Development of platensimycin, platencin, and platensilin overproducers by biosynthetic pathway engineering and fermentation medium optimization

Abstract   The platensimycin (PTM), platencin (PTN), and platensilin (PTL) family of natural products continues to inspire the discovery of new chemistry, enzymology, and medicine. Engineered production of this emerging family of natural products, however, remains laborious due to the lack of practical systems to manipulate their biosynthesis in the native-producing Streptomyces platensis species. Here we report solving this technology gap by implementing a CRISPR-Cas9 system in S. platensis CB00739 to develop an expedient method to manipulate the PTM, PTN, and PTL biosynthetic machinery in vivo. We showcase the utility of this technology by constructing designer recombinant strains S. platensis SB12051, SB12052, and SB12053, which, upon fermentation in the optimized PTM-MS medium, produced PTM, PTN, and PTL with the highest titers at 836 mg L−1, 791 mg L−1, and 40 mg L−1, respectively. Comparative analysis of these resultant recombinant strains also revealed distinct chemistries, catalyzed by PtmT1 and PtmT3, two diterpene synthases that nature has evolved for PTM, PTN, and PTL biosynthesis. The ΔptmR1/ΔptmT1/ΔptmT3 triple mutant strain S. platensis SB12054 could be envisaged as a platform strain to engineer diterpenoid biosynthesis by introducing varying ent-copalyl diphosphate-acting diterpene synthases, taking advantage of its clean metabolite background, ability to support diterpene biosynthesis in high titers, and the promiscuous tailoring biosynthetic machinery. One-Sentence Summary Implementation of a CRISPR-Cas9 system in Streptomyces platensis CB00739 enabled the construction of a suite of designer recombinant strains for the overproduction of platensimycin, platencin, and platensilin, discovery of new diterpene synthase chemistries, and development of platform strains for future diterpenoid biosynthesis engineering.


Introduction
Platensimycin (PTM), platencin (PTN), and platensilin (PTL) are members of an emerging family of bacterial natural products that have been intensively pursued as promising antibacterial and antidiabetic drug leads (Fig. 1 A) (Rudolf et al., 2017 ;Zheng et al., 2021 ;Wang et al., 2006Wang et al., , 2007 ) ). PTM was first isolated from Streptomyces platensis MA7327 (Singh et al., 2006 ;Wang et al., 2006 ), and PTN was isolated from S. platensis MA7339 (Wang et al., 2007 ).Subsequent discovery of PTN production in S. platensis MA7327 (Herath et al., 2008 ) and comparative analysis of the ptm biosynthetic gene cluster (BGC) from S. platensis MA7327 and the ptn Fig. 1.Platensimycin (PTM), platencin (PTN), and platensilin (PTL) structures and biosynthesis.(A) The structures of PTM, PTN, and PTL, as well as the key biosynthetic intermediate 3-amino-2,4-dihydroxybenzoic acid (ADHBA), and their nascent thioacid congeners.(B) Genetic organization of the ptm gene cluster from S. platensis MA7327 and S. platensis CB00739 encoding the production of PTM, PTN, and PTN, and the ptn gene cluster from S. platensis MA7339 encoding the production of PTN only, with the PTM cassette shaded in gray to highlight the difference between the two gene clusters.(C) PTM, PTN, and PTL biosynthesis featuring two diterpene synthases, PtmT1 and PtmT3, that channel the common precursor ent -copalyl pyrophosphate ( ent -CPP) into three distinct scaffolds ent -kauranol, ent -beyerene, and ent -atiserene, and a promiscuous tailoring biosynthetic machinery that acts on all three scaffolds to account for structural diversity of this family of natural products.BGC from S. platensis MA7339 revealed S. platensis MA7327 as a PTM and PTN dual producer (Fig. 1 B) (Smanski et al., 2009(Smanski et al., , 2011 ) ). Motivated by the search for alternative producers with enhanced genetic amenability to expedite in vivo manipulation of PTM and PTN biosynthesis, we identified six alternative PTM and PTN dual producers by mining the Actinobacteria strain collection of the Natural Products Discovery Center at the Wertheim UF Scripps Institute (Hindra et al., 2014 ). S. platensis CB00739, one of the six alternative producers, was subsequently developed into a platform strain for PTM and PTN biosynthesis.S. platensis SB12029, a recombinant strain of S. platensis CB00739, massively overproducing PTM and PTN (Hindra et al., 2014 ;Shi et al., 2016 ), enabled (1) the discovery of the minor product PTL (Zheng et al., 2021 ) and (2) the revelation of thioplatensimycin (thioPTM), thioplatencin (thioPTN), and thioplatensilin (thioPTL) as the nascent final products of the PTM, PTN, and PTL biosynthetic machinery (Fig. 1 A, C) (Dong et al., 2016(Dong et al., , 2018 ; ;Zheng et al., 2021 ).
Consisting of a varying diterpenoid-derived ketolide moiety linked via a flexible propionamide chain to 3-amino-2,4dihydroxybenzoic acid, the PTM, PTN, and PTL family of natural products continues to inspire the discovery of novel biosynthetic chemistry and enzymology (Fig. 1 A) (Rudolf et al., 2017 ;Zheng et al., 2021 ).Among the many characteristic features of the PTM, PTN, and PTL biosynthetic machinery are (1) two diterpene synthases, PtmT3 and PtmT1, that partition a common diterpene precursor ent -copalyl diphosphate ( ent -CPP) into three distinct diterpene scaffolds, for example, ent -kauranol, ent -atiserene, and ent -beyerene, and (2) a common tailoring biosynthetic machinery, consisting of minimally 14 enzymes, that acts on all three scaffolds parallelly to afford thioPTM, thioPTN, and thioPTL as the nascent products, which undergo nonenzymatic hydrolysis into PTM, PTM, and PTL during isolation (Fig. 1 B, C).The tailoring biosynthetic machinery displays remarkable substrate promiscuity, a biosynthetic logic that serves as an inspiration to access natural product diversity and engineer designer analogues by combinatorial biosynthesis and synthetic biology (Smanski et al., 2016 ;Teijaro et al., 2019 ).
PTM, PTN, and PTL biosynthesis and production have benefited greatly from a suite of engineered overproducers (Table 1 ).The first generation PTM and PTN overproducers were engineered in S. platensis MA7327 or S. platensis MA7339 by replacing the pathwayspecific negative transcriptional regulator ptmR1 or ptnR1 with the apramycin resistance gene aac(3)IV , as exemplified by S. platensis SB12001 and SB12002, or S. platensis SB12600, respectively (Smanski et al., 2009 ;Yu et al., 2010 ).These ptmR1 or ptnR1 recombinant strains overproduce PTM and PTN, but they sporulate poorly and are recalcitrant to further in vivo manipulation.The secondgeneration overproducers were engineered in S. platensis CB00739 similarly by replacing ptmR1 with aac(3)IV but followed by removing the aac(3)IV cassette via a λ-RED and FLP/FRT-mediated site-specific recombination, as exemplified by S. platensis SB12029 (Gust et al., 2003(Gust et al., , 2004 ; ;Rudolf et al., 2015 ).While S. platensis SB12029 overproduces PTM and PTN, the removal of aac(3)IV via the λ-RED and FLP/FRT-mediated method required weeks of tedious screening, making the generation of multiple mutations impractically laborious; it also leaves an 81-bp scar within each of the targeted genes, the effect of which on functional expression of the engineered BGCs is uncertain.As a result, no markerless double mutant has been constructed in S. platensis CB00739 to date.Expression of the ptn BGC from S. platensis MA7339 in selected model Streptomyces hosts has also been explored to circumvent the challenges in manipulating PTM and PTN biosynthesis in their native producers, as exemplified by S. lividans SB12606 that harbors the ptn BGC with the ptnR1 mutation in Streptomyces lividans K4-114.However, S. lividans SB12606 produces PTN at very low titers (Smanski et al., 2012(Smanski et al., , 2016 ) ).
Systems based on clustered regularly interspaced short palindromic repeats (CRISPR) and CRISPR-associated (Cas) proteins have been recently introduced for genetic manipulation in selected Streptomyces species that allow for fast and precise genome editing (Cobb et al., 2015 ;Huang et al., 2015 ;Liu et al., 2020 ;Tong et al., 2015Tong et al., , 2019 ; ;Zeng et al., 2015 ).Implementation of a CRISPR/Cas9 system in S. platensis CB00739 would overcome the limitations of the λ-RED and FLP/FRT-mediated methods and facilitate efficient and precise in vivo manipulation of the PTM, PTN, and PTL biosynthetic machinery in S. platensis .Herein, we report the CRISPR/Cas9-mediated engineering of S. platensis CB00739 to generate recombinant strains that overproduce PTM ( S. platensis SB12053), PTN ( S. platensis SB12052), or all three metabolites of PTM, PTN, and PTL ( S. platensis SB12051).Medium and fermentation optimization further highlight the utility of these engineered strains for PTM, PTN, and PTL production.The power of this CRISPR/Cas9-based genetic engineering strategy is finally showcased by the construction of the first triple mutant strain, S. platensis SB12054, setting the stage to exploit the PTM, PTN, and PTL biosynthetic machinery to engineer designer analogues for this emerging family of natural products.

Bacterial Strains, Plasmids, and Chemicals
Lists of oligonucleotides, plasmids, and bacterial strains used in this study are provided ( Supplementary Tables S1, S2).Oligonucleotides were purchased from Sigma Aldrich (Boston, MA, USA) and Integrated DNA Technologies (Coralville, IA, USA).Kits from Omega Bio-Tek (Norcross, GA, USA) were used for gel extraction of DNA and plasmid preparation.Restriction endonucleases, Q5 high-fidelity DNA polymerase, Gibson Assembly Master Mix, and T4 DNA ligase were all purchased from New England Biolabs (Ipswich, MA, USA) and used according to the manufacturer's instructions.The DIG High Prime DNA Labeling and Detection Starter Kit I from Roche (Basel, Switzerland) was used for Southern blot analyses.Other chemicals and components for culture media were purchased from standard commercial sources and used as is.DNA Sanger sequencing was performed by Genewiz (South Plainfield, NJ, USA).

General Experimental Procedures
Escherichia coli strains were cultured in lysogeny broth with appropriate antibiotics (Sambrook & Russel, 2001 ).The wild-type and recombinant strains of S. platensis CB00739 were grown at 28°C on solid ISP2 and ISP4 medium for sporulation (Shirling & Gottlieb, 1966 ) or liquid tryptic soy broth for growth of mycelium, seed cultures, and isolation of genomic DNA (gDNA) with appropriate antibiotics.High-performance liquid chromatography-mass spectrometry (HPLC-MS) was performed using an Agilent (Santa Clara, CA, USA) 1260 Infinity LC coupled to a 6230 TOF (high-resolution electrospray ioinization) equipped with an Agilent Poroshell 120 EC-C18 column (4.6 × 50 mm, 2.7 μm).

Genetic Engineering of S. Platensis CB00739 and Mutant Variants Using CRISPR/Cas9
Identification of potential 20-nt single guide RNA (sgRNA) target sequences in the S. platensis CB00739 chromosome was carried out using CasOT (Xiao et al., 2014 ).Candidate sequences with a predicted low probability of off-target effects were selected for construction of the pCRISPomyces-2-based editing plasmids, which followed the standard protocol (Cobb et al., 2015 ).In brief, for each target gene, a set of paired oligonucleotides representing the selected sgRNA sequence were annealed and inserted into pCRISPomyces-2 via Golden Gate assembly.Simultaneously, two ∼2-kb homology arms, one from each end of the targeted gene, were amplified via polymerase chain reaction (PCR) from the gDNA of S. platensis CB00739.The sgRNA-containing plasmids were then digested by Xba I and assembled with the two homology arms in a three-piece Gibson assembly to afford the final disruption plasmids pBS12129 ( ptmR1 ), pBS12130 ( ptmT3 ), and pBS12131 ( ptmT1 ).
The disruption plasmids were introduced into S. platensis CB00739 and its mutant variants via intergeneric conjugation from the methylation-deficient E. coli strain ET12567/pUZ8002 (Hopwood et al., 2000 ).Exconjugants were restruck onto ISP2 plates supplemented with 30 μg mL −1 apramycin and grown at 28°C for 3 days.Single colonies were grown in TSB medium for gDNA isolation.The target region was PCR-amplified with primers binding outside of the target region, and the products were analyzed via gel electrophoresis.Colonies that showed a pattern reflecting the desired mutation were selected for curing of the editing plasmid.The colonies were first subjected to two rounds of growth on nonselective ISP2 at 37°C before being replica-plated on both selective and nonselective ISP2 plates.Colonies showing the desired sensitivity to apramycin were grown in TSB to isolate gDNA.The genotypes of the recombinant strains were confirmed via diagnostic PCR, Southern blot analysis, and Sanger sequencing ( Supplementary Figs.S1-S3).While the ptmR1 strain of S. platensis SB12051 was constructed by deleting ptmR1 in S. platensis CB00739, the double mutants of S. platensis SB12052 ( ptmR1 and ptmT3 ) and S. platensis SB12053 ( ptmR1 and ptmT1 ) were constructed by deleting ptmT3 and ptmT1 in S. platensis SB12051,  respectively.The triple mutant in S. platensis SB12054 ( ptmR1 , ptmT3 , and ptmT1 ) was constructed by deleting ptmT1 from S. platensis SB12052.

Fermentation and HPLC-MS Analysis of S. Platensis SB12051, SB12052, SB12053, and SB12054
Fermentation of S. platensis SB12051, SB12052, SB12053, and SB12054, with S. platensis CB00739 as a control, followed previous protocols (Rudolf et al., 2015 ;Smanski et al., 2009 ;Zheng et al., 2021 ).Briefly, spores of the S. platensis strains of interest were inoculated into TSB medium and cultured at 28°C and 250 rpm for 2 days to yield a seed culture.The original production medium, known as PTM medium and renamed PTMM here for clarity, consisted of 40 g L −1 dextrin, 40 g L −1 α-lactose, and 5 g L −1 yeast extract, pH 7.0 (Smanski et al., 2009 ).In 250 mL baffled flasks, 50 mL of PTMM was charged with 3% (w/v) Amberlite XAD-16 resin (Sigma Aldrich, Boston, MA, USA), autoclaved, and inoculated with 4% (v/v) seed culture.The fermentation cultures were incubated at 28°C and 250 rpm for 7 days.Resin was separated from mycelium by diluting the culture with water, allowing the resin to settle, and decanting cell debris.This was repeated until no visible debris remained.The resin was extracted with methanol (3 × 10 mL).An aliquot of the extract was diluted 1:10 in methanol and centrifuged prior to analysis.Liquid chromatography for HPLC-MS analysis was performed using an 18-min solvent gradient from 5% to 95% methanol in water containing 0.1% formic acid at a flow rate of 0.4 mL min −1 .The peak area at 254 nm was used to quantify PTM, PTL, and PTN titers using absorbance coefficients obtained from standard calibration curves generated using pure compounds.Reported titers represent mean values from at least three independent biological replicates.

Engineering of PTM, PTN, PTL Biosynthesis by Applying CRISPR-Cas9 System to S. Platensis CB00739
The feasibility of applying the CRISPR/Cas9 system to engineer the PTM, PTN, and PTL biosynthetic machinery was demonstrated in S. platensis CB00739 by constructing the scarless ptmR1 mutant strain S. platensis SB12051.Introduction of the pCRISPomyces-2-based ptmR1 -targeting plasmid pBS12129 and subsequent curing of the plasmid yielded the scarless ptmR1 mutant strain S. platensis SB12051 ( Supplementary Fig. S1).This strategy avoided the laborious process of inserting and screening for removal of the acc(3)IV resistance cassette via the λ-RED and FLP/FRT-mediated methods used previously to construct the ptmR1 mutant strain S. platensis SB12029 (Rudolf et al., 2015 ).The growth phenotype of S. platensis SB12051 matched that of S. platensis SB12029, and most importantly, S. platensis SB12051 displayed dense sporulation, implying that it avoids the drawback of the overproducers engineered previously, whose poor sporulation phenotype has hampered further genetic manipulation (Fig. 2 A) (Hindra et al., 2014 ;Smanski et al., 2009 ). S. platensis SB12051 was fermented under established fermentation conditions with S. platensis SB12029 as a control, and crude extracts of the fermentation cultures were subjected to HPLC-MS analysis to compare metabolite profiles (Hindra et al., 2014 ;Smanski et al., 2009 ). S. platensis SB12051 overproduced PTM (168 mg L −1 ), PTN (41 mg L −1 ), and PTL (20 mg L −1 ), comparable to those of S. platensis SB12029 (Fig. 2 B, C, Table 1 ).Taken together, these findings confirm the compatibility of the CRISPR/Cas9 system with S. platensis CB00739 and set the stage to further manipulate PTM, PTN, and PTL biosynthesis by iterative application of the CRISPR/Cas9 system to the engineered overproducer S. platensis SB12051.

Engineering of a PTN Overproducer by Deleting ptmT3 in S. Platensis SB12051
As the two diterpene synthases PtmT3 and PtmT1 partition the common precursor ent -CPP into the three PTM, PTN, and PTL scaffolds in their biosynthesis (Smanski et al., 2011 ;Zheng et al., 2021 ), we first applied the CRISPR/Cas9 system to delete ptmT3 in S. platensis SB12051, affording the ptmR1 / ptmT3 double mutant strain S. platensis SB12052 ( Supplementary Fig. S2).S. platensis SB12052 exhibited similar growth characteristics as S. platensis SB12051, including dense sporulation (Fig. 2 A). S. platensis SB12052 was similarly fermented under the established conditions, with S. platensis SB12051 as a control, and crude extracts of the fermentation cultures were analyzed by HPLC-MS for changes in the metabolite profile (Hindra et al., 2014 ;Smanski et al., 2009 ).Production of PTM and PTL was completely abolished in S. platensis SB12052, with a concomitant increase ( > 6-fold) in PTN titer (271 mg L −1 ) (Fig. 2 B, C, Table 1 ).The fact that S. platensis SB12052 produces PTN with titers that are > 12-fold higher than S. platensis SB12600 engineered from the PTN producer S. platensis MA7339 (Yu et al., 2010 ) further supports S. platensis SB12051 as a superior strain for engineered production of the PTM, PTN, and PTL family of natural products (Table 1 ).Consistent with the role of PtmT1 as a dedicated ent -atiserene synthase (Smanski et al., 2011 ), together with the finding that the PTN titer in S. platensis SB12052 is comparable to the overall titers of PTM, PTN, and PTL in S. platensis SB12051, these findings also reveal that the metabolic flux of the PTM, PTN, and PTL biosynthetic machinery could be exploited to engineer designer analogues by manipulating the gatekeeping diterpene synthases as exemplified by PtmT1 and PtmT3, taking advantage of the inherent substrate promiscuity of the tailoring biosynthetic machinery (Fig. 1 C) (Dong et al., 2019 ;Wang et al., 2018 ;Zheng et al., 2021 ).

Engineering of a PTM Overproducer by Deleting ptmT1 in S. Platensis SB12051
Complementary to the ptmR1 / ptmT3 double mutant strain S. platensis SB12052 that overproduced PTN, we next applied the Fig. 2. Engineered S. platensis recombinant strains that overproduce PTM, PTN, and PTN.(A) Morphology of the S. platensis CB00739 wild-type strain in comparison with the CRISPR-Cas9-enabled recombinant strains S. platensis SB12051, SB12052, SB12053, and SB12054 and λ-RED and FLP/FRT-mediated recombinant strain S. platensis SB12029, all of which sporulated well on several media as exemplified with ISP4.Metabolite profiles, following fermentation in PTMM medium, of S. platensis SB12051, SB12052, SB12053, and SB12054 in comparison with S. platensis CB00739 upon HPLC-MS analysis with (B) UV detection at 254 nm or (C) negative-mode electrospray ionization mass spectrometry and ion extraction at m / z = 424 (black solid line, the expected mass of PTN and PTL) and m / z = 440 (dashed red line, the expected mass of PTM).Also see Supplementary Fig. S4 for full-length chromatograms with authentic standards.PTM ( ♦), PTN ( •), and PTL ( ).CRISPR/Cas9 system to delete ptmT1 in S. platensis SB12051, affording the ptmR1 / ptmT1 double mutant strain S. platensis SB12053 ( Supplementary Fig. S3).Like S. platensis SB12051 and S. platensis SB12052, S. platensis SB12053 also grew and sporulated well (Fig. 2 A).Upon fermentation of S. platensis SB12053 under the established conditions, the crude extracts of the fermentation cultures were analyzed by HPLC-MS for changes in the metabolite profile (Hindra et al., 2014 ;Smanski et al., 2009 ).As expected, S. platensis SB12053 overproduced PTM with a significantly increased titer (286 mg L −1 ), but PTL titer (16 mg L −1 ) remained largely unchanged, and most surprisingly, it still produced PTN, albeit with a significantly reduced titer (14 mg L −1 ) (Fig. 2 B, C, Table 1 ).The latter observation revealed that PtmT3 is capable of producing ent -atiserene as a minor product, a well-known property of plant ent -kauranol synthases (Xu et al., 2007 ).Nature has apparently evolved two distinct chemistries, that is, PtmT1 and PtmT3, for PTN biosynthesis (Fig. 1 B, C).This subtlety of the PTM, PTN, and PTL biosynthetic machinery escaped detection in the ptmT3 mutant strain S. platensis SB12008 made previously from the original S. platensis MA7327 due to low overall titers (Smanski et al., 2011 ), underscoring the advantage of the recombinant strains enabled by the CRISPR/Cas9-based system in the current study for future investigation of PTM, PTN, and PTL biosynthesis and engineered production.We lastly confirmed that PtmT1 and PtmT3 control the channeling of ent -CPP into PTM, PTN, and PTL biosynthesis in S. platensis by applying the CRISPR/Cas9 system to delete both ptmT1 and ptmT3 from S. platensis SB12052, affording the ptmR1 / ptmT1/ ptmT3 triple mutant strain S. platensis SB12054 (Fig. 1 B, C, and Supplementary Fig. S3).S. platensis SB12054 exhibited the same growth and sporulation characteristics as the other CRISPR/Cas9-enabled recombinant strains (Fig. 2 A).Fermentation of S. platensis SB12054 under the established conditions, followed by HPLC-MS analysis of the crude extracts of the fermentation cultures (Hindra et al., 2014 ;Smanski et al., 2009 ), demonstrated complete abolishment of PTM, PTN, and PTL production (Fig. 2 B, C).Notably, S. platensis SB12054 is the first triply mutated recombinant S. platensis strain engineered to date, highlighting the power of the CRISPR/Cas9 system for genetic manipulation of the PTM, PTN, and PTL biosynthetic machinery.Complementary to the bottom-up strategy of accessing diterpenoid structural diversity by introducing varying diterpene synthases into a suite of engineered E. coli chassis strains (Cyr et al., 2007 ), S. platensis SB12054 could be similarly exploited as a platform strain by introducing varying diterpene synthases to engineer diterpenoid biosynthesis (Fig. 1 B, C).

Medium Optimization of S. Platensis SB12051, SB12052, and SB12053 for PTM, PTN, and PTL Production
We finally further improved PTM, PTN, and PTL production titers by subjecting S. platensis SB12051, SB12052, and SB12053 to fermentation medium optimization (Fig. 3 , Table 1 ).Most of the published studies on PTM, PTN, and PTL biosynthesis and production to date used the production medium known as PTM medium (herein renamed as PTMM), which was developed for the S. platensis MA7327 and MA7339 wild-type, as well as their ptmR1 or ptnR1 mutant strains (Smanski et al., 2009 ;Yu et al., 2010 ).An optimized production medium, known as "PTM-SS" medium (Shi et al., 2016 ), was also developed by systematically varying the carbon and nitrogen sources, as well as the inorganic salts, for S. platensis SB12026 to produce PTM on a pilot scale, affording an impressive PTM titer of 1,560 mg L −1 in 15-L fermenters (Shi et al., 2016 ).We therefore first subjected S. platensis SB12051, SB12052, and SB12053, together with S. platensis SB12029 as a control, to fermentation in the "PTM-SS" medium.Significant increases in PTM, PTN, and PTL titers, varying between 1.5 and 2fold, were observed across all the strains tested (Fig. 3 , Table 1 ).Encouraged by these findings, we next focused on fine-tuning the carbon and nitrogen sources for their effect on production by S. platensis SB12052 and SB12053 while keeping the inorganic salts and pH of the "PTM-SS" medium unchanged.Among the five carbon and six nitrogen sources examined, the highest titers for both strains were enabled by the combination of α-maltose and soybean flour as the preferred carbon and nitrogen sources ( Supplementary Tables S3, S4).We named this optimized production medium as "PTM-MS" medium and finally demonstrated further improved production of PTM, PTN, and PTL in shake flasks using the "PTM-MS" medium.S. platensis SB12052 produced PTN exclusively with titers reaching 791 mg L −1 , while S. platensis SB12053 produced PTM as the predominant product, with titers reaching 836 mg L −1 , and PTN and PTL as minor products, with titers reaching 40 mg L −1 each, respectively (Fig. 3 , Table 1 ).These findings highlight the power of combining contemporary metabolic pathway engineering strategies with traditional medium and fermentation optimization methods for strain development and titer improvement, setting the stage for future production of the PTM, PTN, and PTL family of natural products, as well as their engineered analogues, at pilot scales for further development as promising antibacterial and antidiabetic drug leads.

Conclusions
Biosynthesis and engineering of the PTM, PTN, and PTL family of natural products have been hampered by the lack of practical systems to manipulate their biosynthesis in the native-producing S. platensis species (Rudolf et al., 2017 ;Zheng et al., 2021 ).Expression of the BGC encoding PTN biosynthesis in model Streptomyces hosts has fallen short of overcoming the challenge, producing PTN only at a very low titer ( < 2.0 mg L −1 ) (Smanski et al., 2012 ).We have now solved this technology gap by implementing the CRISPR-Cas9 system in S. platensis CB00739 and developed an expedient method to manipulate the PTM, PTN, and PTL biosynthetic machinery in vivo (Fig. 2 ).We showcased the utility of this technology by constructing designer recombinant strains S. platensis SB12051, SB12052, and SB12053 which, upon fermentation in the optimized PTM-MS medium, produced PTM, PTN, and PTL with the highest titers at 836 mg L −1 , 791 mg L −1 , and 40 mg L −1 , respectively (Table 1 ).Comparative analysis of PTM, PTN, and PTL biosynthesis between S. platensis SB12052 ( ptmR1 / ptmT3 ), SB12053 ( ptmR1 / ptmT1 ), and SB12054 ( ptmR1 / ptmT1 / ptmT3) also revealed two distinct chemistries, catalyzed by the two diterpene synthases PtmT1 and PtmT3, that nature has evolved for PTM, PTN, and PTL biosynthesis (Fig. 1 B, C).Further, S. platensis SB12054 could be exploited as a platform strain by introducing varying ent -CPP-acting diterpene synthases to engineer diterpenoid biosynthesis, taking advantage of its clean metabolite background, ability to support diterpene biosynthesis in high titers, and the promiscuous tailoring biosynthetic machinery that can transform the nascent diterpene scaffolds to further enrich structural diversity.
not discovered until 2021(Zheng et al., 2021 ), and PTL titers therefore were not determined in the earlier studies.ND, not determined.b Boldface highlights the engineered strains that afford the highest PTM, PTN, and PTL titers when fermented under the optimized PTM-MS medium.

Fig. 3 .
Fig.3.PTM, PTN, and PTL titers from fermentation of S. platensis SB12051, SB12052, and SB12053 in three media highlighting PTM-MS as the optimized medium for production.Bar heights reflect the mean titers of at least three biological replicates.See Table1for standard deviation.

Table 1 .
Development of PTM, PTN, and PTL Overproducers by Biosynthetic Pathway Engineering and Fermentation Medium Optimization of Streptomyces platensis