Synthetic Biology in Natural Product Biosynthesis

Abstract

Synthetic biology has played an important role in the renaissance of natural products research during the post-genomics era. The development and integration of new tools have transformed the workflow of natural product discovery and engineering, generating multidisciplinary interest in the field. In this review, we summarize recent developments in natural product biosynthesis from three different aspects. First, advances in bioinformatics, experimental, and analytical tools to identify natural products associated with predicted biosynthetic gene clusters (BGCs) will be covered. This will be followed by an extensive review on the heterologous expression of natural products in bacterial, fungal and plant organisms. The native host-independent paradigm to natural product identification, pathway characterization, and enzyme discovery is where synthetic biology has played the most prominent role. Lastly, strategies to engineer biosynthetic pathways for structural diversification and complexity generation will be discussed, including recent advances in assembly-line megasynthase engineering, precursor-directed structural modification, and combinatorial biosynthesis.

Figure 1.

Representative natural products from eight different classes.

Figure 2.

Simplified scheme of RiPP biosynthesis.

Figure 3.

Representative natural products that contain features from multiple classes (hybrids).

Figure 4.

Representative natural products that do not belong to the eight classes.

Figure 5.

(A) Organization and constituents of a typical microbial biosynthetic gene cluster. (B) Examples of the erythromycin and lovastatin BGCs are shown as examples of bacterial and fungal BGCs, respectively.

Figure 6.

Simplified workflow of synthetic biology approaches to study and engineer natural products.

Figure 7.

Distribution of more than 30 different natural product classes of BGCs from over 1000 Streptomyces strains enables strain-level comparison between number and classification of BGCs. Reprinted from Belknap, K. C.; Park, C. J.; Barth, B. M.; Andam, C. P. Genome Mining of Biosynthetic and Chemotherapeutic Gene Clusters in Streptomyces Bacteria. Sci Rep 2020, 10 (1), 2003. Copyright 2020 Springer Nature. This is an open access article distributed under the terms of the Creative Commons CC BY license https://creativecommons.org/licenses/by/4.0/.

Figure 8.

Examples of natural products from different RiPPs subfamilies.

Figure 9.

BGCs encode self-resistance enzymes (SREs) that are insensitive to the bioactive natural product. (A) Scheme of SRE colocalization and function; (B) Recent examples of natural products with SREs colocalized in the respective BGCs. The target/SRE functions are shown in blue.

Figure 10:

Examples of reactions catalyzed by enzymes that are initially annotated as hypothetical proteins or incorrect functions based on sequence alignments.

Figure 11.

BGCs of tripyrrole natural products contain a condensation enzyme such as PigC that condense MBC and MAP to afford prodigiosin. By using the condensation enzyme PigC to build an SSN, new BGCs that condense MBC to different monopyrroles were discovered.

Figure 12.

Chimeric LuxR is engineered as an activator in Pseudomonas to upregulate BGC expression.

Figure 13.

New natural products discovered from HiTES.

Figure 14.

Examples of fungal natural products discovered via overexpression of BGC-specific transcription factors.

Figure 15.

Examples of fungal natural products discovered from genetic manipulation of native hosts.

Figure 16.

Examples of natural products discovered using CRISPR-based tools.

Figure 17.

New endolides and beauvericin identified using MassQL-integrated approaches.

Figure 18.

HypoRiPPAtlas predicts hypothetical RiPPs structures from a query spectrum, enabling matches to be identified from entries in a spectral database.

Figure 19.

Cyclic peptides identified from an environmental Symploca extract using SMART 2.0.

Figure 20.

Selected examples of natural products that have been heterologously produced.

Figure 21.

Activity of phosphopantetheinyl transferase (PPTase) to convert carrier protein (CP) from apo- to holo- form. A serine residue is phosphopanthetheinylated with coenzyme A in the presence of PPTase.

Figure 22.

Examples of heterologous expression in E. coli. (A) 99 and 1 from the erythromycin pathway. (PDB: 7M7I); (B) taxadiene (90) and taxadien-5α-ol (100) which are precursors to paclitaxel (7); and (C) mycosporine-like amino acids such as shinorine (102).

Figure 23.

Examples of natural products heterologously produced in Streptomyces hosts.

Figure 24.

Examples of natural products heterologously produced in other bacterial hosts.

Figure 25.

Examples of natural products heterologously expressed in bacterial hosts using cosmid libraries.

Figure 26.

Examples of natural products produced in heterologous hosts using BACs.

Figure 27.

Examples of natural products produced in heterologous hosts after BGC capturing using CATCH.

Figure 28.

Examples of natural products produced in heterologous hosts after BGC capturing using CAPTURE. (A) Overview of CAPTURE workflow. (B) Compounds discussed in text.

Figure 29.

Examples of natural products produced in heterologous hosts after BGC capturing using TAR. (A) Overview of TAR workflow; (B) Compounds discussed in text.

Figure 30.

Examples of natural products heterologously expressed using recombineering-based strategies.

Figure 31.

Examples of polyketides produced in bacterial heterologous hosts.

Figure 32.

A diiodotetrayne 157 is a universal precursor to enediyne natural products such as 154 and 155.

Figure 33.

Increase polyketide precursor acetyl-CoA through controlled breakdown of triacylglycerols.

Figure 34.

Examples of nonribosomal peptides produced in bacterial heterologous hosts.

Figure 35.

Examples of RiPPs produced in E. coli that led to elucidation of key PTM enzymes.

Figure 36.

Biosynthesis of amino acid derived metabolites using PEARL enzymes.

Figure 37.

Phosphorylation-based two-step IPP/DMAPP biosynthetic pathway for isoprenoid production. (A) This approach starts with feeding of prenol and isoprenol alcohols; (B) Representative compounds that have been produced in bacteria using this approach.

Figure 38.

Spatial engineering to improve P450 functions in E. coli. (PDB: 7SUX (blue) and 3EJB (yellow)).

Figure 39.

Examples of natural products biosynthesized without known core enzymes. Heterologous expression played an important role in pathway elucidation. (A) Biosynthetic pathway of 19; (B) biosynthetic pathway of 20.

Figure 40.

Examples of natural products from BGCs without well-known core enzymes.

Figure 41.

Production of 94 in yeast followed by semisynthesis to 8. (A) biosynthetic pathway to 94 is reconstituted in yeast at high titer; (B) Sanofi’s semisynthetic preparation of 8 from 94.

Figure 42.

Examples of sesquiterpenes and diterpenes that have been produced in yeast.

Figure 43.

Examples of triterpene and sterol natural products that have been produced in yeast.

Figure 44.

Heterologous biosynthesis of THIQ alkaloids in yeast.

Figure 45.

Heterologous biosynthesis of tropane alkaloids in yeast.

Figure 46.

Production of MIAs in yeast. (A) Biosynthetic pathway to 50, precursor to MIAs; (B) Examples of MIAs produced in yeast.

Figure 47.

Heterologous biosynthesis of ergot alkaloids in yeast.

Figure 48.

Representative polyketides and NRPs produced in engineered yeast.

Figure 49.

Heterologous biosynthesis of cannabinoids in yeast.

Figure 50.

Examples of flavonoids produced heterologously in yeast. (A) Early stages of flavonoid biosynthesis from L-phenylalanine; (B) Selected examples.

Figure 51.

Examples of natural products produced in Y. lipolytica or P. pastoris.

Figure 52.

Endogenous secondary metabolites in A. nidulans.

Figure 53.

Three-plasmid system used to perform heterologous expression experiments in A. nidulans A1145.

Figure 54.

Examples of natural products heterologously produced in A. nidulans.

Figure 55.

Examples of natural products biosynthesized by multiple core enzymes.

Figure 56.

Revision of AoiQ function using A. nidulans for pathway reconstruction.

Figure 57.

Representative natural products not derived from well-known core enzymes in A. nidulans.

Figure 58.

A. nidulans as a host for reconstituting BGCs contain an SRE.

Figure 59.

BGCs from phytopathogenic fungi expression in engineered A. nidulans resulted in new natural products.

Figure 60.

Biosynthesis of 96 in engineered A. nidulans with an emphasis on compartmentalization.

Figure 61.

Fungal natural products of which biosynthesis involves a pericyclase-catalyzed reaction.

Figure 62.

Examples of pericyclic reactions in natural product biosynthesis that have been identified and characterized using heterologous expression in A. nidulans.

Figure 63.

PLP-dependent enzymes as scaffold building enzymes in fungal natural product biosynthesis. (A) PLP-dependent pipecolic and picolinic acid formation. (B) Cycloleucine biosynthesis investigated via heterologous expression in A. nidulans.

Figure 64.

Representative natural products heterologously biosynthesized in A. nidulans.

Figure 65.

Representative natural products produced in engineered A. oryzae.

Figure 66.

Representative natural products produced in engineered A. oryzae.

Figure 67.

Representative natural products biosynthesized from A. niger.

Figure 68.

Overview of Agrobacterium-mediated transient expression of heterologous genes in N. benthamiana. Created in BioRender. Dror, M. (2024) https://BioRender.com/c51d443

Figure 69.

Biosynthetic pathway of 11, identified using heterologous expression in N. benthamiana.

Figure 70.

Biosynthetic pathway of colchicine (91) reconstituted in N. benthamiana.

Figure 71.

Biosynthetic pathway of limonoids 336 and 337 reconstituted in N. benthamiana.

Figure 72.

Biosynthetic pathways of 341 and 191b characterized using heterologous expression in N. benthamiana.

Figure 73.

Early stages of the paclitaxel biosynthetic pathway as characterized by heterologous expression in N. benthamiana.

Figure 74.

Late-stage biosynthetic pathway of paclitaxel characterized by heterologous expression in N. benthamiana.

Figure 75.

Notable natural products produced heterologously in N. benthamiana.

Figure 76.

Evolutionary module boundaries are more robust in module swapping than traditional boundaries. (A) Comparison of traditional boundaries and new, evolutionary boundaries used to define modules for venemycin PKS. (B) Hybrid Pik and Vem PKSs constructed using both traditional and new boundaries result in synthesis of 351. However, the hybrid PKS from new boundaries is far more robust when the turnover rates are measured.

Figure 77.

A four-module PKS built using evolutionary boundaries led to efficient synthesis of tetraketide 352.

Figure 78.

The engineering strategy employed by Piel and coworkers enabled the generation of a library of 353a-b derivatives 353c-j.

Figure 79.

NRPS engineering using XU, XUC and XUT exchangeable units developed by the Bode lab. (A) Types of modular units used; (B) Application of the exchange units towards de novo synthesis of the proteasome inhibitor 34b inspired from 34.

Figure 80.

Earliest example of PDB involved feeding simple precursors to liquid fermentation cultures of P. chrysogenum. Diverse R-groups were incorporated in the penicillin products.

Figure 81.

PDB of 380 antimycin derivatives by exploring the substrate promiscuity of three pathway enzymes.

Figure 82.

Application of PDB to enhance the bioactivities of natural products.

Figure 83.

Examples of mutasynthesis. (A) Generation of prodigiosin (55) analogs; (B) generation of physostigmine (359a) analogs.

Figure 84.

Biotransformation of (A) 360 and (B) 361 using A. eureka as a host.

Figure 85.

Feeding of 362 to X. feejeensis enabled the generation of 363 with an unprecedented chemical scaffold.

Figure 86.

Examples of natural products that have undergone structural diversification via biotransformation.

Figure 87.

PDB of type I PKS using acyl-SNACs as substrate mimics. (A) Scheme of intercepting type I PKS with unnatural starter and extender units; (B) PDB to produce 1b.

Figure 88.

PDB with narbonolide PKS with advanced acyl thioesters.

Figure 89.

Unnatural polyketide extender unit can be generated via feeding of unnatural carboxylic acids and esters to strains expressing promiscuous enzymes including (A) malonate-CoA synthetase; (B) crotonyl-CoA reductase/carboxylase; or (C) acyl-CoA carboxylase.

Figure 90.

Incorporation of alkyne-containing extender unit into polyketide structures using the JamABC cassette.

Figure 91.

Promiscuous AT can be explored to generate fluorinated macrolides.

Figure 92.

Using engineered trans-AT to alter extender unit specificity.

Figure 93.

Mutasynthesis of NRP through inactivation of pathways for nonproteingenic amino acid biosynthesis.

Figure 94.

Enduracidin A analogs (387b-e) were generated by swapping A domains in its NRPS assembly-line.

Figure 95.

Swapping of NRPS starter condensation domains (C_S) enabled generation of new lipopeptides.

Figure 96.

Promiscuity of C-terminal condensation domain allows the selective incorporation of putrescine into nonribosomal peptides.

Figure 97.

PCY1 is a promiscuous head-to-tail peptide macrocyclase.

Figure 98.

SyncM performs lanthipeptide cyclization with diverse substrates. The lines indicate lanthionine and methyllanthionines linkages.

Figure 99.

DarE is an oxidative crosslinking maturation enzyme. Residue substitutions tolerated by DarE are noted on the right.

Figure 100.

PlpX and PlpY excise tryptamine from precursor peptides. (A) Scheme of the reaction and (B) a precursor peptide variant was used to synthesize 394.

Figure 101.

Cyclization of FPP and analogs by the sesquiterpene cyclase Bot2.

Figure 102.

Methylated isoprenyl pyrophosphates as building blocks for unnatural terpene synthesis. (A) IPP methyltransferases facilitates synthesis of methylated zeaxanthin analogs; (B) SpSodMT and engineered mutants facilitate synthesis of C16 terpenoids.

Figure 103.

Cross-coupling of halogenated violacein derived from PDB using tryptophan analog.

Figure 104.

Biosynthesis of halogenated MIAs through feeding of halogenated tryptophan or tryptamine. Numbers on ring labeled represent sites of diversification.

Figure 105.

Swapping tryptophan halogenases led to the production of 411 analogs.

Figure 106.

Using different l-tryptophan halogenases led to the production of 412b-d.

Figure 107.

Biocatalytic access to MIA analogs via unnatural supply of haloindoles as a precursor to halotryptophan.

Figure 108.

The use of fluorinated SAM analogs for fluoromethylation of natural products. Biocatalytic production of (A) 413 and (B) 416. (C) Fluoromethylation of natural products.

Figure 109.

Generation of 420 analogs using engineered SAM synthetase and RapM.

Figure 110.

Select natural products targeted for sugar diversification.

Figure 111.

Biocatalytic prduction of NDP-sugars for vancomycin analog generation.

Figure 112.

Modifying the sugar groups on polyketide scaffolds. (A) 426a as an example; (B) 428a as an example.

Figure 113.

Glycosyltransferase swap between BGCs of lenoremycin and endusamycin enabled generation of 432.

Figure 114.

Genearation of aminoglycoside derivatives through exchange of glycosyltransferases.

Figure 115.

Biosynthesis of the semisynthetic 439 using engineered acyltransferase and small molecule acyl thioester 440.

Figure 116.

Targeting acyl transfer steps in biosynthesis for product diversification.

Figure 117.

PDB was used to generate 444 that is the precursor to the magnetic-nanoparticle/ansamitocin conjugate 443.

Figure 118.

Early example of combinatorial biosynthesis by Hopwood and coworkers.

Figure 119.

Coexpression of ox4 with cbmA-E resulted in the generation of new hybrid PoTeMs 451 and 452.

Figure 120.

Access to new sesquiterpenoids using (A) the pericyclase from the BGC of 38 to modify the eupenifeldin core; or (B) tailoring enzymes from the BGC of 288 to modify 456.

Figure 121.

SptF is a promiscuous oxygenase that modifies meroterpenoids. (A) Modification of andilesin C (464) and (B) oxidation of other fungal meroterpenoids.

Figure 122.

The mixing and matching of two sesquiterpene cyclases and their associated P450s enabled the generation of a wide variety sesquiterpenoids.

Figure 123.

Combinatorial expression of oxidative enzymes that act on 484.

Figure 124.

Engineering of a P450-T2 mutant enabled the enzymatic “hybridization” of 485 and 160 to 486 via cyclopropane formation.