CENPF Upregulation is Associated with Immunosuppressive Status and Predicts Poor Clinical Outcomes in Lung Adenocarcinoma Validated by Qrt- PCR

Background: CENPF was differentially expressed in various cancers. However, the relationship between CENPF and immune inltrates in lung adenocarcinoma was previously unknown. Methods: We implemented a comprehensive analysis of expression of CENPF in the GEO and TCGA databases. CENPF was evaluated for its prognostic value combining clinical samples from the GEPIA2 and TCGA databases. The Metascape together with WebGestalt databases were used for enrichment analysis of genesets that were most postively associated with CENPF. We retrieved the score for immune cell inltration in TCGA data and examined the correlation between CENPF expression and the inltration of immune cell by R software. Results: The outcomes exhibited that up-regulated CENPF mRNA expression was evidently related to poor PFS, DSS and OS in patients with lung adenocarcinoma. Moreover, high expression of CENPF was markedly related to genes associated with immune checkpoint. Further analysis showed that T cell CD4+ Th2 inltration increased in lung adenocarcinoma samples with high CENPF expression. Conclusions: Our study indicated that CENPF led to T cell CD4+ Th2 inltration through oncogenic activity and may be employed as a biomarker for the prediction of prognosis in lung adenocarcinoma.


Introduction
In the respiratory system, lung adenocarcinoma (LUAD) is one of the most prevalent malignant tumors [1].
It is prone to recurrence and metastasis, as well as the development of treatment resistance, and owns a high mortality [2,3]. There are currently no early tumor indicators or diagnostic techniques for LUAD that are enough effective [4]. A growing number of important genes have been identi ed as a result of the rapid development of transcriptomic research together with high-throughput sequencing techniques [5,6].
However, further studies to identify more signi cant oncogene, particularly those that may in uence the con guration of the immune microenvironment in LUAD, are now required.
CENPF encodes a protein that is dynamically expressed throughout the cell cycle [7]. For instance, CENPF has been found to be a latent biomarker for the pancreatic cancer both diagnostically and prognostically [8]. Meanwhile, CENPF serves as a novel biomarker for the nonmuscle invasive bladder cancer that may be utilized to predict disease prognosis and progression, in accordance with the weighted gene co-expression network analysis [9]. Additionally, elevated CENPF expression substantially reduces overall survival in individuals with hepatocellular cancer [10]. Nonetheless, the particular CENPF effect in LUAD is mostly unknown at the moment.
In our research, we investigated at the expression of CENPF in a variety of tumor types from three different cohorts, including TCGA, GTEx, and GEO, and explored its correlation with patient outcomes. The overexpression of CENPF was discovered in LUAD and high expression of CENPF was related to poorer PFS, DSS and OS of LUAD patients. What's more, CENPF was predicted to be implicated in pathways related to certain cancer. Since immune cell in ltration affects the prognosis of LUAD patients, we investigated the correlation between the level of immune cell in ltration and CENPF expression and found that T cell CD4+ Th2 in ltration was greater when the expression level of CENPF was high. Our outcomes exhibited that CENPF may exerts a functional effect in LUAD, implying a molecular mechanism by which CENPF affected T cell CD4+ Th2 in ltration in the tumor microenvironment.

Methods And Materials
Data collection The CENPF differential expression in normal samples and 33 cancer was explored using GTEx and TCGA data [11,12]. All expression pro les are displayed in log 2 (TPM+1). The datasets used, including GSE75037, GSE40791 and GSE32863, were from GEO database, and the download data format was MINIML [13,14]. Additional information on three GSE datasets was included in Table 1.

Cell culture
The H1299, A549, HBE, H1975 and PC9 cell lines were provided by the Cell Bank of the Chinese Academy of Sciences, which were cultured with 5% CO2 under a temperature of 37 °C in the RPMI-1640 medium (Gibco) added with penicillin/streptomycin (1%, Gibco) and fetal bovine serum (10%, Biological Industries).

qRT-PCR
TRIzol reagent (Invitrogen, USA) was utilized for extracting the total RNA respectively from H1299, A549, H1975 and PC9 cells. The rst-strand cDNA could be generated from the total RNA with reverse transcriptase and oligo-dT primers (Invitrogen, USA). QuantiTect SYBR Green PCR Master Mix (Qiagen, Germany) was employed to conduct qRT-PCR. On the basis of CENPF mRNA sequence, primers could be designed in GenBank. Additional information regarding primer sequences was available in Table 2. As an endogenous control, β-actin was identi ed in each experimental sample. All responses were performed three times. Table 2 The sequences information of qRT-PCR primers

Survival analysis
With the aim of investigating the latent prognostic effect of CENPF in TCGA pan-cancer, the associations between the expression of CENPF and OS in patients with LUAD were implemented via GEPIA2 database [17]. From TCGA, the clinical data of 33 cancer patients could be downloaded, univariate cox regression was applied to analyze PFS and DSS. The R software package "forestplot" was employed for visualizing 95% con dence interval, HR, and p value of each variable utilizing a forest map.

Correlation and enrichment analysis
To conduct enrichment analysis, the top 300 genes with the positively pearson correlation coe cient in LUAD from TCGA data were selected to reveal the function of CENPF. Then, the Ontology (GO) and genome encyclopedia (KEGG) analyses of the above gene sets were performed using Metascape [18]. Meanwhile, utilizing WebGestalt, Gene set enrichment analysis (GSEA) was implemented in order to further investigate CENPF function [19,20].

Immune cell in ltration analysis
We divided LUAD samples from TCGA into two groups in accordance with the median expression level s(high CENPF and low CENPF) for assessing the level of immune cell in ltration in LUAD tissues. Then, with xCell algorithm, the immune cells in ltration were estimated according to R package "Immuneeconv". CENPF subgroups were represented on the horizontal axis of heat maps, distinct immune scores were displayed on the vertical axis, and correlation coe cients were shown in various colors. In addition, utilizing TIMER2.0 database, the association between the expression pro le of CENPF and the proportion of immune in ltration of T cell CD+ Th2 in LUAD was con rmed [21].
Correlation between CENPF expression and immune checkpoint-related genes, TMB or MSI The relationship between genes related to immune checkpoint and CENPF were studied using LUAD mRNA sequencing data from TCGA. Furthermore, the association between the expression of CENPF and MSI or TMB in LUAD were analyzed using the R package "ggstatsplot" based on TCGA data, respectively. All the above analysis results were displayed by R packages "ggplot2" and "pheatmap".

Statistical analysis
All of the data were performed with R version 4.0.3. The levels of CENPF that were differentially expressed between malignant and non-malignant lung samples were estimated by the Student t-test. The relationships between the immune cells in ltration level and CENPF were investigated through exploiting spearman's correlation. The Spearman's correlation was exploited to express the relationships between the immune cells in ltration level and CENPF. Then p < 0.05 were regarded statistically signi cant.

Results
Pan-cancer CENPF expression analysis  Figure 1A-1D). In addition, CENPF was highly expressed in LUAD based on GSE75037, GSE32863 and GSE40791 ( Figure  1E−1G). The analysis of UALCAN database showed that the expression of CENPF was up-regulated in the tumor tissues of LUAD patients as in contrast to the adjacent normal tissues (Figure 2A). Simultaneously, the CENPF expression level in LUAD increased with the progression of tumor grade ( Figure 2B). Moreover, IHC outcomes in the HPA database exhibited that the expression of CENPF was evidently up-regulated in LUAD lungs in contrast to controls containing normal histology ( Figure 2C). The CENPF expression was measured through qRT-PCR in H1299, A549, H1975 and PC9 cells. According to our ndings, CENPF was overexpressed in H1299, A549, H1975 and PC9 cells in comparison with the HBE cells ( Figure 2D-2F).

Association between CENPF expression and cancer patient prognosis
We

Correlation and enrichment analysis
To ascertain the role of CENPF, we utilized TCGA data to conduct a correlation study between CENPF and other LUAD genes. For enrichment analysis, the rst 300 genes remarkably associated with CENPF were chose (Supplementary Table 1). We next used Metascape to thoroughly investigate biological pathways with above 300 genes. Gene Ontology (GO) analysis revealed that mitotic cell cycle process, cell cycle process regulation together with microtubule skeleton organization were signi cantly enriched in BP; microtubule binding, ATP-dependent activity and chromatin binding were markedly abundant in MF; spindle, spindle pole, as well as chromosomal region were signi cantly enriched in CC ( Figure 4A-4C). Moreover, the study of KEGG pathway suggested that DNA replication, fanconi anemia pathway and cell cycle were obviously abundant ( Figure 4D). GSEA was employed to determine Reactome and KEGG pathways, which revealed that the progesterone-mediated oocyte maturation and the degradation of mitotic proteins pathways mediated by APC/C and APC/C:Cdc20 activation were signi cantly enriched ( Figure 4E-4F). These ndings showed that CENPF was participated in various pathways related to cancer in LUAD.

CENPF expression and immune cell in ltration analysis
Further investigation of LUAD immune cell in ltration score in TCGA revealed that the level of T cell CD4+ Th2 in ltration was substantially higher in the group with high CENPF expression ( Figure 5A-5B). This result indicated that high expression CENPF promoted the in ltration level of T cell CD4+ Th2, which demonstrated a correlation with the immunosuppressive state in LUAD [22]. T cell CD4+ Th2 expression was signi cantly enhanced in tumor patients, which promoted the occurrence and development of malignancies by decreasing cellular immune function [23]. Likewise, we discovered that three indicators scored substantially lower on groups of high CENPF expression, including immune score, microenvironment score, and stroma score ( Figure 5C-5E). As well, we further veri ed the signi cantly positive correlation between the T cell CD4 + Th2 in ltration level and CENPF expression in the LUAD tissues through Timer 2.0 database ( Figure 6A).
Correlation between CENPF and immune checkpointrelated genes, TMB or MSI in LUAD Subsequently, the signi cantly positive correlation between TMB/MSI score and the CENPF expression level in LUAD was found ( Figure 6B-6C). Then, we analyzed the correlation between CENPF and immune checkpoint-related genes using LUAD data in TCGA and found that CENPF was also highly positively correlated with immune checkpoint-related genes in LUAD, such as CD274, KIR2DL3, LAG3, LARS1, NECTIN2, PDCD1, TGFBR1 and TIGIT ( Figure 6D). Following these ndings, it was hypothesized that increased CENPF expression was strongly associated with immunosuppressive state in patients with LUAD [24]. Moreover, several immune checkpoints were chosen with the aim of exploring differential expression levels between high CENPF expression and low CENPF expression, and found that CD274, LAG3, PDCD1 and TIGIT were substantially increased in the group with high CENPF expression ( Figure   6E). These outcomes exhibited that high expression of CENPF was associated with tumor immunosuppression in LUAD patients. The results presented CENPF was positively correlated with many known immune checkpoints so as to inhibit the immune state of tumor. However, the score of TMB/MSI indicated that there may be potential immune checkpoints to be found and the immunosuppression status of LUAD could be effectively improved in patients with LUAD containing high CENPF expression.

Discussion
Although CENPF was identi ed in the development of many malignancies, it had not been widely studied in various types of cancer [25,26]. Hence, the effect of CENPF in the cancer progression and prognosis must be clari ed as rapidly as possible. CENPF has been involved in a variety of cancers, according to previous research. For instance, CENPF can stimulate the metastasis of breast cancer bone through triggering PI3K-AKT-mTORC1 signaling, suggesting it as a new target for breast cancer therapy [27]. Meanwhile, Ping-An Zou et al. found that higher CENPF levels led to osteosarcoma cell proliferation through regulating apoptosis and the cell cycle [28]. Based on current study, CENPF boosted the progression of papillary thyroid carcinoma through controlling cell proliferation and apoptosis [29].
According to our ndings, CENPF was signi cantly expressed in COAD, CHOL, BRCA, LAML, GBM, ESCA, LUAD, LIHC, LGG, STAD, PAAD, OV, UCS and TGCT, while it was expressed at a low level in HNSC, KICH, KIRC, KIRP, READ, SKCM, and THCA according to TCGA and GTEx data. Additionally, the GSE75037, GSE40791, and GSE32863 datasets revealed that CENPF was signi cantly expressed in LUAD. There could be a discernible difference in CENPF expression across tumor types, which might also suggest different underlying actions and mechanisms. Furthermore, we discovered that upregulation CENPF was related to worse prognosis in individuals with tumors expressing a high level of CENPF. However, its expression was related to poor prognosis in THKM. These ndings revealed CENPF might be a bene cial biomarker for the prediction of the cancer patients prognosis.
Tumor microenvironment immune cells were a critical component of tumor tissues, with mounting data demonstrating their clinicopathological role in predicting tumor individual survival and treatment effectiveness [30,31]. We also discovered that when the connection between CENPF and immune cell in ltration was analyzed, the amount of T cell CD4+ Th2 in ltration was substantially greater in the group containing high expression of CENPF. CENPF also revealed a remarkable correlation with genes involved in immune checkpoint-related genes, indicating a function in tumor immunology regulation. Speci cally, the in ltration level of T cell Cd4+ Th2 promotes the progression of LUAD with high CENPF.

Conclusion
In summary, CENPF is likely to be involved in in the in ltration of immune cell and may be a useful predictive biomarker for lung adenocarcinoma.      The CENPF expression correlated with immune in ltration. (A) The expression of CENPF remarkably associated with a variety of immune cells in ltration levels according to xCell. (B) T cell CD4+ Th2 in ltration level was substantially higher in LUAD tissues with high CENPF expression than in LUAD tissues with low CENPF expression. (C-E) The microenvironment, stroma and immune scores in high CENPF expression group was evidently lower than the scores in low CENPF expression group. * p < 0:05, * * p < 0:01, and * * * p < 0:001.