12,13-Aziridinyl Epothilones. Stereoselective Synthesis of Trisubstituted Olefinic Bonds from Methyl Ketones and Heteroaromatic Phosphonates and Design, Synthesis, and Biological Evaluation of Potent Antitumor Agents

Article

pubs.acs.org/JACS

12,13-Aziridinyl Epothilones. Stereoselective Synthesis of

Trisubstituted Oleﬁnic Bonds from Methyl Ketones and

Heteroaromatic Phosphonates and Design, Synthesis, and Biological

Evaluation of Potent Antitumor Agents

K. C. Nicolaou,*^,†Derek Rhoades,^†,∥Yanping Wang,^†,∥Ruoli Bai,^‡Ernest Hamel,^‡Monette Aujay,^§

Joseph Sandoval,^§and Julia Gavrilyuk^§

^†Department of Chemistry, BioScience Research Collaborative, Rice University, 6100 Main Street, Houston, Texas 77005, United

States

^‡Screening Technologies Branch, Developmental Therapeutics Program, Division of Cancer Treatment and Diagnosis, Frederick

National Laboratory for Cancer Research, National Cancer Institute, National Institutes of Health, Frederick, Maryland 21702,

United States

^§Abbvie Stemcentrx, LLC, 450 East Jamie Court, South San Francisco, California 94080, United States

S

* Supporting Information

ABSTRACT: The synthesis and biological evaluation of a series of

12,13-aziridinyl epothilone B analogues is described. These

compounds were accessed by a practical, general process that

involved a 12,13-oleﬁnic methyl ketone as a starting material obtained

by ozonolytic cleavage of epothilone B followed by tungsten-induced

deoxygenation of the epoxide moiety. The attachment of the aziridine

structural motif was achieved by application of the Ess−Kurti−Falck

̈

aziridination, while the heterocyclic side chains were introduced via

stereoselective phosphonate-based oleﬁnations. In order to ensure

high (E) selectivities for the latter reaction for electron-rich

heterocycles, it became necessary to develop and apply an

unprecedented modiﬁcation of the venerable Horner−Wadsworth−

Emmons reaction, employing 2-ﬂuoroethoxyphosphonates that may

prove to be of general value in organic synthesis. These studies

resulted in the discovery of some of the most potent epothilones reported to date. Equipped with functional groups to

accommodate modern drug delivery technologies, some of these compounds exhibited picomolar potencies that qualify them as

payloads for antibody drug conjugates (ADCs), while a number of them revealed impressive activities against drug resistant

human cancer cells, making them desirable for potential medical applications.

abandoned after an early clinical trial provided evidence of an

ineﬃcient delivery construct.⁷Ixabepilone (5) and 12,13-

aziridinyl epothilone A (8) were prepared from epothilone B³

(2) and epothilone A⁴(1), respectively, through semisynthesis

(see Figure 2A).

The rapidly expanding paradigm of antibody-drug conjugates

(ADCs) and other targeted therapeutic approaches oﬀer a way

to improve the therapeutic index of compounds whose high

potencies preclude them from being viable drugs due to toxicity

issues.⁸It was with this reasoning that we undertook the

present study with the intention of achieving high potencies,

while at the same time installing handles in the molecule for

conjugation to linkers and subsequent attachment to antibodies

or other suitable delivery systems. In this Article, we describe

1. INTRODUCTION

The epothilone class of natural products, represented by

epothilones A, B, C, and D (1−4, Figure 1), enjoys a rich

history of synthetic studies stemming from total synthesis and

semisynthesis expeditions.¹These investigations led to several

drug candidates with one receiving approval as a clinical agent

for the treatment of metastatic or locally advanced breast cancer

(5, ixabepilone, marketed as Ixempra, Figure 1). The latter is

produced by semisynthesis from the naturally occurring

epothilone B (2, Figure 1).^2,3Other notable epothilone B

analogues are the methylthio epothilone B (ABJ879, 6, Figure

1), the aminomethyl epothilone B (BMS-310705, 7, Figure 1),

the C12−C13 aziridinyl epothilone A analogue (8, Figure 1),⁴

and its N-alkylated derivative (BMS-748285, 9, Figure 1),⁵the

former two entering clinical trials but failing, presumably due to

unacceptable therapeutic windows.⁶Conversely, BMS-748285

(9, Figure 1) was explored as a folate conjugate but was

Received: March 16, 2017

A

DOI: 10.1021/jacs.7b02655

J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

attachment as payloads to delivery systems for assembling

ADCs and other vehicles for targeted cancer therapies.

2. RESULTS AND DISCUSSION

2.1. Initial Explorations. Inspired by the required presence

of a basic nitrogen atom in a certain position in highly active

epothilones and motivated by the necessity for suitable

functional groups as points of attachment for conjugation

studies,^1,9we decided to focus on epothilone analogues in

which the epoxide moiety of epothilone B is replaced with an

aziridine moiety. Given that most synthetic routes to

epothilones proceed through the corresponding C12−C13

oleﬁn and mindful of the power of the recently disclosed Ess−

Kurti−Falck aziridination,¹⁰we initiated an exploration to test

̈

whether epothilones C and D (3 and 4, Figure 1) could serve as

substrates for this reaction. It should be noted that the synthesis

of C12−C13 aziridinyl epothilone B has not been reported,

although the corresponding aziridinyl epothilone A (8) was

previously prepared from epothilones A and C (1 and 3, Figure

2A), albeit through a lengthy process.⁴In contrast, and as

shown in Scheme 1, we were able to prepare both the aziridinyl

epothilone A analogue 8 (70% yield) and aziridinyl epothilone

Figure 1. Molecular structures of epothilones A−D (1−4),

ixabepilone (5), methylthio epothilone B (ABJ879, 6), aminomethyl

epothilone B (BMS-310705, 7), aziridinyl epothilone A (8), and its N-

alkylated analogue (BMS-748285, 9).

B analogue 10 (66% yield) through the Ess−Kurti−Falck

̈

method [O-(2,4-dinitrophenyl)hydroxylamine (DPH),

Rh₂(esp)₂cat.]¹⁰directly and in one step from epothilones C

and D (3 and 4, Scheme 1), respectively. Furthermore, the

aziridination reaction was found to be completely regio- and

stereoselective, leading to the desired epothilone conﬁguration

as proven by comparison of the NMR data of aziridine 8 with

those reported by BMS for the same compound.⁴By analogy,

we assigned the β-conﬁguration for aziridine 10 and the other

aziridine compounds obtained in this study. In addition, the

absolute stereochemistry was unambiguously proven via X-ray

crystallographic analysis of aziridinyl epothilone B 10 (mp

112−114 °C, see ORTEP in Scheme 1 and the Supporting

Information for details). These results contrast with the

epoxidation reaction [DMDO (dimethyldioxirane) or TFDO

(methyl(triﬂuoromethyl)dioxirane)], which generates epothi-

lone B (2) from its corresponding oleﬁnic precursor,

epothilone D (4), a process that displayed moderate

diasteroselectivity (ca. 5:1 dr).¹¹Subsequent alkylation of

these aziridines with 2-bromoethanol (K₂CO₃) aﬀorded N-

hydroxyethyl aziridinyl epothilones A (9, 97% yield) and B (11,

95% yield), respectively, as shown in Scheme 1. The primary

hydroxyl group on these structures may serve as a convenient

functionality to attach linkers for potential ligation to antibodies

and other delivery systems.

2.2. Attempted Stille Coupling Approach for the

Synthesis of Aziridinyl Epothilone B Side Chain

Analogues. Having established the feasibility of the

aziridination reaction with these rather complex substrates

and demonstrated its excellent regio- and stereoselectivity as

well as its tolerance of the thiazole moiety, we proceeded to

explore its applicability to other substrates such as those

featuring a vinyl iodide moiety (e.g., 71; for its ability to serve

as a precursor to a wide range of analogues) and various

heterocyclic side chains (e.g., N-methyl-5-methylthiopyrazole;

for its ability to impart high potency),¹²as depicted in Scheme

2. Thus, triol iodide precursor 69, readily available by total

synthesis,¹³was converted to iodide 71 (NaBH₃CN, 80% yield)

via bis-iodide 70, the latter obtained from 69 through its

tosylate counterpart (Ts₂O, Et₃N, DMAP; then TBAI, 88%

yield). Iodide 71, however, left much to be desired as a

Figure 2. (A) Previous syntheses of ixabepilone (5) and 12,13-

aziridinyl epothilone A (8) from epothilone B (2) and epothilones A

(1) or C (3), respectively. (B) General synthetic strategy for accessing

aziridinyl epothilone B analogues I from β-heteroaromatic phospho-

nates II and aziridinyl methyl ketone III, the latter to be derived from

oleﬁn methyl ketone IV, and ultimately epothilone B (2). HWE =

Horner−Wadsworth−Emmons.

the details of our investigations that led to (a) a practical and

eﬃcient semisynthetic route to a versatile key aziridine

precursor (i.e., methyl ketone III, Figure 2B) from epothilone

B (2); (b) a direct aziridination of epothilones C (3) and D (4)

to produce the corresponding aziridinyl epothilones and their

N-substituted analogues; (c) the development of a highly

stereoselective synthesis of trisubstituted oleﬁns from ketones

and β-heteroaromatic phosphonates; and (d) the application of

these synthetic strategies and technologies to the synthesis and

biological evaluation of a series of designed aziridinyl

epothilone B analogues (8−40, Figure 3) from the correspond-

ing methyl ketone III and β-heteroaromatic phosphonates

(41−68, Figure 4) equipped with appropriate functionalities for

B

DOI: 10.1021/jacs.7b02655

J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

Figure 3. Synthesized aziridinyl epothilone analogues 8−40.

C

DOI: 10.1021/jacs.7b02655

J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

Scheme 2. Synthesis of Epothilone D Analogue 73 from

Vinyl Iodide 69 and Attempted Aziridination of Vinyl Iodide

a

71 and Epothilone 73

Figure 4. Synthesized β-heteroaromatic phosphonates 41−68.

a

Reagents and conditions: (a) Ts₂O (3.0 equiv), Et₃N (5.0 equiv),

DMAP (1.0 equiv), CH₂Cl₂, 0 °C, 20 min; then TBAI (5.0 equiv), 0

°C, 20 min, 88%; (b) NaBH₃CN (12 equiv), DMPU, 25 °C, 40 min,

80%; (c) DPH (1.1 equiv), Rh₂(esp)₂(0.1 equiv), TFE, 25 °C, 16 h,

no desired products 40 or 72; (d) 74 (2.5 equiv), Pd₂(dba)₃(0.5

equiv), AsPh₃(1.0 equiv), CuI (2.0 equiv), DMF, 0 °C, 1 h, 67%. dba

= dibenzylideneacetone; DMAP = 4-dimethylaminopyridine; DMPU

= 1,3-dimethyl-3,4,5,6-tetrahydro-2(1H)-pyrimidinone; TBAI = n-

tetrabutylammonium iodide; Ts = 4-toluenesulfonyl.

Scheme 1. Aziridination of Epothilones C (3) and D (4) and

Synthesis of N-Hydroxyethyl Aziridinyl Epothilone A and B

a

Analogues 9 and 11 via N-Alkylation of 8 and 10

produce the desired aziridine (40), presumably due to oxidative

degradation of the electron-rich methylthiopyrazole moiety.¹⁴

2.3. Development of a New Synthetic Route to

Substituted Aziridinyl Epothilone B Analogues. Faced

with these dead ends, we then turned our attention to the

rupture of the side chain of readily available (via fermentation)

epothilone B (2) and deoxygenation of its epoxide moiety to

provide a viable precursor onto which both the aziridine moiety

and various side chains could be installed. Previous studies by

Hoﬂe¹⁵and Bristol−Myers Squibb (BMS) scientists¹⁶indicated

̈

that the former objective could be accomplished through

ozonolysis, whereas the latter could be realized through an in

situ generated reducing metal (e.g., W or Mg/Ti).

Thus, and as shown in Scheme 3, epothilone B (2) was

converted by ozonolysis (O₃; then Me₂S) to methyl ketone 75

(94% yield), which was then exposed to TESOTf and 2,6-

lutidine to aﬀord bis-TES ether 76 in 84% yield. Compounds

75 and 76 were tested in the subsequent deoxygenation of the

epoxide moiety with WCl₆/n-BuLi, revealing their worthiness

as viable substrates for the preparation of the required oleﬁnic

methyl ketones 77 [85% yield, (Z):(E) ca. 5:1] and 78 [86%

yield, (Z) only], respectively. The latter proved to be the

preferred substrate due to its exclusive geometrical selectivity

for the desired (Z) isomer. The aziridination of both substrates

77 or 78 also proved pleasantly successful under the standard

conditions,¹⁰with 77 aﬀording aziridine 79 in 87% yield and 78

furnishing aziridine 80 in 90% yield, both with complete

stereoselectivity as desired. Compound 80 was converted to

epothilone B analogue 10 (Scheme 1), whose NMR

spectroscopic data matched those of a sample derived from

a

Reagents and conditions: (a) DPH (1.1 equiv), Rh₂(esp)₂(0.05

equiv), TFE, 25 °C, 4 h, 70% for 8, 66% for 10; (b) 2-bromoethanol

(5.0 equiv), K₂CO₃(6.0 equiv), DMF, 50 °C, 48 h, 97% for 9, 95% for

11. DMF = N,N-dimethylformamide; DPH = O-(2,4-dinitrophenyl)-

hydroxylamine; esp = α,α,α′,α′-tetramethyl-1,3-benzenedipropionic

acid; TFE = 2,2,2-triﬂuoroethanol.

substrate for aziridination under the Ess−Kurti−Falck con-

̈

ditions, and aziridine 72 could not be obtained in meaningful

quantities due to the capricious nature of the reaction in this

instance. To circumvent this problem, we then prepared

substrate 73 through Stille coupling of vinyl iodide precursor

71 with pyrazolyl stannane 74¹²[Pd₂(dba)₃cat., CuI, AsPh₃,

67% yield], as shown in Scheme 2. Disappointingly, however,

exposure of this substrate to the aziridination reaction failed to

D

DOI: 10.1021/jacs.7b02655

J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

furnish, through the ensuing stereoselective Horner−Wads-

worth−Emmons (HWE) oleﬁnation,^17,18the expected pro-

tected aziridinyl epothilone B analogue 83 in 60% yield. From

the latter, desilylated product aziridinyl methylthio epothilone

B analogue 12 was liberated by treatment with HF·py (79%

yield). In a similar manner, precursor 82 was coupled with

phosphonate 42, this time under the inﬂuence of NaHMDS, to

aﬀord protected aziridinyl epothilone B analogue 84 (68%

yield). Global deprotection (HF·py; then TFA) provided

aziridinyl aminomethyl epothilone B analogue 13 in 48%

overall yield, as summarized in Scheme 3.

Scheme 3. Synthesis of Aziridinyl Epothilone B Analogues

12 and 13 from Epothilone B (2)

a

This synthetic route clearly oﬀers distinct advantages over

previously described routes (e.g., 8⁴and 9,⁵Figure 1) due to its

versatility and the practical availability of the common

intermediate (i.e., methyl ketone 82, Scheme 3), from which

it can diverge to a wide range of imaginable side chain aziridinyl

epothilone B analogues. Furthermore, this synthetic work

represents the ﬁrst application of the Ess−Kurti−Falck

̈

aziridination on substrates of high complexity, an observation

that bodes well for its future application for the synthesis of

complex natural products and designed molecules.

2.4. Synthetic Route Extension to Free Aziridinyl

Epothilone B Analogues. To enrich the repertoire of

synthetic strategy options for the synthesis of aziridinyl

epothilone B analogues with varying side chains, we carried

out a brief exploration of direct attachment of epothilone side

chains using the naked (80) or protected (86 and 87) aziridinyl

methyl ketones as substrates, as summarized in Scheme 4. It

was delightful to observe that the free aziridine bis-TES

protected methyl ketone 80 could serve as a substrate for the

HWE reaction with methylthiothiazolyl phosphonate 41 (n-

BuLi, 59% yield) to aﬀord, with complete geometrical control

a

Reagents and conditions: (a) O₃; then Me₂S (10.0 equiv), CH₂Cl₂,

−78 °C, 5 min, 94%; (b) TESOTf (2.4 equiv), 2,6-lutidine (3.0 equiv),

CH₂Cl₂, −78 °C, 15 min, 84%; (c) WCl₆(2.0 equiv), n-BuLi (4.0

equiv), THF, −78 → 25 °C, 40 min; then 75 or 76 (1.0 equiv), −20

→ 0 °C, 2 h, 85% for 77 [(Z):(E) ca. 5:1], 86% for 78 [(Z) only]; (d)

DPH (1.5 equiv), Rh₂(esp)₂(0.02 equiv), TFE, 25 °C, 30 min, 87%

for 79, 90% for 80; (e) 2-bromoethanol or 2-bromoethanol TBS ether

(5.0 equiv), K₂CO₃(4.0 equiv), DMF, 70 °C, 12 h, 29% for 81, 90%

for 82; (f) 41 (15 equiv), n-BuLi (12 equiv); then 82 (1.0 equiv),

THF, −78 → 0 °C, 3 h, 60%; or 42 (8.3 equiv), NaHMDS (6.8

equiv); then 82 (1.0 equiv), THF, −78 °C, 2.5 h, 68%; (g) HF·py (xs),

THF, 0 → 25 °C, 1 h, 79%; (h) HF·py (xs), THF, 0 → 25 °C, 5 h;

then TFA (xs), CH₂Cl₂, 0 → 25 °C, 2.5 h, 48% overall. HMDS =

hexamethyldisilazide; OTf = triﬂuoromethanesulfonate; py = pyridine;

TBS = tert-butyldimethylsilyl; TES = triethylsilyl; TFA = triﬂuoro-

acetic acid; THF = tetrahydrofuran; xs = excess.

Scheme 4. Synthesis of Free Aziridinyl Epothilone B

Analogue 14 from Methyl Ketone 80 or N-Protected Methyl

a

Ketone 87

epothilone B¹⁶(or epothilone D) as described in Scheme 1

(see also the Supporting Information for further details),

thereby supporting the stereochemical assignment of the

aziridination of methyl ketones 77 and 78. Ultimately, the

absolute conﬁguration of aziridinyl methyl ketone 80 was

unambiguously conﬁrmed by X-ray crystallographic analysis of

epothilone 39 (see Scheme 9 for the ORTEP of epothilone 39,

which was derived from precursor 80 through intermediates 96

and 97).

The subsequent alkylation of aziridines 79 (free hydroxyl

groups) and 80 (TES-protected hydroxyl groups) with 2-

bromoethanol (with or without TBS protection of the hydroxyl

group) and K₂CO₃[79 + 2-bromoethanol → 81 (29% yield);

80 + 2-bromoethanol TBS ether → 82 (90% yield)], however,

distinguished the TES-protected compound (i.e., 80) and 2-

bromoethanol TBS ether as the preferred substrates in terms of

overall yield and selectivity. Precursor 82 was then coupled with

side chain phosphonate 41 under the inﬂuence of n-BuLi to

a

Reagents and conditions: (a) 41 (9.6 equiv), n-BuLi (7.7 equiv); then

80 (1.0 equiv), THF, −78 → 25 °C, 1.5 h, 59%; (b) HF·py (xs), THF,

0 → 25 °C, 1 h, 93%; (c) Boc₂O (3.0 equiv), Et₃N (3.0 equiv), DMAP

(0.2 equiv), MeCN, 0 °C, 5 min, 78%; (d) SEMCl (1.5 equiv), i-

Pr₂NEt (2.0 equiv), CH₂Cl₂, 0 °C, 2 h, 59%; (e) 41 (23 equiv), n-BuLi

(19 equiv); then 86 or 87 (1.0 equiv), THF, −78 → 10 °C, 3 h, 88

isolated with impurities; 60% for 89; (f) TFA (xs), CH₂Cl₂, 0 → 25

°C, 1.5 h, 75%. Boc = tert-butyloxycarbonyl; SEM = 2-(trimethylsilyl)-

ethoxymethyl.

E

DOI: 10.1021/jacs.7b02655

J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

a

Scheme 5. Synthesis of N-Hydroxyethyl Aziridinyl Epothilone B Analogues 15−26 from Methyl Ketone 82

a

Reagents and conditions: (a) 43 (14 equiv), NaHMDS (14 equiv); then 82 (1.0 equiv), THF, −78 → 0 °C, 4 h, 69%; (b) HF·py (xs), THF, 0 →

25 °C, 5 h; then TFA (xs), CH₂Cl₂, 0 → 25 °C, 6 h, 80% overall; (c) 44 (12 equiv), NaHMDS (11 equiv); then 82 (1.0 equiv), THF, −78 → − 40

°C, 1 h; then HF·py (xs), THF, 0 → 25 °C, 3 h, 37% overall [(E):(Z) ca. 88:12], plus 27% recovered 82; (d) 45 (15 equiv), NaHMDS (13 equiv);

then 82 (1.0 equiv), THF, −78 °C, 30 min, 65%; (e) HF·py (xs), THF, 0 → 25 °C, 3 h, 95%; (f) Ac₂O (5.0 equiv), DMAP (3.0 equiv), CH₂Cl₂, 0

→ 25 °C, 25 min, 74%; (g) 46 (12 equiv), NaHMDS (9.7 equiv); then 82 (1.0 equiv), THF, −78 → 0 °C, 2.5 h, 50%; 47 (12 equiv), NaHMDS (9.7

equiv); then 82 (1.0 equiv), THF, −78 → 0 °C, 2.5 h, 45%; (h) HF·py (xs), THF, 0 → 25 °C, 4 h, 90%; HF·py (xs), THF, 0 → 25 °C, 5 h; then

TFA (xs), CH₂Cl₂, 0 → 25 °C, 3 h, 71% overall; (i) 48 (28 equiv), n-BuLi (22 equiv); then 82 (1.0 equiv), THF, −78 → 25 °C, 2 h, 94%; (j) HF·py

(xs), THF, 0 → 25 °C, 5 h, 93%; (k) 49 (21 equiv), n-BuLi (18 equiv); then 82 (1.0 equiv), THF, −78 → 25 °C, 4 h, 67%, plus 28% recovered 82;

(l) HF·py (xs), THF, 0 → 25 °C, 5 h, 74%; (m) 50 (11 equiv), NaHMDS (4.9 equiv); then 82 (1.0 equiv), THF, −78 °C, 1 h; then t-BuOK (5.0

equiv), THF, − 20 °C, 5 min, 28% overall [(E):(Z) ca. 30:70]; (n) HF·py (xs), THF, 0 → 25 °C, 5 h, 82%; (o) 51 (13 equiv), n-BuLi (10 equiv);

then 82 (1.0 equiv), THF, −78 → 10 °C, 1.5 h, 65%; (p) HF·py (xs), THF, 0 → 25 °C, 9 h, 81%; (q) 52 (11 equiv), n-BuLi (10 equiv); then 82

(1.0 equiv), THF, −78 °C, 1.5 h; then t-BuOK (5.0 equiv), THF, − 20 °C, 5 min, 73% overall, plus 10% recovered 82; (r) HF·py (xs), THF, 0 → 25

°C, 3 h, 95%; (s) Ac₂O (2.0 equiv), i-Pr₂NEt (2.0 equiv), DMAP (0.05 equiv), CH₂Cl₂, 0 °C, 1.5 h, 71%.

[(E):(Z) > 98:2], the corresponding protected aziridinyl

epothilone B 85, whose desilylation (HF·py, 93% yield) led

smoothly to the targeted aziridinyl methylthio epothilone B

analogue 14. Be that as it may, reproducibility issues, as well as

a narrow substrate scope with unpredictable results, prompted

us to screen various protecting groups. Thus, further

experimentation with Boc-protected aziridinyl methyl ketone

86 (prepared from 80 and Boc₂O in the presence of Et₃N and

DMAP, 78% yield, Scheme 4) as a substrate for the HWE

oleﬁnation reaction did not give the expected product (i.e., 88),

demonstrating its unsuitability to serve fruitfully, presumably

due to activation of the aziridine moiety imparted by the

carbamate group.¹⁹On the other hand, the use of the SEM

group as a protective device on the nitrogen atom (substrate

87, prepared from 80, as indicated in Scheme 4) led to success

in the HWE oleﬁnation reaction with phosphonate 41,

furnishing the corresponding aziridinyl epothilone B derivative

89 in 60% yield as a single geometrical isomer [(E):(Z) >

98:2]. Global desilylation of the latter (TFA) gave the targeted

aziridinyl methylthio epothilone B analogue 14 in 75% yield.

These ﬁndings broadened the scope of this synthetic strategy

toward other designed free aziridinyl epothilone B analogues

(see Schemes 7 and 12 below).

2.5. Molecular Design of Aziridinyl Epothilone B

Analogues. Having developed eﬃcient and stereoselective

synthetic strategies and protocols for the construction of

aziridinyl epothilones with appropriate side chains, as

demonstrated with the synthesis of 12,13-aziridinyl epothilone

B analogues 8−14 (Schemes 1, 3, and 4), we were in a position

to apply them to the synthesis of a variety of additional

aziridinyl epothilone B analogues. During the course of these

eﬀorts, and as we shall see below, unexpected obstacles arose

along the way that were overcome but not before a novel

modiﬁcation to the HWE oleﬁnation reaction was developed.

To this end, we designed 26 additional analogues (15−40,

Figure 3) as well as a series of corresponding β-heteroaromatic

F

DOI: 10.1021/jacs.7b02655

J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

a

Scheme 6. Synthesis of N-Hydroxyethyl Aziridinyl Epothilone B Analogues 27−31 from Methyl Ketone 82

a

Reagents and conditions: (a) 53 (31 equiv), n-BuLi (29 equiv); then 82 (1.0 equiv), THF, −78 → 0 °C, 3 h, 32%; 54 (60 equiv), n-BuLi (48

equiv); then 82 (1.0 equiv), THF, −78 → 0 °C, 3 h, 71%; (b) HF·py (xs), THF, 0 → 25 °C, 3 h, 80% for 27, 76% for 28; (c) 56 (15 equiv),

NaHMDS (14 equiv); then 82 (1.0 equiv), THF, −78 → 15 °C, 2.5 h, 33%; (d) HF·py (xs), THF, 0 → 25 °C, 3 h, 95%; (e) 55 (11 equiv),

NaHMDS (10 equiv); then 82 (1.0 equiv), THF, −78 → − 60 °C, 1 h, 57% [(E):(Z) ca. 75:25]; (f) HF·py (xs), THF, 0 → 25 °C, 3 h, 90%; (g) 57

(15 equiv), NaHMDS (14 equiv); then 82 (1.0 equiv), THF, − 60 °C, 1 h, 35% [(E):(Z) ca. 86:14], plus 34% recovered 82; (h) HF·py (xs), THF, 0

→ 25 °C, 3 h, 97%.

phosphonates (see Figure 4; for the preparation of these

phosphonates, see the Supporting Information). In designing

this 12,13-aziridinyl family of epothilones, we were mindful of

previous and encouraging results from our laboratories,²⁰

Altmann,²¹and BMS,¹⁵which indicated that the epoxide

moiety could be replaced with other isosteres, such as the

cyclopropane or aziridine structural motifs. The previous

aziridination eﬀorts, however, ended with the epothilone A

explorations due to synthetic methodology limitations at the

time.⁴Since epothilone B exhibits higher potencies than

epothilone A in general (ca. 10-fold),¹it was desirable to direct

our eﬀorts toward analogues of the former structure, a goal that

we can now attain and have broadly expanded upon with the

installation of the aziridine moiety and a variety of heterocyclic

side chains into the epothilone B precursor molecule. With

regard to the side chains, we also relied on previous studies

from our laboratories that suggested, for example, the

methylthiopyrazole¹²and methylthiothiazole²²moieties as

potency-enhancing structural motifs. The requirement of a

basic heteroatom (e.g., N) capable of acting as a hydrogen bond

acceptor as it is found in the natural epothilones was of course

maintained in all designs due to its proven essentiality. We also

incorporated a number of new ideas for binding optimization as

described below. Finally, and in order to fulﬁll the requirement

of sites for linker attachment or direct conjugation to

appropriate delivery systems, we adopted additional functional

groups, such as primary hydroxyl and amino moieties, among

others, into our designed molecules. A further rationale for

certain epothilone designs will be discussed with their syntheses

below.

82 and the corresponding phosphonates 43−52 through the

two-step sequence of oleﬁnation (NaHMDS or n-BuLi, see

Scheme 5) followed by global deprotection (HF·py or HF·py;

then TFA) in overall yields ranging from 32−87%, as indicated

in Scheme 5. Analogues possessing the pyridine (21),

benzothiazole (24), and N-methyl-5-methylthiopyrazole (23)

were synthesized (see Scheme 5) due to the reported high

potencies found in previous studies with such epothilone B side

chain analogues by Altmann and our group.^14,23In addition, the

aminothiazole (15), hydroxyethylthiazole (19), and amino-

ethylthiazole (20) containing analogues represent new

modiﬁcations to the natural product side chain, while the

analogues containing the underexplored methyloxazole

(16)^19,24and unknown methylthiooxazole (17) structural

motifs were also prepared. Furthermore, the design and

synthesis of bis-thiazolyl analogue 22 and 1,1-dipyrazolylmethyl

analogue 25 were based on the hypothesis that a well-

positioned second basic nitrogen atom in the vicinity of the

essential thiazole nitrogen atom (crucial for hydrogen bonding

of epothilones to β-tubulin)²⁵might result in an increase of

potencies for these compounds. Lastly, the peracetylated (18)

and monoacetylated (26) analogues were prepared by standard

acetylation conditions from analogues 17 and 25, respectively,

in order to explore the likelihood of enhanced potency as a

result of increased cellular membrane permeability.²⁶

Scheme 6 summarizes the preparation of N-hydroxyethyl

aziridinyl epothilone B analogues 27−31 possessing unnatural

heterocyclic moieties (i.e., oxadiazole 27, thiadiazoles 28 and

30, and isoxazoles 29 and 31) from the protected methyl

ketone 82 and phosphonates 53−57, respectively, in an eﬀort

to further probe structure−activity relationships (SARs) within

the side chain region. In a similar fashion to the synthesized

analogues described in Scheme 5, this protocol relied on the

standard oleﬁnation (NaHMDS or n-BuLi)/global deprotec-

tion (HF·py) strategy, providing good overall yields, as

indicated in Scheme 6. It is noteworthy that, while the majority

of the HWE reactions aﬀorded the desired (E) isomer

exclusively [(E):(Z) > 98:2], oxazole [analogue 16, Scheme

2.6. Synthesis of Designed N-Hydroxyethyl Aziridinyl

Epothilone B Analogues 15−31. The synthesis of designed

aziridinyl epothilone B analogues 15−31 (Figure 3) relied on

the optimal intermediates and conditions described above for

the synthesis of the aforementioned aziridinyl epothilone B

analogues (i.e., 12−14, Schemes 3 and 4). Scheme 5

summarizes the construction of N-hydroxyethyl aziridinyl

epothillone B analogues 15−26 from protected methyl ketone

G

DOI: 10.1021/jacs.7b02655

J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

5, (E):(Z) ca. 88:12] and isoxazole [analogues 29 and 31,

Scheme 6, (E):(Z) ca. 75:25 and 86:14, respectively]

phosphonates (i.e., 44, 55, and 57, respectively) also rendered

a small amount of the corresponding epothilone (Z) isomer but

were still satisfactorily diastereoselective. Through these

observations, it became apparent that the electron density of

the heterocyclic moiety plays a pivotal role in determining the

stereochemical outcome of this reaction (discussed further

below).

Scheme 8. Synthesis of Aziridinyl Epothilone B Analogues

34−36 from Analogue 12

a

2.7. Synthesis of Designed Free Aziridinyl Epothilone

B Analogues 32 and 33. Our aforementioned protecting

group strategy for the construction of free aziridinyl epothilone

B analogues (i.e., Scheme 4) was applied accordingly to the

synthesis of analogues 32 and 33 from SEM-protected methyl

ketone precursor 87, as shown in Scheme 7. Thus, oleﬁnation

Scheme 7. Synthesis of Free Aziridinyl Epothilone B

a

Analogues 32 and 33 from Methyl Ketone 87

a

Reagents and conditions: (a) Ts₂O (5.0 equiv), Et₃N (4.0 equiv),

CH₂Cl₂, 0 → 25 °C, 45 min; then NaN₃(4.0 equiv), DMF, 25 °C, 17

h, 40% overall for 34; (b) Ts₂O (2.0 equiv), Et₃N (2.0 equiv), CH₂Cl₂,

0 °C, 5 min; then NaSH (2.0 equiv), DMF, 0 → 15 °C, 1.5 h, 54%

overall for 35; (c) AcCl (2.0 equiv), i-Pr₂NEt (2.0 equiv), CH₂Cl₂, 0

°C, 1 h, 88%.

NaN₃(34, 40% overall yield) or NaSH (35, 54% overall yield);

selective acetylation of 12 with AcCl, i-Pr₂NEt (36, 88% yield),

see Scheme 8].

The syntheses of N-aminoethyl, N-cyclopropylmethyl, and

N-homopropargylic aziridinyl epothilone B analogues 37−39

are summarized in Scheme 9. Thus, N-alkylation of bis-TES

protected aziridinyl epothilone B analogue 85 (for its synthesis,

see Scheme 4) with bromide 90 led to precursor 91 (K₂CO₃,

32% yield), whose global deprotection (HF·py; then TFA, 65%

overall yield) aﬀorded the targeted analogue 37 equipped with

a primary amino group appropriate for conjugation to suitable

delivery systems. The synthesis of cyclopropyl analogue 38

started with aziridinyl methyl ketone 80 and bromide 92, whose

coupling under basic conditions (K₂CO₃, 92% yield) furnished

intermediate methyl ketone 93. Oleﬁnation of the latter with

phosphonate 41 (n-BuLi, 65% yield) then led exclusively to

precursor 94 [(E):(Z) > 98:2], whose desilylation with HF·py

aﬀorded the targeted analogue 38 in 92% yield. A similar

pathway from the same precursor (i.e., aziridinyl methyl ketone

80), employing homopropargylic bromide 95 as the alkylating

agent (K₂CO₃, 90% yield) provided methyl ketone inter-

mediate 96. The latter, upon oleﬁnation with the same

phosphonate (i.e., 41) under similar conditions, furnished

TMS-protected acetylenic precursor 97 (63% yield) from

which the desired analogue 39 was generated by global

deprotection (HF·py; then TBAF/AcOH, 68% overall yield), as

shown in Scheme 9. X-ray crystallographic analysis of 39 (mp

89−91 °C, see ORTEP in Scheme 9 and the Supporting

Information for details) unambiguously established the absolute

conﬁguration of this compound and its parent aziridine derived

from the newly developed synthetic route (i.e., 80, Scheme 3).

2.9. Observation and Isolation of Stable β-Hydrox-

yphosphonate Adducts. An intriguing feature of the HWE

reaction observed in the course of these investigations was the

formation of stable β-hydroxyphosphonate adducts with certain

substrates (Schemes 10 and 11).^27,28Thus, and as shown in

Scheme 10, treatment of methyl ketone substrates 82 and 87

with the carbanion generated from phosphonate 52 and n-BuLi

a

Reagents and conditions: (a) 49 (18 equiv), n-BuLi (15 equiv); then

87 (1.0 equiv) THF, −78 → 25 °C, 4 h, 34%, plus 50% recovered 87;

(b) TFA (xs), CH₂Cl₂, 0 → 25 °C, 2 h, 85%; (c) 52 (15 equiv), n-

BuLi (14 equiv); then 87 (1.0 equiv), THF, −78 °C, 40 min; then t-

BuOK (3.0 equiv), THF, − 20 °C, 5 min, 53% overall, plus 28%

recovered 87; (d) TFA (xs), CH₂Cl₂, 0 → 25 °C, 2 h, 52%.

of 87 with phosphonates 49 and 52 (n-BuLi) followed by

deprotection (TFA) delivered free aziridinyl epothilone B

analogues 32 (29% overall yield) and 33 (28% overall yield),

respectively, as indicated in Scheme 7. The lower yields were

compensated by the recovery of considerable amounts of

starting material (i.e., 87) for these challenging substrates (see

caption, Scheme 7).

2.8. Synthesis of Designed N-Substituted Aziridinyl

Epothilone B Analogues 34−39. To further explore the

SARs of this new library of epothilones, and in order to install a

variety of potential conjugation sites, we synthesized a series of

analogues with diﬀerent functional groups on the aziridine ring,

while maintaining a highly potent moiety in the side chain

region (i.e., methylthiothiazole). The results of these endeavors

are depicted in Schemes 8 and 9. Thus, and as shown in

Scheme 8, N-azidoethyl (34), N-thioethyl (35), and N-

acetoxyethyl (36) aziridinyl epothilone B analogues were

prepared from N-hydroxyethyl aziridinyl epothilone 12

(prepared as described above, Scheme 3) through selective

functionalization of the primary hydroxyl group [selective

tosylation with Ts₂O followed by tosylate displacement with

H

DOI: 10.1021/jacs.7b02655

J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

Scheme 9. Synthesis of Aziridinyl Epothilone B Analogue 37 from Protected Aziridinyl Epothilone B Analogue 85 and Synthesis

a

of Aziridinyl Epothilone B Analogues 38 and 39 from Methyl Ketone 80

a

Reagents and conditions: (a) 90 (6.0 equiv), K₂CO₃(5.0 equiv), DMF, 75 °C, 12 h, 32%, plus 35% recovered 85; (b) HF·py (xs), THF, 0 → 25

°C, 2 h; then TFA (xs), CH₂Cl₂, 0 °C, 1 h, 65% overall; (c) 92 (6.0 equiv), K₂CO₃(5.0 equiv), DMF, 75 °C, 16 h, 92%; (d) 41 (13 equiv), n-BuLi

(10 equiv); then 93 (1.0 equiv), THF, −78 → 10 °C, 1.5 h, 65%; (e) HF·py (xs), THF, 0 → 25 °C, 3.5 h, 92%; (f) 95 (5.0 equiv), K₂CO₃(4.0

equiv), DMF, 80 °C, 18 h, 90%; (g) 41 (20 equiv), n-BuLi (17 equiv); then 96 (1.0 equiv), THF, −78 → 0 °C, 3 h, 63%; (h) HF·py (xs), THF, 0

°C, 2 h; then TBAF/AcOH (1:1, xs), THF, 25 °C, 8 h, 68% overall.

provided β-hydroxyphosphonate adducts 98 and 99, respec-

tively. These adducts were observed by thin layer chromatog-

raphy and high resolution mass spectrometry (HRMS) but

were not puriﬁed. Rather, the crude material was subsequently

treated with t-BuOK to furnish protected analogues 100 (73%

overall from 82, plus 10% recovered 82) and 101 (53% overall

from 87, plus 28% recovered 87), respectively, with exclusive

(E) geometry [(E):(Z) > 98:2].²⁹Note: Compounds 100 and

101 are the intermediates through which analogues 25 and 33

were produced from 82 and 87, respectively, as summarized in

Schemes 5 and 7. This result is reminiscent of the pioneering

work of Seyden-Penne and Bottin-Strzalko, who reported that

the diethyl ester of phenylmethylphosphonic acid with

benzaldehyde aﬀorded only trans-stilbene.³⁰However, further

experimentation in our studies showed that the base-induced

elimination of β-hydroxyphosphonates derived from β-hetero-

aromatic phosphonates and methyl ketones does not always

provide high (E) selectivity, as we discuss below.

Thus, and as shown in Scheme 11, treatment of phosphonate

50 with NaHMDS followed by transfer of the resulting

carbanion into a solution of methyl ketone 82 delivered adduct

102 as a mixture of diastereoisomers (ca. 3:7), which proved

chromatographically stable, allowing its isolation in pure form

(but still as an inseparable diastereomeric mixture) in 37%

yield. Subsequently, treatment of this mixture (102) with t-

BuOK produced protected aziridinyl epothilone B analogue

103 (75% yield) as an inseparable mixture of geometrical

isomers [(E)-103:(Z)-103 ca. 3:7, Scheme 11]. The conserved

diastereomeric ratio (ca. 3:7) of 102 and 103 is in accordance

with the classical mechanistic studies of β-ketophospho-

nates^17,31(i.e., the 3:7 dr for 102 corresponds to the 3:7 syn/

anti adduct distribution; see Figure 5). Studies by Petrova and

co-workers have deﬁned this conserved ratio phenomenon as

“chemical proof ” of the relative conﬁgurations of the two β-

hydroxyphosphonate adducts.^27aThis observation suggests that

equilibration to the syn β-hydroxyphosphonate does not occur

with this particular substrate (see Figure 5), a result that

mirrors previous studies from Tsuge and co-workers,^28awho

reported the reaction of β-furanylphosphonates with methyl

ketones to provide oleﬁns with geometrical ratios matching the

diastereomeric ratio of the corresponding β-hydroxyphospho-

nate adducts. Interestingly, the major diastereoisomer of adduct

102 is presumably the anti isomer, a result that stands in

contrast to previous reports from Petrova, who observed that

the reaction of benzylphosphonates with aldehydes or ketones

displayed marked syn selectivity.^27a,d

2.10. Ethoxy Group Substituent Eﬀects on the HWE

Reaction of β-Heteroaromatic Phosphonates with

Methyl Ketones 82 and 87. As described above, the majority

of the standard ethoxyphosphonates (i.e., 41−49, 51−57)

performed well in their reactions with the corresponding

methyl ketone epothilone precursors, both in terms of yield and

geometrical selectivity. The lack of selectivity and room for

yield improvement of the standard ethoxyphosphonate 50

employed for the synthesis of the coveted N-methyl-5-

methylthiopyrazole analogue 23 (see Scheme 11), however,

prompted us to undertake an investigation in search of a more

stereoselective HWE reaction for this β-heteroaromatic

phosphonate. We suspected that the latter predicament was

due to an unfavorable imbalance between the electron-donating

nature of the electron-rich methylthiopyrazole moiety and the

electron-withdrawing ability of the phosphonate group. To

I

DOI: 10.1021/jacs.7b02655

J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

Scheme 10. Observation of HWE Adducts 98 and 99 and

Treatment with Base to Provide Protected Aziridinyl

a

Epothilone B Analogues 100 and 101

Figure 5. Classic Horner−Wadsworth−Emmons mechanism for oleﬁn

geometry formation.³²Geometrical selectivity is a function of (1) the

stereochemistry of the initial carbonyl addition step and (2) the ability

of the intermediates to equilibrate. Equilibration of syn and anti β-

hydroxyphosphonate adducts favors the trans oxaphosphetane,

ultimately leading to the desired thermodynamic product [(E) oleﬁn].

EWG = electron-withdrawing group.

Table 1. Deuterium Labeling Studies of Phosphonate 50

c

entry

base

solvent

THF

ϑ (°C)

t (h) % D

a

Reagents and conditions: (a) 52 (11 equiv), n-BuLi (10.0 equiv);

1

2

3

4

5

6

NaHMDS

0 → 25

−78

25

0.5

0.3

2.5

2

25

45

30

35

80

85

a

then 82 (1.0 equiv), −78 °C, 1.5 h; then t-BuOK (5.0 equiv), THF,

−20 °C, 5 min, 73% overall, plus 10% recovered 82; (b) 52 (15

equiv), n-BuLi (14 equiv); then 87 (1.0 equiv), THF, −78 °C, 40 min;

then t-BuOK (5.0 equiv), THF, −20 °C, 5 min, 53% overall, plus 28%

recovered 87.

t-BuLi

Et₂O

a

n-BuLi

THF/HMPA

hexanes/HMPA

THF

a

t-BuLi

b

n-BuLi/t-BuOK

0.3

2

b

KH

THF

a

b

Reactions were performed using 0.8 equiv of base. Reactions were

Scheme 11. Detection and Isolation of HWE Adduct 102 and

Treatment with Base to Provide Protected Aziridinyl

c

performed with excess base. Approximate D incorporation, as

determined by ¹H NMR spectroscopic analysis of crude reaction

mixtures. See the Supporting Information for details.

a

Epothilone B Analogue 103

discussed above, was ineﬀective (25% D incorporation after

0.5 h), even at ambient temperature (Table 1, entry 1). More

surprisingly, treatment of phosphonate 50 with t-BuLi or n-

BuLi in a variety of solvents, in the absence or presence of

HMPA (see Table 1, entries 2−4) at −78 °C, led only to partial

deprotonation, as evidenced from low deuterium incorporation

(25−45%) in our experiments. In contrast, exposure of

phosphonate 50 to Schlosser’s base (n-BuLi/t-BuOK)³³in

THF at −78 °C or KH in DMF at 25 °C led to 80 and 85%

deuterium incorporation, respectively, indicating a high degree

of deprotonation (Table 1, entries 5 and 6, respectively).

Unfortunately, however, these conditions proved too harsh for

the methyl ketone substrate (i.e., 82), precluding the formation

of the expected product due to substrate decomposition.

These failures (see Table 1 and Table 2, entries 1 and 2)

motivated us to undertake an exploration of an alternative

avenue to activate the phosphonate (i.e., 50) by decreasing its

pK_ain order to enhance its reactivity. Reasoning that electron-

withdrawing residues on the alkyl groups of the phosphonate

moiety would accomplish this goal, we investigated a number of

ﬂuorinated and chlorinated derivatives, as shown in Table 2.

After our fruitless attempts with standard phosphonate 50, we

then employed the 2,2,2-triﬂuoroethoxyphosphonate 58

(adopted from the Still−Gennari method³⁴and prepared as

described in the Supporting Information), which reacted with

a

Reagents and conditions: (a) 50 (11 equiv), NaHMDS (4.9 equiv);

then 82 (1.0 equiv), THF, −78 °C, 1 h, 37%; (b) t-BuOK (5.0 equiv),

THF, − 20 °C, 5 min, 75%.

verify this hypothesis, we subjected phosphonate 50 to the

action of a series of bases in diﬀerent solvents, and assessed its

relative deprotonation by D₂O quenching of the resulting

carbanion and subsequent ¹H NMR spectroscopic analysis (see

the Supporting Information for further details). These results

are listed in Table 1. The use of NaHMDS, which has been

suitable for the deprotonation of several phosphonates

J

DOI: 10.1021/jacs.7b02655

J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

Table 2. Ethoxy Group Substituent Eﬀects for the Synthesis of Protected Aziridinyl Epothilone B Analogue 103

a

b

entry

R

% D [t (h)]

base

ϑ (°C)

−78

−78 → 25

−78

t (h)

[(E):(Z)] (yield)

1

2

3

4

5

6

7

8

CH₂CH₃(50)

CH₂CF₃(58)

25 (0.3)

45 (0.5)

70 (0.5)

40 (0.3)

20 (0.4)

30 (1.5)

80 (0.3)

n-BuLi

0.2

1

c

NaHMDS

n-BuLi

d

2

1:1 (70%)

CH₂CH₂CF₃(59)

CH₂CCl₃(60)

CH₂CHF₂(61)

CH₂CH₂F (62)

n-BuLi

2

c

n-BuLi

3

e

NaHMDS

2.5

2

c

1:1 (52%)

50 (0.5)

1

0.5

c

a

Approximate D incorporation as determined by H NMR spectroscopic analysis of crude sample after quenching the phosphonate/base mixture

b

c

after the designated time with D₂O. See the Supporting Information for details. Reactions were performed using 0.8 equiv of base. Unidentiﬁed

mixture. Formation of β-hydroxyphosphonate adducts (102, see Scheme 11). Decomposition of phosphonate observed.

d

e

methyl ketone 82 in the presence of n-BuLi to aﬀord the

desired product (i.e., 103) in 70% yield, but unfortunately as a

1:1 geometrical ratio of inseparable oleﬁnic isomers at the

newly formed trisubstituted oleﬁnic bond (Table 2, entry 3).

Interestingly, and despite its impact on organic synthesis, the

traditional Still−Gennari stereoselective oleﬁnation reaction

with β-heteroaromatic phosphonates has not been reported to

our knowledge. Furthermore, we are not aware of a systematic

exploration of the eﬀect of various halogen substituents on

diethoxyphosphonates in any context outside the original Still−

Gennari protocol. As such, our observations with ﬂuorinated

phosphonate 58 and methyl ketone 82 (Table 2, entry 3) led us

to investigate the reactivities of additional phosphonates

containing ﬂuorine and chlorine substituents in search of a

stereoselective oleﬁnation of methyl ketone substrates 82 and

87.

We then prepared 3,3,3-triﬂuoropropyloxy- and 2,2,2-

trichloroethoxyphosphonates 59 and 60 (for preparation

details, see the Supporting Information), but unfortunately,

they did not perform well in their intended coupling with

substrate 82, as shown in Table 2 (entries 4−6). Interestingly,

the 2,2-diﬂuorophosphonate 61 (for preparation details, see the

Supporting Information) in combination with NaHMDS

reacted smoothly with methyl ketone 82 at −78 °C to give

the desired oleﬁnic product (i.e., 103) in 52% yield, albeit as a

mixture of geometrical isomers [(E):(Z) ca. 1:1, Table 2, entry

7], echoing the result with the 2,2,2-triﬂuoroethoxyphospho-

nate 58 (Table 2, entry 3). Lastly, we decided to employ 2-

ﬂuoroethoxyphosphonate 62 (Table 2, entry 8). While its

attempted reaction with NaHMDS and substrate 82 led to its

decomposition (Table 2, entry 8), this phosphonate provided,

upon further experimentation, a solution to the geometrical

selectivity we were seeking in our quest for aziridinyl

epothilone B 103, as discussed below.

−78 °C for 10 min followed by addition of the resulting

carbanion into a solution of methyl ketone 87 and then

quenching the reaction mixture after 1 h at −78 °C aﬀorded

adduct 104 in 28% yield (plus 42% recovered 87) as an

inseparable mixture of diastereoisomers (dr ca. 3:7), which

proved chromatographically stable. Remarkably, exposure of

this adduct (104) to t-BuOK delivered the desired protected

epothilone analogue 105 in 63% yield and with exclusive (E)

oleﬁnic geometry [(E):(Z) > 98:2]. Further reﬁnement of these

conditions led to a practical, one-step generation of the oleﬁnic

product (105) from the reaction of methyl ketone 87 and

phosphonate 62, as summarized in Scheme 12B. Thus,

treatment of methyl ketone 87 with the carbanion generated

from phosphonate 62 and n-BuLi at −78 to 0 °C provided,

after global deprotection (TFA), aziridinyl epothilone B

analogue 40 (52% overall yield) in a highly stereoselective

fashion [(E):(Z) ca. 92:8]. Similarly, and as shown in Scheme

12B, methyl ketone 82 reacted (at −78 to −40 °C) with the

anion of phosphonate 62, generated in the presence of n-BuLi

(−78 °C), to aﬀord precursor (E)-103 (62% yield), whose

global desilylation with HF·py furnished the targeted aziridinyl

epothilone B analogue (E)-23 in 82% yield. In these cases, and

in contrast to our result in Scheme 12A (HWE reaction of 87

plus 62 quenched at −78 °C), the HWE coupling between

methyl ketones 87 or 82 and phosphonate 62 was allowed to

warm to 0 or −40 °C (Scheme 12B), inducing the

corresponding in situ formed β-hydroxyphosphonates (e.g.,

104) to undergo elimination to provide directly protected

precursors 105 and (E)-103 of aziridinyl epothilone B

analogues 40 and (E)-23, respectively.

Collectively, the results shown in Scheme 12 demonstrate

the thermodynamic control of this reaction, with higher

temperatures favoring equilibration of the reaction intermedi-

ates and/or other thermodynamic factors (e.g., pK_aeﬀects),

such as epimerization at the phosphorus-bearing carbon center,

that promote (E) oleﬁnic bond formation with high stereo-

selectivity [(E):(Z) > 98:2]. These results led us to propose

mechanistic rationales for the diﬀerent stereochemical out-

comes of the reactions of phosphonates 50 [(E):(Z) ca. 3:7]

and 62 [(E):(Z) > 98:2] with methyl ketone substrates 82 and

87 (Figures 6 and 7, respectively). Thus, and as shown in

Scheme 12A shows our experiments that led to the isolation

and identiﬁcation of the intermediate β-hydroxyphosphonates

(i.e., 104, inseparable mixture of diastereoisomers, dr ca. 3:7)

formed from methyl ketone 87 and phosphonate 62

[synthesized in 78% overall yield from phosphonate 50 by

stepwise treatment with TMSCl, (COCl)₂, 2-ﬂuoropropanol, as

summarized in Scheme 12]. Thus, reaction of 62 with n-BuLi at

K

DOI: 10.1021/jacs.7b02655

J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

Scheme 12. Synthesis of Phosphonate 62 from Phosphonate

50 and Detection, Isolation, and Treatment of HWE Adduct

104 with Base to Provide Protected Aziridinyl Epothilone B

Analogue 105 and Stereoselective Syntheses of Analogues 23

a

and 40

a

Reagents and conditions: (A) (a) TMSCl (5.1 equiv), 80 °C, 72 h;

then (COCl)₂(2.5 equiv), CH₂Cl₂, 0 → 25 °C, 4 h; then 2-

ﬂuoropropanol (4.0 equiv), Et₃N (6.0 equiv), DMAP (0.02 equiv),

CH₂Cl₂, 0 → 25 °C, 12 h, 78% overall; (b) 62 (22 equiv), n-BuLi (18

equiv), 10 min; then 87 (1.0 equiv), THF, −78 °C, 1 h, 28%, plus 42%

recovered 87; (c) t-BuOK (5.0 equiv), THF, − 20 °C, 5 min, 70%. (B)

(d) 62 (22 equiv), n-BuLi (18 equiv); then 87 (1.0 equiv), THF, −78

→ 0 °C, 4 h, 63%; (e) TFA (xs), CH₂Cl₂, 0 → 25 °C, 2 h, 82%; (f) 62

(16 equiv), n-BuLi (13 equiv); then 82 (1.0 equiv), THF, −78 → − 40

°C, 1 h, 62%; (g) HF·py (xs), THF, 0 → 25 °C, 5 h, 82%. TMS =

trimethylsilyl.

Figure 6. Proposed rationale for the (Z)-selective synthesis of

protected aziridinyl epothilone B analogue 103.

Figure 6, substrate 82 reacts with the carbanion generated from

ethoxyphosphonate 50 (NaHMDS) to aﬀord a mixture of syn

and anti β-hydroxyphosphonates (syn-102 and anti-102, dr ca.

3:7). As can be seen from their Newman projections, these

diastereomers are poised to form, under the inﬂuence of t-

BuOK, the corresponding oxaphosphetane intermediates A and

B, respectively, which faithfully collapse to the (E)- and (Z)-

oleﬁnic products through a trans and cis elimination,

respectively (see Figure 6), and with preservation of the initial

diastereoselectivity [(E):(Z) ca. 3:7]. It is the stability of

diastereoisomers syn-102 and anti-102 toward C17 epimeriza-

tion under the reaction conditions and/or the resistance of

these adducts to revert back to starting materials (i.e., 82 and

50) that accounts for their stereospeciﬁc conversion to the

corresponding (E) and (Z) oleﬁns. Had one or both of these

events occurred [e.g., Figure 6, anti-102 equilibrating with syn-

102 (via reversal back to 82 and 50, respectively) or with 17-

epi-anti-102 (via epimerization) and their conversion to oleﬁn

(E)-103], then the faithful conversion of these β-hydrox-

yphosphonate adducts would have been disturbed.

Similar to the ethoxyphosphonate 50, and as shown in Figure

7, the bis(2-ﬂuoroethoxy)phosphonate 62 also reacts with

methyl ketone 87, this time under the inﬂuence of n-BuLi, to

initially aﬀord β-hydroxyphosphonates syn-104 and anti-104 in

a similar ratio (dr ca. 3:7, isolated, structures tentatively

assigned). To explain the exclusive conversion of this mixture of

adducts, after further treatment with t-BuOK, to (E)-oleﬁnic

product (E)-105, we envision epimerization of the anti-

diastereoisomer (anti-104) to 17-epi-anti-104 (see Newman

projection), which ﬁnds its way to (E)-oleﬁn (E)-105 via

oxaphosphetane F, the latter undergoing trans elimination. We

note that oxaphosphetane intermediates D and F (Figure 7)

L

DOI: 10.1021/jacs.7b02655

J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

This exploration resulted in a unique application of the HWE

reaction of β-heteroaromatic 2-ﬂuoroethoxyphosphonate 62 in

the synthesis of oleﬁnic heterocyclic compounds [e.g., (E)-103

and (E)-105]. Further applications of these types of

phosphonates, including other β-hetereocyclic and β-keto-

phosphonates, are expected to ﬁnd applications in synthetic

organic and medicinal chemistry.

While we were fortunate to observe an exceptionally broad

substrate scope in terms of heterocyclic moieties throughout

our investigations, it should be noted that phosphonates 63−68

(Figure 4) proved recalcitrant, thus far, toward delivering the

desired oleﬁnation products. Speciﬁcally, reaction of methyl

ketone 82 with thiadiazolyl phosphonate 63 and oxadiazolyl

phosphonate 64 produced an unidentiﬁable mixture of side

products, while imidazolyl phosphonate 65 and N-arylpyrazolyl

phosphonates 66−68 resulted only in recovered starting

materials.

2.11. Biological Evaluation of Synthesized Aziridinyl

Epothilone B Analogues and Structure−Activity Rela-

tionships (SARs). The synthesized aziridinyl epothilone

analogues (8−40) were subjected to biological evaluation

with regard to their cytotoxicity against several human cancer

cell lines (including drug resistant cell lines, see Tables 3 and 4)

and tubulin polymerization properties (Table 3). Epothilone D

methylthiopyrazolyl analogue 73 (Scheme 2) and aziridinyl

truncated epothilone intermediates 79 and 81 (Scheme 3)

lacking the heterocyclic side chain, epothilones A−D (1−4,

Figure 1), ixabepilone (5, Figure 1), the tubulin stabilizing

agent monomethyl auristatin E [MMAE, the payload of

brentuximab vedotin (Adcetris)], and N-acetyl calicheamicin

γ^I₁[NAC, the payload of gemtuzumab ozogamicin (Mylotarg)]

were also tested alongside a number of these synthesized

compounds for comparison purposes (see Tables 3 and 4).

As can be seen in Table 3 [National Cancer Institute (NCI)

data], there is considerable correlation (but not always)

between tubulin polymerization and cytotoxic potencies for

the synthesized compounds, with aziridinyl epothilone B

analogues 9, 14, 17, 23, 34, 36, and 39 exhibiting the most

impressive potencies against an array of human cancer cell lines,

including a highly drug resistant human ovarian cancer cell line

[NCI/ADR-RES, GI₅₀= 3.2−8.8 nM vs paclitaxel: GI₅₀= 4800

nM; ixabepilone (5): GI₅₀= 1400 nM; see Table 3]. Table 4

(Abbvie-Stemcentrx data) shows the potencies of the

synthesized aziridinyl epothilone B analogues against human

uterine sarcoma cell line MES SA, human uterine sarcoma cell

line with marked multidrug resistance MES SA DX, and human

embryonic kidney cancer cell line HEK 293T. The most

impressive potencies in these studies were exhibited by

analogues 10, 12, 13, 17, 24, 34, and 36−39, all in the

Figure 7. Proposed rationale for the stereoselective synthesis of

protected aziridinyl epothilone B analogue (E)-105.

subnanomolar range (IC₅₀down to 0.02 nM vs MMAE: IC₅₀

≥

0.068 nM; NAC: IC₅₀≥ 0.166 nM; paclitaxel: IC₅₀≥ 1.76 nM;

see Table 4) against all three cell lines.

represent enantiomeric transition states, thus providing the

same desired (E) oleﬁnic product upon trans elimination.

We propose that the facile equilibration of anti-104 and 17-

epi-anti-104 is a consequence of the electron-withdrawing eﬀect

of the ﬂuorine residues on phosphonate 62, which renders the

17H more acidic than its counterpart in the ethoxyphosphonate

50, which apparently does not undergo epimerization under

similar conditions (see Figure 6). However, we cannot exclude

the possibility that these electron-withdrawing groups cause

reversibility to occur (i.e., retro-HWE) in the system, thereby

promoting the thermodynamically favorable formation of the

(E) oleﬁn (see Figure 5).

Many of these compounds demonstrated low picomolar

activities against the tested cell lines, including the MES SA DX

cell line with marked multidrug resistance (IC₅₀down to 0.51

nM vs MMAE: IC₅₀= 88.19 nM; NAC: IC₅₀= 15.31 nM;

paclitaxel: IC₅₀> 400 nM; see Table 4). The high potencies of a

number of these aziridinyl epothilone B analogues and their

reactive chemical handles qualify them as promising payloads

for antibody−drug conjugates.

It was also interesting to note the relatively high potencies of

the 12,13-oleﬁnic pyrazolyl epothilone D (lacking the epoxide

moiety) analogue 73 (for structure see Scheme 2; GI₅₀= 5.0−

M

DOI: 10.1021/jacs.7b02655

J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

Table 3. Tubulin Assembly Induction (EC₅₀) and Cytotoxicity (GI₅₀) Assay Data

a

b

EC₅₀is the drug concentration yielding an unbound protein supernatant 50% that of controls. GI₅₀is the compound concentration required to

c

d

e

decrease cell growth by 50%. Human breast cancer cell line. Human ovarian cancer cell line. Highly drug resistant human ovarian cancer cell line.

f

g

h

Human melanoma cell line. Human glioblastoma cell line. Not tested. For more details, see the Supporting Information. These data were

obtained at the Screening Technologies Branch of the Frederick National Laboratory for Cancer Research, National Cancer Institute, National

Institutes of Health. Colored background rows highlight notably potent compounds.

52 nM, Table 3, and 0.17−39.18 nM, Table 4). More intriguing

is the observation of signiﬁcant cytotoxicity of truncated methyl

ketone aziridine epothilone 81 (as opposed to 79, for structures

see Scheme 3; see Tables 3 and 4) despite its lack of a

heterocyclic side chain. These studies allowed us to formulate

further structure−activity relationships (SARs) within the

epothilone structural class, which conﬁrmed and expanded

previously developed SARs.^1bThus, the C12−C13 aziridinyl

epothilone B analogues proved to be generally more potent

than their corresponding epoxide analogues, which in turn are

known to be more potent than their oleﬁnic precursors (with

the exception of epothilone D analogue 73). These

observations are in line with the notion that the C12−C13

region of the molecule is not involved in critical binding

interactions with β-tubulin, the biological target of the

epothilones. This concept was previously conﬁrmed by X-ray

crystallographic analysis of an epothilone−tubulin complex.^25b

Most importantly, the impressively wide range of functional

group tolerability on the nitrogen atom of the aziridine moiety

is remarkable. Figure 8A demonstrates this tolerance within a

family of these analogues that carry the same heterocyclic side

chain (i.e., methylthiothiazole). In addition, the N-substituents

can vary from nonpolar (e.g., cyclopropylmethyl) to polar (e.g.,

aminoethyl) residues, as reﬂected in Figure 8B. Interestingly,

there seems to be no strict correlation between substituent

polarity and potency, as deduced by inspection of the potency−

structural motif trend shown in Figure 8C. The broad

functional group tolerance of the aziridine substituents is also

evident within those analogues exhibiting strong actions against

multidrug resistant human cancer cell lines (e.g., 12, 14, 34, 36,

and 39, see Tables 3 and 4), making this domain of the

epothilone molecule fertile ground for further exploitation

toward much needed anticancer therapies.

N

DOI: 10.1021/jacs.7b02655

J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

Table 4. Cytotoxicity Data against the Cancer Cell Lines

MES SA, MES SA DX, and HEK 293T for Epothilone

Analogues

Figure 8. (A) Summary of C12−C13 aziridine SAR. (B) Aziridines

ranked in order of increasing polarity. (C) Aziridines ranked in order

of increasing potency. A wide variety of functionality is tolerated, and

functional group polarity does not necessarily correlate with potency.

3. CONCLUSION

A practical, hitherto not achieved,⁴process for the synthesis of

C12−C13 aziridinyl epothilone B analogues from the natural

and readily available epothilone B was developed. Involving a

short degradative sequence to an oleﬁnic methyl ketone

substrate, followed by Ess−Kurti−Falck aziridination¹⁰and

̈

subsequent HWE coupling of a β-heteroaromatic phosphonate,

this process provided facile entry to a series of novel aziridinyl

epothilone B analogues. Biological evaluation of the synthesized

analogues led to the discovery of a number of highly potent

epothilones, some possessing low picomolar potencies against

human cancer cell lines, including drug resistant lines. Of

particular interest are analogues that qualify them, by virtue of

their unprecedented potencies and appropriate handles for

conjugation, as payloads for antibody-drug conjugates and

other delivery systems. In addition, the described chemical and

biological investigations revealed a number of useful SARs,

thereby expanding our knowledge and enhancing our ability to

design additional analogues for further biological studies and

drug discovery eﬀorts. Alongside these drug discovery eﬀorts,

our extensive investigations into the HWE coupling of β-

heteroaromatic phosphonates with our methyl ketone sub-

strates led to the development of 2-ﬂuoroethoxyphosphonates

for the (E)-selective synthesis of trisubstituted oleﬁns with an

electron-rich heteroaromatic system. Finally, this work also

provides an inspiring demonstration of the power of the newly

a

IC₅₀is the 50% inhibitory concentration of the compound against cell

b

growth. For more details, see the Supporting Information. Human

uterine sarcoma cell line. MES SA cell line with marked multidrug

resistance. Human embryonic kidney cancer cell line. These data

were obtained at Abbvie Stemcentrx. Colored background rows

highlight notably potent compounds.

c

d

developed Ess−Kurti−Falck aziridination reaction¹⁰in the

̈

synthesis of complex molecules relevant to biology and

medicine.

In contrast to the C12−C13 region, we found that the

substituent tolerance for high potency was much more

restricted in the side chain region of the molecule. Thus,

only a select few of the 20 heterocyclic side chains within a

speciﬁc C12−C13 aziridinyl epothilone B family (i.e., N-

hydroxyethyl aziridine) proved to be highly active, all

possessing the requisite basic nitrogen atom within the

heterocycle in a nearly identical position. Despite the rather

restricted range of tolerable side chain variations, however,

these studies revealed a new and highly empowering

heterocyclic structural motif, that of the methylthiooxazole

included in the structure of analogue 17, one of the most

potent epothilones ever reported.

ASSOCIATED CONTENT

* Supporting Information

The Supporting Information is available free of charge on the

ACS Publications website at DOI: 10.1021/jacs.7b02655.

■

S

Experimental procedures and characterization data for all

compounds; cytotoxicity data and tubulin polymerization

studies (NCI, Frederick); and HEK 293T, MES SA, and

MES SA DX (Abbvie Stemcentrx) data (PDF)

Crystallographic information for epothilone 10 (CIF)

Crystallographic information for epothilone 39 (CIF)

O

DOI: 10.1021/jacs.7b02655

J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

(d) Vite, G. D.; Lee, F. Y.; Leamon, C. P.; Vlahov, I. R. United States

Patent US 20070276018 A1, 2007. (e) Vite, G. D.; Lee, F. Y.; Leamon,

C. P.; Vlahov, I. R. World Patent WO 2007140297 A2, 2007. (f) Vite,

G. D.; Lee, F. Y.; Leamon, C. P.; Vlahov, I. R. World Patent WO

2007140298 A1, 2007. (g) Parlanti, L.; Yu, J. World Patent WO

2008147941 A1, 2008.

AUTHOR INFORMATION

Corresponding Author

*kcn@rice.edu

ORCID

K. C. Nicolaou: 0000-0001-5332-2511

Derek Rhoades: 0000-0002-9536-424X

■

(5) (a) Vite, G. D.; Lee, F. Y.; Leamon, C. P.; Vlahov, I. R. United

States Patent US 20070275904 A1, 2007. (b) Kim, S.-H.; de Mas, N.;

Author Contributions

Parlanti, L.; Lyngberg, O. K.; Strohlein, G.; Guo, Z.; Dambalas, K.;

̈

^∥D.R., Y.W.: These authors contributed equally to this work.

Rosso, V. W.; Yang, B.-S.; Girard, K. P.; Manaloto, Z. A.; D’Arasmo,

G.; Frigerio, R. E.; Wang, W.; Lu, X.; Bolgar, M. S.; Gokhale, M.;

Thakur, A. B. Org. Process Res. Dev. 2011, 15, 797. (c) Gokhale, M.;

Thakur, A.; Rinaldi, F. Drug Dev. Ind. Pharm. 2013, 39, 1315.

(6) (a) Clinical trial data for 6 (Novartis) was not publicly disclosed.

Notes

The authors declare no competing ﬁnancial interest.

ACKNOWLEDGMENTS

■

̀

(b) For a phase I clinical trial of 7, see: Sessa, C.; Perotti, A.; Llado, A.;

Cresta, S.; Capri, G.; Voi, M.; Marsoni, S.; Corradino, I.; Gianni, L.

Ann. Oncol. 2007, 18, 1548 10.1093/annonc/mdm198.

(7) Peethambaram, P. P.; Hartmann, L. C.; Jonker, D. J.; de Jonge,

M.; Plummer, E. R.; Martin, L.; Konner, J.; Marshall, J.; Goss, G. D.;

Teslenko, V.; Clemens, P. L.; Cohen, L. J.; Ahlers, C. M.; Alland, L.

Invest. New Drugs 2015, 33, 321.

We thank Drs. Lawrence B. Alemany and Quinn Kleerekoper

(Rice University) for NMR-spectroscopic assistance, Drs.

Christopher L. Pennington (Rice University) and Ian

Riddington (University of Texas at Austin) for mass

spectrometric assistance, and Professor Arnold Rheingold

(University of California, San Diego) for the X-ray crystallo-

(8) (a) Chari, R. V. J.; Miller, M. L.; Widdison, W. C. Angew. Chem.,

Int. Ed. 2014, 53, 3796. (b) Srinivasarao, M.; Galliford, C. V.; Low, P.

S. Nat. Rev. Drug Discovery 2015, 14, 203.

(9) (a) Nicolaou, K. C.; He, Y.; Ninkovic, S.; Pastor, J.; Roschangar,

F.; Sarabia, F.; Vallberg, H.; Vourloumis, D.; Winssinger, N.; Yang, Z.;

King, N. P.; Finlay, M. R. V. World Patent WO 9825929 A1, 1998.

(b) Nicolaou, K. C.; King, N. P.; Finlay, M. R. V.; He, Y.; Roschangar,

F.; Vourloumis, D.; Vallberg, H.; Bigot, A. World Patent WO 9967252

A2, 1999. (c) Nicolaou, K. C.; Hepworth, D.; Finlay, M. R. V.; King,

N. P. World Patent WO 9967253 A2, 1999. (d) Nicolaou, K. C.;

graphic analyses. We also thank Prof. Laszlo Kurti (Rice

́

̈

University) for helpful discussions. This work was supported by

the National Institutes of Health (USA) (grant AI055475), the

Cancer Prevention & Research Institute of Texas (CPRIT),

and The Welch Foundation (grant C-1819). Bristol−Myers

Squibb is gratefully acknowledged for their generous gift of

epothilones A−D. Disclaimer by the NCI co-authors (R.B. and

E.H.): The content of this paper is solely the responsibility of

the authors and does not necessarily reﬂect the oﬃcial views of

the National Institutes of Health, USA.

Namoto, K.; Ritzen

Patent WO 03018002 A2, 2003. (e) Namoto, K.; Nicolaou, K. C.;

Ritzen, A. World Patent WO 2004014919 A1, 2004. (f) Nicolaou, K.

́

, A.; Ulven, T.; Shoji, M.; Altmann, K.-H. World

REFERENCES

́

■

C.; Pratt, B.; Arseniyadis, S. World Patent WO 2007062288 A2, 2007.

(g) Nicolaou, K. C.; Rhoades, D.; Wang, Y.; Totokotsopoulos, S.; Bai,

R.; Hamel, E. ChemMedChem 2015, 10, 1974. (h) Rhoades, D.

Synthesis of Natural and Designed Antitumor Agents: Epothilones and

Thailanstatin A. Ph.D. Thesis, University of California, San Diego, Aug

2016. (i) Nicolaou, K. C.; Rhoades, D.; Wang, Y. U.S. Patent

Application US 2016057093, 2016.

(1) (a) Nicolaou, K. C.; Snyder, S. A. Epothilones A and B. Classics in

Total Synthesis II: More Targets, Strategies, Methods; Wiley-VCH:

Weinheim, Germany, 2003; pp 161−210. (b) Altmann, K.-H.; Pfeiffer,

B.; Arseniyadis, S.; Pratt, B. A.; Nicolaou, K. C. ChemMedChem 2007,

2, 396. (c) Altmann, K.-H.; Hofle, G.; Muller, R.; Mulzer, J.; Prantz, K.

̈

In The Epothilones: An Outstanding Family of Anti-Tumor Agents;

Kinghorn, A. D., Falk, H., Kobayashi, J., Eds.; Progress in the

Chemistry of Organic Natural Products 90; Springer: Berlin, Germany,

2009. (d) Pfeiffer, B.; Kuzniewski, C. N.; Wullschleger, C.; Altmann,

K.-H. Macrolide-Based Microtubule-Stabilizing Agents − Chemistry

and Structure-Activity Relationships. In Tubulin-Binding Agents.

Synthetic, Structural, and Mechanistic Insights; Carlomagno, T., Ed.;

Topics in Current Chemistry 286; Springer: Berlin, Germany, 2009;

pp 5−33. (e) Altmann, K.-H.; Gaugaz, F. Z.; Schiess, R. Mol. Diversity

2011, 15, 383. (f) Pfeiffer, B.; Gaugaz, F. Z.; Schiess, R.; Altmann, K.-

H. Epothilones as Lead Structures for New Anticancer Drugs. In Drug

Discovery from Natural Products; Genilloud, O., Vicente, F., Eds.; RSC

Drug Discovery Series 25; The Royal Society of Chemistry:

Cambridge, U.K., 2012; pp 339−373. (g) Altmann, K.-H.; Schinzer,

D. Chemistry and Biology of Epothilones. In Natural Products in

Medicinal Chemistry, 1st ed.; Hanessian, S., Ed.; Methods and

Principles in Medicinal Chemistry 60; Wiley-VCH: Weinheim,

Germany, 2014; pp 83−125. (h) Schiess, R.; Altmann, K.-H.

Epothilones. In Macrocycles in Drug Discovery; Levin, J., Ed.; RSC

Drug Discovery Series 40; The Royal Society of Chemistry:

Cambridge, U.K., 2015; pp 78−108.

(10) Jat, J. L.; Paudyal, M. P.; Gao, H.; Xu, Q.-L.; Yousufuddin, M.;

Devarajan, D.; Ess, D. H.; Kurti, L.; Falck, J. R. Science 2014, 343, 61.

̈

(11) Nicolaou, K. C.; Ninkovic, S.; Sarabia, F.; Vourloumis, D.; He,

Y.; Vallberg, H.; Finlay, M. R. V.; Yang, Z. J. Am. Chem. Soc. 1997, 119,

7974.

(12) Nicolaou, K. C.; Pratt, B. A.; Arseniyadis, S.; Wartmann, M.;

O’Brate, A.; Giannakakou, P. ChemMedChem 2006, 1, 41.

(13) Nicolaou, K. C.; King, N. P.; Finlay, M. R. V.; He, Y.;

Roschangar, F.; Vourloumis, D.; Vallberg, H.; Sarabia, F.; Ninkovic, S.;

Hepworth, D. Bioorg. Med. Chem. 1999, 7, 665.

(14) Zenzola, M.; Doran, R.; Degennaro, L.; Luisi, R.; Bull, J. A.

Angew. Chem., Int. Ed. 2016, 55, 7203.

(15) (a) Sefkow, M.; Kiffe, M.; Schummer, D.; Hofle, G. Bioorg. Med.

̈

Chem. Lett. 1998, 8, 3025. (b) Hofle, G.; Glaser, N.; Leibold, T.;

Sefkow, M. Pure Appl. Chem. 1999, 71, 2019.

̈

(16) Johnson, J.; Kim, S.-H.; Bifano, M.; DiMarco, J.; Fairchild, C.;

Gougoutas, J.; Lee, F.; Long, B.; Tokarski, J.; Vite, G. Org. Lett. 2000,

2, 1537.

(17) (a) Horner, L.; Hoffmann, H. M. R.; Wippel, H. G. Chem. Ber.

1958, 91, 61. (b) Horner, L.; Hoffmann, H. M. R.; Wippel, H. G.;

Klahre, G. Chem. Ber. 1959, 92, 2499. (c) Wadsworth, W. S., Jr.;

Emmons, W. D. J. Am. Chem. Soc. 1961, 83, 1733. (d) Wadsworth, D.

H.; Schupp, O. E., III; Seus, E. J.; Ford, J. A., Jr. J. Org. Chem. 1965, 30,

680.

(2) Vahdat, L. Oncologist 2008, 13, 214.

(3) Borzilleri, R. M.; Zheng, X.; Schmidt, R. J.; Johnson, J. A.; Kim,

S.-H.; DiMarco, J. D.; Fairchild, C. R.; Gougoutaz, J. Z.; Lee, F. Y. F.;

Long, B. H.; Vite, G. D. J. Am. Chem. Soc. 2000, 122, 8890.

(4) (a) Vite, G. D.; Kim, S.-H.; Hoﬂe, G. World Patent WO 9954319

̈

(18) For selected reviews, see: (a) Maryanoff, B. E.; Reitz, A. B.

A1, 1999. (b) Regueiro-Ren, A.; Borzilleri, R. M.; Zheng, X.; Kim, S.-

H.; Johnson, J. A.; Fairchild, C. R.; Lee, F. Y. F.; Long, B. H.; Vite, G.

D. Org. Lett. 2001, 3, 2693. (c) Regueiro-Ren, A.; Borzilleri, R. M.;

Vite, G. D.; Kim, S.-H. World Patent WO 02098868 A1, 2002.

Chem. Rev. 1989, 89, 863. (b) Nicolaou, K. C.; Harter, M. W.;

̈

Gunzner, J. L.; Nadin, A. Liebigs Ann./Recl. 1997, 1997, 1283. (c) Gu,

Y.; Tian, S.-K. Olefination Reactions of Phosphorous-Stabilized

P

DOI: 10.1021/jacs.7b02655

J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society

Article

Carbon Nucleophiles. In Stereoselective Alkene Synthesis; Wang, J., Ed.;

Topics in Current Chemistry 327; Springer: Berlin, Germany, 2012;

pp 197−238. (d) Blakemore, P. R. Olefination of Carbonyl

Compounds by Main-Group Element Mediators. In Comprehensive

Organic Synthesis, 2nd ed.; Knochel, P., Molander, G. A., Eds.; Elsevier:

Oxford, U.K., 2014; pp 729−817. (e) Bisceglia, J. A.; Orelli, L. R. Curr.

Org. Chem. 2015, 19, 744.

(19) Attempted HWE reaction with a 4-methoxybenzyl (PMB)-

protected aziridine moiety also proved unsuccessful (see the

Supporting Information for details).

(20) (a) Nicolaou, K. C.; Finlay, M. R. V.; Ninkovic, S.; King, N. P.;

He, Y.; Li, T.; Sarabia, F.; Vourloumis, D. Chem. Biol. 1998, 5, 365.

́

(b) Nicolaou, K. C.; Namato, K.; Ritzen, A.; Ulven, T.; Shoji, M.; Li, J.;

D’Amico, G.; Liotta, D.; French, C. T.; Wartmann, M.; Altmann, K.-

H.; Giannakakou, P. J. Am. Chem. Soc. 2001, 123, 9313. (c) Nicolaou,

(28) For examples of β-hydroxyphosphonate HWE adducts from β-

heteroaromatic phosphonate precursors, see: (a) Tsuge, O.;

Kanemasa, S.; Suga, H. Bull. Chem. Soc. Jpn. 1988, 61, 2133.

(b) Harusawa, S.; Koyabu, S.; Inoue, Y.; Sakamoto, Y.; Araki, L.;

Kurihara, T. Synthesis 2002, 2002, 1072.

(29) It is noteworthy to mention that the corresponding 2-

ﬂuoroethoxyphosphonate of compound 52 (see the Supporting

Information for its preparation) did not perform adequately in the

HWE reaction, and only starting materials were recovered. This

suggests that the success of 2-ﬂuoroethoxyphosphonates in the HWE

reaction between ketones and β-heteroaromatic phosphonates is

highly dependent on the electronic nature of the heterocyclic moiety.

(30) (a) Deschamps, B.; Lefebvre, G.; Seyden-Penne, J. Tetrahedron

1972, 28, 4209. (b) Bottin-Strzalko, T.; Seyden-Penne, J. Bull. Soc.

Chim. Fr. 1984, 161.

(31) (a) Denmark, S. E.; Dorow, R. L. J. Am. Chem. Soc. 1990, 112,

864. (b) Zarges, W.; Marsch, M.; Harms, K.; Haller, F.; Frenking, G.;

Boche, G. Chem. Ber. 1991, 124, 861. (c) Narasaka, K.; Hidai, E.;

Hayashi, Y.; Gras, J.-L. J. Chem. Soc., Chem. Commun. 1993, 102.

(d) Brandt, P.; Norrby, P.-O.; Martin, I.; Rein, T. J. Org. Chem. 1998,

63, 1280. (e) Ando, K. J. Org. Chem. 1999, 64, 6815.

́

K. C.; Namato, K.; Li, J.; Ritzen, A.; Ulven, T.; Shoji, M.; Zaharevitz,

D.; Gussio, R.; Sackett, D. L.; Ward, R. D.; Hensler, A.; Fojo, T.;

Giannakakou, P. ChemBioChem 2001, 2, 69. (d) Nicolaou, K. C.;

́

Ritzen, A.; Namato, K.; Buey, R. M.; Díaz, J. F.; Andreu, J. M.;

Wartmann, M.; Altmann, K.-H.; O’Brate, A.; Giannakakou, P.

Tetrahedron 2002, 58, 6413.

́

(32) Kurti, L.; Czako, B. Horner−Wadsworth−Emmons Oleﬁnation.

̈

(21) (a) Cachoux, F.; Isarno, T.; Wartmann, M.; Altmann, K.-H.

Synlett 2006, 2006, 1384. (b) Kuzniewski, C. N.; Gertsch, J.;

Wartmann, M.; Altmann, K.-H. Org. Lett. 2008, 10, 1183. (c) Altmann,

K.-H.; Cachoux, F.; Feyen, F.; Gertsch, J.; Kuzniewski, C. N.;

Wartmann, M. Chimia 2010, 64, 8. (d) Schiess, R.; Gertsch, J.;

Schweizer, W. B.; Altmann, K.-H. Org. Lett. 2011, 13, 1436.

(e) Gaugaz, F. Z.; Redondo-Horcajo, M.; Barasoain, I.; Díaz, J. F.;

Cobos-Correa, A.; Kaufmann, M.; Altmann, K.-H. ChemMedChem

2014, 9, 2227.

Strategic Applications of Named Reactions in Organic Synthesis; Elsevier:

Oxford, U.K., 2005; pp 212−213.

(33) Schlosser, M. Pure Appl. Chem. 1988, 60, 1627.

(34) Still, W. C.; Gennari, C. Tetrahedron Lett. 1983, 24, 4405.

́

(22) Nicolaou, K. C.; Ritzen, A.; Namoto, K.; Buey, R. M.; Díaz, J. F.;

Andreu, J. M.; Wartmann, M.; Altmann, K.-H.; O’Brate, A. Tetrahedron

2002, 58, 6413.

(23) (a) Nicolaou, K. C.; Scarpelli, R.; Bollbuck, B.; Werschkun, B.;

Pereira, M. M. A.; Wartmann, M.; Altmann, K.-H.; Zaharevitz, D.;

Gussio, R.; Giannakakou, P. Chem. Biol. 2000, 7, 593. (b) Nicolaou, K.

C.; Sasmal, P. K.; Rassias, G.; Reddy, M. V.; Altmann, K.-H.;

Wartmann, M.; O’Brate, A.; Giannakakou, P. Angew. Chem., Int. Ed.

2003, 42, 3515.

(24) Nicolaou, K. C.; Sarabia, F.; Finlay, M. R. V.; Ninkovic, S.; King,

N. P.; Vourloumis, D.; He, Y. Chem. - Eur. J. 1997, 3, 1971.

(25) For X-ray crystallographic and NMR spectroscopic structural

studies of epothilones with β-tubulin, see: (a) Carlomagno, T.;

Blommers, M. J. J.; Meiler, J.; Jahnke, W.; Schupp, T.; Petersen, F.;

Schinzer, D.; Altmann, K.-H.; Griesinger, C. Angew. Chem., Int. Ed.

2003, 42, 2511. (b) Nettles, J. H.; Li, H.; Cornett, B.; Krahn, J. M.;

Snyder, J. P.; Downing, K. H. Science 2004, 305, 866. (c) Reese, M.;

́

Sanchez-Pedregal, V. M.; Kubicek, K.; Meiler, J.; Blommers, M. J. J.;

Griesinger, C.; Carlomagno, T. Angew. Chem., Int. Ed. 2007, 46, 1864.

(d) Prota, A. E.; Bargsten, K.; Zurwerra, D.; Field, J. J.; Díaz, J. F.;

Altmann, K.-H.; Steinmetz, M. O. Science 2013, 339, 587.

(26) (a) Rautio, J.; Kumpulainen, H.; Heimbach, T.; Oliyai, R.; Oh,

D.; Jarvinen, T.; Savolainen, J. Nat. Rev. Drug Discovery 2008, 7, 255.

̈

(b) Huttunen, K.; Raunio, H.; Rautio, J. Pharmacol. Rev. 2011, 63, 750.

(27) For examples of β-hydroxyphosphonate HWE adducts from β-

aromatic phosphonate precursors, see: (a) Petrova, J.; Vassilev, N. G.;

Kirilov, M. Phosphorus, Sulfur Silicon Relat. Elem. 1990, 47, 457.

(b) Amer, A.; Ho, D.; Rumpel, K.; Schenkel, R. I.; Zimmer, H. J. Org.

Chem. 1991, 56, 5210. (c) Angelova, O.; Macicek, J.; Vassilev, N. G.;

Momchilova, S.; Petrova, J. J. Crystallogr. Spectrosc. Res. 1992, 22, 253.

(d) Petrova, J.; Momchilova, S.; Vassilev, N. G. Phosphorus, Sulfur

Silicon Relat. Elem. 1992, 68, 45. (e) Vassilev, N. G. Phosphorus, Sulfur

Silicon Relat. Elem. 1993, 85, 49. (f) Vassilev, N. G.; Dimitrov, V. S.

Magn. Reson. Chem. 1994, 32, 639. (g) Mizuno, M.; Fujii, K.; Tomioka,

K. Angew. Chem., Int. Ed. 1998, 37, 515. (h) Modro, A. M.; Modro, T.

A.; Mphahlele, J.; Perlikowska, W.; Pienaar, A.; Sales, M.; van Rooyen,

P. H. Can. J. Chem. 1998, 76, 1344. (i) Takaki, K.; Itono, Y.; Nagafuji,

A.; Naito, Y.; Shishido, T.; Takehira, K.; Makioka, Y.; Taniguchi, Y.;

Fujiwara, Y. J. Org. Chem. 2000, 65, 475.

Q

DOI: 10.1021/jacs.7b02655

J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Article Doi

12,13-Aziridinyl Epothilones. Stereoselective Synthesis of Trisubstituted Olefinic Bonds from Methyl Ketones and Heteroaromatic Phosphonates and Design, Synthesis, and Biological Evaluation of Potent Antitumor Agents

DOI: 10.1021/jacs.7b02655

Source and publish data:

Authors:

Article abstract of DOI:10.1021/jacs.7b02655

Full text of DOI:10.1021/jacs.7b02655

Products guided by the article

R&D Labs maybe for 1001419-35-7

Relevant to this article

Hot Product