Sylwia Tarcza,
Xiulan Xieb and
Shu-Ming Li*a
aInstitut für Pharmazeutische Biologie und Biotechnologie, Philipps-Universität Marburg, Deutschhausstrasse 17A, 35037 Marburg, Germany. E-mail: shuming.li@staff.uni-marburg.de
bFachbereich Chemie, Philipps-Universität Marburg, Hans-Meerwein-Strasse, 35032 Marburg, Germany
First published on 7th April 2014
Prenylated xanthones are secondary metabolites with a broad spectrum of biological activities including antimicrobial and antitumor activities. One xanthone O-prenyltransferase XptB has been identified in Aspergillus nidulans. Recently, we characterised a bisindolyl benzoquinone C- and N-prenyltransferase AstPT from Aspergillus terreus with unusually high substrate specificity towards both the prenyl donor dimethylallyl diphosphate and acceptor bisindolyl benzoquinone. In this study, we demonstrate the acceptance of a number of hydroxyxanthones by AstPT in the presence of not only dimethylallyl but also geranyl and farnesyl diphosphate. Structural elucidation by HR-MS and NMR analyses proved the O-prenylation of all thirteen isolated enzyme products with different prenyl moieties. These results demonstrated the remarkable substrate and catalytic promiscuity of AstPT, which was recognised as a specific enzyme prior to this study.
By using recombinant protein from E. coli, we demonstrated that AstPT from Aspergillus terreus was responsible for the transfer of a dimethylallyl moiety from dimethylallyl diphosphate (DMAPP) to N-1 and C-2 of asterriquinone D (AQ D, Fig. 1), a methylated bisindolyl benzoquinone derivative.8 In the presence of AQ D, AstPT accepted only DMAPP but not geranyl (GPP) or farnesyl diphosphate (FPP) as prenyl donor. This observation is consistent with the relatively high substrate specificity of the members of the DMATS superfamily9 to which AstPT also belongs. In contrast to most of the enzymes of the DMATS superfamily with relaxed substrate specificities towards aromatic substrates, AstPT did not accept tryptophan, tryptophan-containing cyclic dipeptides, hydroxynaphthalenes or flavonoids as prenylation substrates.8 Interestingly, product formation was detected in the incubation mixtures of AstPT with several hydroxyxanthones in the presence of DMAPP, GPP and FPP as prenyl donors. In this article we report its behavior towards hydroxyxanthones and the identification of the enzyme products as O-prenylated derivatives.
AstPT from Aspergillus terreus was shown to be responsible for the prenylation of AQ D.8 Although AstPT did not accept tryptophan or tryptophan-containing cyclic dipeptides as prenylation substrates, we carried out incubations of this enzyme with xanthone and ten hydroxylated derivatives (Table 1) in the presence of DMAPP, GPP or FPP. After incubation with 20 μg of AstPT in 100 μL assays at 37 °C for 16 h, the ethyl acetate extracts of the reaction mixtures were analysed on HPLC. HPLC analyses revealed notable product formation in the incubation mixtures of four hydroxyxanthones (1a–4a), not only with DMAPP but also with GPP and even with FPP (Fig. 2). In comparison to the results of negative controls with heat-inactivated protein, one additional predominant peak each was observed in the HPLC chromatograms of enzyme assays of 1a–3a with all tested prenyl donors and of 4a with GPP and FPP. In the assay of 4a with DMAPP three additional peaks were observed. No product formation was detected in the incubation mixtures with heat-inactivated protein (Fig. 2) or in the absence of the prenyl donors (data not shown). 1a was accepted as the best xanthone derivative with conversion yields of 23.1, 26.1 and 18.8% for DMAPP, GPP and FPP, respectively. 2a and 4a were accepted by AstPT with relative activities of 30 to 75% of those with 1a and the respective prenyl donors (Table 1). Remarkably, the relative activities in the assays with different prenyl donors for a given xanthone derivative did not change significantly. For 1a and 2a, GPP was slightly better accepted than DMAPP and FPP. In the case of 4a FPP was the best tested prenyl donor.
Substrate | Absolute conversion [%] | ||||||||
---|---|---|---|---|---|---|---|---|---|
R1 | R2 | R3 | R4 | R5 | R6 | DMAPP | GPP | FPP | |
a Incubations were carried out using 20 μg of recombinant protein for 16 h at 37 °C. Conversion yields for substrates 1a–4a are calculated from peak areas of the substrate and products in HPLC chromatograms and their integrals in the 1H NMR spectra (as incubation mixtures for isolation of the enzyme products). n.d.: not determined. | |||||||||
Xanthone | H | H | H | H | H | H | <0.1 | n.d. | n.d. |
3-Hydroxyxanthone | H | OH | H | H | H | H | <0.1 | n.d. | n.d. |
1-Hydroxy-6,8-dimethylxanthone | OH | H | H | CH3 | H | CH3 | 0.3 | n.d. | n.d. |
1,3-Dihydroxyxanthone | OH | OH | H | H | H | H | 0.9 | n.d. | n.d. |
1,7-Dihydroxy-6-methylxanthone (1a) | OH | H | H | CH3 | OH | H | 23.1 | 26.1 | 18.8 |
1,7-Dihydroxy-6,8-dimethylxanthone | OH | H | H | CH3 | OH | CH3 | <0.1 | n.d. | n.d. |
1,7-Dihydroxy-5,6,8-trimethylxanthone | OH | H | CH3 | CH3 | OH | CH3 | 0.4 | n.d. | n.d. |
1,7-Dihydroxy-6-methyl-8-hydroxy-methylxanthone | OH | H | H | CH3 | OH | CH2OH | <0.1 | n.d. | n.d. |
1,3,6-Trihydroxyxanthone (2a) | OH | OH | H | OH | H | H | 11.0 | 11.2 | 6.2 |
1,3,7-Trihydroxyxanthone (3a) | OH | OH | H | H | OH | H | 6.1 | 1.7 | 2.7 |
1,3,6,8-Tetrahydroxyxanthone (4a) | OH | OH | H | OH | H | OH | 11.8 | 8.4 | 14.0 |
The intriguing results of AstPT with hydroxyxanthones prompted us to compare its amino acid sequence with that of XptB again and to carry out enzyme reactions of XptB with its natural substrate 1,7-dihydroxy-6-methyl-8-hydroxymethylxanthone6,7 and the natural substrate of AstPT AQ D in the presence of DMAPP, GPP and FPP. AstPT and XptB share a sequence identity of only 23% on the amino acid level (Table S1†). XptB did not accept AQ D as aromatic substrate by using DMAPP, GPP or FPP as prenyl donor at 37 °C for 16 h. Product formation was only detected in the XptB assay of 1,7-dihydroxy-6-methyl-8-hydroxymethylxanthone with DMAPP, but not with GPP or FPP (Fig. S1†).
Compound | Chemical formula | [M]+ | Deviation [ppm] | |
---|---|---|---|---|
Calculated | Measured | |||
1b | C19H18O4 | 310.1205 | 310.1227 | 7.1 |
1c | C24H26O4 | 378.1831 | 378.1817 | −3.7 |
1d | C29H34O4 | 446.2487 | 446.2457 | −6.7 |
2b1 + 2b2 | C18H16O5 | 312.0998 | 312.0994 | −1.3 |
2c1 + 2c2 | C23H24O5 | 380.1624 | 380.1632 | 2.1 |
2d1 + 2d2 | C28H32O5 | 448.2224 | 448.2250 | 5.8 |
3b1 + 3b2 | C18H16O5 | 312.0998 | 312.0979 | −6.1 |
4c | C23H24O6 | 396.1573 | 396.1595 | 5.6 |
4d | C28H32O6 | 464.2199 | 464.2158 | −8.8 |
Comparison of the NMR data of 1b, 1c and 1d revealed that the aromatic protons have identical chemical shifts with nearly same coupling constants indicating the same prenylation position in their structures. Relatively strong downfield shift of +0.06 ppm was observed for the signals of H-5, probably caused by a prenylation at OH-7 rather than OH-1. To confirm the prenylation position, HMBC spectrum was taken for 1d. Correlations were clearly observed between H-1′ and C-7, proving the attachment of the farnesyl moiety to OH-7 (Fig. S15 and S16†). Therefore, 1d was identified as 6-methyl-7-farnesyloxy-1-hydroxyxanthone (Fig. 2). The position of the prenyl moieties in 1b and 1c was assigned analogously as the same changes were found in these 1H NMR spectra as for 1d. Hence, 1b and 1c were identified as 6-methyl-7-dimethylallyloxy-1-hydroxyxanthone and 6-methyl-7-geranyl-oxy-1-hydroxyxanthone, respectively (Fig. 2).
Inspection of the 1H NMR spectra of the isolated peaks 2b–2d revealed the presence of two prenylated products each (Table S3, Fig. S7–S9†). The ratios of 2b1 to 2b2 from the incubation mixture of 2a with DMAPP and 2c1 to 2c2 from that of 2a with GPP were found to be 1:4.3 and 4.3:1, respectively. Similar ratio of 6.1:1 was calculated for 2d1 to 2d2 from the incubation with FPP. Unfortunately, these compounds could not be separated from each other and their structures were elucidated by interpretation of the NMR spectra of the mixtures. This was possible due to the different intensities of the signals for protons of both compounds. The chemical shifts of H-7 and H-8 in 2b1, 2c1 and 2d1 have almost not changed in comparison to those of 2a indicating that the prenylation was unlikely at OH-6. This conclusion was supported by the significant changes of chemical shifts for H-2, H-4 and H-5. The prenylation position in 2b1, 2c1 and 2d1 was therefore assigned to OH-3 (Fig. 2). In comparison to those of 2b1, 2c1 and 2d1, the chemical shifts for H-5 and H-7 in 2b2, 2c2 and 2d2 differed clearly from those in 2a. In contrast, only slight changes have been observed for H-2. Therefore, the prenyl moieties in 2b2, 2c2 and 2d2 were assigned to OH-6.
Using 3a as substrate, interpretable spectra were only obtained for the products of the incubation mixture with DMAPP (Table S4, Fig. S11†). This sample was found to be a mixture of 3b1 and 3b2 with a ratio of 1:3.2. Comparison of the spectrum of 3b1 with that of 3a, significant changes of chemical shifts were found for H-2 and H-4 and nearly no changes for H-5, H-6 and H-8. In the spectrum of 3b2 obvious changes were found for H-5 and H-6 with upfield shifts of 0.05 and 0.06 ppm to that of 3a, respectively. These proved the prenylation at OH-3 in 3b1 and at OH-7 in 3b2.
4a has a symmetric structure; the 1H NMR spectrum shows therefore only two coupling doublets of two aromatic protons (Table S5, Fig. S12†). Interpretable spectra were obtained for 4c from incubation mixture with GPP and 4d with FPP. Two different coupling systems were observed in the spectra of both compounds (Fig. S13 and S14†). In the spectra of 4c and 4d, obvious changes of chemical shifts were found for H-2 and H-4. Thus, the prenylation position was assigned to OH-3 in both cases.
Compound | KM [μM] | kcat [s−1] | kcat/KM [s−1 M−1] |
---|---|---|---|
a Kinetic parameters for aromatic substrates were determined using DMAPP as prenyl donor.b Kinetic parameters for prenyl donors were determined using 1a as aromatic substrate.c Kinetic parameters were not determined, because maximal velocity could not be reached before inhibition occurred under our conditions. | |||
1aa | 17.3 | 0.001 | 55.2 |
2aa | —c | —c | —c |
3aa | 168 | 0.0007 | 4.2 |
4aa | 28.6 | 0.0003 | 8.6 |
DMAPPb | 97.3 | 0.001 | 8.5 |
GPPb | 12.1 | 0.002 | 169.6 |
FPPb | —c | —c | —c |
Very recently, we successfully cloned, overexpressed and purified AstPT from Aspergillus terreus and showed that this enzyme catalysed the transfer of a dimethylallyl moiety from DMAPP to AQ D, a methylated bisindolyl benzoquinone.8 AstPT was proven to be very specific for its prenyl donor DMAPP and acceptor AQ D. It accepted no further aromatic substrates of known prenyltransferases and no other prenyl donors such as GPP and FPP. In this study we demonstrated that AstPT was in fact able to transfer prenyl moieties not only to AQ D, but also to hydroxyxanthones. Both AQ D (Fig. 1) and hydroxyxanthones share similar structural features, i.e. a planar aromatic system as well as carbonyl groups. As hydroxyl groups in AQ D are blocked by methyl groups, a prenylation of these groups is not possible. On the other hand, hydroxyxanthones contain no heterocyclic nitrogen and an N-prenylation is not possible. Recently, the O-prenyltransferase SirD was shown capable of transferring prenyl moieties to N-, C- or S-atoms, proving that the prenylation position can be switched in the presence of unnatural substrates.21 A multiple sequence alignment of AstPT and seven prenyltransferases of the DMATS superfamily revealed that several conserved prenyl diphosphate binding moieties are present in the sequence of AstPT, but no suggestions can be made for the binding of the aromatic substrates or a mechanism for substrate activation. The results obtained in this study raised an important question, if bisindolyl benzoquinone and xanthone prenyltransferases share close biochemical relationships. Therefore, we tested the acceptance of AQ D by the xanthone O-prenyltransferease XptB from Aspergillus nidulans in the presence of different prenyl donors. No activity was observed in these assays (Fig. S1†).
Another interesting feature of AstPT refers to the acceptance of DMAPP, GPP and FPP in the presence of hydroxyxanthones. GPP was an even better substrate than DMAPP (Fig. 2, Tables 1 and 3). As aforementioned, GPP was not accepted by AstPT in the presence of AQ D. For the reaction of AstPT with AQ D and DMAPP, KM and kcat were determined at 33.5 μM and 0.02 s−1, respectively.8 In comparison, these values are 12.1 μM and 0.002 s−1 for the reaction with 1a and GPP providing evidence that AQ D is very likely the natural substrate for AstPT rather than hydroxyxanthones.
From the structures of the enzyme products (Fig. 2), it is obvious that OH-1 of xanthones was a poor prenylation position for AstPT, probably due to a steric hindrance by the keto group. No O1-prenylated derivative was identified in this study. This hypothesis explains the fact that only one main product was found in the enzyme assays of 1a with DMAPP, GPP and FPP each as well as those of 4a with GPP and FPP. In contrast, two products each with prenylations at OH-3, OH-6 or OH-7 were identified in the reaction mixtures of 2a and 3a. With DMAPP as prenyl donor, O6- or O7-prenylated derivative was predominant product in the reaction mixtures of 2a and 3a. In the reaction mixtures of 2a with GPP and FPP, O3-prenylated derivatives are predominant products.
For isolation and structure elucidation of the enzyme products, incubations were carried out in 50 mL at 37 °C for 16 h. The incubation mixtures contained 50 mM Tris HCl (pH 7.5), 5 mM CaCl2, 0.5 mM hydroxyxanthone derivative, 2 mM DMAPP, 0.5 mM GPP or FPP, 2.5% (v/v) DMSO, 2.6% (v/v) glycerol and 12.0 mg of purified recombinant protein.
Conversion yields were calculated from the peak area of substrates and enzyme products in the HPLC chromatogram. Yields were corrected according to the ratio between substrates and products taken from the peak area from HPLC chromatograms and integrals from 1H NMR spectra of the incubation mixtures for isolation of enzyme products.
For analysis of incubation mixtures of AQ D and XptB, water (solvent A) and acetonitrile (solvent B) were used as solvents. A linear gradient of 50–100% (v/v) solvent B in 20 min was used. The column was then washed with 100% solvent B for 5 min and equilibrated with 50% (v/v) solvent B for 5 min. Detection was carried out by a photo diode array detector and absorptions at 277 nm were illustrated.
For isolation of the enzyme products, the same HPLC equipment was used. Isolation of 1b and 2b1 + 2b2 was carried out on a Multospher 120 RP 18-5 column (250 × 4 mm, 5 μm, CS-Chromatographie Service) at a flow rate of 1 mL min−1. The same method was used as for the analysis of the enzyme assays. All other enzyme products were isolated on a Multospher 120 RP 18 HP column (250 × 10 mm, 5 μm, CS-Chromatographie Service). For isolation of 2c1 + 2c2 and 3b1 + 3b2, a linear gradient of 80–100% (v/v) solvent B in 15 min was followed by 100% (v/v) solvent B for 10 min at a flow rate of 2.5 mL min−1. The column was then equilibrated with 80% (v/v) solvent B for 5 min. For isolation of 2d1 + 2d2 and 4c, the elution profile was initiated with 90% (v/v) solvent B for 2 min at a flow rate of 3 mL min−1. Subsequently, a linear gradient of 90–100% (v/v) solvent B in 12 min was used and followed by 100% (v/v) solvent B for 5 min. The column was then equilibrated with 90% (v/v) solvent B for 5 min. The enzyme products 4d, 1c and 1d were isolated using 100% (v/v) solvent B in an isocratic run for 20 min (4d), 25 min (1c) or 33 min (1d) at a flow rate of 2.5 mL min−1.
For NMR analysis, the isolated products were dissolved in 0.75 mL of (CD3)2CO (Euriso-Top, Saarbrücken, Germany). Spectra were recorded at room temperature on a JEOL ECA-500 spectrometer (JEOL, Akishima, Tokyo, Japan). Heteronuclear multiple bond correlation (HMBC) spectrum of 1d in 0.30 mL (CD3)2SO (Euriso-Top) was recorded with standard methods26 on a Bruker Avance 500 MHz spectrometer. Spectra were processed with MestReNova.6.0.2 (Metrelab Research, Santiago de Compostella, Spain) and chemical shifts were referenced to the signal of (CD3)2CO at 2.05 ppm or to the 1H (2.50 ppm) and 13C signal (39.52 ppm) of (CD3)2SO. NMR data are given in Tables S2–S5.†
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c4ra00337c |
This journal is © The Royal Society of Chemistry 2014 |