Biosynthetic Pathways of Dimeric Natural Products Containing Bisanthraquinone and Related Xanthones

Abstract

Many dimeric natural products containing bisanthraquinone and related xanthones with diverse structures and versatile bioactivities have been isolated over the years. However, the complicated biosynthetic pathways of such natural products, which have remained elusive until recently, negatively impact their mass bioproduction and biosynthetic structural modification for drug discovery. In this concept, we summarize the recent progress in gene cluster mining and biosynthetic pathway elucidation of natural products containing bisanthraquinone and related xanthones. These pioneering works may pave the way for further biosynthetic pathway elucidation and structure modification of dimeric natural products through gene and protein engineering.

Mechanistic Investigation on C–C Bond Cleavage of Anthraquinone Catalyzed by an Atypical Nonheme Iron-Dependent Dioxygenase BTG13

“…As shown in Figure 4a and Figure S7, the septet state is the ground state, and the energies of the quintet, triplet, and openshell singlet states are higher than that of the septet state by 1.0, 4.0, and 9.9 kcal/mol, respectively (Figure 4a and Table S6). Moreover, the calculation results show that the spin populations on Fe in 7 S7). These electronic structures are further confirmed by the analysis of spin natural orbitals (SNO) (Figure 4b).…”

“…6 The effect of the carboxyl-group modification of lysine on the reaction energy barrier was first studied. Two models with Kcx and Lys as the iron ligands were built and named the KcxO differences are observed between the optimized geometries of 5 (7) (E•S•O 2 ) I-KcxO and 5 (7) (E•S•O 2 ) I-LysN (Figure 7a and b). Compared with 5 (7) (E•S•O 2 ) I-KcxO , the Fe−O a distance in 5 (7) (E•S•O 2 ) I-LysN (2.30 Å (2.85 Å)) is much longer than that of 5 (7) (E•S•O 2 ) I-KcxO (2.00 Å (2.32 Å)), but the O a −O b distance (1.24 Å (1.22 Å)) is shorter than that of 5 (7) (E•S•O 2 ) I-KcxO (1.29 Å (1.26 Å)).…”

“…Two models with Kcx and Lys as the iron ligands were built and named the KcxO differences are observed between the optimized geometries of 5 (7) (E•S•O 2 ) I-KcxO and 5 (7) (E•S•O 2 ) I-LysN (Figure 7a and b). Compared with 5 (7) (E•S•O 2 ) I-KcxO , the Fe−O a distance in 5 (7) (E•S•O 2 ) I-LysN (2.30 Å (2.85 Å)) is much longer than that of 5 (7) (E•S•O 2 ) I-KcxO (2.00 Å (2.32 Å)), but the O a −O b distance (1.24 Å (1.22 Å)) is shorter than that of 5 (7) (E•S•O 2 ) I-KcxO (1.29 Å (1.26 Å)). Moreover, changing the axial ligand from Kcx to Lys could significantly change the spin distributions on Fe (from 4.33 (4.23) to 3.90 (3.83)) and dioxygen (from −0.73(1.44) to 0.10 (1.90); Figure 7a and b).…”

“…Without the carboxylation modification, the iron−dioxygen complex in 5 (7) (E•S•O 2 ) I-LysN is more like an Fe II −O 2 species with dioxygen at the singlet or triplet state rather than the Fe III −O 2…”

“…•− species in 5 (7) (E•S•O 2 ) I-KcxO , in which the dioxygen is activated. In addition, a larger quintet−septet energy gap for 5 (7) (E•S•O 2 ) I-LysN is observed, where the energy of 7 (E•S• O 2 ) I-LysN is 11.0 kcal/mol lower than that of 5 (E•S•O 2 ) I-LysN (Figure 7b and Table S15).…”