Abstract

   The prognosis and treatment efficacy of lung adenocarcinoma (LUAD), a
   disease with a high incidence, remains unsatisfactory. Identifying new
   biomarkers and therapeutic targets for LUAD is essential. Chromosomal
   assembly factor 1B (CHAF1B), a p60 component of the CAF-1 complex, is
   closely linked to tumor incidence and cell proliferation. However,
   CHAF1B's biological role and molecular mechanism in LUAD remain
   unclear. Here, CHAF1B expression in LUAD was examined using the GEPIA2
   and UALCAN databases. Using The Cancer Genome Atlas (TCGA) LUAD
   database, we analyzed the diagnostic and prognostic significance of
   CHAF1B and its association with immune infiltration and immunological
   checkpoints. Gene ontology (GO) enrichment and single-cell function
   analyses were employed to investigate CHAF1B's possible biological
   roles. Drug sensitivity analysis predicted CHAF1B's effect on
   chemotherapeutic drug sensitivity. We also predicted
   lncRNAs-miRNA-CHAF1B axis to explore the molecular mechanism of CHAF1B
   in LUAD. Preliminary in vitro studies using qRT-PCR, CCK8, Transwell,
   glucose, and lactate metabolism confirmed CHAF1B's expression and role
   in LUAD. Its expression is associated with drug sensitivity,
   immunological checkpoints, and immune cell infiltration. We predicted
   that three miRNAs (miR-29c-3p, miR-145-5p, miR-1247-5p) and three
   lncRNAs ([36]AL139287.1, NEAT1, SHG1) may be target miRNAs and target
   lncRNAs that regulate CHAF1B. In vitro tests showed that CHAF1B
   suppression decreased LUAD's migration, invasion, proliferation, and
   glycolysis. Overall, CHAF1B may be an innovative biomarker and
   therapeutic target for LUAD.

   Keywords: CHAF1B, Lung adenocarcinoma, Biomarkers, Immunotherapy
   targets, Drug sensitivity, Prognosis

Introduction

   Lung cancer is among the most prevalent malignant tumors worldwide,
   with the highest global fatality rate and second-highest incidence
   rate. Consequently, there is an urgent need to address this serious
   public health issue [[37]1]. Non-small cell lung cancer (NSCLC)
   accounts for 85% of all lung cancers, with LUAD being the most
   prevalent form [[38]2]. Despite significant advances in the treatment
   of LUAD over the past few decades, challenges such as postoperative
   tumor recurrence, chemotherapy resistance, and targeted therapy
   resistance persist. On average, fewer than 20% of patients survive
   beyond five years [[39]3]. These limitations are partly owing to the
   lack of reliable biomarkers or therapeutic targets [[40]4]. Therefore,
   investigating the molecular mechanisms of LUAD and identifying
   innovative biomarkers and treatment targets are essential to extend
   survival and enhance patient quality of life.

   The chromosome assembly factor (CAF-1) complex comprises three
   subunits: large (CAF-1-P150, CHAF1A), small (RbAp48, P48), and medium
   (CAF-1-P60, CHAF1B). This complex facilitates ribosome assembly by
   enlisting histones H3 and H4 and is implicated in DNA synthesis
   [[41]5]. CHAF1B, the p60 subunit of CAF-1, plays crucial roles in DNA
   replication, chromosomal assembly, and DNA damage repair [[42]6]. Its
   expression is substantially downregulated in quiescent cells, making it
   an effective marker for differentiating between proliferative and
   quiescent cells [[43]7]. CHAF1B has been linked to the growth of
   various malignant cancers, including hepatocellular carcinoma[[44]8],
   acute myeloid leukemia [[45]9], myeloproliferative neoplasm [[46]10],
   leukemia [[47]11] and prostate cancer [[48]12]. Furthermore, CHAF1B
   shows potential as a biomarker for cervical cancer [[49]13],
   hepatocellular carcinoma [[50]14] and glioma [[51]15].

   Prior research indicates that high levels of CHAF1B in NSCLC tissues
   may stimulate disease growth and inhibit apoptosis, resulting in a poor
   prognosis [[52]16]. However, the precise mechanism by which CHAF1B
   functions in LUAD, along with its biological role and expression
   pattern, remain unclear. Therefore, we examined CHAF1B expression in
   LUAD and assessed its impact on tumor identification, onset,
   progression, and prognosis. Additionally, we explored its role in LUAD
   and its effect on drug sensitivity. To predict the molecular mechanism,
   we constructed a lncRNAs-miRNA-CHAF1B network. Using siRNA to
   downregulate CHAF1B expression in A549 cells, we confirmed its gene
   expression and biological roles. This study aimed to highlight the
   potential diagnostic and prognostic significance of CHAF1B in LUAD,
   along with its association with immune infiltration and drug
   sensitivity. The findings offer novel avenues for the clinical
   diagnosis and treatment of LUAD.

Materials and methods

Data download and analysis

   We downloaded the TCGA-LUAD dataset from the publicly accessible TCGA
   database ([53]http://cancergenome.nih.gov/), containing 598 LUAD and 59
   normal tissues. Log2 (value + 1) conversion was performed on the per
   million base (FPKM) RNA-seq data, and TPM-formatted and clinical data
   were extracted. Using the GEPIA2 online database [[54]17]
   ([55]http://gepia.cancer-pku.cn/), we examined CHAF1B expression
   differences in malignant and normal tissues and its relationship with
   survival time in patients with LUAD. The UALCAN database [[56]18]
   ([57]http://ualcan.path.uab.edu) was used to confirm CHAF1B expression
   levels in tumor tissues versus normal tissues and its connection to the
   TP53 mutation.

Association between clinicopathological characteristics of patients with LUAD
and CHAF1B expression

   We used the" Wilcoxon signed-rank test" in the R package to examine the
   association between CHAF1B expression levels and the
   clinicopathological characteristics of patients with LUAD. Results were
   visualized with ' ggplot2' [3.3.6] and a significance level of p < 0.05
   was considered meaningful.

Diagnostic and prognostic value of CHAF1B

   ROC analysis was performed on TCGA-LUAD dataset using the R package
   'pROC ' [1.18.0], with results visualized by 'ggplot2 ' [3.3.6].
   Diagnostic accuracy was deemed high if the area under the ROC curve
   (AUC) was higher than 0.9 [[58]19]. Proportional hazards and survival
   regression were analyzed using GEPIA2 and the R package "survival"
   [version 3.6]. The ' ggplot2 ' [3.3.6] package and 'survmine' package
   were used to visualize the results. In this study, KM was used to
   examine the relationship between CHAF1B expression and the overall
   survival (OS) of patients with LUAD. Patient samples were classified
   into two groups based on median gene expression: high and low. Survival
   risk was evaluated using risk ratios (HR) and log-rank p-values.

Correlation and gene enrichment analysis

   The connection between CHAF1B and other gene mRNAs in LUAD was analyzed
   using TCGA-LUAD data, visualized by the R package “ggplot2” [3.3.6]. In
   addition, we used Metascape [[59]20]
   ([60]http://metascape.org/gp/index.html) to perform GO enrichment and
   PPI network analyses of the top 100 genes related to CHAF1B.

Single cell RNA-seq analysis

   The Human Protein Atlas (HPA) [[61]21]
   ([62]http://www.proteinatlas.org/) database provides a variety of
   reports and tables of tissue, cell, and pathology, as well as
   subcellular localization of genes, which can be used to search for the
   expression of CHAF1B in different single-cell types. The biological
   function of CHAF1B at the single-cell level in LUAD was investigated
   using the CancerSEA database [[63]22]
   ([64]http://biocc.hrbmu.edu.cn/CancerSEA/), which extensively studies
   the various functions of cancer cells at the single-cell level. The
   database includes 41900 cancer single cells from 25 human cancers.

Immune infiltration analysis

   Using the ssGSEA method of the R package “GSEA,” the immune
   infiltration analysis determines the immune infiltration of each sample
   in the TCGA-LUAD data set using the 24 immune cell characteristic genes
   supplied by Bindea et al. [[65]23]. In addition, we used the TCGA-LUAD
   database to analyze the Spearman correlation between the expression of
   CHAF1B and seven clinically commonly used LUAD immune checkpoints. The
   “ggplot2” package was used to visualize the results above. P < 0.05 was
   considered statistically significant.

Drug sensitivity analysis

   Data were downloaded from the CellmMiner database
   ([66]https://discover.nci.nih.gov/cellminer/) regarding drug
   sensitivity and CHAF1B mRNA expression profiles from 60 distinct cancer
   cell lines. The “limma” and “impute” package of the R language were
   used to perform correlation analysis on the sensitivity score of
   FDA-approved medications and the expression profile data of CHAF1B
   mRNA. The “ggplot2” and “ggpubr” packages were then used to illustrate
   the analysis results.

Construction of LncRNA-miRNA-CHAF1B regulatory network

   Five online databases, DIANA-mircoT
   ([67]http://diana.imis.athena-innovation.gr/DianaTools/index.php?r=micr
   oT_CDS/index), miRcode ([68]http://www.mircode.org/index.php), miRwalk
   ([69]http://mirwalk.umm.uni-heidelberg.de/), Targetscan
   ([70]http://www.targetscan.org/vert_80/) and ENCORI
   ([71]https://rnasysu.com/encori/), were used to predict CHAF1B-targeted
   microRNAs (miRNAs). The “ggplot2” tool and the “VennDiagram” package
   were used to illustrate the unique and common elements of the data from
   each group. At least two miRNA databases were used to identify target
   miRNAs. The ENCORI database ([72]https://rnasysu.com/encori/) was used
   to analyze the interaction between lncRNA and miRNA, the expression
   levels of miRNA and lncRNA, and the interaction between miRNA and
   CHAF1B in LUAD. Finally, SangerBox3.0
   ([73]http://sangerbox.com/tool.html) was used to make a Sankey diagram
   of the lncRNA-miRNA-CHAF1B network.

Tissue specimen collection

   Thirteen patients with LUAD, including cancerous tissues and nearby
   tissues (more than 5 cm from the margin of the mass), were recruited
   from the Department of Thoracic Surgery at the Second Affiliated
   Hospital of Xiamen Medical College between July 2017 and February 2023.
   Tissue samples were immediately frozen in liquid nitrogen and stored in
   a refrigerator at − 80 °C. All tissue specimens were confirmed as LUAD
   by at least two pathologists, with no patients having a history of
   other tumors, preoperative radiotherapy, or chemotherapy. Written
   informed consent was obtained from each patient. The study was approved
   by the Second Affiliated Hospital Ethics Committee of Xiamen Medical
   College, conducted in compliance with the Declaration of Helsinki and
   best clinical practices.

Cell culture and transfection

   Three lines (A549, H1975, 95D) and a normal human bronchial epithelial
   cell line (16HBE) were purchased from Shanghai Institute of
   Biochemistry. A549 and 95D cells were cultured in FBS-supplemented
   DMEM, while H1975 and 16HBE cells were cultured in FBS-supplemented
   RPMI 1640 medium. A549 and H1975 cells were seeded in 6-well plates,
   and si-nc (control) and si-CHAF1B (experimental group) were transfected
   at 30–50%. After 48 h, knockdown efficiency was determined by qRT-PCR.
   The target sequence of si-CHAF1B1 is GACGGATCTTTGCTTCTC, that of
   si-CHAF1B2 is GCAAGAAGCTACCGGATGT, and si-CHAF1B3’s is
   GTCCAATCTTGCTCGTCAT. All the siRNAs were purchased from Ruibo
   (Guangzhou, China).

qRT-PCR

   TRIzol reagent (RK30129, Abclonal) was used to extract RNA from tissues
   and cells, and a reverse transcription kit (ABclonal, RK20433) was used
   to convert RNA into cDNA. Amplification was performed using a qPCR kit
   (RK21203; ABclonal). The relative expression of CHAF1B mRNA in tissues
   and cells was compared by 2^−ΔΔ Ct method with GAPDH as internal
   reference. The primers for qPCR were as follows: GAPDH (forward primer:
   5′- ACAACAGCCTCAAGATCATCAGC-3′, reverse primer: 5′-
   GCCATCACGCCACAGTTTCC-3′) and CHAF1B (forward primer:
   5′-CCATCGCTCATCTTCCATGTCCTG-3′, reverse primer:
   5′-CTCCACCTGTTTCCACCACTG-3′).

Cell activity detection

   After 48 h of transfection, A549 and H1975 cells were resuspended by
   trypsin digestion and seeded into 96-well plates at a density of
   4 × 10^3 cells/well. Each group consisted of three holes. The cells
   were cultured in the incubator for 24 h, 48 h, and 72 h, and 10 µl
   CCK-8 solution (MA0218, MeilunBio) was added to the corresponding
   holes. The absorbance value was determined at 450 nm wavelength after
   2 h.

Transwell experiment

   Migration experiment Following a 48-h transfection period, A549 and
   H1975 cells were reconstituted via trypsin digestion and introduced
   into the upper chamber (200 µl/well) of a transwell plate (3422-A,
   Corning) with serum-free medium at a cell density of 2 × 10^4
   cells/well, and 600 µl of complete medium containing 10% fetal bovine
   serum was added dropwise to the lower chamber. Three duplicate wells
   were used for each group, and cells were cultured for 24 h. After the
   culture, the medium was removed, PBS was washed, paraformaldehyde
   (P0099, Beyotime) was fixed for 30 min, 0.5% crystal violet solution
   (G1065, Solarbio) was stained for 30 min, photographed, and counted.

   Invasion experiment Matrigel (354262, BD Biosciences) was diluted with
   serum-free medium at 1 of 6, and 60 µl/well was spread in the upper
   chamber of the transwell for 6 h. The cells were inoculated in the
   upper chamber (200 µl/well) of transwell plate (3422-A, Corning)
   according to the number of 5 × 10^4 cells/well, and 600 µl complete
   medium containing 10% fetal bovine serum was added in the lower
   chamber. Three duplicate wells were used for each group and cultured
   for 24 h. After the culture, the medium was removed, PBS was washed,
   paraformaldehyde (P0099, Beyotime) was fixed for 30 min, 0.5% crystal
   violet solution (G1065, Solarbio) was stained for 30 min, photographed,
   and counted.

Glucose and lactic acid determination

   After 48 h of transfection, glucose and lactic acid concentrations in
   the medium of the A549 and H1975 cells were measured using a
   multifunctional microplate reader and glucose detection kits (A154-1-1,
   Nanjing Jiancheng), and lactic acid detection kit (A019-2-1, Nanjing
   Jiancheng).

Data analysis

   GraphPad Prism 9.2.0, and R software (version 4.1.3) were used for
   statistical analysis, with image analysis performed using Image J
   software. Data were expressed as mean ± standard deviation (x ± s). The
   two groups were compared using the t test while multiple group
   comparisons were performed using one-way analysis of variance, and at
   least three repeated experiments were performed. P < 0.05 was
   considered statistically significant.

Results

Expression of CHAF1B mRNA was up-regulated in LUAD

   First, the expression levels of CHAF1B mRNA in various malignant tumor
   types were determined using the GEPIA2 database. These findings
   indicated that CHAF1B mRNA was up-regulated in BLCA, BRCA, CESC, COAD,
   DLBC, LUAD, LUSC, READ, SKCM, STAD and THYM, but downregulated in LAML
   cells (Fig. [74]1A, B). According to the UALCAN database, CHAF1B mRNA
   expression level in LUAD tissues was considerably higher than that in
   normal tissues (Fig. [75]1C). Additionally, patients with TP53
   mutations exhibited high CHAF1B mRNA expression levels (Fig. [76]1D).
   According to the ROC curve, the AUC of CHAF1B mRNA in LUAD was 0.963,
   suggesting its potential as a diagnostic tool for early detection
   (Fig. [77]1E). KM analysis indicated that patients with high levels of
   CHAF1B mRNA expression exhibited short survival times [hazard ratio
   (HR) = 1.5, p = 0.0098] (Fig. [78]1F). Consequently, CHAF1B may
   influence the prognosis of LUAD and could be used to develop novel
   biomarkers for LUAD.

Fig. 1.

   [79]Fig. 1
   [80]Open in a new tab

   CHAF1B mRNA is up-regulated in LUAD. A The GEPIA2 database shows CHAF1B
   mRNA expression in a variety of malignant tumor types. B, C GEPIA2 and
   UALCAN databases showed that CHAF1B mRNA was up-regulated in LUAD. D
   The UALCAN database revealed a correlation between CHAF1B mRNA
   expression and TP53 mutation. E ROC curve of differentially expressed
   CHAF1B in patients with LUAD. F The GEPIA2 database showed the survival
   curve of differentially expressed CHAF1B in patients with LUAD. *
   P < 0.05

CHAF1B expression correlates with clinicopathological parameters in patients
with LUAD

   We examined the relationship between CHAF1B expression and
   clinicopathological characteristics of patients with LUAD to
   investigate the clinical implications of high CHAF1B expression in
   LUAD. The findings demonstrated that the overexpression of CHAF1B was
   positively correlated with the clinicopathological stage (I, II and
   III, IV), N stage (N0, N1 and N2, N3), and T stage (T1 and T2) of LUAD,
   with higher expression of CHAF1B in patients at a later stage
   (Fig. [81]2A–C). Additionally, we examined the prognostic correlation
   between CHAF1B and patients with LUAD. The findings indicated that in
   patients with stage I and II, N0, and M0 stages, the shorter the
   survival time, the greater the expression of CHAF1B (P < 0.05)
   (Fig. [82]2D–F). CHAF1B is crucial for LUAD onset and progression and
   has good prognostic value for patients with stage N0, M0, stage I and
   stage II disease.

Fig. 2.

   [83]Fig. 2
   [84]Open in a new tab

   Expression of CHAF1B is related to the clinicopathological features of
   LUAD patients. A−C The differential expression of CHAF1B in different
   clinicopathological stages. A pathological grade. B N stage. C T stage.
   D–F the correlation between CHAF1B and the prognosis of LUAD patients
   with different clinical stages. D Stage I and II. E N0. F M0.
   *P < 0.05,**P < 0.01

The co-expression pattern of CHAF1B in LUAD

   We searched for genes co-expressed with LUAD and produced a volcano map
   to identify genes that can interact with CHAF1B in LUAD and facilitate
   the understanding of the mechanism of CHAF1B (Fig. [85]3A). The top 50
   most important genes that were positively or negatively correlated with
   CHAF1B (Fig. [86]3B, C) are also shown.

Fig. 3.

   [87]Fig. 3
   [88]Open in a new tab

   Results of CHAF1B co-expression analysis. A Volcano map, red points
   indicate positive correlation of genes, and blue points indicate
   negative correlation of genes. B The heat map showed the first 50 genes
   positively correlated with CHAF1B. C Heatmap showed the top 50 genes
   negatively correlated with CHAF1B

The role of CHAF1B co-expressed genes in LUAD

   We used pathway enrichment analysis to gain further insight into the
   mechanism of CHAF1B co-expressed genes in LUAD. According to the
   Metascape website, the top 100 genes co-expressed with CHAF1B were
   primarily clustered in the mitotic cell cycle, cell cycle regulation,
   chromosome organization, and cell cycle checkpoints (Fig. [89]4A, B).
   According to GO enrichment analysis, CHAF1B was mostly engaged in
   cellular processes, regulation of biological processes, and metabolic
   processes. (Fig. [90]4C). The PPI network of the top 100 genes
   co-expressed with CHAF1B was mainly concentrated in the mitotic cell
   cycle, mitotic cell cycle process, and the cell cycle (Fig. [91]4D, E).
   Additionally, MCODE analysis indicated that CHAF1B and its adjacent
   genes may affect ATR activation in response to replication stress,
   activation of the pre-replication complex, and DNA replication
   (Fig. [92]4F, G). In conclusion, these findings suggest that genes
   co-expressed with CHAF1B are mostly involved in the mitotic cell cycle
   and cell cycle activities in LUAD. Based on these findings, we
   hypothesized that CHAF1B is involved in the regulation of cell
   proliferation.

Fig. 4.

   [93]Fig. 4
   [94]Open in a new tab

   CHAF1B and co-expressed genes in LUAD were analyzed using GO and PPI. A
   & B Cluster pathway analysis of genes co-expressed with CHAF1B. C
   CHAF1B involved in GO analysis bar chart. D & E CHAF1B and its adjacent
   genes involved in the protein–protein interaction network. F & G CHAF1B
   and adjacent genes involved in MCODE components

Analysis of CHAF1B at the single cell level in LUAD

   The HPA database showed that the top three single cells with CHAF1B
   expression in lung cancer tissues were alveolar cells type 2,
   macrophages, and club cells. CHAF1B is mainly expressed in alveolar
   cells type 2, with 14.0 standardized transcripts per million
   protein-coding genes (nTPM) in this cell (Fig. [95]5A). Simultaneously,
   owing to the different expression levels of CHAF1B, each cell was
   divided into different types of cell clusters. CHAF1B is primarily
   expressed in c-8 cell clusters (Alveolar cells type 2) (Fig. [96]5B).
   We analyzed the correlation between the expression level of CHAF1B and
   the biological function of single LUAD cells using the CancerSEA
   database. The results showed that CHAF1B expression was positively
   correlated with DNA repair and the cell cycle, and negatively
   correlated with quiescence, inflammation and differentiation
   (Fig. [97]5C). These results further reveal the potential biological
   function of CHAF1B in LUAD at the single-cell level.

Fig. 5.

   [98]Fig. 5
   [99]Open in a new tab

   Single-cell level analysis of CHAF1B in LUAD. A HPA database showed the
   single cell expression level of CHAF1B in lung cancer tissues. B The
   expression level of CHAF1B in different cell clusters of lung cancer. C
   CancerSEA database shows the correlation between CHAF1B and LUAD single
   cell biological function. *P < 0.05,**P < 0.01,***P < 0.001

The correlation between CHAF1B and immune infiltration in LUAD was analyzed

   This study investigated the association between CHAF1B expression and
   immune cell infiltration in patients with LUAD. Spearman correlation
   analysis was conducted to examine the relationship between CHAF1B
   expression and immune cell infiltration, as well as immune checkpoints.
   The results indicated a significant positive correlation between CHAF1B
   expression and Th2 cell infiltration (R = 0.543, P < 0.001), and
   negative correlations with CD8 + T cell infiltration (R = −0.232,
   P < 0.001), B cell infiltration (R = − 0.241, P < 0.001), iDC cell
   infiltration (R = − 0.258, P < 0.001), and DC cell infiltration
   (R = − 0.218, P < 0.001) (Fig. [100]6A–F). Furthermore, there was a
   significant positive correlation observed between CHAF1B expression and
   the immune checkpoints PDCD1 (R = 0.127, P = 0.003), CD274 (R = 0.189,
   P < 0.001), LAG3 (R = 0.223, P < 0.001), and SIGLEC15 (R = 0.145,
   P < 0.001) as depicted in Fig. [101]7A–E. These findings suggest that
   CHAF1B may influence the tumor immune microenvironment and potentially
   serve as a novel target for immunotherapy in patients with LUAD.

Fig. 6.

   [102]Fig. 6
   [103]Open in a new tab

   LUAD immune cell infiltration and CHAF1B relationship. A A comparison
   of 24 different immune cell types' infiltration and CHAF1B expression.
   B−F The correlation scatter plot between immune cell infiltration and
   CHAF1B expression. B Th2 cells. C CD8 T cells. D B cells. E iDC cells.
   F DC cells

Fig. 7.

   [104]Fig. 7
   [105]Open in a new tab

   Immune checkpoints and CHAF1B correlate in LUAD. A The heat map
   demonstrating association between seven different types of
   immunological checkpoints and CHAF1B expression. B−E CHAF1B expression
   and immune checkpoint correlation scatter plot. B PDCD1. C CD27. D
   LAG3. E SIGLEC15

Correlation analysis between CHAF1B and drug sensitivity

   In the CellMiner database, CHAF1B expression was significantly
   associated with the sensitivity of 12 FDA-approved drugs (Fig. [106]8).
   The CHAF1B expression level was positively correlated with the half
   maximal inhibitory concentration (IC50) of Nelarabine (Fig. [107]8A),
   Vorinostat (Fig. [108]8B), Methylprednisolone (Fig. [109]8C) and
   6-Thioguanine (Fig. [110]8i). This finding suggests that patients with
   high CHAF1B expression may enhance Nelarabine, Vorinostat,
   Methylprednisolone and 6-Thioguanin resistance. In contrast, the
   expression level of CHAF1B was negatively correlated with IC50 of
   Alectinib (Fig. [111]8D), Defactinib (Fig. [112]8E), Celecoxib
   (Fig. [113]8F), Elesclomol (Fig. [114]8G), PF-06463922(lorlatinib)
   (Fig. [115]8H), Isotretinoin (Fig. [116]8J), brigatinib (Fig. [117]8K)
   and Olaparib (Fig. [118]8L).

Fig. 8.

   [119]Fig. 8
   [120]Open in a new tab

   Correlation analysis between CHAF1B expression and drug sensitivity.
   A–L Correlation between CHAF1B expression and drug sensitivity scatter
   plot. A Nelarabine. B Vorinostat. C Methylprednisolone. D Alectinib. E
   Defactinib. F Celecoxib. G Elesclomol. H PF-06463922(lorlatinib). I
   6-Thioguanine. J Isotretinoin. K brigatinib. L Olaparib

Construction of LncRNAs-miRNA-CHAF1B network in LUAD

   Some lncRNAs can act as competitive endogenous RNA (ceRNAs) to
   competitively adsorb onto common sites of miRNAs and mRNA, thereby
   regulating mRNA expression [[121]24, [122]25]. To further explore the
   molecular mechanism underlying the biological function of CHAF1B, we
   predicted the lncRNAs-miRNA-CHAF1B network using software tools. The
   ENCORI, miRWalk,TargetScan, DIANA-miroT, and miRcode databases were
   used to identify target miRNA of CHAF1B. These databases identified 18,
   1711,700, 77, and 22 miRNAs, respectively. A total of 495 miRNAs were
   identified at the intersections of at least two databases
   (Fig. [123]9A). We selected three miRNAs that may be negatively
   correlated with mRNA expression, namely hsa-miR-29c-3p, hsa-miR-145-5p,
   and hsa-miR-1247-5p (Fig. [124]9B–D). According to the results shown in
   the ENCORI database, we found that hsa-miR-29c-3p, hsa-miR-145-5p, and
   hsa-miR-1247-5p were down-regulated in cancer tissues of LUAD patients
   (Fig. [125]9E–G). Next, we used the ENCORI database to predict the
   target lncRNA of miRNA and made a Sankey diagram. The findings
   indicated that six lncRNAs (AL13928711, [126]AC016717.2, MIR4458HG,
   [127]AC007036.3, NEAT1, and OIP5-AS1) might be the target lncRNAs of
   hsa-miR-29c-3p, NEAT1 and SNHG22 might be the target lncRNAs of
   hsa-miR-1247-5p, and three lncRNAs (SNHG1, MALAT1, [128]AC006064.5)
   might be the target lncRNAs of hsa-miR-145-5p (Fig. [129]9H). Among
   them, [130]AL139287.1, NEAT1 and SHG1 were highly expressed in cancer
   tissues of LUAD patients. They are more likely to be the lncRNA
   molecules that regulate the expression of CHAF1B (Fig. [131]9I–K).

Fig. 9.

   [132]Fig. 9
   [133]Open in a new tab

   Construction of LncRNAs-miRNA-CHAF1B network in LUAD. A ENCORI
   database, miRWalk database,Targetscan database, DIANA-miroT database
   and miRcode database were used to identify the Wayne diagram of miRNAs
   binding to CHAF1B. B−D CHAF1B and target miRNA correlation scatter
   plot. B hsa-miR-29c-3p. C hsa-miR-145-5p. D hsa-miR-1247-5p. E–G
   Expression level of miRNA in cancer tissues and normal tissues of LUAD
   patients. E hsa-miR-29c-3p. F hsa-miR-145-5p. G has-miR-1247-5p. H
   Predicted Sankey diagram of lncRNAs-miRNA-CHAF1B network. I–K
   Expression level of lncRNA in cancer tissues and normal tissues of LUAD
   patients. I [134]AL139287.1. J NEAT1. K SHG1.
   *P < 0.05,**P < 0.01,***P < 0.001

CHAF1B promotes the proliferation, migration, invasion and glycolysis of LUAD
cells

   The bioinformatics analyses provided a preliminary understanding of
   CHAF1B’s biological functions in LUAD. To validate our results, we used
   13 clinical samples of LUAD cancer tissues and adjacent tissues to
   detect CHAF1B mRNA expression in lung adenocarcinoma tissues using
   qRT-PCR. Our findings demonstrated that lung cancer tumors expressed
   more CHAF1B mRNA than normal tissues (Fig. [135]10A). Additionally,
   CHAF1B was highly expressed in A549, H1975, and 95D cells relative to
   16HBE cells with statistical significance observed in in A549 and H1975
   cells (Fig. [136]10B). Given the high expression of CHAF1B in both LUAD
   tissues and cells, and its association with poor prognosis as indicated
   by bioinformatic analysis, we hypothesized that CHAF1B plays a
   significant influence in cancer progression in LUAD. To verify this
   hypothesis, we used si-CHAF1B to knock down CHAF1B mRNA expression in
   A549 and H1975 cells. We also constructed three siRNA-transfected cell
   lines to ensure knockdown efficiency. In A549 and H1975 cells,
   si-CHAF1B1 and si-CHAF1B3 demonstrated higher knockdown efficiencies in
   both cell lines (Fig. [137]10C, D). Thus, we selected si-CHAF1B1 and
   si-CHAF1B3 to transfect A549 and H1975 cells respectively. Using the
   CCK8 test, we discovered that the cell viability of si-CHAF1B3 and
   si-CHAF1B1 groups was reduced in A549 and H1975 cells
   (Fig. [138]10E−F). Transwell migration and invasion assays indicated
   that the knockdown of CHAF1B3 and CHAF1B1 resulted in decreased
   migration and invasion of A549 and H1975 cells (Fig. [139]10G−H).
   Furthermore, CHAF1B knockdown led to decreased glucose consumption and
   lactate metabolism in A549 and H1975 cells (Fig. [140]10I–L).
   Therefore, the knockdown of CHAF1B may inhibit the proliferation,
   migration, invasion, and glycolysis of A549 and H1975 cells.

Fig. 10.

   [141]Fig. 10
   [142]Open in a new tab

   CHAF1B promotes the proliferation, migration, invasion and glycolysis
   of LUAD cells. A CHAF1B mRNA expression in LUAD along with adjacent
   tissues was found with qRT-PCR. B The expression of CHAF1B mRNA in
   A549, H1975 and 95D cells was detected by qRT-PCR. C−D qRT-PCR was used
   to detect the interference effects of si-CHAF1B1, si-CHAF1B2 and
   si-CHAF1B3. E−F CCK8 was used to determine the effect of CHAF1B on the
   proliferation of LUAD cells. G–H Transwell assay was used to detect the
   effects of CHAF1B on the migration and invasion ability of LUAD cells.
   I–L Glycolysis assay was used to detect the effect of CHAF1B on the
   glycolysis of LUAD
   cells.*P < 0.05,**P < 0.01,***P < 0.001,****P < 0.0001

Discussion

   LUAD is the most common and aggressive form of lung cancer,
   characterized by the rapid growth and spread of cancerous cells in the
   lungs. Its diagnosis and treatment are relatively complex and highly
   heterogeneous. Therefore, identifying new biomarkers of LUAD is crucial
   detection and treatment targets [[143]26, [144]27]. As a member of the
   CAF-1 family, CHAF1B is a key regulator of chromosomal assembly after
   DNA synthesis and repair [[145]28]. Previous studies have indicated
   that CHAF1B is significantly overexpressed in melanoma tissues and is
   correlated with the spread of cancer to the skin, lymph nodes, and
   distant organs. This suggests that CHAF1B plays a critical role in the
   aggressive behavior of melanoma cells, potentially serving as a
   biomarker of metastatic potential [[146]29]. Additionally, high
   expression of CHAF1B has been shown to reduce the sensitivity of LUAD
   cells to cisplatin by inducing NCOR2 degradation; however, its
   biological function and molecular mechanism in the development of LUAD
   have not been fully understood [[147]30]. On this basis, we used TCGA,
   GEPIA2, Metascape, other databases, qRT-PCR, CCK8, Transwell, and other
   experimental methods to explore CHAF1B as a potential biomarker and
   therapeutic target for LUAD. Our research focused on gene expression,
   predictive ability, immune infiltration, drug sensitivity, single-cell
   biological function, construction of the LncRNAs-miRNA-CHAF1B axis and
   molecular mechanisms.

   Based on GO enrichment analysis and PPI network analysis, CHAF1B and
   its co-expressed genes were enriched in some pathways, such as the
   mitotic cell cycle, cell cycle, cell cycle regulation, chromosome
   organization, and cell cycle checkpoints. The CancerSEA database also
   showed that the biological function of CHAF1B in LUAD single cells was
   mainly related to DNA repair and the cell cycle. These findings suggest
   that CHAF1B may increase the malignancy of tumors by promoting the
   proliferation and cell cycle of LUAD cells.

   The tumor immune microenvironment (TME) is typically defined as the
   milieu surrounding the tumor, consisting predominantly of immune cells
   and cytokines produced by tumor cells. Studies have indicated a
   correlation between the extent of immune cell infiltration in tumor
   tissues and tumor malignancy of tumors [[148]31]. Zeng et al. [[149]32]
   observed a positive correlation between Th2 cell infiltration and
   glycolysis in LUAD cells. Patients with LUAD with high levels of Th2
   cell expression generally have a poor prognosis and reduced sensitivity
   to immunotherapy. Our results showed that CHAF1B expression positively
   correlated with Th2 cell infiltration and glycolysis. Therefore, we
   hypothesized that increased CHAF1B expression in LUAD cells may
   exacerbate glycolysis by promoting Th2 cell infiltration, ultimately
   leading to poor prognosis and immune resistance. Immune checkpoints are
   essential for tumor development and checkpoint inhibition can block the
   immune escape of cancer cells [[150]33]. Our study showed that CHAF1B
   expression was positively correlated with the immune checkpoints PDCD1,
   CD27, LAG, and SIGLEC. These correlations suggest that CHAF1B could
   become a target for LUAD immunotherapy.

   Using the CellMiner database, we analyzed the correlation between
   CHAF1B and the sensitivity of drugs. The results showed that patients
   with high expression of CHAF1B may increase resistance to Nelarabine,
   Vorinostat, Methylprednisolone, and 6-Thioguanine. In addition,
   patients with overexpression of CHAF1B may enhance drug sensitivity to
   Alectinib, Defactinib, Elesclomol, PF-06463922(lorlatinib), Celecoxib
   Isotretinoin, brigatinib, and Olaparib. Clinical trials have shown that
   Alectinib, PF-06463922(lorlatinib), and brigatinib have high efficacy
   in patients with advanced anaplastic lymphoma (ALK)-positive NSCLC
   [[151]34–[152]36]. It can be seen that advanced ALK-positive LUAD
   patients with higher CHAF1B expression levels may be more suitable for
   Alectinib, PF-06463922(lorlatinib), and brigatinib treatment.

   To confirm the reliability of the bioinformatics analysis, we performed
   in vitro experiments to verify the expression and biological function
   of CHAF1B in LUAD. The experimental results matched the predictions
   from bioinformatics analysis, showing significant CHAF1B expression in
   LUAD tissues. After interfering with CHAF1B expression using siRNA, the
   viability, migration, invasion ability, and glycolysis level of LUAD
   cells decreased, further proving the accuracy of our previous
   prediction results. This indicates that CHAF1B plays an important
   biological role in the occurrence and development of LUAD.

   LncRNAs, a subclass of non-coding RNA molecules with lengths exceeding
   200 nucleotides, function as competitive endogenous RNA to sequester
   miRNAs and modulate the expression of downstream target mRNAs
   [[153]37–[154]39]. We explored the target miRNAs and target lncRNAs
   that regulate CHAF1B. The results showed that hsa-miR-29c-3p,
   hsa-miR-145-5p and hsa-miR-1247-5p were lowly expressed in LUAD, which
   may be the target miRNA regulating CHAF1B. LncRNA [155]AL139287.1,
   NEAT1 and SHG1 are highly expressed in LUAD, which may be the target
   lncRNA regulating the expression of CHAF1B. Previous studies have shown
   that miR-29c-3p is lowly expressed in LUAD and is associated with poor
   prognosis in LUAD patients. Overexpression of miR-29c-3p can inhibit
   the proliferation, migration, invasion and glycolysis of LUAD cell
   lines [[156]40]. It has been reported that LncRNA NEAT1 is highly
   expressed in LUAD and is associated with poor prognosis of patients.
   Knockdown of NEAT1 expression can inhibit the malignant progression of
   LUAD [[157]41]. MiR-29c-3p and lncRNA NEAT1 were identified as miRNAs
   and lncRNAs for subsequent major analysis, respectively. We will verify
   the interaction between them and the related molecular mechanism in
   future experiments.

   Our study found that the expression of CHAF1B was increased in LUAD and
   may be used as a biomarker for diagnosis and prognosis of LUAD.
   Additionally, interfering with CHAF1B expression effectively suppressed
   the proliferation, migration, invasion, and glycolysis of LUAD cells.
   Notably, CHAF1B expression was also found to correlate with immune cell
   infiltration and drug sensitivity in LUAD, suggesting its potential
   role in predicting response to immunotherapy and chemotherapy. These
   results suggest that CHAF1B can be used as a potential biomarker and
   new therapeutic target for LUAD. However, owing to the lack of in vitro
   and in vivo experiments to verify the effect of CHAF1B on LUAD cell
   immunity, the research has limitations. In future studies, we will
   further verify the effect of CHAF1B on LUAD cell immunity and drug
   sensitivity through in vitro and in vivo experiments.

Conclusion

   Through comprehensive bioinformatics analysis and experimental
   validation, we examined the expression, biological function, and
   molecular mechanism of CHAF1B in LUAD and identified it as a possible
   biomarker and novel therapeutic target.

Acknowledgements