Characterisation of the Broadly-Specific O-Methyl-transferase JerF from the Late Stages of Jerangolid Biosynthesis

We describe the characterisation of the O-methyltransferase JerF from the late stages of jerangolid biosynthesis. JerF is the first known example of an enzyme that catalyses the formation of a non-aromatic, cyclic methylenolether. The enzyme was overexpressed in E. coli and the cell-free extracts were used in bioconversion experiments. Chemical synthesis gave access to a series of substrate surrogates that covered a broad structural space. Enzymatic assays revealed a broad substrate tolerance and high regioselectivity of JerF, which makes it an attractive candidate for an application in chemoenzymatic synthesis with particular usefulness for late stage application on 4-methoxy-5,6-dihydro-2H-pyran-2-one-containing natural products.


Introduction
α-Pyrones and γ-pyrones are abundant structures in microbial natural products. Prominent examples for biologically active α-pyrones are coumarin and isocoumarin derivatives, which are extensively used as fragrances, anticoagulants or rodenticides [1]. More recently, 4-hydroxyl-α-pyrones have also been identified as signalling molecules in bacterial communication [2].
O-Methylated pyrones occur in polyketide natural products with highly interesting biological activity such as enterocin (2) and (+)-(R)-aureothin (6), which are produced by Streptomyces thioluteus and Streptomyces maritimus, respectively (see Scheme 1). For both compounds, methyl transfer takes place during the final tailoring steps of their biosynthesis and is catalyzed by class I S-(5 -adenosyl)-L-methionine (SAM)-dependent O-methyltransferases (O-MTs; EncK for 2, AurI for 6) [3]. Although EncK and AurI possess very high sequence similarity, they react with different chemoselectivity on the same structural element [4,5].

Synthesis of Substrate Surrogates
To confirm the postulated biosynthetic role of JerF and to evaluate its potential for chemoenzymatic synthesis, we set out to investigate the catalytic activity of JerF in vitro. Synthetic precursor surrogates of varying complexity (rac-14a-h) were synthesised that covered a broad structural space (Scheme 2; Figures S1-S26). Two series of compounds were obtained, containing either the 3-methyl-6-vinyldihydro-2H-pyran-2,4(3H)-dione (rac-14d-h) that is present in the biosynthetic precursor or 3-desmethyl analogs (rac-14a-c). All substrates were readily available by vinylogous aldol reaction of β-ketoester enolates 11a or 11b, respectively, followed by lactonization under basic conditions (Scheme 2a) [22][23][24][25]. The relative orientation of the substituents on ring positions 3 and 6 in rac-14d-h was shown to be predominantly syn, according to 1 H-NMR spectroscopy and NOE correlation spectroscopy (see NMR spectra and Figure S26 in the Supplementary Materials). The methylenolethers rac-15a-g were synthesised from rac-14a-g by O-methylation using MeI and K2CO3 or NaH, respectively. To unambiguously assign the methylation activity to the jerF gene product, we conducted a series of control experiments. Methylation occurred only in the presence of MgCl2, SAM tosylate and the lysate from the jerF-pColdI expression (Figure 1f). If any of these components was left out, no formation of rac-15d was obtained (Figure 1b,c). The same was true if the lysate was denatured by heat treatment prior assaying, which attributes the activity to a component of the cell lysate ( Figure 1d). The lysate of a pColdI vector expression (devoid of jerF) in E. coli BL21 did also not cause the formation of rac-15d, clearly highlighting that the expression product of jerF is responsible for the observed activity ( Figure 1e).

Comparative Assaying of Synthetic Substrates and Assay Upscaling
Under the conditions established for rac-14d, compound rac-14h was completely methylated, highlighting the broad substrate tolerance of JerF ( Figure 2). The higher conversion compared to rac-14d reflects the closer structural similarity of rac-14h to the proposed biosynthetic precursor 9.
Insertion of the other synthetic substrate surrogates rac-14b, rac-14c and rac-14e-g under similar conditions showed partial conversion of all substrates containing a 3-Me group (rac-14e-g) To unambiguously assign the methylation activity to the jerF gene product, we conducted a series of control experiments. Methylation occurred only in the presence of MgCl 2 , SAM tosylate and the lysate from the jerF-pColdI expression (Figure 1f). If any of these components was left out, no formation of rac-15d was obtained (Figure 1b,c). The same was true if the lysate was denatured by heat treatment prior assaying, which attributes the activity to a component of the cell lysate ( Figure 1d). The lysate of a pColdI vector expression (devoid of jerF) in E. coli BL21 did also not cause the formation of rac-15d, clearly highlighting that the expression product of jerF is responsible for the observed activity ( Figure 1e).

Comparative Assaying of Synthetic Substrates and Assay Upscaling
Under the conditions established for rac-14d, compound rac-14h was completely methylated, highlighting the broad substrate tolerance of JerF ( Figure 2). The higher conversion compared to rac-14d reflects the closer structural similarity of rac-14h to the proposed biosynthetic precursor 9.
Molecules 2016, 21,1443 5 of 23 acts highly chemo-and regioselectively. However, for the non-branched substrates rac-14b and rac-14c, the absolute amount of formed product was unexpectedly low. Overnight incubation of compounds rac-14b, rac-14c and rac-14e-g in the absence of JerF revealed that all substrates undergo slow, spontaneous degradation at pH 8.8. This trend is more pronounced for the non-branched lactones rac-14b and rac-14c and seems to be accelerated by uncharacterised components of the lysate. Accordingly, attempts to conduct the reaction with rac-14d on the semi-preparative scale (up to 7 mg starting material) led to hardly reproducible results and yields below 10%.
A markedly increased stability of the lactones as well as the corresponding methylenolethers rac-15b, rac-15c and rac-15e-g was observed at near-neutral pH. In comparative enzymatic assays with compound mixture rac-14e and JerF at different pH values, the best results in terms of conversion and compound stability were observed at pH 7.5 ( Figure S59). The experiments were thus repeated at this pH value with substrates rac-14b, rac-14c and rac-14e-g. Full conversion was reproducibly obtained for compounds rac-14b, rac-14c, rac-14e and rac-14f ( Figure 3, for unprocessed spectra see Figures S60-S64). The compound mixture rac-14g was also methylated to a large extend, however the presence of small amounts of starting material was still visible. The 3-desmethyl substrates rac-14b and rac-14c were fully methylated according to HPLC-MS analysis. However, complete degradation of rac-14b and rac-14c without any conversion into rac-15b and rac-15c was observed in some repetitions of the experiment, indicating that destructive side reactions caused by the lysate could be partially responsible for this positive result. (a) synthetic rac-14h after overnight incubation in buffer; (b) incubation of rac-14h with cell-free extract containing JerF; (c) incubation of rac-14h with cell-free extract containing JerF and SAM tosylate in reaction buffer. The minimal conversion is probably caused by residual amounts of SAM in the cell-free extract. (d) HR-MS analysis of the assay product of (b); (e) HR-MS analysis of the assay product of (c); Conditions of the conversion assay: 0.25 mM rac-14h, 5 mM MgCl 2 , 1.0 mM SAM-tosylate, 0.2 mL cell-free extract (3.0 mg/mL), 28 • C, 16 h, pH 8.8.
Insertion of the other synthetic substrate surrogates rac-14b, rac-14c and rac-14e-g under similar conditions showed partial conversion of all substrates containing a 3-Me group (rac-14e-g) and full conversion of the non-branched substrates rac-14b and rac-14c ( Figures S54-S58). In all cases, an exclusive methylation on the 4-O and no reaction on 2-O or C-3 occurred, showing that JerF acts highly chemo-and regioselectively. However, for the non-branched substrates rac-14b and rac-14c, the absolute amount of formed product was unexpectedly low.
Overnight incubation of compounds rac-14b, rac-14c and rac-14e-g in the absence of JerF revealed that all substrates undergo slow, spontaneous degradation at pH 8.8. This trend is more pronounced for the non-branched lactones rac-14b and rac-14c and seems to be accelerated by uncharacterised components of the lysate. Accordingly, attempts to conduct the reaction with rac-14d on the semi-preparative scale (up to 7 mg starting material) led to hardly reproducible results and yields below 10%.
A markedly increased stability of the lactones as well as the corresponding methylenolethers rac-15b, rac-15c and rac-15e-g was observed at near-neutral pH. In comparative enzymatic assays with compound mixture rac-14e and JerF at different pH values, the best results in terms of conversion and compound stability were observed at pH 7.5 ( Figure S59). The experiments were thus repeated at this pH value with substrates rac-14b, rac-14c and rac-14e-g. Full conversion was reproducibly obtained for compounds rac-14b, rac-14c, rac-14e and rac-14f ( Figure 3, for unprocessed spectra see Figures S60-S64). The compound mixture rac-14g was also methylated to a large extend, however the presence of small amounts of starting material was still visible. The 3-desmethyl substrates rac-14b and rac-14c were fully methylated according to HPLC-MS analysis. However, complete degradation of rac-14b and rac-14c without any conversion into rac-15b and rac-15c was observed in some repetitions of the experiment, indicating that destructive side reactions caused by the lysate could be partially responsible for this positive result.  Reaction upscaling was also more successful at pH 7.5 and gave reproducible results (Table 1, Figures S43-S47). Approximately 4 mg of compounds rac-14c, rac-14e and rac-14f were individually incubated with the cell-free extract from a jerF expression in a total assay volume of 10 mL. Conversions of 27%-42% into the respective methylenolethers rac-15c, rac-15e and rac-15f were obtained. The crude products of rac-15e and rac-15f were partially purified by column chromatography on silica gel. In both cases, an aliphatic impurity was co-purified, which could not be removed.  for m/z of substrates rac-14b, rac-14c and rac-14e-g (blue traces) and the respective O-methylated products rac-15b, rac-15c and rac-15e-g (red traces). General conditions of the conversion assays: cell-free extract of the JerF expression ( Reaction upscaling was also more successful at pH 7.5 and gave reproducible results (Table 1, Figures S43-S47). Approximately 4 mg of compounds rac-14c, rac-14e and rac-14f were individually incubated with the cell-free extract from a jerF expression in a total assay volume of 10 mL. Conversions of 27%-42% into the respective methylenolethers rac-15c, rac-15e and rac-15f were obtained. The crude products of rac-15e and rac-15f were partially purified by column chromatography on silica gel. In both cases, an aliphatic impurity was co-purified, which could not be removed. The conversion was determined by integration of the signals of the protons at position 6 of the dione in the 1 H-NMR spectra; b The e.e. was calculated from the peak areas of the chiral HPLC chromatograms. c Crude yield is given after column chromatography on silica gel. A non-removable, aliphatic impurity was co-purified (see Figures S45 and S47).
The crude products of the semi-preparative conversions as well as the samples from column chromatography were analysed by chiral HPLC. In the cases of rac-14c and rac-14e, the racemic starting material was converted into product enriched in one stereoisomer with an enantiomeric ratio of 92:8 and 71:29, respectively (Table 1). For rac-14f, an only negligible enantiomeric excess was observed. JerF thus discriminates between the inserted stereoisomers, however, with a strong dependence on the substrate structure.

Discussion
We were able to characterise the O-methyltransferase JerF from jerangolid biosynthesis by assaying of the enzyme in bioconversion experiments with synthetic substrate surrogates. JerF is the first characterised case of an O-methyltransferase that forms a cyclic, non-aromatic methylenolether. The enzyme shows promising substrate tolerance. It accepts aliphatic and aromatic residues (R in Scheme 3) of varying size as well as substrates that lack the methyl group on C-3, which is present on the natural precursor. Furthermore, the enzyme is fully selective for methylation on 4-O and is not reactive towards the other two potential methylation sites at 2-O and C-3. During chemical synthesis of substrates rac-15a-h, only C-3-methylated side products were obtained, suggesting that an inherent selectivity for 4-O-methylation exists under the conditions applied. In this context, it would be interesting to evaluate if the O-MT AurI keeps its confirmed 2-O-selectivity in the reaction with this kind of substrates. The crude products of the semi-preparative conversions as well as the samples from column chromatography were analysed by chiral HPLC. In the cases of rac-14c and rac-14e, the racemic starting material was converted into product enriched in one stereoisomer with an enantiomeric ratio of 92:8 and 71:29, respectively (Table 1). For rac-14f, an only negligible enantiomeric excess was observed. JerF thus discriminates between the inserted stereoisomers, however, with a strong dependence on the substrate structure.

Discussion
We were able to characterise the O-methyltransferase JerF from jerangolid biosynthesis by assaying of the enzyme in bioconversion experiments with synthetic substrate surrogates. JerF is the first characterised case of an O-methyltransferase that forms a cyclic, non-aromatic methylenolether. The enzyme shows promising substrate tolerance. It accepts aliphatic and aromatic residues (R in Scheme 3) of varying size as well as substrates that lack the methyl group on C-3, which is present on the natural precursor. Furthermore, the enzyme is fully selective for methylation on 4-O and is not reactive towards the other two potential methylation sites at 2-O and C-3. During chemical synthesis of substrates rac-15a-h, only C-3-methylated side products were obtained, suggesting that an inherent selectivity for 4-O-methylation exists under the conditions applied. In this context, it would be interesting to evaluate if the O-MT AurI keeps its confirmed 2-O-selectivity in the reaction with this kind of substrates. Reactions with the enzyme could be conveniently performed using the cell-free extract from a jerF expression. Reactions on the analytical scale with substrates rac-14a-h proceeded with high to complete conversion. Problems arising from slow spontaneous degradation during the reaction could be reduced by performing the reaction at pH 7.5.
Upscaling of the enzymatic reactions with substrate mixtures rac-14c, rac-14e and rac-14f was successful at pH 7.5, giving conversions of 27%-42%. A further optimisation of the reaction conditions will probably improve this result. It is furthermore known that SAM-dependent MTs are often inhibited by S-adenosylhomocysteine (SAH) that is formed during the reaction. Addition of a SAH-hydrolase or SAH nucleosidase could thus also be helpful. Analysis of the semi-preparative scale conversions by chiral HPLC revealed a discrimination of JerF between the inserted stereoisomers. The degree of selectivity strongly depended on the constitution of the inserted substrates.
The insights gained about relevant features of JerF like substrate tolerance and chemoselectivity suggest further investigations on its applicability in chemoenzymatic synthesis. Future studies will concentrate on a thorough optimisation of the enzyme overexpression and the reaction conditions on the analytical and the (semi)preparative scale. Further studies on its substrate tolerance will help Scheme 3.
Reactions with the enzyme could be conveniently performed using the cell-free extract from a jerF expression. Reactions on the analytical scale with substrates rac-14a-h proceeded with high to complete conversion. Problems arising from slow spontaneous degradation during the reaction could be reduced by performing the reaction at pH 7.5.
Upscaling of the enzymatic reactions with substrate mixtures rac-14c, rac-14e and rac-14f was successful at pH 7.5, giving conversions of 27%-42%. A further optimisation of the reaction conditions will probably improve this result. It is furthermore known that SAM-dependent MTs are often inhibited by S-adenosylhomocysteine (SAH) that is formed during the reaction. Addition of a SAH-hydrolase or SAH nucleosidase could thus also be helpful. Analysis of the semi-preparative scale conversions by chiral HPLC revealed a discrimination of JerF between the inserted stereoisomers. The degree of selectivity strongly depended on the constitution of the inserted substrates.
The insights gained about relevant features of JerF like substrate tolerance and chemoselectivity suggest further investigations on its applicability in chemoenzymatic synthesis. Future studies will concentrate on a thorough optimisation of the enzyme overexpression and the reaction conditions on the analytical and the (semi)preparative scale. Further studies on its substrate tolerance will help to evaluate the scope of the enzyme. Finally, the enzyme will be applied in the chemoenzymatic total synthesis of 4-methoxy-5,6-dihydro-2H-pyran-2-one-containing natural products.

Chemistry Methods and Materials
All reactions were performed in oven dried glassware under an atmosphere of nitrogen gas unless otherwise stated. Dry solvents were purchased from Sigma-Aldrich Chemie GmbH (Steinheim, Germany) and Acros (Geel, Belgium) or taken out of a solvent system from M. Braun (Garchingen, Germany). Dry reagents were ordered from Sigma-Aldrich, Arcos, abcr GmbH (Karlsruhe, Germany) and Roth (Karlsruhe, Germany). NMR spectra were recorded with DRX-500, DPX-400 and AVANCE-400 instruments (Bruker, Billerica, MA, USA) with the residual solvent signal as internal standard (CHCl 3 = 7.26 ppm). multiplicities are described using the following abbreviations: s = singlet, d = doublet, t = triplet, q = quartet, m = multiplet, b = broad. 13 C-NMR spectra are reported as values in ppm relative to residual solvent signal (CHCl 3 = 77.06 ppm) as internal standards. The multiplicities are elucidated using the distortionless enhancement by polarisation transfer (DEPT) spectral editing technique, with secondary pulses at 90 • and 135 • . Multiplicities are reported using the following abbreviations: q (quarternary carbon), t (tertiary carbon = methine), s (secondary carbon = methylene), p = (primary carbon = methyl). High resolution mass spectra are obtained with a Micromass LCT via loop-mode injection from an Alliance 2695 HPLC system (Waters, Milford, MA, USA). Alternatively, a Micromass Q-TOF in combination with a Waters Acquity Ultra performance LC system is employed. Ionisation is achieved by ESI or APCI. Modes of ionisation, calculated and found mass are given. Reversed phase-HPLC-applications were performed with membrane-filtrated and double distilled water as well as commercial available HPLC-grade solvents (methanol or acetonitrile). Semi-preparative HPLC was performed with a Merck Hitachi

Biochemistry Methods and Materials
All chemicals and antibiotics were purchased from Sigma-Aldrich and Roth. Cell disruption was conducted by sonication (Sonoplus Typ UW3100) from Bandelin (Berlin, Germany). His-bind nickel chelate chromatography resin was purchased from Novagen. Millipore Amicon ® ultra Molecules 2016, 21, 1443 9 of 23 centrifugal filters (10,000 and 30,000 MW Cut-off) and PD-10 desalting columns from GE Healthcare (Buckinghamshire, UK) were used for protein concentration and buffer exchange respectively.

General Procedures
Aldol Reaction A solution of LDA was freshly prepared by adding n-BuLi (2.5 M in hexane, 2.5 equiv.) to diisopropylethylamine (0.7 M, 2.5 equiv.) in THF at −78 • C. The solution was stirred at room temperature for 30 min and after cooling to −78 • C, DMPU (1.0 equiv.) was added. Methyl-2-methyl-3-oxobutanoate or methyl-3-oxobutanoate (1 M, 1.0 equiv.), respectively, in THF was added and the solution was stirred for further 50 min. The aldehyde (1.1 equiv.) was added and the solution stirred for 2 h. The reaction was quenched by addition of 2 M HCl. After separation of the layers, the aqueous layer was extracted by Et 2 O and the combined organic layers were dried over MgSO 4 . The solvent was removed in vacuo.

Lactonization
The product of the aldol reaction was dissolved in 1 M KOH and stirred at room temperature for 5 h. After cooling to 0 • C, 2 M HCl was added until a pH value of 0 was reached. The resulting solid was filtrated, washed with H 2 O and the desired product purified by column chromatography.

O-Methylation of the Dihydropyran-2,4-diones
After dissolving the lactone (0.2 M, 1.0 equiv.) in DMF abs , the solution was cooled to 0 • C and MeI (1.0 equiv.) and K 2 CO 3 (1.5 equiv.) were added. After 1 h, the solution was warmed to room temperature and stirred overnight. The layers were separated and the aqueous layer was extracted with EtOAc. The combined organic layers were washed with brine, dried over MgSO 4 and the solvent was removed in vacuo. The crude product was purified by column chromatography.

O-Methylation of the 3-Methyldihydropyran-2,4-diones
After dissolving the lactone (0.2 M, 1.0 equiv.) in THF abs , the solution was cooled to 0 • C and NaH (60% in mineral oil, 1.3 equiv.) was added. After 1 h, MeI (1.2 equiv.) was added and the solution was warmed to room temperature. After stirring overnight, the reaction was quenched by the addition of H 2 O. After separation of the layers, the aqueous layer was extracted with EtOAc, the combined organic layers were washed with brine and dried over MgSO 4 . The solvent was removed in vacuo and the crude product was purified by column chromatography.

Substrate Synthesis
Compounds rac-14a-g and rac-15b-h were synthesised according to the route shown in Schemes 4-6.

Lactonization
The product of the aldol reaction was dissolved in 1 M KOH and stirred at room temperature for 5 h. After cooling to 0 °C, 2 M HCl was added until a pH value of 0 was reached. The resulting solid was filtrated, washed with H2O and the desired product purified by column chromatography.

O-Methylation of the Dihydropyran-2,4-diones
After dissolving the lactone (0.2 M, 1.0 equiv.) in DMFabs, the solution was cooled to 0 °C and MeI (1.0 equiv.) and K2CO3 (1.5 equiv.) were added. After 1 h, the solution was warmed to room temperature and stirred overnight. The layers were separated and the aqueous layer was extracted with EtOAc. The combined organic layers were washed with brine, dried over MgSO4 and the solvent was removed in vacuo. The crude product was purified by column chromatography.

O-Methylation of the 3-Methyldihydropyran-2,4-diones
After dissolving the lactone (0.2 M, 1.0 equiv.) in THFabs, the solution was cooled to 0 °C and NaH (60% in mineral oil, 1.3 equiv.) was added. After 1 h, MeI (1.2 equiv.) was added and the solution was warmed to room temperature. After stirring overnight, the reaction was quenched by the addition of H2O. After separation of the layers, the aqueous layer was extracted with EtOAc, the combined organic layers were washed with brine and dried over MgSO4. The solvent was removed in vacuo and the crude product was purified by column chromatography.
Under nitrogen atmosphere, sodium hydride (60% suspension, 900 mg, 22.5 mmol, 1.2 equiv.) was suspended in 40 mL dry THF and cooled to 0 °C. To this reaction mixture, methyl acetoacetate (2.0 mL, 18.7 mmol, 1.0 equiv.) was added dropwise. After gas evolution had ceased, the reaction was cooled to −78 °C followed by dropwise addition of n-BuLi (8.3 mL, 20.6 mmol, 1.1 equiv., 2.5 M in hexanes). The reaction was allowed to warm to 0 °C for 30 min and then cooled again to −78 °C. (E)-crotonaldehyde (1.7 mL, 20.6 mmol, 1.1 equiv.) was added dropwise over 5 min and the reaction was stirred for 30 min at r.t. The reaction was quenched by addition of 20 mL saturated NH4Cl/H2O (1:1) at 0 °C. The aqueous solution was washed twice with 50 mL EtOAc and then acidified to pH 1 using concentrated hydrochloric acid. The resulting precipitate was dissolved in 20 mL EtOAc and the aqueous phase was extracted twice with 20 mL EtOAc. The combined organic phases were dried over Na2SO4 and filtered. The solvents were removed under reduced pressure. Compound rac-14a was obtained as a pale-yellow solid (20%, 588 mg, 4.50 mmol).
Under nitrogen atmosphere, sodium hydride (60% suspension, 900 mg, 22.5 mmol, 1.2 equiv.) was suspended in 40 mL dry THF and cooled to 0 • C. To this reaction mixture, methyl acetoacetate (2.0 mL, 18.7 mmol, 1.0 equiv.) was added dropwise. After gas evolution had ceased, the reaction was cooled to −78 • C followed by dropwise addition of n-BuLi (8.3 mL, 20.6 mmol, 1.1 equiv., 2.5 M in hexanes). The reaction was allowed to warm to 0 • C for 30 min and then cooled again to −78 • C. (E)-crotonaldehyde (1.7 mL, 20.6 mmol, 1.1 equiv.) was added dropwise over 5 min and the reaction was stirred for 30 min at r.t. The reaction was quenched by addition of 20 mL saturated NH 4 Cl/H 2 O (1:1) at 0 • C. The aqueous solution was washed twice with 50 mL EtOAc and then acidified to pH 1 using concentrated hydrochloric acid. The resulting precipitate was dissolved in 20 mL EtOAc and the aqueous phase was extracted twice with 20 mL EtOAc. The combined organic phases were dried over Na 2 SO 4 and filtered. The solvents were removed under reduced pressure. Compound rac-14a was obtained as a pale-yellow solid (20%, 588 mg, 4.50 mmol).
Crude methyl ester rac-13d (414 mg, 2.10 mmol, 1.0 equiv.) was dissolved in 20 mL 1 M potassium hydroxide solution and was stirred at r.t. for 7 h. The solution was cooled to 0 • C followed by addition of 6 M hydrochloric acid to pH 1.0. The resulting precipitate was filtered using a glass frit and the obtained crystals were dried under reduced pressure to give compound rac-14d as yellow crystals (263 mg, 1.50 mmol, 70%, syn:anti = 5:1).
Methyl-(R)-3-((tert-Butyldimethylsilyl)oxy)-2-methylpropanoate (I) (R)-Methyl-3-hydroxy-2-methylpropionate (10.0 mL, 90.0 mmol, 1.0 equiv.), DMAP (110 mg, 0.90 mmol, 0.01 equiv.) and imidazole (9.9 g, 144 mmol, 1.6 equiv.) were dissolved in 100 mL CH 2 Cl 2 . TBSCl was dissolved in 20 mL CH 2 Cl 2 and added dropwise to the reaction mixture at 0 • C. The resulting suspension was stirred at r.t. for 1 h. The reaction was quenched by the addition of 150 mL H 2 O. The aqueous phase was extracted three times with 50 mL CH 2 Cl 2 . The combined organic phases were washed once with H 2 O and dried over MgSO 4 . After filtration, the solvent was removed under reduced pressure. The compound I was obtained as a pale-yellow oil (20.1 g) and was used without further purification. To a suspension of bromo(isopropyl)triphenyl-λ5-phosphane (2.95 g, 7.65 mmol, 2 equiv.) in 15 mL THF was slowly added n-BuLi (2.60 mL, 6.50 mmol, 1.7 equiv., 2.5 M in hexane) at −78 • C. It was stirred at room temperature for 30 min until a dark-red colour appeared. It was cooled to −78 • C and aldehyde III (774 mg, 3.83 mmol, 1 equiv.) in a small amount of THF was added dropwise. After 16 h, water was added to the white suspension. The aqueous layer was three times extracted with CH 2 Cl 2 . The combined organic layers were dried over MgSO 4 and the solvent removed in vacuo. After column chromatography on silica gel (PE), compound IV (600 mg, 2.63 mmol, 69% yield) was obtained.
(S)-2,4-Dimethylpent-3-en-1-ol (V) Alkene IV (570 mg, 2.50 mmol, 1 equiv.) was solved in 10 mL MeOH and PPTS (3.76 g, 14.97 mmol, 6 equiv.) was added. After stirring for 24 h at 50 • C, the solvent was removed under reduced pressure. The residual solid was washed with 3 mL water and the aqueous layer was three times extracted with CH 2 Cl 2 . The combined organic layers were washed with saturated NaHCO 3 solution and saturated NaCl solution and then dried over MgSO 4 . After removal of the solvent under reduced pressure and column chromatography on silica gel (PE/Et 2 O = 5:1), alcohol V (262 mg, 2.30 mmol, 92% yield) was obtained as a colourless oil.
(R,E)-4,6-Dimethylhepta-2,5-dien-1-ol (VIII) Ester VII (65 mg, 0.36 mmol, 1 equiv.) was solved in 6 mL THF and cooled to −78 • C. A solution of DIBAL-H (1.10 mL, 1.10 mmol, 3.08 equiv., 1 M in hexane) that was cooled to 0 • C was added dropwise. It was stirred for 2 h at −78 • C, for 1 h at room temperature, 3 mL of a solution of saturated K-Na-tartrate solution was added together with 1 mL Et 2 O. It was stirred for further 23 h. The aqueous layer was three times extracted with Et 2 O. The combined organic layers were washed with saturated NaCl solution, dried over MgSO 4 and the solvent was removed under reduced pressure. After column chromatography on silica gel (PE/Et 2 O = 5:1), alcohol VIII (47 mg, 0.34 mmol, 94% yield) was isolated.
((R,E)-3,5-Dimethylhexa-1,4-dien-1-yl)-3-methyldihydro-2H-pyran-2,4(3H)-dione (rac-14h) Under argon atmosphere, diisopropylamine (52 µL, 0.37 mmol, 2.5 equiv.) was dissolved in 0.5 mL dry THF and the resulting solution was cooled to −78 • C. To this solution, n-BuLi (150 µL, 0.37 mmol, 2.5 M in hexanes, 2.5 equiv.) was added dropwise. The reaction was allowed to warm to 0 • C for 30 min. At −78 • C, DMPU (19 µL, 0.15 mmol, 1.0 equiv.) was added dropwise and the resulting mixture was stirred for 30 min. Methyl-3-oxobutanoate (19.4 mg, 0.15 mmol, 1.0 equiv.) was dissolved in 250 µL dry THF and this solution was added to the reaction mixture followed by stirring for 1 h at −78 • C. Aldehyde IX (21 mg, 0.15 mmol, 1.0 equiv.), dissolved in 250 µL dry THF, was then added followed by stirring for 2 h at −78 • C. The reaction was quenched by addition of 1 mL 3 M hydrochloric acid at −78 • C and allowed to warm to r.t. The aqueous phase was extracted three times with 10 mL Et 2 O. The combined organic phases were dried over Na 2 SO 4 and filtered. The solvents were removed under reduced pressure. The crude product was taken up in 4 mL 1 M potassium hydroxide solution and the resulting solution was stirred for 7 h at r.t. At 0 • C, the solution was acidified to pH 1.0 by addition of 6 M hydrochloric acid. The aqueous phase was extracted three times with 10 mL EtOAc. The combined organic phases were dried over Na 2 SO 4 and filtered. The solvents were removed under reduced pressure. Purification by flash column chromatography (EtOAc/PE = 1:10) yielded 40% of compound rac-14h as a pale-yellow oil (16 mg, 0.15 mmol).

Analytical Data
The solvents were removed under reduced pressure. The crude product was taken up in 4 mL 1 M potassium hydroxide solution and the resulting solution was stirred for 7 h at r.t. At 0 °C, the solution was acidified to pH 1.0 by addition of 6 M hydrochloric acid. The aqueous phase was extracted three times with 10 mL EtOAc. The combined organic phases were dried over Na2SO4 and filtered. The solvents were removed under reduced pressure. Purification by flash column chromatography (EtOAc/PE = 1:10) yielded 40% of compound rac-14h as a pale-yellow oil (16 mg, 0.15 mmol).  potassium hydroxide solution and the resulting solution was stirred for 7 h at r.t. At 0 °C, the solution was acidified to pH 1.0 by addition of 6 M hydrochloric acid. The aqueous phase was extracted three times with 10 mL EtOAc. The combined organic phases were dried over Na2SO4 and filtered. The solvents were removed under reduced pressure. Purification by flash column chromatography (EtOAc/PE = 1:10) yielded 40% of compound rac-14h as a pale-yellow oil (16 mg, 0.15 mmol).