Abstract

Simple Summary

   Lung adenocarcinoma is the main pathological type of lung cancer with a
   very low 5-year survival rate. In the present study, through analyzing
   the mRNA, miRNA, and DNA methylation sequencing data from TCGA,
   combined with the downloaded clinical data, we accidentally found that
   both methylation and gene expression of some genes were up-regulated or
   down-regulated, which is in contradiction with our general
   understanding. To probe the mechanism, we selected MNDA associated with
   lung cancer prognosis for further analysis. The results showed that the
   imbalance of methylase and demethylase resulted in the demethylation of
   MNDA. Moreover, the expression of SPI1, the main transcription factor
   of MNDA, was down-regulated, while the two miRNAs hsa-miR-33a-5p and
   hsa-miR-33b-5p, which directly targeted MNDA, were up-regulated, thus
   inhibiting the expression of MNDA. In conclusion, the abnormal
   expression of MNDA in lung cancer is the result of the combined effects
   of transcriptional and post-transcriptional regulation.

Abstract

   Lung adenocarcinoma (LA) is the main pathological type of lung cancer
   with a very low 5-year survival rate. In the present study, after
   downloading the mRNA, miRNA, and DNA methylation sequencing data from
   TCGA, combined with the downloaded clinical data, comparative analysis,
   prognostic analysis, GO and KEGG analysis, GSEA analysis, methylation
   analysis, transcriptional regulation and post-transcriptional
   regulation were performed. We found that both methylation and gene
   expression of MNDA in LA were down-regulated, while high expression of
   MNDA was associated with good overall survival in LA. To probe the
   mechanism, further analysis showed that SPI1 was the main transcription
   factor of MNDA, but it was also down-regulated in LA. At the same time,
   the expression of eight target miRNAs of MNDA was significantly
   up-regulated, and the expression of hsa-miR-33a-5p and hsa-miR-33b-5p
   were verified to directly target MNDA. In conclusion, the abnormal
   expression of MNDA in LA is the result of the combined effects of
   transcriptional and post-transcriptional regulation.

   Keywords: lung adenocarcinoma, myeloid cell nuclear differentiation
   antigen, gene methylation, SPI1, miRNA

1. Introduction

   According to the latest report of the World Health Organization (WHO)
   and the International Agency for Cancer Research (IARC), lung cancer is
   the most common malignant tumor type in human beings, accounting for
   11.6% of total cancer and 18.4% of total cancer deaths [[32]1]. In lung
   cancer, non small cell lung cancer (NSCLC) is the most common, which is
   further divided into lung adenocarcinoma (LA) and squamous cell
   carcinoma. LA is the main pathological type of lung cancer, accounting
   for about 40% of the primary lung tumors and prone to brain metastasis
   [[33]2,[34]3]. Due to the lack of obvious early symptoms, the diagnosis
   of NSCLC is often in the late stage of cancer. Although the research on
   lung cancer has made great progress with technological advances, the
   pathogenic mechanism is very complicated, and the 5-year survival rate
   is only 16% [[35]4]. Therefore, it is necessary to find better
   biomarkers for the diagnosis and treatment of lung cancer.

   Myeloid cell nuclear differentiation antigen (MNDA) is a member of the
   immune-related hematopoietic IFN-inducible nuclear protein containing a
   200-amino-acid repeat (HIN-200) family, which can act as a pattern
   recognition receptor to recognize foreign double-stranded DNA [[36]5].
   The N-terminus of MNDA is a highly helical pyrin domain (PYD). PYD
   belongs to one of the four subfamilies of the evolutionarily highly
   conserved death domain protein superfamily. The other three are death
   domains, death effector domains and caspase recruitment domain (CARD).
   These domains play important roles in innate immunity, inflammation,
   differentiation, apoptosis, and cancers through interacting to activate
   a variety of effector proteins, such as caspases and transcription
   factors [[37]6]. The C-terminus of MNDA contains two adjacent
   oligonucleotide/oligosaccharide-binding (OB) domains, which are
   involved in nucleic acid recognition, and proteins with this structural
   feature are involved in DNA replication, recombination, repair and
   telomere maintenance and other processes [[38]6]. Therefore, MNDA is
   mainly involved in autoimmune diseases by affecting the immune system
   [[39]7]. However, the role of MNDA in LA is unclear and needs to be
   elucidated.

   In recent years, epigenetic regulation has been increasingly considered
   to be an important goal of oncology [[40]8]. Researchers have found
   that hypermethylation of tumor suppressors and hypomethylation of
   proto-oncogenes play important roles in the diagnosis and prognosis of
   tumors. DNA methylation involving the transfer of methyl to the C5
   position of cytosine to form 5-methylcytidine, leading to gene
   inhibition or inhibition of the binding of transcription factors to DNA
   [[41]9]. Abnormal DNA methylation can lead to the pathological process
   of organisms, carcinogenesis and development [[42]10]. DNA methylation
   is often considered to be a marker of inhibiting gene transcription, by
   blocking the binding of transcription factors to gene promoter region
   in spatial structure. However, when we analyzed The Cancer Genome Atlas
   (TCGA) data, we accidentally found that both methylation and gene
   expression of some genes were up-regulated or down-regulated, which is
   in contradiction with our general understanding. This study attempted
   to explore whether this phenomenon is a special case or a sense of
   universality by means of bioinformatics, and describe the possible
   mechanism of this phenomenon.

2. Materials and Methods

2.1. TCGA Data Acquisition and Analysis

   We used the University of California Santa Cruz (UCSC) xena
   ([43]http://xena.ucsc.edu/; accessed on 20 February 2019) database to
   obtain the sequencing data and corresponding clinical information
   [[44]11]. The UCSC database is derived from Level 3 data downloaded
   from the TCGA data coordination center. We downloaded the gene
   expression profiles of TCGA-LUAD containing 535 tumor samples and 59
   normal samples, and TCGA-LUSC containing 502 tumor samples and 51
   normal samples. Gene expression profiles were experimentally measured
   using the Illumina HiSeq 2000 RNA sequencing platform at the University
   of North Carolina (UNC) TCGA Center for genome characterization.
   Normalized counts for Fragments Per Kilobase of exon model per Million
   mapped fragments (FPKM) were performed on these datasets. The miRNA
   RNAseq data of TCGA-LUAD and the 450K methylated site data of TCGA-LUAD
   were also downloaded from UCSC xena. The miRNA RNAseq data including
   450 tumors and 45 normal samples were normalized in TPM format.
   Additionally, the 450K methylated site data containing 460 tumor
   samples and 32 normal samples were measured experimentally using the
   Illumina Infinium HumanMethylation450 platform. The 450K integrated
   data and clinical information were downloaded from cBioPortal
   ([45]http://www.cbioportal.org/; accessed on 20 February 2019). We used
   R language and limma package to analyze the difference between gene
   expression, miRNA expression and DNA methylation [[46]11]. The
   differential mRNA was screened out according to the criteria of p <
   0.05 and log2 ≥ |1|, and the differential miRNA was screened out
   according to the criteria of p < 0.05 and log2 ≥ |1|. The significant
   differentially different genes with methylated changes were screened
   under the conditions of p < 0.05 and | methylation change ≥ |0.2|. All
   of the above data were from the TCGA database, and their use and access
   were based on TCGA′s data access policy and publishing guidelines
   ([47]https://cancergenome.nih.gov/publications/publicationguidelines;
   accessed on 20 February 2019). Therefore, no additional ethics
   committee approval was required.

2.2. Comparative Analysis of MNDA between LA and Normal Tissues

   Paired and unpaired analytical methods were used to compare MNDA
   expression between LA and normal lung tissue. The UALCAN database was
   used to analyze the expression of MNDA protein
   ([48]http://ualcan.path.uab.edu/; accessed on 6 May 2022) and the
   methylation changes at its promoter position in MNDA [[49]12,[50]13].
   The Human Protein Atlas (HPA) ([51]https://www.proteinatlas.org/;
   accessed on 6 May 2022) database was used to assess protein expression
   of MNDA by immunohistochemistry [[52]14]. The relationship between the
   methylating site of MNDA and the expression of MNDA in 450K was
   analyzed.

2.3. Intersection of Differential mRNA and Differential Methylation

   Intersection of up-regulated mRNA expression and methylation and
   down-regulated mRNA expression and methylation was obtained. Then, the
   genes for further study were screened under the conditions that genes
   with hypermethylation and high expression were related to good
   prognosis, or genes with low methylation and low expression were
   associated with good prognosis.

2.4. Oncomine Analysis of the Expression of MNDA

   The expression of MNDA in published research was studied by using
   online database Oncomine [[53]15] ([54]https://www.oncomine.org/;
   accessed on 21 August 2019) under the condition of Gene: MNDA; Analysis
   Type: Lung Cancer vs. Normal Analysis; Data Type: mRNA and ORDER BY:
   Under-expression: Fold Change.

2.5. Functional Analysis of MNDA in LA

   The samples in LA were divided into the low expression and high
   expression groups according to the expression level of MNDA for
   differential gene analysis. p < 0.05 and log2 ≥ |1| were used as the
   screening conditions for significant differences. GSEA analysis was
   performed to identify pathways based on the obtained transcript
   sequences of TCGA [[55]16,[56]17]. Gene Ontology (GO) function and
   Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment
   analysis were performed to analyze the function and pathways of
   MNDA-related DEGs. The ssGSEA algorithm in the GSVA package was used to
   evaluate the immune-infiltrating relationship of MNDA [[57]18].

2.6. TF Chip-Seq Analysis of SPI1 Binding to the MNDA Promoter

   The Encode ([58]https://www.encodeproject.org/; accessed on 10
   September 2019) database was used to analyze the binding of SPI1 to the
   MNDA promoter and visualized with the WashU Epigenome Browser online
   tool ([59]http://epigenomegateway.wustl.edu/browser/; accessed on 10
   September 2019) [[60]19,[61]20].

2.7. Prediction of MNDA Target miRNA by TargetScanHuman

   The predictive target miRNAs of MNDA mRNA were selected using the
   online database TargetScanHuman
   ([62]http://www.targetscan.org/vert_72/; accessed on 19 September 2019)
   [[63]21].

2.8. Cell Culture and Real-Time RT-PCR

   Human normal lung epithelial BEAS-2B cells and lung cancer cell lines
   PC9, H1299, and A549 cells were cultured in DMEM plus 10% fetal bovine
   serum (FBS) at 37 °C and 5% CO[2]. Total RNA was extracted from cells
   using TRIzol reagent (Takara, Kyoto, Japan). Primers were synthesized
   by Sangon Biotech Co., Ltd. (Shanghai, China), and the MNDA primers
   were 5′-ACTGACATCGGAAGCAAGAGGG-3′ (forward) and
   5′-TGCAGATGTGCTGGCTCCTGAG-3′ (reverse), the SPI1 primers were
   5′-GACACGGATCTATACCAACGCC-3′ (forward) and 5′-CCGTGAAGTTGTTCTCGGCGAA-3′
   (reverse). Each sample was reverse-transcribed into cDNA using
   TransScript Uni All-in-One First-Strand cDNA Synthesis SuperMix for
   qPCR (Transgen, Tianjing, China). Then, the cDNA was synthesized by
   reverse transcription and amplified using 2× SYBR Green qPCR Master Mix
   (Bimake, Houston, TX, USA) according to the manufacturer’s
   instructions. qPCR was performed at 95 °C for 3 min, and 40 cycles of
   95 °C for 15 s, 60 °C for 30 s, and 72 °C for 30 s. GAPDH and β-actin
   were the internal references for the target genes. The relative