Skip to main content


  • Research
  • Open Access

Biosynthesis of phlorisovalerophenone and 4-hydroxy-6-isobutyl-2-pyrone in Escherichia coli from glucose

Microbial Cell Factories201615:149

  • Received: 20 June 2016
  • Accepted: 22 August 2016
  • Published:



Type III polyketide synthases (PKSs) contribute to the synthesis of many economically important natural products, which are typically produced by direct extraction from plants or synthesized chemically. For example, humulone and lupulone (Fig. 1a) in hops (Humulus lupulus) account for the characteristic bitter taste of beer and display multiple pharmacological effects. 4-Hydroxy-6-methyl-2-pyrone is a precursor of parasorboside contributing to insect and disease resistance of plant Gerbera hybrida, and was recently demonstrated to be a potential platform chemical.
Graphical Abstract image
Fig. 1

Examples of phloroglucinols (a) and 2-pyrones (b) synthesized by type III PKS. PIBP phlorisobutyrophenone; PIVP phlorisovalerophenone; TAL 4-hydroxy-6-methyl-2-pyrone (triacetic acid lactone); HIPP 4-hydroxy-6-isopropyl-2-pyrone; HIBP 4-hydroxy-6-isobutyl-2-pyrone


In this study, we achieved simultaneous biosynthesis of phlorisovalerophenone, a key intermediate of humulone biosynthesis and 4-hydroxy-6-isobutyl-2-pyrone in Escherichia coli from glucose. First, we constructed a biosynthetic pathway of isovaleryl-CoA via hydroxy-3-methylglutaryl CoA followed by dehydration, decarboxylation and reduction in E. coli. Subsequently, the type III PKSs valerophenone synthase or chalcone synthase from plants were introduced into the above E. coli strain, to produce phlorisovalerophenone and 4-hydroxy-6-isobutyl-2-pyrone at the highest titers of 6.4 or 66.5 mg/L, respectively.


The report of biosynthesis of phlorisovalerophenone and 4-hydroxy-6-isobutyl-2-pyrone in E. coli adds a new example to the list of valuable compounds synthesized in E. coli from renewable carbon resources by type III PKSs.


  • Isovaleryl-CoA
  • Phlorisovalerophenone
  • 4-Hydroxy-6-isobutyl-2-pyrone
  • Valerophenone synthase
  • Escherichia coli


A large number of natural products are synthesized by type III polyketide synthases (PKSs). These compounds play an important role in human nutrition and health, and have recently expanded their roles as platform chemicals. Acylphloroglucinol derivatives have been isolated from a number of plants [13]. Their chemical structures and intriguing biological activities have attracted increasing attention in recent years. Humulone and lupulone (Fig. 1a) in hops (Humulus lupulus) account for the characteristic bitter taste of beer and display multiple pharmacological effects [4]. Hyperforin (Fig. 1a) is one of the main active constituents in extracts of Hypericum perforatum used for the treatment of depression [5]. The 4-hydroxy-2-pyrone analogues display multiple biological activities and are valuable pharmaceutical precursors [6]. 4-Hydroxy-6-methyl-2-pyrone, also known as triacetic acid lactone (TAL, Fig. 1b) is a precursor of gerberin and parasorboside, which could protect Gerbera hybrid from fungal pathogens and attack by insects [7]. Fistupyrone (4-hydroxy-6-isovaleryl-2-pyrone) (Fig. 1b) was isolated from Streptomyces sp. TP-A0569 and showed inhibition of the infection of Chinese cabbage by a fungus Alternaria brassicicola [8]. Germicidin homologs (Fig. 1b) arisen from streptomyces inhibit spore germination and Na+/K+ ATPase [9, 10]. More recently, 4-hydroxy-2-pyrones were reported to be potential platform chemicals (Additional file 1: Figure S1a) [1113].

Advances in synthetic biology and metabolic engineering greatly promoted biosynthesis of a variety of valuable compounds, for instance, flavonoids produced by type III PKSs [14, 15] in microbial hosts. Biological production of phloroglucinol (Fig. 1a) and TAL (Fig. 1b) from glucose has also been accomplished using genetically modified microorganisms [6, 1619]. However, biosynthesis of acylphloroglucinols and TAL analogues from renewable feedstocks in Escherichia coli has not been well investigated.

The acylphloroglucinol cores, phlorisovalerophenone (PIVP) and phlorisobutyrophenone (PIBP) are formed by condensation of three malonyl-CoA-derived acetate units with isovaleryl-CoA or isobutyryl-CoA as the starter units [4, 20]. The enzyme valerophenone synthase (VPS) involved in the formation of PIVP in the biosynthesis of humulone has been characterized [2123]. VPS is a homologue of chalcone synthase (CHS). They share the same reaction mechanism namely claisen condensation, but differ in substrate specificity. CHS catalyzes condensation of one molecule of p-coumaroyl-CoA and three molecules of malonyl-CoA to form naringenin-chalcone. However, VPS preferentially uses isovaleryl-CoA or isobutyryl-CoA instead of p-coumaroyl-CoA as the starter unit to condense with three molecules of malonyl-CoA to form PIVP [21, 23]. Recently, dual functional CHS/VPS have also been reported. A CHS, FvCHS2-1 from strawberry (Fragaria vesca) was identified to be responsible for acylphloroglucinols synthesis in strawberry fruit [3]. Intriguingly, in vitro experiments demonstrated that type III PKSs such as VPS and CHS usually also synthesize TAL analogues, 4-hydroxy-6-isobutyl-2-pyrone (HIBP) and 4-hydroxy-6-isopropyl-2-pyrone (HIPP) (Fig. 1b), carrying isobutyl or isopropyl at position 6 with isovaleryl-CoA or isobutyl-CoA as the starter unit and malonyl-CoA as the extender unit.

In this work, we engineered the synthesis of isovaleryl-CoA in E. coli by recruiting a biosynthetic pathway via hydroxy-3-methylglutaryl CoA (HMG-CoA), an intermediate of the mevalonate pathway (Fig. 2a) [24]. The VPS from hops (Humulus lupulus) HlVPS, the CHS from strawberry (Fragaria vesca) FvCHS2-1 and a newly cloned CHS from H. perforatum HpCHS were introduced into the isovaleryl-CoA producing E. coli strain separately, led to the simultaneous production of PIVP and HIBP at different ratios. The highest titers of PIVP and HIBP produced by the recombinant strains reached 6.4 and 66.5 mg/L, respectively. This work adds new products to the list of valuable compounds biosynthesized in E. coli using type III PKSs and lays a foundation for microbial synthesis of not only acylphloroglucinol derivatives, but also potential platform chemicals.
Fig. 2
Fig. 2

The biosynthetic pathway of phlorisovalerophenone (PIVP) and 4-hydroxy-6-isobutyl-2-pyrone (HIBP) constructed from glucose in E. coli. a The biosynthetic pathway of isovaleryl-CoA. Red arrows represent the constructed isovaleryl-CoA biosynthetic pathway via HMG-CoA followed by dehydration, decarboxylation and reduction. Black arrows represent the biosynthesis of isovaleryl-CoA catalyzed by the BCDH complex. Dotted arrows represent the mevalonate-dependent pathway of isoprenoid. b The formation of PIVP and HIBP catalyzed by VPS. PDH pyruvate dehydrogenase complex, ACC acetyl-CoA carboxylase, ERG10 acetyl-CoA acetyltransferase, ERG13 HMG-CoA synthase, LiuC HMG-CoA dehydratase, AibA/B MG-CoA decarboxylase, AibC DMA-CoA reductase, Als acetolactate synthase, IlvC ketol-acid reductoisomerase, IlvD dihydroxy-acid dehydratase, LeuA α-isopropylmalate synthase, LeuCD α-isopropylmalate isomerase complex, LeuB isopropylmalate dehydrogenase, BCDH branched-chain α-keto acid dehydrogenase complex, BCAT branched-chain amino acid transaminase, VPS valerophenone synthase


Strains, plasmids and medium

Strains and plasmids used in this study are listed in Table 1. Luria–Bertani (LB) medium was used for the propagation of E. coli cells for plasmid construction. Terrific Broth (TB) medium (12 g/L tryptone, 24 g/L yeast extract, 4 mL/L glycerol, 2.3 g/L KH2PO4, and 12.5 g/L K2HPO4) and modified M9 medium (1 × M9 minimal salts, 20 g/L glucose, 10 g/L yeast extract, 2 mM MgSO4, and 0.1 mM CaCl2) were used for protein expression and PIVP production, respectively. Ampicillin (100 mg/L) and streptomycin (100 mg/L) were added to the medium as needed.
Table 1

Bacterial strains and plasmids






pBR322 ori with PT7; AmpR



CDF ori with PT7; SmR



pETDuet-1 carrying aibA, aibB, aibC and liuC

This study


pCDFDuet-1 carrying ERG10 and ERG13

This study


pCDFDuet-1 carrying ERG10, ERG13 and HlVPS syn

This study


pCDFDuet-1 carrying ERG10, ERG13 and FvCHS2-1 syn

This study


pCDFDuet-1 carrying ERG10, ERG13 and HpCHS

This study



Myxococcus xanthus (CGMCC 1.3865)



Saccharomyces cerevisiae BY4742



E. coli BL21 (DE3) with pETDuet-1 and pCDFDuet-1

This study


E. coli BL21 (DE3) with pET-A and pCDF-E

This study


E. coli BL21 (DE3) with pET-A and pCDF-EV1

This study


E. coli BL21 (DE3) with pET-A and pCDF-EV2

This study


E. coli BL21 (DE3) with pET-A and pCDF-EV3

This study

CGMCC China general microbiological culture collection center

Construction of plasmids

Oligonucleotide primers used in this study are summarized in Additional file 1: Table S1. AibA, aibB, aibC and liuC were PCR-amplified from Myxococcus xanthus (CGMCC 1.3865) genome with primer pairs aibA-F/aibA-R, aibB-F/aibB-R, aibC-F/aibC-R or liuC-F/liuC-R, respectively. The amplified gene products of aibA, aibB and aibC were digested with restriction enzyme and simultaneously subcloned into the multiple cloning sites (MCS) 1 of pETDuet-1 by Golden Gate [25] cloning using AarI. Subsequently, the gene product of liuC was digested and ligated into the above resulting plasmid using HindIII/AflII to yield the final plasmid pET-A. Here, the sequences of T7 promoter, lac operator and ribosome binding site (RBS) from pETDuet-1 were incorporated into the primer design, so each gene was driven by a single T7 promoter. ERG10 and ERG13 were PCR-amplified from Saccharomyces cerevisiae BY4742 genome with primer pairs ERG10-F/ERG10-R or ERG13-F/ERG13-R, respectively. First, the ERG13 fragment was digested and cloned into NcoI/BamHI sites of pCDFDuet-1. The ERG10 fragment was digested and ligated into the above plasmid via NdeI/XhoI sites, generating the recombinant plasmid pCDF-E. The HlVPS gene from hops (GenBank: AB015430.1) was codon optimized for expression in E. coli and synthesized by Shanghai Generay Biotech Co., Ltd. (Shanghai, China), designated as HlVPS syn (Additional file 1: Table S2). HlVPS syn was PCR-amplified with primer pair VPS-F/VPS-R, and cloned into pET28a through NdeI/AflII sites generating the intermediate expression vector pET28a-HlVPS. The T7-HlVPS syn fragment with a T7 promoter and an N-terminal His tag was PCR-amplified from pET28a-HlVPS with primer pair T7VPS-F/T7VPS-R, digested and ligated into pCDF-E via NotI/AflII to generate the final recombinant plasmid pCDF-EV1. FvCHS2-1 from strawberry (GenBank: XM_004306495.1) was codon optimized for expression in E. coli and synthesized by Shanghai Generay Biotech Co., Ltd. (Shanghai, China), designated as FvCHS2-1 syn (Additional file 1: Table S2). HpCHS was obtained from the mRNA of H. perforatum by RT-PCR. RNA of H. perforatum callus was isolated by RNAprep pure Plant Kit from Tiangen Biotech Co., Ltd. (Beijing, China). Reverse transcription was performed using the First Strand cDNA Synthesis Kit from Toyobo Co., Ltd. (Osaka, Japan). PCR reactions were performed using High Fidelity PCR system (Fermentas, Germany) with primer pair HpCHS-F/HpCHS-R. The HpCHS sequence was submitted into NCBI database (GenBank: KU180217). FvCHS2-1 syn and HpCHS were ligated into pCDF-E individually using the same method as pCDF-EV1 construction, yielding pCDF-EV2 and pCDF-EV3, respectively.

Biosynthesis of isovaleryl-CoA in E. coli

The plasmids, pET-A and pCDF-E, were electroporated into BL21 (DE3), generating the recombinant strain APG-IV. The strain APG-0 harboring pETDuet-1 and pCDFDuet-1 was used as the control. A 1 mL aliquot of the overnight cultured single colony was inoculated into 50 mL LB medium containing antibiotics and cultivated at 37 °C, 200 rpm with shaking. When OD600 of the culture reached 0.6, 0.1 mM isopropyl-β-D-thiogalactoside (IPTG) was added to induce recombinant protein expression at 30 °C for 12 h. Subsequently, the cells were centrifuged (4000 rpm, 30 °C, 10 min), washed and resuspended in 50 mL modified M9 medium and cultured at 30 °C for 3 h. The cell culture was chilled on ice for 10 min. Cells were harvested by centrifugation (4000 rpm, 4 °C, 10 min) and resuspended in 1.5 mL extraction buffer (methanol:acetonitrile:water = 2:2:1, v/v, with 0.1 % formic acid). After sonication and centrifugation, 50 μL of the supernatants were used for HPLC and LC–MS analysis.

Biosynthesis of PIVP and HIBP in E. coli

Escherichia coli strain BL21 (DE3) was transformed with plasmids pET-A and pCDF-EV1, pET-A and pCDF-EV2 or pET-A and pCDF-EV3 to generate the recombinant strains APG-1, APG-2 or APG-3, respectively. Two stage fermentation processes were applied to achieve the production of PIVP and HIBP. The first stage is for pathway genes overexpression, and recombinant strains were cultivated in TB medium. Single colonies were cultured in 3 mL LB medium with antibiotics at 37 °C for 7 h as seed culture, 1 mL of which was transferred into 50 mL TB medium with antibiotics and cultured in a shake flask at 37 °C until an OD600 of 1.2 followed by addition of 0.1 mM IPTG. To determine optimal temperature for fermentation conditions at the stage, the culture was incubated at 16, 23 or 30 °C for 16 h. Different concentrations of IPTG (0.01, 0.05, 0.1, 0.2 or 0.3 mM) were also added in the TB medium to determine the optimal concentrations. The cell pellets were harvested by centrifugation and resuspended in an equal volume of modified M9 medium and fermented at 30 °C, 200 rpm for secondary metabolite production as the second stage. Time course for production of PIVP and HIBP by strain APG-1 was determined for 12, 24, 36, 48 and 60 h. Subsequently, the culture broth was centrifuged at 4000 rpm for 10 min. The supernatants were extracted with an equal volume of ethyl acetate twice, concentrated by rotary evaporator, and re-dissolved in 1.5 mL methanol. To extract the intracellular metabolites, cell pellets were resuspended in 10 mL 80 % acetone, sonicated, centrifuged and the supernatant was evaporated to remove the acetone. The procedure was repeated once. Subsequently, the remaining water phase was also extracted with an equal volume of ethyl acetate twice and re-dissolved in 1.5 mL methanol as the supernatant of the fermented broth. Then 20 µL extraction samples of the fermentation broth or the cell pellets were analyzed by HPLC–MS separately.

Compounds purification

2 L fermentation broth of the recombinant strain APG-1 was centrifuged to separate into supernatant and cell pellets. After crude extraction using the above protocol, purification of HIBP and PIVP was conducted by semi-preparative HPLC performed on a Shimadzu LC-6 AD with SPD-20A detector. The HPLC conditions were as follows: solvent A = H2O; solvent B = methanol; flow rate: 4 mL/min, 0–5 min 90 % A and 10 % B, 6–40 min 90 % A and 10–100 % B (linear gradient), 41–45 min 100 % B. A YMC-pack ODS-A (10 × 250 mm; particle size, 5 μm) was used as the stationary phase.

Chemical analysis and quantification

HPLC–MS was performed on an Agilent 1260 system with 1260 Infinity UV detector and a Bruker microQ-TOF II mass spectrometer equipped with an ESI ionization probe. The innoval C18 column (4.6 × 250 mm; 5 μm particle size) was used in this study. The HPLC conditions for isovaleryl-CoA, HIBP and PIVP were as follows: solvent A = H2O (containing 20 mM ammonium acetate for the analysis of isovaleryl-CoA, or containing 0.1 % formic acid for the analysis of HIBP and PIVP); solvent B = methanol; flow rate = 1 mL/min; 0–5 min 95 % A and 5 % B, 6–45 min 95 % A and 5 % B to 100 % B (linear gradient). All these products were detected at 254 nm. Standard calibration curve of isovaleryl-CoA was generated with a series of known concentrations of the isovaleryl-CoA standard purchased from Sigma-Aldrich (Milwaukee, WI, USA). Standard calibration curves of HIBP and PIVP were generated with a series of known concentrations of the purified compounds from large scale fermentation. All experiments were carried out in triplicate and repeated at least twice. The titer was presented as mean ± SD.

NMR analysis

NMR experiments were performed on a Bruker Avance 400 (Karlsruhe, Germany). The sample were dried by evaporation and dissolved in 500 µL of DMSO-d 6 and transferred into 2.5 mm NMR tube. Chemical shifts were expressed in δ (ppm) and coupling constants (J) were given in Hertz (Hz) (Additional file 1: Figure S2).

HIBP 1H-NMR (DMSO-d 6, 400 MHz) δ 5.94 (d, J = 2.1 Hz, 1H), 5.21 (d, J = 2.1 Hz, 1H), 2.30 (d, J = 7.2 Hz, 2H), 1.93 (m, 1H), 0.89 (d, J = 6.6 Hz, 6H);

PIVP 1H-NMR (DMSO-d 6, 400 MHz), δ 5.79 (s, 2H), 2.86 (d, J = 6.8 Hz, 2H), 2.13 (m, 1H), 0.91 (d, J = 6.7 Hz, 6H).


Engineered synthesis of isovaleryl-CoA in E. coli

First, we generated isovaleryl-CoA in E. coli cells employing a pathway from M. xanthus via hydroxy-3-methylglutaryl CoA (HMG-CoA), a key intermediate of mevalonate pathway in the biosynthesis of isoprenoids [24]. Short branched-chain acyl-CoAs are usually derived from the branched-chain amino acids as a mixture via transamination and subsequent oxidative decarboxylation catalyzed by the branched-chain α-keto acid dehydrogenase complex (BCDH) [26]. In current work, acetyl-CoA is converted into HMG-CoA as in the mevalonate pathway (Fig. 2a). Enzymes ERG10 (acetyl-CoA acetyltransferase) and ERG13 (HMG-CoA synthase) from S. cerevisiae were used for efficient biosynthesis of HMG-CoA, as these two enzymes yielded enhanced levels of mevalonate in E. coli [2729]. The three enzymes from M. xanthus, including LiuC (HMG-CoA dehydratase), AibA/B (3-methylglutaconyl-CoA (MG-CoA) decarboxylase) and AibC (3,3-dimethylacrylyl CoA (DMA-CoA) reductase), were used to transform HMG-CoA into isovaleryl-CoA consecutively via MG-CoA and DMA-CoA in the above E. coli strain (Fig. 2a). Acyl-CoAs were extracted from the engineered E. coli strain APG-IV harboring pET-A and pCDF-E and analyzed by HPLC–MS with strain APG-0 harboring pETDuet-1 and pCDFDuet-1 as the negative control (Fig. 3). This procedure confirmed the production of isovaleryl-CoA by comparing the retention time (t R  = 24.5 min) and the molecular ion [(M + H)+ = 852.1802] with those of the isovaleryl-CoA standard (Fig. 3b, c). The titer of isovaleryl-CoA from this new pathway was 1421.8 ± 129.2 nmol/g wet weight, about sixteen times higher than the titer of our previous engineered E. coli strain using the BCDH complex under the same fermentation and detection conditions (Fig. 3d) [26].
Fig. 3
Fig. 3

Engineered synthesis of isovaleryl-CoA by E. coli strains. a HPLC analysis of chemical profiles produced by E. coli strain APG-0 harboring empty vectors. b HPLC analysis of isovaleryl-CoA produced by E. coli strain APG-IV carrying pET-A and pCDF-E. c The isovaleryl-CoA standard. d MS chromatogram showing the synthesis of isovaleryl-CoA. e Titers of isovaleryl-CoA produced by the recombinant strain APG-IV and our previous E. coli expressing the BCDH complex [26]

Engineered synthesis of PIVP and HIBP in E. coli carrying HlVPS syn and optimization of fermentation conditions

The HlVPS syn from hops was synthesized and introduced into the isovaleryl-CoA producing E. coli strain, generating the recombinant strain APG-1. After fermentation, the broth supernatants and cell pellets were analyzed by HPLC–MS. As shown in the chromatogram, the recombinant strain APG-1 produced two new compounds (Fig. 4c, d), compared with the control strain APG-0 (Fig. 4a, b). The first compound with a t R of 35.6 min was initially identified as HIBP by LC–MS (Additional file 1: Figure S2a, c; [M + H]+ = 169.0763). In the previous enzymatic studies, the production of HIBP was confirmed only by LC–MS analysis probably due to no enough products for NMR analysis [3, 23]. In this work, about 74.9 mg HIBP was isolated from 2 L scale culture, and we were able to confirm the structure by 1D-NMR spectroscopy analysis (Additional file 1: Figure S2a, c). The second product with a t R of 39.7 min was identified as PIVP (Additional file 1: Figure S2b, [M + H]+ = 211.0925). About 6.5 mg HIBP was isolated from 2 L culture, and the PIVP structure was further confirmed by 1H-NMR spectroscopy analysis (Additional file 1: Figure S2d). The localization of the two products was also tested. PIVP mainly accumulated in the cell pellets (Fig. 4c), while HIBP was distributed in the fermentation broth as well as the cell pellets (Fig. 4c, d).
Fig. 4
Fig. 4

Engineered synthesis of 4-hydroxy-6-isovaleryl-2-pyrone (HIBP) and phlorisovalerophenone (PIVP) by E. coli strains. a, b Strain APG-0 harboring empty vectors as negative control. c, d Strain APG-1 harboring plasmids pET-A and pCDF-EV1. e, f Strain APG-2 harboring plasmids pET-A and pCDF-EV2. g, h Strain APG-3 harboring plasmids pET-A and pCDF-EV3. a, c, e, g HPLC analysis of metabolites extracted from cell pellets. b, d, f, h HPLC analysis of metabolites extracted from the fermentation broth. i The titers of HIBP produced by the strains APG-1, APG-2 and APG-3. j The titers of PIVP produced by the strains APG-1, APG-2 and APG-3

We further optimized the production of PIVP and HIBP by varying fermentation conditions for the recombinant strain APG-1 such as temperature and IPTG concentration at the first stage. The recombinant strain APG-1 was incubated in TB medium for 16 h at varying temperature (16, 23 or 30 °C) with a fixed IPTG concentration of 0.1 mM. The cell pellets were harvested and resuspended in modified M9 medium and incubated at 30 °C for 36 h. Combining both the fermentation broth and cell pellets, the titerss of PIVP and HIBP exhibited best at 23 °C (Fig. 5a). After that, different concentrations of IPTG, 0.01, 0.05, 0.1, 0.2 and 0.4 mM were used to induce protein expression at 23 °C. The titers of PIVP and HIBP were highest when the concentration of IPTG was 0.05 mM as shown in Fig. 5b. Furthermore, we monitored titers of PIVP and HIBP produced by APG-1 cultivated in modified M9 medium at different fermentation time. The results are summarized in Fig. 5c. The titers of HIBP and PIVP reached the highest in 24 h, which were 52.0 ± 7.9 and 6.4 ± 0.6 mg/L, respectively. Under these concentrations, no inhibition on cell growth by the compounds HIBP and PIVP was observed.
Fig. 5
Fig. 5

Optimization of fermentation conditions. a The titers of PIVP and HIBP by strain APG-1 at different protein induction temperature. b The titers of PIVP and HIBP by strain APG-1 at different IPTG concentration. c The titers of PIVP and HIBP by strain APG-1 at different fermentation time

Biosynthesis of HIBP and PIVP in E. coli using CHSs

In addition to HlVPS, a CHS from strawberry (F. vesca), FvCHS2-1 played a role as a bifunctional CHS/VPS enzyme [3]. We performed codon optimization for the FvCHS2-1 gene, synthesized and introduced the FvCHS2-1syn into the isovaleryl-CoA producing E. coli strain, generating APG-2 and fermented under the optimized conditions. The protein expression was induced at 23 °C for 16 h with 0.05 mM IPTG, and then the fermentation was conducted in modified M9 medium at 30 °C for 24 h. HPLC analysis showed that the strain APG-2 also produced HIBP and PIVP, although the titer of PIVP was much lower than that carrying HlVPS syn , only 1.4 ± 0.2 mg/L (Fig. 4e, f, j). However, the titer of HIBP reached 53.3 ± 9.0 mg/L, which was similar to that of the strain harboring HlVPS syn (Fig. 4i).

VPS was involved in the formation of PIVP in the biosynthesis of humulone and could also produce PIPB using isobutyryl-CoA as the starter unit [2123]. PIBP is the core structure of hyperforin, an antidepressant extracted from Hypericum species. The enzyme catalyzing formation of PIBP is still unknown. It was assumed that there may be a VPS-like or dual functional CHS/VPS enzyme in the Hypericum species responsible for the formation of PIBP, and more than likely, this enzyme could also use isovaleryl-CoA as the starter molecular to form PIVP and HIBP. Based on the reported type III PKSs gene sequences of the Hypericum species [30] in the NCBI database, several candidate genes were amplified from the callus of H. perforatum by homology cloning. Among these genes, a suspected CHS gene showing 96.2 % similarity with FvCHS2-1 at the amino acid level was designated as HpCHS (GenBank: KU180217), and introduced into the above isovaleryl-CoA producer. The recombinant E. coli strain APG-3 harboring the isovaleryl-CoA pathway genes and HpCHS was fermented as described above. The HIBP titer produced by this strain reached 66.5 ± 13.1 mg/L, which was the highest among all the three producers (Fig. 4g–i). The PIVP titer in APG-3 was 2.8 ± 0.5 mg/L, which was lower than that of the APG-1 harboring HlVPS syn (Fig. 4g, h, j).


Short branched-chain acyl-CoAs are important building blocks for a large number of valuable compounds [5, 24, 31]. However, native E. coli metabolism does not produce these short branched-chain acyl-CoAs, which hampers the heterologous production of those economically important chemicals in this host. Previously engineered short branched-chain acyl-CoA producing E. coli strains usually biosynthesized isobutyryl-CoA, isovaleryl-CoA and 2-methyl-butyryl-CoA as a mixture derived from the metabolic pathways of branched-chain amino acids catalyzed by the BCDH complex [26, 32, 33]. In this study, we successfully demonstrated the feasibility of synthesizing isovaleryl-CoA in E. coli via HMG-CoA (Fig. 2a) recruiting five enzymes from yeast and myxobacteria [24]. Isovaleryl-CoA was the only product in this pathway which makes the engineered E. coli strain more suitable for producing compounds such as humulone and lupulone with isovaleryl-CoA as the building block. In the BCDH dependent pathway, isovaleryl-CoA and isobutyryl-CoA are derived from α-ketoisocaproate or α-ketoisovalerate directly, and α-ketoisovalerate is an intermediate for the formation of α-ketoisocaproate. Thus the formation of isobutyryl-CoA competes with isovaleryl-CoA in the pathway (Fig. 2a). Meanwhile short branched-chain keto acids would also be transformed into corresponding amino acids such as leucine, isoleucine and valine beyond acyl-CoAs (Fig. 2a). HMG-CoA formation with ERG10 and ERG13 derived from acetyl-CoA has been well established in the mevalonate pathway (Fig. 2a) in E. coli [2729]. By recruiting LiuC, AibA/B and AibC, the HMG-CoA was further uniquely diverted into the formation of isovaleryl-CoA via MG-CoA and DMA-CoA. This engineered strain represents a novel “factory” for biosynthesis of a series of valuable chemicals with isovaleryl-CoA as the building block, including fatty acids, biofuels and natural products [26, 32, 33].

Introduction of VPS from hops, CHSs from strawberry or H. perforatum to the above isovaleryl-CoA producing E. coli strain resulted in the biosynthesis of PIVP and HIBP. PIVP was detected in hop extracts [34], and is the key intermediate of humulone, which contribute the characteristic bitter flavor of beer and display multiple pharmacological effects [4, 31]. The work described here may pave the way for synthesis of humulone and other derivatives. HIBP has not been isolated from any natural resources [1113]. As an analogue of TAL, HIBP could be used as a potential platform chemical to produce various chemical intermediates and end products with short branched-chains (Additional file 1: Figure S1a, b) [1113, 35]. More experiments are needed to test whether HIBP itself has any antibacterial or fungicidal activity or not.

In this study, the VPS from hops was the best enzyme for PIVP formation. The newly cloned CHS from H. perforatum was the most appropriate enzyme for the synthesis of HIBP. The ratios of HIBP and PIVP were calculated based on the titers produced by strains APG-1 (52.0 over 6.4 mg/L), APG-2 (53.3 over 1.4 mg/L) and APG-3 (66.5 over 2.8 mg/L), which were about 8:1, 37:1 and 24:1, respectively. In the in vitro enzyme activity analysis of HlVPS and FvCHS, PIVP was the major product, instead of HIBP [3, 22]. It was reported that the PIVP and HIBP content was affected by the relative ratio of isovaleryl-CoA/malonyl-CoA in the in vitro experiment [3]. In our recombinant E. coli, the production of isovaleryl-CoA and malonyl-CoA both derived from acetyl-CoA, and there may be a dynamic balance between the concentration of isovaleryl-CoA and malonyl-CoA in vivo (Fig. 2a), which may affect the ratio of HIBP and PIVP.

Amino acids sequence alignment of HlVPS, FvCHS2-1 and HpCHS showed high similarity (Additional file 1: Figure S4). The three essential catalytic amino acids Cys-His-Asn, the active site loop of CHS enzymes GFGPG, and two Phe residues, important in determining the substrate specificity of CHS are all well conserved in these three proteins (Additional file 1: Figure S4). However, ratios between PIVP and HIBP produced by these three enzymes are quite different (Fig. 4). Several protein engineering studies have elucidated the basis of starter molecule selectivity and the control of polyketide length of type III PKSs. For example, Jez and colleagues reported that a triple mutant (T197L/G256L/S338I) of CHS, generated an enzyme that was functionally identical to 2-PS, catalyzing the synthesis of the TAL from an acetyl-CoA starter molecule and two malonyl-CoAs [36]. In 2005, Abe and co-workers reported that a point mutation (M207G) in a pentaketide chromone synthase (PCS) expanded the volume of the catalytic cavity, to convert PCS into octaketide synthase [37]. Thus, modeling analyses and site-directed mutagenesis experiments are needed to improve the product specificity and increase the synthesis of PIVP or HIBP.

Directed evolution may be also used to increase the activity of HlVPS, FvCHS2-1 or HpCHS. Using directed evolution, Rao and colleagues obtained a series of thermostable PhlD mutants with improved phloroglucinol productivity [19]. Recently, Tang and colleagues engineered the E. coli regulatory protein AraC to activate gene expression in response to TAL, developed an endogenous TAL reporter system in E. coli. Using this system, they conducted in vivo directed evolution of 2-PS in E. coli, and obtained a 2-PS variant conferring ~20-fold higher TAL production [6]. Similar reporter systems may also be constructed in our recombinant E. coli to facilitate high throughput screening of HlVPS, FvCHS2-1 or HpCHS for improved productivity of HIBP or PIVP. At the same time, precursor supply including isovaleryl-CoA and malonyl-CoA may be improved for the production of HIBP and PIVP. The availability of intracellular malonyl-CoA may be enhanced by over-expressing the acetyl-CoA carboxylase (Acc) and acetyl-CoA synthase (Acs) genes and deletion of competing pathways, which was applied in improving phloroglucinol productivity by engineered E. coli strain [19, 38]. In the case of HIBP and PIVP, as both substrates isovaleryl-CoA and malonyl-CoA were derived from acetyl-CoA, there may be a dynamic balance between these two substrates and may need more elaborate regulation.


In this work, we firstly constructed a biosynthetic pathway of isovaleryl-CoA in E. coli by recruiting a route via hydroxy-3-methylglutaryl CoA, and then investigated the potential of bio-renewable production of PIVP and HIBP with E. coli as the host using type III PKSs. Even though still in infancy, the work described here may pave the way for microbial synthesis of not only acylphloroglucinols derivatives, but also “unnatural” TAL analogues.



2-pyrone synthase


branched-chain α-keto acid dehydrogenase complex


chalcone synthase


3,3-dimethylacrylyl CoA






hydroxy-3-methylglutaryl CoA






pentaketide chromone synthase






polyketide synthases


triacetic acid lactone


valerophenone synthase


Authors’ contributions

TL, HB, WZ, YZ, YB and YM designed the project. WZ, HB, YZ and YB performed the experiments. WZ, HB, YZ and TL analyzed the data. TL, HB, WZ and YZ wrote the manuscript. All authors read and approved the final manuscript.



Competing interests

The authors declare that they have no competing interests.

Availability of data and materials

Codon-optimized sequences generated in this project are listed in Additional file 1: Table S2. The HpCHS sequence was submitted into Genbank ( and the accession number was indicated in the “Methods” section.

Ethics approval and consent to participate

This manuscript dose not report data collected from humans or animals.


This work was supported by grants from the 973 Program of China (2012CB721100), the National Natural Science Foundation of China (31300040 and 21302214).

Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (, which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver ( applies to the data made available in this article, unless otherwise stated.

Authors’ Affiliations

Tianjin Institute of Industrial Biotechnology, Chinese Academy of Sciences, Tianjin, 300308, China
Key Laboratory of Systems Microbial Biotechnology, Chinese Academy of Sciences, Tianjin, 300308, China
University of Chinese Academy of Sciences, Beijing, China


  1. Bohr G, Gerhauser C, Knauft J, Zapp J, Becker H. Anti-inflammatory acylphloroglucinol derivatives from Hops (Humulus lupulus). J Nat Prod. 2005;68:1545–8.View ArticleGoogle Scholar
  2. Crispin MC, Hur M, Park T, Kim YH, Wurtele ES. Identification and biosynthesis of acylphloroglucinols in Hypericum gentianoides. Physiol Plant. 2013;148:354–70.View ArticleGoogle Scholar
  3. Song C, Ring L, Hoffmann T, Huang FC, Slovin JP, Schwab W. Acylphloroglucinol biosynthesis in strawberry fruit. Plant Physiol. 2015;15:00794.Google Scholar
  4. Clark SM, Vaitheeswaran V, Ambrose SJ, Purves RW, Page JE. Transcriptome analysis of bitter acid biosynthesis and precursor pathways in hop (Humulus lupulus). BMC Plant Biol. 2013;13:12.View ArticleGoogle Scholar
  5. Barnes J, Anderson LA, Phillipson JD. St John’s wort (Hypericum perforatum L.): a review of its chemistry, pharmacology and clinical properties. J Pharm Pharmacol. 2001;53:583–600.View ArticleGoogle Scholar
  6. Tang SY, Qian S, Akinterinwa O, Frei CS, Gredell JA, Cirino PC. Screening for enhanced triacetic acid lactone production by recombinant Escherichia coli expressing a designed triacetic acid lactone reporter. J Am Chem Soc. 2013;135:10099–103.View ArticleGoogle Scholar
  7. Eckermann S, Schroder G, Schmidt J, Strack D, Edrada RA, Helariutta Y, Elomaa P, Kotilainen M, Kilpelainen I, Proksch P, et al. New pathway to polyketides in plants. Nature. 1998;396:387–90.View ArticleGoogle Scholar
  8. Igarashi Y, Ogawa M, Sato Y, Saito N, Yoshida R, Kunoh H, Onaka H, Furumai T. Fistupyrone, a novel inhibitor of the infection of Chinese cabbage by Alternaria brassicicola, from Streptomyces sp. TP-A0569. J Antibiot (Tokyo). 2000;53:1117–22.View ArticleGoogle Scholar
  9. Petersen F, Zahner H, Metzger JW, Freund S, Hummel RP. Germicidin, an autoregulative germination inhibitor of Streptomyces viridochromogenes NRRL B-1551. J Antibiot (Tokyo). 1993;46:1126–38.View ArticleGoogle Scholar
  10. Aoki Y, Matsumoto D, Kawaide H, Natsume M. Physiological role of germicidins in spore germination and hyphal elongation in Streptomyces coelicolor A3(2). J Antibiot (Tokyo). 2011;64:607–11.View ArticleGoogle Scholar
  11. Chia M, Schwartz TJ, Shanks BH, Dumesic JA. Triacetic acid lactone as a potential biorenewable platform chemical. Green Chem. 2012;14:1850–3.View ArticleGoogle Scholar
  12. Dumesic JA, Chia M. Production of 2,4-diones from 4-hydroxy-6-substituted-2-pyrones. 2013; United State Patent Application, No. 20120283477A1.Google Scholar
  13. Schwartz TJ, O’Neill BJ, Shanks BH, Dumesic JA. ChemInform abstract: bridging the chemical and biological catalysis gap: challenges and outlooks for producing sustainable chemicals. ChemInform. 2014;45:2060–9.Google Scholar
  14. Mora-Pale M, Sanchez-Rodriguez SP, Linhardt RJ, Dordick JS, Koffas MA. Metabolic engineering and in vitro biosynthesis of phytochemicals and non-natural analogues. Plant Sci. 2013;210:10–24.View ArticleGoogle Scholar
  15. Trantas EA, Koffas MA, Xu P, Ververidis F. When plants produce not enough or at all: metabolic engineering of flavonoids in microbial hosts. Front Plant Sci. 2015;6:7.View ArticleGoogle Scholar
  16. Xie D, Shao Z, Achkar J, Zha W, Frost JW, Zhao H. Microbial synthesis of triacetic acid lactone. Biotechnol Bioeng. 2006;93:727–36.View ArticleGoogle Scholar
  17. Achkar J, Xian M, Zhao H, Frost JW. Biosynthesis of phloroglucinol. J Am Chem Soc. 2005;127:5332–3.View ArticleGoogle Scholar
  18. Cardenas J, Da Silva NA. Metabolic engineering of Saccharomyces cerevisiae for the production of triacetic acid lactone. Metab Eng. 2014;25:194–203.View ArticleGoogle Scholar
  19. Rao G, Lee J-K, Zhao H. Directed evolution of phloroglucinol synthase PhlD with increased stability for phloroglucinol production. Appl Microbiol Biotechnol. 2013;97:5861–7.View ArticleGoogle Scholar
  20. Adam P, Arigoni D, Bacher A, Eisenreich W. Biosynthesis of hyperforin in Hypericum perforatum. J Med Chem. 2002;45:4786–93.View ArticleGoogle Scholar
  21. Novák P, Matoušek J, Bříza J. Valerophenone synthase-like chalcone synthase homologues in Humulus lupulus. Biol Plant. 2003;46:375–81.View ArticleGoogle Scholar
  22. Okada Y, Ito K. Cloning and analysis of valerophenone synthase gene expressed specifically in lupulin gland of hop (Humulus lupulus L.). Biosci Biotechnol Biochem. 2001;65:150–5.View ArticleGoogle Scholar
  23. Paniego NB, Zuurbier KW, Fung SY, van der Heijden R, Scheffer JJ, Verpoorte R. Phlorisovalerophenone synthase, a novel polyketide synthase from hop (Humulus lupulus L.) cones. Eur J Biochem. 1999;262:612–6.View ArticleGoogle Scholar
  24. Li Y, Luxenburger E, Muller R. An alternative isovaleryl CoA biosynthetic pathway involving a previously unknown 3-methylglutaconyl CoA decarboxylase. Angew Chem Int Ed Engl. 2013;52:1304–8.View ArticleGoogle Scholar
  25. Engler C, Marillonnet S. Golden Gate cloning. Methods Mol Biol. 2014;1116:119–31.View ArticleGoogle Scholar
  26. Bi H, Bai Y, Cai T, Zhuang Y, Liang X, Zhang X, Liu T, Ma Y. Engineered short branched-chain acyl-CoA synthesis in E. coli and acylation of chloramphenicol to branched-chain derivatives. Appl Microbiol Biotechnol. 2013;97:10339–48.View ArticleGoogle Scholar
  27. Martin VJ, Pitera DJ, Withers ST, Newman JD, Keasling JD. Engineering a mevalonate pathway in Escherichia coli for production of terpenoids. Nat Biotechnol. 2003;21:796–802.View ArticleGoogle Scholar
  28. Steen EJ, Chan R, Prasad N, Myers S, Petzold CJ, Redding A, Ouellet M, Keasling JD. Metabolic engineering of Saccharomyces cerevisiae for the production of n-butanol. Microb Cell Fact. 2008;7:36.View ArticleGoogle Scholar
  29. Zhu F, Zhong X, Hu M, Lu L, Deng Z, Liu T. In vitro reconstitution of mevalonate pathway and targeted engineering of farnesene overproduction in Escherichia coli. Biotechnol Bioeng. 2014;111:1396–405.View ArticleGoogle Scholar
  30. Karppinen K, Hokkanen J, Mattila S, Neubauer P, Hohtola A. Octaketide-producing type III polyketide synthase from Hypericum perforatum is expressed in dark glands accumulating hypericins. FEBS J. 2008;275:4329–42.View ArticleGoogle Scholar
  31. Van Cleemput M, Cattoor K, De Bosscher K, Haegeman G, De Keukeleire D, Heyerick A. Hop (Humulus lupulus)-derived bitter acids as multipotent bioactive compounds. J Nat Prod. 2009;72:1220–30.View ArticleGoogle Scholar
  32. Atsumi S, Hanai T, Liao JC. Non-fermentative pathways for synthesis of branched-chain higher alcohols as biofuels. Nature. 2008;451:86–9.View ArticleGoogle Scholar
  33. Howard TP, Middelhaufe S, Moore K, Edner C, Kolak DM, Taylor GN, Parker DA, Lee R, Smirnoff N, Aves SJ, Love J. Synthesis of customized petroleum-replica fuel molecules by targeted modification of free fatty acid pools in Escherichia coli. Proc Natl Acad Sci USA. 2013;110:7636–41.View ArticleGoogle Scholar
  34. Zuurbier KWM, Fung S-Y, Scheffer JJC, Verpoorte R. Formation of aromatic intermediates in the biosynthesis of bitter acids in Humulus lupulus. Phytochemistry. 1995;38:77–82.View ArticleGoogle Scholar
  35. Burns MJ, Ronson TO, Taylor RJ, Fairlamb IJ. 4-Hydroxy-6-alkyl-2-pyrones as nucleophilic coupling partners in Mitsunobu reactions and oxa-Michael additions. Beilstein J Org Chem. 2014;10:1159–65.View ArticleGoogle Scholar
  36. Jez JM, Austin MB, Ferrer J, Bowman ME, Schroder J, Noel JP. Structural control of polyketide formation in plant-specific polyketide synthases. Chem Biol. 2000;7:919–30.View ArticleGoogle Scholar
  37. Abe I, Utsumi Y, Oguro S, Morita H, Sano Y, Noguchi H. A plant type III polyketide synthase that produces pentaketide chromone. J Am Chem Soc. 2005;127:1362–3.View ArticleGoogle Scholar
  38. Zha WJ, Rubin-Pitel SB, Shao ZY, Zhao HM. Improving cellular malonyl-CoA level in Escherichia coli via metabolic engineering. Metab Eng. 2009;11:192–8.View ArticleGoogle Scholar


© The Author(s) 2016