ISSN 00036838, Applied Biochemistry and Microbiology, 2010, Vol. 46, No. 2, pp. 148–153. © Pleiades Publishing, Inc., 2010. Published in Russian in Prikladnaya Biokhimiya i Mikrobiologiya, 2010, Vol. 46, No. 2, pp. 161–165.

Function Analysis of a New Type I PKSSAT Domain by SatEat Domain Replacement1 Y. L. Jiaoa, L. H. Wangb, B. H. Jiaob, S. J. Wanga, Y. W. Fanga, and S. Liuc a

College of Marine Sciences, HuaiHai Institute of Technology, Lianyungang, 222005, China b Department of Biochemistry and Molecular Biology, College of Basic Medical Sciences c Second Military Medical University, Shanghai, 200433, China email: [email protected] Received March 31, 2009

Abstract—The function of a new starter unit acyltransferase (SAT) domain SATEF080951 (GenBank accession number) encoded in a new type I polyketide synthase (PKS) gene cluster EF568935 (GenBank accession number) isolated for this study was analyzed by domain replacement with an extender unit AT (EAT) domain of avermectin PKS. It was shown that the SATEF080951 incorporated malonylCoA specif ically in vivo, which contradicted the specificity that we had previously determined by substrate binding test in vitro. The result of this study indicates that type I PKSSAT can alter its specificity in vivo and functions well in extender units and proved the feasibility of the SATEAT domain replacement in type I PKS. We pro pose that SATEAT replacement strategy could be a novel route for increasing the diversity of new polyketides combinatorially biosynthesized. The new type I PKSSATEF080951 studied herein may be further employed for related studies on enzymology or combinatorial biosynthesis of polyketides. DOI: 10.1134/S0003683810020043 1

Polyketides form an enormous class of natural products including antibiotics, pigments, and other biologically active compounds [1, 2]. They are biosyn thesized from sequential condensation of short chain acyl coenzyme A (CoA) precursors or ACPlinked precursors by multifunctional enzymes called polyketide synthases (PKSs) [3, 4]. Due to its striking twolevel sequential arrangement about gene of mod ules and domains, modular type I PKS is the most suitable enzyme system for combinatorial biosynthesis of new compounds [5, 6]. SAT domain has broad sub strate specificity while EAT domain has stricter struc tural and stereochemical specificity for substrates and exercises a tight control over the nature of the polyketide chain extension unit [7, 8]. Thus the dever sity of polyketides biosynthesized is mainly deter mined by the broad substrate specificity of SAT domain. So we presume that when SAT domain is placed in extender units, it can increase the diversity of polyketides biosynthesized by PKS. But what should be known first is that whether SAT can functions well in extender unit. In contrast to the well studied keto synthase (KS) domain [9], less is known about whether SAT can work as EAT and whether the speci ficity in vivo will be different from that in vitro. 1

The article is published in the original.

Abbreviations: Acyl carrier protein—ACP; coenzyme A—CoA; cetyl trimethylammonium bromide—CTAB; extender unit acytransferase—EAT; ketosynthase—KS; polyketide synthase— PKS; starter unit acytransferase—SAT.

In previous work, we had isolated a 7981 bp long new PKS partial gene cluster EF568935 from metage nome DNA of marine sediment samples [10]. The gene structure of EF568935 was analyzed with bioin formatics software (Fig. 1). A 1026 bp long SAT gene SATEF080951 in EF568935 was heterologously expressed and acetylCoA specificity of it was deter mined in vitro by substrate binding test with 4 sub strates CoA, acetylCoA, malonylCoA and methyl malonylCoA. In this study, the specificity of SATEF080951 domain was further analyzed by domain replacement in vivo. Targeting plasmid for replacement was con structed and transformated into protoplast of S. aver mitilis A1 strain. Gene encoding a malonylCoA spe cific AT domain AT2 (3) in avermectin PKS gene clus ter (as shown in Fig. 2) was replaced with SAT EF080951 gene in the host S. avermitilis A1 strain by double crossover homologous recombination. The mutant strain was verified by PCR detection and the metabolites were analyzed by HPLC. If SAT EF080951 incorporates acetylCoA specifically the extension of polyketide chain will terminate, and no avermectin will be detected; if SATEF080951 can incorporate malonylCoA the product will remain unchanged. The objectives of this study are to clarify if there is a difference of the specificity of type I PKSSAT domain between in vivo and in vitro, and whether SAT domain can functions as well in extender units, and thus to provide a novel possible strategy for increasing

148

FUNCTION ANALYSIS OF A NEW TYPE I PKSSAT DOMAIN

149

Attenuator 2502–2545 –35 box 2824 –10 box 2844 P450A 11

2868 trsanscriptial KS (DQ924531) AT (EF080951) start

P450B

887 1178 1636

3508

4322

4681

5625

ACP 6755

6899 7981 bp

Fig. 1. Analysis on gene structure of EF568935.

aveA1

aveA2

translate

translate

AVES1 Loading domain module 1

AVES2 module 2

module 3 module 4

DH KR KR KR AT ACP KS AT ACP KS AT ACP KS ATACP KS AT ACP

O

S

S

S

S

O

O

OH OH

module 6

DH KR KR KS AT ACP KS AT ACP

S

O

O

module 5

S

KS

C

O

O

OH

OH

O

OH

OH

O

OH

OH OH

AT2(3) The AT incorporating malonylCoA

O

O OH OH

Fig. 2. Gene structure of avermectin PKS. (The arrow indicates the target AT domain for the replacement).

the diversity of new polyketides by combinatorial syn thesis. MATERIALS AND METHODS Reagents. pMALc2X vector (NEB, USA), pMD18T vector (Takara, Japan), pMLA vector (apramycin resistant, constructed from pMALc2X vector), E. coli BL21 (Novagen, USA), S. avermitilis A1 (isolated from soil and conserved in this lab); apra mycin (Bio Basic Inc.), all rengents were from Sigma (USA); PCR regents and enzymes were from Takara; protoplast buffer, transformation buffer, all microb culture media (R2, R2YE, RM14, YEME, et al.) were prepared according to reference [11]; actinomycetes DNA exraction buffer preparation: 1.4 mol/l NaCl, 2% CTAB, 0.02 mol/l EDTA, 0.1 mol/l TrisHCl. APPLIED BIOCHEMISTRY AND MICROBIOLOGY

DNA preparation for plasmid construction. Tweentyfive micrograms of S. avermitilis A1 cells har vested from YEME medium were grand to debris, and then added 500 μl DNA extraction buffer, 50 μl SDS (10%) and 2 μl proteinase K (20 g/l) sequentially. The mixture was incubated in 37°C water bath and then added 75 μl NaCl (5 mol/l) solution, equal volume of chloroform: isoamyl alcohol 24:1 sequentially. Fol lowed by centrifugation at 12000 g for 15 min and repeated 2–3 times. The supernatant was transferred and added 2 volume icecold isopropanol and placed for an h, and then centrifiiged at 10000 g for 15 min. The DNA pellet was washed by 70% icecold ethanol and disolved in 100 μl TE buffer. RNA was removed by adding RNase A and incubating at 37°C for an h. DNA was then centrifuged to pellet, washed by 70% icecold ethanol and finally dissolved in 100 μl TE buffer.

Vol. 46

No. 2

2010

150

JIAO et al.

Primers used in the experiments DNA fragments Primers HmL

Sequences

Size, bp

LHMFP GCGAATTCGTGCGGGTGAGTGCTCGATG(EcoR I)

1076

LHMRP CGCGGCGCCAAGCAGCCTCTAGACTCCCCGGTCCCGGTCCCGC(Xba I) HmR

RHMFP GGCTGCTTGGCGCCGCGCTGCAGACCCACCACAACCTCCCCAAC(Pst I)

1084

RHMRP GCAAGCTTCCCCGACCACGACCGACACC(Hind III) ATEF080951

ATFP

CTTCTAGAGGCTTCGGCGAGGCGGGCTTC(Xba I)

ATRP

CACTGCAGGGGCTCCGCACGGAACCAG(Pst I)

5' recombinant region

FTFP

GCGTCATCTCCGGTCGTATCTCG

FTRP

CCGCCGCCGCCTCACGCAGC

3' recombinant region

RTFP

TCCTCGAAACCTCCCTGCAC

RTRP

GAAACCGAGACAGCACGCAAC

Linker regions flanking both sides of AT2(3) domain gene in aveA1 of avermectin PKS gene were selected as homologous arms. A1 kb 5' homologous arm (HmL, with 5' EcoR I and 3' Xba I sites) and a 1 kb 3' homologous arm (HmR, with 5' Pst I and 3' Hind III sites) were amplified with primers shown in table from the genomic DNA of S. avermitilis A1, cloned into pMD18T vector and verified by DNA sequencing. HmL were linked to HmR in 5' to 3' direction by recombinant PCR, resulting a 2.2 kb DNA fragment which was then verified by DNA sequencing. ATEF080951 fragment was amplified by PCR with primers ATFP and ATRP (table) from pMAL ATEF080951 expression vector. 5' and 3' ends of the ATEF080951 fragment were added Xba I and Pst I sites respectively. ATEF080951 fragment was then verified by DNA sequencing, and cloned into pMD18T vector. Construction of plasmid pMLAHmLRSAT EF080951 for AT domain replacement. pMLA vector (Aprr) and recombinant fragment HmLR were both double digested by EcoR I and Hind III, and then HmLR was inserted into pMLA vector. Plasmid pMLAHmLR and pMD18TATEF080951 were both digested by Xba I and Pst I, and then AT EF080951 was cloned into pMLAHmL R, resulting targeting plasmid pMLAHmLRATEF080951. Transformation of S. avermitilis A1 and screening for the mutant. The cells of S. avermitilis A1 cultured on YEME protoplast medium for 5 days were har vested and centrifuged at 4000 g for 5 min. The cells were resuspended by P buffer and rinsed 5 times. Lysozyme was added to a final concentration of 1 mg/ml and the mixture was slowly vibrated for 5 h, resulting the protoplast cells. The cells was filtered through sterilized cotton wool and centrifuged at 3000 g for 5 min, and then resuspended by P buffer. Ten mil

1026

1569

1344

liliter of TE dissolved pMLAHmLRATEF080951 plasmid and 0.5 ml T buffer (fresh prepared) were added into 100 μl protoplast cells. After 30 s, the mix ture was diluted with 3 ml soft agar, and then overlaid on RM14 regeneration medium. After incubation at 28°C for 18 h, the plate was covered with 3 ml RM14 soft agar containing apramycin (100 μg/ml final), and then continued incubation at 28°C for 10 days. The domain replacement by homologous recombinant was represented in Fig. 3. Positive clones were picked and colony PCR was carried for screening of the mutants. Two pairs of primers were designed targeting the 5' and 3' recombinant region (as shown in table). The identi fied mutant was named AT2(3). HPLC analysis of metabolites. Synchronous spores of the wildtype strain S. avermitilis A1 and the mutant S. avermitilis AT2(3) were collected and both inocu lated with the same size into media for seed culture. After incubation at 280 C for 24 h, they were synchro nously inoculated with the same size into fermenta tion media, and then continued incubation at speed of 180 rpm for 10 days. One micro liter of each of the cul ture broths were added 4 ml methanol respectively and then shook for 20 min, followed by centrifugation at 11000 g for 10 min. Forty ml of each of the superna tant were subjected to HPLC analysis respectively. The experimental conditions were as follow: the column was Agilent ZORBAX Extend C18 (4.6 × 250 mm, 5 μm); detection wavelength was 245 nm; temperature of the column was 20°C; flow rate was 1 ml/min; sam ple volume was 40 μl; mobile phase was methanol water (80 : 20). RESULTS AND DISCUSSION Generation of SATEAT domain replaced mutant. To avoid altering the original structure of avermectin PKS as much as possible, region of interdomainal

APPLIED BIOCHEMISTRY AND MICROBIOLOGY

Vol. 46

No. 2

2010

FUNCTION ANALYSIS OF A NEW TYPE I PKSSAT DOMAIN aveA1

module 2 aveAT2(3)

module 1

S. avermitilis A1 chromosomal DNA

151

XbaI Pst I

EcoRI

HindIII

HmL ATEF080951 HmR in EF568935 pMLAHmLRATEF080951

Double crossover

aveA1

module 1

module 2 ATEF080951

Chromosomal DNA of replacement mutant Fig. 3. Schematic representation of the replacement of the gene encoding AT2(3) in strain A1 chromosome with ATEF080951 in plasmid pMLAHmLRATEF080951, which contains a 3.2kb EcoRl–HindIII fragment with the foreign AT region and the homologous regions flanking AT2(3).

linker peptides at both terminus of the foreign SAT EF080951 domain were both no more than 14 amino acid residues and the native interdomainal linker pep tides were retained maximally, which facilitated the replacement of EAT with SAT and resulted a success ful domain replacement experiment in this study. Region of inserted foreign SATEF080951 DNA sequence was 1032 bp and that of replaced native AT2(3) DNA sequence was 837 bp. Nterminal pep tide remaining of SATEF080951 domain was GFGEAGFGSARRTA and Cterminal peptide remaining was GQVRWARWFRAEP, while both replaced terminal peptides of native domain AT2(3) were AAGKTA and DNTLTTL respectively. Targeting vector pMLAHmLRSATEF080951 was verified by restriction enzymes digestion test and DNA sequenc ing (Fig. 4). Colony PCR was used for screening the correct domain replaced mutant with two pairs of primers. For each pair of primers, one primer binds to SAT EF080951 and the other to the downstream of HmR or upstream of HmL on avermectin PKS gene. Approximately 900 apramycin resistant clones were picked and tested by PCR amplification and two of them gave the expected amplicons of 1569 bp and 1344 bp. To further confirm the identity of the mutants, the PCR fragments of expected size were verified by DNA sequencing and they were both SAT replaced mutants. Metabolites in the domainswapped mutants. The metabolites produced in the mutants as well as in the wildtype of S. avermitilis were analyzed by high HPLC. The yield of avermectin B1a in the extracts from the mutant was less than that of the wildtype strain (Fig. 5) while two unknown substance peaks 1 and 2 remain the same. By absorption equation and peak area of avermectin B1a, the concentration of B1a in the fermentation broths of the wildtype A1 and the mutant AT2(3) were calculated as 0.103; and 0.056 μg/μl respec APPLIED BIOCHEMISTRY AND MICROBIOLOGY

tively. As parallel control, the culture medium without inoculation was detected by HPLC with the same method and no peak 1 or 2 was found. The key function of type I PKS is the synthesis of long chains of carbon atoms by decarboxylative βcondensa tion of small organic acids in their active forms CoA esters, so if an acetylCoA specific AT domain exists in extender units of type I PKS the chain elongation may terminate [3]. Analysis on the sequence of EF080951 gene in the partial PKS gene cluster EF568935 showed that it was a SAT (Fig. 1), and was concluded as acetyl CoA specific by substrate binding test in vitro, which agreed with the theoretical function of SAT but con flicted with the prediction that SATEF080951 was malonylCoA specific by bioinformatics software (as Clustal × 1.83), so an AT domain AT2(3) with specific ity consistent with software prediction was chosen and replaced with SATEF080951. Substrate binding test is an effective and simple method which has been adopted by many researchers in various studies about enzyme’s substrate specificity [12, 13]. This study pre sented an evidence supporting existence of the differ

Vol. 46

M

1

2

3

4

4.5 kb 3 kb 2 kb 1.2 kb 0.8 kb

Fig. 4. Double digestion test of pMLAHmLRAT EF080951 vector. M—marker; 1—Pst I and Hind III double digestion; 2—Xba I and Pst I double digestion; 3— EcoR I and Xba I double digestion; 4—EcoR I and Hind III double digestion. No. 2

2010

152

JIAO et al.

25

40

25

36.792

B1a

35 (c)

2 29.852

37.378

B1a 28.793

24.243 30

30

26.856

1

(b) 20.007

28.724

2 26.957

1 24.002

B1a

36.902

(a)

30

35

Fig. 5. HPLC analysis of the metabolites: (a) of standard avermectin B1a; (b) of fermentation broth of S. avermitilis A1; (c) of fermentation broth of mutant S. avermitilis AT2(3).

ence of type I SAT specificity between in vitro and in vivo and showed that the results of the two methods can be contradictory. Using substrate feeding strategy in vitro, Chaitan Khosla et al. have found that type I PKSSAT domain has broad substrate specificity [7]. In this study SATEF080951 exhibited broad but dif ferent specificity and complemented their conclusions in another way. When placed in an extender unit in vivo, SAT EF080951 was possibly be interfered with some unknown cell factors and conversed to be malonyl CoA specific, while ATEF080951 might not totally lost its ability of incorporating acetylCoA and caused the lower production of avermectin, and if SAT EF080951 still holds its acetylCoA specificity strin gently when replaced in an extender unit, the PKS will be destroyed and no normal polyketide can be synthe sized. The results showed that SAT still have broad substrate specifictity when placed in an extender unit, and thus gained insight into a novel approach for increasing the diversity of novel polyketides biosynthe sized by genetic manipulation. Nevertheless, there was another possibility: SATEF080951 might altered the original stereo chemical structure of avermectin PKS protein, resulting a hybrid PKS with decreased enzy matic activity. Gokhale R.S. et al. [14] have found that intermod ular and interpolypeptide linkers are crucial compo nents of PKS assembly while less was known about interdomainal linkers which play an important role in combinatorial biosynthesis manipulation of new polyketides. This study showed a successful domain replacement by controlling the foreign interdomainal linker length within 14 animo acid residues. However the domain replacement method is technically diffi

cult and has not high operability because many factors may alter the structure of native PKS protein such as the constitution, length of foreign domain interdo mainal linkers and the homologous recombination regions selected. When the alteration interfers func tional amino acid sites of the original PKS a hybrid PKS with new enzymatic activity will be produced. Though not as simple and feasible as substrate binding test in vitro this method may be more accurate and is undoutedly important in studies on domain function identification and combinatorial biosynthesis of new polyketides. ACKNOWLEDGMENTS This work was supported in part by Natural Science Funds of the People’ Republic of China (grant no. 30670043). REFERENCES 1. Alekhova, T.A. and Novozhilova, T.Y., Appl. Biochem. Microbiol., 2001, vol. 37, no. 3, pp. 267–273. 2. Bentley, R. and Bennett, J.W., Annu. Rev. Microbiol., 1999, vol. 53, pp. 411–446. 3. Hopwood, D.A., PLoS Biol., 2004, vol. 2, no. 2, p. E35. 4. Shen, B., Curr. Opin. Chem. Biol., 2003, vol. 7, no. 2, pp. 285–295. 5. Reeves, C.D., Crit. Rev. Biotechnol., 2003, vol. 23, no. 2, pp. 95–147. 6. Floss, H.G., J. Biotechnol., 2006, vol. 124, no. 1, pp. 242–257. 7. Khosla, C., Gokhale, R.S., Jacobsen, J.R., and Cane, D.E., Annu. Rev. Biochem., 1999, vol. 68, pp. 219–225.

APPLIED BIOCHEMISTRY AND MICROBIOLOGY

Vol. 46

No. 2

2010

FUNCTION ANALYSIS OF A NEW TYPE I PKSSAT DOMAIN 8. Yadav, G., Gokhale, RS, Mohanty, D. J. Mol. Biol., 2003, vol. 328, no. 2, pp. 335–363. 9. Castonguay, R., He, W., Chen, A.Y., Khosla, C., and Cane, D.E. J. Am. Chem. Soc., 2007, vol. 129, no. 44, pp. 13 758–13 769. 10. Jiao, Y.L., Wang, L.H., Dong, X.Y., Chen, Y.F., Zong, Y., Gao, Y., Ren, N., Guo, A.Y., Zhang, X.Q., and Jiao, B.H., Appl. Biochem. Biotechnol., 2008, vol. 149, no. 1, pp. 67–78. 11. Kieser, T., Bibb, M.J., Buttner, M.J., Chater, K.F.., and Hopwood, D.A., Practical Streptomyces Genetics: A

APPLIED BIOCHEMISTRY AND MICROBIOLOGY

153

Laboratory Manual, Norwich: The John Innes Foun dation, 1985. 12. Reeves, C.D., Chung, L.M., Liu, Y.Q., Xue, Q., Car ney, J.R., Revill, W.P., et al., J. Biol. Chem., 2002, vol. 277, no. 11, pp. 9155–9159. 13. Funa, N., Ohnishi, Y., Ebizuka, Y., and Horinouchi S., J. Biol. Chem., 2002, vol. 277, no. 7, pp. 4628–4635. 14. Gokhale, R.S., Sankaranarayanan, R., and Mohanty, D., Curr. Opin. Struct. Biol., 2007, vol. 17, no. 6, pp. 736– 743.

Vol. 46

No. 2

2010