Concise Synthesis of Tunicamycin V and Discovery of a Cytostatic DPAGT1 Inhibitor

Concise Synthesis of Tunicamycin V and Discovery of a

Cytostatic DPAGT1 Inhibitor

Katsuhiko Mitachi

[a]

David Mingle

[a]

Wendy Effah

[b]

Antonio Sánchez-Ruiz

[c]

Kirk E.

Hevener

[a]

Ramesh Narayanan

[b]

William M. Clemons Jr.

[d]

Francisco Sarabia

[e]

Michio

Kurosu

[a]

Department of Pharmaceutical Sciences, College of Pharmacy, University of Tennessee Health

Science Center, 881 Madison Avenue, Memphis, Tennessee 38163, United State

[b]

Department of Medicine, University of Tennessee Health Science Center, 19 S. Manassas,

Room 120, Memphis, 38103, United State

[c]

Faculty of Pharmacy, Campus de Albacete, Universidad de Castilla-La Mancha, Avda. Dr. José

María Sánchez Ibáñez S/N 02008, Albacete, Spain

[d]

Division of Chemistry and Chemical Engineering, California Institute of Technology, 1200 E.

California Blvd., Pasadena, California 91125, United State

[e]

Department of Organic Chemistry, Faculty of Sciences, Universidad de Málaga, Campus de

Teatinos 29071 Málaga, Spain

Abstract

A short total synthesis of tunicamycin V (

), a nonselective phosphotransferase inhibitor, is

achieved via a Büchner-Curtius-Schlotterbeck type reaction. Tunicamycin V can be synthesized

in 15 chemical steps from D-galactal with 21% overall yield. The established synthetic

scheme is operationally very simple and flexible to introduce building blocks of interest. The

inhibitory activity of one of the designed analogues

against human dolichyl-phosphate

-acetylglucosaminephosphotransferase 1 (DPAGT1) shows 12.5-times greater than

. While

tunicamycins are cytotoxic molecules with a low selectivity, the novel analogue

displays

selective cytostatic activity against breast cancer cell lines including a triple-negative breast cancer.

Graphical Abstract

A Büchner–Curtius–Schlotterbeck-type reaction enabled the total synthesis of tunicamycin V

(TN-V), a nonselective phosphotransferase inhibitor, in just 15 steps from D-galactal in 21%

overall yield. The short and high-yielding synthetic route was also adapted for the synthesis of

mkurosu@uthsc.edu .

Supporting information for this article is given via a link at the end of the document.

HHS Public Access

Author manuscript

Angew Chem Int Ed Engl

. Author manuscript; available in PMC 2023 August 01.

Published in final edited form as:

Angew Chem Int Ed Engl

. 2022 August 01; 61(31): e202203225. doi:10.1002/anie.202203225.

Author Manuscript

analogues, thus leading to the discovery of a novel cytostatic tunicamycin analogue, TN-TMPA,

with high DPAGT1 selectivity.

Keywords

Tunicamycins; Total synthesis; DPAGT1 inhibitors; Antimigration effect; Cytostatic activity

Introduction

The tunicamycins inhibit dolichyl-phosphate

-acetylglucosamine-phosphotransferase 1

(DPAGT1), which is the integral membrane enzyme responsible for the committed step

-linked glycosylation.

[

]

Since their discovery in 1971,

[

]

tunicamycins have been

broadly applied in research of

-linked glycosylation and protein misfolding fields.

The tunicamycins inhibit 1)

-linked glycosylation of cancer cells at relatively low

concentrations, resulting in ER stress-mediated apoptosis, and 2) growth of a wide

range of mammalian cell lines in a nonselective fashion with narrow therapeutic index.

[

]

Due to their low selectivity against DPAGT1, the tunicamycins have remained as a

research tool.

[

]

To advance tunicamycins into clinically useful drugs, their promiscuous

cytotoxicities need to be attenuated to be selective cytotoxic or cytostatic agents. The general

structure of tunicamycins is less complicated than the other nucleoside antibiotics (

e.g

muraymycin, caprazamycin, and mureidomycin). Thus, tunicamycin is the potential scaffold

for identifying selective DPAGT1 inhibitors that can be synthesized without sophisticated

synthetic protocols.

[

]

Total synthesis of the tunicamycins has received much attention in

the context of development of the unique synthetic strategy.

[

]

To date, only four groups

have reported the total synthesis of tunicamycin V (Suami et al. 1984,

[

]

Myers et al.

1994,

[

]

Li et al. 2015,

[

]

and Yamamoto et al. 2018

[

]

). Each synthesis is unique in the

C-C bond connection at the C5’-position. However, their syntheses require 21–35 chemical

steps with 0.037–3.5% overall yields in the longest linear sequences. Considering the facts

that 1) selective deacylation of tunicamycins to modify the disaccharide core structure is an

extremely low yielding approach

[

]

and 2) an efficient purification method of tunicamycins

from a

Streptomyces lysosuperficus

culture medium has not yet been established,

[

]

we have

developed a practical synthetic route to tunicamycin V via a Büchner-Curtius-Schlotterbeck

type reaction (Figure 1).

[

]

The proposed synthetic route enables to synthesize the designed

analogues in a systematic manner with fewer number of chemical steps. A structure-based

design using a co-crystal structure with tunicamycin V is applied to discover a selective

DPAGT1 inhibitor.

[

]

Here, we report a concise total synthesis of tunicamycin V and the

discovery of a cytostatic selective DPAGT1 inhibitor of anti-cancer tunicamycin analogue.

Results and Discussion

Büchner-Curtius-Schlotterbeck (BCS) reaction has been used to synthesize

-keto esters

from the condensation between aldehydes and diazo esters.

[

]

Simple diazoalkanes (non-

stabilized diazo compounds) have also been applied to react with aldehydes and ketones

in carbon-chain elongation reactions.

[

]

However, prediction of the ratio of the reaction

products (

e.g

., carbonyls and epoxides) including 1,2-rearrangement remains difficult in

Mitachi et al.

Page 2

Angew Chem Int Ed Engl

. Author manuscript; available in PMC 2023 August 01.

Author Manuscript

application for complex molecules.

[

]

Synthetic studies towards the tunicamycins using the

diazo chemistry have been explored by Sarabia et al.

[6o-q]

Accordingly, if the diazoalkyl-

disaccharide

is stable in solution and reacts with aldehyde

, producing ketone

exclusively through a 1,2-hydride shift, an unprecedently-short chemical synthesis of the

tunicamycins will be achieved (Figure 1). In addition, theoretical studies to obtain insights

into the prediction and rationale of the BCS products have not been reported. Therefore,

we reinvestigated the stereochemical course of a model diazo coupling between

and

(Scheme 1) which may represent a useful model study for

and

in Figure 1. The

unstablized diazo intermediate

was generated in situ from the

-nitrosoacetamido

with

40% KOH (20 equiv) in a mixture of Et

O and MeOH (10/1) and into the reaction mixture

the aldehyde

was added. Careful analysis of the isolated product(s) revealed that this

model BCS reaction afforded exclusively the desired ketone

, without generation of

cis

trans

-epoxides (

). Conformational analysis of the aldehyde

was performed

to obtain insight into the observed stereochemical course (Figure 2). The result of the

theoretical calculation of

suggested that there is one preferred conformation with the

minimum energy (conformer

) as depicted in Figure 2 (see Supporting Information).

According to the Felkin-Ahn model,

[

]

the electronegative atom adopts a 90° angle relative

to the carbonyl group. The energy barriers for the interconversion between the minimum

energy conformer (

) and the two possible Felkin-Ahn conformers (

FA-1

and

FA-2

) were

calculated. We determined that the rotational barriers between

FA-1

FA-2

and

have

5.03 or 7.57 Kcal/mol. Such an insignificant difference between these two energy levels

allows free rotation around the C(1)-C(2) bond of the aldehyde

. The

FA-2

conformer has

lower steric demand in nucleophilic attacks than the conformer

FA-1

, leading to addition of

from the less hindered face of the aldehyde

as per the trajectory illustrated for

FA-2

Figure 2. The experimental results may support that the

FA-2

conformer is the predominant

specie in this model BCS reaction.

The reaction in Scheme 1 will proceed through the transition state [

A’

]

where the

-face

of the nucleophilic diazo carbon of

approaches the

-face of the carbonyl group of

to deliver the

pro

-5’

,6’

zwitterionic intermediate

A’

. BCS reactions are believed

to proceed via a two-step mechanism instead of a [2+2] cycloaddition-type reaction.

[

]

Furthermore, it is widely accepted that, in 1,2-addition reactions to carbonyls with ylids,

the zwitterionic moieties on adjacent atoms are oriented in an eclipsed conformation (

cisoid

arrangement).

[

]

According to this scenario, the kinetically favored intermediate

A’

has an

antiperiplanar arrangement to minimize van der Waals repulsions and smoothly undergoes

the 1,2-hydride shift to form the C5’-ketone

. In contrast, the epoxide formation requires

a high-energy conformer [

A”

], whose alkoxy and diazonium groups are oriented in an

antiperiplanar arrangement and where the bulkiest substituents are in a synclinal disposition.

Indeed, the (5’

, 6’

cis

-epoxide

12b

was not isolated in the reaction examined in Scheme

1. On the other hand, the formation of the

trans

-epoxides, such as

13b

, requires the

si-re

approach of the diazo specie and aldehyde through the intermediate [

B”

]. However, the

kinetically controlled (initial) transition state [

B’

]

could have a very high potential energy.

Thus, its formation is disfavored. If the R

and R

groups do not have stereochemical

disadvantage in the transition state [

B’

]

, a lower energy conformer [

B”

] with

pro

-5’

, 6’

configuration would lead the epoxide. This pathway represents the general mechanism of

Mitachi et al.

Page 3

Angew Chem Int Ed Engl

. Author manuscript; available in PMC 2023 August 01.

Author Manuscript

epoxide formations in BCS reactions with simple substrates.

[

]

In Scheme 1, the epoxides

were not isolated, indicating that BCS reaction between

and

provides the kinetic

intermediate

A’

exclusively.

To achieve an efficient convergent synthesis, it is essential to establish BCS reaction

with diazoalkyl-disaccharide

(

→

in Scheme 2); complex diazonium species of

disaccharides have not previously been applied in carbonyl addition reactions.

[

]

Moreover,

the following chemical transformations are required to investigate: 1) selective 11’-

-”-

-trehalose-type linkage with a convenient thioglycoside (

e.g

), 2) selective mono

-nitrosylation of the diacetamide (

→

), 3) stereoselective reduction of the C5’-

carbonyl (

→

), and 4) acid-catalyzed deprotections without cleavage of the 11’-

-1”-

-glycoside linkage. We have an extensive knowledge of mild acid-cleavable protecting

groups of the uridine ureido nitrogen.

[

]

The proposed protecting group strategy

illustrated in Figure 1, 4,4’-(bisfluorophenyl)methoxymethyl (BFPM) was selected as this

group displays an excellent relative stability in chemical transformations for advancing

nucleosides, but can be cleaved with 80% AcOH.

[

]

An optimized synthetic route for the

tunicamycin core structure

is illustrated in Scheme 2. The dinitrobenzenesulfonamide

derivative

was synthesized from D-galactal (

) in three steps including tosylation,

acetonide formation and 2,4-dinitrobenzenesulfonamide formation in 70% overall yield.

Azidonitration of the galactal derivative

was performed with excess ceric ammonium

nitrate and sodium azide in acetonitrile at −20

C, producing an

-mixture of 2-azide-2-

deoxy-D-galactosamine derivative

in 75% yield.

[

]

The anomeric nitrate ester and

azide groups of

was hydrolyzed with NH

•AcOH and hydrogenated with 10%

Pd-C to furnish the 2-amino-2-deoxy-D-galactosamine intermediate, whose free-amine was

protected as the corresponding phthalimide by use of the phthaloylating agent, ethyl 1,3-

dioxoisoindoline-2-carboxylate.

[

]

Overall, the glycosyl acceptor

was synthesized in 7

steps from D-galactal in 45% yield. The glycosyl donor

was synthesized by acetylation

of azido-tolylthioglycoside

, which was converted from D-glucosamine (

) in 5 steps

according to the reported procedures (Supporting Information).

[

]

Thioglycosides have

several advantages over other types of glycosyl donors in context from stability for storage

and easy synthesis.

[

]

However, their applications to the synthesis of 11’-

-1”-

-glycoside

linkage have not been reported. The thioglycoside

(2 equiv) could be activated with NIS

in the presence of either TfOH, BF

•OEt

, or AgOTf and reacted with

. However, the

selectivity was 1/1–5/1.

[6k,6s]

A combination of NIS and AgBF

in CH

at 0

C furnished

the desired 11’-

-1”-

-glycoside

in 85% yield with over 15/1

-selectivity.

[

]

The

isolation yield of

was improved (from 85% to 92% yield) by increasing the amount of the

donor

(3 equiv). The stereochemistry of the 11’-

-1”-

-glycoside linkage was determined

by the coupling constants for H1”: 4.70 (

= 5.6 Hz) and H11’: 5.35 (

=8.9 Hz) in

H-NMR. The stereochemistries of the synthetic intermediates generated in Scheme 2 were

unequivocally determined by comparison of physical data of the synthetic TN-V (

) with

those of natural tunicamycin V. Reduction of the azide group and cleavage of the DNs group

, followed by acetylations were simultaneously conducted with AcSH in pyridine,

yielding the diacetamide

in 90%.

H-NMR analysis of

revealed that the chemical

shift of C6’-NAc (2.11 ppm) is separated from that of C2”-NAc (1.95 ppm); this difference

indicates that the C6’-acetamide nitrogen is less nucleophilic than the other acetamide

Mitachi et al.

Page 4

Angew Chem Int Ed Engl

. Author manuscript; available in PMC 2023 August 01.

Author Manuscript

nitrogen. Interestingly, nitrosylation of

with NaNO

-Ac

O in AcOH at 0

C provided

the mono-

-nitrosoacetamide

in 91% yield. The structure of

was confirmed by the

downfield shift of C6’-NAc (C6’-N(NO)Ac: 2.83 ppm) in

H-NMR and LC-MS analysis

(+NO). The observed selectivity could be attributable to a steric reason; the less hindered

C6’-NAc nitrogen was nitrosylated under a kinetically controlled condition. Gratifyingly,

diazomethyl-disaccharide

generated from

with 40% KOH in Et

O/MeOH could be

extracted with Et

O and was stable in solution for over 6 h. Removal of excess KOH was

essential to attain a high-yielding coupling reaction with the base-labile

. BCS reaction

between

and

provided the desired ketone

in 90% yield from

. As demonstrated

in Scheme 1, formation of the potential by-products, the corresponding epoxides, were not

observed. Similar to the studies in Figure 2, the stereochemical outcome of BCS reaction

between

and

is illustrated in the Supporting Information.

In order to generate the desired 5’

-alcohol

, several reducing agents were examined

(Table 1). NaBH

reduction of

in MeOH provided a 2.5/1 ratio of 5’

- and 5’

-alcohols

(

and

). Reduction under Luche’s condition (NaBH

, CeCl

•7H

[

]

improved

the isolation yield but the 5’

/5’

selectivity was not much increased. Reduction of

with Zn(BH

)

in Et

O changed the stereochemical course; the 5’

/5’

selectivity was

1/5.

[

]

In our attempt to isolate only the desired

, two chiral reducing agents (

-Me

CBS and RuCl[Ts-DPEN] (

-cymene)) were explored. (

-Me-CBS-catalyzed BH

•SMe

reduction

[

]

at −20

C furnished the desired

in 96% yield. Alternatively, (

Me-CBS-BH

•SMe

reduction provided

in 95% yield. Asymmetric Meerwein-Ponndorf-

Verley reduction of

with RuCl[(

R,R

)-Ts-DPEN] (

-cymene)

[

]

PrOH-HCO

H-Et

at rt furnished

exclusively in quantitative yield. The same reduction with RuCl[(

S,S

)-Ts-

DPEN](

-cymene) gave

without contamination of

. Due to the operational convenience

of conducting the reaction at room temperature, enough

and

towards tunicamycin

V and its 5’

-diastereomer were synthesized separately with the chiral RuCl[Ts-DPEN](

cymene).

Introduction of the fatty acid and global deprotections towards tunicamycin V (TN-V,

) are

illustrated in Scheme 3. The C10’-phthaloyl group and C3”-OAc of

were simultaneously

removed by ethylenediamine in EtOH, yielding the amino-alcohol

in 91% yield. Amide-

formation of

with (

)-13-methyltetradec-2-enoic acid (

) was performed using EDCI,

NMM, and NHS in DMF to provide

in 87% yield. The ketals and BFPM group of

were deprotected with 80% AcOH at 60

C; these conditions do not affect the 11’-

-1”-

-glycosidic bond, furnishing

in 89% yield after reverse-phase HPLC purification.

H- and

C-NMRs and [

]

data for

showed good agreement with those for natural

tunicamycin V. To understand influence of the stereochemistry of the C5’-position in

DPAGT1 enzyme and cytotoxic effects, TN-5’

epimer

was synthesized from

[6k]

We performed docking studies using the human DPAGT1 with bound tunicamycin (PDB:

6BW6) (Supporting Information, Table S1).

[

]

The tunicamycins and related congeners

have poor water-solubility that hamper the drug formulation for biological studies

in vivo

We have demonstrated that ((((trifluoromethoxy)phenoxy)piperidin-1-yl)phenyl)methoxy

methyl or its related group (

in Scheme 3) is a water-soluble hydrophobic mimetic

that binds to the putative hydrophobic tunnel created by the transmembrane segments of

Mitachi et al.

Page 5

Angew Chem Int Ed Engl

. Author manuscript; available in PMC 2023 August 01.

Author Manuscript

DPAGT1.

[

]

One of good docking scored molecules (Glide score: −15.85)

[29c]

corresponded

to TN-TMPA (

), which was synthesized from

by the same procedure for

with the

glycolyl ether

(Scheme 3).

In reference to the natural tunicamycin V, TN-V (

), TN-5’S (

), and TN-TMPA (

)

synthesized in Scheme 3 were evaluated in enzyme inhibitory assays against bacterial

phosphotransferases, MraY and WecA, and archaeal and human dolichyl-phosphate

GlcNAc-1-phosphotransferases, AglH and DPAGT1 (Table 2).

[

]

Synthetic TN-V (

)

displayed equal phosphotransferase inhibitory activity compared to a natural sample of

TN-V. The 5’

-epimer of TN-V did not show any phosphotransferase inhibitory activity

even at 50 μM concentration (Table 2). Water-solubility of TN-TMPA (

) was significantly

improved to 12.5 mg/mL (for neutral form) and 35.0 mg/mL (for HCl salt), whereas TN-V

has a poor solubility of <0.2 mg/mL (water).

[

]

Thus, TN-TMPA could conveniently be

formulated with saline. TN-TMPA did not bind to bacterial phosphotransferases, MraY

and WecA, but increased more than 10-times the DPAGT1 and AglH inhibitory activity

compared to those of TN-V. The selectivity observed in the enzyme inhibitory assays

were correlated with the antibacterial activity;

and

did not inhibit growth of

TN-V susceptible strains [

Mycobacterium tuberculosis

(H37Rv, MIC 6.3 μg/mL (for TN-

V)),

Mycobacterium smegmatis

(ATCC607, MIC 6.3 μg/mL ),

Streptococcus salivarius

(ATCC7752, MIC 25.0 μg/mL), and

Bacillus cereus

(ATCC1479, MIC 0.4 μg/mL) even

at 50.0 μg/mL (Supporting Information). TN-V displays rapid cytotoxic activity against

healthy (normal) and cancer cell lines at IC

between 0.5–2.5 μM within 24 h.

[4a]

this article, we evaluated cell viability of tunicamycin analogues against normal kidney

and breast cells (Vero and MCF10A) and breast cancer cell lines (MCF7, SKBR3, and

MDA-MB-231 (an established model of triple-negative breast cancer)). TN-V showed

cytotoxic activity against all breast cancer cell lines with the selectivity index of 1.2–1.8

(IC

MCF10A / IC

cancer cells). All cells lost their viability at 2 × IC

concentrations

(TN-V) in 24 h. TN-5’

(

) exhibited neither antibacterial (Supporting Information, Table

S3) nor cytotoxic activities at the concentrations examined in Table 3. TN-TMPA (

) did

not show cytotoxic effect against all (normal and cancer) cell lines in Table 3. However,

exhibited “cytostatic” effect only against the cancer cell lines; growth of the monolayers of

these cancer cells were suppressed at 1.25–4.50 μM concentrations.

TN-TMPA induced early apoptosis in MDA-MB-231 in 24 h determined by image-based

cell death assays (Figure 3) and the flow cytometry experiment (Supporting Information).

Image-based assays using fluorescent dye-stains (deep red anthraquinone 5 (DRAQ5):

a DNA dye, 10-nonyl acridine orange (NAO): a cardiolipin binder in mitochondria,

annexin V-FITC: a selective phosphatidylserine binding protein in plasma membrane to

identify apoptotic cells) revealed that treatment of

changed the lipid components of

the mitochondrial inner membrane and plasma membrane.

[

]

Although a majority of MDA-

MB-231 cells retained their viability at 48 h post-treatment (determined by MTT), the cells

treated with

showed significant difference in staining with NAO (a1 Vs. a4 in Figure 3A)

and were stained positively for annexin V-FITC (a2 Vs.a5 in Figure 3). The monolayer of

the untreated cells formed a multiple-layered structure in 120 h (a3). The

-treated cells

remained as a monolayer with negative to the NAO-staining (a6); the cells in a6 (Figure

Mitachi et al.

Page 6

Angew Chem Int Ed Engl

. Author manuscript; available in PMC 2023 August 01.

Author Manuscript

3A) completely lost their viability in the MTT assay. In an endpoint migration assays via

Boyden chambers,

[

]

the cell migrations of MDA-MB-231 treated with

were inhibited in

a concentration dependent manner; less than 50% of the migrated cells were counted even

at 0.2 μM (~IC

against DPAGT1) of

. At 3.5 and 10.0 μM concentrations, compared

to the control 82% and 96% of the cell migrations were supressed (Figure 3B). Thus, a

non-cytotoxic antimetastatic tunicamycin analogue was identified for the first time.

Conclusion

In conclusion, we established a short synthetic scheme for tunicamycin V and its congeners.

It could be achieved by exploring a highly efficient Büchner-Curtius-Schlotterbeck (BCS)

reaction that enables to couple between the complex disaccharide and uridine segments

in a wet solvent. In our convergent synthesis, no protecting group exchange process

(deprotection followed by reprotection) is required. Through this synthetic study, several

unprecedented selectivities were observed in scheme 2; 1) nitrosylation took place only

at the C6’-NHAc nitrogen, and 2) the 11’-

-1”-

-glycoside linkage of the disaccharide

could be constructed in high yield with >15:1 selectivity with a convenient thioglycoside.

The BFPM group introduced for the uridine ureido nitrogen could be deprotected with

under mild acidic conditions (80% AcOH) in which all protecting groups were cleaved

simultaneously at the last step. The tunicamycin V synthesis summarized here demonstrated

a 15-chemical step synthesis in the longest linear sequence with 21% overall yield from

D-galactal. The synthetic flexibility was demonstrated by synthesizing the TN-V epimer,

TN-5’

(

) and a designed analogue TN-TMPA (

) via

in-silico

method. The bioassays of

these analogues in reference to the synthetic TN-V revealed that the C5’-stereochemistry

has to be the natural configuration of

to display the phosphotransferase inhibitory

activity (Table 2). Replacing the C15-fatty acid with a water-soluble fatty acid mimetic,

((((trifluoromethoxy)phenoxy)piperidin-1-yl)phenyl)methoxy methyl (TMPA) group in TN-

V led to the loss of its bacterial phosphotransferase (MraY and WecA) inhibitory activity,

but

improved 12- and 52-fold increase in human DPAGT1 and archaeal AglH inhibitory

activity. It was demonstrated that

does not possess the cytotoxic effect observed for TN-

V, but has selective cytostatic activity against the breast cancer cells over the normal cells.

Tunicamycin V treated MDA-MB-231 cells are arrested in the cell cycle at G

, whereas,

TN-TMPA induces G

arrests (Supporting Information, Figure S13). These differences

in cell cycle arrests by tunicamycins and our DPAGT1 inhibitors have been observed in

pancreatic cancer cell lines.

[4a]

The DPAGT1 inhibitor

serves as a strong antimetastatic

agent that induces apoptosis in a triple-negative breast cancer (MDA-MB-231) cell line.

TN-TMPA is a cell-permeable molecule, which was demonstrated in Figure 3B; the

antimigration activity was observed at the IC

(DPAGT1 enzyme) concentration level.

These data could imply that TN-V’s rapid cytotoxic activity is not due to its DPAGT1

inhibitory activity. We are currently performing a structure-based design and synthesis

to improve the lead

to be a low nM DPAGT1 inhibitor. A short chemical synthesis

of TN-V and its analogues summarized in scheme 2 and 3 enables us to perform SAR

studies. We have reported that DPAGT1 crosstalks to Snail (a transcription factor that

promotes the repression of the adhesion molecule E-cadherin) in intracellular signaling of

pancreatic cancers.

[4a]

Triple-negative breast cancers are among the most aggressive and

Mitachi et al.

Page 7

Angew Chem Int Ed Engl

. Author manuscript; available in PMC 2023 August 01.

Author Manuscript

potentially metastatic malignancies. MDA-MB-231 cells are known to reduce the E-cadherin

expression level and increase mesenchymal markers, providing the cell’s mobility and

invasive characteristics.

[

]

Thus, we are very interested in understanding the molecular

mechanisms of antimetastatic effect of DPAGT1 inhibitors in MDA-MB-231 cells. We will

report SAR and comparative glycoproteomic data to understand antimetastatic activity of

a cytostatic DPAGT1 inhibitor of tunicamycin analogue in triple-negative breast cancers

elsewhere.

Supplementary Material

Refer to Web version on PubMed Central for supplementary material.

Acknowledgements

Research reported in this publication was partially supported by the National Institute of General Medical Sciences

of the National Institutes of Health under award number R01GM114611. M.K. thanks UTRF (University of

Tennessee Health Science Center) for generous financial support (Innovation award R079700292). F. S. thanks

Ministerio de Ciencia, Innovación y Universidades (Spain) for financial support (RTI2018-098296-B-I00). R.N.

thanks NCI for financial support (R01 CA229164). NMR data were obtained on instruments supported by the

NIH Shared Instrumentation Grant. M.K. would like to thank Dr. Michael McNeil (Colorado State University) for

providing

E. coli

B21 WecA strain. The authors gratefully acknowledge Drs. Seok-Yong Lee (Duke) and Gustavo

Miranda-Carboni (University of Tennessee Health Science Center) for providing purified DPAGT1 and the breast

cancer cell lines, respectively.

References

[1]. a)Kurosu M, Future Med. Chem 2019, 11, 927–933; [PubMed: 30907628] b)Kurosu M, J. Mol.

Pharm. Org. Process Res 2018, 6, 1.

[2]. a)Takatsuki A, Arima K, Tamura G, J. Antibiot 1971, 24, 215–223;b)Takatsuki A, Tamura G, J.

Antibiot 1971, 24, 224–231;c)Takatsuki A, Tamura G, J. Antibiot 1971, 24, 232–238;d)Takatsuki

A, Tamura G, J. Antibiot 1971, 24, 785–794;e)Ito T, Takatsuki A, Kawamura K, Sato K, Tamura

G, Agric. Biol. Chem 1980, 44, 695–698;f)Kamogashira T, Takegawa S, Sugiura K, Agric. Biol.

Chem 1988, 52, 859–861.

[3]. a)Huang S, Wang D, Zhang S, Huang X, Wang D, Ijaz M, Shi Y, Cancer Chemother. Pharmacol

2017, 80, 685–696; [PubMed: 28779263] b)Banerjee A, Martinez JA, Longas MO, Zhang

Z, Santiago J, Baksi K, Banerjee DK, Adv. Exp. Med. Biol 2015, 842, 355–374; [PubMed:

25408354] c)Nami B, Donmez H, Kocak N, Exp. Toxicol. Pathol 2016, 68, 419–426; [PubMed:

27350212] d)Hou H, Sun H, Lu P, Ge C, Zhang L, Li H, Zhao F, Tian H, Zhang L, Chen T,

Yao M, Li J, Mol. Cancer Ther 2013, 12, 2874–2884; [PubMed: 24130050] e)Price NP, Hartman

TM, Li J, Velpula KK, Naumann TA, Guda MR, Yu B, Bischoff KM, J. Antibiot 2017, 70,

1070–1077;f)Morin MJ, Bernacki RJ, Cancer Res 1983, 43, 1669–1674; [PubMed: 6339042]

g)Lim EJ, Heo J, Kim YH, Apoptosis 2015, 20, 1087–1098. [PubMed: 26022098]

[4]. a)Mitachi K, Kansal RG, Hevener KE, Gillman CD, Hussain SM, Yun HG, Miranda-Carboni

GA, Glazer ES, Clemons WM Jr, Kurosu M, J. Med. Chem 2020, 63. 10855–10878; [PubMed:

32886511] b)Elbein AD, Trends Biochem. Sci 1981, 6, 219–221.

[5]. a)Mitachi K, Aleiwi BA, Schneider CM, Siricilla S, Kurosu M, J. Am. Chem. Soc 2016, 138,

12975–12980; [PubMed: 27617631] b)Mitachi K, Kurosu SM, Eslamimehr S, Lemieux MR,

Ishizaki Y, Clemons WM Jr, Kurosu M, Org. Lett 2019, 21, 876–879. [PubMed: 30698984]

[6]. a)Suami T, Sasai H, Matsuno K, Chem. Lett 1983, 12, 819–822;b)Suami T, Sasai H, Matsuno

K, Suzuki N, Fukuda Y, Sakanaka O, Tetrahedron Lett 1984, 25, 4533–4536;c)Suami T, Sasai

H, Matsuno K, Suzuki N, Carbohydr. Res 1985, 143, 85–96;d)Danishefsky SJ, Barbachyn M,

J. Am. Chem. Soc 1985, 107, 7761–7762;e)Sasai H, Matsuno K, Suami T, J. Carbohydr. Chem

1985, 4, 99–112;f)Suami T, Bioact. Mol 1986, 1, 265–276;g)Danishefsky SJ, DeNinno SL, Chen

SH, Boisvert L, Barbachyn M, J. Am. Chem. Soc 1989, 111, 5810–5818;h)Myers AG, Gin DY,

Mitachi et al.

Page 8

Angew Chem Int Ed Engl

. Author manuscript; available in PMC 2023 August 01.

Author Manuscript

Widdowson KL, J. Am. Chem. Soc 1991, 113, 9661–9663;i)Ramza J, Zamojski A, Tetrahedron

1992, 48, 6123–6134;j)Myers AG, Gin DY, Rogers DH, J. Am. Chem. Soc 1993, 115, 2036–

2038;k)Myers AG, Gin DY, Rogers DH, J. Am. Chem. Soc 1994, 116, 4697–4718;l)Karpiesiuk

W, Banaszek A, Bioorg. Med. Chem. Lett 1994, 4, 879–882;m)Karpiesiuk W, Banaszek A,

Carbohydr. Res 1994, 261, 243–253; [PubMed: 7954514] n)Zamojski A, Banaszek A, Wiadom.

Chemi 1995, 49, 683–702;o)Sarabia-García F, López-Herrera FJ, González MSP, Tetrahedron

1995, 51, 5491–5500;p)Sarabia-García F, López-Herrera FJ, Tetrahedron 1996, 52, 4757–

4768;q)Sarabia F, Martín-Ortiz L, López-Herrera FJ, Org. Biomol. Chem 2003, 1, 3716–3725;

[PubMed: 14649903] r)Li J, Yu B, Angew. Chem. Int. Ed 2015, 54, 6618–6621;s)Yamamoto

K, Yakushiji F, Matsumuru T, Ichikawa S, Org. Lett 2018, 20, 256–259; [PubMed: 29256622]

t)Karpiesiuk W, Banaszek A, Carbohydr. Res 1997, 299, 245–252. [PubMed: 9175272]

[7]. Dong YY, Wang H, Pike ACW, Cochrane SA, Hamedzadeh S, Wyszyński FJ, Bushell SR, Royer

SF, Widdick DA, Sajid A, Boshoff HI, Park Y, Lucas R, Liu WM, Lee SS, Machida T, Minall

L, Mehmood S, Belaya K, Liu WW, Chu A, Shrestha L, Mukhopadhyay SMM, Strain-Damerell

C, Chalk R, Burgess-Brown NA, Bibb MJ, Barry Iii CE, Robinson CV, Beeson D, Davis BG,

Carpenter EP, Cell 2018, 175, 1045–1058. [PubMed: 30388443]

[8]. Hayakawa M, Kajiura T, Nonomura H, J. Ferment. Bioeng 1991, 72, 327–333.

[9]. a)Buchner E, Curtius T, Chem. Ber 1885, 18, 2371–2377;b)Schlotterbeck F: Chem. Ber 1907, 40,

1826–1827.

[10]. Yoo J, Mashalidis EH, Kuk ACY, Yamamoto K, Kaeser B, Ichikawa S, Lee SY, Nat. Struct. Mol.

Biol 2018, 25, 217–224. [PubMed: 29459785]

[11]. Ford A, Miel H, Ring A, Slattery CN, Maguire AR, McKervey MA, Chem. Rev 2015, 115,

9981–10080. [PubMed: 26284754]

[12]. Candeias NR, Paterna R, Gois PMP, Chem. Rev 2016, 116, 2937–2981. [PubMed: 26854865]

[13]. Hagen S, Anthonsen T, Kilaas L, Tetrahedron 1979, 35, 2583–2589.

[14]. a)Chérest M, Felkin H, Prudent N, Tetrahedron Lett. 1968, 9, 2199–2204;b)Ahn NT, Eisenstein

O, Tetrahedron Lett. 1976, 17, 155–158.

[15]. Aggarwal VK, Charmant JPH, Fuentes D, Harvey JN, Hynd G, Ohara D, Picoul W, Robiette R,

Smith C, Vasse JL, Winn CL, J. Am. Chem. Soc 2006, 128, 2105–2114. [PubMed: 16464113]

[16]. Aggarwal VK, Harvey JN, Richardson J, J. Am. Chem. Soc 2002, 124, 5747–5756. [PubMed:

12010049]

[17]. Büchner-Curtius-Schlotterbeck (BCS) reaction of the

-methyl 2,3-O-isopropylidene-

-D-ribo-

pentonodialdo-1,4-furanoside provided a 1 : 1 mixture of the ketone and epoxide. Its selectivity

outcome is summarised in the Supporting Information.Few examples are found in the literature

about the synthesis and applications of unstabilized diazo compounds in the context of

the carbohydrates: a)Farkaš J, Šorm F, Collect. Czech. Chem. Commun 1972, 37, 2798–

2803;b)Alexander MS, Horton D, Carbohydr. Res 2007, 342, 31–43; [PubMed: 17125754]

c)Buda S, Crich D, J. Am. Chem. Soc 2016, 138, 1084–1092. [PubMed: 26731511]

[18]. Wang Y, Kurosu M, Tetrahedron 2012, 68, 4797–4804. [PubMed: 22711936]

[19]. Mitachi K, Mingle D, Kurosu M, Synthesis 2021, 53, 2643–2650. [PubMed: 35431351]

[20]. a)Lemieux RU, Ratcliffe RM, Can. J. Chem 1979, 57, 1244–1251;b)Seeberger PH, Roehrig S,

Schell P, Wang Y, Christ WJ, Carbohydr. Res 2000, 328, 61–69. [PubMed: 11005576]

[21]. McArthur CR, Worster PM, Okon AU, Synth. Commun 1983, 13, 311–318.

[22]. a)Yan RB, Yang F, Wu Y, Zhang LH, Ye XS, Tetrahedron Lett. 2005, 46, 8993–8995;b)Singh

L, Lam A, Premraj R, Seifert J, Tetrahedron Lett. 2008, 49, 4854–4856;c)Glibstrup E, Pedersen

CM. Org Lett. 2016. 18, 4424–4427. [PubMed: 27552399]

[23]. a)Fügedi P, Garegg PJ, Lönn H, Norberg T, Glycoconj. J 1987, 4, 97–108;b)Garegg PJ, Adv.

Carbohydr. Chem. Biochem 1997, 52, 179–205; [PubMed: 9218334] c)Codée JDC, Litjens

REJN, van den Bos LJ, Overkleeft HS, van der Marel GA, Chem. Soc. Rev 2005, 34, 769–782.

[PubMed: 16100617]

[24]. Kurosu M, Li K, Crick DC, Org. Lett 2009, 11, 2393–2396. [PubMed: 19405507]

[25]. Gemal AL, Luche JL, J. Am. Chem. Soc 1981, 103, 5454–5459.

[26]. Kurosu M, Marcin LR, Grinsteiner TJ, Kishi Y, J. Am. Chem. Soc 1998, 120, 6627–6628.

Mitachi et al.

Page 9

Angew Chem Int Ed Engl

. Author manuscript; available in PMC 2023 August 01.

Author Manuscript

[27]. a)Corey EJ, Bakshi RK, Shibata S, J. Am. Chem. Soc 1987, 109, 5551–5553;b)Kurosu M, Kishi

Y, J. Org. Chem 1998, 63, 6100–6101. [PubMed: 11672234]

[28]. a)Hashiguchi S, Fujii A, Takehara J, Ikariya T, Noyori R, J. Am. Chem. Soc 1995, 117, 7562–

7563;b)Fujii A, Hashiguchi S, Uematsu N, Ikariya T, Noyori R, J. Am. Chem. Soc 1996, 118,

2521–2522.

[29]. a)Schrödinger Release 2021–4: Glide, Schrödinger, LLC, New York, NY, 2021;b)Friesner RA,

Murphy RB, Repasky MP, Frye LL, Greenwood JR, Halgren TA, Sanschagrin PC, Mainz DT,

J. Med. Chem 2006, 49, 6177–6196; [PubMed: 17034125] c)Halgren TA, Murphy RB, Friesner

RA, Beard HS, Frye LL, Pollard WT, Banks JL, J. Med. Chem 2004, 47, 1750–1759; [PubMed:

15027866] d)Friesner RA, Banks JL, Murphy RB, Halgren TA, Klicic JJ, Mainz DT, Repasky

MP, Knoll EH, Shelley M, Perry JK, Shaw DE, Francis P, Shenkin PS, J. Med. Chem 2004, 47,

1739–1749. [PubMed: 15027865]

[30]. a)Siricilla S, Mitachi K, Skorupinska-Tudek K, Swiezewska E, Kurosu M, Anal. Biochem 2014,

461, 36–45; [PubMed: 24939461] b)Mitachi K, Siricilla S, Yang D, Kong Y, Skorupinska-Tudek

K, Swiezewska E, Franzblau SG, Kurosu M, Anal. Biochem 2016, 512, 78–90; [PubMed:

27530653] c)Mitachi K, Yun HG, Kurosu SM, Eslamimehr S, Lemieux MR, Klaicć L, Clemons

WM Jr, Kurosu M, ACS Omega 2018, 3, 1726–1739. [PubMed: 29503973]

[31]. Lipinski C, Methods Principles Med. Chem 2008, 37, 257–282.

[32]. Collins TJ, Ylanko J, Geng F, Andrews DW, Assay Drug Dev. Technol 2015, 13, 547–557.

[PubMed: 26422066]

[33]. a)Gomez-Lopez N, Vadillo-Ortega F, Estrada-Gutierrez G, PLoS One 2011, 6, e28771; [PubMed:

22174892] b)Doerr ME, Jones JI, J. Biol. Chem 1996, 271, 2443–2447. [PubMed: 8576205]

[34]. a)Saatci O, Kaymak A, Raza U, Ersan PG, Akbulut O, Banister CE, Sikirzhytski V, Tokat UM,

Aykut G, Ansari SA, Dogan HT, Dogan M, Jandaghi P, Isik A, Gundogdu F, Kosemehmetoglu K,

Dizdar O, Aksoy S, Akyol A, Uner A, Buckhaults PJ, Riazalhosseini Y, Sahin O, Nat. Commun

2020, 11, 2416; [PubMed: 32415208] b)Noyan S, Andac Ozketen A, Gurdal H, Gur Dedeoglu B,

Cell. Signal 2021, 83, 109996. [PubMed: 33798630]

Mitachi et al.

Page 10

Angew Chem Int Ed Engl

. Author manuscript; available in PMC 2023 August 01.

Author Manuscript

Figure 1.

A proposed high-yielding chemical synthesis of tunicamycin V from D-galactal -

Retrosynthetic analysis.

BFPM: (4,4’-bisfluorophenyl)methoxymethyl, DNs: 2,4-dinitrobenzenesulfonyl

Mitachi et al.

Page 11

Angew Chem Int Ed Engl

. Author manuscript; available in PMC 2023 August 01.

Author Manuscript

Figure 2.

Conformational analysis of the aldehyde

and rationale of BCS reaction with

Mitachi et al.

Page 12

Angew Chem Int Ed Engl

. Author manuscript; available in PMC 2023 August 01.

Author Manuscript

Figure 3.

Growth and migration inhibition of a triple-negative breast cancer MDA-MB-231 by TN-

TMPA (

[a] All assay protocols are summarized in SI. Purple (a1–6): DNA stained with DRAQ5

(CY5 filter: Ex: 635/18, Em: 692/40), Green (a1, a3, a4, and a6): mitochondria stained with

NAO (GFP filter: NAO Ex: 482/25, Em: 524/24), Green (a2 and a5): plasma membrane and

mitochondria stained with annexin V-FITC (GFP filter: NAO Ex: 482/25, Em: 524/24). [b]

Cell migration assays using the Boyden Chamber. Data represent the mean ±SD (

< 0.01).

Mitachi et al.

Page 13

Angew Chem Int Ed Engl

. Author manuscript; available in PMC 2023 August 01.

Author Manuscript

Scheme 1.

A model Büchner-Curtius-Schlotterbeck (BCS) reaction towards tunicamycins.

PMB:

-methoxybenzyl, Phth: phthaloyl

Mitachi et al.

Page 14

Angew Chem Int Ed Engl

. Author manuscript; available in PMC 2023 August 01.

Author Manuscript

Scheme 2.

Synthesis of the core structure

for tunicamycins.

Mitachi et al.

Page 15

Angew Chem Int Ed Engl

. Author manuscript; available in PMC 2023 August 01.

Author Manuscript

Scheme 3.

Synthesis of tunicamycin V (TN-V,

), its 5’

-diastereomer (TN-5’

), a water-soluble

hydrophobic mimetic analogue (TN-TMPA,

). NMM:

-Methylmorpholine, EDCI: 1-

Ethyl-3-(3-dimethylaminopropyl)carbodiimide, N

HS:

-hydroxysuccinimide

Mitachi et al.

Page 16

Angew Chem Int Ed Engl

. Author manuscript; available in PMC 2023 August 01.

Author Manuscript

Mitachi et al.

Page 17

Table 1.

Stereoselective reduction of

Reagent and conditions

21 (5’

)

[c]

/ 22 (5’

)

Yield (%)

NaBH

/ MeOH (0

2.5 / 1

NaBH

, CeCl

•7H

O / MeOH (0

3 / 1

Zn(BH

)

/ Et

O (−20

1/5

(

-Me CBS,

[a]

•SMe

/ THF (−20

1 / 0

(

-Me CBS, BH

•SMe

/ THF (−20

0 / 1

RuCl[(

R,R

)-Ts-DPEN] (

-cymene)

[b]

PrOH-HCO

H-Et

N (rt)

1 / 0

>99

RuCl[(

S,S

)-Ts-DPEN] (

-cymene)/

PrOH-HCO

H-Et

N (rt)

0/1

>99

[a]

Oxazaborolidine-catalyzed reduction (widely known as the CBS reduction).

[b]

RuCl[(

R,R

)-Ts-DPEN] (

-cymene): Chloro(

-cymene)[(

R,R

-toluenesulfonyl)-1,2-diphenylethylenediamine]ruthenium(II).

[c]

The natural form. Ts:

-Toluenesulfonyl, DPEN: 1,2-Diphenylethylenediamine

Angew Chem Int Ed Engl

. Author manuscript; available in PMC 2023 August 01.

Author Manuscript

Mitachi et al.

Page 18

Table 2.

Phosphotransferase enzyme inhibitory activities of tunicamycin analogues.

[a]

Compound

(μM)

DPAGT1

[d]

AglH

[e]

WecA

[f]

MraY

[g]

TN-V

(natural)

[b]

1.5

13.2

0.15

2.9

TN-V

(

, synthetic)

[c]

1.5

13.0

0.15

2.5

TN-5’

(

)

[c]

>50

TN-TMPA

(

)

[c]

0.12

0.25

>50

[a]

All assay protocols are summarized in SI.

[b]

60–70% purity of crude natural product was purified by HPLC.

[c]

Synthesized in Scheme 3.

[d]

Human.

[e]

Methanococcus jannaschii

[f]

Escherichia coli

[g]

Hydrogenivirga

sp.

< 0.01 (

= 3).

Angew Chem Int Ed Engl

. Author manuscript; available in PMC 2023 August 01.