Page 51 - GPD-3-1

Gene & Protein in Disease Prognostic potential of LMNB2 in LPS

2.5. Analysis of immune infiltration by CIBERSORT genes. Among them, only LMNB2 exhibited a |FC| <1 and

CIBERSORT (https://cibersortx.stanford.edu/) is a tool P < 0.05 (P = 0.013, HR = 3.117, 95% confidence interval
designed for the deconvolution of expression matrices of [CI, 1.271 – 7.645]), while other genes did not meet this
human immune cell subtypes. It operates on the principle statistical criterion (Figure S1). Therefore, we initially
of linear support vector regression, using gene expression considered LMNB2 as a potential LPS biomarker and
data to estimate the abundance of cell types within a proceeded with further investigations.
mixed cell population. 37,38 After uploading a microarray 3.2. Significantly upregulated LMNB2 in LPS tissues
or sequencing expression matrix and a reference dataset,
CIBERSORT generates outputs indicating the proportion To investigate the potential prognostic value of LMNB2 as
of immune cell infiltration based on the reference dataset. a biomarker for LPS, we analyzed the expression of LMNB2
In addition, it provides statistics such as P-values, R, and in different subtypes of LPS tissues and normal tissues using
RSME, which sum to one for all cell types under default GSE21122 and GES30929 datasets. The results revealed a
parameters. In our study, CIBERSORT was employed significant increase in LMNB2 expression in LPS tissues
39
to analyze immune infiltration in tumor tissues obtained compared with normal controls (Figure 3A). Moreover,
from 60 LPS patients in the TCGA dataset. This analysis the expression of LMNB2 was upregulated across different
yielded insights into the percentage of immune cell subtypes of LPS tissues compared with normal tissues,
infiltration based on the reference dataset. The results suggesting a close association between the increased
obtained were further visualized using GraphPad software, expression of LMNB2 and the development of each subtype
which provides new insights into biomarkers associated of LPS (Figure 3B). In addition, significant differences were
with LPS pathogenesis and prognosis. observed in the expression of LMNB2 among different LPS
subtypes, especially between WDLPS and other subtypes.
2.6. Statistical analysis Notably, the expression of LMNB2 was significantly

IBM SPSS Statistics 26 (SPSS Inc., USA) and GraphPad higher in DDLPS or PLPS than Myxoid/round cell LPS,
Prism 8.4.3 (GraphPad Inc., USA) were used for statistical indicating a potential correlation between the expression
analysis. Univariate and multivariate Cox regression of LMNB2 and the occurrence of different LPS subtypes
models were developed using SPSS software. In the (Figure 3B and C). The observed significant differences in
univariate analysis, biomarkers significantly associated the expression of LMNB2 between normal and LPS tissues,
with LPS prognosis were identified, with risk factors as well as among different LPS subtypes, suggest a potential
screened (P < 0.05). Subsequently, meaningful risk role for LMNB2 in the occurrence and development of
40
factors identified in the univariate analysis were also LPS. In addition, analysis of LMNB2 expression in DDLPS
subjected to a multifactorial analysis, with a significance patients revealed significantly higher expression of LMNB2
level of P < 0.05 employed as the criterion for statistical in patients with copy number amplification (n = 21) than
significance. Survival curves for LPS patients from the those with normal diploid copy number (n = 32) (P = 0.003)
TCGA or GSE30929 dataset were plotted by GraphPad (Figure S2), indicating a consistent relationship between
software with a 50% cutoff value. DNA copy number of LMNB2 and its expression level.
3. Results Next, receiver operating characteristic curves were
generated based on the expression levels of LMNB2 in LPS
3.1. Preliminary screening of LPS prognostic and normal tissues of each subtype. The results revealed a
biomarker significant increase in LPS tissues compared with normal
Through GEO2R, DEGs were obtained from GSE21122 controls (Figure 3D). Furthermore, the expression of
and GSE159659 (with a significance threshold of P < 0.05 LMNB2 was increased across various subtypes of LPS tissues
and |FC| >2). After overlapping, 192 DEGs were acquired compared with normal tissues (Figures 3E and S3A-S3B).
by intersecting 318 DEGs in GSE159659 and 896 DEGs In addition, significant differences in LMNB2 expression
in GSE21122 (Figure 2). The resulting gene list was then were observed among different LPS histological subtypes
imported into the KOBAS database for pathway enrichment (Figures 3F and S3C-S3F). These results indicate that
analysis. A meaningful apoptosis pathway was identified the expression of LMNB2 differs significantly between
in the exported result file (false discovery rate [FDR] normal and LPS tissues, as well as among LPS histological
<0.05), comprising 14 genes, including LMNB2, LMNB1, subtypes. This finding suggests that the expression of
FOS, GADD45B, NFKBIA, JUN, BAX, MAP3K5, PIK3R3, LMNB2 demonstrates a good discrimination ability for
MCL1, PIK3R1, ATF4, ITPR1, and BIRC5. Subsequently, diagnosing LPS or differentiating LPS subtypes, and it
SPSS software was used for univariate analysis of these 14 might be related to the occurrence and development of LPS.

Volume 3 Issue 1 (2024) 5 https://doi.org/10.36922/gpd.2607