Page 37 - TD-3-3

Tumor Discovery LncRNA HA117 in osteosarcoma regulation

genes. Finally, we selected 11 differentially expressed target io/hisat2/) and Bowtie2 (v2.2.5), respectively. 30,31 Gene
genes for survival analysis. expression profiles of the 51 osteosarcoma samples were
calculated using RSEM (http://deweylab.github.io/RSEM/;
2. Materials and methods accessed on September 25, 2022), and the abundance
32
of all genes was normalized using the fragments per kb
2.1. Study design and data preprocessing
of transcript per million mapped reads (FPKM) method.
Initially, we downloaded batches of human and mouse All differential gene expression analyses were conducted
osteosarcoma and multiple myeloma RNA-seq sequencing using edgeR (version 3.32.1). Only genes that met
33
data from NCBI Sequence Read Archive (SRA) and the threshold criteria of log2|fold change|>1 and false
the European Bioinformatics Institute to test HA117 discovery rate (FDR) <0.001 were considered significant
expression (Table S1). We found that only the data from DEGs. To meticulously compare the differences in HA117
whole transcriptome library sequencing could detect expression between tumor and normal tissues in fresh
the expression of HA117 in osteosarcoma. Therefore, samples, as well as the differences between chemotherapy
we decided to utilize a dataset comprising 51 whole and non-chemotherapy FFPE samples, we extracted
transcriptome sequencing samples (PRJNA389279) for the FPKM values of HA117 and conducted a Wilcoxon
subsequent analysis. Concurrently, we constructed a test. A P < 0.05 was considered to indicate a statistically
customized reference gene database that includes all human significant difference.
mRNA sequences and HA117 for gene quantification and
differential analysis. To accurately predict the target genes 2.3. Prediction of HA117 target genes
regulated by HA117, we combined two methods: one LncRNA can regulate gene expression in both cis and trans
based on sequence binding sites and the other determined manners. Previous studies have confirmed that the cis-
26
by coexpression relationships. Based on the differential regulatory target gene of HA117 is RGS6. Here, we focus
expression of these HA117 target genes, we conducted on predicting the trans-target gene for HA117. Briefly,
multiple bioinformatics analyses on these key target genes. correlation analysis was performed between HA117 and
The 51 whole transcriptome sequencing data were all mRNAs in fresh samples and FFPE samples according
downloaded from the SRA (https://www.ncbi.nlm.nih. to the FPKM value. We first calculated the Spearman rank
gov/bioproject/PRJNA389279), belonging to osteosarcoma correlation coefficients between HA117 and all expressed
bone samples. The dataset were divided into two parts: genes. If the Spearman rank correlation coefficient was
One part consisted of fresh-frozen samples from 18 greater than 0.8 and the significance was P < 0.05, we
osteosarcoma patients, comprising 18 pairs of tumor and considered it to be a robust correlation. The mode of
non-tumor samples; the other part included 15 formalin- action of lncRNA is often more complex than that of
fixed paraffin-embedded (FFPE) osteosarcoma samples. miRNA; hence, lncRNA may regulate gene expression
We used the SRA toolkit to obtain single-end reads in through base pairing complementarity with sequences
fastq format. Additional metadata, such as sample name, such as mRNA, miRNA, or DNA. As a supplement, we
34
grouping, and chemotherapy status, were obtained from used LncTar software (http://www.cuilab.cn/lnctar) to
the literature published by Ho et al. 28 predict the complementary pairing relationship between
HA117 and mRNA, predicting the best binding sites based
2.2. Basic analysis of whole transcriptome data on the minimum free energy, with parameters as follows:

We first performed quality control on the raw data to perl LncTar.pl -p 1 -l HA117.txt -m mra.txt -d-0.1 -s T -o
obtain clean reads for downstream analysis. Briefly, the output.txt. Only mRNAs supported by both methods
downloaded raw sequencing data were first processed were considered regulatory target genes for HA117. The
using fastp (version 0.19.6 [https://github.com/OpenGene/ network relationships and visualization of HA117 with
fastp]) according to the following criteria: (i) Removal of its corresponding target genes were processed using
29
reads containing sequencing adapters, (ii) removal of reads Cytoscape software (version 3.9.0). 35
with more than 10% N bases, and (iii) removal of reads
where the proportion of N bases with a quality score lower 2.4. GO and Reactome enrichment analysis
than 10 exceeds 20%. To accurately detect the expression To further explore the molecular biological functions of
of HA117, we merged the human hg19 transcripts with HA117 target genes, 94 genes were annotated using GO
36
HA117 to serve as the custom reference gene database. and Reactome pathways databases with the DAVID
Subsequently, we mapped the clean reads to the human functional annotation tool (https://david.ncifcrf.gov/).
37
reference genome (GRCh37/hg19) and the custom gene Through the DAVID online platform, the target genes of
database using HISAT2 (http://daehwankimlab.github. HA117 were subjected to GO enrichment and Reactome

Volume 3 Issue 3 (2024) 3 doi: 10.36922/td.3670