Chemistry Research

ABSTRACT

Epothilones, a new class of microtubule stabilizing agents, are important because of their biological activity, specifically they are effective against cancer cells which have multiple drug resistance. There have been a couple of syntheses published over the last year.^17-18 There is a great deal of interest in the possible biological activity of epothilone derivatives. The objective of this project is based on the synthesis of these derivatives utilizing the completed synthesis of epothilone-A published in 1996 by Dr. Samuel Danishefsky (Sloan-Kettering Institute for Cancer Research). The focus of this research is to develop a new synthesis for derivatives of 3-benzoxy-2-methyl propanal, one of the Danishefsky's precursors for epothilone-A, beginning with 3-chloro-1,2- propanediol.

Structures of Epothilone-A and Its Precursors

INTRODUCTION

The clinical and commercial success of paclitaxol (Taxol) has stimulated an intense search for chemicals with similar modes of action and superior properties. The search was potentially successful in the isolation of a new class of naturally cytotoxic compounds, the epothilones, produced by a cultured strain of the myxobacterium Sorangium cellulosum.¹ The epothilones ability against cancer cell lines was discovered by researchers at Merck in 1996.

The mode of action of the epothilones is similar to the mode of action of Taxol. Taxol kills cancer cells by interfering with one of the fundamental structures comprising the cytoskeleton of eukaryotic cells, the microtubules. Taxol preferentially binds the polymeric microtubule form of tubulin in a 1:1 stoichiometry with the a,b-tubulin heterodimer subunit.² This stabilizes the microtubulin by reducing the rate of a,b-tubulin dissociation. The cell death occurs because of mitotic arrest, blocking the transition between the G2 and M phase.

Because of the mode of action of Taxol and epothilone, they have been found to be promising treatments not only for cancer but in antimalarial chemotherapy as well. This is because parasite replications and the establishment of new infections are dependent on microtubules. Both Taxol and epothilone work in this manner by preventing the establishment of new infections. In multidrug resistant malaria, it has been found that epothilone was nearly 10 fold more effective that Taxol.³

The epothilones have several advantages over Taxol. Even though Taxol has been found affective against a variety of cancers, its clinic usefulness is limited by its side effect profile (neutropenia, peripheral neuropathy, and alopecia) and the fact that its low solubility necessitates delivery of the drug by Cremophor. Cremophor delivery can in itself affect cardiac function and cause severe hypersensitivity responses.² Epothilone A has a higher solubility than Taxol possibly preventing delivery by Cremophor.

A major advantage of the epothilones over Taxol is that epothilone is not a P-glyocoprotein substrate. P-glycoprotein pumps cytotoxic compounds out of multiple drug resistant cells hampering treatment. The P-glycoprotein prevents a significant concentration to build up in the cell and cause cell death. Taxol is a substrate for P-glycoprotein, limiting its ability against many multiple drug-resistant cancer cell lines.

Two point mutations in the b-tubulin gene have also been found to cause resistance to Taxol. In one cell line, there is a change from phenylalanine to valine; in a second line, a change from alanine to threonine was found. The changes cause a 24 fold increase in Taxol resistance but only a 1.3 to 4 fold cross-resistance to epothilone B.⁴

The availability of the two drugs is also an important issue. Taxol comes from the pacific yew tree, which limits its availability. Epothilone availability is not so limited. The completed compound can be produce by fermentation of the myxobacterium.

Another potential advantage of the epothilones is that the drug does not stimulate macrophage activity. It has been found that Taxol has properties similar to endotoxins that stimulate the synthesis of proinflammatory cytokines and nitric oxide by macrophages. Researchers found that epothilone B produced the desired affect on microtubules without triggering the endotoxin signaling pathway and therefore macrophage synthesis of proinflammatory cytokines and nitric oxide.⁵ It is believed that the endotoxin activity of Taxol may be the cause of some of the nonhematological clinical side effects such as myalgia, or muscle weakness and pain, and arthralgia, or joint pain (rarely severe). These side effects may not occur with the use of an epothilone compound in cancer treatment.

The epothilones have been shown to act as competitive inhibitors of taxol binding, suggesting overlapping binding sites on the a,b-tubulin dimer. Epothilones, like Taxol induce the formation of hyperstable tubulin polymers. The two primary epothilones are A and B, (Figure 1). The difference between A and B is a methyl group at the C-12 position which possibly enhances a hydrophobic interaction between the drug and tubulin.⁶ In activity studies, epothilone B has been found to be more active than epothilone A. The structures of epothilone A was published in July of 1996, by Gerhard Holfe from the National Biotechnology Research Institute in Braunschweig, Germany.

Figure 1: Structures of Epothilone and Its Derivatives

It has been found in vitro that epothilone is highly effective against microtubules, but it has also been found to be highly toxic. This has lead to extensive structure-activity studies. It has been found that changing the epoxide on epothilone B to a double bond results in desoxyepothilone B (Figure 2). This change decreases the toxicity of the drug without decreasing the effectiveness of the compound.⁸ Replacing the oxygen molecule in the epoxide moiety with a carbon molecule led to a total loss of activity.⁹ It has been found that the arm of the compound, carbon 15, can be changed considerably without a loss of function. The same has been found for carbons 8-15.⁸ It has also been found that changes involving carbon 1-8 leads to a total loss of activity leading to the conclusion that this portion of the compound is involved with b-tubulin binding. Carbons 1-8 have also been identified using molecular modeling as essential to epothilone activity when compared to Taxol.

Figure 2: Desoxyepothilone

The similar mode of action between Taxol and epothilone and the results of the competitive binding studies indicating a possible overlap of binding sites lead to the use of molecular mechanics software to search for regions of steric and functional similarity between epothilone and Taxol by Winkler and Axelson (Figure 3).¹⁰

Figure 3: Taxol and Its Derivatives

The regions of Taxol known to be essential for biological activity and a region of absolute stereochemistry shared by both Taxol and epothilone were used to define the parameters of the software. Winkler and Axelson found thirteen of the fifteen ring atoms and most of the side chain atoms in epothilone can be superimposed onto corresponding atoms in taxol, all with the proper stereochemistry ( Figure 3).¹⁰

The similar conformations between taxol and epothilone can be used to apply the same derivation studies of taxol to epothilone. In a study conducted at the National Cancer Institute, a C-2 derivative of paclitaxol (Taxol) was more active in promoting tubulin assembly than the original structure. The analogs of paclitaxol found to have increased activity were those with an m-azido group on the benzoyl residue at the C-2 carbon and those with nonreactive meta groups (methoxy and chloro) at the same position (Figure 3).¹¹

This project involves synthesizing the corresponding methoxy derivative of epothilone A (1c), which should have improved biological activity as the related Taxol derivative. This synthesis was completed by modeling a known epothilone precursor. After careful study of all synthetic pathways published for epothilone12-15, our lab selected a precursor synthesized by Danishefsky.¹² Danishefsky first synthesized structures 4 and 5, shown in Figure 4 and then joined them using a ring closing macrolactonization and completed the epothilone synthesis through an epoxidation reaction.

Figure 4: Danishefsky's Precursors We're Interested In

The carbon of interest (labeled with an asterisk in figure 4) locates the specific site our lab functionalized with a methoxy group rather than the naturally-occurring methyl group. Danisheshefsky's precursors for structure 5 are 3-benzoyl-2-S-methyl propanal (6) and 2-methoxy-4-trimethylsilylhex-2,4-diene (7), both of which are shown in Figure 5.

Figure 5: How Our Target Compound Would Be Used In Danishefsky's Synthesis

Structure 6 modified on the C-2 position (*) was our target compound. The synthetic approach utilized is shown in Figure 6.

Figure 6: Our 5-Step Synthesis

The 5-step synthesis starts with a readily available starting material, 3-chloro-1,2-propanediol (8) and ends with the target compound (13) where the asterisked carbon is the site of interest. This target compound (13) is identical to Danishefsky's structure 6 (see Figure 5) except for substitution of a methyl for a methoxy functionality. Step two allows not only for methylation at this site (as shown above) but would also allows for protection of this secondary alcohol followed by later replacement with either an azide or a chloro group. All threee of these derivatives (methoxy, azide, and chloro) have already shown increased biological activity in Taxol, thus expected to have the same result in epothilone A.

The first step involves selectively protecting the primary alcohol as a silyl ether (9). The second step methylates the secondary alcohol (10), and the third step is a condensation reaction with benzyl alcohol yielding a benzyl ether intermediate (11). The last two steps involve the deprotection of primary silyl ether (12) followed by oxidation of the resulting primary alcohol into an aldehyde (13).

EXPERIMENTAL

Preparation of 3-chloro-1-t-butyldimethylsilyloxy-2-propanol (9)

To a round bottom flask propanediol (4.3 mL, 50 mmol) and triethylamine (7.0 mL, 50 mmol) were added to dichloromethane (5 mL). 4-Dimethylaminopyridine (0.25 g, 2.0 mmol), was added as a transfer catalyst. t-Butyldimethylsilyl chloride was then added (7.5 g, 50 mmol) to the reaction mixture. Nitrogen was used to flush out all oxygen and moisture from the reaction set-up. The set-up was then closed and the reaction was left at room temperature for 24 hours. The reaction was quenched with water (5 mL) and dichloromethane (5 mL) was added. The organic layer was washed with water (2 x 5 mL), then saturated ammonium chloride (3 x 5 mL), and finally dried with sodiumsulfate. The organic solvent was left to evaporate resulting in a clear oil. TLC (toluene:methanol, 9:1, R_f 0.63) revealed some polar impurities, thus the product mixture was passed through a pad of silica gel (toluene:methanol, 9:1) and the solvent was allowed to evaporate resulting in a clear colorless oil (8.24 g, 36.8 mmol, 73.5%); bp 81-83 C; Rf 0.857 (toluene:methanol, 7:3); IR (neat): 3600-3170 (O-H), 2950, 2920, 2890, 2849 (alkane C-H), 1460, 1360 (CH₃), 1255, 1104, 1072 (C-O), 835, 775 (C-Cl); ¹H-NMR (CDCl3): 0.5 ppm (s, 6H), 1.34 ppm (s 9H), 3.4-3.75 ppm (m, 5H), 5.23 ppm (s, 1H). GCMS: 225 [M⁺], 175, [M⁺ - ClCH₃], 167, 159, 149, 131 [t-butyldimethyl silyloxy], 115 [t-butyldimethyl silyl], 89, 75, 57 [t-butyl], 45, 29.

Synthesis of 3-chloro-1-t-butyldimethylsilyloxy-2-methoxypropane (10)

To dichloromethane (10 mL), product (9) (6.1 g, 27 mmol), sodium hydroxide (0.104 g, 2.60 mmol), water (0.104 g, 5.78 mmol), and tetrabutyl ammonium hydrogen sulfate (3.4 mg, 0.010 mmol) were added to the reaction mixture. The mixture was stirred vigorously for 30 minutes at room temperature, then cooled in an ice-water bath. Over 1 hour, dimethyl sulfate (0.15 g, 1.2 mmol) was added dropwise. The reaction was stirred vigorously for 3 hours until the reaction was complete by TLC (toluene:methanol, 9:1, Rf 0.80). The reaction mixture was quenched by addition of conc. ammonium hydroxide (0.04 mL) followed by extraction with water (2 x 10 mL). The organic layer was then dried with magnesium sulfate and the solvent was allowed to evaporate resulting in a yellow oil. TLC (toluene:methanol, 9:1) revealed some polar impurities, thus the product mixture was passed through a pad of silica gel (toluene:methanol, 9:1) and the solvent was allowed to evaporate resulting in a clear, slightly yellow oil (4.897g, 20.6 mmol, 82.4%); bp 205-207 C; R_f 0.902 (toluene:methanol, 7:3); IR (neat): 2921 & 2848 (alkane C-H), 1500, 1352 (CH₃), 1250, 1181 (C-O), 830,770 (C-Cl); ¹H-NMR (CDCl₃): 0.5 (s, 6H), 1.34 (s, 9H), 4.29 (s, 3H), 3.68-4.21 (m, 5H). GCMS: 239 [M⁺], 137 [M⁺ - t-butyl - 3CH₃], 113 [t-butyldimethylsilyl], 99 [t-butylmethylsilyl], 85 [t-butylsilyl], 71, 57 [t-butyl], 43, 29.

Preparation of 3-benzoxy-1-t-butyldimethylsilyloxy-2-methoxy-propane (11)

Product (10) (2.392 g, 10.05 mmol) was placed in a round bottom flask. Benzyl alcohol (1.2 mL, 12 mmol) and N,N-diisopropylamine (2.0 mL, 10 mmol) were added to the reaction mixture. The mixture was left at 0C for 1 hour, then allowed to come to room temperature. The reaction was left to react for 72 hours. The product was extracted with ether and dried with anhydrous sodium sulfate. The organic layer was run through a silica gel column with ether as the solvent. The ether was allowed to evaporate resulting in a yellow oil (1.14 g, 3.88 mmol, 38.6%); bp 108-110 C; R_f 0.906 (toluene:methanol, 7:3); IR (neat): 3065, 3059, 3025 (alkene C-H), 2955, 2950, 2888, 2858, 2735, 2709 (alkane C-H), 1469, 1385 (CH₃), 1360, 1255, 1113, 1069 (C-O), 838, 781, 737,698; ¹H-NMR (CDCl₃): 0.82 (s,6H), 1.58 (s, 9H), 3.45-3.70 (m, 5H), 4.92 (s, 3H), 5.37 (s, 2H), 7.95-8.10 (m, 5H). GCMS: 310 [M⁺], 212, 167 [M⁺ - CH₂OTBS], 105 [PhCH₂O], 91 [PhCH₂], 77 [Ph}, 65, 51, 39, 28.

Synthesis of 4-benzoxy-2-methoxy-1-butanol (12)

Tetrabutylammonium fluoride (2.2 eq., 0.383 g) in methylene chloride (3 mL) was added dropwise over one hour to the reaction flask containing the product (11) (0.15 g, 0.50 mmol). The reaction mixture was left open and stirring overnight. An equal amount of ether and water (2.5 mL each) was added to the flask. The organic layer was removed and dried over anhydrous sodium sulfate. The ether was evaporated off resulting in a yellowish colored oil (0.057 g, 0.40 mmol, 80%); bp 210-212 C; R_f 0.962 (toluene:methanol, 7:3); IR (neat): 3349 (O-H), 3065, 3059, 3024 (alkene C-H), 2955, 2950, 2858 (alkane C-H), 1721 (C=O tautomer), 1459, 1250,1255, 1046 (C-O), 837, 781, 736,698; ¹H-NMR (CDCl₃): 1.78 (s, 1H), 3.40-3.70 (m 5H), 4.60 (s, 3H), 5.28 (s, 2H), 6.95 (m 5H). GCMS: 194 [M⁺], 173, 147 [M⁺ - OCH₃ - OH], 131, 106 [PhCH₂O], 89, 78, [Ph], 64, 45.

Preparation of 3-benzoxy-2-methoxypropanal (13)

Chromium trioxide (0.24 g, 2.0 mmol) was added to a stirring solution of dry pyridine (0.40 mL, 4.8 mmol) in dry dichloromethane (4 mL). The reaction was stoppered with a drying tube containing drierite. The deep burgundy solution was stirred at room temperature for 15 minutes. After the 15 minute, period a solution of compound (12) in dichloromethane (1 mL) was added in one portion to the reaction mixture. Immediately a black deposit separated out of solution. The reaction was left to mix for 30 minutes and then decanted off. The deposit was washed with ether and the organic solution was filtered. The organic solution was washed with 4 mL portions of 5% NaOH, 5% HCl, 5% sodium hydrogen carbonate, and finally with saturated sodium chloride. The organic solvent was allowed to evaporate, resulting in a clear colorless oil (28 mg, 0.19 mmol, 50%); bp 266-268 C; R_f 0.973 in (toluene:methanol, 7:3); IR (neat) 3059, 3025 (alkene C-H), 2955, 2857, 2888, 2875 ( alkane C-H), 2726 (aldehyde C-H), 1705 (C=O aldehyde), 1255, 1071 (C-O), 837, 782; ¹H-NMR (CDCl₃): 3.75 (d, 4H), 4.61 (s, 3H), 4.71 (dt, 1H), 5.12 (s, 2H), 7.30-7.47 (m, 5H), 9.54 (s,1H). GCMS: 194 [M⁺], 164 [M⁺ - CO], 152 [M⁺ - CO - CH₃], 122 [M⁺ - CH₃OCHCHO], 108 [M⁺ - CH₃OCH(CH₂)CHO], 92 [M⁺ - CH₃OCH(CH₂O)CHO], 78 [Ph], 64, 46, 34.

DISCUSSION

Based on FT-IR analysis (Perkin Elmer Spectrum BX-II, Heidelberg College) and ¹H-NMR analysis (300 MHz, Miami University, Ohio) each intermediate has been positively identified. The purity of each product was determined by TLC analysis. Some impurities were not easily identified by TLC analysis, making purification of the desired compounds difficult. Dr. Richard Taylor from Miami University graciously agreed to purity and molecular weight analysis of our products by GCMS. The results of this analysis confirmed the presence of the desired products but showed substantial amounts of impurities at each step. Almost every impurity was identified as starting materials from the previous steps.

The removal of the silyl protecting group with 5.5 equivalents of tetrabutylammonium fluoride as reported ¹⁶ caused side reactions with the loss of methoxy functional group. Decreasing the number of equivalents to 2.2 equivalents minimized this problem.

Final analysis of the target compound was complicated by the possible tautermerization of the final product (Figure 7).

Figure 7: Tautomers of Our Final Product

This was predicted due to the presence of a vinylic H on ¹H-NMR at ~8.1 ppm; yet the presence of these tautomers was not confirmed by GCMS analysis. If both tautomers were present, the GCMS data should have indicated two different peaks with the same molecular weight.

REFERENCES

1. Balog, A., Meng, D., Kamenecka, T., Bertinato, P., Su, D., Sorensen, E., Danishefsky, S., Angew. Chem. Int. Ed. Engl., (1996), 35, 2801-2803
2. Bollag, D.M., McQuency, P.A., Zhu, J., Hensens, O., Koupal, L., Liesch, J., Goetz, M., Lazarides, E., Woods, C.M., Cancer Research, (1995), 55, 2325-2333
3. Taraschi, T.F., Trelka, D., Schneider, T., Matthews, I., Experimental Parasitology, (1998), 88, 184-193
4. Giannakakout, P., Sackett, D.L., Kang, Y., Zhan,Z., Buters, J.T.M., Fojo, T., Poruchynsky, M.S., J. Biol. Chem., (1997), 272, 17118-17125
5. Muhlradt, P.F., Sasse, F., Cancer Research, (1997), 57, 3344-3346
6. Holfe, G., Bedorf, N., Steinmetz, H., Schomburg, D., Gerth, K., Reichenbach, Angew. Chem. Int. Ed. Engl., (1996), 35, 1567-1569
7. Gerth, K., Bedorf, N., Hofle, G., Irschik, H., Reichnebach, H., The Journal of Antibiotics, (1996), 49, 560-563
8. Chou, T.C., Zhang, X.G., Balog, A., Su, D.S., Meng, D., Savin, K., Bertino, J.R., Danishefsky, S.J., Proc. Natl. Acad. Sci. USA, (1998), 95, 9642-9647
9. Nicolaou, K.C., Finlay, M.R.V., Ninkovic, S., King, N.P., He, Y., Li, T., Sarabia, F., Vourloumis, D., Chemistry & Biology, (1998), 5, 365-372
10. Winkler, J.D., Axelson, P.H., Bioorganic & Medicinal Chemistry Letters, (1996), 6, 2963-2966
11. Kowalski, R.J., Giannakokov, P., Hamel, E., J. Biol. Chem., (1997), 272, 2534-2541
12. Balog, A., Meng D.F., Kamenecka, T., Bertinato P., Su, D.S., Sorensen, E.J., Danshefsky, S., Angew., Chem. Int Ed. Engl., (1997), 35, 2801-2803
13. Nicolaou, K.C., Sarabia, F., Ninkovic, S., Yang, Z., Angew. Chem. Int. Ed. Engl., (1997), 36, 525-52
14. Schnizer, D., Limberg, A., Baur, A., Bohm, O.M., Cordes, M., Angew. Chem. Int. Ed. Engl., (1997), 36, 523-524
15. Yang, Z., He, Y., Vourloumis, D., Vallerg, H., Nicolau, K.C., Angew. Chem. Int. Engl., (1997), 36, 116-168
16. Ito, Y., Nakatsuka, M., Saegusa, J. Am. Chem. Soc., (1980), 102, 863-865
17. S. J. Danishefsky et al., J. Org. Chem., 1996, 61, 8000-8001
18. K. C. Nicolaou et al., Angew. Chem. Int. Ed. Engl., 1997, 36, 166-168

Chemistry Home Page

Last modified: July 19, 1999. JG.