Fig 1.
Workflow to integrate computational network biology and experimental testing for characterizing the anticancer activity of DPP4 inhibitors.
The schematic outlines the two-phase strategy of this study. Phase 1 (Computational Studies) begins with the construction of a DPP4-centric protein-protein interaction (PPI) network using Cytoscape and the STRING database (confidence score of selected nodes ≥ 0.4). Pathway enrichment analysis of this network identifies putative cancer-associated pathways. Key hub genes from this initial analysis are selected to generate an expanded PPI network, which undergoes a second enrichment analysis to pinpoint central mechanisms. Concurrently, the genomic alteration landscape of DPP4 is analyzed in patient data from TCGA. These analyses collectively generate a testable hypothesis on the anticancer pathways of DPP4 inhibitors. Phase 2 (Experimental Testing) involves the validation of these computational predictions. A suite of in vitro assays—including cellular viability, anchorage-independent growth, apoptosis, cell cycle analysis, and gene expression profiling—is performed using both commercial and synthesized DPP4 inhibitors to functionally validate the hypothesized mechanisms.
Fig 2.
Protein-protein interaction networks for DPP4.
A) Initial direct interactions network using DPP4 as a seed querying the STRING database with a medium confidence score threshold of 0.400 and limited to 10 first-shell interactors to define a high-confidence core interactome. The top 10 significantly enriched KEGG pathways (FDR ≤ 0.05) are displayed. The FDR value for each pathway is indicated. Pathways strongly associated with cancer mechanisms, such as ‘Proteoglycans in cancer’, ‘Focal adhesion’, and ‘ECM-receptor interaction’, are prominently featured. B) Expanded direct interactions network using DPP4 plus three additional gene seeds identified from the initial network: FN1, ITGB1 and CAV1. The additional seeds were selected based on their high connectivity (hub status) in the first network and their established biological relevance to the enriched cancer pathways. The same confidence and interactor limits were applied. Enrichment analysis of the expanded network confirmed the cancer-related pathways with greater statistical robustness and identified the PI3K-Akt signaling pathway as the most significantly enriched.
Fig 3.
Genomic alterations of DPP4 in cancer.
A) Genetic alterations of DPP4 across various cancer types, derived from a dataset of 10,967 samples from 10,953 cancer patients. B) mRNA expression levels of DPP4 in colorectal cancer patients, based on 594 samples. C) Frequency of DPP4 genetic mutations in colorectal cancer patients, analyzed from 594 samples. All patient data were sourced from TCGA PanCancer Atlas Studies [38] and analyzed using cBioPortal [39]. Source: National Cancer Institute as the source.
Table 1.
The 50% inhibitory concentration (IC50, µM) values for DPP4 inhibitor treatment in SW620, HCT116, SW480, and Caco2 colorectal cancer cells for 48 hours. Values are expressed as the mean ± SD of three independent experiments in duplicates (n = 6).
Table 2.
Combination index (CI) values for combined treatment of DPP4 inhibitors and doxorubicin in CRC cell lines. CI < 1, = 1, and > 1 indicate synergistic, additive, and antagonistic effects, respectively. µM: micromolar.
Table 3.
Combination index (CI) values for combined treatment of DPP4 inhibitors and 5FU in CRC cell lines. CI < 1, = 1, and > 1 indicate synergistic, additive, and antagonistic effects, respectively. µM: micromolar.
Fig 4.
The percentage reduction of IC50 for chemotherapy in combination with DPP4 inhibitor treatment in HCT116, SW620, Caco2, and SW480 colorectal cancer cells.
Cells were cultured and allowed to attach overnight. The next day, cells were treated with different concentrations of selected compounds of DPP4 inhibitors with either doxorubicin or 5-FU, for 48 h. After that, the cell viability was determined using an MTT assay. *P < 0.05 significantly different from respective cisplatin treatment. ****P < 0.0001.
Fig 5.
Effect of DPP4 inhibitors (Saxagliptin, Sitagliptin, AE-AMID, and PA-AMID) on colony count and size of HCT116 using colony formation assay.
P-value < 0.05 expresses significantly different from respective untreated cells’ status; while asterisk: ns (not-significant) P > 0.05; * P ≤ 0.05; ** P ≤ 0.01; *** P ≤ 0.001; **** P ≤ 0.0001 (according to GraphPad Prism 9). The standard deviation of values did not exceed 5%. IC50: The 50% inhibitory concentration.
Fig 6.
Histogram and partial summary of the effects of sub-IC₅₀ concentrations of Saxagliptin, Sitagliptin, AE-AMID, and PA-AMID on HCT116 cell cycle progression after 48 hours of treatment.
DNA content was analyzed by flow cytometry following PI staining, with histograms indicating the distribution of cells across G0/G1, S, and G2/M phases. Gating was performed based on fluorescence intensity profiles, using the untreated control as a reference. Minor adjustments were made to accommodate treatment-induced shifts while maintaining consistency in phase identification.
Fig 7.
A) Dot plot for annexin V-FITC/ PI staining expressing the apoptotic effect of DPP4 inhibitors, Saxagliptin, Sitagliptin, AE-AMID, and PA-AMID (double IC50) for 48 hours treatment against HCT116 cells.
Where Q3 showed viable cells, Q1 necrotic cells, Q2 late apoptotic, and Q4 early apoptotic. B) Percentages of healthy, apoptotic, and necrotic cells expressed as mean, SD did not exceed 5%. P-value <0.05 indicates statistical significance in comparison to untreated control, while asterisk: ns (not-significant) P > 0.05; * P ≤ 0.05; ** P ≤ 0.01; *** P ≤ 0.001; **** P ≤ 0.0001 (according to GraphPad Prism 9).
Fig 8.
Effect of DPP4 inhibitors Saxagliptin, Sitagliptin, AE-AMID, and PA-AMID on multiple RNA expression of specific genes in HCT116 colorectal cancer cell line.
CD26: the gene for dipeptidyl peptidase enzyme; Bcl-2: the gene for Bcl-2; VEGF: the gene for vascular endothelial growth factor; Fold difference expressed as mean±SD and was measured using ΔΔCt method. All experiments were run in duplicates and with three independent experiments. P-value < 0.05 express significantly different from respective untreated cells’ status; while asterisk: ns (not-significant) P > 0.05; * P ≤ 0.05; ** P ≤ 0.01; *** P ≤ 0.001; **** P ≤ 0.0001 (according to GraphPad prism 9). IC50: The 50% inhibitory concentration.