Team:SCUT-China/Project/Synthetic TKL

Overview

In this project, we propose that DNA coding for polyketide synthase domains can be standardized and reconstructed to generate chimeric polyketide synthases that produce new-to-nature polyketides, through which we can dramatically expand the library of antibiotic candidates. In order to demonstrate the feasibility of our proposal, we conducted “proof of concept” experiments to synthesize a simple polyketide— triketide lactone 1B—using a truncated erythromycin-producing polyketide synthase assembled with biobricks constructed by standardizing essential domains required for the synthesis of the triketide lactone. The domains we standardized are the loading domain, domains of DEBS1 and the thioesterase (TE) domain from DEBS3 (Fig. 1). The loading domain is required for the direct loading of propionyl-CoA to form a starter unit for triketide lactone. DEBS1 is a large multifunctional polypeptide composed of two modules that incorporate (2S)-methylmalonyl-CoA to form extender units. The TE domain catalyzes the intramolecular cyclization of the triketide to form the target triketide lactone. [1]The target product is expected to be detected and identified from the culture broth.

Design

Candidate selection

Biosynthesis of polyketides like erythromycin, tacrolimus and many others are guided by modular PKS genes. These genes encode large enzymes consisting of modules of domains, forming an "assembly line" to extend the ketide units to form polyketides. Among the known PKS genes, the genes coding for erythromycin producing polyketide synthethase have been extensively and deeply studied, and there are numerous information sources. Therefore, we chose a truncated erythromycin-producing polyketide synthase which has been proved to produce a simple triketide lactone 1B.[2]

Standardization of domains

Figuring Out the Boundary

With the idea of engineering PKS genes to be interchangeable domains, we firstly develop certain design principles for the engineering of PKS genes. Although many PKSs are highly modular proteins consisting of modules of domains, figuring out the boundary of each domain should be carefully considered because slight changes may result in lost of function.

To achieve this goal, we searched for more information from two database ASMPKS (Analysis System for Modular Polyketide Synthases:http://gate.smallsoft.co.kr:8008/~hstae/asmpks/pks_prediction.pl ) and MAPSI (Management and Analysis for Polyketide Synthase type I: http://gate.smallsoft.co.kr:8008/pks/ ). According to the resource provided by these websites, we separated the sequences of DEBS1 with centering on domains.

Standardization of Interchangeable Domains

Once the boundaries of the domains have been defined, standard prefix or suffix should be assigned to each end of the domains to make them interchangeable. The standardization of sufficient numbers of domains from numerous organisms would allow customized replacement, rearrangement, and/or recombining of different domains to produce novel polyketides. In this project, we tested customized synthesis of triketide lactones by replacing the loading domain and swapping KR+ACP domains between modules to test the compatibility of the standardized domains.

The “proof-of-concept” product--triketide lactone 1B

Assemble standardized domains

To validate our proposal we attempted to express the varient -- DEBS1+TE in E.coli. [3] Initially, we tried to amplify the coding sequences for each individual domain from Saccharopolyspora erythraea genomic DNA by PCR using primers consisting of RFC 23. However, after laborious work of weeks, we had failed to obtain proper PCR products due to the high G+C content (≥70%) of the interest DNA sequences. Afterward, to achieve our goal in a more efficient way, we contracted GENEWIZ, an international genomics service company, to synthesize the genes. With the coding genes under the control of T7 promoter and double T7 terminator, we successfully assembled three constructs for our design.

We were able to assemble all of our constructs and continued our work with our synthetic DEBS1+TE, DEBS1’+TE (KR of module2 in DEBS1 was replaced with KR of module1) and RFP+DEBS1+TE (DEBS1 was fused with red fluorescent protein) by transforming them into the E. coli strain BAP1. (Fig. 2)

If the right product triketide lactone can be generated, we can conclude that the standardized domains can work relatively independently and generate normal product.

Also, we changed the loading domains which selected different substrates resulting in product of different polyketides.

Furthermore, we can shuffle the domains as units between different DEBSs.

Induction of PKSs

12-48 hours after induction with IPTG, culture broth samples of the recombinant strain expressing the chimeric polyketide synthetases were collected and subjected for further analysis.

Results

1. DNA Electrophoresis

We used restriction enzyme digestion and/or PCR to validate the constructs (Fig. 3). The constructs carrying DEBS1+TE and DEBS1’+TE were isolated and digested with EcoRI (Fig. 3, lane 1 and 2) or double digested with EcoRI+PstI (data not shown). The construct carrying RFP +DEBS1+TE was validated with PCR (Fig. 3 lane 3) and EcoRI+PstI double digestion (data not shown). The results showed that all of the three constructs were successfully created.

2. SDS-PAGE

DEBS1 is a large protein of 370kDa. We performed SDS-PAGE to confirm the expression of the recombiant DEBS1+TE and DEBS1’+TE. However, none of DEBS1+TE, DEBS1’+TE (data not shown) and RFP+DEBS1+TE (Fig. 4) could be detected after Coomassie blue staining.

Figure 4 SDS-PAGE of the recombinant protein RFP+DEBS1+TE. M, protein size markers; RFP+DEBS1+TE(1-3), total proteins from RFP +DEBS1+TE transformants; BL21, total protein from BL21.

3.RT-PCR validation on transcriptional level

After failure of detecting target proteins using SDS-PAGE, we used RT-PCR to examine the expression of the target genes on transcriptional level. The results were shown in Fig. 5. We used 16S rRNA gene as an house keeping gene to compare the expression level of target genes. The results showed that DEBS1 gene was successfully expressed, albeit the transcription level was relatively low compared with other target genes (acc, pcc and sfp) whose transcription are controlled by T7 promoter as well. The low transcription level of DEBS1 gene is in consistent with our SDS-PAGE results. This result might due to the large size of DEBS1, which could hamper both the transcriptional and translational efficiency.

4.Fluorescence microscopy analyses

To further confirm the expression of DEBS1, we monitored the fluorescence of RFP under Zeiss LSM710 confocal microscope (Fig. 6). The results showed that after excitation with lazer source, red fluorescence was monitored in the RFP+DEBS1+TE transformant, while in the BL21 strain, no fluorescence was detected. These results indicate that RFP+DEBS1+TE was successfully expressed in the transformant.

Figure6: Results of fluorescence microscopy

5.UPLC-MS/MS Detection of TKL 1B

To test if RFP+DEBS1+TE was functionally expressed, the transformant was grown in Luria Broth (LB) liquid medium and induced with IPTG. After 48h of induction, cells were harvested and the supernatant was used for the extraction of the product TKL1B. After extraction, the samples were concentrated by 5 times and subjected into a triple quadrupole mass spectrometry and monitored under an MRM mode using the parent ion /product ion=173.1/155.1 channel. The results were shown in Fig. 7. The retention time of the target product in the UPLC column was 6.42 min, using water/methanol gradient elution protocol. We monitored a strong response at m/z = 155.1 ion pair in the RFP +DEBS1+TE transformant sample. The ion pair of 173.1/155.1 is specific to the expected product TKL1B, and is in consistent with previous studies. These results demonstrated that TKL1B was successfully produced from the transformant.[4]

Conclusion

In this project, we realized the design by standardizing domains from the gene cluster of DEBSs. To our knowledge, it was the first time to assemble domains as individual units.

We performed separations of the coding sequence resulted in standardized domains and assembled these domains in correct order for generating our target product triketide lactone.[5] Evaluation of recombinant DEBS+TE expression was performed by RT-PCR analyses, fluorescence microscopy analyses, SDS-PAGE and final polyketide product was detected using Mass Spectrometry. Our results demonstrated that standardized domains could express correctly folded and posttranslationally modified modular PKSs, we concluded that domains in PKSs can work relatively independently. Furthermore, RFP may influence the function of protein, but in our project the product could be generated via BAP1 harboring plasmids with RFP-DEBS1-TE. Also, there is one scar (Thr-Arg) every two parts produced in every ligations because of RFC 23 .We had worried these additional amino acids may affect the expression of our synthetase. However, the complete construction with 20 additional amino acids of RFP-DEBS1-TE, expressed well from the IPTG-inducible promoter present in pSB1C3 in E. coli strain BAP1. According to these results, we demonstrated the compatibility as well as interchangeability between different standardized domains. [6]

Based on investigation of standardized domains coming from semi- rational separation, the obtained results represented an important milestone toward the ultimate goal of creating novel PKSs by assembling standardized domains. [7]Additionally, these results showed a new approach to generating triketide lactone, which can be an important source of advanced intermediates of complex polyketides.[8]

References

[1] Keatinge-Clay, Adrian T. "The structures of type I polyketide synthases."Natural product reports 29.10 (2012): 1050-1073. [2] Menzella, Hugo G., John R. Carney, and Daniel V. Santi. "Rational design and assembly of synthetic trimodular polyketide synthases." Chemistry & biology 14.2 (2007): 143-151. [3]Pfeifer, Blaine A., et al. "Biosynthesis of complex polyketides in a metabolically engineered strain of E. coli." Science 291.5509 (2001): 1790-1792. [4] Menzella, Hugo G., et al. "Combinatorial polyketide biosynthesis by de novo design and rearrangement of modular polyketide synthase genes." Nature biotechnology 23.9 (2005): 1171-1176. [5] Jiang, Ming, and Blaine A. Pfeifer. "Metabolic and pathway engineering to influence native and altered erythromycin production through E. coli." Metabolic engineering 19 (2013): 42-49. [6] Yan, John, et al. "Functional Modular Dissection of DEBS1-TE Changes Triketide Lactone Ratios and Provides Insight into Acyl Group Loading, Hydrolysis, and ACP Transfer." Biochemistry 51.46 (2012): 9333-9341. [7] Jiang, Ming, and Blaine A. Pfeifer. "Metabolic and pathway engineering to influence native and altered erythromycin production through E. coli." Metabolic engineering 19 (2013): 42-49. [8] Hughes, Diarmaid. "Exploiting genomics, genetics and chemistry to combat antibiotic resistance." Nature reviews genetics 4.6 (2003): 432-441.