Modulatory effect of metformin and its transporters on immune infiltration in tumor microenvironment: a bioinformatic study with experimental validation

Abstract

Metformin is a traditional antidiabetic drug for type 2 diabetes mellitus. However, it showed antitumor activity in many types of tumors, and it also has an influence on tumor metastasis in several types of tumors. It is transported through organic cationic transporters (OCTs), OCT1, OCT2, and OCT3, into the cells or into tumor microenvironment (TME). The complex interaction of metformin and its transporters on immune infiltration in TME of different types of tumors of The Cancer Genomic Atlas (TCGA) is not yet studied. The objective of this study is to identify the most suitable therapeutic target of tumors and immune infiltrates for metformin and its transporters in the TME. TIMER2.0, a bioinformatic tool, and other computational analysis were used to investigate this complex interaction; moreover, the identification of metformin target protein in TME is also investigated. The results revealed that the most suitable therapeutic target for metformin and OCTs among 32 types of TCGA data tumor types is Breast Invasive carcinoma (BRCA), and the most relevant immune infiltrate among 14 types of immune infiltrates that yields better prognosis and better therapeutical effect in TME is Macrophage M1. Furthermore, metformin showed a cytotoxic effect and an inhibitory effect on Urokinase Plasminogen Activator (uPA) gene expression in a concentration dependent fashion in MDA-MB-231 breast cancer cell line. This may suggest that metformin is a promising antitumor drug, stimulant for natural antitumor immune infiltrates, and inhibitor for metastasis in breast cancer.

Keywords: Immune infiltrates; Metformin; Organic cationic transporter; TIMER2.0; Tumor microenvironment; Urokinase plasminogen activator.

Fig. 1

Graphical Design for the bioinformatic study of Metformin and experimental validation

Fig. 2

Representative figure for the Outcome module of Macrophage M1, as an example for the 14 immune infiltrates, against 32 tumor types of TCGA data. Kaplan–Meier Curve displays immune infiltrate clinical outcome on tumor types. Split Infiltration Percentage of Patients = 50%. Survival Time Between in months = 200. Significant result (p−value < 0.05). Increase risk result (Z score > 0) and decrease risk (Z score < 0)

Fig. 3

Frequency of OCT genes mutations in different tumor types. This analysis was done using Mutation module in TIMER2.0 website (http://timer.cistrome.org/). SLC22 A1 or OCT1 = Organic Cationic Transporter 1. SLC22 A2 or OCT2 = Organic Cationic Transporter 2. SLC22 A3 or OCT3 = Organic Cationic Transporter 3

Fig. 4

OCT genes expression in BRCA as compared to normal controls. The analysis was done by using Gene_DE module in TIMER2.0. Statistical significance computed by the Wilcoxon test that is annotated by the number of stars (*p−value < 0.05; ***p−value < 0.001). SLC22A1 or OCT1 = Organic Cationic Transporter 1. SLC22A2 or OCT2 = Organic Cationic Transporter 2. SLC22A3 or OCT3 = Organic Cationic Transporter 3. BRCA Normal = Breast normal tissue

Fig. 5

Possible Metformin target proteins. SwissTargetPrediction website has been used to elucidate the predicted results. Chembl ID (CHEMBL3286) is the id number of Urokinase−type plasminogen activator (uPA) in ChEMBL database website (https://www.ebi.ac.uk/chembl/target_report_card/CHEMBL3286/). Uniprot ID (P00749) is the id number of Urokinase−type plasminogen activator (uPA) in Uniprot website (https://www.uniprot.org/uniprotkb/P00749/entry.). Plasminogen Activator Urokinase (PLAU) = Urokinase−type Plasminogen Activator (uPA)

Fig. 6

Frequency of uPA gene mutations in different tumor types. This analysis was done using Mutation module in TIMER2.0 website (http://timer.cistrome.org/). PLAU = uPA = Urokinase Plasminogen Activator

Fig. 7

Interaction of uPA protein with Metformin. a and b show uPA protein structure. c and d show the 3 poses of metformin in hydrophobic pocket of uPA protein. e shows the yellow highlighted amino acids which are the targets of metformin molecule 3 poses in the hydrophobic pocket of chain A in uPA protein by molecular docking of Schrodinger suite version2023−3

Fig. 8

Metformin conformations with uPA protein. a shows all 3 poses of metformin combined in the hydrophobic pocket. b shows the structure of metformin in 3D. c and d show pose 1 only in hydrophobic pocket. e and f show pose 2 only in hydrophobic pocket. g and h show pose 3 only in hydrophobic pocket

Fig. 9

MDS of metformin with uPA protein. a, b, and c are the three poses of interaction of metformin (Arrow) with uPA protein. The stable interaction should last at least 100 ns

Fig. 10

Metformin cytotoxicity in MDA−MB−231 cancer cell line. Metformin was incubated in cells for 48 h, and % of cell viability was presented as mean ± SD of triplicate cultures using MTT Assay

Fig. 11

Effect of Metformin on uPA gene expression. Metformin was added in concentrations of 0 (Control), 10% of its IC50 (770.2 µM), and 20% of IC50 (1540.4 µM) in triplicate cultures, and data presented as mean ± SD. Statistical analysis was done by One Way ANOVA and Tukey Kramer as Post ANOVA test. Significance was accepted at p ≤ 0.05. @ Significantly different from control. # Significantly different from 10% of IC50. (**p−value < 0.01; ****p−value < 0.0001)

Fig. 12

Effect of Metformin on uPA protein content in culture media of MDA−MB−231. Metformin was added in concentrations of 0 (Control), 10% of its IC50 (770.2 µM), and 20% of IC50 (1540.4 µM) in triplicate cultures, and data presented as mean ± SD. a % of uPA protein expression. b Western Plot of uPA. Statistical analysis was done by One Way ANOVA and Tukey Kramer as Post ANOVA test. Significance was accepted at p ≤ 0.05. @ Significantly different from control. # Significantly different from 10% of IC50. (**p−value < 0.01; ****p−value < 0.0001)

Fig. 13

Proposed role of Metformin and its transporters in Tumor Microenvironment