https://orcid.org/0000-0003-3768-6352
Figures
Abstract
Long non-coding RNAs (lncRNAs) have emerged as key regulators of cellular senescence by transcriptionally and post-transcriptionally modulating the expression of many important genes involved in senescence-associated pathways and processes. Among the different lncRNAs associated to senescence, Senescence Associated Long Non-coding RNA (SALNR) was found to be down-regulated in different cellular models of senescence. Since its release in 2015, SALNR has not been annotated in any database or public repository, and no other experimental data have been published. The SALNR sequence is located on the long arm of chromosome 10, at band 10q23.33, and it overlaps the 3’ end of the HELLS gene. This investigation helped to unravel the mystery of the existence of SALNR by analyzing publicly available short- and long-read RNA sequencing data sets and RT-PCR analysis in human tissues and cell lines. Additionally, the expression of HELLS has been studied in cellular models of replicative senescence, both in silico and in vitro. Our findings, while not supporting the actual existence of SALNR as an independent transcript in the analyzed experimental models, demonstrate the expression of a predicted HELLS isoform entirely covering the SALNR genomic region. Furthermore, we observed a strong down-regulation of HELLS in senescent cells versus proliferating cells, supporting its role in the senescence and aging process.
Citation: Consiglio A, Venturin M, Briguglio S, Rossi C, Grillo G, Bellosta S, et al. (2023) Lost in HELLS: Disentangling the mystery of SALNR existence in senescence cellular models. PLoS ONE 18(5): e0286104. https://doi.org/10.1371/journal.pone.0286104
Editor: Jacopo Sabbatinelli, Universita Politecnica delle Marche, ITALY
Received: February 8, 2023; Accepted: May 9, 2023; Published: May 30, 2023
Copyright: © 2023 Consiglio et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All the relevant data are present in the manuscript and the Supporting information data can be found in UNIMI Dataverse [https://dataverse.unimi.it/dataverse/SALNR_HELLS] and DOI [https://doi.org/10.13130/RD_UNIMI/JY8KN5]. Short-read RNA-seq mapping tracks on UCSC genome browser: https://genome.ucsc.edu/s/cnr.itb.ba/short-read_RNA-seq. Long-read RNA-seq mapping tracks on UCSC genome browser: https://genome.ucsc.edu/s/cnr.itb.ba/long-read_RNA-seq.
Funding: This work was supported by ‘Piano di sostegno alla Ricerca 2020’ Grant (Università degli Studi di Milano, award no. PSR2020_BATTAGLIA_LINEA_B to CB and MV; award no. PSR2020_CATTANEO_LINEA_C to MGC), ‘NutrAge’ Grant (CNR-FOE 2021, award no. DBA.AD005.225) to AC. Bioinformatics analysis has been executed on ICT infrastructure funded by CNR-BiOmics project (Italian MUR, PON R&I 2014-2020 D.D. n. 424 28/02/2018) to FL. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
The progressive increase of the average human lifespan has drawn attention on the cellular and molecular determinants of aging, the time-dependent physiological impairment of cellular and tissue functions that has emerged as the main risk factor for several high prevalence pathologies (e.g., cancer, cardiovascular disorders, and neurodegenerative diseases), now referred to as “age-related disorders” [1]. Among the different hallmarks of aging, cellular senescence recapitulates many molecular features of aging and senescent cells were demonstrated to cause age-related phenotypes by accumulating throughout the body over time [2,3]. Cellular senescence is a highly heterogeneous and permanent state characterized by many cellular and molecular alterations such as permanent cell cycle arrest, senescence-associated secretory phenotype (SASP) that promotes the secretion of pro-inflammatory molecules, increased senescence-associated β-galactosidase (SA-β-gal) activity, altered cellular and nuclear morphology, induction of anti-apoptotic genes, DNA damage, epigenetic modifications, altered proteostasis, changes in cellular metabolism and accumulation of dysfunctional mitochondria [3,4]. The senescence state can be induced in vitro by prolonged cell culturing, as first described by Hayflick and Moorhead [5], or by other numerous stimuli among which DNA damage, oxidative stress, oncogene overexpression and mitochondrial dysfunction [2,3].
In the complex regulatory network orchestrating cellular senescence, long non-coding RNAs (lncRNAs), a large and heterogeneous class of non-coding RNAs longer than 200 bp, seem to play a key role by transcriptionally and post-transcriptionally modulating the expression of many important genes involved in the above mentioned senescence-associated pathways and processes, and often showing an altered expression during senescence and aging [6,7]. Among the different lncRNAs associated to senescence, SALNR (“Senescence Associated Long Non-coding RNA”), was first described by Wu et al. [8] by microarray experiments aimed at identifying deregulated lncRNAs during replicative senescence of lung fibroblasts. Authors pointed to SALNR as a consistently down-regulated lncRNA not only in replicative senescence but also in other cellular models of induced senescence. Moreover, they showed that SALNR regulates senescence by interacting with the NF90 RNA-binding protein: NF90 acts as repressor of senescence in the nucleus by inhibiting the biogenesis of senescence-inducing miRNAs through the binding to their corresponding pri-miRNAs. In senescent cells, NF90 translocates in the nucleolus, losing its inhibitory activity on miRNA biogenesis; in this scenario SALNR, located in the nucleus, would prevent NF90 translocation, preserving its repressive activity on senescence [8].
Despite these evidences, since its release in 2015, SALNR has not been annotated in any database or public repository and no other experimental data have been published. SALNR gene should be located in the long arm of chromosome 10, within the 10q23.33 band, overlapping the HELLS gene. HELLS gene encodes for a helicase belonging to the SNF2 family of proteins involved in lymphoid lineage cellular proliferation. In the last years, next-generation RNA sequencing (RNA-seq) approaches enabled the sequencing of protein-coding and non-coding RNAs together with the assessment of differential expression transcriptomic analysis in perturbed biological conditions. The upgrading of long-read RNA-seq platforms allowed the production of reads containing kilobase-sized RNA fragments, thus opening the way to tackle challenging regions of the human genome and to investigate RNA isoforms [9].
In this paper, we have tried to disentangle the mystery of SALNR existence by exploiting both in silico and in vitro approaches. We show that SALNR transcript is completely overlapped to a HELLS transcriptional isoform, the RefSeq transcript XR_007061960. We validated the existence of this HELLS isoform by analysis of publicly available RNA-seq datasets and by semi quantitative RT-PCR experiments, while there is no evidence of the existence of SALNR as separate, independent transcript in our experimental setting. Moreover, we found a strong down-regulation of HELLS in cellular models of replicative senescence.
Materials and methods
Identification of SALNR genomic location and annotation
SALNR was identified in 2015 [8] as cDNA clone AK091544 and its location on the human genome was situated on the long arm of chromosome 10, on the Crick strand, chr10: 96353372–96356258 on NCBI36/hg18. We first converted its genomic location on the GRCh38/hg38 assembly on chromosome 10, chr10: 94603625–94606511. Then we explored both the Ensembl Release 108 (Oct 22) [https://www.ensembl.org/index.html] and NCBI reference genome and transcriptome [NCBI Homo sapiens Annotation Release 110 (2022-04-12)] to extract its putative sequence and the annotated portions of DNA overlapping or flanking its transcription area. To search for lncRNA annotations, we also checked LNCipedia v.5.2, a comprehensive compendium of human long non-coding RNAs [10] [https://lncipedia.org/].
Search for SALNR/HELLS transcripts in public RNA-seq datasets
As first instance, we have searched for RNA-seq public datasets with accessible FASTQ files proving the transcription of SALNR/HELLS sequences and focusing on the study of replicative senescence in open repositories such as Gene Expression Omnibus (GEO, https://www.ncbi.nlm.nih.gov/geo/) and NCBI Sequence Read Archive (SRA, https://www.ncbi.nlm.nih.gov/sra). Overall, we selected five RNA-seq datasets from projects that compare senescent and proliferative human cell line experiments, listed in Table 1, including one dataset comparing proliferative and senescent WI-38 lung fibroblasts.
Moreover, to deeply investigate the structure of the genomic region located on chromosome 10 (chr10: 94603625–94606511 on GRCh38/hg38) including HELLS and SALNR transcripts, we checked for availability of FASTQ files of long-reads RNA-seq runs focusing on normal human tissues or cell lines in SRA.
To get comprehensive annotation on HELLS, three open resources have been consulted: Human Protein Atlas (https://www.proteinatlas.org/, Version: 22.0. Atlas updated: 2022-12-07), GTEx (https://www.gtexportal.org/home/, v8, update 2022-07-21) and GenAge (The Ageing Gene Database, https://genomics.senescence.info/).
Bioinformatic analysis
RNA-seq FASTQ files downloaded from SRA were quality checked with FastQC tool [17] (see Data availability for FastQC output files). Both short (Table 1) and long read datasets (S1 Table) were mapped against the Ensembl genome and transcriptome (Release 108). Short reads (50-150bp) were mapped with STAR [18], the expressions were estimated by RSEM v1.3.3 [19] and MultiDEA v1.1.0 [20] in order to manage multiread issues. Differential expression analysis of senescent and proliferative samples was performed with DESeq2 v1.36.0 [21], and the results were filtered by mean read count >50 for genes and >20 for isoforms, absolute log2 Fold Change (log2FC) >1 and adjusted p-value (FDR) <0.05. Long-read FASTQ files were mapped with deSALT v1.5.6 [22].
Below is a list of non-default parameters used for each tool. STAR was run through RSEM with—star option, and—paired-end was used for paired end samples, while deSALT was run with—gtf to use GTF annotation files and with -x clr for PacBio reads (PacBio SMRT CLR reads, error rate 15%) or -x ont1d for Nanopore reads (Oxford Nanopore 1D reads, error rate > 20%). A table containing read ID, transcript ID and mapping quality was extracted from the BAM output of mapping and used as input for MultiDEA’s module mapping_output_2_trapezoids.sh. DESeq2 was applied to raw read counts by testing design = ~ condition, where condition is “proliferative” or “senescent”. We applied DESeq2 at gene level to RSEM read counts and at isoform level to both RSEM and MultiDEA read counts, with the aim of correlating the differential expression events to the different isoforms of HELLS.
Since SALNR transcript totally or partially overlaps HELLS transcripts, some ad-hoc Bash scripts were developed to explore in deeper detail the characteristics of the reads mapping to their genomic location.
The results were loaded into the UCSC genome browser [https://genome.ucsc.edu/], and the corresponding URLs are listed in the Data availability section.
Human cell lines
Normal Human Dermal Fibroblasts (NHDF) were purchased from Sigma (C-12302, Sigma). Cells were cultured in EMEM (ECB2071L, Euroclone), 10% FBS (ECS0180L, Euroclone), 1% NEAA (ECB3054D, Euroclone), 1% 200 mM glutamine (ECB3000D, Euroclone), 1% Penicillin/Streptomycin (ECB3001D, Euroclone). Human aortic smooth muscle cells (HSMC) were purchased from the American Tissue Culture Collection (PCS-100-012, ATCC, Manassas, USA). Cells were cultured in ATCC Vascular Cell Basal Medium (PCS-100-030, ATCC; 500 mL added with 500 μL ascorbic acid, 500 μL rh EGF, 500 μL rh insulin and rh FGF-b, 25 mL glutamine), 5% FBS (ATCC Vascular Smooth Muscle Growth Kit), and 5 mL Penicillin-Streptomycin 100X (Euroclone, Milan, Italy). Human umbilical vein endothelial cells (HUVEC) were purchased from Lonza (Bend, OR, USA) and routinely grown in Endothelial Growth Medium (EGM™-2) as indicated by the provider. The neuroblastoma cell line SH-SY5Y was purchased from Sigma (ECACC 94030304, Sigma) and cultured in EMEM (ECB2071L, Euroclone) and HAM’S F-12 (ECB7502L, Euroclone) 1:1, 10% FBS (ECS0180L, Euroclone), 1% NEAA (ECB3054D, Euroclone), 1% 200 mM glutamine (ECB3000D, Euroclone), 1% Penicillin/Streptomycin (ECB3001D, Euroclone). All cultures were maintained at 37°C in a 5% CO2 incubator.
In vitro cellular models of replicative senescence
We set up the replicative senescence (RS) by prolonged culturing of NHDF and HSMC cells. We considered as young the cells at the sixth passage corresponding to Population Doubling Level (PDL) 20 and as old/senescent the cells at sixteenth passage corresponding to PDL60-65. HSMCs were also used at the sixth passage as young and at sixteenth as old. The senescence condition in the two cellular models was assessed by monitoring the morphological changes and by β-galactosidase (β-gal) staining with the Senescence Cells Histochemical Staining Kit (CS0030, Sigma-Aldrich). After reaching the preferred condition, cells were collected in TRIzol reagent (15596018, Invitrogen) and stored at -80°C.
RNA isolation from human cell lines
Total RNA was isolated from different human cell lines using TRIzol reagent according to the manufacturer’s instructions. Concentration and purity of RNA were measured using the Nanodrop 1000 spectrophotometer (ThermoFisher Scientific). All RNA samples had an A260/280 value of 1.8–2.1. The quality of RNA was also evaluated using the Bioanalyzer 2100 instrument (G2939BA, Agilent) or Tape Station 2200 (Agilent). All the samples had a RIN value ≥ 7.
RT-PCR and Sanger sequencing
RNA isolated from cells (1 μg) was digested using the RQ1 RNase-Free Dnase (M6101, Promega) and reverse-transcribed using the High-Capacity cDNA Reverse Transcription Kit (4368814, Applied Biosystems), according to the manufacturer’s instructions. High-quality RNA from 12 human tissues (bladder, brain, cervix, esophagus, kidney, liver, lung, ovary, skeletal muscle, small intestine, spleen, and testis; First Choice Human Total RNA Survey Panel, AM6000, Ambion) was reverse-transcribed using Maxima H Minus cDNA Synthesis Master Mix with dsDNase kit (M1682; Thermo Fisher Scientific), according to the manufacturer’s protocol.
RT-PCR was performed using the GoTaq G2 DNA Polymerase (M7845, Promega) with 3 μl of reverse transcriptase product and the following primers: for HELLS_XR_007061960 amplicon (811 bp) fw: TTTCAAAGGTGGTCAGTCTGGA; rev: TCCTACCGACAGCCAAGTCT; for HELLS_XR_007061960_3’ amplicon (906 bp) fw: GGCTTTTCATGGGGAGAGATTG; rev: ATCACGAGTTTCAAGCAAGCCTA. PCR fragments were analyzed by gel electrophoresis with 1.8% agarose gels in TBE buffer containing 20000x SYBR Safe (S33102, Invitrogen). 5 μl of PCR products were loaded in each well. The bands were visualized using the Uvitec transilluminator (Cambridge). The two PCR products were directly sequenced in both directions using the Big Dye Terminator Cycle Sequence Ready Reaction Kit v3.1 (4337455, Applied Biosystems) and resolved on a 3130xl Genetic Analyzer (Life Technologies). The electropherograms were analyzed by means of the ChromasPro software (v2.1.5) (Technelysium Pty Ltd).
Quantitative RT-PCR assays
Quantitative RT-PCR (qPCR) was performed with the QuantStudio 5 thermocycler (Applied Biosystems) in 384-well plates using the GoTaq qPCR Master Mix (A6002, Promega). Primers were designed by the Primer-BLAST tool (https://www.ncbi.nlm.nih.gov/tools/primer-blast/index.cgi) (S2 Table). A total of 10 μl PCR reactions were prepared containing 2 μl of reverse transcriptase product and 0.2 μl of each primer (10 μM). The PCR mixtures were incubated at 95°C for 3 min, followed by 40 cycles of 95°C for 10 s and 60°C for 1 m. The results were analyzed with the QuantStudio Design & Analysis Software v1.5.2 (Applied Biosystems). All the experiments were performed at least in triplicate and the calculation of gene expression levels was based on the 2−ΔΔCt method. The geometric mean of the expression values of CYC1, EIF4A2 and RPSA was taken as normalizer gene set; then the fold change obtained as ratio of normalized expression levels in old/senescent cells compared to young cells was calculated. Descriptive statistics, unpaired t-test and regression analysis of qPCR data were carried out by Microsoft Excel and/or by using Prism 8 (GraphPad Software, La Jolla, CA, United States).
Results
In silico analysis reveals the overlapping of SALNR/AK091544 and HELLS isoforms
The SALNR lncRNA was published and characterized by Wu et al. in their 2015 paper [8]. The cDNA clone code corresponding to SALNR is AK091544, located on chromosome 10 (chr10: 94603625–94606511 on GRCh38/hg38), later extended via a 5’-RACE from which they saw it is actually longer at 5’ end (chr10: 94602724–94606511). Since its publication, SALNR has been repeatedly cited among human lncRNAs, but its actual origin and transcriptomic products have not been further confirmed. Currently, the only annotations overlapping this genomic location are those of some transcripts of the HELLS gene (Ensembl ID: ENSG00000119969, Entrez Gene ID: 3070), located upstream on the same strand. Several HELLS isoforms/transcripts have been annotated in genomic databases: specifically, 21 isoforms in Ensembl GENCODE v42 annotation and 11 isoforms in NCBI RefSeq gene annotation release 110. Fig 1 shows the overlap among SALNR/AK091544 and HELLS isoforms. In particular, we found an overlap in the 3’-UTR portion of two Ensembl transcripts (ENST00000698800 and ENST00000419900), in the introns of ENST00000371327 and of ENST00000432120 lncRNA (ENSG00000244332 gene) and in an exon of a predicted RefSeq transcript (XR_007061960). Noteworthy, the RefSeq transcript XR_007061960 was added to HELLS RefSeq isoforms in April 2022 (NCBI Homo Sapiens Annotation Release 110, 2022-04-12) and it completely covers SALNR genomic location. Looking at the chr10: 94603625–94606511 genomic region (GRCh38/hg38) in the LNCipedia database we only retrieved the ENST00000432120 transcript (referred to LNCipedia ID: lnc-TBC1D12-1:1, https://lncipedia.org/db/transcript/lnc-TBC1D12-1:1), but no further annotation was available.
A screenshot from UCSC genome browser showing AK091544 clone (highlighted in cyan, while the 5’-RACE result is highlighted in yellow) and the annotated transcripts of HELLS overlapping or flanking it from Ensembl GENCODE (only 14 out of 21 isoforms are included in this partial view) and NCBI RefSeq. GENCODE color legend: Blue, coding; green, non-coding; red, problem. NCBI RefSeq color legend: Blue, reviewed; lilac, predicted.
RNA-seq data collected from public repositories were mapped against the Ensembl human genome and the results were loaded into the UCSC genome browser (genome release GRCh38/hg38). Short-reads revealed an unexpected read coverage over the 3’ end of HELLS gene, with a considerable number of reads mapped on SALNR/AK09154 region (see some examples in S1 Fig, and see Data availability section for online URL). Furthermore, long-read datasets were used to search for transcribed RNAs overlapping SALNR/AK09154 location as well as to identify the isoforms transcribed in the selected samples (S1 Table).
In general, we found that long-read datasets showed poor coverage of HELLS genomic portion, and its transcripts are too long to be entirely represented, even by long-read sequencing (see Data availability section for online URL of long-read samples). However, as depicted in Fig 2, some interesting results are visible in the samples. Overall we observe that: the reads mapping to the 3’ end of HELLS gene confirm the presence of at least one isoform that includes that portion, and the 3’ end of XR_007061960 and ENST00000371327 is confirmed to be the definitive limit for the transcription of that genomic locus (chr10, position 94613905) (Fig 2.i); the reads including the last intron of XR_007061960 suggest the presence of another isoform (Fig 2.ii and 2.iii) the reads excluding such intron confirm XR_007061960 3’ configuration (Fig 2.iii); the read in SRR12524789 run from SRP278890 confirms the presence of an isoform that extends the 3’ of an HELLS transcript over SALNR location (Fig 2.iv); none of the reads exactly matches the SALNR/AK091544 location starting and ending in its candidate boundaries.
The figure shows 3 representative samples of long-read RNA-seq and their alignment on HELLS genomic location. SALNR/AK091544 location is highlighted in cyan and yellow. Green boxes highlight that (i) reads mapping to the 3’ end of HELLS gene confirm the end of the gene and the transcription of such portion; (ii) reads including the last intron of RefSeq XR_007061960 (see Fig 1) suggest the presence of another isoform and (iii) the reads excluding such intron confirm XR_007061960 3’ configuration; (iv) the read in SRR12524789 run from SRP278890 confirms the presence of an isoforms that extends the 3’ of an HELLS transcript over SALNR/AK091544 location. For the other samples see Data availability section.
RT-PCR confirms the presence of the predicted isoform of HELLS in human cell lines and tissues
Working with HELLS gene can be challenging because of its low expression in human tissues as depicted by the Human Protein Atlas (https://www.proteinatlas.org/ENSG00000119969-HELLS). Moreover, as shown by the GTEx database, the data availability of the expression levels of HELLS isoforms in human samples is limited to few of them (https://gtexportal.org/home/gene/HELLS). Therefore, to confirm the existence of the predicted HELLS isoform that completely overlaps the SALNR/AK091544 sequence, we designed primers that specifically amplify the RefSeq XR_007061960 predicted isoform (HELLS_XR_007061960 and HELLS_XR_007061960_3’) (S2 Fig) and carried out reverse-transcription PCR (RT-PCR) on RNA extracted from several human cell lines (NHDF, HSMC, HUVEC, SH-SY5Y) and from 12 human tissues (see methods section). The semi-quantitative RT-PCR analysis yielded in all cases a specific amplification product of the expected size (Fig 3). Sanger sequencing carried out on two representative RT-PCR products obtained from HSMC and SH-SY5Y cells confirmed the correspondence between the sequence of the amplicons and the one of the predicted isoform (S1 File).
The figure shows the gel electrophoresis analysis of RT-PCR products specific for the HELLS XR_007061960 isoform in human tissues and cell lines. A, scheme of the primers designed to amplify the HELLS XR_007061960 isoform and their genomic location in HELLS locus (see methods); B-C: HELLS_XR_007061960 amplicon, 811 bp; D-E: HELLS_XR_007061960_3’ amplicon, 906 bp. PCR products were analyzed through a 1,8% gel in TBE buffer. Panel B, D: M. MW Marker, 1. Bladder, 2. Brain, 3. Cervix, 4. Esophagus, 5. Kidney, 6. Liver, 7. Lung, 8. Ovary, 9. Skeletal muscle, 10. Small intestine, 11. Spleen, 12. Testis. Panel C,E: M. MW Marker, 1. SH-SY5Y, 2. HSMC, 3. NHDF, 4. HUVEC.
In silico differential expression analysis reveals under-expression of HELLS transcripts in senescence cellular models
Wu et al. observed that SALNR/AK091544 amplicon shows a reduced expression in cellular senescence [8]. Since SALNR/AK091544 genomic location is overlapped to the actual annotations of HELLS gene, we have performed a differential expression analysis of the RNA-seq reads expressed by the genomic location that included HELLS and SALNR/AK091544, on the datasets listed in Table 1, representing cellular models of replicative senescence. Since Ensembl and RefSeq annotations have a few different transcripts, we defined a single annotation for HELLS by manually adding the XR_007061960 sequence to the Ensembl GTF annotation file. SALNR/AK091544 is absent in both the annotations and we decided to not consider it in this step. Firstly, we mapped the dataset with STAR and estimated HELLS expression (at the gene level) with RSEM. Then, we performed differential expression analysis with DESeq2 (Table 2), observing a decreased read count in senescent cells for each dataset. The statistical analysis resulted in a fairly reproducible down-regulation of HELLS at gene level, and in four out of eight datasets the Log2FC values reached the statistical significance (Table 2).
At this point, to isolate the source of under-expression events, we tried to distinguish the expression of the different HELLS isoforms. We analyzed a total of 23 isoforms: 21 Ensembl HELLS isoforms, ENSG00000244332/ENST00000432120 and RefSeq XR_007061960. Specifically, the SALNR genomic locus overlapped three isoforms of HELLS. Due to the evident overlap between transcripts, RSEM failed in distributing the reads among the different isoforms, because some of them turned out to have null expression even though there are reads mapping to their transcripts. MultiDEA provided a more detailed picture of the ambiguity in the mapping, and the only isoforms that achieve a significant number of uniquely mapping reads (>20) are XR_007061960 and ENST00000419900 (S3 Table). Moreover, the differential expression analysis performed on both RSEM and MultiDEA isoform expression estimations showed that under-expression events are not limited to those HELLS isoforms overlapping with SALNR genomic area (i.e., on ENST00000371327, ENST00000698800, ENST00000419900 and XR_007061960) but they involve all HELLS transcripts (S4 Table). This result contributes to question the existence of SALNR as an independent gene and to confirm the involvement of HELLS in senescence.
qPCR confirms the down-regulated expression of HELLS in senescence cellular models
To verify the results obtained from in silico analyses, we evaluated the expression levels of HELLS during senescence by qPCR. We exploited two different senescence cellular models, NHDF and HSMC cells in replicative senescence, i.e., serially cultured until replicative exhaustion (passage p16). The senescent condition was validated by morphological changes (cellular enlargement and flattening with a concomitant increase in the size of the nucleus) and positive β-gal staining (S3 Fig). In addition, we confirmed their senescent status by detecting the gene expression levels of well-known markers of senescence in the old cells as compared to the young cells by qPCR. Overall, we observed in both cell lines the up-regulation of CDKN1A/p21, CXCL8 and MMP3, and the down-regulation of HGMB1 and LMNB1 in the old cells compared to the young cells (Table 3), consistently with their senescent status.
(afold change obtained as ratio of expression levels in old senescent cells compared to young cells; bSALNR amplicon overlapped with the XR_007061960 HELLS isoform).
Focusing on HELLS/SALNR expression, we designed two set of qPCR primers (S2 Table): the first one was able to amplify a consensus region shared by all HELLS isoforms, while the second one was designed to amplify the region comprising the SALNR/AK091544 amplicon overlapping the XR_007061960 HELLS isoform (for abbreviation “SALNR” amplicon). The results of the qPCR showed a strong down-regulation of HELLS gene expression levels in the senescent cells compared to the young proliferating cells, thus confirming the in silico analysis. Similarly, the levels of SALNR were down-modulated in all the senescence cellular models (Table 3).
As expected, we found a strong correlation with a Pearson coefficient of 0.947 (Fig 4) between the expression levels measured with the qPCR amplicon amplifying the consensus region of major HELLS isoforms and those obtained with the SALNR amplicon overlapping the predicted XR_007061960 HELLS isoform (Table 3).
Linear regression of gene expression levels (FC) of HELLS and SALNR/XR_007061960 HELLS amplicons of all senescent cells.
Discussion
In this investigation, we applied an extensive in silico analysis of several publicly available RNA-seq datasets, including short- and long-read experiments, with the aim of disentangling the mystery of SALNR. We demonstrated that SALNR/AK091544 transcript is completely overlapped to a predicted HELLS transcriptional isoform and we did not find evidence of SALNR/AK091544 existence as independent transcript. By wet approaches (RT-PCR and Sanger sequencing), we confirmed the presence of this predicted HELLS isoform in human cells and tissues of different origin. Moreover, by re-analysis of RNA-seq datasets as well as by qPCR, we assessed the down-regulation of HELLS gene in old/senescent compared to young/proliferative human cells.
SALNR was first described by Wu et al. [8], who performed microarray experiments aimed at identifying deregulated lncRNAs during replicative senescence of lung fibroblasts and pointed to SALNR/AK091544 as a consistently down-regulated lncRNA in replicative senescence and in other cellular models of induced senescence. In our study, thanks to the availability of several publicly available RNA-seq datasets, we could deeply investigate the HELLS genomic region. Even if long-read datasets showed poor coverage of HELLS genomic portion, we found the presence of enough reads enabling a deep analysis. Our data are not necessarily at odds with the results shown by Wu and colleagues about SALNR. According to our results, two main hypotheses can be advanced. The first one supposes that SALNR does not exist as an independent transcriptional unit but it is a portion of the XR_007061960 HELLS isoform. Under this hypothesis, XR_007061960 should be considered as an HELLS sense-overlap lncRNA and the findings obtained by Wu and colleagues about SALNR could instead be attributed to the region of the XR_007061960 HELLS isoform overlapping SALNR rather than to the potential SALNR non-coding transcript. Failure to amplify the very long XR_007061960 transcript by RACE experiments could be caused by the intrinsic limitations of these techniques. For instance, 5’-RACE often yields truncated products when trying to obtain the amplification of regions longer than 1 Kb, due for example to the presence of secondary structures in the transcript that cause the detachment of reverse transcriptase. Of note, the XR_007061960 HELLS isoform overlapping SALNR/AK091544 was only added in the latest human genome annotation release of April 2022 and the RNA-seq datasets providing evidence on the existence of this isoform were all made available after the publication of Wu et al. paper. The second hypothesis is that both SALNR and the XR_007061960 HELLS isoform exist. In this case, the failure to detect SALNR transcripts could be caused by several reasons. Many lncRNAs may be particularly difficult to detect by analysis of steady-state RNA levels because they are expressed at very low levels and they are highly unstable. An example is provided by lncRNAs generated through premature transcription termination of protein-coding genes that are not polyadenylated and rapidly degraded [23]. Moreover, they often display a restricted tissue-specific or cell-specific expression pattern. In our study, we could not evaluate the existence of the XR_007061960 HELLS isoform in the cell line used by Wu and colleagues as model for cellular senescence. But despite the limitations discussed above, at least in the context of the experimental conditions adopted in our study, our findings do not support the existence of SALNR as a separate, independent transcript.
By in silico re-analysis of eight RNA-seq datasets (see Table 1) associated to transcriptomic investigation of replicative senescence in different human cells such as fibroblasts, smooth muscle cells, and endothelial cells, we found that the normalized read counts of HELLS were significantly down-regulated in senescent cells compared to young cells. In the same direction, by using in vitro cellular models of replicative senescence, we confirmed the down-regulation of HELLS in senescent NHDFs, as well as in HSMCs, compared to proliferating/young cells. What about the biological function of HELLS? Does HELLS play any role in senescence? HELLS gene (also known as LSH, PASG, SMARCA6) encodes for a helicase of 838 amino acids (Uniprot ID: Q9NRZ9) belonging to the SNF2 family of proteins and playing an essential role in normal development and involved in the regulation of lymphoid lineage expansion and survival. It is a chromatin remodeling protein with a role in the epigenetic regulation acting at different levels, including remodeling of nucleosomes, histone modifications, and de novo or maintenance DNA methylation [24]. It has also been shown to have a role in double strand DNA repair [25]. It may play a role in the formation and organization of heterochromatin, implying a functional role in the regulation of transcription and mitosis [26]. Noteworthy, the relevance of epigenetic regulation in aging and senescence has increased by time [27]. Several pieces of evidence suggest the involvement of HELLS in senescence and aging: indeed, its expression is transcriptionally repressed by p53 via p21 and down-regulated during senescence [28]. Zhou et al. demonstrated that high HELLS/Lsh expression in young cells represses p16INK4a expression, thereby maintaining the proliferative state of 2BS cells, while loss of Lsh expression may lead to increased p16INK4a expression in senescent 2BS cells, contributing to the onset of cellular senescence [29,30]. In the same line, Hells knock-out mice display premature aging and the fibroblasts derived from these mice show a senescent phenotype [31]. Moreover, HELLS has been annotated in the database of genes related to aging, GenAge, with potential relevance to the human aging process.
Conclusion
Overall, in this study, we demonstrated that SALNR/AK091544 sequence completely overlaps HELLS transcriptional isoforms; moreover, in the context of the experimental conditions adopted, it was not possible to validate the existence of SALNR as a separate, independent transcript in the human genome. This study indicates caution when working with genomic loci hosting lncRNA transcripts and suggests to adopt a deep assessment by in silico approaches before proceeding with further investigation on a specific transcript. Note that our investigation has revealed a coherent and fairly reproducible down-regulation of the HELLS gene in human cell models undergoing replicative senescence. Recently, the role of epigenetics in promoting senescence has been highlighted. In this scenario, the function of HELLS as an epigenetic regulator of senescence deserves further exploration.
Supporting information
S1 Fig. Short-read RNA-seq.
Visualization of sequence reads (from short-read RNA-seq) aligned to the genomic portion including HELLS and SALNR, using UCSC genome browser. It is not possible to isolate SALNR expression from this type of reads, but the heavy presence of reads covering the 3’ end of the HELLS gene suggests an interesting production of known and novel isoforms in that location. This figure represents the sample with highest read count for each dataset (from proliferative cells). The GSE157867 track (the dataset with 2x150bp reads) is plotted in “squish” display mode, while the others are plotted in “dense” mode. The short-read RNA-seq mapping tracks can be visualized on UCSC genome browser: https://genome.ucsc.edu/s/cnr.itb.ba/short-read_RNA-seq. See Data Availability section.
https://doi.org/10.1371/journal.pone.0286104.s001
(TIF)
S2 Fig. XR_007061960 primers.
Genomic location of the two amplicons used to amplify the XR_007061960 isoform of HELLS and corresponding primer sequences.
https://doi.org/10.1371/journal.pone.0286104.s002
(TIF)
S3 Fig. β-gal staining.
Representative images of colorimetric β-gal assay in young (p6) (A) and old (p16) (B) HMSCs, and in young (p6) (C) and old (p16) (D) NHDFs (x10 magnification).
https://doi.org/10.1371/journal.pone.0286104.s003
(TIF)
S1 Table. Long-read RNA-seq dataset.
List of all long-read datasets that were mapped to the human genome to search for reads mapping to HELLS genomic location to identify the isoform that originates them.
https://doi.org/10.1371/journal.pone.0286104.s004
(XLSX)
S2 Table. qPCR primers.
List of primers used for quantitative RT-PCR experiments.
https://doi.org/10.1371/journal.pone.0286104.s005
(XLSX)
S3 Table. RSEM and MultiDEA read count estimations.
Read count estimation computed by RSEM and MultiDEA for each isoform in HELLS genomic location, for each sample and for each dataset. Despite the diffuse overlapping among the isoforms, the Bayesian model included in RSEM tends to prefer only a small subset of isoforms, changing in each dataset. MultiDEA clearly shows that the only isoforms to achieve a significant number of uniquely mapping reads (>20) are XR_007061960 and ENST00000419900, while all the other reads map to more than one reference transcript. RSEM and MultiDEA columns list the isoform expression estimation computed by each tool; [A,B,C,D] is the multiread detail computed by MultiDEA (trapezoid vertices, A = uniquely mapping reads, B = reads having the isoform as unique best match; C = reads having that isoform as non-unique best match; D is the total number of reads that map to that isoform). Isoforms are sorted by start position on chr10; XR_007061960 is an additional isoform of HELLS proposed by RefSeq and ENST00000432120 is a lncRNA that overlaps HELLS.
https://doi.org/10.1371/journal.pone.0286104.s006
(XLSX)
S4 Table. DESeq2 on RSEM and MultiDEA expressions.
Results of DESeq2 differential expression analysis performed on RSEM and MultiDEA isoform expression estimation. The isoforms are sorted by start position on chr10, and even if RSEM and MultiDEA show some different results, it is possible to deduce that the senescence expression decrease influences both the transcript overlapping SALNR and the ones not overlapping it. Statistically significant count is the number of times that the isoform resulted significantly down-regulated in the 8 tested datasets; XR_007061960 is an additional isoform of HELLS proposed by RefSeq and ENST00000432120 is a lncRNA that overlaps HELLS; a result is considered statistically significant (*) if mean read count >20, absolute log2 Fold Change >1 and adjusted p-value (FDR) <0.05.
https://doi.org/10.1371/journal.pone.0286104.s007
(XLSX)
S1 File. Nucleotide sequences in FASTA format.
Results of Sanger sequencing carried out on RT-PCR products (HELLS_XR_007061960 and HELLS_XR_007061960_3’ amplicons) obtained from HSMC and SH-SY5Y cells exported as sequences in FASTA format.
https://doi.org/10.1371/journal.pone.0286104.s008
(FASTA)
Acknowledgments
We thank Paride Pelucchi, Elena Battaglioli, Emanuela Toffolo, Lisa Giovannelli for their help with setting cell culture models and for providing technical support. We thank Nicola Losito for maintaining the ICT infrastructure for bioinformatic analysis. SB is recipient of a scholarship of PhD in Experimental Medicine, Università degli Studi di Milano. CR is recipient of a scholarship of PhD in Pharmacological Biomolecular Sciences, Experimental and Clinical, Università degli Studi di Milano.
References
- 1. Niccoli T, Partridge L. Ageing as a risk factor for disease. Vol. 22, Current Biology. 2012. pmid:22975005
- 2. López-Otín C, Blasco MA, Partridge L, Serrano M, Kroemer G. The hallmarks of aging. Vol. 153, Cell. 2013. pmid:23746838
- 3. Hernandez-Segura A, Nehme J, Demaria M. Hallmarks of Cellular Senescence. Vol. 28, Trends in Cell Biology. 2018.
- 4. Martínez-Cué C, Rueda N. Cellular Senescence in Neurodegenerative Diseases. Vol. 14, Frontiers in Cellular Neuroscience. 2020. pmid:32116562
- 5. Hayflick L, Moorhead PS. The serial cultivation of human diploid cell strains. Exp Cell Res. 1961;25(3). pmid:13905658
- 6. Montes M, Lund AH. Emerging roles of lncRNAs in senescence. FEBS Journal. 2016. pmid:26866709
- 7. Puvvula PK. Lncrnas regulatory networks in cellular senescence. Vol. 20, International Journal of Molecular Sciences. 2019. pmid:31141943
- 8. Wu CL, Wang Y, Jin B, Chen H, Xie BS, Mao Z Bin. Senescence-associated long non-coding RNA (SALNR) delays oncogene-induced senescence through NF90 regulation. J Biol Chem. 2015;290(50). pmid:26491010
- 9. Logsdon GA, Vollger MR, Eichler EE. Long-read human genome sequencing and its applications. Vol. 21, Nature Reviews Genetics. 2020. pmid:32504078
- 10. Volders PJ, Anckaert J, Verheggen K, Nuytens J, Martens L, Mestdagh P, et al. Lncipedia 5: Towards a reference set of human long non-coding rnas. Nucleic Acids Res. 2019;47(D1). pmid:30371849
- 11. Suda M, Shimizu I, Katsuumi G, Yoshida Y, Hayashi Y, Ikegami R, et al. Senolytic vaccination improves normal and pathological age-related phenotypes and increases lifespan in progeroid mice. Nat Aging. 2021;1(12). pmid:37117524
- 12. Zhang C, Zhang X, Huang L, Guan Y, Huang X, Tian XL, et al. ATF3 drives senescence by reconstructing accessible chromatin profiles. Aging Cell. 2021;20(3). pmid:33539668
- 13. Ohori M, Nakayama Y, Ogasawara-Shimizu M, Toyoshiba H, Nakanishi A, Aparicio S, et al. Gene regulatory network analysis defines transcriptome landscape with alternative splicing of human umbilical vein endothelial cells during replicative senescence. BMC Genomics [Internet]. 2021; Available from: pmid:34856941
- 14. Uryga AK, Grootaert MOJ, Garrido AM, Oc S, Foote K, Chappell J, et al. Telomere damage promotes vascular smooth muscle cell senescence and immune cell recruitment after vessel injury. Commun Biol. 2021;4(1). pmid:34021256
- 15. Marthandan S, Priebe S, Baumgart M, Groth M, Cellerino A, Guthke R, et al. Similarities in Gene Expression Profiles during In Vitro Aging of Primary Human Embryonic Lung and Foreskin Fibroblasts. Biomed Res Int [Internet]. 2015;2015:1–17. Available from: http://www.hindawi.com/journals/bmri/2015/731938/. pmid:26339636
- 16. Marthandan S, Baumgart M, Priebe S, Groth M, Schaer J, Kaether C, et al. Conserved Senescence Associated Genes and Pathways in Primary Human Fibroblasts Detected by RNA-Seq. Esteban FJ, editor. PLoS One [Internet]. 2016 May 3 [cited 2020 Aug 3];11(5):e0154531. Available from: https://dx.plos.org/10.1371/journal.pone.0154531. pmid:27140416
- 17.
Andrews S. FastQC—A quality control tool for high throughput sequence data. http://www.bioinformatics.babraham.ac.uk/projects/fastqc/. Babraham Bioinforma. 2010.
- 18. Dobin A, Davis CA, Schlesinger F, Drenkow J, Zaleski C, Jha S, et al. STAR: Ultrafast universal RNA-seq aligner. Bioinformatics. 2013;29(1). pmid:23104886
- 19. Li B, Dewey CN. RSEM: Accurate transcript quantification from RNA-Seq data with or without a reference genome. BMC Bioinformatics. 2011;12.
- 20. Consiglio A, Mencar C, Grillo G, Marzano F, Caratozzolo MF, Liuni S. A fuzzy method for RNA-Seq differential expression analysis in presence of multireads. BMC Bioinformatics. 2016;17. pmid:28185579
- 21. Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15(12). pmid:25516281
- 22. Liu B, Liu Y, Li J, Guo H, Zang T, Wang Y. deSALT: fast and accurate long transcriptomic read alignment with de Bruijn graph-based index. Available from: pmid:31842925
- 23. Nojima T, Proudfoot NJ. Mechanisms of lncRNA biogenesis as revealed by nascent transcriptomics. Nat Rev Mol Cell Biol [Internet]. 2022 Jun 25 [cited 2023 Mar 17];23(6):389–406. Available from: www.nature.com/nrm. pmid:35079163
- 24. Peixoto E, Khan A, Lewis ZA, Contreras-Galindo R, Czaja W. The Chromatin Remodeler HELLS: A New Regulator in DNA Repair, Genome Maintenance, and Cancer. 2022 [cited 2023 Jan 26]; pmid:36012581
- 25. Chen X, Li Y, Rubio K, Deng B, Li Y, Tang Q, et al. Lymphoid-specific helicase in epigenetics, DNA repair and cancer. Vol. 126, British Journal of Cancer. 2022. pmid:34493821
- 26. Muegge K. Lsh, a guardian of heterochromatin at repeat elements. In: Biochemistry and Cell Biology. 2005. pmid:16094458
- 27. Cheng LQ, Zhang ZQ, Chen HZ, Liu DP. Epigenetic regulation in cell senescence. Vol. 95, Journal of Molecular Medicine. 2017. pmid:28887637
- 28. Schuller S, Sieker J, Riemenschneider P, Köhler B, Drucker E, Weiler SME, et al. HELLS Is Negatively Regulated by Wild-Type P53 in Liver Cancer by a Mechanism Involving P21 and FOXM1. Cancers (Basel). 2022;14(2). pmid:35053620
- 29. Niu J, Chen T, Han L, Wang P, Li N, Tong T. Transcriptional activation of the senescence regulator Lsh by E2F1. Mech Ageing Dev. 2011;132(4). pmid:21453717
- 30. Zhou R, Han L, Li G, Tong T. Senescence delay and repression of p16INK4a by Lsh via recruitment of histone deacetylases in human diploid fibroblasts. Nucleic Acids Res. 2009;37(15). pmid:19561196
- 31. Sun LQ, Lee DW, Zhang Q, Xiao W, Raabe EH, Meeker A, et al. Growth retardation and premature aging phenotypes in mice with disruption of the SNF2-like gene, PASG. Genes Dev. 2004;18(9). pmid:15105378