Identifying purgative targets of sennoside A via in situ biotransformation of prodrug-based probes

Highlights

• •A strategy for in situ metabolically synthesized active drug-based probes was proposed.
• •The potential purgative targets of SA were successfully hooked and identified.
• •The work provided a new insight for studying the direct targets of unstable active drugs.

Graphical abstract

Prodrugs need to be converted to active drugs to exert their pharmacological activities. Identifying the direct targets of active drugs is essential to elucidate the pharmacological mechanisms of prodrugs, but remains challenging, especially for active drugs with low stability. The activity-based probes have been the primary technology for drug target identification because of their main advantage in identifying the direct drug targets [1,2]. Yet, this technology is unsuitable for prodrugs whose active drugs have low stability, as the instability of active drugs results in difficulty in synthesizing active drug-based probes. Here, we report a strategy for in situ metabolically synthesized active drug-based probe via biotransformation of prodrug-based probe with gut microbes to directly identify the mechanism-of-action targets. Sennoside A (SA), a typical prodrug, is converted into active pharmacological rheinanthrone by gut microbes [3,4]. As the rheinanthrone is highly susceptible to oxidation, its direct targets remain unclear [4,5]. Here, we synthesize a SA-based probe (SAP) in which a bio-orthogonal reactive group, alkyne is covalently conjugated to SA. The alkyne group has emerged as preferred tags incorporated into probe design for the enrichment of target proteins based on the bio-orthogonal reaction. Moreover, due to its small size with minimal structural perturbation and its chemical inertness in organisms, the alkyne will not be metabolized but can be transferred from the prodrug-based probe to the active drug-based probe. The rheinanthrone-based probe (RAP, the active drug-based probe) is successfully converted from the SAP by the gut microbes in an anaerobic condition. Coupling this probe with mass spectrometry (MS), the mechanism-of-action targets of SA were identified. Meanwhile, the direct interaction between the identified target protein and rheinanthrone is evaluated by surface plasmon resonance (SPR). This study provides an example for studying the direct targets of unstable active drugs generated by prodrugs. All animal experiments were approved by the Animal Ethics Committee of the Beijing University of Chinese Medicine (Approval No.: BUCM–4–2018091304–3023), and all procedures were performed by institutional guidelines and ethical standards.

To verify gut microbes are the essential medium for the biotransformation of SA, we investigated the purgative effect of SA with antibiotic (ampicillin) interfering treatment (Fig. S1A). Compared with the astriction model mice (AM) group, the time to first defecation of the SA-treated mice (SAT) significantly decreased. However, with the antibiotic interference (Amp), the time to first defecation significantly increased compared with the SAT group (Fig. S1B). And the time to first defecation had no significant difference between the Amp group and the AM group (Table S1). These results demonstrate that antibiotic weaken the purgative effect of SA. To further confirm that gut microbes are the essential medium to exert the purgative effect of SA, we detected the effect of gut microbes on the production of rheinanthrone by MS. As rheinanthrone is easily oxidized to rhein, we took the derivatization of rheinanthrone with p-nitrosodimethylaniline (pNDA). To quantify the content of rheinanthrones, we mainly detected two rheinanthrone related products, namely rhein and azolylmethyl-rheinanthrone (Fig. S2A). Firstly, we built a valid method for determining the content of the two rheinanthrone-related products, and the results of the method validation was shown in Tables S2 and S3. After incubating SA with gut microbes (SA + GM), the content of rheinanthrone-related products showed significant increase with the prolongation of incubating time while SA exhibited high stability in culture medium (SA + CM) (Figs. S2B−G). However, with the ampicillin interference (SA + GM + Amp), the products almost could not be detected (Figs. S2B−G). These results show that the SA was converted into rheinanthrone by the gut microbes.

To hook the mechanism-of-action targets of SA, we first designed and synthesized the prodrug-based probes. In the structural design of the prodrug-based probe, SA was used as a recognition group to bind target proteins, and the alkyne functional group was used to enrich the target proteins (Fig. 1A). We synthesized the SA-based probes (SAPs) via amide reaction of SAs with propargylamines. The products were purified and subsequently characterized by MS and nuclear magnetic resonance (NMR) (Figs. 1B and C, and S3). With the successful synthesis of the SAPs, we next detected the bioconversion products of SAPs by gut microbes in an anaerobic condition to confirm whether the SAPs can be converted into RAPs (Fig. S4A). To accurately quantify the production of RAPs, we took the derivatization of the bioconversion products of SAPs with pNDA. We first optimized the collision energies of MS (Figs. S4B and Table S4) and built a valid semi-quantitative method to detectthe products of RAPs (Figs. S4C-F and Table S5). After incubating SAPs with gut microbes (SAP + GM), the content of the SAP significantly decreased with the prolongation of incubating time (Fig. 1D) while the SAP presented high stability in culture medium (SAP + CM) (Fig. S5). The content of RAP-derivatives and oxidation products (RAP-oxides) significantly increased in the SAP + GM (Figs. 1E and F). These results suggest the prodrug-based probes could be metabolized by gut microbes.

Fig. 1.

Synthesis, characterization, and biotransformation of sennoside A (SA)-based probe (SAP) and the identification of the mechanism-of-action targets of SA. (A) The synthetic route of SAP and the scheme of potential purgative target identification. (B, C) The structural characterization of SAP. The analysis results of mass spectrometry (MS) (B) and proton nuclear magnetic resonance (¹H NMR) (C). (D–F) The investigation of the conversion of SAP to rheinanthrone-based probe (RAP) by gut microbes. Extracted ion chromatograms (EIC) of SAP (D), azolylmethyl-rheinanthrone-2-carboxylic acid prop-2-ynylamides (RAP-derivatives) (E), and rhein-2-carboxylic acid prop-2-ynylamides (RAP-oxides) (F). (G) The results of the purgative effect of SAP in astriction model mice (AM). (H–J) The molecular docking results of potential purgative proteins binding with rheinanthrone and the interaction details in the active site: Nucleobindin-1 (H), Calmodulin-3 (CALML3) (I), and Actin, gamma-enteric smooth muscle (J). (K) The binding affinity of the interaction between CALML3 and rheinanthrone by surface plasmon resonance (SPR) analysis. Rheinanthrone at 3.125−200 μM was injected over the chip surface and the responses were recorded (colorful lines). The data were fitted to a l:l binding mode using the Biacore insight evaluation software (black lines). EDC: 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride; HOBT: 1-hydroxybenzotriazole; DIPEA: N, N-diisopropylethylamine; LC-HRMS: liquid chromatography-high resolution mass spectrometry; DMSO: dimethyl sulfoxide; SAP + CM: SAP with culture medium; SAP + GM: SAP with gut microbes; NC: normal control mice; SAT: SA-treated mice; SAPT: SAP-treated mice; Lys: lysine; His: histidine; Val: valine; Asn: asparagine; Thr: threonine; Arg: arginine; Leu: leucine; Nag2: the cofactor of actin, gamma-enteric smooth muscle; Pro: proline; Phe: phenylalanine; la: aalanine. ^∗P < 0.1, ^∗∗P < 0.01, ^∗∗∗P < 0.001; ns: no significance.

Having proved the validation of the in situ metabolically synthesizing RAPs, we next evaluated the purgative action of SAPs to assess the ability of RAPs to bind target proteins. As shown in Fig. 1G and Table S1, compared with the AM, the time to first defecation of SAP-treated mice (SAPT) significantly decreased, and there was no significant difference between the SAPT and SAT. These results illustrate that the RAPs generated by SAPs may exert a laxative effect by binding the target proteins in the colon. Thus, the metabolically synthesized RAPs could be used for hooking laxative targets in an anaerobic condition.

We performed a pull-down experiment to enrich the target proteins by utilizing RAPs combined with streptavidin beads and then identified these proteins by MS. First, we incubated the SAPs with gut microbes to obtain the RAPs in an anaerobic condition. Then, RAPs were incubated with proteins digested from colon tissue in an anaerobic condition for 1 h, followed by adding the streptavidin beads into the incubating solution to isolate the proteins labelled by probes. A total of 178 proteins were identified (Fig. S6A). To eliminate the non-specific adsorption effects of streptavidin beads, we mixed streptavidin beads with proteins digested from colon tissue, and 156 proteins were identified (Fig. S6A). To exclude the interference of the non-converted SAP, we further mixed SAPs with proteins digested from colon tissue and 249 proteins were identified (Fig. S6A). In summary, we identified 25 unique proteins in the RAPs group (Table S6). Pathway enrichment analysis of these proteins revealed a top-ranking functional cluster of “Smooth Muscle Contraction”, which is closely related to diarrhea (Fig. S6B). According to the functional analysis of the 25 proteins in the Uniprot database and papers, we found 3 potential target proteins of rheinanthrone to exert its purgative action, including Actin, gamma-enteric smooth muscle, Calmodulin-3, and Nucleobindin-1. Studies have shown that these potential target proteins may contribute to diarrhea by different pathways, such as promoting intestinal peristalsis and regulating bidirectional transmembrane water flow. These results indicate that rheinanthrone may exert purgative action by multi-target regulation.

To determine whether the potential purgative proteins can bind the rheinanthrone, we chose Calmodulin-3 (PDBID: 7CR4), Nucleobindin-1 (PDBID: 1SNL), and Actin, gamma-enteric smooth muscle (PDBID: 3W3D) as the reporters and then docked these proteins with the rheinanthrone. Here, we selected the optimum conformation (lowest binding free energy) of protein-rheinanthrone binding poses to analyze their interactions (Figs. 1H–J and Table S7). The rheinanthrone binds to the protein residues by potent hydrogen bond and hydrophobic interactions. The binding affinity of rheinanthrone and recombinant Calmodulin-3 homologous protein was analyzed by SPR assay. And, a dissociation constant (K_D) was determined to be 67.9 μM, indicating a strong interaction between the active intermediate and the target protein (Fig. 1K). These results further indicate that Calmodulin-3 may be the potential target of rheinanthrone.

In this study, we utilized an in situ microbial biotransformation strategy to convert the SAPs into RAPs by gut microbes. This strategy overcomes the challenge in synthesizing active drug-based probes based on the unstable active intermediates produced by prodrugs. Coupling with chemical proteomics, we successfully identified three potential target proteins of rheinanthrone, including Calmodulin-3, Nucleobindin-1, and Actin, gamma-enteric smooth muscle. SPR was applied to evaluate the interactions of rheinanthrone with the identified proteins. This work shows a major advance in identifying the mechanism-of-action targets of SA. Overall, this study provides a new idea to identify the direct targets of unstable bioactive intermediates converted by prodrugs.

CRediT author statement

Zhen Liu: Writing – original draft, Software, Investigation, Formal analysis. Xinyue Geng: Writing – original draft, Software, Investigation, Formal analysis. Xinyue Liu: Resources. Mengru Li: Investigation, Formal analysis. Xiang Li: Visualization. Zhixin Zhang: Validation. Gan Luo: Writing – review & editing. Ying Wang: Writing – review & editing. Xiaoyan Gao: Writing – review & editing, Project administration, Funding acquisition.

Declaration of competing interest

The authors declare that there are no conflicts of interest.

Appendix A. Supplementary data

The following is the Supplementary data to this article: