In an effort to identify novel inhibitors of Chikungunya (CHIKV), Dengue (DENV) and Zika (ZIKV), a systematic study with 311 extracts from tropical Euphorbiaceae species was performed in a virus-cell-based assay for CHIKV and DENV and ZIKA NS5 inhibition assays.
The French Guianese species Sandwithia guyanensis and Sagotia racemosa, from which bark extracts exhibited significant anti-CHIKV activities, were first investigated. Following a classical bio-guided isolation workflow, more than 20 new diterpenes were characterized but none of them showed significant antiviral activity. [1]
To address this issue, a Feature-Based Molecular Network (FBMN) was built from the LC-MS2 data acquired from the chromatographic fractions of both extracts. FBMN is a computational method that bridges data processing tools for LC-MS2 and molecular networking (MN) analysis on GNPS [2]. Then the Network Annotation Propagation (NAP) workflow that improve in silico fragmentation candidate structure ranking using spectral networks to propagate information from spectral library matching, allowed highly reliable identification of phorbol analogues. Those compounds present in trace amounts in both extracts, provides a plausible explanation for the loss of the biological activity observed during the bioassay-guided isolation procedure [3].
In a second study, a bioactive prioritization approach based on the merging of taxonomical and bioassays data over the MN built from the initial set of Euphorbiaceae extracts led to the targeted isolation and characterization of several dual inhibitors from Codiaeum peltatum [4]. Both studies exemplify how MN and recent advances in molecular annotation can be implemented in phytochemical studies to understand drawbacks and improve the bioactive compound discovery process.
Fig. 1
