A specific cytochrome P450 hydroxylase in herboxidiene biosynthesis

Bioorganic & Medicinal Chemistry Letters 24 (2014) 4511–4514

Contents lists available at ScienceDirect

Bioorganic & Medicinal Chemistry Letters journal homepage: www.elsevier.com/locate/bmcl

A specific cytochrome P450 hydroxylase in herboxidiene biosynthesis Dayu Yu a,b,⇑, , Fuchao Xu b, , Lei Shao b, Jixun Zhan b,⇑ a b

Department of Applied Chemistry and Biological Engineering, College of Chemical Engineering, Northeast Dianli University, Jilin, Jilin 132012, China Department of Biological Engineering, Utah State University, 4105 Old Main Hill, Logan, UT 84322, United States

a r t i c l e

i n f o

Article history: Received 2 July 2014 Revised 25 July 2014 Accepted 29 July 2014 Available online 6 August 2014 Keywords: Herboxidiene Polyketide biosynthesis Cytochrome P450 Hydroxylation Specificity

a b s t r a c t The anti-cholesterol natural product herboxidiene is synthesized by a noniterative modular polyketide synthase (HerB, HerC and HerD) and three tailoring enzymes (HerE, HerF and HerG) in Streptomyces chromofuscus A7847. In this work, the putative monooxygenase HerG was expressed in Escherichia coli and the purified enzyme was subjected to biochemical studies. It was identified as a cytochrome P450 enzyme responsible for the stereospecific hydroxylation at C-18. This enzyme is highly substrate-specific as it efficiently hydroxylates 18-deoxy-25-demethyl-herboxidiene, but showed no activity towards 18-deoxy-herboxidiene. The kcat/Km value for the HerG-catalyzed hydroxylation of 18-deoxy-25-demethyl-herboxidiene was determined to be 1669.70 ± 47.36 M 1 s 1. In vitro co-reaction of HerG with the methyltransferase HerF and analysis of the product formation in S. chromofuscus A7847 revealed that the biosynthetic intermediate 18-deoxy-25-demethyl-herboxidiene is successively hydroxylated at C-18 by HerG and methylated at 17-OH to yield the final product herboxidiene. The minor metabolite 18-deoxy-hereboxidiene is a byproduct of the biosynthetic pathway. Ó 2014 Elsevier Ltd. All rights reserved.

Herboxidiene (1, Fig. 1) is a novel anti-cholesterol compound from Streptomyces chromofuscus A7847 (ATCC 49982).1 It was also found to have antitumor2 and herbicidal activities,3,4 thus representing a promising bioactive molecule. Structurally, 1 is a polyketide that features conjugated diene, oxycarbonyl, and epoxide groups. Recently, the gene cluster responsible for herboxidiene biosynthesis has been identified from S. chromofuscus by our group. It contains seven open reading frames (ORFs), including a putative regulatory gene herA, three type I polyketide synthase (PKS) genes (herB, herC and herD) and three tailoring enzyme genes herE, herF and herG. The noniterative modular PKS consisting of three polypeptides HerB, HerC and HerD assembles the nonaketide intermediate (2, Fig. 1), which was subjected to three tailoring reactions to yield the final product 1. In our previous work, we have confirmed the involvement of the PKS in herboxidiene biosynthesis by disrupting an acyltransferase (AT) domain in HerC.5 We have also identified HerF as a dedicated methyltransferase that is responsible for the methylation of 17-OH.6 To fully understand the biosynthetic pathway of 1, the two remaining tailoring reactions, C-14,C-15-epoxidation and C-18 hydroxylation, need to be elucidated. This work focuses on the characterization of the enzyme in charge of C-18 hydroxylation through heterologous expression and biochemical studies. ⇑ Corresponding authors. Tel.: +1 435 797 8774; fax: +1 435 797 1248.  

E-mail addresses: [email protected] (D. Yu), [email protected] (J. Zhan). These authors contributed equally to this work.

http://dx.doi.org/10.1016/j.bmcl.2014.07.078 0960-894X/Ó 2014 Elsevier Ltd. All rights reserved.

BLAST analysis revealed that both HerE and HerG are cytochrome P450 (CYP) enzymes, a superfamily of heme-thiolate proteins widely distributed in living organisms. CYPs catalyze a variety of reactions such as hydroxylation, deamination, dealkylation, dehalogenation, and epoxidation.7,8 HerE showed significant sequence similarity to many epoxidases, while HerG is homologous to a number of putative CYP monooxygenases, including a previously reported CYP hydroxylase PikC in picromycin biosynthesis. PikC is a tailoring CYP hydroxylase that accepts both 12and 14-membered ring macrolide as the substrates.9 HerG and PikC share 49% identity and 60% similarity. Accordingly, we propose that HerG is the C-18 hydroxylase in herboxidiene analysis. We aligned the amino acid sequence of HerG with several previously reported CYP hydroxylases (Fig. 2). PdmJ is a C-5 hydroxylase in the pradimicin biosynthetic pathway,10,11 while EryF is the first well-characterized natural product hydroxylase involved in erythromycin biosynthesis.12,13 As shown in Figure 2, several conserved regions are present in HerG. The glutamic acid295 (E295) and arginine298 (R298) residues of the highly conserved EXXR motif are proposed to stabilize the core and heme-binding.14 The characteristic CYP consensus sequence F360XXGXXXCXG369 in the heme-binding loop is also found in HerG. The conserved cysteine367 (C367) serves as the fifth ligand to the heme iron.15 The (G/A)GX(D/E)T conserved in CYPs is present in HerG and the threonine260 (T260) in this motif is supposed to participate in the formation of the oxygen-binding pocket.16,17 These sequence characteristics further suggested that HerG is a CYP monooxygenase.

4512

D. Yu et al. / Bioorg. Med. Chem. Lett. 24 (2014) 4511–4514 20

O

6x

O

HO

3

S-CoA

Methylmalonyl-CoA

+ O

3x

HO

PKS: HerB-HerC-HerD

COOH

HerE

O

HerG

COOH

OH

OH epoxidation

OH O

hydroxylation

(2)

O

1

methylation 19 18 OH

22

O

12

14 24

23

Herboxidiene (1)

H F er

x

O

HerG

COOH OCH 3

25

OCH 3

OH 25-Demethyl-herboxidiene (5)

18-Deoxy-25-demethyl-herboxidiene (3)

21

COOH

COOH

HerF

S-CoA

Malonyl-CoA

O

O

O

O

7

O

18-Deoxy-herboxidiene (4)

Figure 1. Assembly of herboxidiene (1) through a type I polyketide biosynthetic pathway.

To functionally identify HerG, we amplified the corresponding gene (1269 bp) from the genome of S. chromofuscus A7847 and ligated it into the Escherichia coli expression vector pET28a, yielding pLS17. This plasmid was introduced into E. coli BL21(DE3) for protein expression. SDS–PAGE analysis of the induced culture of E. coli BL21(DE3)/pLS17 revealed that HerG can be expressed solubly in this heterologous host (Fig. 3A). The C-His6-tagged HerG was purified by Ni-NTA chromatography to near homogeneity (Fig. 3A). We first tested the biochemical properties of HerG. Purified HerG was subjected to spectroscopic analysis. An absorption spectrum of 10 lM Fe(III)-HerG in 50 mM Tris–HCl buffer (pH 7.9) generated at room temperate showed the major Soret or c band at 418 nm and the smaller a and b bands at 535 and 570 nm, respectively (Fig. 3B). Addition of 2.5 mM sodium dithionite (DT) to the enzyme saturated with CO resulted in the change of the absorption peak at 418 to a new peak at 448 nm immediately, indicating that HerG is a CYP enzyme (Fig. 3B). However, the P450 co-adduct is unstable and rapidly changed back to the P420 form over 15 min. After confirming that HerG is really a CYP enzyme, we next conducted in vitro enzymatic reactions18 to identify its role in herboxidiene biosynthesis. Two herboxidiene analogues that lack the corresponding 18-OH were used in the reactions as potential substrates.9,10 18-Deoxy-25-demethyl-herboxidiene (3, Fig. 1) is a compound isolated from the engineered strain S. chromofuscus A7847/pLS7 created in our previous work.6 18-Deoxy-herboxidiene (4, Fig. 1) is a minor metabolite in wild type S. chromofuscus A7847.5 These two compounds were respectively reacted with HerG in the presence of spinach ferredoxin and spinach ferredoxin–NADP reductase at 30 °C. The reactions were quenched by

addition of an equal volume of methanol. After centrifugation, the supernatants were subjected to LC–MS analysis. As shown in Figure 3C, 3 was efficiently converted into a new product 5 in 10 min (trace i), while no product was observed from the incubated of 4 with HerG (trace ii). No reaction occurred even with elongated time (overnight, data not shown). Incubation of the substrate (3 or 4) with the reaction buffer for the same time period was used as the negative controls, and no products were detected from these controls (traces iii and iv, Fig. 3C). ESI-MS spectra of 5 showed the ion peaks [M+H]+ at 425 m/z and [M H] at 423 m/z (Supplementary Fig. 1), indicating that the molecular weight of this compound is 424. This is 16 mass units larger than the substrate, suggesting that it is a hydroxylated product of 3. A comparison of this product with the standard of 25-demethyl-herboxidiene (trace v, Fig. 3C) from our previous work indicated that they are identical. Thus, 5 was identified as 25-demethyl-herboxidiene. The conversion of 3 to 5 by HerG confirmed that this CYP enzyme is the C-18 hydroxylase. Furthermore, this enzyme showed strict substrate specificity, as it has no activity towards 4 that has the same structure as 3 except an additional CH3. We next studied the kinetics of the HerG-catalyzed hydroxylation of 3 with the substrate concentrations ranging from 0.025 to 0.4 mM. The results indicated that the enzyme has a high affinity (65.88 ± 3.49 lM) and turnover number 0.11 ± 0.01 s 1 towards 3, as shown in Table 1. Consequently, the kcat/Km value for the hydroxylation of 3 by HerG was determined to be 1669.70 ± 47.36 M 1 s 1. Our previous work revealed that the methyltransferase HerF can also use 3 as a substrate for the O-methylation. However, the efficiency is much lower than

HerG 1 MTETCPARDLYTP--AYFKDPYPA---LTRLRDAGPVHRVERPDGLVVWLITRYAEAQAALGDPRLSMDGEVVQKALGAFAYGYLD-PE---NEAPHTLLSSDPPDH 98 EryF 1 MT-TVP--DLESD--SFHVDWYRT---YAELRETAPVTPV-RFLGQDAWLVTGYDEAKAALSDLRLSSDPKKKYPGVEVEFPAYLGFPEDVRNYFATNMGTSDPPTH 98 PikC 1 MRRTQQGTTASPPVLDLGALGQDFAADPYPT---YARLRAEGPAHRVRTPEGDEVWLVVGYDRARAVLADPRFSKDWRN-STTPLTEAEAALN----------HNMLESDPPRH 100 PdmJ 1 MPSSKDAPTVDPRPDVTPAFPFRPDDPFQPPCEHARLRASDPVAKVVLPTGDHAWVVTRYADVRFVTSDRRFSKEAVT-RPGAPRLIPMQRG---------SKSLVIMDPPEH 103 HerG 99 TRLRRLVNRTFTARRIQALRPRVQELMDGLLDALGP-DAAHADLIEAVAAPLPIAVICELLGVPPEDYDSFKLWTTTMFVLPA-DVGDGMSPTDAMRNLRRYLSDLIAAKRA 208 EryF 99 TRLRKLVSQEFTVRRVEAMRPRVEQITAELLDEVG--DSGVVDIVDRFAHPLPIKVICELLGVDEKYRGEFGRWSSEILVMDP-ERAEQRG--QAAREVVNFILDLVERRRT 205 PikC 101 TRLRKLVAREFTMRRVELLRPRVQEIVDGLVDAMLAAPDGRADLMESLAWPLPITVISELLGVPEPDRAAFRVWT-DAFVFPD-DPAQAQT---AMAEMSGYLSRLIDSKRG 207 PdmJ 104 TRMRKIVSRAFTARRVEGMRAHVRDLTSGFVDEMVE-HGPPADLIAHLALPLPVTVICEMLGVPPEDRPRFQDWTDRMLTIGAPALAQADEIKAAVGRLRGYLAELIDAKTA 214 HerG 209 ERPETGQGAAGTEESGDLLSALIAVRDTDEGRLSEHELVSMAVQLLIAGHETTVNGIGNAVLNLLRHPEQLAALRAEPALLPRAVDELLRF-EGPLETAILRVATEPIPLG 318 EryF 206 E-P-------GD----DLLSALIRVQDDDDGRLSADELTSIALVLLLAGFEASVSLIGIGTYLLLTHPDQLALVRRDPSALPNAVEEILRY-IAPPETTT-RFAAEEVEIG 302 PikC 208 Q-D-------GE----DLLSALVRTSDEDGSRLTSEELLGMAHILLVAGHETTVNLIANGMYALLSHPDQLAALRADMTLLDGAVEEMLRY-EGPVESATYRFPVEPVDLD 305 PdmJ 215 A-P-------AD----DLLSLLSRAHADDG--LSEEELLTFGMTLLAAGYHTTTAAITHSVYHLLREPSRYARLREDPSGIPAAVEELLRYGQIGGGAGAIRIAVEDVEVG 311 HerG 319 DQVVPAGALVKVVLAAANRDPDRFAAPDTLDITRKNEGHLQFGHGIHNCLGAFLARMETEIAISSLLRRYPGLSLGVPEDEIRWREIAIMRALAELPVTLT-----GPV 422 EryF 303 GVAIPQYSTVLVANGAANRDPKQFPDPHRFDVTRDTRGHLSFGQGIHFCMGRPLAKLEGEVALRALFGRFPALSLGIDADDVVWRRSLLLRGIDHLPVRLD-----G 404 PikC 306 GTVIPAGDTVLVVLADAHRTPERFPDPHRFDIRRDTAGHLAFGHGIHFCIGAPLARLEARIAVRALLERCPDLALDVSPGELVWYPNPMIRGLKALPIRWRRGREAGRRTG 416 PdmJ 312 GTLVRAGEAVIPLFNAANRDPEVFADPEELDLGRTDNPHIALGHGIHYCLGAPLARLELQVVLETLVERTPALRLAIDDADITWRPGLAFARPDALPIAW 411 Figure 2. Amino acid sequence alignment of HerG with three reported natural product CYP hydroxylases. EryF, GenBank accession number AAA26496; PikC, GenBank accession number AAC68886; PdmJ, GenBank accession number ABM21756. Identical residues are highlighted in grey. Three conserved regions in HerG are boxed.

4513

D. Yu et al. / Bioorg. Med. Chem. Lett. 24 (2014) 4511–4514 Table 1 Steady state kinetic parameters for the hydroxylation of 3 by HerGa

A

a

B

0.8

Δ Absorbance

0.3

0.6

0.2 420 nm 445 nm

0.1 0 0

5

10

15

Absorbance

-0.1 -0.2

Time (min)

0.4

HerG + CO (0 min) HerG + CO + DT (1 min) 0.2

0 350

HerG + CO + DT (15 min)

400

450

500

550

600

Kinetic parameter

Value

kcat Km kcat/Km

0.11 ± 0.01 s 1 65.88 ± 3.49 lM 1669.70 ± 47.36 M

1

s

1

Data were from three independent experiments.

insights into the actual biosynthetic pathway of 1, an in vitro coreaction of HerG and HerF with 3 was set up, which was subjected to time-course analysis.19 Because the kcat/Km value for the HerGcatalyzed hydroxylation of 3 is 28 and 6 times those for the HerF-catalyzed methylation of 3 and 5, respectively, it was hard to observe the methylated products when both enzymes were present at the same molar concentration with small amounts of 3. Accordingly, the ratio of HerG to HerF in the co-reaction mixture was changed to 1:5 and excess substrate was used to observe all the products. As shown in Figure 4A, the hydroxylated product 5 was immediately synthesized by HerG after 10 min. This compound was then converted to the final product 1 by HerF. At 30 min, three products were observed including 1, 4, and 5, which were identified by ESI-MS analysis (Supplementary Fig. 1) and a comparison with authentic samples. Among the products, 4 was synthesized from 3 by HerF, while 1 was formed by the actions of both HerF and HerG. The amount of 5 decreased with the reaction time, together with the accumulation of 1 and 4. It is apparent that 5 was further converted to 1, but 4 is not a precursor of 1, which is consistent with the reaction result shown in trace ii of

650

Wavelength (nm)

C

Figure 3. Biochemical characterization of HerG. (A) SDS–PAGE analysis of the expression and purification of HerG. M: protein ladder; 1: soluble fraction; 2: insoluble fraction; 3: flow through; 4–6: eluent with buffer A + 10 mM, 25 mM and 250 mM imidazole. (B) Spectral analysis of HerG. HerG (10 lM) in 50 mM Tris–HCl buffer (pH 7.9) was degassed with argon. The sample was saturated with CO in an anaerobic cuvette and the UV spectrum was recorded on a UV/Vis spectrophotometer. 2.5 mM DT was then added to the enzyme solution, and the UV spectrum was collected over 15 min. (C) HPLC analysis (230 nm) of the in vitro reactions of HerG with 3 and 4. The reactions were conducted in 100 mM Tris–HCl buffer for 10 min. (i) Incubation of HerG with 3 in the reaction buffer; (ii) incubation of HerG with 4 in the reaction buffer; (iii) incubation of 3 with the reaction buffer; (iv) incubation of 4 with the reaction buffer; (v) standard of 5.

the hydroxylation, as reflected by the kcat/Km value of 59.50 ± 0.17 M 1 s 1,6 suggesting that 3 is a natural substrate for HerG in the host S. chromofuscus A7847. Since both HerF and HerG displayed activities to 3, it will be interesting to understand how these two enzymes work collaboratively/competitively during herboxidiene biosynthesis. To gain

Figure 4. HPLC analysis (230 nm) of in vitro and in vivo formation of 1 and 4. (A) Time-course analysis of the co-reaction of HerF and HerG with 3. (B) Time-course analysis of the production of 1 and 4 in S. chromofuscus A7847.

4514

D. Yu et al. / Bioorg. Med. Chem. Lett. 24 (2014) 4511–4514

Figure 3C. We also conducted a time-course analysis of product formation during the fermentation of S. chromofuscus A7847.20 As shown in Figure 4B, 1 was produced as the major product, with 4 as a minor product. The amounts of the two products increased simultaneously with the fermentation time, and reached a stable level at 6 days. The final titer of 1 was 5.26 mg/L. Both the in vivo and in vitro results consistently confirmed that 4 can accumulate without being converted to 1 due to the strict substrate specificity of HerG. Thus, it is a byproduct instead of a precursor of 1. In summary, through heterologous expression and in vitro biochemical approaches, HerG was identified as a CYP hydroxylase that is responsible for the stereoselective hydroxylation at C-18 in herboxidiene biosynthesis. This enzyme is specific to the herboxidiene biosynthetic intermediate 3. From the time-course analysis results of the in vitro co-reaction of HerF and HerG with 3 and in vivo biosynthesis of 1 and 4 in S. chromofuscus A7847, it can be concluded that 1 is synthesized from 3 through successive C-18 hydroxylation and C-17 O-methylation catalyzed by HerG and HerF, respectively, as shown in Figure 1. Due to the relaxed substrate specificity of HerF, 4 was produced as a minor product when 3 is methylated. Thus, this work not only characterized a key tailoring enzyme in herboxidiene biosynthesis, but also revealed the sequence of the last two tailoring enzymes and provided biochemical clues for the formation of 4 during the biosynthetic process. Acknowledgments This work was supported by a National Scientist Development Grant (09SDG2060080 to J.Z.) from the American Heart Association and a Grant from the National Natural Science Foundation of China (31170763 to D.Y.). We thank Mr. Andrew Fielding and Professor Lance Seefeldt of the Department of Chemistry and Biochemistry, Utah State University for the assistance in the spectroscopic analysis of HerG. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.bmcl.2014.07. 078.

References and notes 1. Koguchi, Y.; Nishio, M.; Kotera, J.; Omori, K.; Ohnuki, T.; Komatsubara, S. J. Antibiot. 1997, 50, 970. 2. Hasegawa, M.; Miura, T.; Kuzuya, K.; Inoue, A.; Won Ki, S.; Horinouchi, S.; Yoshida, T.; Kunoh, T.; Koseki, K.; Mino, K.; Sasaki, R.; Yoshida, M.; Mizukami, T. ACS Chem. Biol. 2011, 6, 229. 3. Isaac, B. G.; Ayer, S. W.; Elliott, R. C.; Stonard, R. J. J. Org. Chem. 1992, 57, 7220. 4. Miller-Wideman, M.; Makkar, N.; Tran, M.; Isaac, B.; Biest, N.; Stonard, R. J. Antibiot. 1992, 45, 914. 5. Shao, L.; Zi, J.; Zeng, J.; Zhan, J. Appl. Environ. Microbiol. 2012, 78, 2034. 6. Yu, D.; Xu, F.; Zhang, S.; Shao, L.; Wang, S.; Zhan, J. Bioorg. Med. Chem. Lett. 2013, 23, 5667. 7. Bernhardt, R. J. Biotechnol. 2006, 124, 128. 8. Syed, K.; Mashele, S. S. PLoS One 2014, 9, e95616. 9. Xue, Y.; Wilson, D.; Zhao, L.; Liu, H.-W.; Sherman, D. H. Chem. Biol. 1998, 5, 661. 10. Napan, K. L.; Zeng, J.; Takemoto, J. Y.; Zhan, J. Bioorg. Med. Chem. Lett. 2012, 22, 606. 11. Zhan, J.; Qiao, K.; Tang, Y. ChemBioChem 2009, 10, 1447. 12. Podust, L. M.; Sherman, D. H. Nat. Prod. Rep. 2012, 29, 1251. 13. Weber, J. M.; Leung, J. O.; Swanson, S. J.; Idler, K. B.; McAlpine, J. B. Science 1991, 252, 114. 14. Hasemann, C. A.; Kurumbail, R. G.; Boddupalli, S. S.; Peterson, J. A.; Deisenhofer, J. Structure 1995, 3, 41. 15. Werck-Reichhart, D.; Feyereisen, R. Genome Biol. 2000, 1, S3003. 16. Poulos, T. L.; Finzel, B. C.; Howard, A. J. J. Mol. Biol. 1987, 195, 687. 17. Danielson, P. B. Curr. Drug Metab. 2002, 3, 561. 18. A typical 80-lL reaction of HerG contained 100 mM Tris–HCl buffer (pH 7.9), 4 lg of spinach ferredoxin, 0.02 U spinach ferredoxin–NADP reductase, 1.5 mM NADPH, 1 lM HerG, and 100 lM substrate (3 or 4). Reactions were initiated by the addition of substrate at 30 °C and quenched by an equal volume of methanol. The reaction mixtures were vortexed and centrifuged at 15,000 rpm for 5 min. The supernatants were analyzed on an Agilent 1220 HPLC using an Agilent ZORBAX XD-C18 column (5 lm, 4.6 mm  250 mm), eluted with 65– 100% methanol–water (containing 0.1% formic acid) at 1 mL/min over 30 min. 19. The co-reaction of HerF and HerG with 3 contained 100 mM Tris–HCl buffer (pH 7.9), 1 mM S-adenosylmethionine (SAM), 4 lg of spinach ferredoxin, 0.02 U spinach ferredoxin–NADP reductase, 1.5 mM NADPH, 5 lM HerF, 1 lM HerG, and 400 lM 3. The reaction was conducted in triplicate. The 600-lL reactions were initiated by the addition of substrate at 30 °C. Samples (80 lL) were taken at 0 min, 10 min, 30 min, 1 h, 2 h, 4 h, and overnight. Each sample was treated with the same volume of methanol and subjected to HPLC analysis and the representative results are shown in Figure 4A. 20. S. chromofuscus A7847 was grown in 500 mL of YM medium (0.4% glucose, 0.4% yeast extract, 1% malt extract, pH 7.3) at 28 °C with shaking at 250 rpm. Samples (100 mL) were taken from the fermentation broth on 2, 4, 6 and 8 days, and were extracted with an equal volume of ethyl acetate. The extracts were dried under reduce pressure, redissolved in methanol, and analyzed using the same HPLC method as that for the in vitro reactions.