Tải bản đầy đủ - 0 (trang)
1 Introduction: Discovery and Biological Activity

1 Introduction: Discovery and Biological Activity

Tải bản đầy đủ - 0trang

84



3 Chemistry and Biology of Epothilones



H€

ofle from the myxobacterium Sorangium cellulosum Sc 90 [8,9]; in addition, a

number of related structures have been isolated in smaller quantities from

myxobacteria, including the C12, C13 deoxy variants epothilone C (Epo C) and

epothilone D (Epo D) [10].

R



O



S



HO



N

O

O



R = H: Epo A

R = Me: Epo B



OH O

R

S



HO



N

O

O



R = H: Epo C

R = Me: Epo D



OH O



The mode of action underlying the strong antiproliferative activity of Epo A

and B, that is, the stabilization of cellular microtubules, was unraveled by

Bollag et al. [11] only 8 years after the compounds’ original discovery. Since

then, other natural products have been recognized to be MSA (for reviews, see

Refs [12–15]; for the most recent example, see Ref. [16]), thus providing a

whole set of potential new lead structures for anticancer drug discovery.

Remarkably, all potent MSA known to date are natural products or natural

product derived (for reviews, see Refs [12–15]).

The biochemistry, cell biology, and pharmacology of epothilones have been

covered in a number of excellent review articles [17–22] and only some of the most

basic aspects of their biological profile shall be summarized here. In cell-free

systems, epothilones prevent the Ca2ỵ- or cold-induced depolymerization of

preexisting MT [23], and they promote the polymerization of tubulin heterodimers

into MT polymers under otherwise destabilizing conditions [11,23]. Epothilone

binding to MT is competitive with taxol [11,23], and the binding of Epo A to the

taxol site on b-tubulin (whose location had been established previously) has also

been directly demonstrated by electron crystallography of a tubulin–Epo A complex

as part of Zn2ỵ-stabilized tubulin polymer sheets [24].

Low nanomolar concentrations of epothilones result in aberrant mitotic spindle

formation [11,23], cell cycle arrest in mitosis, and apoptosis of cancer cells. IC50

values for the in vitro inhibition of cancer cell proliferation by Epo A and B are in

the nanomolar (Epo A) or even subnanomolar (Epo B) range. In contrast to taxol,

however, these antiproliferative effects extend to different types of multidrugresistant (MDR) cancer cell lines [11,23,25], with IC50 values often being identical

with, or at least close to, those that are observed for drug-sensitive cancer cells

[11,20,23,26]; this includes both cell lines that are taxol resistant either due to

overexpression of the Pgp170 efflux pump or due to specific tubulin mutations

[11,23]. However, acquired or inherent resistance of cancer cells to epothilones can



3.1 Introduction: Discovery and Biological Activity



arise through Pgp-independent mechanisms, such as tubulin mutations [18,27–29]

or the overexpression of alternative transporters like the MRP7 efflux pump [30].

The clinical significance of such alternative mechanisms of epothilone resistance

has not been investigated.

Epo B (as well as a number of its analogs) have been demonstrated to possess

potent in vivo antitumor activity in a number of drug-sensitive as well as multidrugresistant human tumor models in mice [25,31–33] and also in syngeneic rat models

[31]. Epo B is rapidly degraded in rodent plasma in vitro [34], but based on the

available in vivo data, tissue distribution must be significantly more rapid than

plasma metabolism. In humans, blood concentrations of Epo B after a short

infusion were found to decline in a biphasic manner with a terminal half-life of 4

days [35], which clearly indicates that plasma stability is not critical for human

therapy.

Up to this point, at least ten epothilone-type compounds have been advanced to

clinical studies in humans. These include BMS-247550 (ixabepilone, the lactam

analog of Epo B; Bristol-Myers Squibb (BMS)), natural Epo B (EPO906, patupilone;

developed by Novartis), the fully synthetic analog ZK-Epo (sagopilone, Bayer),

Epo D (deoxyEpo B, KOS-862; Kosan/Roche/BMS), BMS-310705 (C21-aminoEpo B; BMS), ABJ879 (C20-desmethyl-C20-methylsulfanyl-Epo B; Novartis), 9,10didehydro-Epo B (KOS-1584 (Kosan/Roche/BMS)), and, most recently, an analog of

unknown structure (UTD1), which is reported to have entered phase I clinical trials

in China [36]. In addition, the tumor-targeted folate conjugate BMS-753493 (BMS)

has been evaluated in phase I studies by BMS [37]. The latest addition to the stream

of epothilone-based clinical candidates appears to be an isoxazole-containing

derivative of C26-trifluoro-E-9,10-didehydro-Epo D (iso-fludelone); at the time of

writing of this chapter, patient recruitment for a phase I trial with this compound

was ongoing.1) As the sole epothilone so far, ixabepilone has obtained regulatory

approval in the United States for the treatment of advanced and metastatic breast

cancer [38]. The only other epothilone that has reached the level of phase III clinical

evaluation is the natural product Epo B (patupilone); unfortunately, the compound

did not show superiority over liposomal doxorubicin in phase III trials in ovarian

cancer and Novartis has decided not to file for registration of the compound in this

indication [39,40]. The current development status of patupilone is unknown (to

the public) and the same is true for Bayer’s sagopilone, which has undergone

extensive phase II clinical investigations. Clinical development of ABJ-879 [41],

BMS-310705, KOS-862, KOS-1584, and also BMS-753493 appears to have been

terminated or at least put on hold.2)



1) ClinicalTrials.gov Identifier: NCT01379287.

2) BMS-310705 KOS-862 or KOS-1584 are not

part of BMS’ (published) development

pipeline: http://www.bms.com/research/

pipeline/Pages/default.aspx. While KOS-862

and KOS-1584 were initially developed by

Kosan Biosciences, the company has been



acquired by BMS. No active clinical trials could

be identified for any of these compounds in the

NCI webspace. As of December 31, 2012, the

pipeline included an unspecified MT stabilizer

for neuroscience. This compound may be KOS862 (Epo D).



85



86



3 Chemistry and Biology of Epothilones



The number of epothilone-type structures that have been promoted to clinical

development reflects the surge of interest in these compounds after the discovery

of their taxol-like mode of action in 1995. In excess of 30 total syntheses of Epo A

or B have been developed since H€ofle et al. disclosed the absolute configuration

of Epo B in 1996 [42] (for reviews, see Refs [43–49]); the total synthesis work

has laid the foundation for the preparation of a host of synthetic analogs

for comprehensive structure–activity relationship (SAR) studies (reviewed in

Refs [8,17–22,43,44,46]). In the case of sagopilone, chemical synthesis has even

been the means for the production of drug substance for clinical trials [50]. The

information available from these previous accounts cannot be reiterated here

completely. Instead, this chapter will outline and selectively exemplify some of the

essential concepts that have been developed for the synthesis of natural

epothilones and their semisynthetic and synthetic analogs, including structures

that are no longer based on a canonical polyketide backbone and hybrid structures

with other natural products. Reference will be made to the biological activity of

specific analogs, or groups of analogs, thereby highlighting the most pertinent

features of the epothilone SAR. Finally, models that have been developed for the

epothilone pharmacophore and the bioactive conformation of epothilones will be

discussed briefly.



3.2

Synthesis of Natural Epothilones



Three major macrocyclization paradigms have been followed to form the

16-membered macrolactone ring in epothilones from advanced acyclic precursors

(Scheme 3.1), namely, the formation of a C12/C13 double bond through ringclosing olefin metathesis (RCM) (A), ring-closure through intramolecular ester

bond formation (B), or, most recently, RCM between C9 and C10 followed by

reduction of the resulting disubstituted double bond (C).

In all three paradigms, ring closure is generally followed by global deprotection

(to produce deoxyepothilones, such as Epo C and D) and epoxidation of the

trisubstituted C12–C13 double bond. Deviations from this general theme have also

been reported, however; for example, Mulzer’s second-generation synthesis of Epo

B involved installation of the epoxide moiety in an acyclic precursor prior to

macrolactonization and even before the construction of the C6–C7 bond [51], while

Sun and Sinha have reported the successful RCM of an epoxide-containing diene

substrate [52]. RCM-based ring closure was at the heart of two of the three early

total syntheses of Epo A by Schinzer et al. [53] and Nicolaou et al. [54], respectively

(Scheme 3.2); in contrast, Danishefsky and coworkers’ initial approach to Epo A

involved ring closure by a selective intramolecular aldol reaction with aldehyde 3

(Scheme 3.2) [55].

The RCM-based approach A and Danishefsky’s intramolecular aldol approach

have been used less frequently since this initial pioneering work, mainly in the



3.2 Synthesis of Natural Epothilones

R



O



S



HO



N

O

O



R = H: Epo A

R = Me: Epo B



OH O

Epoxidation

R

S



HO



N

R = H: Epo C

R = Me: Epo D



O

O



OH O



C



A

B



R



12



R



PGO



10

9



S



13



N



S



O

O



S



OGP



R



N

OGP



OPG O



15



O



N



O



OPG O



OPG OH

O



OH

O



Scheme 3.1



S

TBSO



N

O

O



Grubbs I, CH2Cl2

RT, 12 h, 94%

E/Z = 1:1 [55]

TBSO

or



N

O



Grubbs I, CH2Cl2

RT, 8 h, 50% [56]



OTBS O



S



O



1



2



S

TBSO



N

O



TIPSO



O



O



3

Scheme 3.2



OTBS O



KHMDS, THF

-78°, 30 min

43%



S

TBSO



N

O



TIPSO



OH O



4



87



88



3 Chemistry and Biology of Epothilones

O



LDA, THF

-78°, 70%



5

+



HO



Single isomer

O



O



O



O



O



7



O



6

OTrt



OTrt

S



O



N

+



OTBS



8



LDA, THF

-78° to -40°

71%



S

HO



N

OTBS



dr >>10

O

O



OTBS OTBS



OTBS OTBS



9



10



Scheme 3.3



case of RCM-based macrocyclization, due to unsatisfactory E/Z ratios. However,

as illustrated by the elegant work of F€

urstner et al., this problem can be

overcome by employing ring-closing alkyne metathesis (RCAM) to establish the

C12–C13 linkage followed by selective reduction of the resulting alkyne to the

required Z olefin [56]. Most recently, Schrock and coworkers have developed a

new metathesis catalyst that has enabled the fully selective synthesis of Epo C

by RCM [57].

A variety of reagents have been employed for the epoxidation of the C12–C13

double bond in deoxyepothilones, including meta-chloroperbenzoic acid (mCPBA),

dimethyldioxirane (DMDO), or trifluoromethyl-methyldioxirane (reviewed in Refs

[43,44,46]). In general, Epo D is epoxidized more selectively than Epo C; the most

selective method reported so far for the conversion of Epo D into Epo B involves the

use of DMDO at À50  C (!20 : 1 selectivity, 97% yield) [58]. In addition, Mulzer

and coworkers have developed an approach to the introduction of the epoxide

moiety in Epo B that does not rely on the epoxidation of a double bond, but rather

involves the intramolecular displacement of a C12 mesylate by a C13 hydroxyl

group [59].

A plethora of approaches have been described for the construction of different

variants of the linear precursors required for RCM or macrolactonization. A

common feature of the majority of these approaches is the stereoselective

formation of the C6–C7 bond by means of aldol chemistry. Among the various

ethyl ketones that have been employed in this step, acetonide 6 has been shown to

deliver the desired 6R,7S product with the highest selectivity (>20 : 1) (Scheme 3.3)

[53,60].

Excellent selectivities have also been reported with bis-TBS ether 9 (under very

specific experimental conditions (Scheme 3.3) [61], and more recently, for phenyl

ester 11 (although the latter has only been employed in the context of analog

synthesis) [62]. In contrast, the dianion of carboxylic acid 12, which was employed



3.2 Synthesis of Natural Epothilones



in the first-generation syntheses of epothilones by Nicolaou et al. [63,64], generally

gives low selectivity (2 : 1–3 : 1).

O

O



OH



OTBS O



O



OTBS O



11



12



An intriguing long-range effect on aldol selectivity was observed by

Danishefsky and coworkers in the reaction of ketone 14 with aldehyde 13

(Scheme 3.4) [65], which gave the desired aldol product with significantly higher

selectivity than the related saturated aldehyde (5.5 : 1 for 13 versus 1.3 : 1 for the

saturated aldehyde). It has been suggested that this effect arises from favorable

transition state interactions between the terminal double bond and the aldehyde



ÀO group in 13.

The profound impact of the structure of the aldehyde on aldol selectivity in the

establishment of the C6–C7 bond in epothilones is also apparent from the reaction

between ketone 18 and aldehyde 17, which proceeded with >95 : 5 selectivity in

favor of the desired C6-R/C7-S-isomer (Scheme 3.5) [51]. The selectivity of this

reaction, thus, is significantly higher than what is usually observed for analogous

reactions with the unsaturated aldehyde with a double bond in place of the epoxide

moiety.

Although aldol chemistry has been used most frequently to achieve stereoselective bond formation between C6 and C7, alternative approaches have also been

pursued [55,58,66].



O



LDA, THF

-120° 70%



13 +



HO

O



15 : 16 = 5.5 : 1

O



OLi OTESO



O



14



+



HO

O

O



OTESO



OTESO



15



16



Scheme 3.4



O



S



O



17 +



N

OTES

OTBS



O



Scheme 3.5



18



O

LDA, THF

-78°, 92%



S



HO



N



dr = 95:5



OTBS OTES

O



19



89



90



3 Chemistry and Biology of Epothilones



A second critical problem in the assembly of the carbon skeleton of epothilones

is the stereoselective construction of the trisubstituted C12–C13 double bond.

Again, a number of solutions to this problem have been developed, the most

important of which will be discussed in Section 3.3.2.2.



3.3

Synthesis and Biological Activity of Non-Natural Epothilones

3.3.1

Semisynthetic Derivatives



Semisynthesis has played a crucial role in the elucidation of the epothilone SAR,

with the most impressive indicator for the significance of this work being the

discovery of the semisynthetic Epo B derivative ixabepilone (also known as

BMS-247550, 21) as the only epothilone-based FDA-approved anticancer drug so

far (marketed by BMS as Ixempra1) [38]. Ixabepilone is the lactam analog of

Epo B and the BMS group has devised a highly original strategy for the

preparation of this compound from the (fermentatively produced) natural

product (Scheme 3.6). The approach exploits the allylic nature of the ester

group in the macrolactone ring, which allows the preparation of 21 in only three

steps from Epo B [34].

Thus, the Pd-catalyzed opening of the lactone ring in Epo B produces azide 20

with complete retention of configuration. Subsequent reduction of the azide moiety

to an amino group followed by intramolecular amide bond formation under

standard conditions then furnishes the desired macrolactam 21. As an alternative

to this semisynthetic approach, the total synthesis of 21 has been reported by

Danishefsky and coworkers [67] (also refer to Ref. [68]).

Compound 21 was reportedly conceived to overcome the limited metabolic

stability of Epo B that was detected in rodent plasma [34]. However, all other



O



Epo B



S



Pd(PPh3)4, NaN3 HO

45°, 60-70%



N

OH N3

O



OH



20



O

O



1. Me3P, 71%

2. EDC, HOBt, 65%



N

NH

O



Scheme 3.6



S



HO



OH O



21



3.3 Synthesis and Biological Activity of Non-Natural Epothilones

O



S



HO



TiCp2Cl2, Mg(s)

THF, 80%



N



S

HO



N



O

O



O

Epo A



OH O



Br



O



Br



S



TBSO



N

O

O



22



OTBS O



OH O



1. Bu3SnH, AIBN

C6H12, 70°, 76%

2. 20% CF3COOH/CH2Cl2

-15 °C, 90%



Epo C



1. TBSOTf, CH2Cl2

lutidine, 0°, 69%

2. Benzyltriethyl-ammonium

chloride, 50% NaOH (aq.),

CHBr3, 45 °C, 12%



S

HO



N

O

O



OH O



23



Scheme 3.7



epothilones that have been advanced to human clinical trials are lactone based; in

particular, clinical data for Epo B indicate that the ester linkage as part of the

macrolactone ring is sufficiently stable in humans to achieve promising therapeutic

effects.

The BMS group has also elaborated processes for the conversion of Epo A

and B into Epo C and D. The latter could then be used for the preparation of

cyclopropyl epothilones, although yields in the actual cyclopropanation steps

are very low (12% for the conversion of Epo C into cyclopropyl-Epo A (23))

(Scheme 3.7) [69].

Conversion of Epo A/B into the corresponding O3, O7-bis-formyl derivatives and

treatment of the latter with ammonia provides access to 3-deoxy-2,3-didehydro-Epo

A/B 24a/24b [70].

R



O



S



HO



N

O

O



O



91



R = H, Me



24a/24b



Quite remarkably, analogs 24a/24b retain most of the parent natural products’

activity; for example, 3-deoxy-2,3-didehydro-Epo B (24b) is only fourfold less potent

than Epo B against the human colon carcinoma cell line HCT-116 [70]. Saturated

3-deoxy derivatives of Epo A and B, which lack a direct conformational constraint

about the C2–C3 bond, have been investigated by Altmann et al. and shown to

retain low nanomolar IC50 values for human cancer cell growth inhibition. These

analogs are discussed in more detail in Section 3.3.2.

Numerous other semisynthetic epothilone derivatives have been prepared that

cannot be discussed here. This chemistry has been reviewed recently [71].



92



3 Chemistry and Biology of Epothilones



3.3.2

Fully Synthetic Analogs

3.3.2.1 Polyketide-Based Macrocycles

The research on synthetic epothilone analogs has been summarized in several

excellent review articles [17–22,43,44,46] and only selected aspects of this work will

be discussed in this chapter. Synthetic work on epothilone analogs was largely

driven by the desire to decipher the structural parameters that are essential (or

dispensable) for biological activity. In this context, the first structures to be

investigated were those that were available as intermediates in the total synthesis of

Epo A or B (or were accessible from such intermediates in a straightforward

manner). The most important analogs that emerged from this early SAR work were

the natural products Epo C and D, as the substrates for the final epoxidation step in

the synthesis of Epo A and B, respectively. As an example, Scheme 3.8 summarizes

Danishefsky’s second-generation synthesis of Epo D that was developed in

response to an increasing demand for material for the extensive preclinical

profiling of this analog [65,72] (for Danishefsky’s first-generation approach to Epo

B, refer to Refs [55,58,73]). Key steps of this improved synthesis of Epo D are (i) the

aldol reaction between ketone 14 and aldehyde 13, which proceeds with $5.5/1

selectivity (see above); (ii) the Suzuki–Miyaura coupling of terminal olefin 26 with

vinyl iodide 27; and (iii) the highly selective Noyori reduction of the C3-keto group

in 28. The overall optimization process also included the development of efficient

routes for the synthesis of the individual building blocks 13, 14, and 27. Obviously,

this approach implicitly also provides improved access to Epo B.

The biological activity of epothilones C and D is similar to that of the epoxidecontaining congeners [73–77] both at the level of tubulin polymerization induction

and the inhibition of cancer cell proliferation. Thus, Epo D inhibits tumor cell

growth with low nanomolar IC50 values, and, like Epo B, it retains full activity

against Pgp-overexpressing multidrug-resistant cells. (For example, the IC50 of Epo

D against the multidrug-resistant human cervical carcinoma cell line KB-8511 is

1.44 versus 0.19 nM for Epo B) [25].) As shown by the Sloan-Kettering group, the in

vitro activity of Epo D translates into potent in vivo efficacy and the compound

efficiently inhibits the growth of different types of solid tumors in mouse models of

human cancer (although a very particular dosing schedule is required to achieve

efficacy at tolerated dose levels) [78,79]. Epo D (as KOS-862) has been investigated

in a number of phase I and II clinical trials [80,81]; however, the clinical

development of the compound in oncology has been terminated. More recent

data indicate that Epo D may have potential for the treatment of Alzheimer’s

disease [82].

The potent biological activity of Epo C and D indicates that microtubule

stabilization and inhibition of cancer cell growth by epothilone-type macrolides did

not depend on the presence of an epoxide moiety, thus suggesting that other

epoxide-free epothilone variants may also display significant activity. This

presumption turned out to be true, as demonstrated by independent studies by the

group at BMS and the Nicolaou group at Scripps on cyclopropane-based analogs of



94



3 Chemistry and Biology of Epothilones



Epo A and B (see also Section 3.3.1). In these studies the replacement of the

epoxide ring in Epo A or B by a cyclopropane moiety was found to be well tolerated

and the corresponding analogs were essentially equipotent with the epoxidederived natural products [69,83]. (For example, IC50 values against the human

colon carcinoma cell line HCT-116 are 1.4 nM for 12,13-cyclopropyl-Epo A and

0.7 nM for 12,13-cyclopropyl-Epo B compared to 4.4 and 0.8 nM for Epo A and B,

respectively) [69].) It thus appears that the epoxide moiety in epothilones merely

serves to stabilize the bioactive conformation of the macrolactone ring, rather than

acting as a reactive electrophile or a hydrogen bond acceptor. More recently, Buey et

al. [84] have shown that the replacement of the epoxide oxygen by a methylene

group also leads to an (entropy-driven) increase in microtubule binding affinity (for

cyclopropyl analogs derived from both Epo A and B).

Cyclopropyl epothilones can be accessed either by semisynthesis from fermentatively produced Epo A or B or by total chemical synthesis. The former approach,

which has been pioneered by the BMS group and involves deoxygenation of Epo A/

B to Epo C/D and subsequent cyclopropanation [69], has been discussed in

Section 3.3.1. The chemical synthesis of cyclopropane-based epothilone analogs

has been spearheaded by Nicolaou et al.; this work has also included structures

incorporating additional modifications (apart from the epoxide to cyclopropane

exchange) [83,85,86]. As an example, Scheme 3.9 summarizes Nicolaou’s synthesis

OMe

N N

LDA, -120°

50 - 60%



4 steps



OH



30



31



1. LDA, 33, THF, 0°, 6 h

32, -98° to 10°, 14 h, 87%



I



76%



OH



33



OBn



32



2. a) MeI, refl, 3 h

b) 3M HCl/pentane, 3 h

91% (2 steps)



OTBS

O



OTBS



9



LDA, 9, THF/Et2O

-78°, 1 h; -40°, 0.5 h

O



OBn



HO



OBn

OTBS



34 at -78°, 5 min, 80%

O



34

S

I



N



OTBS



6 steps



O



TBSO

OTBS



63%



O



35



OTMSE



36



O



S



37



CrCl2, NiCl2

4-tBu-pyridine

37, DMSO



S

TBSO



N

OTBS OH

O



OTMSE

O



Scheme 3.9



38



S



S

4 steps



HO



N

O



10% (from 36)

O



OH O



39



S



Tài liệu bạn tìm kiếm đã sẵn sàng tải về

1 Introduction: Discovery and Biological Activity

Tải bản đầy đủ ngay(0 tr)

×