Science - USA (2022-02-25)

INSIGHTS | PERSPECTIVES

GRAPHIC: N. CARY/

SCIENCE

science.org SCIENCE

By Myles Smith

W

ith the emergence of drug-resis-

tant bacteria that can evade cur-

rent frontline therapeutics, the

need for antibiotics with new

modes of action is increasingly

urgent ( 1 ). One emerging strategy

to address this need is to revisit previously

identified antibiotics of promise and ap-

ply modern medicinal chemistry and bio-

chemical tools to improve their

activity and unveil their often

ill-defined modes of action (2,

3 ). Among such natural products

is himastatin ( 1 ), a dimeric pep-

tide antibiotic isolated in 1990

that shows good activity against

Gram-positive bacteria, albeit

through a largely unknown mode

of action ( 4 ). On page 894 of this

issue, D’Angelo et al. ( 5 ) develop

an efficient chemical synthesis

of himastatin through a bioin-

spired dimerization of its native

monomeric units. This approach

provides a blueprint to access not

only this complex target and its

derivatives, but also molecular

probes that allow for its mode of

action to be elucidated ( 5 ).

The most intuitive simplifica-

tion in the retrosynthesis of a

challenging dimeric target such

as 1 would be to break its struc-

ture into two units of its mono-

mer, depsipeptide 2 , by cleaving

the central C5–C5 9 bond. This ap-

proach reduces the synthetic task

to the preparation of monomer

2 , provided a method could be

found to unite the halves at the

correct position. Such a dimer-

ization is used by the producing

organism in the biosynthesis of 1 through

P450-catalyzed oxidative dimerization of 2.

Being able to effect such a transformation

by chemical means is a daunting challenge.

Any developed method would need to toler-

ate the numerous functional groups in the

complex peptide backbone of 2 , which might

interfere through their reactive nitrogen and

oxygen atoms, as well as forge the necessary

carbon–carbon bond selectively at one spe-

cific position on each monomer. Prior ap-

proaches to natural products related to 1 had

unsuccessfully attempted similar late-stage

dimerizations, and lengthy two-directional

strategies were used in which a simplified

dimeric core is grown outward into each

macrocyclic half. This approach was initially

showcased in Kamenecka and Danishefsky’s

inaugural synthesis of 1 ( 6 ) and in later ef-

forts for the synthesis of chloptosin ( 7 , 8 ).

From a biological standpoint, the dimeric

structure of 1 is vital for its antibacterial

properties, with monomeric compounds like

2 showing no activity ( 6 ).

Inspired by the simplicity of the biosyn-

thetic route, D’Angelo et al. began by de-

veloping a dimerization process on simpler

cyclotryptophan and cyclotryptamine com-

pounds using either a silver(I)- or copper(II)-

based oxidant to selectively arrive at C5–C5 9

dimers in useful yields without the need for

C5 prefunctionalization. Mechanistic studies

showed that such couplings likely proceed

through dimerization of an aniline radical

cation, forming the new C–C bond at the

most accessible C5 position. To prepare the

necessary monomer 2 , they combined the

convenience of solid-phase synthesis of a

pentadepsipeptide with solution-phase cou-

pling of the more tricky cyclotryptophan

unit and a final macrolactamization to form

the 18-membered ring in a highly efficient

manner (see the figure). With monomer 2 in

hand, the authors performed the dimeriza-

tion with a stoichiometric copper(II) oxidant

to deliver 1 in 40% yield.

Given the complexity of the substrate,

the obtained yield is surprisingly high, and

the efficiency of the dimerization is notable

given the absence of any protecting groups

in 2 that would typically be used to mask

reactive groups that might otherwise risk

being oxidized themselves or sequester the

metal oxidant. The convergent synthesis of 2

ORGANIC CHEMISTRY

Dimerization decrypts antibiotic activity

Direct dimerization simplifies the synthesis of himastatin and elucidates its mode of action

Department of Biochemistry, University of Texas

Southwestern Medical Center, Dallas, TX 75390, USA.

Email: [email protected]

N

H

N

HN O

N

NH

N

H

O

O

O

OOO

Me Me

Me

Me

H

N

OH

OH Me

Me

HO

H

H

N

N

NH

O

N

HN

H

N

OOO

O

O

O

Me Me

Me

Me

N

H

HO

Me

Me

OH

Me

OH

H

5

5

5'

Me

N

H

N

HN O

N

N NH

H

OOO

OOO

Me Me

Me

Me

H

N

OH

Me

Me

OH

Me

HO

H

N

H

N

HO

H

H

N

N

OH

H

N

H

NH

TESO

H

CO 2 Allyl

HO 2 C

HO 2 C

FmocHN

NHFmoc

O

N

NHFmoc

O

O

Me Me

Me

Teoc

N

OTBS

Me

Ot-Bu Me

Me

Me

O O

+

+

Abbreviations: Me, methyl; t-Bu, tert-butyl; Teoc, 2-(trimethylsilyl)ethoxycarbonyl; Fmoc, 9-fluorenylmethoxycarbonyl; TBS, tert-butyl(dimethyl)silyl;

TES, triethylsilyl; and DTBMP, 2,6-di-tert-butyl-4-methylpyridine.

Himastatin: 1

Himastatin monomer: 2

Coupling reactions

(25% overall)

(40%)

Radical-radical coupling

Dye-tag

site

Direct dimerization

of unprotected monomers

Cu(SbF 6 ) 2 DTBMP

(oxidation)

Hybrid solid- and

solution-phase synthesis

Solid

support

Cyclization

Streamlining dimerization

Chemical routes to forming large peptide dimers like himastatin ( 1 ) from its monomer ( 2 ) usually are unsuccessful.

D’Angelo et al. report a bioinspired oxidative coupling route that performs this step in useful yields.

820 25 FEBRUARY 2022 • VOL 375 ISSUE 6583