Abstract

Background

   Long non-coding RNAs (lncRNAs) are importantly involved in the
   initiation and progression of non-small cell lung cancer (NSCLC).
   However, the classification and mechanisms of lncRNAs remain largely
   elusive.

Aim

   Hence, we addressed this through bioinformatics analysis.

Methods and results

   We utilized microarray technology to analyze lncRNAs and mRNAs in
   twenty paired NSCLC tumor tissues and adjacent normal tissues. Gene set
   enrichment analysis, Kyoto Encyclopedia of Genes and Genomes, and Gene
   Ontology were conducted to discern the biological functions of
   identified differentially expressed transcripts. Additionally, networks
   of lncRNA-mRNA co-expression, including cis-regulation,
   lncRNA-transcription factor (TF)-mRNA, trans-regulation, and
   lncRNA-miRNA-mRNA interactions were explored. Furthermore, the study
   examined differentially expressed transcripts and their prognostic
   values in a large RNA-seq dataset of 1016 NSCLC tumors and normal
   tissues extracted from the Cancer Genome Atlas (TCGA). The analysis
   revealed 391 lncRNAs and 344 mRNAs with differential expression in
   NSCLC tumor tissues compared to adjacent normal tissues. Subsequently,
   43,557 co-expressed lncRNA-mRNA pairs were identified, including 27
   lncRNA-mRNA pairs in cis, 9 lncRNA-TF-mRNA networks, 34 lncRNA-mRNA
   pairs in trans, and 8701 lncRNA-miRNA-mRNA competing endogenous RNA
   (ceRNA) networks. Notably, these lncRNAs were found to be involved in
   immune-related pathways. Six significant transcripts, including NTF4,
   PTPRD-AS, ITGA11, HID1-AS1, RASGRF2-AS1, and TBX2-AS1, were identified
   within the ceRNA network and trans-regulation.

Conclusion

   This study brings important insights into the regulatory roles of
   lncRNAs in NSCLC, providing a fresh perspective on lncRNA research in
   tumor biology.

   Keywords: Non-small cell lung cancer, Long non-coding RNAs,
   cis-Regulation, trans-Regulation, Tumorigenesis

Highlights

     * •
       Genome-wide profiling of long non-coding RNAs (lncRNAs) in NSCLC
       datasets.
     * •Cis- and trans-lncRNAs were identified in NSCLC.
     * •Both cis- and trans-lncRNAs regulate key immune-related pathways
       in NSCLC.

1. Introduction

   Lung cancer is one of the most prevalent cancer types and is a leading
   cause of cancer-related mortality [[35]1]. In 2022, lung cancer leads
   to approximately 870,982 new cases and 505,618 deaths in China [[36]2].
   Over 80% of lung cancer have been diagnosed as non-small cell lung
   cancer (NSCLC), representing a significant and serious subtype of the
   disease [[37]3]. Multiple treatment strategies have been developed for
   cancer management, aimed at enhancing the five-year survival (OS) rates
   of patients with non-small cell lung cancer (NSCLC) over the past
   decades [[38][4], [39][5], [40][6]]. Despite advancements in treatment
   strategies, the five-year overall survival (OS) rate for non-small cell
   lung cancer (NSCLC) is still low, ranging from 10% to 20% according to
   the Global Surveillance of Trends in Cancer Survival 2000–2014
   (CONCORD-3) [[41]7]. These findings underscore the need to explore
   additional diagnostic and therapeutic biomarkers for NSCLC, thereby
   significantly enhancing the clinical outcomes for patients with this
   disease.

   In recent years, numerous studies highlighted the pivotal roles of
   lncRNAs in cancer, demonstrating their regulation of tumor initiation
   and progression, and their correlation with the prognosis and survival
   outcomes of patients [[42][8], [43][9], [44][10]]. Especially, lncRNA
   pro-transition associated RNA (PTAR) can be an oncogene in NSCLC by
   increasing cell proliferation, migration, and invasion [[45]11]. While
   lncRNAs have been implicated in various oncogenic processes,
   fetal-lethal non-coding developmental regulatory RNA (FENDRR) acts as a
   tumor suppressor in NSCLC, inhibiting tumor cell growth, migration, and
   invasion [[46]12]. Although numerous lncRNAs exert vital roles in
   NSCLC, the focus regarding attention on the mechanism of action is
   their RNA products in the past years. Recent research indicates that
   not only the products of RNA species exert important roles, but also
   the lncRNA locus separately exerts regulatory functions
   [[47]13,[48]14].

   About 10%–20% of human transcripts are identified as protein-coding
   RNAs, and 80%–90% of transcripts are non-protein-coding RNAs
   (non-coding RNAs) [[49]15]. LncRNAs, a subclass of non-coding RNAs
   characterized by transcripts longer than 200 nucleotides, do not encode
   proteins but possess extensive regulatory capabilities, influencing a
   broad spectrum of cellular processes [[50]16,[51]17]. LncRNAs are
   heterogeneous molecules in various cancers due to their diverse locus
   and function [[52]18]. LncRNAs are broadly classified into different
   categories according to relative genomic position to the protein-coding
   RNAs, including sense, antisense, intronic, intergenic, and
   bidirectional [[53]16,[54]19,[55]20]. Indeed, lncRNAs play multifaceted
   roles in both cis and trans regulation, organization of nuclear, and
   gene modulation, sometimes even encoding small proteins themselves
   [[56][21], [57][22], [58][23]]. It has been challenging to understand
   the biological function of several lncRNAs in regulating biological and
   pathological processes mediated by the lncRNAs themselves. However, to
   date, our understanding of the expression and regulatory mechanisms of
   lncRNAs in NSCLC remains limited. In recent years, advancements in
   next-generation sequencing (NGS) have enabled the confirmation of
   numerous biomarkers and underlying mechanisms in lung cancer through
   bioinformatics analysis [[59][24], [60][25], [61][26], [62][27]].

   This study aims to identify the landscape of lncRNA expression and
   explore the function of lncRNAs in NSCLC using bioinformatics analysis.
   We identified lncRNAs and protein-coding transcripts that showed
   differential expressions based on microarray analysis. The cis- and
   trans-regulatory lncRNAs and their regulated biological pathways were
   then explored in NSCLC. Importantly, TBX2-AS1, NTF4, PTPRD-AS, ITGA11,
   RASGRF2-AS1, and HID1-AS1 were identified as potential therapeutic
   biomarkers for NSCLC.

2. Material and methods

2.1. Specimen

   Three tumor samples and normal tissues were collected from patients
   diagnosed with NSCLC. The cohort consisted of two females and one male,
   with a median age of 64 years (50–70). All patients have signed
   informed consent for their inclusion in this study. Importantly,
   samples were procured before any pre-operative radiation or
   chemotherapy treatments. After the pathological examination by three
   pathologists, the samples were saved in liquid nitrogen. Our research
   has been supported by the Ethics Committee of the First Hospital of
   Kunming Medical University.

2.2. RNA extraction, cDNA library preparation, and microarray analysis

   Total RNA was isolated from fresh NSCLC tumor tissues and normal
   tissues using TRzol reagent (Thermo Fisher Scientific, Waltham, MA,
   USA) following the manufacturer's protocol. The concentration and
   quality of RNA were assessed using the NanoDrop ND-2000
   spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA), and the
   integrity of RNA was evaluated using an Agilent Bioanalyzer 2100
   (Agilent Technologies, Santa Clara, CA, USA) according to the
   manufacturer's instructions. Subsequently, purified single-strand RNA
   was reverse transcribed to generate double-strand cDNA and labeled with
   Cyanine 3-CTP (Cy3). The Cy3-labeled cDNA was then hybridized onto the
   Agilent Human ceRNA Microarray 2019 (4∗180k, Design ID:086188).
   Finally, the hybridized slides were scanned using the G2505C Agilent
   Scanner (Agilent Technologies).

2.3. Quality control and data standardization

   After screening the original images of slides, the raw data was
   extracted using the Feature Extraction software (version 10.7.1.1,
   Agilent Technologies). Subsequently, the raw data were normalized and
   analyzed using the R software package. The coefficient of variable (CV)
   value of each sample was used to the stability of the signal detected
   by the chip repeat probe. The correlation between samples was used to
   measure the correlation between these samples. Principal component
   analysis (PCA) was used to reveal the distribution of the main
   information of each sample.

2.4. Differential analysis for lncRNAs and mRNAs

   LncRNAs and mRNAs with different expressions were identified using the
   student's t-test [[63]28], the threshold adjusted as log2(Fold change,
   FC)≥2 and P ≤ 0.05. The volcano plots and heatmaps were drawn to show
   the significantly differential expressed lncRNAs and mRNAs.

2.5. Biofunction analysis of DE-mRNAs

   Gene Ontology (GO) analysis [[64]29,[65]30], encompassing biological
   process (BP), cellular component (CC), and molecular function (MF), was
   conducted using bioinformatics tools available at OECloud
   ([66]https://cloud.oebiotech.cn). The significance of each GO term was
   assessed utilizing the Hypergeometric distribution algorithm, as
   previously described [[67]31]. KEGG pathway enrichment analysis was
   carried out using the Hypergeometric distribution algorithm based on
   the KEGG database [[68]32].

2.6. Correlation analysis for co-expression lncRNA-mRNA pairs

   The correlation between lncRNAs and mRNAs with differential expressions
   (DE-mRNAs) was evaluated using the Pearson correlation analysis.
   LncRNA-mRNA pairs were revealed based on a correlation coefficient
   greater than 0.8 and P < 0.05. The lncRNA-mRNA pairs were visualized
   using the Circos tool [[69]32].

2.7. Cis- and trans-regulation analysis

   Furthermore, we explored the regulatory manners of lncRNAs according to
   the distances between the protein-coding transcripts and lncRNAs.
   First, we investigated the potential cis-regulation of lncRNAs on their
   neighboring protein-coding transcripts, focusing on those located
   within a 100 kb distance from the lncRNA locus. The neighboring
   protein-coding transcripts of lncRNAs were screened using the FEELnc
   tool [[70]33], and the cis-regulatory lncRNA-mRNA pairs were identified
   by overlapping with the above lncRNA-mRNA pairs. Second, we explored
   the trans-regulatory lncRNAs on protein-coding transcripts using
   RIsearch-2.0 according to the filter criteria, the direct binding sites
   of lncRNAs sequences on protein-coding transcript sequences more than
   10 bases, and the binding free emery less than 100 [[71]34]. The
   trans-regulatory lncRNA-mRNA pairs were screened by overlapping with
   the above lncRNA-mRNA pairs and visualized using a network tool
   [[72]35]. Third, lncRNAs recruited TFs to the specific DNA sequence
   region (such as promoters) to initiate transcriptional activity, or
   multiple TFs bind to a lncRNA to regulate molecular mechanisms. We
   analyzed transcription factor (TF) binding motifs within the regulatory
   regions of lncRNAs, specifically within a range of ± 1 kb from the
   lncRNA start sites. Potential TFs binding to lncRNAs were predicted
   using JASPAR [[73]36]. We constructed the lncRNA-TF-mRNA ternary
   regulatory network based on data from the Gene Transcription Regulation
   Database (GTRD) [[74]37] and the above lncRNA-mRNA pairs, and the
   network was visualized network tool [[75]35].

2.8. Pathway and network enrichment analysis

   The protein-coding transcripts of the cis-, trans-regulatory
   lncRNA-mRNA pairs, and ceRNA network were annotated and enriched via
   the clusterProfiler package.

2.9. TCGA data analysis

   The RNA-seq dataset was obtained from 1016 NSCLC tumors and normal
   tissues from the Cancer Genome Atlas (TCGA) database, accessible at
   [76]https://www.cancer.gov/about-nci/organization/ccg/research/structur
   al-genomics/tcga. This dataset includes 515 samples of TCGA-LUAD (lung
   adenocarcinoma) and 501 samples of TCGA-LUSC (lung squamous cell
   carcinoma). The relationship between significant transcripts and
   clinical phenotypes, including age, gender, stages, TNM, lymph node
   count information, and the relationship between expression of the
   transcripts and survival were investigated.

3. Results

3.1. Identification of the DE-lncRNAs and DE-mRNAs in NSCLC

   In this study, a Microarray analysis was carried out to investigate the
   transcriptome alteration in NSLCLC. After quality control and data
   standardization ([77]Fig. S1), 391 lncRNAs were identified, which were
   preferably distributed in Chr1, Chr2, Chr6, Chr10, Chr11, and Chr17
   ([78]Fig. S2A). More than 70 % of lncRNAs are longer than 500 bp
   ([79]Fig. S2B). We found that the highest percentage of lncRNAs located
   in 1–3 exons ([80]Fig. S2C). The types of lncRNAs were identified as
   exon type (27.55 %), intron type (23.14 %), and intergenic region type
   (49.31 %) ([81]Figs. S2D–E). Furthermore, 304 DE-lncRNAs (145
   upregulated and 159 downregulated) were identified [log2(FC) > 2 and
   P < 0.05] ([82]Fig. 1A–C, [83]Table S1). Similarly, a total of 344
   DE-mRNAs (232 upregulated and 112 downregulated) were identified
   [log2(FC) > 2 and P < 0.05] ([84]Fig. 1D–F, [85]Table S2) in NSCLC
   tumor tissues compared to that in normal tissues.

Fig. 1.

   [86]Fig. 1
   [87]Open in a new tab

   Identification of the DE-lncRNAs and DE-mRNAs in NSCLC. (A). The
   volcano plots illustrated the different distributions of the
   DE-lncRNAs. Red points represented upregulated lncRNAs. Point
   represented downregulated lncRNAs. (B). The scatter plots exhibited the
   expression levels of DE-lncRNAs between NSCLC tumor and adjacent normal
   tissues. Red points represented upregulated lncRNAs. Point represented
   downregulated lncRNAs. (C). Heatmap of the unsupervised hierarchical
   clustering showing the distinction of DE-lncRNAs between NSCLC tumor
   and adjacent normal tissues. (D). The volcano plots exhibited the
   distribution of DE-mRNAs. Red points represented upregulated mRNAs.
   Point represented downregulated mRNAs. (E). The scatter plots exhibited
   the expression levels of DE-mRNAs between NSCLC tumor and adjacent
   normal tissues. Red points represented upregulated mRNAs. Point
   represented downregulated mRNAs. (F). Heatmap of the unsupervised
   hierarchical clustering showing the distinction of DE-mRNAs between
   NSCLC tumor and adjacent normal tissues. (For interpretation of the
   references to color in this figure legend, the reader is referred to