We employed the TIMER (https://cistrome.shinyapps.io/timer/) (Li et al., 2017), Oncomine (https://www.oncomine.org), and GEPIA databases (http://gepia.cancer-pku.cn/) (Tang et al., 2017) to analyze the expression of IQGAP3 in pan-cancer, and the CCLE tools (https://portals.broadinstitute.org/ccle/) (Ghandi et al., 2019) were employed to examine the expression of IQGAP3 in diverse cancer cells lines. UALCAN tools (http://ualcan.path.uab.edu/) (Chandrashekar et al., 2017) were used to analyze the protein of IQGAP3 in different cancers. The expression of let-7c-5p was analyzed by using starBase (Li et al., 2014). The Kaplan–Meier plotter (http://kmplot.com/analysis/) (Hou et al., 2017) was employed to examine the prognosis of let-7c-5p in pan-cancer.

We employed the GEPIA (http://gepia.cancer-pku.cn/) and prognostic databases (http://dna00.bio.kyutech.ac.jp/PrognoScan/index.html) (Tang et al., 2017; Mizuno et al., 2009) to analyze the OS and RFS of IQGAP3 in pan-cancer; additionally, the correlation between the tumor stage and IQGAP3 expression was analyzed by using GEPIA. Tumor stage, lymph node metastasis, and expression of let-7c-5p were analyzed by using UALCAN tools (http://ualcan.path.uab.edu/) (Chandrashekar et al., 2017). We also employed the prognosis tools to verify the prognosis of IQGAP3 in pan-cancer.

The DNA methylation of IQGAP3 was analyzed by Ualcan tools (http://ualcan.path.uab.edu/) (Chandrashekar et al., 2017), the correlation between the OS and DNA methylation level was analyzed by the Methsurv database (https://biit.cs.ut.ee/methsurv/) (Modhukur et al., 2018). The mutation information of IQGAP3 in pan-cancer was analyzed by the cbioportal database (https://www.cbioportal.org/) (Cerami et al., 2012).

We employed the starBase database (http://starbase.sysu.edu.cn/) to forecast the potential miRNAs of IQGAP3 (Li et al., 2014), and examine the expression, prognosis, and correlation between let-7c-5p and IQGAP3, we also used the starbase to predict the binding with between the miRNA, mRNA, and lncRNA.

We employed the CancerSEA database (http://biocc.hrbmu.edu.cn/CancerSEA/) analysis the function of IQGAP3 in pan-cancer (Yuan et al., 2019), the LinkedOmics (http://www.linkedomics.org/admin.php) was employed to analyze the KEGG pathway of IQGAP3 in LUAD (Vasaikar et al., 2018).

We employed the STRING database (https://string-db.org/cgi) to construct the protein interaction networks of IQGAP3 in cancer (Szklarczyk et al., 2017), and the GeneMANIA (http://genemania.org/) was employed to analyze the interaction gene with the IQGAP3 (Franz et al., 2018).

We employed the TIMER (https://cistrome.shinyapps.io/timer/) and XCELL tools (https://xcell.ucsf.edu/) to analyze the immunological roles of IQGAP3 (Li et al., 2017; Aran et al., 2017), including the correlation between diverse immune cells and the immune regulator. The TISIDB (http://cis.hku.hk/TISIDB/) was adopted to analyze the relationship between IQGAP3 expression and 28 tumor-infiltrating lymphocytes, 45 immune stimulators, 24 immune inhibitors, 41 chemokines, 18 receptors, and 21 MHC molecules in pan-cancer (Ru et al., 2019). The TMB and MSI scores were obtained from TCGA. Correlation analysis between the IQGAP3 expression and TMB or MSI was performed using Spearman’s method.

We employed the Genomics of Drug Sensitivity in Cancer (GDSC) (www.cancerRxgene.org) and CTRP databases to analyze the correlation between IQGAP3 expression and drug sensitivity (Basu et al., 2013; Yang et al., 2013).

We employed the lncLocator (www.csbio.sjtu.edu.cn/bioinf/lncLocator) and CPC2 (http://cpc2.cbi.pku.edu.cn) to examine the subcellular localization and the protein-coding ability of IQGAP3 (Kang et al., 2017; Cao et al., 2018).

The BEAS-2B cell line was purchased from the cell bank of Kunming Institute of Zoology and cultured in BEGM media (Lonza, CC-3170). HEK-293T was obtained from ATCC. Lung cancer cell lines, including A549, H1299, and H1975 were purchased from Cobioer, China, with the STR document, and the A549, H1299, and H1975 cells were all cultured in an RPMI1640 medium (Corning) supplemented with 10% fetal bovine serum (Cat# 10099141C, Gibco, United States) and 1% penicillin/streptomycin.

The qRT-PCR assay was performed as documented (Yang et al., 2013). The primer sequences are as follows: IQGAP3-F: GCA​GCC​TAT​GAA​CGC​CTC​A, IQGAP3-R: GGA​GGG​TGC​AAA​ACA​GTG​G, β-actin-F: CTTCGCGGGCGACGAT, and β-actin-R: CCA​TAG​GAA​TCC​TTC​TGA​CC. The expression quantification was obtained with the 2−ΔΔCt method.

Immunohistochemistry staining assay was performed as documented (Basu et al., 2013). Briefly, lung cancer tissue and normal lung tissues were obtained from advanced-stage lung cancer patients from The Third Affiliated Hospital of Kunming Medical University (Yunnan Tumor Hospital), Kunming, China. These tissues were used to perform immunohistochemistry (IHC), primary antibody overnight incubation, and second antibody incubation, and finally, developed using the instrument. The detailed information of antibodies employed in our study is as follows: IQGAP3 antibody (Rabbit polyclonal to IQGAP3, ab219354, 1:500).

Analysis of the IQGAP3 expression pan-cancer was estimated using t-tests. For survival analysis, the HR and p-value were calculated employing univariate Cox regression analysis. Kaplan–Meier analysis was employed to examine the survival time of patients stratified according to high or low levels of the IQGAP3 expression. p-values less than 0.05 were considered statistically significant. For all figures, ∗, ∗∗, and ∗∗∗ indicate p < 0.05, p < 0.01, and p < 0.001, respectively.

To examine IQGAP3 expression in different types of human cancer, we first employed the TIMER database (http://timer.comp-genomics.org/). We discovered that IQGAP3 had high expression in BLCA, BRCA, CESC, CHOL, COAD, ESCA, GBM, HNSC, KIRC, KIRP, LUAD, LUSC, PAAD, PCPG, PRAD, READ, STAD, THCA, and UCEC (Figure 1A). We also analyzed IQGAP3 expression in human cancer based on Gene Expression Omnibus (GEO) datasets by employing the Oncomine database (www.oncomine.com/). IQGAP3 expression was increased in the cancer of the bladder, breast, colorectum, stomach, kidney, liver, and lung (Figure 1B). Owing to a lack of data for normal tissue in The Cancer Genome Atlas (TCGA) database (www.cancer.gov/about-nci/organization/ccg/research/structural-genomics/tcga/), we used the GEPIA database (http://gepia.cancer-pku.cn/) to explore IQGAP3 expression in human cancer types. A high expression of IQGAP3 was observed in BLCA, BRCA, CESC, COAD, ESCA, GBM, HNSC, LIHC, LUAD, LUSC, OV, PAAD, READ, SKCM, STAD, Thymoma (THYM), UCEC, and uterine carcinosarcoma (UCS) (Figure 1C). Finally, to determine IQGAP3 expression in different cancer cells lines, we used the Cancer Cell Line Encyclopedia (CCLE) database (https://sites.broadinstitute.org/ccle/). We found that IQGAP3 expression was upregulated in different cancer cells lines (Figure 1D). Overall, these results showed that IQGAP3 expression was upregulated in different types of human cancer.

IQGAP3 expression was associated significantly with the pathological stage of ACC, BRCA, KICH, KIRC, KIRP, LIHC, LUAD, LUSC, OV, PAAD, and THCA (Figure 2 and Supplementary Figure S1). These findings demonstrated that IQGAP3 expression was correlated significantly with the pathological stage of different types of cancer.

Lymph node metastasis plays a crucial part in cancer progression. We examined IQGAP3 expression in lymph node metastasis of different types of cancer. We discovered that IQGAP3 expression was positively correlated with lymph node metastasis of BLCA, BRCA, CESC, CHOL, COAD, ESCA, HNSC, KICH, KIRC, KIRP, LIHC, LUAD, LUSC, MESC, PAAD, PRAD, READ, STAD, and THCA (Supplementary Figures S2A–C). These findings showed that IQGAP3 expression was significantly related to lymph node metastasis in different types of cancer.

To ascertain the prognostic role of IQGAP3 in different types of cancer, we ascertained the OS, DFS, progression-free survival (PFS), disease-specific survival (DSS), and relapse-free survival (RFS) in human cancer types. A high expression of IQGAP3 was not only related to poor OS in ACC, KIRC, KIRP, LAML, LGG, LIHC, LUAD, MESO, and UVM (Figure 3 and Supplementary Figure S3) but also associated with poor DFS in ACC, KICH, KIRP, LGG, LIHC, MESO, PRAD, SKCM, and UVM (Figure 4). Cox regression analysis showed that a high expression of IQGAP3 was related to poor PFS in ACC, KICH, KIRC, KIRP, LGG, LIHC, MESO, PAAD, PCPG, PRAD, THCA, UCEC, and UVM (Supplementary Figure S4). A high expression of IQGAP3 was associated with poor DSS in KIRP, LIHC, PAAD, PRAD, THCA, and UCEC (Supplementary Figure S4). To verify the results shown above, we employed a prognostic database (http://dna00.bio.kyutech.ac.jp/PrognoScan/index.html) to analyze the OS, RFS, and DFS in diverse GEO cohorts: identical results were obtained (Table 1). Collectively, these data indicated that IQGAP3 expression was closely related to the prognosis of patients with different cancer types.

We showed above that a high expression of IQGAP3 was correlated with the prognosis of patients with different cancer types. Next, we investigated if IQGAP3 could act as a biomarker for different cancer types. We undertook an analysis of receiver operating characteristic (ROC) curves of IQGAP3 expression to obtain the area under the ROC curve (AUC) values. The AUC values for BRCA, PRAD, LUSC, LUAD, KIRP, KIRC, COAD, READ, THCA, GBM, LGG, PAAD, SKCM, LIHC, STAD, and ESCA are given in Supplementary Figure S5. The latter showed that IQGAP3 could be used as a biomarker to diagnose different types of cancer with high sensitivity and specificity.

According to different features, human cancers can be divided into immune subtypes and molecular subtypes. Immune subtypes can be classified further into six types: C1 (wound healing), C2 (interferon-gamma dominant), C3 (inflammatory), C4 (lymphocyte depleted), C5 (immunologically quiet), and C6 (transforming growth factor-β dominant) (Hu et al., 2021). IQGAP3 showed high expression in C1 and C2 and low expression in C3 in LUAD (Supplementary Figure S6). For molecular subtypes, IQGAP3 displayed different expressions in different cancer types (Supplementary Figure S7). These results suggested that IQGAP3 had different expression patterns in human cancer types.

TMB is the number of noninherited mutations per million bases of an investigated genomic sequence. TMB has emerged as a specific and sensitive biomarker of the response to immune checkpoint inhibitors (Addeo et al., 2021). We examined the correlation between IQGAP3 expression and TMB of human cancers. IQGAP3 expression was markedly positively correlated with TMB of ACC, PAAD, STAD, KICH, LUAD, CHOL, and PRAD and negatively correlated with TMB of DLBC and THYM (Figure 5A).

MSI represents a hyper-mutable state of DNA sequences caused by a lack of activity of DNA repair (Boland and Goel, 2010). We explored the correlation between IQGAP3 expression and MSI in human cancers. IQGAP3 expression was markedly positively correlated with MSI in LUSC, SRAC, and UCEC and negatively correlated with MSI in DLBC (Figure 5B). Collectively, these data implied that IQGAP3 may influence antitumor immunity by regulating the composition and immune mechanism in the tumor microenvironment.

We wished to explore the gene-mutation information of IQGAP3 in human cancer types. We used the cBioPortal (www.cbioportal.org/) database, and the main information is shown in Figure 6A. The frequency of IQGAP3 alterations (>12%) was the highest in hepatocellular carcinoma, with amplification being the main type of alteration (Figure 6B). Amplification was the main reason why the mRNA of IQGAP3 was upregulated in different cancer types (Figure 6C). IQGAP3 mutation was significantly correlated with genes such as FLG, SPTA1, NES, GON4L, INSPP, ASH1L, BCAN, ARHGEFLL, MEF2D, and TTC24 (Figure 6D).

We also examined the mutation type and base mutation in cancer types. A missense substitution and base mutation C > T were the most common in different cancer types (Figure 6E). The mutation of IQGAP3 affected the prognosis of patients with hepatocellular carcinoma, with a poor OS and DFS being noted (Figures 6F, G). Finally, we showed that the CNV of IQGAP3 in different cancer types was positively correlated with its expression (Figures 6H, I and Supplementary Table S1). Overall, these results suggested that the CNV of IQGAP3 was positively correlated with IQGAP3 expression in different types of human cancer.

DNA methylation has a significant role in the regulation of gene expression (Klutstein et al., 2016). Next, we explored if a high expression of IQGAP3 was attributed to low DNA methylation in the promoter region of IQGAP3. We discovered that DNA hypo-methylation in IQGAP3 was negatively correlated with IQGAP3 expression in ACC, BLCA, BRCA, CHOL, COAD, KIRC, LGG, LIHC, LUAD, LUSC, PAAD, READ, SARC, SKCM, STAD, testicular germ cell tumors (TGCT), THCA, UCEC, UCS, and UVM (Supplementary Table S2). We also examined if the methylation level of IQGAP3 affected the prognosis of patients with different types of cancer. A low DNA-methylation level in IQGAP3 was correlated with a better poor prognosis in different cancer types (Table 2). These findings demonstrated that IQGAP3 expression was significantly correlated with DNA methylation in different types of human cancer.

After discovering that IQGAP3 was correlated markedly to the prognosis, tumor stage, and lymph node metastasis, we explored IQGAP3 functions in human cancer types using CancerSEA (http://biocc.hrbmu.edu.cn/). IQGAP3 was involved mainly in angiogenesis, apoptosis, cell cycle, cell differentiation, DNA damage, epithelial–mesenchymal transition (EMT), hypoxia, inflammation, invasion, metastasis, and proliferation in human cancers (Figure 7A). A high expression of IQGAP3 was positively correlated with the cell cycle (r = 0.74), cell proliferation (r = 0.62), DNA damage (r = 0.53), DNA repair (r = 0.39), EMT (r = 0.30) and inflammation (r = –0.30) in different cancer types (Figure 7B). These findings suggested that IQGAP3 has a pivotal role in the initiation and prognosis of diverse types of human cancer.

We employed the GeneMANIA (https://genemania.org/) and Search Tool for the Retrieval of Interacting Genes/Proteins (STRING; https://string-db.org/) databases to construct gene-interaction and protein-interaction networks with IQGAP3. The genes most closely associated with IQGAP3 were those of CDC42, MYL6B, MYH2, RAC1, PPP1R16A, MEF2A, IST1, NDC80, HIF1A, ITPRIPL2, CALM1, RAC3, RAC2, SRGAP3, KIFAP3, COG2, NPHP4, IQGAP1, IQGAP2, and CLUAP1 (Figure 7C). The proteins most closely associated with IQGAP3 were those for CDC42, IQGAP2, KIF20, CDH1, MEN1, CTNNB1, CTNNA1, IQGAP1, RAC1, and CLIP1 (Figure 7D). These proteins have been reported (Modhukur et al., 2018) to have crucial roles in the cell proliferation and cell cycles of different cancer types.

We wished to examine the expression pattern of genes that interact with IQGAP3. We employed the Gene Set Cancer Analysis (GSCA) (http://bioinfo.life.hust.edu.cn/GSCA/#/) database to analyze the CNV and DNA methylation of genes that interact with IQGAP3 in different cancer types.

CTNNB1 had the highest mutation rate (29%) (Figure 8A), and the CNV was markedly positively correlated with CTNNB1 expression in human cancers (Figure 8B). The gene CNV of IQGAP3 interaction genes significantly affected the prognosis of diverse patients (Figure 8C). Next, we examined the DNA-methylation level of genes in different cancer types. DNA methylation of these genes was significantly negatively correlated with mRNA expression in different cancer types (Figure 8D).

We wished to explore if an “IQGAP3 axis” plays an important part in cancer progression. We used the LinkedOmics (www.linkedomics.org/) database and Kyoto Encyclopedia of Genes and Genomes (www.genome.jp/kegg/) database to analyze which signaling pathways were enriched in different cancer types.

High expression of IQGAP3 was mainly involved in “cell cycle,” “ECM-receptor interaction,” and “miRNA in cancer” in BRCA (Figure 9A); “spliceosome” and “human T cell virus” in COAD (Figure 9B); “cell cycle” and “lysosomes” in KIRP (Figure 9C); “PI3K-AKT signaling pathway” and “rap1 signaling pathway” in LGG (Figure 9D); “cell cycle” and “RNA transport” in LIHC (Figure 9E) “cell cycle,” “spliceosome,” and “RNA transport” in LUAD (Figure 9F); “cell adhesion” and “cell cycle” in LUSC (Figure 9G); “Nod-like receptor signaling pathway” and “tight junctions” in PAAD (Figure 9H). These results demonstrated that IQGAP3 has a crucial role in the development of different cancer types.

Transcription factors have indispensable roles in controlling gene expression (Bradner et al., 2017). We explored the transcription factors that could regulate IQGAP3 transcription. We employed the JASPAR (https://jaspar.genereg.net/), PROMO (http://alggen.lsi.upc.es/cgi-bin/promo_v3/), ConTrav3 (http://bioit2.irc.ugent.be/contra/v3/#/step/1/), and UCSC (https://genome.ucsc.edu/) databases. PTBP1 was positively correlated with IQGAP3 expression in diverse cancer types (Supplementary Figures S8A–8D and Table 3). We used the KNOCK-TF (www.licpathway.net/KnockTF/) database to verify the above result. The depletion of PTBP1 reduced IQGAP3 expression in LIHC (Supplementary Figure S8E). A high expression of PTBP1 was correlated with a poor prognosis in patients with ESCA, KIRP, LIHC, LUAD, PAAD, or SARC, and related to better prognosis in patients with BLCA, ESCC, OV, READ, STAD, or THYM (Supplementary Figures S9A–9D). We examined GEO datasets and obtained identical results (Table 4). Overall, these results demonstrated that PTBP1 may be a transcription factor for IQGAP3 in different cancer types.

miRs have crucial roles in regulating the expression of messenger (m)RNA. According to the competing endogenous RNA (ceRNA) hypothesis, there is a negative correlation between miR expression and mRNA expression. If IQGAP3 shows high expression in cancer, the miR expression in cancer should be low. We used public databases to analyze the upstream miRs of IQGAP3. Using different datasets to obtain intersections, we identified four miRs (miR-196b-5p, miR-422a, miR-18b-5p, let-7c-5p). Among these miRs, only the expression of let-7c-5p was decreased significantly in human cancers.

Next, we analyzed the expression of let-7c-5p in human cancers and investigated its prognostic value. Let-7c-5p had low expression in CHOL, BRCA, BLCA, UCEC, THCA, SATD, LUSC, LUAD, LIHC, KICH, HNSC, and COAD (Supplementary Figure S10 and Supplementary Table S3). A high expression of let-7c-5p was correlated with a good prognosis in patients with BRCA, CECS, ESCA, HNSC, KIRP, LIHC, LUAD, or LUSC, but correlated with a poor prognosis in patients with BLCA, PAAD, or STAD (Supplementary Figure S11). A low expression of let-7c-5p was correlated with the tumor stage in diverse cancer types (Supplementary Figure S12). These results showed that let-7c-5p expression was decreased in most cancer types and that increased expression of let-7c-5p was correlated with a better prognosis in many cancer types. We also analyzed the correlation between let-7c-5p and IQGAP3 in human cancers. Let-7c-5p expression was negatively correlated with IQGAP3 expression in human cancers (Supplementary Table S4). These results suggested that let-7c-5p was the most potential binding IQGAP3 in cancer patients.

We wished to identify the upstream lncRNAs of let-7c-5p in human cancers. We employed the starBase (https://starbase.sysu.edu.cn/) and LncRNAbase databases (http://carolina.imis.athena-innovation.gr/diana_tools/web/index.php?r=lncbasev2%2Findex-predicted) to predict the lncRNAs that may bind with let-7c-5p. We discovered five possible lncRNAs: AC234582.1, AL590666.2, AL590666.2, MIR29B2CHG, and IQGAP3AR. Among these lncRNAs, only the expression of IQGAP3AR was upregulated significantly in BRCA, BLCA, UCEC, STAD, PRAD, LUSC, LUAD, LIHC, KIRP, CHOL, KICH, HNSC, ESCA, and COAD (Supplementary Figures S13A–13D and Supplementary Table S5). We also identified a drug that was negatively correlated with IQGAP3AR expression (Supplementary Figure S13E). Further study revealed that the high expression of IQGAP3AR was not only associated with a poor prognosis in patients with ACC, BRCA, CESC, COAD, KIRC, or LIHC but also correlated with the tumor stage in COAD, LIHC, and OV. The expression of the other lncRNAs was not significantly different in different cancer types (Figures 10A–C).

According to the ceRNA hypothesis, lncRNA expression should have a negative correlation with let-7c-5p expression and positive correlation with IQGAP3 expression. The Spearman correlation analysis revealed IQGAP3AR expression to be significantly negatively correlated with let-7c-5p expression (Supplementary Figures S14A–14C and Supplementary Table S6) and positively correlated with the expression of IQGAP3 in pan-cancer (Supplementary Figures S15A–15C and Supplementary Table S7). The target sites between IQGAP3AR, let-7c-5p, and IQGAP3 were predicted by using TarBase (https://carolina.imis.athena-innovation.gr/) (Figure 10D).

The subcellular localization of IQGAP3AR was identified using the lncLocator (www.csbio.sjtu.edu.cn/bioinf/lncLocator/) database. IQGAP3AR was located mainly in the cytoplasm (Figure 10E). We also analyzed the protein-coding potential of IQGAP3AR by performing the coding potential calculator: IQGAP3AR did not possess the protein-coding ability (Figure 10F). Our results suggested that IQGAP3AR may be upstream of the lncRNA let-7c-5p, which regulates IQGAP3 expression in different cancer types.

The infiltration of immune cells has an indispensable role in cancer progression. Next, we explored the relationship between IQGAP3 expression and immune cell infiltration in different cancer types. The TIMER database showed that IQGAP3 expression was correlated significantly with the abundance of cluster of differentiation (CD)8+ T cells in 25 types of cancer, CD4⁺ T cells in 27 cancer types, neutrophils in 29 cancer types, dendritic cells in 30 cancer types, macrophages in 29 cancer types, and B cells in 29 cancer types (Supplementary Figure S16A).

To verify these results, we employed the xCell (https://xcell.ucsf.edu/) database to assess the correlation between IQGAP3 expression and immune cell infiltration in diverse cancer types. IQGAP3 expression was positively correlated with 38 types of immune cells in 27 cancer types and negatively correlated with 38 types of immune cells in one cancer type (Supplementary Figure S16B). These findings indicated that IQGAP3 expression was significantly correlated with the infiltration of immune cells in human cancer.

We wished to further understand the relationship between IQGAP3 and the tumor microenvironment. We examined the correlation between IQGAP3 expression and immune checkpoint–related genes using the TCGA database. IQGAP3 expression was positively correlated with immune checkpoint–related genes in 31 cancer types. These immune checkpoint–related genes were CD274, CTLA4, HAVCR2, LAG3, PDCD1, PDCD1LG2, SIGLEC15, and TIGIT (Supplementary Figure S17). The TISIDB (http://cis.hku.hk/TISIDB/) database showed that IQGAP3 expression was positively correlated with 28 tumor-infiltrating lymphocytes, 45 immune stimulators, 24 immune inhibitors, 41 chemokines, 18 receptors, and 21 major histocompatibility complex (MHC) molecules in different cancer types (Figures 11A–F). These findings indicated that IQGAP3 had an indispensable role in the regulation of the immune response in human cancers.

The results detailed above suggested that IQGAP3 may have roles in cancer progression. Next, we explored the correlation between IQGAP3 expression and sensitivity to different drugs in different cancer cell lines from the GDSC database and Cancer Therapeutics Response Portal (CTRP) (https://portals.broadinstitute.org/ctrp/).

IQGAP3 expression was positively correlated with sensitivity to the drugs TPCA-1, vorinostat, methotrexate, PHA-793887, PIK-93, XMD13-2, BHG712, AR-42, CUDC-101, ispinesib mesylate, SNX-2112, OSI-027, and vinblastine in the GDSC database (Figure 12A and Supplementary Table S8). In the CTRP database, we observed IQGAP3 expression to be positively correlated with sensitivity to the 53 drugs shown in Figure 12B and Supplementary Table S9. In summary, these results demonstrated that IQGAP3 expression was significantly correlated with sensitivity to many drugs in different cancer cell lines.

We showed that IQGAP3 expression was upregulated significantly in NSCLC using TGCG/LUAD/LUSC data. To verify this finding, we measured the mRNA and protein expressions of IQGAP3 in NSCLC. Real-time RT-qPCR and IHC assays showed that IQGAP3 expression was increased in NSCLC cell lines and lung cancer tissues compared with normal lung cells and normal lung tissues, respectively. These findings demonstrated that IQGAP3 expression was upregulated in NSCLC and indicated that IQGAP3 may have a crucial regulatory role in NSCLC progression (Figure 13).