Metagenomics combined with activity-based proteomics point to gut bacterial enzymes that reactivate mycophenolate

Abstract

Mycophenolate mofetil (MMF) is an important immunosuppressant prodrug prescribed to prevent organ transplant rejection and to treat autoimmune diseases. MMF usage, however, is limited by severe gastrointestinal toxicity that is observed in approximately 45% of MMF recipients. The active form of the drug, mycophenolic acid (MPA), undergoes extensive enterohepatic recirculation by bacterial β-glucuronidase (GUS) enzymes, which reactivate MPA from mycophenolate glucuronide (MPAG) within the gastrointestinal tract. GUS enzymes demonstrate distinct substrate preferences based on their structural features, and gut microbial GUS enzymes that reactivate MPA have not been identified. Here, we compare the fecal microbiomes of transplant recipients receiving MMF to healthy individuals using shotgun metagenomic sequencing. We find that neither microbial composition nor the presence of specific structural classes of GUS genes are sufficient to explain the differences in MPA reactivation measured between fecal samples from the two cohorts. We next employed a GUS-specific activity-based chemical probe and targeted metaproteomics to identify and quantify the GUS proteins present in the human fecal samples. The identification of specific GUS enzymes was improved by using the metagenomics data collected from the fecal samples. We found that the presence of GUS enzymes that bind the flavin mononucleotide (FMN) is significantly correlated with efficient MPA reactivation. Furthermore, structural analysis identified motifs unique to these FMN-binding GUS enzymes that provide molecular support for their ability to process this drug glucuronide. These results indicate that FMN-binding GUS enzymes may be responsible for reactivation of MPA and could be a driving force behind MPA-induced GI toxicity.

Diverse but desolate landscape of gut microbial azoreductases: A rationale for idiopathic IBD drug response

“…Processed IBDMDB metaproteomics data did not contain AzoR proteins when protein function was assigned using EC annotations 42 . To confirm the absence of AzoR proteins, raw IBDMDB metaproteomics files (directly retrieved from https://ibdmdb.org/tunnel/public/summary.html ) were re-analyzed using all AzoR protein sequences identified using structural metagenomics as the reference database using proteomics data analysis methods as described previously 55 , 63 .…”

The Effect of the Gut Microbiota on Transplanted Kidney Function

“…Simpson and colleagues examined faecal samples of patients after kidney transplantation and identified differences in bacterial β-glucuronidase function compared to healthy individuals. They were able to find a higher abundance of a flavin mononucleotide-binding ortholog of β-glucuronidase, causing reactivation of mycophenolic acid [ 39 ]. Moreover, Khan et al revealed that there is a difference in GUS (β-glucuronidase) activity in stool between hematopoietic cell and kidney transplantation patients, possibly affecting mycophenolate mofetil dosing requirements.…”

Activity‐Based Protein Profiling – Enabling Phenotyping of Host‐Associated and Environmental Microbiomes

“…This is principally due to enterohepatic recirculation of the glucuronidated form of the metabolized drug by bacterial β‐glucuronidases. Using a β‐glucuronidase specific probe, the Redinbo group identified and quantified a specific β‐glucuronidase enzyme that binds to flavin mononucleotide resulting in the reactivation of the active form of the drug, mycophenolic acid [37] . In a related approach described by Whidbey et al., ABPP was used to selectively isolate and characterize active subpopulations of microbes displaying β‐glucuronidase activity by coupling their ABP approach with fluorescence‐activated cell sorting (FACS) [38] .…”

“…Using a β-glucuronidase specific probe, the Redinbo group identified and quantified a specific β-glucuronidase enzyme that binds to flavin mononucleotide resulting in the reactivation of the active form of the drug, mycophenolic acid. [37] In a related approach described by Whidbey et al, ABPP was used to selectively isolate and characterize active subpopulations of microbes displaying β-glucuronidase activity by coupling their ABP approach with fluorescence-activated cell sorting (FACS). [38] Coupling of ABPP to FACS provides a powerful method to functionally deconvolute microbiomes.…”