Efficient Bioconversion of Echinocandin B to Its Nucleus byOverexpression of Deacylase Genes in Different Host Strains

Lei Shao,a Jian Li,a Aijuan Liu,a Qing Chang,a,b Huimin Lin,a Daijie Chena

State Key Lab of New Drugs and Pharmaceutical Process, Shanghai Institute of Pharmaceutical Industry, Shanghai, Chinaa; College of Life Sciences, East China NormalUniversity, Shanghai, Chinab

Anidulafungin, which noncompetitively inhibits �-(1,3)-D-glucan synthase in fungal cell wall biosynthesis, is the newest anti-fungal drug to be developed. Echinocandin B deacylase from Actinoplanes utahensis NRRL 12052 catalyzes the cleavage of thelinoleoyl group of echinocandin B, a key step in the process of manufacturing anidulafungin. Unfortunately, the natural yield ofechinocandin B nucleus is low. In our study, the echinocandin B deacylase gene was systematically overexpressed by genetic en-gineering in its original producer, A. utahensis, and in the heterologous hosts Streptomyces lividans TK24 and Streptomyces al-bus. The introduction of additional copies of the gene, under the control of PermE* or its native promoter, into hosts showedsignificant increases in its transcription level and in the efficiency of the bioconversion of echinocandin B to its nucleus. The con-ditions for the cultivation and bioconversion of A. utahensis have been optimized further to improve production. As a result,while the wild-type strain initially produced 0.36 g/liter, a concentration of 4.21 g/liter was obtained after the generation of astrain with additional copies of the gene and further optimization of the reaction conditions. These results are useful for enhanc-ing echinocandin B nucleus production in A. utahensis. Our study could enable the engineering of commercially useful echino-candin B nucleus-overproducing stains.

Fungal infections are being seen in ever-increasing numbers,largely because of the increase in the size of the population at

risk over the past 20 years. This population includes cancer pa-tients, transplant recipients, and other individuals receiving im-munosuppressive treatment. These people are at greater risk thanothers owing to their weakened immune systems and the chronicnature of diseases (1–3). The increased incidence of invasive fun-gal infections has created a major challenge for health care profes-sionals. Since cell walls are present in fungal cells but absent inanimal cells, the fungal cell wall perhaps represents the ideal targetfor the therapeutic treatment of fungal pathogens in humans (4,5). The development of echinocandins, the first class of antifun-gals to target the fungal cell wall, was a milestone in antifungalchemotherapy (6, 7). Three semisynthetic echinocandin deriva-tives have been developed for clinical use: caspofungin, micafun-gin, and anidulafungin (8). Their strengths include low toxicity,rapid fungicidal activity against most isolates of Candida spp., andpredictable, favorable kinetics allowing once-a-day dosing. BesidesCandida spp., their inhibitory spectrum includes Aspergillus spp. andPneumocystis jirovecii, but not Cryptococcus neoformans (9, 10).

Echinocandin B (ECB), obtained by the fermentation of Asper-gillus nidulans and Aspergillus rugulosus, is known as one of thenatural cyclic hexapeptides that have a linoleoyl side chain, whichinhibits a crucial enzyme in fungal cell wall biosynthesis, �-(1,3)-D-glucan synthase (11). ECB can be modified by enzymatic deacy-lation to a cyclic hexapeptide without a linoleoyl side chain and bysubsequent chemical reacylation to generate a few therapeutic an-tifungal agents for clinical practice, such as anidulafungin (12–15). A deacylase from Actinoplanes utahensis NRRL 12052 cata-lyzes the cleavage of the linoleoyl side chain from ECB (Fig. 1), anessential reaction for the three subsequent synthetic steps (16). Theenzyme is a membrane-associated heterodimer composed of 63-kDaand 18- to-20-kDa subunits, and the expression of its activity is notaffected by any cofactors, metal ion chelators, or reducing agents. Inaddition to that of ECB, this deacylase mediates the cleavage of acu-

leacin A, FR901379, various semisynthetic ECB derivatives, dapto-mycin and its three derivatives, teicoplanin, pseudomycin A, and cap-saicins (17, 18). Thus, it may become increasingly significant as apharmaceutical biocatalyst.

However, enzymatic deacylation was rate-limiting when con-ducted with whole cells of A. utahensis. The low bioconversionyield was presumably related to inadequate production of the ECBdeacylase. Because of the need for improved antifungal agents for thetreatment of systemic fungal infections and the broad specificity ofthe enzyme, we were interested in constructing engineered strainswith high production of the enzyme and in developing a better bio-conversion method to further improve the efficiency of bioconver-sion of ECB. Here we describe the cloning of the gene encoding ECBdeacylase, under the control of its native promoter or a PermE* pro-moter, into the original deacylase-producing strain and two Strepto-myces strains by �C31-directed site-specific recombination in orderto understand the effects of the promoters and gene dosage on theefficiency of the bioconversion of ECB to the ECB nucleus, particu-larly with regard to its potential biotechnological application.

MATERIALS AND METHODS

Bacterial strains, plasmids, and reagents. The bacterial strains and plas-mids used in this paper are listed in Table 1. Streptomyces lividans TK24,Streptomyces albus, and A. utahensis NRRL 12052 were obtained from our

Received 10 September 2012 Accepted 20 November 2012

Published ahead of print 7 December 2012

Address correspondence to Lei Shao, shaolei00@gmail.com, or Daijie Chen,chen_daijie@yahoo.com.

L.S., J.L., and A.L. contributed equally to this work.

Copyright © 2013, American Society for Microbiology. All Rights Reserved.

doi:10.1128/AEM.02792-12

1126 aem.asm.org Applied and Environmental Microbiology p. 1126–1133 February 2013 Volume 79 Number 4

on March 23, 2018 by guest

http://aem.asm

.org/D

ownloaded from

laboratory. Biochemicals, chemicals, media, restriction enzymes, and othermolecular biological reagents were from standard commercial sources.

DNA isolation, manipulation, and sequencing. DNA isolation andmanipulation were performed by standard methods (21). PCR amplifica-tions were conducted on an authorized Thermal Cycler (Eppendorf AG,Hamburg, Germany) using PrimerSTAR HS DNA polymerase (TaKaRa).Primer synthesis and DNA sequencing were carried out at Shanghai In-vitrogen Biotechnology Co.

Plasmid construction. To express the ECB deacylase gene under thecontrol of a PermE* promoter, a 2.8-kb DNA fragment that contains onlythe gene encoding ECB deacylase was amplified by PCR from A. utahensisNRRL 12052 genomic DNA using primers 5=-AAAGAATTCGTGCGGGCCTGAAA-3= and 5=-AAATCTAGAGACTGCGTGAGTTCTGC-3= andwas cloned into the pSP72 vector, yielding pYG2001. The identity of thePCR product with the gene encoding ECB deacylase (GenBank accessionnumber BD226911) was confirmed by sequencing. A 0.5-kb fragment

containing a PermE* promoter was PCR amplified from pYG1003 usingprimers 5=-AAAAGATCTTCTAGAAGCCCGACCCGAGCA-3= and 5=-AAAGAATTCTCCGGAGGTCGCACC-3= and was cloned into the BglII/EcoRI site of pYG2001, yielding pYG2002. (Underlined letters in primersequences represent restriction sites.) The 3.3-kb XbaI/XbaI fragmentcontaining the gene coding for ECB deacylase and a PermE* promoter frompYG2002 was inserted into the XbaI site of the pSET152 vector, yieldingpYG2003 for the expression of the gene encoding ECB deacylase under thecontrol of a PermE* promoter by a �C31-directed site-specific recombina-tion event.

For the expression of the gene encoding ECB deacylase under thecontrol of its native promoter, a 4.0-kb DNA fragment that contains thegene encoding ECB deacylase with approximately 1 kb of upstream se-quence and 0.5 kb of downstream sequence was amplified from A. uta-hensis NRRL 12052 genomic DNA by PCR using primers 5=-ATAGAATTCCGTGCCCAGCTGTTC-3= and 5=-AAATCTAGAGACTGCGTGAGT

FIG 1 ECB deacylase-catalyzed reaction.

TABLE 1 Bacterial strains and plasmids used in this study

Strain or plasmid DescriptionSource orreference

StrainsE. coli

DH5� Host for general cloning InvitrogenET12567(pUZ8002) Donor strain for conjugation between E. coli and Actinomyces 3

A. utahensis ECB deacylase-producing strain NRRL 12052DYG2001 A. utahensis derivative with two copies of the ECB deacylase gene (second copy under the control of PermE*) This workDYG2002 A. utahensis derivative with two copies of the ECB deacylase gene under the control of its native promoter This workDYG2003 A. utahensis derivative with three copies of the ECB deacylase gene (second and third copies under the

control of PermE*)This work

S. lividans TK24 19DYG2004 S. lividans derivative with a single copy of the ECB deacylase gene under the control of PermE* This workDYG2005 S. lividans derivative with a single copy of the ECB deacylase gene under the control of its native promoter This workDYG2006 S. lividans derivative with two copies of the ECB deacylase gene under the control of PermE* This study

S. albus 19DYG2007 S. albus derivative with a single copy of the ECB deacylase gene under the control of PermE* This workDYG2008 S. albus derivative with a single copy of the ECB deacylase gene under the control of its native promoter This workDYG2009 S. albus derivative with two copies of the ECB deacylase gene under the control of PermE* This work

PlasmidspSP72 E. coli subcloning vector PromegapSET152 E. coli replicon, Streptomyces �C31 attachment site; Aprr 16pYG1003 2.0-kb fragment containing PermE*-controlled aveBIV in pSET152 20pYG2001 2.8-kb PCR fragment containing the ECB deacylase gene in pSP72 This workpYG2002 3.3-kb PCR fragment containing a PermE*-controlled ECB deacylase gene in pSP72 This workpYG2003 3.3-kb PCR fragment containing a PermE*-controlled ECB deacylase gene in pSET152 This workpYG2004 4.0-kb PCR fragment containing the ECB deacylase gene and its 5= and 3= regulatory sequences in pSP72 This workpYG2005 4.0-kb PCR fragment containing the ECB deacylase gene and its 5= and 3= regulatory sequences in pSET152 This workpYG2006 3.3-kb PCR fragment containing a PermE*-controlled ECB deacylase gene in pSP72 This workpYG2007 Two copies of a 3.3-kb PCR fragment containing a PermE*-controlled ECB deacylase gene in pYG2003 This work

ECB Bioconversion by Overexpression of Deacylase

February 2013 Volume 79 Number 4 aem.asm.org 1127

on March 23, 2018 by guest

http://aem.asm

.org/D

ownloaded from

TCTGC-3= and was cloned into the pSP72 vector, yielding pYG2004. Theidentity of the PCR product with the gene encoding ECB deacylase wasalso confirmed by sequencing. The 4.0-kb EcoRI/XbaI fragment frompYG2004 was inserted into the corresponding sites of pSET152, yieldingpYG2005.

In order to express two copies of the gene encoding ECB deacylaseunder the control of a PermE* promoter, another 3.3-kb DNA fragmentwith terminal EcoRV/EcoRV sites containing the gene encoding ECB de-acylase and a PermE* promoter was amplified by PCR from pYG2002using primers 5=-AAAGATATCAGCCCGACCCGAGCA-3= and 5=-AAAGATATCGACTGCGTGAGTTCTGC-3= and was cloned into the pSP72vector, yielding pYG2006. The identity of the PCR product with the geneencoding ECB deacylase was confirmed. The 3.3-kb EcoRV/EcoRV frag-ment containing the gene encoding ECB deacylase and a PermE* pro-moter from pYG2006 was inserted into the corresponding site of the re-combinant vector pYG2003, yielding the last vector, pYG2007, containingtwo deacylase gene expression cassettes. The correct directions of the geneencoding ECB deacylase and the PermE* promoter in the pYG2003 orpYG2007 vector were verified by restriction enzyme digestion.

Conjugation and generation of recombinant strains. For overex-pression and heterologous expression of the ECB deacylase gene in A.utahensis NRRL 12052 and the two Streptomyces strains (S. lividans TK24and the S. albus strain), expression vectors pYG2003, pYG2005, andpYG2007 were each introduced into hosts by intergeneric conjugationfrom Escherichia coli ET12567(pUZ8002) according to the standard pro-cedure (19, 22). Transformants that were resistant to apramycin wereidentified as the recombinant strains, whose genomic DNAs were inte-grated with the deacylase gene and the apramycin resistance gene by�C31-directed site-specific recombination. The genotypes of the recom-binant strains were further confirmed by PCR amplification with the vec-tor-specific primer pair M13-47 and RV-M.

Culture growth and deacylation procedure. Wild-type and recombi-nant strains were grown on agar plates with a medium consisting of 2%soluble starch, 0.05% NaCl, 0.05% K2HPO4·3H2O, 0.1% KNO3, 0.05%MgSO4·7H2O, 0.001% FeSO4·7H2O, and 2% agar powder (pH 7.4) at28°C for sporulation. For the fermentation of A. utahensis NRRL 12052,an agar piece around 1 cm2 was inoculated into a 250-ml flask containing50 ml of a seed medium consisting of 2.5% sucrose, 2.0% oatmeal, 0.25%yeast powder, 0.1% K2HPO4, 0.05% KCl, 0.05% MgSO4·7H2O, and0.0002% FeSO4·7H2O and was incubated at 28°C and 220 rpm for 3 days.A 250-ml flask containing 50 ml of fresh fermentation medium, consistingof 2% sucrose, 1% peanut meal, 0.1% KH2PO4, and 0.025%MgSO4·7H2O, was then inoculated with 5 ml of the seed culture, andincubation was continued at 28°C and 220 rpm for 4 days. For the fer-mentation of S. lividans TK24 and S. albus, an agar piece around 1 cm2 wasinoculated into a 250-ml flask containing 50 ml of a seed medium con-sisting of 1.0% glucose, 0.5% yeast powder, and 1% peptone and wasincubated at 30°C and 220 rpm for 30 h. A 250-ml flask containing 50 mlof fresh fermentation medium consisting of 2.5% glucose, 1% bean flour,0.3% NaCl, and 0.3% CaCO3 was then inoculated with 1 ml of the seedculture, and incubation was continued at 30°C and 220 rpm for 2 days.

The wet mycelia (15 g) were harvested by centrifugation, washed twicewith double-distilled water, and then resuspended in a 50-ml biotransfor-mation reaction mixture containing 0.1 M Na2HPO4/NaH2PO4 buffer(pH 6.8) in the presence of 2% dimethyl sulfoxide (DMSO). The enzy-matic reaction was initiated by addition of the substrate ECB (1.5 g/liter)and was continued at 30°C and 220 rpm for 5 h before sample analysis.

Different experiments were subsequently conducted to optimize thecultivation medium of A. utahensis by changing the carbon and nitrogensources and their proportions. Maltose, sucrose, glycerol, lactose, andsoybean oil were used individually as the carbon source, at concentrationsof 20 g/liter, instead of glucose. Cottonseed meal, soybean powder, andoatmeal were used individually as the nitrogen source, at concentrationsof 10 g/liter, instead of peanut meal. The optimum concentration of su-crose or cottonseed meal was determined by dosing different amounts of

sucrose (2 to 6%) or cottonseed meal (1 to 5%). The optimum pH forbiotransformation was determined at 30°C. The optimum temperaturefor biotransformation was determined at pH 6.8. Ethanol was used as thecosolvent instead of DMSO. The optimum concentration of the ECB sub-strate was determined by dosing different amounts of ECB (1.5 to 12g/liter) under the optimized conditions.

Analytical methods for the detection of ECB and the ECB nucleus.The reaction was interrupted by the addition of methanol. After low-speed centrifugation to remove precipitated proteins and mycelia, theactivity of the deacylase was determined by monitoring the formation ofthe cyclic hexapeptide (the ECB nucleus) by high-performance liquidchromatography (HPLC) with an Agilent C18 column (250 by 4.6 mm;Agilent, Palo Alto, CA). The column was equilibrated with 92% solvent A(H2O containing 2 mg/ml ammonium acetate) and 8% solvent B (60%CH3CN containing 2 mg/ml ammonium acetate), and the analytical methodwas developed with the following program: from 0 min to 15 min, a lineargradient from 92% A–8% B to 2% A–98% B; from 15 min to 25 min, a lineargradient from 2% A–98% B to 8% A–92% B; and from 25 min to 32 min, aconstant proportion of 92% A–8% B. HPLC was carried out at a flow rate of0.8 ml/min with UV detection at 222 nm using an Agilent 1120 HPLC system(Agilent Technologies, Palo Alto, CA). The identities of compounds wereconfirmed by liquid chromatography-mass spectrometry (LC-MS) analysisperformed on an LCMS-2010A system (Shimadzu, Japan).

Assay of transcript levels by RT-PCR amplification. Total RNAs ofthree wild-type strains and nine recombinant strains were isolated frommycelia in the fermentation cultures. An additional purification step wascarried out by using TRIzol reagent (catalog no. 15596-026; Invitrogen).To obtain cDNAs, DNase treatment and reverse transcription were per-formed by using random primers (catalog no. C1181; Promega) andMoloney murine leukemia virus (M-MLV) reverse transcriptase (catalogno. M1701; Promega) according to the manufacturer’s instructions. Thetranscript levels of the ECB deacylase gene were assayed on a Rotor-GeneRG-3000A instrument (Corbett Research Co., Australia), using real-timequantitative PCR and the 2���CT method (23). Reverse transcription-PCR (RT-PCR) amplification was performed on each 25 �l of the mixture(consisting of 1 �g/ml of template cDNA, 2� SYBR Premix [ShanghaiXinghan Sci-Tech Co., Shanghai, China] and 0.4 �M forward and reverseprimers) with the following program: 95°C for 3 min, followed by 30cycles of 94°C for 20 s, 60°C for 20 s, and 72°C for 15 s. For assessment ofthe transcript level of the ECB deacylase gene, the 244-bp internal frag-ment was amplified from cDNA of the recombinant strains and wild-typeA. utahensis NRRL 12052 using primers 5=-CGCATGTACGAGGACGTCAC-3= and 5=-GCACAGCGGGGACGCTGA-3=. For assessment of thetranscript level of the 16S rRNA gene (as a control), the 242-bp internalfragments were amplified from cDNA of S. lividans TK24 and its recom-binant strains using primers 5=-GCAATCTGCCCTTCACTCTG-3= and5=-TATTCCCCACTGCTGCCTCC-3=; the 285-bp internal fragmentswere amplified from cDNA of S. albus and its recombinant strains usingprimers 5=-GCCGATACTGACGCTGAGGA-3= and 5=-GGACCCTGTCTCCAGAGTTTTC-3=; and the 255-bp internal fragments were amplifiedfrom cDNA of A. utahensis NRRL 12052 and its recombinant strains usingprimers 5=-GGAGGCAGCAGTGGGGAAT-3= and 5=-TGAGCCTCGGGATTTCACATT-3=.

RESULTS AND DISCUSSIONQualitative and quantitative analysis of ECB and the ECB nu-cleus. ECB and its nucleus were extracted from a bioconversionmixture, and their identities were determined according to themethod described above. As shown in Fig. 2, analysis of the sam-ples revealed that the product and the remaining substrate ECBhad the same retention times as those of the standard ECB nucleus(6.75 min) and ECB (22.80 min), respectively. Their identitieswere further confirmed by electrospray ionization-quadrupoletime of flight (ESI–Q-TOF)–MS analysis, showing (M � Na)�

ions at m/z 820.40, consistent with the molecular formula

Shao et al.

1128 aem.asm.org Applied and Environmental Microbiology

on March 23, 2018 by guest

http://aem.asm

.org/D

ownloaded from

C34H51N7O15 for the ECB nucleus, and (M � Na)� ions at m/z1,082.61, consistent with the molecular formula C52H81N7O16 forECB. The concentration of the ECB nucleus in the reaction mix-ture was calculated according to the standard curve of the refer-ence ECB nucleus. The methods described here for qualitative andquantitative analysis of ECB and the ECB nucleus were applied to

the biotransformation reaction mixture of A. utahensis, S. lividans,S. albus, and the recombinant strains obtained in this study.

Overexpression of the ECB deacylase gene in A. utahensisNRRL 12052. Enzymes can be overproduced by increasing thegene copy number. To investigate the effect of deacylase gene dos-age, pYG2003 and pYG2007, which contain one and two deacylasegene expression cassettes, respectively, under the control of aPermE* promoter, were constructed in pSET152 and were intro-duced into A. utahensis NRRL 12052 by a �C31-directed site-specificrecombination event, yielding recombinant strains DYG2001, con-taining two copies of the deacylase gene, and DYG2003, containingthree copies of the deacylase gene. To express deacylase under thecontrol of the native promoter, pTG2005 was constructed inpSET152 and was introduced into A. utahensis NRRL 12052, yieldingrecombinant strain DYG2002, which contained two copies of thedeacylase gene (Fig. 3A). The 3.3-kb, 4.0-kb, and 6.6-kb fragmentscould be amplified separately from genomic DNAs of the recombi-nant strains mentioned above by PCR using primers M13-47 andRV-M. Sequencing showed that these recombinant strains had thedesigned genotypes (Fig. 4). Besides this, there were no apparent dif-ferences in growth characteristics or morphology.

Based on these results, we determined the bioconversion effi-ciency of each recombinant strain. As shown in Fig. 5, the effi-ciency of bioconversion of ECB was doubled in DYG2001 andDYG2002. These observations may mean that insufficient ECBdeacylase activity was present in vivo. This idea is supported byother results. Even higher bioconversion efficiency was observed forDYG2003, containing three copies of the deacylase gene (Fig. 5). In

FIG 2 HPLC analysis of ECB nucleus production in a reaction system withwhole cells of control strain A. utahensis NRRL 12052. Standard ECB (p) andECB nucleus (�) were used as controls. (I) Standard ECB; (II) standard ECBnucleus; (III) A. utahensis NRRL 12052.

FIG 3 Genotypes and phenotypes of the recombinant strains. (A) Predicted fragment sizes for PCR analysis. Each wavy line indicates the DNA fragment of thevector integrated by �C31-directed site-specific recombination. (B) HPLC analysis of ECB nucleus production in a reaction system with ECB (p) and the ECBnucleus (�).

ECB Bioconversion by Overexpression of Deacylase

February 2013 Volume 79 Number 4 aem.asm.org 1129

on March 23, 2018 by guest

http://aem.asm

.org/D

ownloaded from

our study, we found a decline in the level of the ECB substratein the control system containing only the substrate and thebioconversion buffer, while the ECB nucleus product could beaccumulated stably. In the A. utahensis NRRL 12052 biocon-version system, we found both the ECB nucleus product andthe ECB substrate, while no visible ECB substrate was detectedin the three recombinant strains upon HPLC analysis, as shownin Fig. 3B.

Another interesting question is the means of integrating thedeacylase gene into A. utahensis and Streptomyces. As has beenobserved previously with 4=-epidaunorubicin production inStreptomyces coeruleorubidus (20), the ECB deacylase gene couldbe stably integrated into the genomes of host strains by a �C31-directed site-specific recombination event. Thus, it is not neces-sary to use a culture medium with apramycin resistance for the

fermentation of the recombinant strains. Such a medium withoutapramycin helps cut the cost of ECB nucleus production on anindustrial scale.

Heterologous expression of the ECB deacylase gene con-trolled by different promoters. Streptomyces spp. are well-knownproducers of biologically active compounds and enzymes (24–26). Techniques for the cultivation of these organisms on an in-dustrial scale are well established, so these species are especiallyattractive as hosts for heterologous products (27). Different spe-cies of Streptomyces were screened as heterologous expressionhosts by using the same bioconversion method as A. utahensis. Forthe two species examined, S. lividans (strain TK24) and S. albus, noECB nucleus or ECB from whole cells was detected in the reactionmixtures. The PermE* promoter has been used frequently as astrong constitutive promoter for native and heterologous genes in

FIG 4 PCR analysis with genomic DNA from S. lividans TK24, S. albus, A. utahensis NRRL 12052, or mutants as the template. By use of primers M13-47(5=-CGCCAGGGTTTTCCCAGTCACGAC-3=) and RV-M (5=-GAGCGGATAACAATTTCACACAGG-3=), the genotypes of the recombinant strains were con-firmed by PCR to be well in agreement with their designed genetic patterns, demonstrating that the �C31-directed site-specific recombination event had takenplace. Lane 1, no product for wild-type A. utahensis; lane 2, 3.3-kb product for DYG2001; lane 3, 4-kb product for DYG2002; lane 4, 6.6-kb product for DYG2003;lane 5, no product for wild-type S. lividans TK24; lane 6, 3.3-kb product for DYG2004; lane 7, 4-kb product for DYG2005; lane 8, 6.6-kb product for DYG2006;lane 9, no product for wild-type S. albus; lane 10, 3.3-kb product for DYG2007; lane 11, 4.0-kb product for DYG2008; lane 12, 6.6-kb product for DYG2003. AllPCR products were sequenced to confirm each mutation.

FIG 5 Production of the ECB nucleus and ratio of the target gene transcript level to the transcript level of 16S rRNA, used as a control (relative mRNA), inreaction systems with whole cells of the different strains. Error bars represent standard deviations. For each strain, the experiment was carried out independentlythree times.

Shao et al.

1130 aem.asm.org Applied and Environmental Microbiology

on March 23, 2018 by guest

http://aem.asm

.org/D

ownloaded from

Streptomyces species and related bacteria (28). To compare theexpression levels of the ECB deacylase gene under the control of itsnative promoter and under the control of a PermE* promoter, thegenes were cloned into pSET152 and were then transformed intothe two Streptomyces species by a �C31-directed site-specific re-combination event to generate four new strains: DYG2004,DYG2005, DYG2007, and DYG2008. Another expression plas-mid, pYG2007, with two ECB deacylase gene expression cassettescontrolled by PermE*, was constructed and was introduced into S.lividans TK24 and S. albus, respectively, to yield DYG2006 andDYG2009, containing two copies of the deacylase gene (Fig. 3A).The genotypes of these recombinant strains were verified by PCRamplification, whereas no PCR products were isolated from thewild-type strains, as expected (Fig. 4). Subsequently, the biocon-version efficiencies of these recombinant strains were determined,and each reaction mixture was analyzed by HPLC. HPLC analysisrevealed that the bioconversion efficiencies of these recombinantstrains were enhanced over that of the original strain, A. utahensisNRRL 12052. More importantly, the bioconversion efficiencies ofthe recombinant strains DYG2004 and DYG2007, containingPermE*-controlled ECB deacylase, were higher than those ofstrains DYG2005 and DYG2008, in which the ECB deacylase genewas controlled by its native promoter. Strains DYG2006 andDYG2009, containing two copies of the deacylase gene, havehigher bioconversion efficiencies than strains with one deacylase

gene copy (Fig. 5). These findings demonstrated that PermE* wassuperior to the native promoter in the production of ECB deacyl-ase for heterologous expression or overexpression and that therewas potentially an enhancement of bioconversion efficiency andan improvement in the production of ECB deacylase due to theincrease in gene copy number. As has been observed previously,the Streptomyces genome may contain multiple pseudo-attB sites.However, the frequency of integration of pSET152 into the �attBstrain with additional pseudo-attB sites is reduced approximately300-fold (29). In our study, there are strong positive correlationsbetween the gene copy numbers of the recombinant strains andthe transcript levels of the ECB deacylase gene. Therefore, theremay be only one integrated gene. Of all the recombinantstrains, DYG2003 from A. utahensis NRRL 12052 showed thehighest bioconversion efficiency. A possible reason is that theECB deacylase gene can be expressed better in A. utahensisNRRL 12052 and that DYG2003 contains more deacylase genecopies than NRRL 12052.

Optimal conditions for biotransformation of A. utahensis.For the cultivation of A. utahensis and its recombinant strains, thenew fermentation medium formulation based on 3.0% sucrose,2.0% cottonseed meal, 0.12% K2HPO4·3H2O, and 0.05%KH2PO4·3H2O was a good alternative solution for the pigmenta-tion of strain DYG2003 (unpublished data). As a result, thegrowth rate and mycelium volume of A. utahensis and its recom-

FIG 6 Effects of temperature, pH, and the concentration of the ECB substrate on the biotransformation reaction. (A) Temperature course of the biotransfor-mation reaction at pH 6.8. (B) pH course of the biotransformation reaction at 30°C. (C) Effects of different concentrations of the ECB substrate under optimizedconditions (pH 4 and 25°C). Error bars represent standard deviations.

ECB Bioconversion by Overexpression of Deacylase

February 2013 Volume 79 Number 4 aem.asm.org 1131

on March 23, 2018 by guest

http://aem.asm

.org/D

ownloaded from

binant strains increased dramatically. Under the new culture con-ditions, the concentration of mycelium in the fermentation brothwas 30% higher than that under the original conditions in thesame fermentation time. For the biotransformation reaction, theoptimal temperature was 25°C and the optimal pH was 4.5 (Fig.6A and B). The optimum concentration of ECB in the reactionsystem was detected by additions of different ECB quantities un-der the optimized conditions (pH 4 and 25°C) for further evalu-ation of the bioconversion efficiency of the recombinant strainDYG2003 and the original strain A. utahensis NRRL 12052. Withwhole cells of NRRL 12052, the production of the ECB nucleus inreaction mixtures could rise as high as 4.21 g/liter when a suitableconcentration of ECB (8 g/liter) was added to the reaction system(Fig. 6C). Additionally, ethanol is more economical and beneficialthan DMSO as the cosolvent used in the reaction system to sim-plify the separation and purification of the product.

Assay of the transcript levels of the gene by RT-PCR amplifi-cation. To assess the effects of different promoters and gene doseson the increase in the amount of enzyme produced, RT-PCR am-plification for analysis of the transcription levels of the ECB deac-ylase gene was performed on the total RNAs extracted from cells ofeach strain investigated in this study. The transcript levels of thegene fragment that encodes 16S rRNA served as a control. Asshown in Fig. 5, the introduction of one or two additional copiesof the ECB deacylase gene caused significant increases in the tran-script level of the gene. As expected, no ECB deacylase mRNA wasdetected for wild-type S. lividans and S. albus. The ratios of thetranscript levels of the ECB deacylase gene under the control ofPermE* in the S. lividans and S. albus heterologous expressionhosts to the transcript level of the control 16S rRNA gene were ashigh as 0.66 and 0.87, respectively, while those of the gene con-trolled by its native promoter were as high as 0.46 and 0.84, re-spectively. Duplication and triplication of the gene resulted inincreases as high as 2.00-fold and 6.94-fold in A. utahensis NRRL12052. These results indicate that increasing the copy numberof the gene encoding ECB deacylase leads to a significant in-crease in the transcript level of this gene, especially when it iscontrolled by the PermE* promoter. The transcript levels of thegene meet the expectation of increased efficiency of bioconver-sion of ECB to the ECB nucleus in recombinant strains (men-tioned above).

The deacylation of ECB is the starting point for the semi-synthesis of anidulafungin. The current annual production ofanidulafungin is on the order of tons, and anidulafungin iswidely used as an important anti-infective drug. Synthesis ofthe ECB nucleus by chemical deacylation is extremely difficult,and the biotransformation method described here avoids thisproblem. The findings of this study could enable the engineer-ing of commercially useful ECB nucleus-overproducingstrains.

ACKNOWLEDGMENTS

We thank Jia Zeng and S. Gabrielle Gladstone, Department of BiologicalEngineering, Utah State University, who helped with editing.

This work was supported in part by grants from the National Nat-ural Science Foundation of China (grants 81172962, 30801449, and81072557), the Science and Technology Commission of Shanghai Munic-ipality (grant 11QB1406300), and the Ministry of Science and Technologyof China (grant 2009ZX09301-007).

REFERENCES1. Clark TA, Hajjeh RA. 2002. Recent trends in the epidemiology of invasive

mycoses. Curr. Opin. Infect. Dis. 15:569 –574.2. Kriengkauykiat J, Ito JI, Dadwal SS. 2011. Epidemiology and treatment

approaches in management of invasive fungal infections. Clin. Epidemiol.3:175–191.

3. Maertens J, Vrebos M, Boogaerts M. 2001. Assessing risk factors forsystemic fungal infections. Eur. J. Cancer Care 10:56 – 62.

4. Denning DW. 2003. Echinocandin antifungal drugs. Lancet 362:1142–1151.

5. Georgopapadakou NH, Tkacz JS. 1995. The fungal cell wall as a drugtarget. Trends Microbiol. 3:98 –104.

6. Sucher AJ, Chahine EB, Balcer HE. 2009. Echinocandins: the newestclass of antifungals. Ann. Pharmacother. 43:1647–1657.

7. Wagner C, Graninger W, Presterl E, Joukhadar C. 2006. The echino-candins: comparison of their pharmacokinetics, pharmacodynamics andclinical applications. Pharmacology 78:161–177.

8. Cleary JD. 2009. Echinocandins: pharmacokinetic and therapeutic issues.Curr. Med. Res. Opin. 25:1741–1750.

9. Denning DW. 2002. Echinocandins: a new class of antifungal. J. Antimi-crob. Chemother. 49:889 – 891.

10. Fera MT, La Camera E, De Sarro A. 2009. New triazoles and echinocan-dins: mode of action, in vitro activity and mechanisms of resistance. Ex-pert Rev. Anti Infect. Ther. 7:981–998.

11. Nyfeler R, Keller-Schierlein W. 1974. Metabolites of microorganisms.143. Echinocandin B, a novel polypeptide-antibiotic from Aspergillus ni-dulans var. echinulatus: isolation and structural components. Helv. Chim.Acta 57:2459 –2477. (In German.)

12. Debono M, Abbott BJ, Turner JR, Howard LC, Gordee RS, Hunt AS,Barnhart M, Molloy RM, Willard KE, Fukuda D, Butler TF, ZecknerDJ. 1988. Synthesis and evaluation of LY121019, a member of a series ofsemisynthetic analogues of the antifungal lipopeptide echinocandin Ba.Ann. N. Y. Acad. Sci. 544:152–167.

13. Onishi J, Meinz M, Thompson J, Curotto J, Dreikorn S, Rosenbach M,Douglas C, Abruzzo G, Flattery A, Kong L, Cabello A, Vicente F, PelaezF, Diez MT, Martin I, Bills G, Giacobbe R, Dombrowski A, Schwartz R,Morris S, Harris G, Tsipouras A, Wilson K, Kurtz MB. 2000. Discoveryof novel antifungal (1,3)-�-D-glucan synthase inhibitors. Antimicrob.Agents Chemother. 44:368 –377.

14. Petraitiene R, Petraitis V, Groll AH, Candelario M, Sein T, Bell A,Lyman CA, McMillian CL, Bacher J, Walsh TJ. 1999. Antifungalactivity of LY303366, a novel echinocandin B, in experimental dissem-inated candidiasis in rabbits. Antimicrob. Agents Chemother. 43:2148 –2155.

15. Uzun O, Kocagöz S, Cetinkaya Y, Arikan S, Unal S. 1997. In vitroactivity of a new echinocandin, LY303366, compared with those of am-photericin B and fluconazole against clinical yeast isolates. Antimicrob.Agents Chemother. 41:1156 –1157.

16. Boeck LD, Fukuda DS, Abbott BJ, Debono M. 1989. Deacylation ofechinocandin B by Actinoplanes utahensis. J. Antibiot. 42:382–388.

17. Kreuzman AJ, Hodges RL, Swartling JR, Pohl TE, Ghag SK, Baker PJ,McGilvray D, Yeh WK. 2000. Membrane-associated echinocandin Bdeacylase of Actinoplanes utahensis: purification, characterization, heter-ologous cloning and enzymatic deacylation reaction. J. Ind. Microbiol.Biotechnol. 24:173–180.

18. Romano D, Gandolfi R, Guglielmetti S, Molinari F. 2011. Enzymatichydrolysis of capsaicins for the production of vanillylamine using ECBdeacylase from Actinoplanes utahensis. Food Chem. 124:1096 –1098.

19. Kieser T, Bibb MJ, Buttner MJ, Chater KF, Hopwood DA. 2000.Practical Streptomyces genetics. John Innes Foundation, Norwich, UnitedKingdom.

20. Shao L, Huang J, Jing L, Chen JY, Kan SD, Wang M, Li JA, Chen DJ.2010. Overexpression of aveBIV leading to the improvement of 4=-epidaunorubicin production in Streptomyces coeruleorubidus strainSIPI-A0707. Appl. Microbiol. Biotechnol. 87:1057–1064.

21. Sambrook J, Russell DW. 2001. Molecular cloning: a laboratory manual,3rd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

22. Luzhetskii AN, Ostash BE, Fedorenko VA. 2001. Intergeneric conju-gation Escherichia coli–Streptomyces globisporus 1912 using integrativeplasmid pSET152 and its derivatives. Russ. J. Genet. 37:1123–1129.

23. Pfaffl MW. 2001. A new mathematical model for relative quantification inreal-time RT-PCR. Nucleic Acids Res. 29:e45. doi:10.1093/nar/29.9.e45.

Shao et al.

1132 aem.asm.org Applied and Environmental Microbiology

on March 23, 2018 by guest

http://aem.asm

.org/D

ownloaded from

24. Ding W, Lei C, He Q, Zhang Q, Bi Y, Liu W. 2010. Insights into bacterial6-methylsalicylic acid synthase and its engineering to orsellinic acid syn-thase for spirotetronate generation. Chem. Biol. 17:495–503.

25. Shao L, Qu XD, Jia XY, Zhao QF, Tian ZH, Wang M, Tang GL, Liu W.2006. Cloning and characterization of a bacterial iterative type I polyketidesynthase gene encoding the 6-methylsalicyclic acid synthase. Biochem.Biophys. Res. Commun. 345:133–139.

26. Ueda S, Shibata T, Ito K, Oohata N, Yamashita M, Hino M, Yamada M,Isogai Y, Hashimoto S. 2011. Cloning and expression of the FR901379acylase gene from Streptomyces sp. no. 6907. J. Antibiot. 64:169 –175.

27. Inokoshi J, Takeshima H, Ikeda H, O� mura S. 1993. Efficient productionof aculeacin A acylase in recombinant Streptomyces strains. Appl. Micro-biol. Biotechnol. 39:532–536.

28. Pan HX, Li JA, He NJ, Chen JY, Zhou YM, Shao L, Chen DJ. 2011.Improvement of spinosad production by overexpression of gtt and gdhcontrolled by promoter PermE* in Saccharopolyspora spinosa SIPI-A2090.Biotechnol. Lett. 33:733–739.

29. Combes P, Till R, Bee S, Smith MCM. 2002. The Streptomyces genomecontains multiple pseudo-attB sites for the �C31-encoded site-specificrecombination system. J. Bacteriol. 184:5746 –5752.

ECB Bioconversion by Overexpression of Deacylase

February 2013 Volume 79 Number 4 aem.asm.org 1133

on March 23, 2018 by guest

http://aem.asm

.org/D

ownloaded from