In this study, we investigated the expression of MICAL-L2 across various human malignancies and its association with clinical features using bioinformatics analysis of public datasets and immunohistochemistry (IHC) on human KIRC tissue samples. Bioinformatic analysis revealed a close relationship between MICAL-L2 and the HIF family, shedding light on its novel role in tumor angiogenesis. Utilizing kidney cancer as a model, we combined functional studies in cultured KIRC cell lines (786-Oand Caki-1) and human umbilical vein endothelial cells (HUVECs) to investigate the role of MICAL-L2 in tumor angiogenesis. Our findings indicated that MICAL-L2 co-localized with F-actin in KIRC cells upregulated HIF1α expression and activated the HIF1α/mTOR/VEGF pathway, thereby promoting tumor vascularization in co-culture models. Pharmacological inhibition of actin polymerization in cells and genetic inhibition of MICAL-L2 in vitro both reversed these effects. Further bioinformatic analysis predicted that MICAL-L2 expression levels in tumors could serve as a predictor of sensitivity to anti-angiogenic drugs.
The mRNA and protein levels of the MICAL family across various tumors were analyzed using samples from The Cancer Genome Atlas (TCGA, https://www.cancer.gov/tcga) and the Clinical Proteomic Tumor Analysis Consortium (CPTAC, https://proteomics.cancer.gov/programs/cptac) through UALCAN (http://ualcan.path.uab.edu/) webtool. Cox's regression analysis was conducted employing the R package 'survival' based on the TCGA Pan-Cancer (PANCAN, N = 10535, G = 60499) database. Statistical significance was defined as |Log2FC| > 1 and p < 0.05.
The Stromal, immune, and ESTIMATE scores were calculated using the R package 'ESTIMATE', and their correlation with MICAL-L2 expression was analyzed using the 'psych' algorithm. Tumor-immune cell interactions were analyzed using the 'IPS' and 'TIMER' algorithms. The relationship between MICAL-L2 and immune checkpoint inhibitory and stimulatory molecules was examined using the Immune checkpoint gene analysis module of the Sangerbox (http://sangerbox.com/) webtool.
The protein sequence and 3D model of MICAL-L2 were downloaded from EMBL-EBI (http://www.ebi.ac.uk/). STRING (https://string-db.org/) was utilized to predict the most related functional partners of MICAL-L2 and construct a protein-protein interaction network. Gene Set Enrichment Analysis (GSEA) of MICAL-L2 on Gene Ontology-Biological Process (GO-BP) was performed using the Co-Expedia (https://www.coexpedia.org/) webtool. The enrichment analysis of the MICAL-L2 and VEGF signaling pathway was based on KEGG (https://www.kegg.jp/kegg/kegg1.html).
Immunohistochemistry (IHC) images of MICAL-L2 and HIF family members were downloaded from the Human Protein Atlas (HPA, https://www.proteinatlas.org/) database.
Human kidney renal clear cell carcinoma (KIRC) tissue samples and paired adjacent normal kidney tissues were obtained from Yancheng NO.1 People's Hospital (n = 53). The study was approved by the Ethics Committee of Yancheng NO.1 People's Hospital. All patients provided informed consent. The study was conducted in accordance with the principles of the Declaration of Helsinki.
Human KIRC cell lines 786-O, Caki-1, and human umbilical vein endothelial cells (HUVECs) were acquired from the cell bank of the Chinese Academy of Sciences (Shanghai, China). 786-O and Caki-1 cells were cultured in RPMI1640 medium (Gibco, USA), while HUVECs were cultured in DEME medium (KeyGEN BioTECH, China). All media were supplemented with 10% fetal bovine serum (Gibco, USA), 50 U/ml penicillin, and 50 µg/ml streptomycin. The cells were cultured at 37 °C with a humidity of 5% CO.
Cytochalasin B Treatment: experiments involving cultured cells (786-O, Caki-1, HUVECs) were treated with 1 µM of Cytochalasin B (53373ES03, Yesheng, China) for 24 h to inhibit actin polymerization. No cytotoxicity was detected at this drug concentration.
Ridaforolimus Treatment: experiments involving cultured cells (786-O, Caki-1, HUVECs) were treated with 10 nM of Ridaforolimus (53386ES08, Yesheng, China) for 24 h to inhibit the mTOR signaling pathway. No cytotoxicity was detected at this drug concentration.
IHC was performed on paraffin-embedded human KIRC and adjacent normal tissue sections to detect the expression of MICAL-L2 (1:50, Proteintech, China), HIF1α (1:50, Proteintech, China), VEGFA (1:50, Proteintech, China), CD31 (1:200, Proteintech, China), and CD34 (1:50, Proteintech, China).
Both the evaluation of KIRC tissue specimens and the experimental results were scanned and judged using the double-blind method. Immunohistochemical staining was semi-quantitatively assessed based on staining intensity and the percentage of positive tumor cells, according to the following criteria: score 0: no staining or < 10% positive cells; score 1: faint staining in ≥ 10% of cells; score 2: moderate staining in ≥ 10% of cells; score 3: strong staining in ≥ 10% of cells. Samples were subsequently categorized as either low expression (scores 0-1) or high expression (scores 2-3) for statistical analysis. Microvessel density (MVD) was assessed by two experienced pathologists according to a previously described method. Briefly, three areas of highest vascular density (hotspots) within tumor sections stained with CD34 antibody were selected at low magnification (×40). Microvessels were counted in ten separate ×400 fields (0.25 mm per field) within each hotspot. A microvessel was defined as any CD34-positive endothelial cell cluster with a visible lumen. The mean vessel count per field was calculated and recorded as the MVD value (vessels/HPF).
KIRC cell lines (786-O and Caki-1) were plated at 60% confluence in 6-well plates. For gain-of-function: cells were transfected with pcDNA3.1-MICAL-L2 plasmid or empty vector control (pcDNA3.1; RIBO Biotechnology, China) using Hieff Trans Liposomal Reagent (Yeasen, China) and Opti-MEM (Invitrogen, USA). For loss-of-function: cells were transfected with MICAL-L2-targeting siRNA (5'-CGACGTGAATATCTGCAACAT-3') or negative control siRNA (5'-TTCTCCGAACGTGTCACGT-3'; RIBO Biotechnology, China) using identical reagents. All transfections followed manufacturer protocols. Functional assays were initiated 48 h post-transfection.
RIPA lysis buffer (Thermo Fisher Scientific, USA) supplemented with protease inhibitor complete™ III ULTRA (Roche) was utilized for the extraction of total proteins from KIRC cell lines (786-O and Caki-1). Protein quantification was performed using the enhanced BCA protein assay kit (Beyotime, China) prior to separation by 10% SDS-PAGE gel and transfer onto a PVDF membrane (Millipore, Darmstadt, Germany). The membrane was blocked with Tris buffer containing 0.1% Tween-20 and 5% nonfat milk overnight at 4 °C. Rabbit antibodies specific for MICAL-L2 (1:1000, Proteintech, China), GAPDH (1:1000, Proteintech, China), Tubulin (Proteintech, China), HIF1α (D1S7W, 1:1000, Proteintech, China), p-AKT (Ser473, CST, USA), AKT (1:1000, Proteintech, China), ERK (1:1000, Proteintech, China), p-ERK (Thr202/Tyr204, 1:1000, CST, USA), p-mTOR (Ser2448, 1:1000, CST, USA), mTOR (1:1000, CST, USA), VEGFA (Proteintech, China), p-VEGF Receptor 2 (Tyr1175/D5B11, 1:1000, CST, USA), and VEGFR2 (1:1000, Abcam, China) were used to incubate with the membrane overnight, followed by treatment with HRP-conjugated secondary antibody (1:10000; CST, USA). The signal was detected using the enhanced chemiluminescence detection system (Yesheng, China), according to the manufacturer's instructions.
The original, unprocessed blot images were provided in the supplementary materials. The membranes were cut into strips at specific molecular weight positions before antibody incubation to ensure that each section was probed with the intended antibody reactions.
RNA was extracted from KIRC cell lines (786-O and Caki-1) using TRIzol (15596026, Thermo Fisher Scientific, USA) and homogenized. cDNA was synthesized with a HiScript II Reverse Transcriptase kit (Vazyme, China). RT-qPCR (Taq Pro Universal SYBR qPCRMaster Mix, China) was performed on HIF1α and VEGF, using GAPDH as a reference housekeeping gene. Primers for HIF1α (Forward: 5'-GAACGTCGAAAAGAAAAGTCTCG-3', Reverse: 5'-CCTTATCAAGATGCGAACTCACA-3'), VEGFA (Forward: 5'- GGGCAGAATCATCACGA AGT-3', Reverse: 5'-AAATGCTTTCTCCGCTCTGA-3'), and GAPDH (Forward: 5'-GGAGCGAGATCCCTCCAAAAT-3', Reverse: 5'- GGCTGTTGTCATACTTCTCATGG-3') were purchased from Genescript.
After 48 h of transfection with MICAL-L2 siRNA or control siRNA, 786-O cells were seeded into 96-well plates. Then, 50 µL of EdU (Abbkine, USA) culture medium was added to each well and the plates were incubated for 2 h. Following this, 50 µL of DAPI was added for staining, and the plates were incubated at room temperature in the dark for 5 min. The 786-O cells were then observed under a fluorescence microscope, followed by photographing and statistical analysis.
After 48 h of transfection with MICAL-L2 siRNA or pcDNA3.1-MICALL2, 786-O cells were seeded into 96-well plates. Then, 20 µL of CCK-8 solution (Biosharp, China) was added to each well and the plated were incubated for 4 h. The absorbance at 450 nm was measured using a microplate reader. The average absorbance of three replicates was calculated to determine cell viability. Absorbance values were normalized to the respective control group (set as 100%) using the formula: [(OD_treatment - OD_blank)/(OD_control - OD_blank)] × 100%. The formula for calculating cell viability was as follows: Cell viability (%) = [( Experimental group average absorbance ) - ( Blank group average absorbance )] / [( Control group average absorbance )-( Blank group average absorbance) ]×100%.
An indirect non-contact co-culture model using KIRC cell lines (786-O) and HUVECs was employed in this assay. In this model, HUVECs were co-cultured with MICAL-L2-overexpressing 786-O cells or control 786-O cells using a transwell plate with a 0.4 μm pore filter ( Corning). The 786-O cells were plated in the upper chamber at a density of 2 × 10 cells/cm, and HUVECs were plated in the bottom chamber at the same density. They were co-cultured for 72 h. Subsequently, the HUVECs were incubated with 5-chloromethylfluorescein diacetate (CMFDA, Yeasen, China) for 15 min under normal cell growth conditions. The storage solution of CMFDA was previously diluted to the working solution concentration of 1 µM with serum-free medium and preheated to 37 °C. Next, the culture solution was gently replaced with fresh medium. In advance, a 24-well plate was coated with Matrigel (BD Biosciences, Bedford, MA) and polymerized for 30 min at 37 °C. HUVECs (2 × 10 cells/cm) were incubated in 200 µl conditioned medium for 8 h before imaging. The capillary tubes were quantified under a 100× bright-field microscope by measuring the total length of the completed tubule structures. Three independent experiments were conducted for each treatment.
786-O and Caki-1 cells grown on glass coverslips were fixed with 4% formaldehyde in phosphate-buffered saline at room temperature for 15 min. Following permeabilization with 0.2% Triton X-100 in phosphate-buffered saline for 15 min and blocking with 5% goat serum in phosphate-buffered saline for 45 min, the cells were incubated at 4 °C with MICAL-L2 antibodies (1:50, Proteintech, China) overnight and subsequently incubated with Alexa 488- or 594-conjugated secondary antibodies (Abcam, USA) for 30 min. F-actin was labeled using Actin-Tracker Green (Beyotime, China). Fluorescent images were acquired using ZEISS Axio Vert.A1 microscope and ZEISS Axiocam 305 color camera.
The drug sensitivity analysis was conducted utilizing the 'Candidate Agents' module of the BEST webtool (https://rookieutopia.hiplot.com.cn/app_direct/BEST/). Data from bulk samples and cancer cell lines were sourced from the Genomics of Drug Sensitivity in Cancer (GDSC), Cancer Therapy Response Portal (CTRP), and Precision Cancer Medicine Database (PRISM) databases. Drug response prediction was executed using a ridge regression algorithm implemented in the oncoPredict R package. The model was trained using the transcriptional expression profiles and drug response data from cancer cell lines, achieving satisfactory predictive accuracy as evaluated by default 10-fold cross-validation. The importance of drugs was ranked based on their correlations, and the top 30 sensitive and resistant compounds were listed.
All data analysis was conducted using R vision3.6.3 and SPSS vision 25.0. For normally distributed data, independent samples were analyzed with the unpaired t-test, and paired samples with the paired t-test. For non-normally distributed data, the Mann-Whitney U test (for independent samples) or the Wilcoxon signed - rank test (for paired samples) was used. Correlation analysis was performed using Pearson correlation analysis (for continuous variables) or the chi-square test (for categorical variables). A value of P < 0.05 was considered statistically significant. Graphs and charts were created using GraphPad Prism vision 8.2.1 and Adobe Illustrator 2020.