Introduction: Multi-component natural products challenge single-target paradigms in oncology. We present a reproducible network-pharmacology workflow integrating target prediction, druggability filtering, protein–protein interactions (PPI), and GO enrichment to map how three major Crocus sativus bioactives (kaempferol, crocin, picrocrocin) collectively engage breast-cancer–relevant biology. We asked whether multi-compound analysis reveals convergence on cancer hallmarks and identifies hub targets suitable for enrichment and early discovery.
Significance: A druggability-filtered pipeline (SEA→UniProt→Pharos→STRING→ToppFun) yields 86 curated targets (34 cancer-related), hub nodes ESR1 (degree=30) and FGF2 (13), and a clear synergy map guiding multi-compound design.
Methods: SMILES were queried against SEA to collect putative protein targets per compound. Targets were curated to human, reviewed gene names in UniProt. Clinical relevance was prioritized by retaining genes labeled Tclin/Tchem in Pharos (queried with “breast carcinoma”). Intersections among compound target lists were computed (InteractiVenn). We built a STRING v11.5 PPI network (combined compound targets; score ≥0.4; all evidence channels) and visualized it in Cytoscape; isolated nodes and pseudogenes were removed. Degree centrality was used to nominate hubs. GO Biological Process enrichment (ToppFun, Bonferroni-corrected q < 0.01) was performed on the combined study list used in your enrichment runs (compound targets plus curated breast-cancer genes), consistent with your analysis plan.
Results: The curated network comprised 86 unique proteins after filtering. Of these, 34 are flagged as breast-cancer–related in your Cytoscape table. High-degree hubs were: ESR1 (degree 30), IL2 (17), CYP1A1 (15), AKR1B1 (14), FGF2 (13). Compound contributions were asymmetric: kaempferol accounted for 54 nodes; picrocrocin-only 11; crocin-only 3; crocin ∩ picrocrocin 15; all three 2; and kaempferol ∩ picrocrocin 1. Thus, kaempferol supplies breadth and key hubs (e.g., ESR1, CYP1A1), while crocin/picrocrocin reinforce a focused subnetwork (e.g., FGF2, LGALS3).
GO enrichment showed robust convergence on cancer-relevant processes:
Positive regulation of cell migration/motility/locomotion (each 12 hits; Bonferroni q ≈ 3–7×10⁻⁶), with contributors SELL, SELP, FGF1/FGF2, CYP1B1, ABCC1, NOX4, MYLK, TERT, etc.
Angiogenesis/vasculature development (e.g., positive regulation of angiogenesis, 7–8 hits; q ≈ 2–3×10⁻⁴ to 2×10⁻⁴).
Cellular response to xenobiotic stimulus and xenobiotic transport (Bonferroni-significant; hits include ABCB1, ABCG2, ABCC1).
Notably, one-carbon metabolic process was strongly enriched (q ≈ 3×10⁻⁵; 5 hits) driven by carbonic anhydrases (CA9, CA12, CA1, CA3, CA4)—a signature contributed by the kaempferol subset,implicating tumor pH regulation/metabolism. Together, topology and GO profiles indicate coordinated coverage of motility/invasion, angiogenesis, and drug resistance, with metabolic/hormonal breadth added by kaempferol (e.g., ESR1/ESR2, carbonic anhydrases).
Discussion/Impact:
Methodologically, the SEA → UniProt → Pharos (Tclin/Tchem) → STRING (≥0.4) → Cytoscape pipeline yields a compact, clinically oriented network rather than an over-inclusive target list. Degree-based hubs (ESR1, FGF2, CYP1A1) highlight tractable points where multi-compound modulation could be synergistic. The intersection structure suggests strategy: pair kaempferol (broad network penetration; endocrine/metabolic nodes) with crocin/picrocrocin (FGF/lectin/oxidative axes) to maximize coverage of migration/angiogenesis while addressing metabolism and xenobiotic transport. Because your GO run used the combined study list (compound targets + breast-cancer genes), the enriched terms directly reflect clinically relevant biology while preserving compound-specific signals.
Conclusion: Our pipeline reveals that saffron bioactives converge on cell motility/invasion, angiogenesis, and drug resistance, while offering complementary coverage: kaempferol (metabolic/hormonal breadth, ESR1/CA hubs) and crocin/picrocrocin (FGF/oxidative/lectin axes). By focusing on druggable, breast-cancer–linked targets, we highlight ESR1, FGF2, and ABC transporters as tractable nodes for experimental prioritization.
These results justify kaempferol-anchored combinations with crocin/picrocrocin to jointly suppress migration/angiogenesis and drug-transport axes, offering a reproducible framework for natural-product–based cancer drug discovery.
Keywords: Network pharmacology, Saffron bioactives, Breast cancer, Protein-protein interaction, Druggability