Abstract Background Kindlin Family Members have been reported to be aberrantly expressed in various human cancer types and involved in tumorigenesis, tumor progression, and chemoresistance. However, their roles in non-small cell lung cancer (NSCLC) remain poorly elucidated. Methods We analyzed the prognostic value and immune infiltration of Kindlins in NSCLC through Oncomine, GEPIA, UALCAN, CCLE, Kaplan‑Meier plotter, cBioPortal, TIMER, GeneMANIA, STRING, and DAVID database. Additionally, the mRNA expression levels of Kindlins were verified in 30 paired NSCLC tissues and NSCLC cell lines by real-time PCR. Results The expression level of FERMT1 was remarkably increased in NSCLC tissues and NSCLC cell lines, while FERMT2 and FERMT3 were reduced. Kindlins expressions were associated with individual cancer stages and nodal metastasis. We also found that higher expression level of FERMT1 was obviously correlated with worse overall survival (OS) in patients with NSCLC, while higher FERMT2 was strongly associated with better overall survival (OS) and first progression (FP). Additionally, the expression of FERMT2 and FERMT3 were obviously correlated with the immune infiltration of diverse immune cells. Functional enrichment analysis has shown that Kindlins may be significantly correlated with intracellular signal transduction, ATP binding and the PI3K-Akt signaling pathway in NSCLC. Conclusions The research provides a new perspective on the distinct roles of Kindlins in NSCLC and likely has important implications for future novel biomarkers and therapeutic targets in NSCLC. Supplementary Information The online version contains supplementary material available at 10.1186/s12920-021-00967-2. Keywords: Kindlins, Non-small cell lung cancer, Bioinformatics, Prognosis Introduction Lung cancer is the most common cancer and the leading cause of cancer mortality, causing approximately 1,760,000 deaths worldwide each year [[39]1]. Non-small cell lung cancer (NSCLC) accounts for approximately 80–85% of lung cancers, including lung adenocarcinoma (LUAD), lung squamous cell carcinoma (LUSC) and lung large cell carcinoma (LCC) [[40]2]. Although gradual improvements in diagnosis and treatment therapies, the 5-year overall survival rate of NSCLC remains poor [[41]3, [42]4]. Therefore, it is necessary to explore the mechanism underlying the tumorigenesis and progression of NSCLC and to search for novel biomarkers with high sensitivity and specificity. The Kindlin family members are newly discovered focal adhesion proteins consisting of three members (FERMT1, FERMT2 and FERMT3) that share a conserved FERM domain-containing three subdomains (F1, F2 and F3) with an inserted pleckstrin homology domain. Remarkably, Kindlins may function as tumor promoters or suppressors depending on the cancer type. For instance, FERMT2 severs as a tumor-promoting role in gastric cancer [[43]5], pancreatic cancer [[44]6, [45]7] and breast cancer [[46]8, [47]9], but as a tumor-suppressive role in epithelial ovarian cancer [[48]10] and colorectal cancer [[49]11]. Weinstein et al. reported a higher mRNA expression of FERMT1 in lung cancer tissues [[50]12]. Besides, Zhan et al. also found that FERMT1 and FERMT2 had different expressions in lung cancer. FERMT1 was highly expressed in NSCLC, especially in LUSC. On the other hand, FERMT2 was highly expressed in LCC and weakly expressed in LUAD and LUSC [[51]13]. Additionally, Djaafri et al. found that FERMT3 expression was decreased in lung cancer [[52]14]. However, the series of research only reported the expression levels of Kindlins, and few studies have been conducted on the prognosis and mechanism of Kindlins. Rapid advances in online platforms and various databases have contributed to the widely used bioinformatics analysis in the field of cancer research. Up to date, bioinformatics analysis has not been adequately used to investigate the roles of Kindlins in NSCLC. Hence, the present study investigated the role of individual kindlins in NSCLC using RT-PCR combined with bioinformatics. In this study, we aim to investigate the prognostic value and immune infiltration of Kindlins in NSCLC, providing new clues for early diagnosis, prognostic judgments and individualized treatments for NSCLC patients. Materials and methods Expression database of Kindlins in NSCLC Oncomine database ([53]http://www.oncomine.org) is a comprehensive cancer microarray database based on 86,733 samples from 715 databases for gene transcriptome analysis in different cancers [[54]15]. The screening and data mining conditions included: “Gene: FERMT1, FERMT2 and FERMT3”; “Analysis Type: Cancer versus Normal Analysis”; “Cancer Type: Lung Cancer”; “Data Type: mRNA”; “P value < 0.05”; “Fold change:2”; “Gene rank: top 10%”. GEPIA database ([55]http://gepia.cancer-pku.cn/) is a recently developed interactive online platform, applying for analyzing the RNA sequencing (RNA-Seq) expression based on over 9000 tumors from the TCGA and 8000 normal samples from the GTEx [[56]16]. GEPIA was used to analyze Kindlins expression and tumor stage in LUAD and LUSC. UALCAN is a user-friendly web that provides publicly available cancer transcriptome data based on TCGA database [[57]17]. UALCAN was used to confirm the association between mRNA expressions of Kindlins in NSCLC and clinicopathological parameters. CCLE ([58]www.broadinstitute.org/ccle) provides the public information on gene expression, chromosome copy number and mutation profile of 947 human cancer cells [[59]18]. In our research, we mainly use it to verify the expression levels of FERMT1, FERMT2 and FERMT3 in different kinds of cancer cell lines. Kaplan‑Meier plotter Kaplan Meier plotter ([60]http://kmplot.com/analysis) aims to discover and validate survival biomarkers based on a meta-analysis from 11 k samples from 20 different cancer types [[61]19, [62]20]. In the study, we used the KM plotter database from a set of 1926 lung cancer samples to assess Kindlins’ prognostic values, including overall survival (OS), first progression (FP) and post-progression survival (PPS). Moreover, we evaluated the associations of the Kindlins with various clinical parameters of NSCLC, including the histology, clinical stages, gender and smoking history. cBioPortal cBioPortal ([63]http://cbioportal.org) is an open website used to explore, visualize and analyze multilayer cancer genome data from over 5,000 tumors from 20 different cancer studies [[64]21, [65]22]. According to cBioPortal's online instructions, the Kindlins gene alterations information in different cancer types was obtained, including genetic mutations, gene fusions, gene amplifications, deep deletions and multiple alterations. TIMER TIMER ([66]https://cistrome.shinyapps.io/timer/) is a comprehensive database for analysis of the tumor-infiltrating immune cells from 32 cancer types [[67]23]. In our research, Spearman correlation was used in the gene module to explore the correlation between the expression of Kindlins and immune infiltration, including tumor purity and six types of immune cells (B cells, CD8+ T cells, CD4+ T cells, macrophages, neutrophils, and dendritic cells). GeneMANIA and STRING database GeneMANIA ([68]http://www.genemania.org) is a useful and flexible prediction server, displaying a functional interaction network to explore the association between genes and data sets [[69]24]. In the current study, we used GeneMANIA to analyze the relationships between Kindlins concerning the co-expression, co-localization, physical interactions, pathway, genetic interactions, prediction, and shared protein domains. STRING ([70]https://string-db.org/) is an online database, predicting protein–protein interactions network in terms of direct (physical) and indirect (functional) associations [[71]25]. In this study, we constructed the protein–protein interactions network of Kindlins using and the selection criteria: “organism: Homo sapiens”; “the minimum required interaction score > 0.4”; “the max number of interactors: 20”. DAVID database DAVID database ([72]https://david.ncifcrf.gov) is a comprehensive bioinformatics enrichment platform, providing integrative and systematic annotations of biological functions from a series of genes/proteins [[73]26, [74]27]. We applied the DAVID database (version 6.8) for gene ontology (GO) terms analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways enrichment analysis of Kindlins and their related proteins. GO terms covers three aspects: biological processes, cellular components and molecular functions. The P Value < 0.05 was set as a criterion and regarded as significant enrichment. Lung tissue samples In this study, between July 2019 and December 2019, thirty pairs of NSCLC tissue and adjacent normal tissues were collected from the Fuzhou Pulmonary Hospital, China. These tissues were used to detect the expression level of Kindlins mRNA by quantitative real-time PCR (RT-PCR). The collection and use of the samples were approved by the ethics committee of the Second Affiliated Hospital of Fujian Medical University. The approval number is 2019 (ethical research review)-207. Cell culture The NSCLC cell lines (A549, SPCA-1 and H1299) and one normal cell line (BEAS‐2B) were obtained from Procell life science and Technology Co., Ltd (Wuhan, China). Human bronchial epithelial cells (BEAS-2B) were obtained from FenghBio Co., Ltd. (Changsha, China). All the cell lines were cultured at RPMI-1640 medium (GIBCO, Los Angeles, CA, USA) with 10% fetal bovine serum (FBS, Gibco) and grown in a humidified incubator at 37 °C (5% CO2) environment. RT-PCR analysis Total RNA in tissues or cells was extracted using TRIzol reagent (Invitrogen) and cDNA was synthesized using Primescript RT Reagent (Takara Bio Inc., Japan). PCR amplification was performed using the SYBR Green PCR kit (Takara Bio Inc., Japan) in a 7500 PCR system (Thermo Fisher Scientific), and GAPDH was used as an endogenous control. The relative quantification analysis was performed using the comparative CT method. The following PCR primers were used: * FERMT1 forward, 5′-TTGAAGATGGTGAGGTTGCGAGTC-3′ * FERMT1 reverse, 5′-GGGTTGGCTGAATGCGAGGATG-3′ * FERMT2 forward, 5′-TGGCTCTGGACGGGATAAGGATG-3′ * FERMT2 reverse, 5′-TTTGTGCTGAGGGGTGAACTGAAG-3′ * FERMT3 forward, 5′-ACTGCACCGAGGAGGAGATGATG-3′ * FERMT3 reverse, 5′-CCTTGAGGTTGAGCTGCTGAATGG-3′ * GAPDH forward, 5′-CTCCTGCACCACCAACTGCTTAG-3′ * GAPDH reverse, 5′-GACGCCTGCTTCACCACCTTC-3′ Statistical analysis RT-PCR was performed in triplicate. The data were analyzed using the GraphPad Prism (version 8.0). Student’s t-test was used to compare the expression of Kindlins mRNA between NSCLC tissue samples tissues and adjacent normal tissues. P < 0.05 was considered a statistically significant difference. Results Differential expression of Kindlins in NSCLC patients and cell lines We first used ONCOMINE database to explore the expression of Kindlins in NSCLC. Multiple datasets showed that the mRNA expression level of FERMT1 was significantly increased in NSCLC tissues, while FERMT2 and FERMT3 were reduced in NSCLC versus normal tissues (Fig. [75]1a and Table [76]1,). These database include the Hou’s dataset [[77]28], Su dataset [[78]29], Okayama dataset [[79]30], Wachi dataset [[80]31], Selamat’s dataset [[81]32], Landi’s dataset [[82]33], and Stearman’s dataset [[83]34]. Moreover, Meta-analysis of Kindlins genes expression in NSCLC studies from Oncomine databases was consistent with the above results (Fig. [84]1b). Then, we used the GEPIA and UALCAN dataset to further confirm these findings. The results indicated that the expression level of FERMT1 was overexpressed in LUAD and LUSC tissues than in normal tissues, while FERMT2 and FERMT3 were decreased inversely (Fig. [85]1c–e). We also used CCLE to explore expression levels of Kindlins in NSCLC cell lines. As presented in Fig. [86]2a, the expression level of FERMT1 was higher in NSCLC cell lines and FERMT3 was downregulated. These results were consistent with those from Oncomine, GEPIA and UALCAN dataset. Different from these results, FERMT2 expression was increased in NSCLC, which requires further examination. Fig. 1. [87]Fig. 1 [88]Open in a new tab The expression of Kindlins in NSCLC. a The expression of Kindlins in different types of cancers compared with normal tissues (ONCOMINE). b Meta-analysis of FERMT1, FERMT2 and FERMT3 mRNA expression in NSCLC from multiple Oncomine databases. c The expression profile of Kindlins in LUAD and LUSD (GEPIA). Red trace, tumor samples; green trace, normal samples; T, tumor; N, normal. d The expression boxplots of Kindlins in LUSD and LUAD (GEPIA). A t-test was used to compare the expression level differences between tumor and normal tissues (P < 0.01). Y-axis represents log2 (TPM+ 1). Red box, tumor samples; black box, normal samples. T, tumor; N, normal. e The relative expression of Kindlins in LUAD and LUSD (UALCAN) Table 1. The significant changes of Kindlis in transcription level (ONCOMINE database) Comparison groups Fold change P value t-test Sample size References