Algorithmic discovery and characterization of new biosynthetic reactions

Dommaraju, Shravan Rama

Permalink

https://hdl.handle.net/2142/129661 Copy

Description

Title

Algorithmic discovery and characterization of new biosynthetic reactions

Author(s)

Dommaraju, Shravan Rama

Issue Date

2025-02-04

Director of Research (if dissertation) or Advisor (if thesis)

Mitchell, Douglas A

Doctoral Committee Chair(s)

Mitchell, Douglas A

Committee Member(s)

• van der Donk, Wilfred A
• Mehta, Angad P
• Metcalf, William W

Department of Study

Chemistry

Discipline

Chemistry

Degree Granting Institution

University of Illinois Urbana-Champaign

Degree Name

Ph.D.

Degree Level

Dissertation

Keyword(s)

• Biochemistry
• chemical biology
• biosynthesis
• RiPPs
• natural products
• bioinformatics
• computational biology

Abstract

Natural product research lies at the interface of chemistry and biology. Investigation of natural product biosynthesis has the potential to both expand enzyme chemistry and develop understanding of intricate physiological and ecological relationships. One prominent group of natural products is the ribosomally synthesized and post-translationally modified peptides (RiPPs). Studies of RiPP biosynthesis have led to discoveries of new enzyme functions and new microbial physiology. To accelerate these studies, the RiPP field has increasingly relied on genome mining to generate new hypotheses for wet lab investigation. By developing new algorithmic methods to mine microbial genomes for RiPP biosynthetic gene clusters (BGCs), we hoped to discover new modes of biosynthesis and validate them in the lab. Chapter 1 of this dissertation describes the genome mining-led discovery of a new class of RiPPs containing two amino termini. Using RRE-Finder and RODEO for dataset generation, we developed a divisive hierarchical clustering method to sort RiPP Recognition Elements (RREs) into families using biosynthetic first principles. After evaluation of new putative RRE families, we selected a group that was unlike any previously reported RiPP class. Each BGC encoded an alcohol dehydrogenase, aminotransferase, methyltransferase, and an uncharacterized domain fused to an RRE. Using a direct cloning synthetic biology approach, we over-expressed and purified three new RiPPs, called daptides 1-3, containing a native N-terminus and an (S)-N2,N2-dimethyl-1,2-propanediamine (Dmp)-modified C-terminus. Bioactivity evaluation suggests that the daptides interact with membranes synergistically. In vivo co-expression data showed a putative biosynthetic pathway for production of the Dmp-modified terminus. Additional investigation of the first step in biosynthesis demonstrated that the uncharacterized domain-RRE fusion protein was required for the initial oxidative decarboxylation en route to Dmp. Altogether, this study provides a case study in class-independent RiPP genome mining for new biosynthetic enzyme families and led to a new biosynthetic class. One limitation of the initial study on the daptides was the inability to fully investigate their biosynthesis in vitro. Therefore, we sought to characterize additional daptides more thoroughly in terms of metabolic and substrate scope. Chapter 2 of this dissertation describes this investigation. Using a combination of enzymes from two BGCs, we were able to fully reconstitute daptide biosynthesis in a test tube, including determination of required cofactors for each step of biosynthesis. Additional work involved investigation of post-translational modifications (PTMs) installed by a divergent family of YcaO enzymes encoded within daptide BGCs. Through mass spectrometry and NMR experiments, we were able to show the YcaO enzymes capture an amine intermediate and divert Dmp biosynthesis to a 1-amino-1-[1,5-dimethyl-2-imidazolin-2-yl]isopentane (Adi)-modified terminus. Investigation of substrate scope showed high tolerance of the daptide enzymes to sequence variation. Various engineered substrates were also processed, suggesting that the daptide system may be useful for engineering peptides and proteins with unnatural termini. Chapter 3 of this dissertation describes efforts to develop MetaRODEO, an expanded genome mining tool for accessing low-to-medium quality genomic data and genomic data from uncultivated organisms. Large changes were made to the RODEO code to increase efficiency and generate RiPP datasets. Paired post-processing scripts were also developed to enhance data analysis. Using the generated data from a set of >50,000 unannotated NCBI Bioprojects, we began comparing RiPP BGCs from these uncultivated bacteria to cultivated bacteria from NCBI’s Refseq database. Additionally, we validated the biochemistry of three new BGCs in the lab. Further work is currently ongoing to develop and expand these bioinformatic methods. With more comprehensive datasets, we hope to study new modes of biosynthesis and expand knowledge of both enzyme chemistry and microbial physiology. Appendix chapters A-D contain PDF reprints of additional co-authored publications. These works contain contributions to both genome mining methods and structural characterization of RiPPs containing new modifications.

Graduation Semester

2025-05

Type of Resource

Thesis

Handle URL

https://hdl.handle.net/2142/129661

Copyright and License Information

Copyright 2025 Shravan Rama Dommaraju

Owning Collections