of SA‐βgal‐positive cells (Fig. 6G). Together, these results indicate that SAL‐RNA2 and SAL‐RNA3 (XLOC_ 025931 and XLOC_ 025918) likely protect the survival of senescent fibroblasts, while SAL‐RNA1 (XLOC_023166) helps to prevent the untimely onset of senescence.

Cell culture, transfection, small interfering RNAs and SA‐β‐galactosidase activity assay

WI‐38 HDFs were obtained from Coriell Cell Repositories and in DMEM (Invitrogen, Life Technologies, Grand Island, NY, USA) supplemented with 10% (v/v) FBS, antibiotics, and 0.1 mm nonessential amino acids (Invitrogen). All siRNAs were transfected at 20 nm final concentration using Lipofectamine‐2000 (Invitrogen); UUCUCCGAACGUGUCACGUdTdT (Ctrl siRNA) was from Ambion. Duplexes (dsiRNA, each sense and antisense, respectively) obtained from Integrated DNA Technologies were used to silence MALAT1 [GGAGCAGAGAGGUAUGGGAAGCAGA and CCUCGUCUCUCCAUACCCUUCGUU], MIAT [CUGUUUAAACAUUUCCACUUGCCAG and GACAAAUUUGUAAAGGUGAACGGUC], XLOC_023166/SAL‐RNA1 [CAUGCAUAUCAGCUCAGGUCUUAA and GUACGUAUAGUCAGUCCAGAAUU], XLOC_025931/SAL‐RNA2 [UUCUUUCCAGAUUUGUGUCACCUG and AAGAAAGGUCUAAACACAGUGGAG], and XLOC_025918/SAL‐RNA3 [UUUUGCUAAGUUCCCACGAUCAGC and AAAACGAUUCAAGGGUGCUAGUCG]. In each case, 5 days after transfection, cells were split and re‐transfected to ensure the downregulation of SAL‐RNAs. Senescence‐associated β‐galactosidase activity in WI‐38 cells was assessed using a kit from Cell Signaling, following the manufacturer's protocol.

Western blot analysis

Whole‐cell lysates, prepared in RIPA buffer, were separated by SDS‐polyacrylamide gel electrophoresis (SDS‐PAGE) and transferred onto PVDF membranes (Invitrogen iBlot Stack). Incubations with primary antibodies recognizing p21 (Millipore, Billerica, MA, USA), HuR, Sirt1, p53, p16, cleaved PARP, GAPDH (Santa Cruz Biotech, Santa Cruz, CA, USA), or β‐actin (Abcam, Cambridge, MA, USA) were followed by incubations with the appropriate secondary antibodies conjugated with horseradish peroxidase (HRP; GE Healthcare, Pittsburgh, PA, USA) and by detection using enhanced luminescence (GE Healthcare).

RNA isolation, RT‐qPCR, and Sequencing

Trizol (Invitrogen) was used to extract total RNA from proliferating and senescent cells according to the manufacturer's protocol. Total RNA was used for gene expression analysis by reverse transcription (RT) followed by quantitative (q)PCR analysis or by RNA deep sequencing (RNA‐Seq). RT was performed using random hexamers and reverse transcriptase (Maxima Reverse Transcriptase; Fermentas, Pittsburgh, PA, USA), and qPCR was carried out using gene‐specific primers and SYBR green master mix (Kapa Biosystems, Woburn, MA, USA) in an Applied Biosystems 7300 instrument (Applied Biosystems, Life Technologies, Grand Island, NY, USA).

For RNA sequencing, total RNA quality and quantity was assessed using the Agilent 2100‐Bioanalyzer; 100 ng of RNA was used for first‐strand and second‐strand cDNA synthesis followed by single‐primer isothermal amplification using NuGEN Ovation RNA‐Seq System V2 kits according to the manufacturer's protocol. The kit amplified both polyA‐tailed and non‐polyA‐tailed RNA and removed ribosomal RNA. The amplified cDNA was sheared using Bioruptor (Diagenode, Denville, NJ, USA) to an average size of 250–450 bases. The sequencing library was prepared using Illumina ChIp‐Seq kits according to the manufacturer's protocol (Illumina, San Diego, CA, USA). In short, the ends of the fragments were repaired using T4 DNA polymerase, E. coli DNA Pol I large fragment (Klenow polymerase), and T4 polynucleotide kinase and adenines were added to the 3′ end. Adapters were ligated to the DNA fragments, which were size‐selected (250–300 bases) after electrophoresis through a 4% agarose gel. Eighteen cycles of PCR amplification were performed, followed by cluster generation and sequencing with Illumina Genome Analyzer (GA‐II). Sequencing was performed for 42 cycles, and the images generated were analyzed with the Firecrest program followed by base calls using the Bustard program; Firecrest and Bustard are part of the Illumina Analysis Pipeline package.

For RNA‐Seq analysis, the quality of the bases was checked using FASTQC program and called bases were aligned to the human HG19 genome using the Tophat program, the Bowtie1 algorithm, and Ensembl hg19 (v62) as gene model annotations followed by genomic mapping (Trapnell et al., 2010). The aligned reads were assembled into transcripts (both known and novel) using Cufflinks program with Ensembl hg19 (v62) transcripts as a guide (Roberts et al., 2011a). FPKM (fragments per kilobase of exon model per million mapped reads) values were calculated after fragment bias correction and normalization to total hits. Significant changes in transcript expression levels were calculated using Cuffdiff program with a cutoff of fdr <0.1 and minimum number of five alignments. Data were visualized in the UCSC genome browser (Roberts et al., 2011b).

Northern blot analysis

Northern blot analysis was performed as previously described (Abdelmohsen et al., 2007). Briefly, whole‐cell RNA was isolated using Trizol (Invitrogen), denatured, size‐fractionated using 1.2% agarose‐formaldehyde gels, and transferred. Oligonucleotides TTGCCGACCTCACGGATT for MALAT1, CACCAACTCTCCCACTAGGCTATAA for MIAT, and CCAATGGATCCTCGTTAAAGGATTT for 18S rRNA were end‐labeled with [α³²P]dATP and terminal transferase and used to detect these RNAs.

lncRNA PCR sequencing

PCR sequencing primers TGCATGTGTGTGTGTGTGTG and CTCTGGGAATCTGGAACCAA (forward and reverse, respectively) were used to amplify SAL‐RNA1/XLOC‐166. The PCR product was resolved by electrophoresis through a 2% agarose gel, and the DNA was extracted from the gel using QIAquick Gel Extraction Kit (Qiagen, Valencia, CA, USA). The purified PCR product (500 ng) was combined with 6 μL of BigDye Terminator v1.1/3.1 Sequencing Buffer, and 3.2 pmol of forward or reverse primer as used for qPCR. The PCR sample was purified and analyzed using an ABI sequencer; sequences were analyzed by an ABI sequence scanner.

Primer pairs for lncRNA RT‐qPCR

Primer pairs (all forward and reverse, respectively) were as follows:

CTD‐2021J15.2: TGCTCATGACTCCTCTGTGG and GACCAGAAGTCGTGCCTAGC

BX004987.5: CCATCGGCTGTAAACTTGGT and CAGAATATGGGCCAGCCTTA

RP11‐314P12.2: CACAGGGAGGATGTGTTGTG and GAGGCTGCTGCAAGTTCAAGGTC

RP11‐394O4.2: TGCTTCCAGCACTGATGTTC and CCCCATCCTACCCTTCATTT

RP11‐255A11.21: TATGTTGCCACCATCTTGGA and CATAGGCCCT‐GCAGAAACAT

XLOC_023166/SAL‐RNA1: TGCATGTGTGTGTGTGTGTG and CTCTG‐GGAATCTGGAACCAA

OSTN‐AS1‐001: AGACGGGGTTTCACCATATT and GGTGCAGTGGT‐TCACAGTTG

LSAMP‐AS1‐001: ATGATGAGGCAGCAAGAAGG and ACAGTTCTG‐GGGTTGTTGGA

VCAN‐AS1‐001: TGCATTGATCGGTAACCTTATG and AGGCCTGAA‐TGGAGGACTTT

RP11‐346A9.1: CTCTGGACAGGCAGAGGACT and CCTAAGGAGC‐GTGAGGAGTG

AC079949.1: GCCTCGGATACCTCGGATAG and ATGGTTTAGCGC‐CAGGTTC

RP11‐262D11.2: ATGCAAAGAGGGGAAAACCT and TTGGAAGGTG‐GGTCCTAGTG

RP11‐30B1.1: GGCCACATACCTGCAAATCT and GTGCCCTGACCT‐TGTTTGTT

HNRNPA1P29: TGGTGGCATCAAAGAAAACA and GAATCACGGCC‐ATCAAAAGT

CEACAMP10: TGGGAAAACCCTTGTACCTG and AGGCATGAGCAA‐GACAGTT

CEACAMP11: CCCAGGCTAAAGATGTCCAA and GGGTCACTGTG‐GCTGGTACT

XLOC_068343: CACGTTCACCAAGTCGTCTG and GGAGAGGGGCAGGAATTAAG

XLOC_018371: AGGAGCAAAACCATCAGGAA and GGCAGTGAT‐TGTCTCCGAAT

XLOC_007522: GCCAAGGCCTGTGTAAATGT and CTGGCCTGCTG‐ATTTATGGT

XLOC_045655: CAAGGAAGGGCAGGAGATTA and CTTTTTGGGCA‐GCTAGCACT

XLOC_041539: GTCTCCAGCATGACGTCTCA and CCAGGCCTTAGGAGACAGG

XLOC_064216: GCTGGAAGCAAGGTTGAGAA and CATGAGGCTC‐AGTTCTGTCTG

XLOC_014487: CTTCCCCTCAAATCCTGACA and GGACAGAAGAA‐CGGGTGAGA

XLOC_025927: TGATTTCTGGCCATTTCACA and ACACCAGTCCA‐CCTCCTCAC

XLOC_025919: GCACAAATGGAGCAGTCAAG and GACAGGTAGA‐GGTGCACACG

XLOC_024912: ACAGAGAGCCTCGGTGGAT and AGGGTGCCTAC‐ATCCTTGGT

XLOC_025922: GTTGGAGTACAGTGGCATGA and TGGTGGTACACACTTATGGTCCT

TP53: AGGCCTTGGAACTCAAGGAT and TGAGTCAGGCCCTTCTGT‐CT

7SK: GACGACCATCCCCGATAGAG and GGGAGCGCAGCTACTCGTAT

7SL: CAAAACTCCCGTGCTGATCA and GGCTGGAGTGCAGTGGCTAT

Malat1: GCAGGCGTTGTGCGTAGAG and TTGCCGACCTCACGGATT

Xist: GTGTGTGAGTGTACCTACCGCTTT and CGACTAGCCCTAAG‐CCGAGTT

GAPDH: ATGGAAATCCCATCACCATCTT and CGCCCCACTTGATTT‐TGG

MIAT: TCCTGAAACTCCTCTTTGTTTAACTG and CACCAACTCTCCC‐ACTAGGCTATAA

XLOC_025918/SAL‐RNA3: ACTGCTGGGATAACGGTGAC and TCTG‐TGCTCAGCTCTGCAAT

XLOC_025932: TGAAGAAGCAAGAGCCATGA and TGGCACCACT‐GTGACTTGAT

XLOC_025931/SAL‐RNA2: GGATGCTGTGAGCTTTGTGA and GAAA‐CCCCCAGAGCTGAGAC