Introduction

Lung cancer has the highest mortality rate of all cancers, and approximately 2.09 million deaths from lung cancer are reported annually worldwide [1]. The 5-year survival rate for lung cancer is only 16%. Lung adenocarcinoma (LUAD) is the most common lung cancer subtype, accounting for 50% of all lung cancer pathological types, and the number one cause of cancer-related death worldwide [2]. Despite the ongoing development of medical technology and progress in diagnostic imaging, minimally invasive thoracoscopic surgery and postoperative adjuvant therapy, the incidence of LUAD is increasing each year and patient prognosis remains unsatisfactory. Therefore, better understanding of the molecular mechanisms of LUAD is critical to identify new effective therapy modalities and prognostic biomarkers.

Platelet-activating factor acetylhydrolases (PAF-AHs) are a class of diverse isozymes involved in a variety of physiological and pathological processes including prognosis, angiogenesis, inflammatory responses, wound healing, and tumor growth. Previous studies have found that high expression of PAFAH1B3 gene affects the proliferation and migration of gastric cancer cells and is an oncogene of gastric cancer [3]. The results show that PAFAH1B3 promotes the proliferation, migration and invasion capacity of papillary thyroid carcinoma cells(PTC), and the increase of PAFAH1B3 expression is likely to trigger tumor cell lymph node metastasis in PTC patients [4].

Increasing evidence has shown that PAFAH1B3 plays an important role in the development and progression of several human malignancies, including LUAD [5]. However, knowledge of PAFAH1B3 in LUAD is limited, and more research is required.

Some ncRNAs (including miRNAs, lncRNAs and circRNAs) interact through the ceRNA mechanism and participate in the regulation of gene expression [6]. miRNAs bind to the 3′ and 5′ UTRs of target mRNAs and inhibit target gene expression. lncRNAs can increase mRNA expression by competitively binding to shared miRNAs [7]. We speculate that PAFAH1B3 may also be regulated by this mechanism, and therefore we will initially explore its regulatory mechanism.

Methods

Analysis of TCGA data

mRNA expression data from 33 cancer types was retrieved from TCGA database (https://portal.gdc.cancer.gov/) (Fig. 1A). Data were then normalized and evaluated for differential expression of PAFAH1B3 using the R program limma. TCGA database was also used to examine the expression levels of hsa-miR-29c-3p and hsa-miR-101-3p in LUAD and normal controls. TCGA database was also used for survival analysis of LUAD in accordance with hsa-miR-29c-3p expression.

Fig. 1
figure 1

Analysis of PAFAH1B3 expression in cancer. A The expression of PAFAH1B3 in 33 types of human cancer in TCGA data. B–M PAFAH1B3 expression in GEPIA database. ∗P < 0.05; ∗∗P < 0.01; ∗∗∗P < 0.001

Full size image

Analysis of the GEPIA database

GEPIA is a web-based tool for profiling and analyzing interactions between genes in both cancer and normal samples [8]. The expression of PAFAH1B3 and lncRNAs in several types of human malignancies was examined using GEPIA. Survival analysis of PAFAH1B3, including overall survival (OS) and disease-free survival (DFS), was performed in 12 different cancer types using GEPIA. The predictive ability of potential lncRNAs in LUAD was evaluated using GEPIA. The GEPIA database was also used to evaluate the connection between PAFAH1B3 and immune cell biomarkers. The selection criteria |R|> 0.1 and P-value less than 0.05 were used to determine statistical significance.

Prediction of miRNAs

Seven target gene prediction algorithms, including PITA, RNA22, miRmap, DIANA-microT, miRanda, PicTar, and TargetScan, were used to predict the upstream binding miRNAs of PAFAH1B3. We focused on miRNAs that were identified in two or more of the target gene prediction tools mentioned above. The collected miRNAs were considered as candidate miRNAs that regulate PAFAH1B3.

Analysis of Starbase database

Starbase (http://starbase.sysu.edu.cn/) is a database used for studying miRNAs [9]. This database was used for miRNA-PAFAH1B3, lncRNA-hsa-miR-29c-3p, lncRNA-hsa-miR-101-3p, and lncRNA-PAFAH1B3 expression correlation analysis in LUAD. Starbase was used to predict candidate lncRNAs that may bind to let-7c-5p.

TIMER database analysis

TIMER (https://cistrome.shinyapps.io/timer/) is a database used for the analysis of tumor infiltrating immune cells [10]. TIMER was used to analyze the correlation between PAFAH1B3 expression levels and the level of immune cell infiltration in LUAD. The correlation of PAFAH1B3 expression with pyroptosis genes in LUAD was also assessed using the TIMER database.

Immunohistochemistry

We collected 60 cases of postsurgical lung adenocarcinoma tissue and normal lung tissue (25 men, 35 women, aged between 45 and 60 years) from the Department of Thoracic Surgery, Suzhou Hospital, Nanjing Medical University.Tissue sections were first treated in 3% H2O2 for 15 min at room temperature. After blocking with goat serum for 1 h, the slides were treated with primary antibodies against PAFAH1B3 (1:100, Abcam,Cambridge, England, ab241288). The sections were rinsed three times in PBS and incubated with goat-anti-rabbit IgG secondary antibodies (China Fuzhou Maixin Biotech). The slices were washed three times in PBS for 5 min each and then treated with streptavidin-conjugated HRP (China Fuzhou Maixin Biotech). Images were captured with an Olympus camera and corresponding software(Leica: BOND MAX, Switzerland). Two independent pathologists graded the staining results. The staining intensity of the sections was quantified by the IHC tool in Image J software (National Institutes of Health, America). Staining intensity was quantified as 0 (no staining), 1 + (weak staining), 2 + (moderate staining) or 3 + (strong staining) and frequency was quantified as the percentage of relevant cells (e.g., viable tumor cells) exhibiting staining at the appropriate subcellular location. H-scores were then calculated based on the percentage of stained cells at each intensity level: H-score = cell staining intensity score × percentage of staining.

Quantitative real-time PCR

Total RNA was isolated using TRIzol from A549,H460,H1299,H292,BEAS-2B cells (ATCC,America). Reverse transcription was performed using a PCR amplifier (ABI, USA) to generate cDNA. For quantitative PCR experiments, the following reaction conditions were used: pre-denaturation at 95 ℃ for 10 min, and 40 cycles of denaturation at 95 ℃ for 10 s, annealing at 60 ℃ for 20 s, and extension at 72 ℃ for 34 s. The Ct value (threshold cycle) was determined by manually setting the threshold at the lowest point of each logarithmic amplification curve's parallel climb. β-actin mRNA was used as the internal reference. Data were analyzed using the 2-ΔΔCt method. PCR primers were synthesized by Shanghai Bioengineering Technology Service Co., Ltd. and the primer sequences were as follows: PAFAH1B3: F: 5′-CGGATCAAGAGATTCCAGGTTC-3′, R: 5′-CCGGAGGGAGAGGATTCTT-3′.

Transwell invasion assay

A549 cells were inoculated at 5×103 cells/cm2 in cell culture dishes and the medium was changed every 2–3 days until cells achieved 70%–90% confluence. The cells (1 × 105/mL) were then inoculated in Transwell chambers (coated with 500 µL of Matrigel gel). The cells in the top chambers were incubated in serum-free medium and medium containing 10% FBS was included in the bottom chamber. Plates were incubated for 48 h. The chambers were removed, and cells were fixed in paraformaldehyde at 4 ℃ for 30 min and left to stand at room temperature for 20 min. Next, 750 µL of 0.1% crystalline violet solution was added and cells were incubated at room temperature for 20 min. The cells were washed three times with water and the number of invaded cells was counted under a microscope (Zeiss Germany).

CCK-8 assay

The OD of the standard was first measured at a wavelength of 450 nm using the cells to be measured and a standard curve was plotted. Cells were added to a 96-well plate at 100 µL per (105/mL) well and incubated at 37 ℃ with 5% CO2. Next, 10 µL of CCK-8 solution was added to each well and cells were incubated for 2 h. The OD was recorded by measuring the absorbance at 450 nm using a microplate reader. Cell viability was calculated as follows =([As-Ab] / [Ac-Ab]) × 100%.

shRNA and cell transfection

The pLKO.1G GFP-shRNA plasmid for PAFAH1B3 was purchased from Addgene. The PAFAH1B3-specific shRNA sequences were: sh-PAFAH1B3#1: 5′- CACCGGAAGCGAAGGTTCCTGATTCCTCGAGGAATCAGGAACCTTCGCTTCC-3′; and sh-PAFAH1B3#2: 5′- CACCGCCTTCCCACAACATTAAACTCTCGAGAGTTTAATGTTGTGGGAAGGC-3′. Cells were transfected with shRNA (1 µg, 50 pmol) using Lipofectamine 2000 following the manufacturer’s instructions. Cells were incubated for 48 h and then examined by qRT-PCR.

Wound healing assay

Single cell suspensions were inoculated in 6-well culture plates at 0.5 cm intervals. The cells were rinsed three times with PBS to remove the scratched cells and serum-free medium was added. The marker was finally wiped from the back of the 6-well plate and photographed under a 4 × microscope. Cells were photographed at 24 and 48 h. After subsequently opening the images using Image J software, six horizontal lines were randomly scratched and the mean average of the intercellular distances was calculated. Cell migration rate (wound healing rate) = (mean value of initial intercellular distance—mean value of intercellular distance at time t)/mean value of initial intercellular distance.

Statistical analysis

The statistical analyses in this study were generated automatically via the online database stated above. P-values less than 0.05 or log-rank P-values less than 0.05 were considered statistically significant.

Results

Pan-cancer analysis of PAFAH1B3 expression

To evaluate the potential function of PAFAH1B3 in carcinogenesis, we first examined its expression in 33 human cancers using UCSC XENAdata in TCGA database. PAFAH1B3 was significantly elevated in 18 cancer types, including glioblastoma multiforme (GBM), uterine corpus endometrial carcinoma (UCEC), lung adenocarcinoma (LUAD), bladder uroepithelial carcinoma (BLCA), breast invasive carcinoma (BRCA), cervical and endocervical carcinoma (CESC), cholangiocarcinoma (CHOL), colon adenocarcinoma (COAD), esophageal carcinoma (ESCA), squamous cell carcinoma of the head and neck (HNSC), liver hepatocellular carcinoma (LIHC), ovarian serous cystadenocarcinoma (OV), lung squamous cell carcinoma (LUSC), stomach adenocarcinoma (STAD), thyroid carcinoma (THCA) and kidney renal clear cell carcinoma (KIRC), kidney renal papillary cell carcinoma (KIRP), liver hepatocellular carcinoma (LIHC), prostate adenocarcinoma (PRAD), rectal adenocarcinoma (READ), compared with paracancerous tissue. We then examined PAFAH1B3 expression in these 18 cancer types using the GEPIA database. PAFAH1B3 expression was significantly higher in the following cancer types compared with corresponding non-cancerous tissue samples: BLCA, BRCA, CHOL, COAD, LIHC, LUAD, LUSC, PRAD, READ, STAD, THCA, and UCEC (Fig. 1B–M).

The prognostic values of PAFAH1B3 in human cancer

We next performed survival analysis in accordance with PAFAH1B3 expression in BLCA, BRCA, CHOL, COAD, LIHC, LUAD, LUSC, PRAD, READ, STAD, THCA, and UCEC using the GEPIA database. High expression of PAFAH1B3 in LIHC and LUAD was prognostically unfavorable in terms of OS (Fig. 2). Increased expression of PAFAH1B3 in all cancer types also indicated a poor prognosis for LIHC and LUAD in terms of DFS (Fig. 3). PAFAH1B3 did not indicate prognosis in other cancer types. These findings indicated that PAFAH1B3 may be a biomarker for poor prognosis in patients with LUAD.

Fig. 2
figure 2

Overall survival (OS) of cancer patients in accordance with PAFAH1B3 expression in GEPIA database

Full size image
Fig. 3
figure 3

Disease-free survival (DFS) of cancer patients in accordance with PAFAH1B3 expression in GEPIA database

Full size image

PAFAH1B3 is highly expressed in LUAD and correlates with the malignancy of the tumor

Our results above showed that PAFAH1B3 was significantly highly expressed in lung adenocarcinoma and closely associated with the prognosis of lung adenocarcinoma patients. We next performed immunohistochemistry of PAFAH1B3 in tumor tissues and normal lung tissues from 60 lung adenocarcinoma patients and found that PAFAH1B3 was significantly highly expressed in the lung adenocarcinoma tissues compared with normal lung tissues (Fig. 4A, B), which was consistent with our previous findings.

Fig. 4
figure 4

PAFAH1B3 was highly expressed in lung adenocarcinoma and its downregulation reduced malignant activities of lung adenocarcinoma cells. A Immunohistochemistry of PAFAH1B3 in lung adenocarcinoma and normal lung tissues; B PAFAH1B3 expression was significantly lower in tumor tissues compared with normal tissues; C qRT-PCR assay showed that PAFAH1B3 expression was higher in human lung cancer cell lines than in human lung normal epithelial cells; D The expression of PAFAH1B3 in A549 cells was greatly reduced after lentivirus-mediated knockdown of PAFAH1B3; E CCK-8 assay revealed that A549 cell growth was inhibited following PAFAH1B3 knockdown; F Scratch assay showed that the migration ability of A549 cells was significantly reduced after knockdown of PAFAH1B3; G–H Transwell assay showed that the invasive ability of A549 cells was significantly reduced after knockdown of PAFAH1B3

Full size image

We next explored the role of PAFAH1B3 in lung adenocarcinoma through in vitro experiments. qRT-PCR experiments suggested that PAFAH1B3 was expressed at the highest level in the A549 cell line compared with human lung normal epithelial cells BEAS-2B (Fig. 4C). We constructed an A549 cell line with knockdown of PAFAH1B3 and verified the successful knockdown by qRT-PCR (Fig. 4D). CCK-8, wound healing and Transwell assays showed that the proliferation, migration and invasive abilities of A549 cells were significantly inhibited after PAFAH1B3 knockdown (Fig. 4E–G).

Prediction and search for upstream miRNAs of PAFAH1B3

Studies have shown that ncRNAs play a key role in controlling gene expression. To evaluate whether PAFAH1B3 is regulated by miRNAs, we used the starBase database to predict the potential miRNAs that may bind to PAFAH1B3, and the results identified 14 miRNAs. We used Cytoscape software to create a miRNA-PAFAH1B3 regulatory network (Fig. 5A). miRNAs bind the 3′ and 5′ UTRs of target mRNAs to suppress the expression of the target gene. We performed expression correlation analysis and found that PAFAH1B3 was significantly negatively linked with hsa-miR-29c-3p and hsa-miR-101-3p and positively correlated with hsa-miR-301a-3p and hsa-miR-3619-5p in LUAD (Fig. 4B). We then investigated the expression and prognostic value of hsa-miR-29c-3p and hsa-miR-101-3p in LUAD using TCGA LUAD (lung adenocarcinoma) miRNAseq database. The results showed that hsa-mi-R-29c-3p was significantly downregulated in LUAD and its upregulation was positively correlated with patient prognosis (Fig. 5C, D). hsa-miR-101-3p was significantly downregulated in LUAD but its up-regulation was not correlated with patient prognosis (Fig. 5E, F). These findings suggest that hsa-miR-29c-3p may play a role in regulating PAFAH1B3 in LUAD.

Fig. 5
figure 5

hsa-miR-29c-3p is a potential upstream miRNA of PAFAH1B3 in LUAD. A A regulatory network of miRNA and PAFAH1B3 was created using Cytoscape software. B Starbase database analysis of the association between predicted miRNAs and PAFAH1B3 in LUAD. C, D The expression and prognostic significance of hsa-miR-29c-3p in LUAD and normal control samples was examined in TCGA database. E, F The expression and prognostic significance of hsa-miR-101-3p in LUAD and normal control samples was examined in TCGA database

Full size image

Prediction and analysis of lncRNAs upstream of hsa-miR-29c-3

The Starbase database was then used to examine potential upstream lncRNAs of hsa-miR-29c-3p and 30 potential lncRNAs were identified. The Cytoscape program was used to build the hsa-miR-29c-3p regulatory network (Fig. 6A). The expressions of these lncRNAs in LUAD were then assessed using GEPIA. Only double homeobox A pseudogene 8 (DUXAP8) was highly elevated in LUAD compared to normal controls (Fig. 6B). The prognostic significance of DUXAP8 in LUAD was evaluated. LUAD patients with increased DUXAP8 expression had a shorter life expectancy (Fig. 6C). Some lncRNAs function as competitive endogenous RNAs (ceRNAs) by competitively binding to miRNAs resulting in increased mRNA expression. The expression association between DUXAP8 and hsa-miR-29c-3p or PAFAH1B3 in LUAD was further investigated using the Starbase database (Table 1). From the results of expression, survival, and correlation analyses, DUXAP8 may be a promising upstream lncRNA of the hsa-miR-29c-3p/PAFAH1B3 axis in LUAD.

Fig. 6
figure 6

Expression analysis and survival analysis of lncRNAs upstream of hsa-miR-29c-3p in LUAD. A The lncRNA regulatory network of hsa-miR-29c-3p. B, C Expression of hsa-miR-29c-3p in TCGA LUAD compared with “TCGA normal” or “TCGA and GTEx normal” data. D OS analysis of LUAD in accordance with DUXAP8 expression. *P< 0.05

Full size image
Table 1 Correlation analysis between lncRNA and hsa-miR-29c-3p or lncRNA and PAFAH1B3 in LUAD determined by Starbase database
Full size table

The expression of DUXAP8 was negatively correlated with hsa-miR-29c-3p and positively with PAFAH1B3 expression. This fits with the mechanism by which ncRNAs participates in the regulation of gene expression through a ceRNA manner.

PAFAH1B3 corresponds with immune cell infiltration in LUAD

PAFAH1B3 also serve an important role in the immune system. We observed substantial alterations in the levels of B cells, CD4+ T cells, macrophages, neutrophils, and dendritic cell infiltration in LUAD with different copy numbers of PAFAH1B3 (Fig. 7A). Correlation analysis may potentially provide more pertinent insights regarding the function and mechanism of PAFAH1B3. We thus examined the relationship between PAFAH1B3 expression and immune cell infiltration. As shown in Fig. 7B–G, PAFAH1B3 expression showed a significant negative correlation with all immune cells analyzed, including B cells, CD8+ T cells, macrophages, neutrophils and dendritic cells in LUAD.

Fig. 7
figure 7

Relationship between immune cell infiltration and PAFAB1H3 expression in LUAD. A Infiltration levels of various immune cells in LUAD with different copy numbers of PAFAB1H3. B–G Correlation of PAFAH1B3 expression levels in HCC with infiltration levels of B cells (B), CD8+ T cells (C), CD4+ T cells (D), macrophages (E), neutrophils (F) and dendritic cells (G)

Full size image

Correlation of PAFAH1B3 expression with biomarkers of immune cells in LUAD

To further investigate the role of PAFAH1B3 in immune cell infiltration in LUAD, we determined the correlation of PAFAH1B3 expression with immune cell biomarkers in LUAD using the GEPIA database. We observed a correlation between PAFAH1B3 and biomarkers of B cells (CD19 and CD79A), biomarkers of CD8+ T cells (CD8A and CD8B), biomarkers of CD4+ T cells (CD4), biomarkers of M1 macrophages (NOS2, IRF5 and PTGS2), biomarkers of M2 macrophages (CD163, VSIG4 and MS4A4A), biomarkers of neutrophils (ITGAM and CCR7) and biomarkers of dendritic cells (HLA-DPB1, HLA-DRA, HLA-DPA1, CD1C, NRP1 and ITGAX) (Table 2). These findings provide further evidence that PAFAH1B3 has an important association with immune cell infiltration in LUAD.

Table 2 Correlation analysis between PAFAH1B3 and biomarkers of immune cells in LUAD determined by GEPIA database
Full size table

PAFAH1B3 was negatively associated with B-cell biomarkers (CD79A), CD8+ T-cell biomarkers (CD8A), CD4+ T-cell biomarkers (CD4), M1 macrophage biomarkers (IRF5), biomarkers for M2 macrophages (CD163, VSIG4 and MS4A4A), biomarkers for neutrophils (ITGAM and CCR7) and biomarkers of dendritic cells (HLA-DPB1, HLA-DRA, HLA-DPA1, CD1C, NRP1 and ITGAX) were negatively correlated.

Relationship between PAFAH1B3 and pyroptosis genes of LUAD

As we mentioned above, platelet-activated factor acetylhydrolases (PAF-AHs) are a variety of isoenzymes involved in a variety of physiological and pathological processes, including the inflammatory response. We had to consider the possible involvement of PAFAH1B3 in apoptosis, which required further evaluating the relationship between PAFAH1B3 and cell apoptosis in lung adenocarcinoma cells. NOD-like receptor thermal protein domain associated protein 3(NLRP3), NOD-like receptor thermal protein domain associated protein 2(NLRP1), caspase 1(CASP1) and Interferon regulatory factor 2(IRF2) are important cell scorch proteins responsible for pyroptosis. Cell pyroptosis is closely associated with tumor progression. Considering the high expression of PAFAH1B3 as a potential oncogenic and poor prognostic factor in LUAD, we evaluated the relationship between PAFAH1B3 and NLRP3, NLRP1, CASP1 and IRF2 using the TIMER database. As shown in Fig. 8A–D, the expression of PAFAH1B3 was significantly negatively correlated with NLRP3, NLRP1, CASP1 and IRF2 in LUAD. We further used the GEPIA database to verify these associations and found that PAFAH1B3 was significantly negatively correlated with NLRP3, NLRP1, CASP1 and IRF2 in LUAD (Fig. 8E–H). Subsequently, we used TCGA database to further analyze the differential expression of NLRP1, NLRP3, NLRP1 and CASP1 and IRF2 in lung adenocarcinoma tissues and normal lung tissues, and simultaneously explored the survival curves of these pyroptotic genes, and finally found that NLRP 3, NLRP 1, CASP1 and IRF2 were significantly low expression in lung adenocarcinoma tissues and NLRP3 (Fig. 9A-D) and low expression of NLRP1, NLRP3 predicted poor patient prognosis (Fig. 9E-H). Relevant studies have found that NLRP1, NLRP3, IRF2, and CASP1 are widely involved in the pyroptosis and immune infiltration of various tumors [11]. And CASP1 can inhibit the growth and invasion of non-small-cell lung cancer (NSCLC) [12]. These results suggest that the high expression of PAFAH1B3 inhibits the pyroptosis of LUAD cells, leading to tumor progression and poor prognosis.

Fig. 8
figure 8

Correlation of PAFAH1B3 expression with NLRP3, NLRP1, CASP1 and IRF2 expression in LUAD. A–D Spearman correlation analysis of PAFAH1B3 with NLRP3, NLRP1, CASP1 and IRF2 expression in LUAD. E–H Correlation of PAFAH1B3 expression with NLRP3, NLRP1, CASP1 and IRF2 in LUAD as determined by GEPIA database

Full size image
Fig. 9
figure 9

Expression and prognosis analysis of NLRP 3, NLRP 1, and CASP1 and IRF2 in LUAD. A–D The Wilcoxon rank sum test was used to compare the differences between NLRP 3, NLRP 1, CASP1 and IRF2 in LUAD and normal lung tissue in TCGA database, and NLRP 3, NLRP 1, CASP1 and IRF2 were all low expressed in LUAD. E–H The expression of the target genes in LUAD patients is ranked in the TCGA database, with the first 50 percent set as low expression and the last 50 percent as high expression. The OS of NLRP 3, and NLRP 1 high and low expression groups were significantly different

Full size image

Discussion

In this study, we first analyzed PAFAH1B3 expression in pan-cancer using TCGA database and then used the GEPIA database to validate the expression of PAFAH1B3 observed in TCGA pan-cancer data. We further found that LUAD patients with high PAFAH1B3 expression had a poor prognosis. A previous study found PAFAH1B3 was an independent prognostic risk factor for LUAD [13]. This is consistent with the results of our study.

To explore the upstream regulatory miRNAs of PAFAH1B3, we used the starBase database to predict miRNAs that might bind to PAFAH1B3. In total, 14 candidate miRNAs were identified. Some of these miRNAs have been shown to function as tumor suppressor miRNAs in LUAD. For example, hsa-miR-24-3p was shown to decrease AVL9 expression, which inhibited tumor development in non-small cell lung cancer [14]. hsa-miR-101-3p inhibits LUAD cell growth by reducing CEP55 expression [15]. Survival of LUAD patients was highly related with hsa-miR-29c-3p [16]. Through correlation, expression, and survival analyses in this study, hsa-miR-29c-3p was identified as a potential upstream miRNA that suppresses PAFAH1B3.

Furthermore, the putative lncRNAs of the hsa-miR-29c-3p/PAFAH1B3 axis should be carcinogenic lncRNAs, according to the ceRNA hypothesis [17]. Consequently, we explored the lncRNAs upstream of the hsa-miR-29c-3p/PAFAH1B3 axis and identified 30 potential lncRNAs. Expression analysis, survival analysis, and correlation analysis identified DUXAP8 as an lncRNA that potentially regulates the hsa-miR-29c-3p/PAFAH1B3 axis. DUXAP8 has been shown to promote cancer by activating the Akt/mTOR signaling pathway [18]. DUXAP8 also influences the activity of miRNAs [19]. High DUXAP8 expression in NSCLC was shown to be related with poor patient prognosis, and DUXAP8 plays a role in NSCLC cell proliferation, epithelial-mesenchymal transition, and aerobic glycolysis [20]. In addition, it has also been found that DUXAP8 promotes the proliferation and migration of ovarian cancer cells by down-regulating the expression of microRNA-29. It has also been found that the DUXAP8-miR-29-3p axis can influence hepatocellular carcinoma immune infiltration [21]. This laterally validated our predicted DUXAP8-microRNA-29c-3p-PAFAH1B3 pathway [22].

In this study, PAFAH1B3 was found to be strongly adversely linked with numerous immune cells in LUAD. Furthermore, PAFAH1B3 was strongly inversely associated with biomarkers of these invading immune cells. The degree of tumor immune cell infiltration is closely connected to the prognosis of NSCLC [23]. These data suggest that aberrant PAFAH1B3 overexpression leads to abnormal LUAD immune cell infiltration and poor expression of biomarkers of immune cells.

Pyroptosis, a new form of non-apoptotic programmed cell death that is mediated by an inflammatory response, is associated with many diseases [24]. Pyroptosis was found to eliminate anti-apoptotic or anti-necrotic cancer cells in lung cancer [25]. Our results showed that high expression of PAFAH1B3 in LUAD correlated with reduced expression of NLRP3, NLRP1, Casp1 and IRF2, which are related to pyroptosis [26]. This further suggests that high expression of PAFAH1B3 will affect patient prognosis.

In conclusion, we discovered that PAFAH1B3 is highly expressed in a variety of human malignancies, including LUAD, and that its high expression in LUAD correlates with poor prognosis. Our results suggest that the DUXAP8-hsa-miR-29c-3p axis may regulate PAFAH1B3 in LUAD (Fig. 10). Furthermore, we discovered that PAFAH1B3 exerts tumor promoting functions by reducing tumor immune cell infiltration and the expression of related localized death genes. These findings should be verified by additional research and clinical trials.

Fig. 10
figure 10

The DUXAP8-hsa-miR-29c-3p-PAFAH1B3 axis in lung adenocarcinoma

Full size image