Convenient synthesis of C 75 , an inhibitor of FAS and CPT 1 †

C75 is a synthetic racemic α-methylene-γ-butyrolactone exhibiting antitumoural properties in vitro and in vivo as well as to inducing hypophagia and weight loss in rodents. These interesting properties are thought to be a consequence of the inhibition of the key enzymes FAS and CPT1 in lipid metabolism. The need for larger amounts of this compound for biological evaluation prompted us to develop a convenient and reliable route to multigram quantities of C75 from easily available ethyl penta-3,4-dienoate 6. A recently described protocol for the addition of 6 to a mixture of dicyclohexylborane and nonanal followed by acidic treatment of the crude afforded lactone 8, as a mixture of cis and trans isomers, in good yield. The DBU-catalyzed isomerization of the methyl esters 9 arising from 8 gave a 10:1 trans/cis mixture from which the trans isomer was isolated and easily transformed into C75. The temporary transformation of C75 into a phenylseleno ether derivative makes its purification, manipulation and storage easier.


Table of contents entry
A new approach to the enzyme inhibitor C75 and its temporary transformation into a phenylseleno ether derivative is disclosed.This procedure facilitates the purification, manipulation and storage of C75.

Introduction
C75 is a synthetic racemic compound that acts as a potent inhibitor of mammalian type I fatty acid synthase (type I FAS) and, in its CoA-adducted form, inhibits carnitine palmitoyltransferase 1 (CPT1) activity in vivo.Over the last few years, C75 as a racemic mixture of its trans diastereoisomers has been used extensively in the study of fatty acid synthesis in metabolic disorders and cancer. 1 In the course of our ongoing search for potential new drugs against obesity and type 2 diabetes, 2 we required multigram quantities of racemic trans C75 for biological testing and as starting material for the preparation of new analogues.Although C75 is commercially available, its very high price prompted us to consider the preparation of C75 in our own laboratory.Surprisingly, only one procedure for its preparation has been reported by Kuhajda et al. 3 This procedure was based on previous work 4 by Carlson et al. that uses itaconic anhydride and p-methoxybenzyl alcohol as starting materials.Under these conditions C75 is obtained as a separable mixture of C75 and its cis isomer 1.In our hands, the laborious column chromatography process required for this separation caused significant isomerization of the exocyclic double bond to the endocyclic position to give inactive isomer 2 (Fig. 1).Herein, we proposed an alternative straightforward scalable method for the preparation of C75 and its separation from 1 and 2. According to our retrosynthetic analysis of C75, the methylene group would be introduced in the last step from lactone trans-4 that would arise from the appropriate γ-hydroxyester anti-5 (Scheme 1).Taking advantage of our recent studies on the stereoselective addition of allenes to aldehydes, 5 we envisaged that the hydroboration of ethyl penta-3,4-dienoate 6 with dicyclohexylborane in the presence of nonanal could afford the required anti-5. 6rein, we report our findings in this respect.

Results and discussion
The allylation and crotylation of aldehydes with 2-alkenyl boranes provide attractive and versatile routes to homoallylic secondary alcohols.2-Alkenylides derived from alkali and alkaline earth metals generally serve as the starting points for the generation of 2-alkenyl boranes.However, hydroboration of allenes might be a milder alternative for obtaining such boron reagents.Initially, the borane adds to the sterically less hindered face of the allene to form a transient (Z)-2-alkenylborane.However, this kinetically formed borane reagent suffers fast isomerization to the thermodynamically more stable (E)-2-alkenylborane at room temperature.The addition of an aldehyde to this then affords a syn homoallylic alcohol as the major product, through a 6-membered transition state.Recently, we demonstrated that the hydroboration of the allene is also possible in the presence of the aldehyde.5b In such conditions, the (Z)-borane reagent can be trapped immediately by the aldehyde leading to anti adducts.Accordingly, we expected that the treatment of a mixture of ethyl penta-3,4-dienoate 6 and nonanal with Chx 2 BH would led to anti-5 by trapping the transient (Z)-2-alkenylborane, (Z)-7 (Scheme 2).In contrast, the hydroboration of 6 with dicyclohexylborane (Chx 2 BH) followed by the addition of nonanal would afford selectively the γ-hydroxyester syn-5.In practice, the addition of allene 6 to a mixture of Chx 2 BH (1.2 equiv) and nonanal (1.2 equiv) in CH 2 Cl 2 gave, after work-up (triethanolamine) and chromatographic separation, a separable mixture of hydroxyester syn-5 and the lactone trans-8 arising from anti-5 but in a disappointing 6:4 ratio (Scheme 3).Apparently, the minor anti-5 isomer formed suffers a selective lactonization during the work-up and/or chromatographic processes to give trans-8.The unexpected abundance of the syn isomer suggests that after the hydroboration step, isomerization of (Z)-7 to (E)-7 was faster than we expected.It is known that the allylic isomerization of the crotylboranes strongly depends on the steric hindrance of both allylic positions and the isomerization in (Z)-7 seems to be easier than in the more hindered protected allenols used in our previous work.5b When we heated the mixture in MeOH to reflux with acid catalysis, only 8 was obtained in 68% overall yield as a 6:4 cis/trans mixture that was difficult to separate.As expected, isolated syn-5 afforded only cis-8 under acidic conditions.In an effort to improve the proportion of trans-8, the experimental conditions were revised.
Lower temperatures (0 ºC or -20 ºC) slowed down Z/E isomerization of 7 but hydroboration was incomplete and yields diminished.Neither changing the solvent (THF, Et 2 O), or workup (triethanolamine or H 2 O 2 ) nor a fine tuning of the allene/aldehyde/Chx 2 BH ratio were satisfactory.Consequently, we considered the use of more hindered hydroborating agents.Both (-)-Ipc 2 BH 7 (Ipc = isopinocampheyl), and disiamylborane 8 were tested since these bulky reagents are less prone to undergo boratropic isomerization providing a greater opportunity for trapping the Z reagent.Furthermore, by using an enantiopure boron reagent, the reaction could, in principle, become enantioselective.Unfortunately, neither of the reagents improved the trans-8/syn-5 ratio significantly but also led to lower yields (40-45%).In addition, the GC analysis of the trans-8 obtained with (-)-Ipc 2 BH revealed low enantioselectivities (55:44 ratio).On the other hand, when we performed the alternative stepwise process for comparative purposes, i.e. hydroboration of 6 (CH 2 Cl 2 or THF) with (Chx) 2 BH (1.2 equiv.)followed by addition of nonanal at -78 ºC, only complex reaction mixtures containing compounds 5 or 8 in low yield (<15%) were obtained.These results are in sharp contrast with previous ones both from our laboratory 5 and others 9 in which terminal allenes undergo addition to aldehydes in good yields under similar conditions.Our results suggest that, although the starting allene readily disappeared in the hydroboration step (as observed by TLC), the resulting alkenylborane decomposed at 0 ºC giving a mixture of by-products. 10nsequently, we turned our attention to the former one-pot hydroboration with Chx 2 BH (Scheme 3).Having in hand the mixture of lactones 8, the oxidation 11 of the vinyl group was performed with NaIO 4 /RuCl 3 to afford almost quantitatively a mixture of acids 4.
Gratifyingly the transformation of 4 into methyl esters 9 allowed us to improve the diastereomeric ratio to 10:1 trans/cis by heating the mixture with DBU in dry toluene.The compound trans-9 was then easily isolated chromatographically and then hydrolyzed to obtain >95% stereopure trans-4.The transformation of trans-4 into C75 was then quite straightforward (Scheme 5) and its methylenation by a known procedure 12 completed the synthesis of C75 in a very respectable 29% overall yield.
Although C75 can be stored for months at 0 ºC with no loss of activity, it is prone to suffer facile isomerization to the more stable fully conjugated lactone 2 in solution or during chromatographic purifications on silica gel.We envisaged that a phenylseleno ether derivative should be an appropriate protecting group for such a delicate exocyclic double bond.After storage and/or purification, C75 could then be regenerated under mild oxidative conditions. 13This was confirmed by treating a sample of C75 with PhSeSePh and NaBH 4 in EtOH to afford the seleno derivative 3 in 90% yield.It should be noted that 2 remains unchanged under these conditions. 14Derivative 3 could be easily purified by column chromatography 15 and stored in a closed flask at rt for more than one year.Its treatment with 30% H 2 O 2 in THF at rt then regenerated the desired C75.In conclusion, we disclose a useful new method of preparation of C75 in 29% overall yield based on a hydroboration ethyl penta-3,4-dienoate/addition of nonanal tandem process.The unexpected chemical instability of the transient 2-alkenylboranes was mitigated by a recently reported one-pot strategy in which the allene was hydroborated in the presence of the aldehyde.The formation of a considerable amount of undesired syn stereoisomer was overcome by transformation of the adducts into methyl esters 9.These intermediates allowed us minimize and remove the undesired stereoisomer.Finally, we report the temporary transformation of C75 into a phenylseleno derivative as a practical method for its purification and storage.We envisage the application of such phenylseleno derivatives to the preparation of new analogues of C75 in the future.

Experimental General information
All reactions involving moisture-or air-sensitive reagents were performed in ovendried glassware under N 2 . 1 H NMR and 13 C NMR spectra were recorded on Mercury 400 or Varian Inova 300 instruments.Chemical shifts (δ) are quoted in parts per million and referenced to internal TMS for 1 H NMR and to CDCl 3 (δ 77.0 ppm) for 13 C NMR. Column chromatography was performed on silica gel (Merck 230-400 mesh).
HRMS analyses were recorded on an Agilent LC/MSD-TOF mass spectrometer.IR spectra (wave numbers in cm -1 ) were recorded on a Nicolet 6700 FT-IR spectrometer.
The solution was washed with sat.NaHCO 3 , the organic layer dried (Na 2 SO 4 ) and the volatiles removed to give almost pure cis-8 (110 mg, 100%).Colorless oil; R f (hexanes /EtOAc 9:1) = 0.25; After 20 h, water (40 mL) was added and the mixture was extracted with CH 2 Cl 2 (3x40 mL).The organic extracts were dried (MgSO 4 ) and the volatiles were removed.The crude was filtered through a short pad of silica to obtain the corresponding mixture of acid lactones 4 (391 mg, 95%).
Mechanism of hydroboration of 6 followed by nonanal addition.
Preparation of C75 and its temporary transformation into the phenylseleno ether derivative 3.
temperature and stirred for 1 hour.The solution was cooled to -78 ºC and a solution of allene 6 (200 mg, 1.585 mmol) and nonanal (327 µL, 1.902 mmol) in dry CH 2 Cl 2 (1 mL) was added.The mixture was stirred for further 15 minutes at -78 ºC and 20 hours at room temperature.The reaction was quenched by addition of triethanolamine (0.5 CCl 4 /CH 3 CN/water mixture (12 mL) at room temperature.