Lipid metabolism regulatory activity and adverse effects of fungi-derived butyrolactone I

Abstract Butyrolactone I (BTL-I), a butenolide compound isolated from land or marine-derived fungi, has been reported to show diverse activities. To further study the pharmaceutical potential of BTL-I, transcriptome and bioinformatics analysis of BTL-I treated HepG2 cells were taken. BTL-I was revealed with lipid metabolism regulatory activity and confirmed by increasing the mRNA expression of related genes, such as LXRα and its target gene UGT1A1. However, the obvious chemical carcinogenesis of BTL-I was also disclosed. BTL-I could significantly increase the mRNA and protein levels of oncogenes such as CYP1A1. Molecular docking of BTL-I and its analogs were performed to understand the active or toxic effects. Although BTL-I showed attractive activities, enough attention must be paid to its adverse effects in its further development. Graphical Abstract


Introduction
In microbial natural products study, butenolide derivatives, possessing the a,b-unsaturated c-butyrolactone skeleton, were frequently isolated from land or marine-derived fungi, especially the Aspergillus species (Cheng et al. 2019). Several butenolides, especially the well-known compound butyrolactone I (BTL-I), have been reported to show diverse biological activities (Haroon et al. 2013). However, the effects of the butenolides on lipid metabolism regulatory were less frequently reported.
During our study of marine-derived fungi, many butenolide derivatives have been obtained with several different activities (Lin et al. 2019;Qi et al. 2021). The pancreatic lipase inhibitory activities of butenolides revealed in our recent study, prompted us to further investigate their effects on regulating lipid metabolism (Peng et al. 2022). Herein, the most common butenolide compound BTL-I, which has been prepared in large quantities from marine-derived fungi, will be studied with its potential on regulating lipid metabolism, using transcriptome analysis and following molecular biology validation. Due to the potential toxicity of the mycotoxin butanolide, the chemical carcinogenesis of BTL-I will also be revealed and of great concern.

Results and discussion
BTL-I showed no cytotoxicity against HepG2 cells at the concentration of 50 lM (Peng et al. 2022). In order to explore the potential activity of BTL-I (Supplementary material, Figure S1A), we performed transcriptome analysis of the HepG2 cells treated with or without BTL-I (50 lM). DEseq was used to identify the statistically significant differentially expressed genes (DEGs) for the pairwise comparisons between these three groups. The results revealed that there are 169 genes were up-regulated and 113 genes down-regulated (Supplementary material, Figure S1B). The heatmap showed partial DEGs as follows (Supplementary material, Figure S1C). Gene Ontology (GO) analysis highlighted lipid metabolic process as a top three enriched pathway that increased visibly after BTL-I treatment group (Supplementary material, Figure S2A). But the KEGG pathway analysis demonstrated that chemical carcinogenesis of BTL-I was terribly vital (Supplementary material, Figure S2B), to which we must pay more careful attention.
The RNA sequencing data (GSE200538) revealed that BTL-I regulated the lipid metabolism, so we verified the mRNA expression levels of related genes regulating lipid metabolism by qRT-PCR. BTL-I could increase the mRNA levels of LXRa and its target gene UGT1A1 obviously ( Figure 1A). It also seemed to enhance LXRb and LXR target gene ABCA1, although there was no significant difference. The effect of BTL-I on increasing LXRa was further confirmed in primary hepatocytes ( Figure 1B) ). To explore more functions of BTL-I, cells hypoxia/reoxygenation were performed in vitro to simulate hepatic ischemia/reperfusion injury in vivo. BTL-I (50 lM) could protect cells against hypoxia/reoxygenation by decreasing pro-inflammatory cytokines TNFa and CXCL10 in HepG2 cells and apoptosis gene mRNA expression level of BAX in primary hepatocytes ( Figure 1C and D).
Although BTL-I has been revealed with lipid metabolism regulatory activities, its serious deficiency was also appeared in the transcriptome and bioinformatics analysis. BTL-I significantly increased the mRNA and protein levels of oncogenes such as CYP1A1, CYP3A7, ALDH3A1, GSTA2 in HepG2 cells ( Figure 1E and F). CYP1A1 is one of the main cytochrome P450 enzymes, examined extensively for its capacity to activate compounds with carcinogenic properties (Androutsopoulos et al. 2009). Aslo, it upregulated the mRNA expression levels of key inflammatory cytokines including IL6, TNFa and apoptosis-related genes BCL2, BAX indicating cell damage in primary hepatocytes coincide with HepG2 cells ( Figure 1G). These findings suggested overcoming its great chemical carcinogenesis might be impassable for new drug development (Zhang et al. 2021).
To further understand the interaction between BTL-I and the potential active or toxic target, the induce-fit module in the Schr€ odinger suite was employed to perform the molecular docking analysis. Homology models of LXRa, LXRb, CYP1A1, ALDH3A1, GSTA2, and CYP3A7, were selected and subjected to docking analysis with the butenolide molecules. Except BTL-I, some butenolide enantiomers, (þ)/(À)-asperteretals G-I (1a/1b, 2a/2b, 3a/3b) (Supplementary material, Figure S3) (Peng et al. 2022), were also employed. As a result, those butenolide molecules fit comfortably into the binding pocket with the similar binding positions in each model. The 2D and 3D binding modes of BTL-I with LXRa, CYP1A1, ALDH3A1 predicted by molecular docking were showed in Figure S4 (Supplementary material). In the 2D binding model of BTL-I with 3IPQ, one phenolic hydroxyl formed hydrogen bond with the active-site residue MET298, and another benzene moiety played a key role to form a p-p stacking interaction with residue PHE326 of LXRa. Something like that, in the model with 4H80, two phenolic hydroxyl groups formed hydrogen bonds with residues TYR60 and GLU68, and one of the benzene moiety formed a p-p stacking interaction with residue TRP233 of ALDH3A1 (Supplementary material, Figure S4).
The binding free energy values (Glide gscore values) of those seven butenolide molecules with those models of 6 proteins are listed in the Figure S3 (Supplementary material). Most of the butenolide molecules appear little difference of the binding ability to each protein. However, the differences in pairs of enantiomers to individual proteins are obvious relatively, such as 1a/1b and 3a/3b with CYP1A1, 2a/2b with ALDH3A1. Although the enantiomers of chiral substances have the same physicochemical properties, their biochemical activities or toxicities can be different because biochemical processes usually show high stereo-or enantioselectivity (Muller and Kohler 2004). It may give us an indication that different butenolide enantiomers may cause different toxicity.
Our study revealed the lipid metabolism regulatory activity of BTL-I for the first time. However, the chemical carcinogenic effect of the BTL-I is also significant, such as promoting CYP1A1. Because of the easy access by fungal fermentation, BTL-I have been studied with its pharmacological activities intensively, with its adverse effects neglected. Hopefully, our results will lead to a new look at this important fungal metabolite. We think structural modification should be a feasible lead optimization strategy of BTL-I. Otherwise, the chemical carcinogenesis or toxicity issue may be the particular importance or even fatal in its drug development.

Conclusions
We describe here the discovery of lipid metabolism regulatory activity of marinederived butenolide compound BTL-I, by transcriptome analysis and following molecular biology validation. However, the obvious chemical carcinogenesis of BTL-I was also disclosed. BTL-I could significantly increase the mRNA and protein levels of oncogenes such as CYP1A1. Although BTL-I showed attractive biological activities, its adverse effects should not been neglected.