MicroRNA Signatures in Lung Adenocarcinoma Metastases: Exploring the Oncogenic Targets of Tumor-Suppressive miR-195-5p and miR-195-3p

Simple Summary

To identify genes involved in lung cancer brain metastasis, we generated microRNA signatures using lung cancer brain metastasis tissues. Based on the microRNA signatures in lung adenocarcinoma metastases, both strands of pre-miR-195 (miR-195-5p and miR-195-3p) were significantly downregulated in metastatic tissues. A total of 12 genes (ANLN, CDC6, CDCA2, CDK1, CEP55, CHEK1, CLSPN, GINS1, KIF23, MAD2L1, OIP5, and TIMELESS) were identified as pre-miR-195-controlled genes, and these genes greatly contributed to the molecular pathogenesis of lung adenocarcinoma cells. Knockdown assays using siRNAs revealed that ANLN and MAD2L1 facilitated the malignant transformation of lung adenocarcinoma cells. Analysis of the microRNA signature generated in this study will accelerate the identification of genes involved in lung cancer brain metastasis.

Abstract

Background: To improve the prognosis of patients with lung adenocarcinoma (LUAD), revolutionary treatments for metastatic lesions are essential. Methods: To identify genes closely involved in LUAD-cell-derived metastasis, we used RNA sequencing to generate microRNA (miRNA) expression signatures of brain metastatic lesions. Once tumor-suppressive miRNAs are identified, it will be possible to explore the numerous tumor-promoting genes that are regulated by miRNAs. Results: By comparison with a previously created LUAD signature, we identified several miRNAs whose expression was significantly suppressed in brain metastases. We focused on both strands of pre-miR-195 (miR-195-5p and miR-195-3p), which were significantly downregulated in brain metastatic tissues, and confirmed by ectopic expression assays that both strands of pre-miR-195 attenuated the aggressive phenotypes (cell proliferation, migration, and invasion) of LUAD cells. These data suggest that both strands of pre-miR-195 have tumor-suppressive functions in LUAD cells. Next, we explored the target molecules that each miRNA strand regulates in LUAD cells. We identified 159 target genes regulated by miR-195-5p and miR-195-3p, of which 12 genes (ANLN, CDC6, CDCA2, CDK1, CEP55, CHEK1, CLSPN, GINS1, KIF23, MAD2L1, OIP5, and TIMELESS) affect cell cycle/cell division and the prognosis of LUAD patients. Finally, we focused on two genes, ANLN (miR-195-5p target) and MAD2L1 (miR-195-3p target), and demonstrated their oncogenic functions and the molecular pathways they regulate in LUAD cells. Conclusions: The miRNA signature derived from lung cancer brain metastasis will be a landmark in the field, and analysis of this miRNA signature will accelerate the identification of genes involved in lung cancer brain metastasis.

1. Introduction

Lung cancer is the most commonly diagnosed malignant tumor worldwide, with approximately 2.5 million new cases reported in 2022. It is also the leading cause of cancer-related deaths, accounting for 18.7% of cancer-related deaths worldwide [1]. Histologically, lung cancer is classified into non-small-cell lung cancer (NSCLC; which accounts for 85% of cases) and small-cell lung cancer (SCLC; which accounts for 15% of cases). Most NSCLC cases are adenocarcinoma (LUAD) [2]. In recent years, the prognosis of patients with advanced LUAD has improved with the advent of targeted therapies against specific genetic mutations and the introduction of immune checkpoint inhibitors [3,4]. However, despite these advances, distant metastasis remains the primary cause of lung-cancer-related deaths, with the brain being the most common site of dissemination [5,6]. Almost 10% of patients with newly diagnosed NSCLC present with metastasis, and 25–40% develop brain metastases during the disease course [6,7]. The prognosis for patients with brain metastases remains extremely poor, with the median survival ranging from only 3 to 6 months, even after treatment [8]. Therefore, our challenge is to elucidate the molecular mechanisms involved in the brain metastasis of lung cancer cells.

MicroRNAs (miRNAs) are small non-coding RNAs that act as post-transcriptional gene controllers and are essential for cell maintenance [9,10]. Numerous studies have demonstrated that aberrantly expressed miRNAs play important roles in the development, progression, metastasis, and drug resistance of cancer cells [11,12]. miRNA expression signatures provide a wealth of information about the miRNAs dysregulated in cancer cells [13]. Recent advances in RNA-sequencing technology have significantly contributed to the generation of miRNA expression signatures for various types of cancer cells [14]. Our research group has created miRNA expression signatures for various cancer tissues, including LUAD and SCLC, and identified tumor-suppressive miRNAs based on these signatures [15,16].

Our miRNA signatures demonstrated that some passenger strands of miRNAs were actually downregulated or upregulated in brain metastases.. Furthermore, we have demonstrated that some passenger strands, whose functions had previously been unclear, function as tumor-suppressive miRNAs in cancer cells and directly control multiple genes [15,16,17,18,19].

The aim of this study was to identify genes that are essential for lung cancer cells to metastasize to the brain. In this study, we generated a new miRNA expression signature from brain metastatic lesions and compared it with a signature from LUAD tissue that we had previously generated. Our signature showed that multiple passenger strands were downregulated in brain metastatic tissues. In particular, both the guide and passenger strands of the following eight pre-miRNAs were downregulated in metastatic tissues: miR-10a, miR-34b, miR-34c, miR-195, miR-199a, miR-199b, and miR-497. Exploring novel molecular pathways controlled by these miRNAs will provide new insights into the metastasis of LUAD cells. Interestingly, among these downregulated miRNAs, miR-195 and miR-497 are part of a cluster of miRNAs on human chromosome 17p13.1. The target genes of these clustered miRNAs may be involved in malignant transformation of LUAD cells.

To verify the validity of the new signature we created in this study, we focused on both strands of pre-miR-195 (miR-195-5p and miR-195-3p), which were significantly downregulated in brain metastatic tissues, and attempted to search for their target molecules. Our functional assays confirmed that both strands of pre-miR-195 acted as tumor suppressors in LUAD cells. A total of 159 genes were identified as target genes controlled by pre-miR-195 in LUAD cells. We found that 12 of these genes (ANLN, CDC6, CDCA2, CDK1, CEP55, CHEK1, CLSPN, GINS1, KIF23, MAD2L1, OIP5, and TIMELESS) are involved in cell cycle/cell division and affect the prognosis of LUAD patients. Finally, we focused on two genes, ANLN (miR-195-5p target) and MAD2L1 (miR-195-3p target), and demonstrated their oncogenic functions and the molecular pathways regulated by them in LUAD cells.

2. Materials and Methods

2.1. LUAD Patients and LUAD Cell Lines

We obtained surgical specimens from the primary tumor and brain metastatic tissues of patients with LUAD. The background and clinical characteristics of the patients are described in Table S1.

We also employed two LUAD cell lines, A549 and H1299 (American Type Culture Collection, Manassas, VA, USA). The procedures for cell maintenance have been described in our previous publications [15,18].

2.2. Construction of an miRNA Expression Signature for LUAD Based on RNA Sequencing

To assess miRNA expression, LUAD and brain metastatic tissue samples were subjected to sequencing using the Illumina NextSeq 500 system (Illumina, Inc., San Diego, CA, USA). The raw sequencing data have been deposited in the Gene Expression Omnibus (GEO; GEO accession number: GSE230229).Sequence reads matched to the human genome were classified into different types of small RNAs based on their biological features, and the corresponding read counts for each category are presented in Table S2.

2.3. Identification of Oncogenic Targets of miR-195-5p and miR-195-3p in LUAD Cells

We analyzed gene expression data from A549 cells transfected with miR-195-5p or miR-195-3p (GEO accession number: GSE281258) in combination with predictions with TargetScanHuman ver. 8.0 (https://www.targetscan.org/vert_80/; accessed on 25 September 2024) to identify oncogenic targets regulated by miR-195-5p and miR-195-3p. This expression profile represents a comparison of the gene expression between LUAD and normal lung tissues. The molecular functions of the miR-195-5p and miR-195-3p target genes were inferred using GeneCodis4 software (https://genecodis.genyo.es/; accessed on 25 September 2024) [20].

2.4. Analysis of miRNA and Gene Expression and Their Clinical Significance in Patients with LUAD Using an In Silico Database

The expression of miRNAs and genes in LUAD clinical tissues was analyzed using the following publicly available databases: The Cancer Genome Atlas (TCGA) (https://www.cancer.gov/tcga; accessed on 20 September 2024), Genomic Data Commons Data Portal (https://portal.gdc.cancer.gov/; accessed on 20 September 2024), and FIREBROWSE (http://firebrowse.org/; accessed on 20 September 2024). Data related to overall survival were retrieved from OncoLnc (http://www.oncolnc.org/; downloaded on 20 September 2024) and cBioPortal (https://www.cbioportal.org/; accessed on 20 September 2024).

2.5. Functional Assays of miRNAs and miRNA Target Genes in LUAD Cells

miRNAs and small interfering RNAs (siRNAs) were transfected into LUAD cell lines. After transfection, cell proliferation, migration, and invasion were assessed and compared with those of untreated cells. The transfection procedures for the miRNAs and siRNAs were described in our previous studies [15,16,18]. The mock group comprised cells transfected without miRNAs or siRNAs. The cell cycle and apoptotic cells were analyzed using the BD FACSCelesta^TM Flow Cytometer (BD Biosciences, Franklin Lakes, NJ, USA). Details of the cell functional assays are described in our previous papers [15,16,18]. The reagents used for these analyses are listed in Table S3.

2.6. Dual-Luciferase Reporter Assay

The binding sequences of miR-195-5p and miR-195-3p, which were cloned into the psiCHEK2 vector (C8021; Promega, Madison, WI, USA), are shown in Figures S1 and S2. The procedures for the transfection and dual-luciferase reporter assays have been described in our previous studies [15,16,18]. We performed the dual-luciferase reporter assays at 72 h after transfection using the Dual-Luciferase Reporter Assay System (catalog no. E1910, Promega). The reagents used for these analyses are listed in Table S3.

2.7. Western Blotting

Western blotting was conducted following protocols described in our previous publications [15,16,18]. The reagents and antibodies used in the present study are listed in Table S3.

2.8. Statistical Analysis

Statistical analyses were performed using R ver. 4.4.0 (R Core Team, Vienna, Austria; https://www.R-project.org/; accessed on 25 September 2024) and GraphPad Prism 8 (GraphPad Software, La Jolla, CA, USA). Differences between two groups were assessed by the Student’s t-test. Multiple-group comparisons were performed using one-way analysis of variance and Tukey’s test for post hoc analysis. Survival analysis was carried out using Kaplan–Meier survival curves and the log-rank test.

3. Results

3.1. The RNA-Sequencing-Based miRNA Expression Signature Using Clinical Specimens Obtained from LUAD Brain Metastases

In this study, we created a new miRNA expression signature using LUAD brain metastatic tissues. We explored miRNAs whose expression was downregulated in brain metastasis tissues by comparing them with the LUAD signature we had previously created [15]. A total of 48 downregulated miRNAs were identified (

Table 1. Downregulated miRNAs in brain metastasis compared with those in lung adenocarcinoma, identified by RNA sequencing.

MicroRNA	miRBase Accession No.	Guide or Passenger Strand	Chromosomal Location	Log2 FC	p-Value	FDR
hsa-miR-30d-5p	MIMAT0000245	Guide strand	8q24.22	−2.51	0.004	0.018
hsa-miR-34b-5p	MIMAT0000685	Passenger strand	11q23.1	−2.56	0.047	0.117
hsa-miR-4536-5p	MIMAT0019078	Guide strand	Xp11.21	−2.56	0.001	0.005
hsa-miR-6785-3p	MIMAT0027471	Passenger strand	17q25.1	−2.59	0.000	0.000
hsa-miR-3141	MIMAT0015010	Guide strand	5q33.2	−2.61	0.008	0.031
hsa-miR-30b-5p	MIMAT0000420	Guide strand	8q24.22	−2.63	0.003	0.016
hsa-miR-497-3p	MIMAT0004768	Passenger strand	17p13.1	−2.63	0.000	0.003
hsa-miR-4322	MIMAT0016873	Guide strand	19p13.2	−2.64	0.016	0.053
hsa-miR-195-3p	MIMAT0004615	Passenger strand	17p13.1	−2.66	0.000	0.001
hsa-miR-6743-3p	MIMAT0027388	Guide strand	11p15.5	−2.66	0.000	0.002
hsa-miR-148a-5p	MIMAT0004549	Passenger strand	7p15.2	−2.79	0.000	0.000
hsa-miR-195-5p	MIMAT0000461	Guide strand	17p13.1	−2.79	0.000	0.002
hsa-miR-548i	MIMAT0005935	Guide strand	3q21.2	−2.81	0.008	0.032
hsa-miR-6788-3p	MIMAT0027477	Guide strand	18p11.22	−2.82	0.000	0.000
hsa-miR-363-5p	MIMAT0003385	Passenger strand	Xq26.2	−2.85	0.001	0.005
hsa-miR-4536-3p	MIMAT0020959	Passenger strand	Xp11.21	−2.91	0.000	0.000
hsa-miR-497-5p	MIMAT0002820	Guide strand	17p13.1	−2.91	0.000	0.002
hsa-miR-214-5p	MIMAT0004564	Passenger strand	1q24.3	−2.92	0.000	0.001
hsa-miR-95-3p	MIMAT0000094	Guide strand	4p16.1	−2.92	0.008	0.033
hsa-miR-4662a-5p	MIMAT0019731	Guide strand	8q24.13	−2.94	0.008	0.031
hsa-miR-145-5p	MIMAT0000437	Guide strand	5q32	−2.95	0.000	0.002
hsa-miR-130a-3p	MIMAT0000425	Guide strand	11q12.1	−3.02	0.000	0.002
hsa-miR-504-5p	MIMAT0002875	Guide strand	Xq26.3	−3.02	0.012	0.042
hsa-miR-199a-3p	MIMAT0000232	Guide strand	19p13.2	−3.03	0.000	0.002
hsa-miR-34c-3p	MIMAT0004677	Passenger strand	11q23.1	−3.06	0.009	0.036
hsa-miR-6763-5p	MIMAT0027426	Guide strand	12q24.33	−3.07	0.000	0.000
hsa-miR-214-3p	MIMAT0000271	Guide strand	1q24.3	−3.14	0.000	0.002
hsa-miR-199b-3p	MIMAT0004563	Guide strand	9q34.11	−3.15	0.000	0.002
hsa-miR-218-5p	MIMAT0000275	Guide strand	4p15.31	−3.17	0.001	0.006
hsa-miR-34c-5p	MIMAT0000686	Guide strand	11q23.1	−3.22	0.042	0.107
hsa-miR-4470	MIMAT0018997	Guide strand	8q12.3	−3.30	0.000	0.001
hsa-miR-1281	MIMAT0005939	Guide strand	22q13.2	−3.32	0.000	0.000
hsa-miR-4259	MIMAT0016880	Guide strand	1q23.2	−3.33	0.000	0.000
hsa-miR-3064-3p	MIMAT0019865	Passenger strand	17q23.3	−3.41	0.000	0.000
hsa-miR-3926	MIMAT0018201	Guide strand	8p23.1	−3.44	0.000	0.000
hsa-miR-5091	MIMAT0021083	Guide strand	4p15.33	−3.46	0.000	0.001
hsa-miR-199b-5p	MIMAT0000263	Passenger strand	9q34.11	−3.56	0.001	0.007
hsa-miR-602	MIMAT0003270	Guide strand	9q34.3	−3.61	0.000	0.003
hsa-miR-548ar-3p	MIMAT0022266	Passenger strand	13q34	−3.62	0.000	0.000
hsa-miR-10a-3p	MIMAT0004555	Passenger strand	17q21.32	−3.82	0.001	0.004
hsa-miR-10a-5p	MIMAT0000253	Guide strand	17q21.32	−3.82	0.000	0.002
hsa-miR-548w	MIMAT0015060	Guide strand	16p12.1	−3.85	0.000	0.000
hsa-miR-199a-5p	MIMAT0000231	Passenger strand	19p13.2	−3.91	0.000	0.000
hsa-miR-150-5p	MIMAT0000451	Guide strand	19q13.33	−4.00	0.000	0.001
hsa-miR-548a	MIMAT0003251	Guide strand	6p22.3	−4.19	0.000	0.000
hsa-miR-34b-3p	MIMAT0004676	Guide strand	11q23.1	−4.32	0.005	0.023

), and 14 miRNAs were annotated as passenger strands based on the miRBase database (Release 22; http://www.mirbase.org/) [21]. Interestingly, both the guide and the passenger strands derived from the following pre-miRNAs were significantly downregulated: miR-10a, miR-34b, miR-34c, miR-195, miR-199a, miR-199b, and miR-497 (Table 1). Exploring the genes controlled by these miRNAs will provide novel information regarding brain metastasis of LUAD cells.

In the human genome, miR-195 and miR-497 are part of a cluster of miRNAs on human chromosome 17p13.1 (Figure 1A). Both miR-195 and miR-497 were significantly downregulated in brain metastatic tissues compared with LUAD and normal lung tissues (Figure 1B). Analysis of the functional RNA network controlled by this miRNA cluster could potentially elucidate the molecular mechanism of lung cancer cell metastasis. Here, we first focused on miR-195-5p and miR-195-3p to identify their regulated RNA networks in LUAD cells.

3.2. Expression and Clinical Significance of miR-195-5p and miR-195-3p in LUAD Clinical Specimens

We validated the expression of miR-195-5p and miR-195-3p in LUAD clinical specimens using the large clinical cohort database TCGA. The expression levels of miR-195-5p and miR-195-3p were significantly reduced in LUAD tissues compared with normal tissues (Figure 1C). Using the TCGA dataset, we analyzed whether the expression levels of these miRNAs can serve as prognostic markers for LUAD. Low expression of miR-195-3p was associated with a significantly poor prognosis compared with high expression of this miRNA (Figure 1D). A similar analysis showed no significant difference in the expression of miR-195-5p among the tissue types (Figure 1D).

3.3. Tumor-Suppressive Functions of miR-195-5p and miR-195-3p in LUAD Cells

To demonstrate the antitumor functions of both strands of pre-miR-195 in LUAD cells, we ectopically expressed these miRNAs in A549 and H1299 cell lines and examined their effects on cancer cell behavior. Ectopic expression of miR-195-5p or miR-195-3p significantly suppressed the proliferation of LUAD cells (Figure 2A). Cell cycle analysis revealed the typical G0/G1 phase arrest following transfection with either miRNA (Figure 2B). Furthermore, expression of miR-195-5p or miR-195-3p led to a marked increase in the apoptotic cell population (Figure 2C). In addition, the invasive and migratory capacities of LUAD cells were significantly inhibited upon ectopic expression of miR-195-5p or miR-195-3p (Figure 2D,E). Flow cytometry images demonstrating this apoptosis are presented in Figure S3, and representative images of the invasion and migration assays are shown in Figure S4. Based on these results, we conclude that miR-195-5p and miR-195-3p function as tumor-suppressive miRNAs in LUAD cells.

3.4. Identification of miR-195-5p- and miR-195-3p-Regulated Genes in LUAD Cells

As reported previously, we confirmed that both strands of pre-miR-195 function as antitumor miRNAs. Identifying the oncogenic targets regulated by these miRNAs in LUAD cells will aid in the search for genes involved in the malignant progression of lung cancer. Our strategy for the microRNA target search is shown in Figure 3. In this study, we performed a genome-wide gene expression analysis in A549 cells transfected with miR-195-5p or miR-195-3p. The expression data obtained have been deposited in the GEO database and are accessible (accession number GSE281258).

Using a combination of the TargetScanHuman database (release 8.0) and the gene expression profiles obtained from miRNA-transfected LUAD cells, we identified putative target genes regulated by miR-195-5p (95 genes) and miR-195-3p (63 genes). These candidate targets are listed in Tables S4 and S5. To further characterize their potential roles, we analyzed the molecular functions of all 158 genes using the GeneCodis4 database (Table S6). Notably, 27 of these genes were associated with cell cycle regulation (

Table 2. Putative target genes regulated by miR-195-5p or miR-195-3p in A549 cells.

Gene ID	Gene Symbol	Gene Name	Downregulated by miR-195-5p or miR-195-3p	miR-195-5p Transfectant Log2 Fold Change	miR-195-3p Transfectant Log2 Fold Change
402	ARL2	ADP ribosylation factor-like GTPase 2	miR-195-5p	−3.31	0.11
896	CCND3	cyclin D3	miR-195-5p	−3.00	0.74
11339	OIP5	Opa-interacting protein 5	miR-195-5p	−2.49	−0.62
54443	ANLN	anillin, actin-binding protein	miR-195-5p	−2.35	−0.36
55165	CEP55	centrosomal protein 55	miR-195-5p	−2.27	−0.67
983	CDK1	cyclin-dependent kinase 1	miR-195-5p	−1.86	−0.63
63967	CLSPN	claspin	miR-195-5p	−1.85	−0.70
993	CDC25A	cell division cycle 25A	miR-195-5p	−1.80	0.18
990	CDC6	cell division cycle 6	miR-195-5p	−1.78	−0.54
157313	CDCA2	cell division cycle- associated 2	miR-195-5p	−1.78	−0.53
55038	CDCA4	cell division cycle- associated 4	miR-195-5p	−1.75	−0.28
1111	CHEK1	checkpoint kinase 1	miR-195-5p	−1.56	−0.31
6867	TACC1	transforming acidic coiled-coil-containing protein 1	miR-195-5p	−1.52	0.27
9493	KIF23	kinesin family member 23	miR-195-5p	−1.52	−0.11
80010	RMI1	RecQ-mediated genome instability 1	miR-195-3p	−1.52	−1.07
144455	E2F7	E2F transcription factor 7	miR-195-5p	−1.46	−0.01
4085	MAD2L1	mitotic arrest deficient 2-like 1	miR-195-3p	−1.44	−2.13
8914	TIMELESS	timeless circadian clock	miR-195-5p	−1.39	−0.23
79187	FSD1	fibronectin type III and SPRY domain-containing 1	miR-195-5p	−1.26	−0.72
6197	RPS6KA3	ribosomal protein S6 kinase A3	miR-195-5p	−1.24	0.62
27183	VPS4A	vacuolar protein sorting 4 homolog A	miR-195-5p	−1.16	−0.07
9874	TLK1	tousled-like kinase 1	miR-195-5p	−1.12	0.13
9837	GINS1	GINS complex subunit 1	miR-195-3p	−1.08	−1.85
996	CDC27	cell division cycle 27	miR-195-5p	−1.06	−0.11
8243	SMC1A	structural maintenance of chromosomes 1A	miR-195-3p	−0.47	−1.32
7329	UBE2I	ubiquitin-conjugating enzyme E2 I	miR-195-3p	−0.03	−1.00
57804	POLD4	DNA polymerase delta 4	miR-195-3p	0.19	−1.35

).

3.5. Clinical Significance of Genes Regulated by miR-195-5p or miR-195-3p in LUAD

To assess the clinical relevance of the 27 genes potentially regulated by miR-195-5p or miR-195-3p, we analyzed the TCGA-LUAD dataset. Among the 27 target genes, 12 (Table 2, bold) were significantly upregulated in LUAD tissues (n = 516) compared with normal lung tissues (n = 59) (Figure 4A). Furthermore, elevated expression of these genes was significantly associated with a poor prognosis (5-year overall survival rate, p < 0.05) in LUAD patients (Figure 4B). Ectopic expression of miR-195-5p or miR-195-3p in LUAD cells significantly reduced the mRNA levels of these 12 target genes (Figure 5). Among the 12 target genes, we focused on anillin (ANLN), the top prognostic candidate regulated by miR-195-5p, and mitotic arrest deficient 2-like 1 (MAD2L1), the top prognostic candidate regulated by miR-195-3p, for further functional analyses.

3.6. Direct Regulation of ANLN by miR-195-5p and MAD2L1 by miR-195-3p in LUAD Cells

Ectopic expression of miR-195-5p and miR-195-3p in LUAD cells significantly reduced the mRNA and protein expression levels of ANLN and MAD2L1, respectively (Figure 6A,B,E,F). Full-length Western blot images are provided in Figures S5 and S6. Luciferase reporter assays further demonstrated that both miRNAs directly bind to the 3′ untranslated regions (3′-UTRs) of their respective target genes. The predicted miR-195-5p-binding site in the ANLN 3′-UTR and the miR-195-3p-binding site in the MAD2L1 3′-UTR are shown in Figure 6C and G, respectively. Co-transfection of miR-195-5p or miR-195-3p with the corresponding reporter construct led to a marked reduction in luciferase activity, whereas no such reduction was observed after transfection with constructs lacking the respective binding sites (Figure 6D,H). These findings indicate that miR-195-5p and miR-195-3p directly target ANLN and MAD2L1, respectively, and post-transcriptionally suppress their expression in LUAD cells.

3.7. Functional Significance of ANLN in LUAD Cells

To investigate the oncogenic role of ANLN in LUAD cells, we performed siRNA-mediated knockdown assays. Transfection with two independent siRNAs targeting ANLN (siANLN-1 and siANLN-2) resulted in significant reductions in both ANLN mRNA and protein levels in LUAD cells (Figures S7 and S8). Cell proliferation assays revealed that ANLN knockdown inhibited the proliferation of LUAD cells (Figure 7A). Furthermore, flow cytometric analysis showed that ANLN knockdown in LUAD cells induced G0/G1 cell cycle arrest and increased the proportion of apoptotic cells (Figure 7B,C). Notably, ANLN knockdown markedly suppressed both the invasive and migratory capacities of LUAD cells (Figure 7D,E). Flow cytometry images showing the proportion of cells in apoptosis are presented in Figure S9, and representative images of the invasion and migration assays are shown in Figure S10.

3.8. Functional Significance of MAD2L1 in LUAD Cells

Similarly, to investigate the oncogenic role of MAD2L1 in LUAD cells, we performed siRNA-mediated knockdown assays. Transfection with two independent siRNAs targeting MAD2L1 (siMAD2L1-1 and siMAD2L1-2) resulted in significant reductions in both MAD2L1 mRNA and protein expression levels in LUAD cells (Figures S11 and S12). Cell proliferation assays showed that MAD2L1 knockdown slightly inhibited the proliferation of LUAD cells (Figure 8A). Additionally, flow cytometric analysis demonstrated that MAD2L1 knockdown induced G0/G1 cell cycle arrest and increased the proportion of apoptotic cells in LUAD cells (Figure 8B,C). However, in H1299 cells, no increase in the proportion of G0/G1 phase cells was observed; instead, there was a notable increase in the proportion of subG1 cells, indicating enhanced apoptotic activity. Furthermore, MAD2L1 knockdown markedly suppressed the invasive and migratory abilities of LUAD cells (Figure 8D,E). Flow cytometry images showing the proportion of cells in apoptosis are presented in Figure S13, and representative images of the invasion and migration assays are shown in Figure S14.

3.9. ANLN- and MAD2L1-Mediated Cancer Pathways in LUAD Cells

To investigate the signaling pathways influenced by ANLN and MAD2L1 in cancer, we generated genome-wide gene expression profiles using A549 cells transfected with siANLN or siMAD2L1. The resultant expression data have been deposited in the GEO database under accession numbers GSE281585 and GSE281905.

Interestingly, we also observed that siANLN transfection suppressed MAD2L1 expression, while siMAD2L1 transfection suppressed ANLN expression, suggesting the existence of reciprocal regulatory interactions between these two genes (Figure 9). A total of 39 genes were commonly downregulated in both siANLN- and siMAD2L1-transfected cells (

Table 3. Downregulated genes after siANLN-1 or siMAD2L1-1 transfection in A549 cells.

Entrez Gene ID	Gene Symbol	Gene Name	siANLN Transfectant Log2 Fold Change	siMAD2L1 Transfectant Log2 Fold Change
54443	ANLN	anillin actin-binding protein	−6.14	−2.32
84904	ARHGEF39	Rho guanine nucleotide exchange factor 39	−3.61	−2.28
55723	ASF1B	anti-silencing function 1B histone chaperone	−4.52	−2.71
890	CCNA2	cyclin A2	−3.64	−2.09
991	CDC20	cell division cycle 20	−4.26	−2.67
157313	CDCA2	cell division cycle-associated 2	−3.01	−2.03
55536	CDCA7L	cell division cycle-associated 7-like	−1.55	−2.61
983	CDK1	cyclin-dependent kinase 1	−4.60	−1.96
1033	CDKN3	cyclin-dependent kinase inhibitor 3	−2.92	−2.43
2491	CENPI	centromere protein I	−3.73	−2.60
55165	CEP55	centrosomal protein 55	−3.59	−2.03
1164	CKS2	CDC28 protein kinase regulatory subunit 2	−2.68	−1.55
1719	DHFR	dihydrofolate reductase	−4.01	−2.18
79075	DSCC1	DNA replication and sister chromatid cohesion 1	−2.76	−2.39
374393	FAM111B	family with sequence similarity 111 member B	−4.26	−3.13
54478	FAM64A	family with sequence similarity 64 member A	−3.89	−1.92
9837	GINS1	GINS complex subunit 1	−2.36	−3.05
51512	GTSE1	G2 and S-phase-expressed 1	−4.21	−2.36
283120	H19	H19, imprinted maternally expressed transcript	−2.58	−2.98
8479	HIRIP3	HIRA-interacting protein 3	−2.21	−1.93
3161	HMMR	hyaluronan-mediated motility receptor	−3.78	−2.60
55806	HR	hair growth-associated	−1.94	−1.91
56992	KIF15	kinesin family member 15	−4.13	−2.51
10112	KIF20A	kinesin family member 20A	−3.86	−2.05
90417	KNSTRN	kinetochore localized astrin/SPAG5-binding protein	−1.65	−1.25
4085	MAD2L1	MAD2 mitotic arrest deficient-like 1	−2.91	−5.58
4288	MKI67	marker of proliferation Ki-67	−4.47	−3.08
727897	MUC5B	mucin 5B, oligomeric mucus/gel-forming	−1.82	−1.00
23397	NCAPH	non-SMC condensin I complex subunit H	−3.76	−2.26
9768	PCLAF	PCNA clamp-associated factor	−4.43	−2.75
5933	RBL1	RB transcriptional corepressor like 1	−1.87	−2.54
10535	RNASEH2A	ribonuclease H2 subunit A	−2.93	−2.08
200916	RPL22L1	ribosomal protein L22 like 1	−1.07	−2.64
9123	SLC16A3	solute carrier family 16 member 3	−1.13	−1.57
147841	SPC24	SPC24, NDC80 kinetochore complex component	−3.75	−2.40
57405	SPC25	SPC25, NDC80 kinetochore complex component	−4.89	−2.89
7083	TK1	thymidine kinase 1	−4.42	−2.45
8458	TTF2	transcription termination factor 2	−1.57	−1.46
89891	WDR34	WD repeat domain 34	−1.21	−1.80

). These genes were analyzed using GeneCodis4, which revealed that a substantial proportion of the genes are involved in cell-cycle-related pathways (

Table 4. Significantly enriched annotations of downregulated genes after siANLN-1 or siMAD2L1-1 transfection.

Description	p-Value	Genes
cell division	<0.001	MAD2L1, CDC20, NCAPH, CCNA2, ANLN, SPC25, FAM64A, KNSTRN, CDCA2, CEP55, CDK1, SPC24, CDCA7L, CKS2
DNA replication	<0.001	FAM111B, RNASEH2A, KIAA0101, GINS1, DSCC1, CDK1
mitotic spindle assembly checkpoint signaling	<0.001	MAD2L1, CDC20, SPC25, SPC24
chromosome segregation	<0.001	SPC25, KNSTRN, CDCA2, SPC24, CENPI
mitotic cell cycle phase transition	<0.001	CCNA2, CDK1, CKS2
mitotic sister chromatid segregation	<0.001	MAD2L1, KNSTRN, CENPI
regulation of chromosome segregation	<0.001	MKI67, CDCA2
mitotic cytokinesis	<0.001	KIF20A, ANLN, CEP55
G1/S transition in the mitotic cell cycle	<0.001	CCNA2, CDK1, CDKN3
mitotic sister chromatid cohesion	<0.001	CDC20, DSCC1

). Among these genes, the expression of 26 genes had a negative effect on the prognosis of LUAD patients (Figure 10).

4. Discussion

In this study, we focused on both strands of pre-miR-195 (miR-195-5p and miR-195-3p), which were significantly downregulated in brain metastatic tissues. Furthermore, we explored the target molecules that each miRNA controls in LUAD cells. A total of 159 genes were identified as target genes controlled by pre-miR-195. We found that 12 of these genes (ANLN, CDC6, CDCA2, CDK1, CEP55, CHEK1, CLSPN, GINS1, KIF23, MAD2L1, OIP5, and TIMELESS) are involved in cell cycle/cell division and affect the prognosis of LUAD patients. Using in vitro functional assays, we revealed that ANLN and MAD2L1 actually play an oncogenic role.

Our group has previously identified tumor-suppressive miRNAs and their novel regulatory RNA networks in lung cancer cells [15,18,19]. RNA-sequencing-based miRNA signatures provide important information that suggests which miRNAs should be analyzed. Our recent studies revealed that passenger strands derived from pre-miRNAs function as oncogenes or tumor suppressors and actually regulate intracellular target molecules in LUAD cells [18,19]. Similarly to previous results, this study also showed that both the passenger strand and guide strand of miR-195 equally and essentially affect the function of cancer cells (Figure 2). Comprehensive analysis of the passenger strands as well as the guide strands of miRNAs will bring new information to cancer research.

An interesting feature of this signature was that it contained several miRNAs (e.g., miR-10a, miR-34b, miR-34c, miR-195, miR-199a, miR-199b, and miR-497) in which both strands (guide and passenger strands derived from the pre-miRNAs) were significantly downregulated in brain metastatic tissues. Analysis of these miRNAs, including their passenger strands, will shed light on the RNA networks associated with lung cancer metastasis. Among these miRNAs, we focused on the miR-195/miR-497 miRNA cluster on human chromosome 17p13.1. We first investigated the tumor-suppressive functions of miR-195 and its regulated RNA networks in LUAD cells.

Previous reports have shown that the expression of miR-195-5p (the guide strand) is suppressed in various types of cancers. Furthermore, ectopic expression assays demonstrated that miR-195-5p acts as a tumor-suppressive miRNA across cancer types by targeting several oncogenes [22,23,24,25]. Expression of miR-195-5p enhanced the cisplatin or gemcitabine sensitivity of LUAD cells by targeting E2F transcription factor 7 [26]. Recent studies have demonstrated that aberrantly expressed long non-coding RNAs or circulating RNAs may be involved in promoting cancer cell malignant transformation by adsorbing tumor-suppressive miR-195-5p in LUAD cells [27]. Compared with the numerous studies on miR-195-5p, functional assays of miR-195-3p (the passenger strand) are scarce. Our present analysis demonstrates that miR-195-3p is a tumor-suppressive miRNA that regulates many cancer genes in LUAD cells. Combining our data with previous reports, it is clear that both strands of pre-miR-195 (miR-195-5p and miR-195-3p) behave as tumor suppressors in LUAD cells.

Next, we identified cancer-promoting genes among those regulated by pre-miR-195 (miR-195-5p and miR-195-3p) in LUAD cells. Our strategy successfully identified 159 genes as candidate pre-miR-195 targets in LUAD cells. Of these, 12 genes (ANLN, CDC6, CDCA2, CDK1, CEP55, CHEK1, CLSPN, GINS1, KIF23, MAD2L1, OIP5, and TIMELESS) are involved in the cell cycle, cell division, and DNA replication checkpoint-related pathways, and their overexpression has a significant impact on the prognosis of patients with LUAD; therefore, these genes are potential therapeutic targets in LUAD. Among these therapeutic targets, we focused on ANLN (miR-195-5p target) and MAD2L1 (miR-195-3p target) and revealed their cancer-promoting functions in LUAD cells.

ANLN is an actin-binding protein that interacts with actin and cytoskeletal filaments to control the cell cycle [28]. Recent studies show that aberrant expression of ANLN is observed in multiple types of cancer, and that overexpression of ANLN is a useful biomarker predicting cancer cell metastasis and patient prognosis [29]. Knockdown assays targeting ANLN expression revealed dramatically reduced aggressive phenotypes (cell growth, migration, and invasion) of cancer cells, including pancreatic, breast, and lung cancers [30,31,32]. Aberrant expression of ANLN is intricately involved in the molecular pathogenesis of various human cancers and is a potential therapeutic target.

MAD2L1 is an essential component of the spindle assembly checkpoint [33]. In normal cells, MAD2 accumulates at kinetochores before cell division to ensure that spindle microtubules are properly aligned with the kinetochores of each chromosome [33]. Thus, MAD2 is a gatekeeper of cell cycle checkpoints and is essential for maintaining faithful replicative cell division and preventing cancer development [34]. In multiple cancers, overexpression of MAD2L1 is involved in the malignant transformation of cancer cells, and patients overexpressing MAD2L1 have a poor prognosis [35,36,37,38]. More recently, it has been reported that MAD2L1 expression is involved in carboplatin resistance in lung cancer cells [39]. MAD2L1 serves as a therapeutic target molecule in various cancers, including LUAD.

Examination of the interaction between ANLN and MAD2L1 showed that these two genes affect each other. In addition, the existence of genes whose expression is commonly regulated downstream of ANLN and MAD2L1 was revealed. The expression of 26 of these genes had a negative effect on the prognosis of LUAD patients. Our study has shown that suppression of miR-195-5p/-3p expression and overexpression of the ANLN/MAD2L1 genes are among the factors that accelerate the progression of lung cancer cells. Future studies involving in vivo validation or analyses using larger clinical datasets, including independent patient cohorts, will be essential to confirm the biological and therapeutic significance of the identified targets.

Analyses based on the miRNA signature created in this study will facilitate the identification of genes involved in the malignant progression and brain metastasis of LUAD cells. This study is exploratory and based on a limited number of LUAD brain metastasis specimens, which are rare and difficult to obtain. While the findings offer important insights, they should be interpreted with caution and require further validation in larger patient cohorts to confirm their broader applicability.

5. Conclusions

We generated a novel RNA-sequencing-based miRNA signature using clinical specimens obtained from lung cancer brain metastases. Both strands of pre-miR-195 (miR-195-5p and miR-195-3p) were significantly downregulated in brain tissues. Ectopic expression assays demonstrated that the two miRNAs derived from pre-miR-195 attenuated LUAD cell aggressiveness by targeting several oncogenes, particularly ANLN (miR-195-5p target) and MAD2L1 (miR-195-3p target). ANLN and MAD2L1 may be novel therapeutic targets in LUAD cells.