Nitroreductase-Mediated Release of Inhibitors of Lysine-Specific Demethylase 1 (LSD1) from Prodrugs in Transfected Acute Myeloid Leukaemia Cells

ChemBioChem

10.1002/cbic.202000138

FULL PAPER

Nitroreductase-mediated release of inhibitors of

Lysine-Specific Demethylase 1 (LSD1) from prodrugs

in transfected acute myeloid leukaemia cells

[

a]

[a],[b]

[c]

Eva-Maria Herrlinger, Mirjam Hau,

Desiree Melanie Redhaber, Gabriele Greve, Dominica

[

d]

[a]

[c],[e],[f]

Christoph Cornelius Miething,^[^c^]

Willmann, Simon Steimle, Michael Müller, Michael Lübbert,

Roland Schüle,^[^b^]^,^[^d^]and Manfred Jung*^[^a^]^,^[^b^]^,^[^e^]^,^[^f^]

E.M. Herrlinger and M. Hau should be considered joint first authors.

[

a]

E.M. Herrlinger, M. Hau, S. Steimle, Prof. M. Müller, Prof. M. Jung

Department of Chemistry and Pharmacy, University of Freiburg

Institute of Pharmaceutical Sciences

Albertstraße 25, 79104 Freiburg, Germany

E-mail: manfred.jung@pharmazie.uni-freiburg.de

[

b]

c]

d]

M. Hau, Prof. R. Schüle, Prof. M. Jung

CIBSS – Centre for Integrative Biological Signaling Studies, University of Freiburg

Schänzlestr. 18, 79104 Freiburg, Germany

Dr. D. M. Redhaber, Dr. G. Greve, Prof. M. Lübbert, Dr. C. C. Miething

Division of Hematology, Oncology and Stem Cell Transplantation, University of Freiburg Medical Center

Hugstetter Str. 55, 79106 Freiburg, Germany

Dr. D. Willmann, Prof. R. Schüle

Department of Urology and Center for Clinical Research, University of Freiburg Medical Center

Breisacher Str. 66, 79106 Freiburg, Germany

[

e]

f]

Prof. M. Lübbert, Prof. M. Jung

German Cancer Consortium (DKTK), Freiburg, Germany

Prof. M. Lübbert, Prof. M. Jung

[

German Cancer Research Center (DKFZ), Heidelberg, Germany

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

Abstract: Lysine-specific demethylase 1 (LSD1) has evolved as a

promising therapeutic target for cancer treatment, especially in acute

myeloid leukaemia (AML). To approach the challenge of site-specific

LSD1 inhibition, we developed an enzyme-prodrug system with the

bacterial nitroreductase NfsB (NTR) that was expressed in the virally

targeting of prodrugs is mainly realised by two possibilities, either

by delivering the prodrug to specific transporters or receptors on

the cancer cells via antibody-drug conjugates or by a site-specific

drug release (Figure 1).^[²^]Using phenotypic and genotypic

differences between cancer and healthy cells, such as hypoxia or

elevated levels of a specific enzyme, the site-specific activation of

prodrugs to the active drug can be achieved. These endogenous

mechanisms were also exploited for prodrugs of epigenetic

enzyme inhibitors, such as histone deacetylase (HDAC) inhibitor

suberanilohydroxamic acid (SAHA; Vorinostat) and the lysine-

specific histone demethylase 1 (LSD1) inhibitor tranylcypromine

(TCP, Figure 1).^[³^–⁷^]An example of using endogenously elevated

enzyme levels for targeting is the TCP-drug conjugate (Figure 1)

that is activated by LSD1 in cancer cells. In this case, the LSD1

inhibitor itself is used as a prodrug moiety to release attached

anticancer drugs.^[⁸^,⁹^]

+

transfected AML cell line THP1-NTR . The cellular activity of the NTR

was proven with a new luminescent NTR probe. We synthesised a

diverse set of nitroaromatic prodrugs that by design do not affect

LSD1 and are reduced by the NTR to release an active LSD1 inhibitor.

The emerging side products were differentially analysed using

negative controls, thereby revealing cytotoxic effects. The

2-nitroimidazolyl prodrug of a potent LSD1 inhibitor emerged as one

of the best prodrug candidates with a pronounced selectivity window

+

between wild-type and transfected NTR cells. Our prodrugs are

selectively activated and release the LSD1 inhibitor locally, proving

their suitability for future targeting approaches.

In addition to the use of endogenous mechanisms for prodrug

activation in cancer cells, several other strategies are currently

being explored.^[¹^]As schematically shown in Figure 1, the

selective activation of prodrugs can be accomplished by using a

targeted catalyst that is either an exogenous enzyme or a

chemical catalyst. Bioorthogonal uncaging strategies such as

heterogeneous palladium or gold catalysis can be used to release

Introduction

In established cancer chemotherapy, many drugs concurrently

cause side effects through their general toxicity to the whole

organism. Damage to healthy cells and tissues can be diminished

by targeted therapy that exploits differences between healthy and

cancer cells. E.g., therapeutic monoclonal antibodies are used to

target specific surface proteins on cancer cells like HER2 in breast

cancer. Another approach to improve drug-target specificity for a

selected cell type or site of disease is the development of

prodrugs.^[¹^]Prodrugs are pharmacologically inactive forms of an

inhibitor that undergo biotransformation to the active agent. The

the inhibitor at the target site (Figure 1, bioorthogonal

[10,11]

prodrug).

Selective targeting of exogenous enzymes to

tumour cells is mainly accomplished by directed enzyme prodrug

therapy (DEPT) approaches. They allow site-specific release of

active inhibitor by an exogenous enzyme that is either coupled to

an antibody (ADEPT) or encoded by a gene that is targeted to the

_t_u_m_o_u_r_s_i_t_e₍_G_D_E_P_T₎_.[1,12]

1

ChemBioChem

10.1002/cbic.202000138

FULL PAPER

Figure 1. Schematic description of site-specific drug release using an endogenous trigger or a targeted catalyst and recent examples for prodrugs of HDAC inhibitors

and LSD1 inhibitors. The activation of the prodrugs can be triggered chemically (background shaded blue) by using thiols or reactive oxygen species (ROS) that

are present in the cell or by targeted Pd/Au catalysts. Examples for the ROS-triggered activation of aryl-boronate prodrugs are the prodrug of Belinostat and Q-

PAC, which releases the active inhibitor TCP.^[⁶^,⁷^]Prodrug activation by enzymes that are targeted towards the cells or preferentially located in the target cells is also

possible. An example of using endogenously elevated enzyme levels for targeting is NI-SAHA that gets activated by endogenous nitroreductases in hypoxic cells.

Another example is the TCP-drug conjugate (lower section) that is activated by LSD1 in cancer cells. During the mechanism-based irreversible inhibition of LSD1

by TCP derivatives, the imine intermediate is hydrolysed and the nitrogen atom is cleaved off from the cyclopropyl ring, leading to LSD1 triggered release of attached

[

8,9]

anticancer drugs.

Delivery systems for the encoding genes are for example viral

vectors or synthetic vectors such as liposomes.^[¹³^]New

alternatives to ADEPT are for example N-glycan targeting

moieties that preferentially accumulate at cancer cells or

organs.^[¹⁴^,¹⁵^]By employing an exogenous enzyme, the off-target

activity of the prodrug is minimised because the prodrug is likely

not recognised by human enzymes. One prevalent enzyme for

DEPT strategies is the bacterial nitroreductase NfsB (NTR), an

oxygen-insensitive flavin mononucleotide nitroreductase.^[¹^,¹⁶^]This

enzyme has been well studied at both structural and kinetic levels,

showing a ping pong Bi-Bi mechanism with cycles of reduction by

methyl imidazole (9, Scheme 1) that was subsequently used as

prodrug moiety for the cell-specific chemical delivery of several

drugs. Nitrobenzyl and 2-nitroimidazole carbamates were

furthermore used as nitroreductase-labile groups in a modular

activation strategy for cyclopropane-tetrazine ligation.^[²⁹^]

The general concept of NTR-activated prodrugs is based on the

switch from the electron-withdrawing nitro group to the donating

hydroxylamine, which in turn triggers self-immolation and

fragmentation after enzyme reduction. This was predominantly

used for cytotoxic agents so far. We used this system to develop

a research tool to study the effects of an epigenetic inhibitor in

targeted cells which, as a non-cytotoxic agent, also allows the

study of the effect of the Michael acceptor that is formed upon

enzymatic uncaging. Epigenetic modifications, including

modifications of DNA and histone tails, are involved in the

regulation of gene expression. The appropriate regulation is

maintained by enzymes that introduce the modification (writers),

those that remove the marks (erasers), and proteins or domains

that recognise and bind the marks (readers). Mutations and

misdirected recruitment of these epigenetic proteins are often

involved in the development of cancer. Therefore, first inhibitors

NAD(P)H and re-oxidation by the nitroaromatic substrate.^[¹⁷^–²⁰^]

A

large variety of nitroaromatic substrates is converted to the

corresponding hydroxylamines, including nitrofuran antibiotics

and 2,4-dinitrobenzamides such as the prototypical example

CB1954, but also some quinones are recognised and reduced to

hydroquinones.^[¹⁷^,²⁰^]Nitrobenzyl carbamates of a variety of

cytotoxic amines are also converted by NTR to the hydroxylamine

derivatives that subsequently undergo 1,6-elimination and

release CO

2

and the free amine (Figure 2).^[²¹^–²⁴^]Hay and co-

workers extended the substrate spectrum towards 2-alkoxy-4-

nitrobenzylcarbamate and nitroheterocyclic carbamate prodrugs

of 5-aminobenz[e]indole derivatives.^[²⁵^–²⁷^]

of these epigenetic enzymes emerged as promising drug

[30,31]

candidates for cancer therapies over the last years.

Despite

Recently, a series of fluorescein-based fluorophores masked with

different nitroaromatics was synthesised in order to identify the

best general masking group for NTR substrates.^[²⁸^]Fastest

unmasking was observed for fluorescein linked to 2-nitro-N-

extensive research on specific and potent inhibitors, the recent

clinical data suggest that side effects may limit the therapeutic

window of epigenetic inhibitors.^[³⁰^,³²^]The organ- or cell-type-

specific inhibition of epigenetic regulators is a promising approach

2

ChemBioChem

10.1002/cbic.202000138

FULL PAPER

6]

to circumvent these side effects because the effect and function

of those proteins is highly context-dependent. Especially since

tumours have highly heterogeneous characteristics, the targeting

of specific cancer cells in a tumour, e.g. cancer stem cells, may

broaden the therapeutic window. This targeting can be achieved

by prodrugs that are selectively activated. So far, prodrugs for

epigenetic proteins include the DNA methyltransferase inhibitors

azacytidine and decitabine or the examples presented in Figure 1

for HDACs and LSD1, yet no exogenous enzymes were used for

prodrug activation.^[³³^]

carbamate formation was already described by Engel et al.^[They

4

coupled TCP to an aryl boronate trigger (R ) that gets activated

by high hydrogen peroxide levels in glioblastoma cells, releasing

the LSD1 inhibitor and a para-quinone methide (QM), which acts

as a glutathione scavenger (Q-PAC, Figure 1). The carbamate

linker is advantageous as it is more stable than corresponding

esters and carbonates in vivo.^[⁵²^]In addition, the carbamic acid

formed after elimination of the benzylic prodrug moiety undergoes

fast and irreversible self-immolation.^[⁵³^,⁵⁴^]This process is driven

by its positive entropy and the formation of stable products,

In this work, the exogenous enzyme NTR was used to activate

nitroaryl prodrugs of the potent inhibitor 1a (Figure 1) for the

eraser protein LSD1. Dependent on its interaction partners, LSD1

is involved in both activation and repression of gene transcription

by catalysing the demethylation of mono- and dimethylated K4 or

namely CO

2

and the free amine.^[⁵³^]

[

55]

For our prodrug system, the nanomolar LSD1 inhibitor 1a was

selected and the secondary amine was masked via a carbamate

linker with different nitroaromatic alcohols. Figure 2 sketches the

principle of prodrug activation for 4-nitrobenzyl prodrug 1b. In two

reduction steps, the NTR reduces aromatic nitro groups via the

nitroso intermediate to the hydroxylamine derivative using NADH

as co-substrate.^[¹⁹^]The change from the electron-withdrawing

substituent to an electron-donating one enables the self-

immolation of the hydroxylamine by 1,6-elimination in the case of

4-nitrobenzyl prodrugs or 1,4-elimination in the case of

5-membered hetero-aromatics. Thereby, the intermediate

[

34,35]

K9 on histone H3.

Furthermore, LSD1 is able to demethylate

lysine residues at non-histone proteins, such as p53^[³⁶^], DNA

methyltransferase 1^[³⁷^]and E2F1^[³⁸^]. LSD1 was identified as an

important histone modifier in embryonic development,^[³⁹^,⁴⁰^]and is

required for haematopoietic cell lineage determination.^[⁴¹^,⁴²^]

Elevated LSD1 expression levels are associated with poor

prognosis in many types of solid cancers, including prostate

cancer, breast cancer, and non-small-cell lung cancer.^[⁴³^,⁴⁴^]LSD1

is overexpressed in several haematologic malignancies and thus

gained attention as a promising therapeutic target, especially in

acute myeloid leukaemia (AML).^[⁴⁵^]Recently, also a scaffolding

fragments to the active LSD1 inhibitor, CO

acceptor as side product.

2

and a Michael

role of LSD1 was elucidated in AML and small cell lung cancer

SCLC).^[

46–48]

In both malignancies, irreversible LSD1 inhibitors

(

prevent not only the catalytic activity of LSD1 but also its

interaction with the transcription factor GFI1B, thereby activating

silenced genes.^[⁴⁶^–⁴⁸^]The majority of LSD1 inhibitors developed

and used in pre-clinical and clinical studies are TCP derivatives

like 1a that all share a mechanism of irreversible inhibition by

covalently binding the FAD cofactor within the LSD1 active site.^[⁴⁹^]

Herein, an enzyme prodrugs system with the NTR and nitroaryl

prodrugs of LSD1 inhibitor 1a was developed and fully

characterised. Not only the release of 1a and the cell-specific

LSD1 inhibition was shown, but also the effects of the side

products was analysed by using appropriate negative controls.

Additionally, a luminescent probe was developed to confirm the

selective NTR activity in targeted cells. Overall, we established a

new research tool to study the effects of pharmacological LSD1

inhibition in a defined context or area.

Results

Rationale for Prodrug Design. Attachment of a prodrug moiety

to an active small molecule should diminish binding to the target

protein strongly. In the case of LSD1 inhibitors, the cyclopropyl

ring of TCP and its derivatives forms different adducts with C(4a)

or N(5) of the LSD1 cofactor FAD.^[⁵⁰^,⁵¹^]In proposed mechanisms

for this irreversible inhibition, the electron of the amine is involved

in ring-opening of the cyclopropyl ring and subsequent adduct

formation.^[⁵⁰^]Incorporation of the cyclopropylamine into

a

carbamate reduces the basicity and reactivity of the TCP amine,

and therefore the inhibitory activity of these prodrugs is expected

to be significantly reduced. Additionally, the attachment of the

nitroaromatic system to the carbamate could also prevent binding

in the active site close to FAD by steric hindrance. The masking

of the inhibitory activity of TCP-based LSD1 inhibitors by

Figure 2. Prodrug design and principle. Forming a carbamate with the amine

group of LSD1 inhibitor 1a prevents inhibition of LSD1. The NTR reduces the

nitro group of prodrug 1b via the nitroso intermediate to the hydroxylamine

derivative using NADH as co-substrate. The following self-immolative

elimination leads to the release of the LSD1 inhibitor 1a, CO and a Michael

2

acceptor as side product, here the azaquinone methide (AQM) derived from the

nitrobenzyl group.

3

ChemBioChem

10.1002/cbic.202000138

FULL PAPER

The E. coli nitroreductase NfsB has a broad substrate specificity

and thus it was expected that it also reduces sterically more

demanding prodrugs of 1a. Gruber et al. give evidence that

controls, 2c and 2f, were developed (Scheme 1). Amine 2a,

bearing an additional methyl-group at the cyclopropyl of TCP, was

shown to lack inhibitory activity on LSD1 in the low micromolar

range.^[

62]

The attachment of 2-fluoro-4-nitrobenzyl or

2-nitro-N-methyl imidazole is the best masking group for NTR-

mediated prodrug activation.^[²⁸^]However, the most reactive

derivative is not necessarily the best prodrug in cellular

experiments.^[²⁶^,⁵⁶^]Therefore, we synthesised a diverse set of

prodrugs bearing different nitroaromatics. This is also necessary

in order to evaluate the possible interactions of the prodrugs with

LSD1. This undesired interaction is conceivable as the prodrugs

are substituted with a bulky residue at the TCP amine, similar to

potent LSD1 inhibitors and the LSD1-drug conjugate in Figure 1.

2-nitroimidazolyl to the amine of 2a via a carbamate linker gave

the two negative probes 2c and 2f, respectively (Scheme 1). They

release the same reactive intermediates as the prodrugs 1c and

1f, but instead of LSD1 inhibitor 1a, they release the inactive

control 2a, allowing a differentiated analysis of LSD1 inhibition

and effects induced by the released Michael acceptor.

Synthesis of prodrugs and negative controls. Amines 1a and

2

a were synthesised according to common procedures for

62]

In cellular experiments, the reactivity of the released Michael

acceptor side product is different among the nitroaromatics. For

reductive amination with primary amines.^[As trans-TCP is more

potent than the cis-isomer, only prodrugs of trans-1a and trans-

2a were synthesised as racemates and in the case of 1d as

diastereomers, as confirmed by chiral phase HPLC analysis.^[

For the carbamate formation, either activation of the amine or of

the alcohol is necessary in order to introduce the carbonyl moiety.

For prodrugs 1b – 1d, 1f, 1g and negative controls 2c and 2f, the

amine was activated to the corresponding carbamoyl chloride

using triphosgene (Scheme 1). Prior to the addition of the

nitroaromatic alcohol, the hydroxyl group was deprotonated using

sodium hydride to facilitate nucleophilic attack on the carbamoyl

carbonyl. For nitrothiophene 8, this procedure resulted in low yield

and formation of side products. Thus, 8 was activated using

4-nitrophenyl chloroformate to form the carbonate 11

(Scheme S2) that was subsequently added to amine 1a, resulting

in prodrug 1e in good yield (47%).

4-nitrobenzyl prodrugs, the formation of an azaquinone methide

63]

(

AQM; also known as iminoquinone methide; Figure 2) is

described, however, the cellular effects are usually not further

studied.^[²¹^,²²^,⁵⁷^]It is under discussion that these intermediates

react with cellular nucleophiles such as the antioxidant

glutathione, causing cytotoxic effects at higher concentrations.^[⁵⁷^]

So far, the intrinsic cytotoxicity of AQM was not evaluated well as

the activated prodrugs mostly released cytotoxic agents in parallel

and negative controls are missing frequently. For nitro

heteroaromatics that are one of the prevalent bioreductive groups

in hypoxia-activated prodrugs, different mechanisms of

fragmentation, reaction with nucleophiles, and ring-opening are

discussed.^[

58–61]

To distinguish between effects from the released LSD1 inhibitor

and the Michael acceptor, two structurally closely related negative

2

Reagents and conditions: (i) triphosgene, pyridine, CH

nitroaromatic alcohols R -OH and the yield of step (ii) are listed:

2

Cl

2

; (ii) alcohol R -OH, pyridine, NaH, CH

2

Cl2, 24 h. In the following table, the different

2

Compound

Alcohol (R -OH)

Amine

1a

Yield

Compound

1e

Alcohol (R -OH)

Amine

1a

Yield

1b

66%

47%^[^a^]

8

1

2

c

1a

2a

43%

41%

1f

2f

1a

2a

50%

6

7

9

1d

1a

48%

1g

1a

52%

10

[

a] activation of alcohol 8 with 4-nitrophenyl chloroformate and subsequent coupling with amine 1a

Scheme 1. Synthesis of prodrugs of LSD1 inhibitor 1a and the structurally related negative control 2a. Only one enantiomeric structure is shown although it is a

racemic mixture

4

ChemBioChem

10.1002/cbic.202000138

FULL PAPER

Alcohols 6 and 8 were obtained by reduction of the corresponding

Stability tests with the prodrugs in buffer indicated no significant

release of inhibitor 1a that could distort the measured LSD1

inhibition (Figure S2). Schulz-Fincke et al. already described that

the methylated analogue 2a was inactive on LSD1 below 10

µM.^[⁶²^]As expected, also negative controls 2c and 2f were

inactive in the low micromolar range.

carboxylic acid or aldehyde. Compound

according to patent literature, using TMS-CF

7

was prepared

3

for the introduction

of the trifluoromethyl group.^[⁶⁴^]The five-step synthesis scheme for

2

-nitro imidazole 9 described by O’Connor et al. was slightly

modified as described in Scheme S1.^[⁶⁵^]Prior to the addition of

cyanamide in the third step, the pH of the solution was adjusted

to 3 using a NaOAc/HOAc buffer, instead of using 2 M aqueous

NaOH. This stabilization of the pH range avoids double reaction

of the cyanamide which would lead to the formation of guanidine

instead of the amine. This amine is further oxidised to a nitro

group using sodium nitrite and glacial acid. To avoid

decomposition of generated nitrous acid, the sodium nitrite

dissolved in water was cooled to 0 °C prior to the slow addition of

the 2-aminoimidazole in glacial acid. The hydroxymethyl-5-nitro

imidazole 10 was synthesised directly from 1-methyl-5-

In vitro evaluation of prodrug activation and fragmentation.

To establish a general method for analysis of a diverse set of

prodrugs, we developed a protocol to monitor prodrug activation

by the NTR and subsequent fragmentation in one assay. In the

first step, fluorescent measurement of NADH oxidation indicates

the amount of reduced prodrugs as the consumption of co-

substrate NADH directly correlates with the reduction of the nitro

compounds in the reaction.^[⁶⁶^]The assay was optimised in a way

to reach high substrate conversion, necessary for the following

quantitative detection of fragmentation products. The enzyme

concentration was chosen to keep the conversion ratio between

different prodrugs at a constant level and NADH was used in

excess. The prodrugs were compared after 15 minutes, when the

velocity of NADH oxidation of the prodrug samples was equal to

the one from the DMSO control, indicating no further enzymatic

reaction. After extended incubation time to allow quantitative

fragmentation of activated prodrugs, the released inhibitor 1a was

derivatised at its amine functionality using 9-fluorenyl-

methoxycarbonyl chloride (Fmoc-Cl), forming fluorescent

derivative 3 that was further quantified by HPLC (Figure 3A).

Optimised conditions for the derivatisation of amines with Fmoc-

Cl from the literature were adapted for our approach.^[⁶⁷^]

The results from the NADH consumption assay correlate with the

ones obtained in the HPLC fragmentation assay (Figure 3B and

3C). The best-activated prodrug 1f also released the highest

amount of active inhibitor 1a. For 2-nitroimidazole 1f, the

calculated activation was 51.4% and the detected fragmentation

was 59.3%, hence 8% higher than expected. The activation was

calculated after 15 minutes, but the enzymatic reaction was

stopped after two hours. This additional incubation time probably

allows for further slow activation of the prodrugs, leading to an

increased conversion of 1f. Compared with the commonly used

nitroimidazole

by

microwave-assisted

addition

of

paraformaldehyde.

In vitro evaluation of LSD1 inhibition by prodrugs.

A peroxidase-coupled assay was used to check for undesired

inhibition of LSD1 by the prodrugs 1b – 1g and the negative

controls 2a, 2c and 2f. In this assay, the hydrogen peroxide

generated by the LSD1 enzyme during the demethylation reaction

is quantified. The results in Table 1 show that all prodrugs except

1d are indeed, as desired, > 100 fold less potent compared with

the parent inhibitor 1a. The smaller window for 1d and the

inhibition of LSD1 by 1e and 1g in the micromolar range can be

caused either by direct inhibition of LSD1 enzyme by these

carbamates or through inherent instability of these two prodrugs.

The first hypothesis is supported by the fact that the change from

2

-nitro imidazole (1f) to 5-nitro imidazole (1g) leads to a significant

decrease of the IC50. The presence of the trifluoromethyl group in

d also seems to contribute to the interaction with LSD1, leading

1

to a lower IC50 value compared with the nitrobenzyl prodrug 1b.

Table 1. In vitro evaluation of LSD1 inhibition by the LSD1 inhibitor 1a, prodrugs

1b – 1g and negative controls 2a, 2c and 2f using a Peroxidase-coupled assay.

Most of the prodrugs are > 100 fold less potent compared with parent drug 1a.

Determination of higher IC50 values were not possible due to poor solubility.

n.i. = inhibition ≤ 10%.

4-nitrobenzyl 1b, prodrugs 1c, 1e, and 1g showed improved

activation and fragmentation properties. We noticed that

fluorination of the benzyl ring in meta position to the nitro group

LSD1 inhibition

in vitro IC50 [µM]

potency window

prodrug/1a

#

like in 1c improved the prodrug fragmentation rate by a factor of

68]

3

.5, an effect also described by Yang et al.^[Nitrothiophene

prodrug 1e and 5-nitroimidazole 1g were activated to the same

extent, but the fragmentation assay indicates that only 1e

releases the active inhibitor 1a quantitatively. In contrast, for 5-

nitroimidazole prodrug 1g and the nitrobenzyl prodrugs 1b, 1c,

and 1d, only 34 – 60% of the activated prodrug fraction, as

calculated from the NADH assay (Figure 3B), were converted to

the active inhibitor 1a (Figure 3C). Interestingly, the 4-nitrobenzyl

derivative 1d was activated with the lowest rate and showed no

significant fragmentation. To rank the prodrug according to their

activation by the NTR, we compared the fragmentation rates of

each prodrug with the commonly used 4-nitrobenzyl derivative

that showed the poorest NTR-mediated conversion. Overall, we

could improve the prodrug activation and fragmentation from

nitrobenzyl 1b to 2-nitroimidazole 1f by a factor of 17 (Figure 3D).

LSD1 inhibitor

1a

0.094 ± 0.017

n.i. at 10 µM

5.23 ± 0.48

1b

1c

1d

1e

> 110

58 ± 12

288 ± 81

> 110

Prodrugs

26.18 ± 5.55

n.i. at 10 µM

30.50 ± 0.92

n.i. at 10 µM

1f

1g

2a

2c

335 ± 61

> 110

Negative

controls

> 110

2f

> 110

5

ChemBioChem

10.1002/cbic.202000138

FULL PAPER

Figure 3. Scheme of in vitro analysis of prodrug activation by NTR and subsequent fragmentation. A) Prodrug activation leads to release of LSD1 inhibitor 1a that

was further quantitatively derivatised to fluorophore 3 under the conditions (i). (i): FMOC-Cl, NaHCO , H O/ACN, pH 9.0. (B) The activation of the prodrugs was

3

2

visualised by the fluorescent measurement of NADH consumption during the enzymatic reduction. As two reduction steps are necessary to generate the

hydroxylamine that can further fragment, 100% prodrug activation was equated with the consumption of two equivalents of NADH. (C) The fragmentation to 1a was

evaluated by derivatisation to 3 that was quantified by HPLC analysis. The area under the curve was used to calculate the actual concentration of 3, resulting in the

degree of fragmentation displayed in the graph. For the activated fraction of nitrobenzyl-containing prodrugs 1b, 1c and 1d (measured in B), a quantitative

fragmentation to 1a was not observed. In comparison to that, the hetero-aromatic prodrugs 1e and 1f fragmented quantitatively. (D) Representation of the order of

reactivity of the prodrugs with the NTR. The fragmentation data from C) were used to calculate the factors and the commonly used nitrobenzyl derivative 1b was

used as reference.

Synthesis and Evaluation of luminescent NTR probes to

prove NTR activity. In order to determine the expression and

correct function of the NTR in transfected cells, a luminescent

assay was developed. In literature, a variety of nitro-caged

fluorescent and luminescent probes to visualise the activity of

reductases in vitro and in vivo has been described.^[⁶⁹^,⁷⁰^]Many of

these fluorescent probes can be activated under hypoxia by

endogenous oxygen-sensitive nitroreductases, others are used to

detect NTR activity in E. coli or bacterial lysates. For this project,

a sensitive probe that is activated by NTR and that can enter cells

was required. According to Feng et al., firefly luciferin masked with

a nitrobenzyl produced higher luminescent signals than the

[70]

related amino-luciferin masked with a nitrobenzyl carbamate.

Zhou et al. additionally found that the cyanobenzothiazole

6

ChemBioChem

10.1002/cbic.202000138

FULL PAPER

precursor is more suitable for cellular experiments, probably by

increased cell-permeability compared with the negatively charged

luciferin.^[⁷¹^]Furthermore, the introduction of an N,N´-

dimethylethylenediamine linker increased the reactivity towards

the used reductase, in their case a diaphorase.^[⁷¹^]By transferring

this knowledge to the development of a luminescent NTR probe,

the trimethyl lock quinone used by Zhou et al. was replaced by

nitroaromatics as protecting group (Figure 4). To further increase

the reactivity towards NTR, a 5-nitrothiophenyl group was

attached instead of the less reactive but most commonly used 4-

nitrobenzyl prodrug moiety. So far, 5-nitrothiophene was rarely

used as a masking group for the detection of nitroreductase

activity in cells.^[⁷²^]Probe 4 was synthesised by applying standard

Mitsunobu conditions to couple (5-nitrothiophen-2-yl)methanol (8)

to 6-hydroxy-2-cyanobenzothiazole (hydroxy-CBT, Figure 4). For

probe 5, the coupling with the mono-Boc protected diamine linker

was performed as described by Mustafa et al., using bis-

factor of at least 5. At higher enzyme concentration, the difference

in conversion rate was smaller as the conversion of probe 5

stopped to increase linearly with the NTR concentration. Reasons

for the better activation of 5 may be the better accessibility of the

nitroaromatic ring to the NTR active site, but also the generation

of a thermodynamically stable 5-membered ring that forms after

nucleophilic attack of the released amine on the carbonyl group

(Figure 4).

Cellular activity. The cellular activity of our enzyme-prodrug

system was evaluated in the AML cell line THP1. The NTR-

+

expressing cell line THP1-NTR was generated by lentiviral

wt

transfection of wild-type THP1 cells (THP1 ) with an nfsb gene

construct. To confirm the cellular activity of the NTR, the

developed luminescent probes 4 and 5 were tested in both cell

wt

+

lines THP1 and THP1-NTR (Figure 5). The heterologously

expressed NTR activates the probes leading to a release of the

free luciferin precursor only in the transfected cells. After cell lysis,

this precursor can be quantified by addition of luciferin detection

reagent, as described for the in vitro assay. The conversion of the

probes in THP1-NTR⁺cells increased with increasing cell

numbers, whereas in THP1^w^tcells, no increase in luminescent

signal was observed (Figure 5). The diamine linker containing

probe 5 was much better activated than the directly masked

probe 4. One disadvantage of hydroxy-CBT is its instability in cells,

only allowing short incubation times and therefore limiting the

sensitivity. Pre-incubation of hydroxy-CBT in vitro with

recombinant NTR in buffer did not change the luminescent signal,

proving its stability under cell-free conditions. In summary, the

(

pentafluorophenyl) carbonate to activate the hydroxy-CBT

Scheme S2).^[⁷³^]In the following Boc-deprotection with

trifluoroacetic acid (TFA), thioanisol was used as a scavenger to

avoid side reaction of the tert-butyl cation on the cyano group of

hydroxy-CBT.^[⁷³^]The amine was subsequently coupled to the

activated carbonate 11 (Scheme S2) that was used in excess to

minimise intramolecular cyclization under basic conditions. The

conversion was good estimated by TLC but side product

formation necessitated purification via semipreparative HPLC.

Thus only 1% of purified probe 5 were isolated.

+

activity of NTR in the stably transfected THP1-NTR cells can be

proven using the sensitive and rapidly activated probe 5. In

wt

THP1 cells, we did not observe any conversion and thus it can

wt

be expected that no other reductases in THP1 could activate our

pro-luminescent probes. This fast protocol can be a general tool

for the screening of other cell lines in order to identify suitable cells

for the NTR/prodrug approach.

Figure 4. Scheme of luminescent probes for the selective detection of NTR in

cells. After activation by NTR, the reduced intermediate fragments

spontaneously to release the luciferin-precursor hydroxy-CBT (highlighted in

grey) that further reacts with D-cysteine to D-luciferin. This is quantified by the

addition of luciferin detection reagent and subsequent measurement of the

generated luminescent signal.

Prior to cellular tests, the stability of probe 4 and 5 in buffer and

their activation by recombinant NTR was evaluated in vitro. After

incubation of the probes with NTR and excess NADH, the luciferin

detection reagent, containing a luciferase and d-cysteine, was

added. Under basic conditions, the liberated luciferin-precursor

hydroxy-CBT reacts with d-cysteine under a condensation

reaction to D-luciferin that further undergoes enzymatic oxidation

to an excited state, resulting in a measurable luminescent signal

Figure 5. Detection of NTR activity in THP1 cells. In NTR-expressing THP1 cells

(labelled with “+”), probe 4 and 5 are converted to Luciferin whereas in non-

transfected THP1 cells (labelled with “-“), there is no activation. The probe

wt

conversion by NTR increases with increasing cell numbers (#).

_W_e_t_h_e_n_u_s_e_d_t_h_e_t_w_o_c_e_l_l_l_i_n_e_s_T_H_P₁wt and THP1-NTR+ for the

(

Figure 4). At low NTR concentrations (18 – 30 nM), the

cellular testing of the LSD1 inhibitor 1a, the prodrugs 1b – 1g and

the negative controls 2a, 2c, 2f. For the detection of cellular LSD1

conversion of probe 5 exceeded the conversion of probe 4 by a

7

ChemBioChem

10.1002/cbic.202000138

FULL PAPER

activity, we performed

phenotypic assay (Colony Forming Unit assay).

a

functional (CD86-based) and

a

(Table 2, Figure 6). Thus, most prodrugs get completely activated

into the LSD1 inhibitor 1a in THP1-NTR cells. Only 1d had the

+

same EC50 in both cell lines and thus we can conclude that it was

+

CD86-based cell assay. In addition to the catalytic activity on

not activated by NTR in THP1-NTR cells.

methylated proteins, LSD1 has significant scaffolding functions in

Negative controls 2c and 2f did not increase CD86-levels in

cells that still respond to enzyme inhibition by TCP and derivatives. THP1^w^tcells up to 50 µM. 2a showed no LSD1 inhibition in vitro

Maiques-Diaz et al. showed such a role of LSD1 in THP1 cells for

the regulation of GFI1-target genes.^[⁷⁴^]Pharmacological inhibition

of LSD1 disrupts the physical interaction with GFI1 resulting in

altered expression of GFI1-target genes with cd86 being one of

the most highly upregulated genes.^[⁷⁴^]Lynch et al. published

CD86 as a sensitive dose-dependent biomarker for LSD1

inhibition in THP1 cells in 2013.^[⁷⁵^]This change in gene

expression after LSD1 inhibition was utilised as a readout for the

cellular assay to test all compounds for this enzyme-prodrug

system. In this functional assay, CD86-positive cells were

quantified via FACS analysis after the treatment with either the

LSD1 inhibitor 1a, one of the prodrugs or negative controls.

up to 10 µM, however showed a dose-response effect in the

wt

higher micromolar range in THP1 (EC50 = 17.6 ± 0.6 µM) and

+

THP1-NTR (EC50 = 15.3 ± 2.1 µM) cells (Figure 6). 2c and 2f are

prodrugs of the negative control 2a and are expected to have the

same effect as 2a in THP1-NTR⁺cells, particularly since the

prodrugs of 1a were quantitatively activated in this assay. Indeed,

+

2f showed an effect with 33% CD86-positive THP1-NTR cells at

+

10 µM and 2c induced 29% CD86-positive THP1-NTR cells at

the same concentration (2a induced 34% at 10 µM). At

concentrations above 10 µM, the amount of dead cells in the

sample increased. This cytotoxicity, also seen in the viability

assays (Figure S5 and Table S1), impeded the measurement of

+

complete dose-response curves for 2c and 2f in THP1-NTR cells.

Nevertheless, the negative controls 2c and 2f did not affect the

CD86 expression at nanomolar concentrations that are needed

for our enzyme-prodrug system. These results prove that the

Table 2. FACS analysis of CD86-positive cells of non-transfected THP1^w^tcells

and THP1 expressing the NTR (THP1-NTR ). The selectivity window shows the

difference in EC50 between both cell lines. Figure S3 illustrates all individual EC50

curves. n.a. = not accessed.

+

CD86 expression in THP1-NTR cells treated with prodrugs 1b –

1g arise from released LSD1 inhibitor 1a and not from side

selectivity

products of the fragmentation.

wt

THP1-NTR⁺

THP1

window

_C_D₈₆+

EC50 [nM]

_C_D₈₆+

The indirect effect on neighbouring cells resulting from a released

inhibitor that diffuses from one cell type to neighbours which do

not release the inhibitor from prodrugs is called bystander effect.

Since the bystander effect is thought to be important for a

successful GDEPT therapy, the effect was studied for our

prodrugs, too. This bystander effect was studied with mixtures of

#

wt

THP1

/

EC50 [nM]

+

THP1-NTR

LSD1

inhibitor

1a

3.0 ± 0.3

4.1 ± 0.4

1 ± 1

1b

66.5% at 10 µM

783 ± 64

5.8 ± 0.4

3.8 ± 0.2

245 ± 30

4.0 ± 0.2

5.7 ± 0.4

> 1000^[^a^]

206 ± 28

1 ± 1

wt

+

the cell lines THP1 and THP1-NTR and with prodrug

concentrations that only affect the CD86 levels of THP1-NTR⁺

cells. Figure S7 shows that all cells in the mixture of the cell lines

are CD86-positive, which implies that the LSD1 inhibitor released

1c

1d

301 ± 37

Prodrugs

+

wt

in THP1-NTR cells diffuses to “by-standing” THP1 cells without

1e

143 ± 32

36 ± 10

66 ± 8

NTR.

1f

376 ± 17

Colony Forming Unit assay. LSD1 represses promoter and

enhancer activities in the haematopoietic differentiation program.

Therefore, LSD1 inhibition in THP1 cells results in differentiation

and reduced numbers of colonies in the Colony Forming Unit

1

g

617 ± 14

15.1 ± 1.1

41 ± 4

₍₁₇_.₆_±₀_.₆₎_·₁₀3

₍₁₅_.₃_±₂_.₁₎_·₁₀3

2

a

c

1 ± 1

n.a.

Negative

controls

2

> 50 µM

29% at 10 µM

33% at 10 µM

_C_F_U₎_a_s_s_a_y_.[74,76,77] The CFU assay allows the assessment of the

(

capacity of a single progenitor cell to proliferate independently into

a colony. Differentiation, but also cytotoxic effects will result in

reduced numbers of colonies in the assay. In general, irreversible

LSD1 inhibition does not reduce the viability of most cells,

including THP1 cells.^[Therefore, cytotoxic effects should not be

involved with specific, TCP-based LSD1 inhibitors. Cellular

effects of the LSD1 inhibitor 1a, prodrugs 1b – 1g and negative

controls 2a, 2c and 2f were tested in the CFU assay with the two

2

f

n.a.

_a_]_c_a_l_c_u_l_a_t_e_d_w_i_t_h_a_n_e_x_t_r_a_p_o_l_a_t_e_d_d_o_s_e_-_r_e_s_p_o_n_s_e_c_u_r_v_e_i_n_T_H_P₁wt cells.

[

78]

The LSD1 inhibitor 1a showed a dose-dependent increase of

CD86-positive THP1^w^tcells with an EC50 of 3.0 ± 0.3 nM. The

prodrugs of 1a showed undesired dose-dependent effects,

however with distinct lower potencies than the inhibitor 1a. Since

one important feature of our NTR-prodrugs is a diminished activity

wt

+

cell lines THP1 and THP1-NTR .

wt

on LSD1 in THP1 , the most promising prodrug in this assay with

As expected from the nanomolar inhibition of LSD1 in vitro (IC₅₀

of 94 ± 17 nM, Table 1), 1a also affected the colony formation with

an EC₅₀of 68 ± 12 nM (Figure 6). The assumed decreased

potency of the prodrugs on LSD1 was observed in the CFU assay

with potencies at least 50fold higher than the EC₅₀of LSD1

non-transfected THP1^w^tcells is 1b as it is >1000 fold less potent

than the inhibitor 1a. The same assay was performed with THP1-

+

NTR cells, which stably express the prodrug activating enzyme

NTR. In this cell line, the prodrugs are expected to be activated

by the NTR to release LSD1 inhibitor 1a, thereby increasing the

number of CD86-positive THP1-NTR cells. All prodrugs, with the

exception of 1d, regulated CD86 levels in the same range as the

inhibitor 1a (EC50 = 4.1 ± 0.4 nM) in the THP1-NTR⁺cell line

wt

inhibitor 1a (Table 3) in the same cell line THP1 . Among the

+

prodrugs, 1e affected the clonogenic potential of THP1^w^tcells

nearly in the same range as 1a. In vitro data about LSD1 inhibition

(Table 1) and stability data (Figure S2) do not explain this effect

8

ChemBioChem

10.1002/cbic.202000138

FULL PAPER

wt

of 1e in the CFU assay with THP1 , but a rather potent effect was

seen in the CD86-based cell assay, too. The negative controls 2c

and 2f are both inactive in the Peroxidase-coupled assay with

recombinant LSD1 and also in the CFU assay with THP1^w^tcells

at 50 µM. 2a did not inhibit LSD1 in vitro at 10 µM but had an

effect at higher concentrations in the CFU assay resulting in an

EC50 of 45.5 ± 3.2 µM (Figure 6).

wt

Table 3. Colony Forming Unit (CFU) assay with the two cell lines THP1 and

+

THP1-NTR . The selectivity window shows the difference in EC50 between both

cell lines. Figure S4 illustrates all individual EC50 curves.

selectivity

window

THP1^w^t

CFU

EC50 [nM]

THP1-NTR⁺

CFU

EC50 [nM]

#

THP1wt

/

+

THP1-NTR

The inhibitor 1a showed the same potency in both cell lines

THP1^w^t(EC50 = 0.06 ± 0.02 µM) and THP1-NTR (EC50 = 0.04 ±

+

LSD1

inhibitor

1a

1b

0.06 ± 0.02

0.04 ± 0.01

2 ± 1

+

0

.01 µM) in the CFU assay (Figure 6). Since the THP1-NTR cell

line stably expresses the prodrug activating enzyme NTR, LSD1

inhibitor 1a is released after prodrug treatment of THP1-NTR

29.0 ± 1.6

2.9 ± 0.2

2.5 ± 0.3

0.10 ± 0.01

4.4 ± 0.7

5.0 ± 0.3

45.5 ± 3.2

> 50 µM

7.6 ± 0.7

0.4 ± 0.2

4 ± 1

10 ± 5

1 ± 1

+

1

c

d

cells and reduced numbers of colonies should be observed in the

CFU assay with THP1-NTR compared with THP1^w^tcells. This

activation of prodrugs is demonstrated by the increased potencies

+

1

2.6 ± 0.4

Prodrugs

+

in THP1-NTR cells in the CFU assay compared with non-

1e

0.15 ± 0.01

0.33 ± 0.07

1.2 ± 0.3

1 ± 1

transfected THP1^w^tcells, except from 1d and 1e (Table 3). With

1f

14 ± 4

4 ± 1

1

4fold selectivity, the 2-nitroimidazolyl prodrug 1f showed the

wt

best selectivity window between the two cell lines THP1 and

THP1-NTR (Figure 6). Since the thiophene-masked prodrug 1e

_a_l_r_e_a_d_y_a_f_f_e_c_t_e_d_t_h_e_c_l_o_n_o_g_e_n_i_c_p_o_t_e_n_t_i_a_l_i_n_T_H_P₁wt cells, no

1g

2a

+

32.4 ± 1.7

14.9 ± 1.0

0.9 ± 0.2

1 ± 1

increase in potency was obtained in THP1-NTR⁺cells. The

prodrug 1d also showed no change in the EC50 between the two

cell lines, likely because it is not activated by the NTR in THP1-

Negative

controls

2c

> 3 ± 1

> 58 ± 13

2f

> 50 µM

+

NTR cells, which was already seen in in vitro tests and the CD86-

based cell assay with 1d before.

Figure 6. Result of cellular assays with the LSD1 inhibitor 1a and prodrug 1f (A, C, E) and the negative controls 2a and 2f (B, D, F) with both cell lines, non-

transfected THP1^w^t(triangles and straight lines) and stably NTR expressing THP1-NTR⁺(squares and dashed lines). A/B: The CD86-based FACS assay showed

that prodrug 1f and negative control 2f activated in THP1-NTR⁺cells reach the same effect as their parent compounds 1a and 2a. C/D: The Colony Forming Unit

(CFU) assay with the prodrug 1f showed a clear increase in potency in NTR-expressing cells caused by cell-type-specific LSD1 inhibition and cytotoxicity from the

formed side product (C). The assay with the negative control 2f showed the cytotoxic effect from the Michael acceptor as side product involved in the CFU assay.

E/F: The viability assay showed cytotoxic effects in NTR-expressing THP1-NTR⁺cells with the prodrug 1f (E) and negative control 2f (F) due to the formed

Michael acceptor released as side product after prodrug activation.

9

ChemBioChem

10.1002/cbic.202000138

FULL PAPER

Interestingly, the negative controls 2c and 2f showed higher

potencies in THP1-NTR cells than the negative control 2a, which

prodrug activation and not the compound itself causes the

observed GSH (thiol) depletion. The prodrug 1b had no cytotoxic

effect in both cell lines and also did not show GSH depletion in

both cell lines, confirming a correlation between cytotoxicity and

+

is released from 2c and 2f after activation by the NTR. Synergistic

effects of control 2a and the Michael acceptor released as side

product must be involved in the mechanisms of colony reduction

after enzymatic activation of negative controls 2c or 2f in THP1-

NTR⁺cells. As the recombinant NTR activated the 2-

nitroimidazole containing 2f better than 2-fluoro-4-nitrobenzyl-

masked 2c (Figure S1), 2f may release more of the cytotoxic

Michael acceptor, resulting in a higher potency of 2f in THP1-

NTR⁺(EC50 = 0.9 ± 0.2 µM) compared with 2c (EC50 = 14.9 ±

+

GSH-depletion in THP1-NTR cells after prodrug activation.

Discussion

Inhibition of epigenetic proteins is a promising and emerging

strategy in the treatment of cancer. The effect of epigenetic

regulators is highly context-dependent, and as a result, their

global inhibition can lead to diverse and undesired side effects.

To possibly realise a cell-type, organ or tissue-specific inhibition,

we developed an enzyme-prodrug system using NTR and

nitroaromatic prodrugs of LSD1 inhibitor 1a. Together with the

newly developed luminescent probe 5, which enables a fast and

simple detection of NTR activity in any cell line, our system can

be applied with different targeting strategies to selectively inhibit

1

.0 µM). As the structurally related prodrugs 1c and 1f already

affected the colony formation of THP1^w^tin the low micromolar

range, they showed a smaller selectivity window compared with

2c and 2f. Still, 1c and 1f releasing the active LSD1 inhibitor as

parent drug are much more potent than the negative controls 2c

and 2f. The pairs, 1c/2c or 1f/2f, have identical nitroaromatic

prodrug moieties and release the same side products, thus the

increased potencies of the prodrugs 1c and 1f compared with 2c

+

and 2f in THP1-NTR cells must result from the released LSD1

LSD1. By varying the nitroaromatic prodrug moiety, the prodrugs

_w_e_r_e_o_p_t_i_m_i_s_e_d_t_o_b_e_s_t_a_b_l_e_i_n_T_H_P₁wt cells and to release the

inhibitor 1a (Figure 6).

The bystander effect, explained in the CD86-based cell assay

section above, was examined in the CFU assay, too. The assay

LSD1 inhibitor in presence of the prodrug-activating enzyme NTR,

thus reaching a sufficient selectivity window in vitro and in THP1

cells. Prodrug 1f using 2-nitro-N-methyl imidazole (9) as prodrug

moiety showed the best properties in our studies with isolated

NTR as well as in cellular experiments. This result is in

accordance with literature on prodrugs for other targets, where

substrates masked with 9 were recently found to be reduced

fastest by NTR, followed by 5-nitrofuranyl, 5-nitrothiophenyl (8)

and considerably slower the most commonly used 4-nitrobenzyl

_s_u_b_s_t_r_a_t_e_s_.[28] Gruber et al. subsequently demonstrated the

applicability of NTR for cell-specific chemical delivery by the

attachment of 9 to the hydroxy group of a cAMP analogue and to

wt

was performed with mixtures of the cell lines THP1 and THP1-

+

NTR and prodrug concentrations, at which mainly THP1-NTR

cells were affected in the CFU assay. Both prodrugs, 1b at 10 µM

and 1f at 1.0 µM, showed no bystander effect (Figure S8). Due to

the highly diluted cell concentration in the assay, cells have to

proliferate independently without cell-cell communication. Thus,

released inhibitor 1a after prodrug activation in transfected cells

is not able to diffuse significantly to the more remote cells in the

CFU assay. Overall, the in vitro best-activated prodrug 1f showed

the most potent effect in THP1-NTR⁺cell and the biggest

selectivity window between the two cell lines in the CFU assay

the secondary amine of an NMDA receptor antagonist, the latter

_v_i_a_a_c_a_r_b_a_m_a_t_e_l_i_n_k_e_r_.[28] With our enzyme-prodrug system, we

(

Figure 6).

were able to support a broad applicability of NTR in such prodrug

approaches and confirm the favourable properties of the 2-nitro-

N-methyl imidazolyl group (9) as an optimised prodrug moiety for

NTR-mediated release.

Additionally, we showed that 2-fluoro substitution on the benzyl

ring such as in 1c resulted in an increased NTR-mediated

Viability assay and GSH assay. As not only differentiation of

cells, but also cytotoxic effects can result in reduced numbers of

colonies in the CFU assay, all compounds were tested in a

viability assay with both cell lines. In THP1^w^tcells, no cytotoxicity

was observed (Figure S5 and Table S1). Interestingly, the

prodrugs 1c, 1f, 1g and 1e showed cytotoxic effects in the THP1-

NTR⁺cell line. The negative controls 2c and 2f showed

conversion in comparison to the simple nitrobenzyl in 1b, an effect

_a_l_s_o_d_e_s_c_r_i_b_e_d_f_o_r_a_n_i_t_r_o_r_e_d_u_c_t_a_s_e_f_r_o_m_T_._b_r_u_c_e_i_.[68] The fluorine

+

cytotoxicity in THP1-NTR cells, too. Both parent drugs, 1a of the

substitution likely favours an accelerated fragmentation of the

hydroxylamine intermediate, an impact already described for

electron-donating substituents in the 2-position of the

prodrugs and 2a of the negative controls had no cytotoxic effect

in both cell lines in the viability assay (Figure 6). Thus, the

cytotoxicities of prodrugs and negative controls in the two cell

-nitrobenzyl _m_o_i_e_t_y_.[81] Overall, the order of potency and

4

wt

+

lines THP1 and THP1-NTR in the viability assay must rise from

the side products of the enzymatic activation. The QM side

product derived from prodrugs with a p-hydroxy benzyl linker can

deplete the scavenger glutathione (GSH) in cells.^[⁷⁹^]Hulsman et

al. identified a GSH-QM adduct by LC-MS in HT29 cell lysates

after the incubation with their prodrugs.^[⁷⁹^]The cellular GSH

depletion can result in various signal transductions leading to

apoptosis or necrosis.^[⁸⁰^]Similar to QM, the Michael acceptors

released from our activated nitroaromatic molecules, are highly

electrophilic. Therefore, the cytotoxic negative controls 2c and 2f

and prodrugs 1b, 1c and 1f were tested in a GSH-assay with

Ellman’s reagent for thiol reactivity. All toxic compounds in the

selectivity for NTR-positive cells indicates a relationship between

the rate of nitro group reduction and the one-electron reduction

potential [E(1)] of the nitroaromatic groups. Hay et al. discussed

the complexity of this relationship for the NfsB enzyme and

Nivinskas et al. for the related oxygen-insensitive nitroreductase

_f_r_o_m_E_n_t_e_r_o_b_a_c_t_e_r_c_l_o_a_c_a_e_.[26,82] We could confirm a correlation,

nevertheless, the activation is probably also dependent on other

factors such as binding affinity or leaving group characteristics of

the prodrug molecule.

All activated prodrugs, except 1e and 1f, showed an impaired

fragmentation in the HPLC assay. The released amine 1a could

directly form an adduct with the fragmentation side product, the

Michael acceptor, thereby reducing the detectable 1a amount in

the assay. This adduct formation was observed for LSD1 inhibitor

+

viability assay resulted in reduced GSH-levels in THP1-NTR cells

but had no effect in THP1^w^tcells (Figure S6). Thus, enzymatic

1

0

ChemBioChem

10.1002/cbic.202000138

FULL PAPER

prodrug Q-PAC (Figure 1) that releases a QM as a side product.^[⁶^]

The nitrobenzyl prodrugs 1b, 1c, and 1d release an AQM, which

could also react with the released amine 1a. Studies about the

mechanism of reductive activation of the 5-nitroimidazole

containing ronidazole showed that upon reductive metabolism,

ronidazole binds to the sulfhydryl group of cysteine residues with

a subsequent loss of the carbamate group.^[⁶⁰^]Other nucleophiles

such as water are also supposed to bind to the 4-position of

reduced 5-nitroimidazoles to induce carbamate loss and drug

release.^[⁵⁹^,⁶⁰^]Therefore, the fragmentation of 5-nitroimidazole

prodrug 1g could be significantly slower compared with the other

prodrugs that rather fragment spontaneously after reduction.

In the cellular CD86-based assay for LSD1 inhibition, only low

nanomolar compound concentrations are needed for the

depletion by fragmentation products of 2-nitroimidazole, an effect

already investigated by Bérubé et al.^[⁸³^]From the prodrug 1b with

a p-nitrobenzyl alcohol moiety GSH depletion and cytotoxicity was

not observed. As we showed that prodrug 1b gets activated rather

slowly in vitro, compared with the fluorinated derivative 1c or the

2-nitroimidazole substituted prodrug 1f, we hypothesise that the

cells can counteract the slow release of the Michael acceptor

without induction of apoptosis or necrosis. In addition to GSH,

other cellular nucleophiles such as DNA may serve as reaction

partners for the released Michael acceptor which may also

contribute to cytotoxicity.

The negative controls 2c and 2f furthermore helped us to prove

that the released LSD1 inhibitor 1a is responsible for increased

CD86 expression and reduced colony formation of THP1-NTR⁺

cells after prodrug treatment. In the CD86-based assay, the

negative controls were inactive at nanomolar concentrations

relevant for our enzyme-prodrug system. As the released

negative control 2a is not active on LSD1 or cytotoxic, it is

possible to show the impact of the released side product of our

+

induction of CD86-positive THP1-NTR cells. In this concentration

range, all prodrugs except 1d reach the same level of CD86-

positive THP1 cells as 1a, here indicating

a complete

fragmentation of the prodrugs after enzymatic reduction.

Apparently, the released inhibitor does not react with

intermediates formed after prodrug reduction at these low

concentrations, making it available for LSD1 inhibition.

Furthermore, the different in vitro reactivity of the NTR with

prodrugs is negligible for the cellular CD86-assay due to

elongated incubation time and enzyme stability. At higher prodrug

concentrations, as in the cellular CFU assay, we observed

impaired fragmentation, as already shown in the in vitro assay

with recombinant NTR. The EC50 of the LSD1 inhibitor 1a in the

CFU assay was not reached by any prodrug in the NTR-

expressing cells, indicating that the LSD1 inhibitor 1a is not

quantitatively released under these conditions. One possibility is

that the released drug is scavenged by the AQM side product,

which was observed in the in vitro assay, thereby impeding the

efficacy of the prodrugs in the CFU assay.

+

prodrugs in the CFU assay with THP1-NTR cells. The negative

+

controls 2c and 2f reduced the ability of THP1-NTR cells to form

colonies, due to the formed cytotoxic side products, yet with

reduced potency compared with prodrugs 1c and 1f (Figure 6).

This additional colony reduction by the prodrugs is caused by the

released inhibitor 1a. The 2-nitroimidazole prodrug 1f showed the

best selectivity window with dose-dependent colony reduction

from 100 nM – 1 µM and no effect in THP1^w^tup to 2 µM.

The concentration-dependent effects of negative controls 2c and

2f highlight the importance to use an inhibitor with a potency

below the side product for such a prodrug approach. When only

small amounts of released inhibitor are needed to induce the

desired biologic effect, it is more likely that side products or side

reactions have little biological relevance. Overall, we could

identify a well-defined concentration range (10 – 100 nM in the

CD86-based assay) in which only LSD1 inhibition but no

cytotoxicity due to the formed side products is reached with

prodrug candidates 1b, 1c, and 1f. This allows the selective and

specific LSD1 inhibition in NTR-expressing THP1 cells, enabling

further studies on LSD1 and its cellular effects. Our enzyme-

prodrug system can be used as a research tool to elucidate the

cell-specific or organ-specific role of LSD1 in a heterogeneous

environment. Further development for clinical applications is also

possible, e.g. by using DEPT techniques to target the NTR to the

desired cell-type, tissue or organ.

Throughout all experiments, 1d showed no relevant

fragmentation, probably because of the unfavourable

destabilization of the arising benzylic cation by the CF group.

3

Therefore, 1d and in general nitroaromatics with electron-

withdrawing groups at the benzylic position could be used as

negative controls to possibly obtain activated intermediates such

as the nitroso and hydroxylamine that do not further fragment.

With the negative controls 2c and 2f, derived from inactive

inhibitor 2a, it is possible to distinguish for the first time the effects

from the released inhibitor and the electrophilic side product as

always both molecules are present after enzymatic reduction of

self-immolative nitroaromatic prodrugs. Hulsman et al.

demonstrated that, contrary to initial assumptions, the antitumour

effect of the hybrid drug nitric oxide-donating aspirin does not Conclusion

derive from the released NO or aspirin, but solely from the QM

side product.^[⁷⁹^]This example emphasises that negative controls

This study has identified suitable prodrug candidates of an LSD1

are important to avoid a false understanding of mechanisms,

distinguishing between effects from the released drug and

undesired effects from the side products. In many studies on NTR

prodrugs, the effect of the formed Michael acceptor is largely

neglected. Mostly potent cytotoxic agents were released so far,

thereby impeding the specific evaluation of cytotoxic effects

originated from the side product. As the released LSD1 inhibitor

inhibitor for the usage in an NTR-based enzyme-prodrug system

that are selective for NTR expressing cells. We thereby confirmed

the 2-nitro-N-methyl imidazolyl group as favourable prodrug

moiety. Detailed analysis of prodrug activation and fragmentation

revealed the effect of released LSD1 inhibitor and the cytotoxicity

of the formed Michael acceptor derived from the reduced

nitroaromatic prodrug moiety. For the first time, the isolated

cytotoxicity of these reactive fragments deriving from 2-nitro-N-

methyl imidazolyl and 2-fluoro-4-nitrobenzyl moieties was shown

and analysed, detecting GSH depletion as one possible mediator

of events causing cytotoxicity. Even though these prodrug

moieties are in widespread use, their diverse mechanisms of

1a and the negative control 2a are not cytotoxic per se, we could

evaluate the effects of the released Michael acceptor of

nitrobenzyl prodrugs for the first time. To our knowledge, we are

the first to describe GSH depletion by the AQM formed by the

fluorinated p-nitrobenzyl prodrug moiety. We also observed GSH

1

ChemBioChem

10.1002/cbic.202000138

FULL PAPER

cytotoxicity have not been evaluated well so far, as they have

mostly used in conjunction with cytotoxic drugs. Together with our

negative controls for both side products (2c and 2f) and prodrug

activation intermediates (1d), our system can directly be used as

a tool to further study the enzymatic and scaffolding function of

LSD1. Possibly the nitroreductase system of prodrugs, probes

and vector can also be applied to other targets. Another

alternative is to test the nitroaryl prodrugs in hypoxia-based

approaches, in which nitroaromatics are reduced by human

oxygen-sensitive type II nitroreductase selectively in tumour

tissue. Furthermore, our enzyme-prodrug system can be

extended by using different GDEPT approaches to target the NTR

gene towards other cancer cells for intracellular activation. The

most widely used transfection methods for gene therapy are viral

vectors.^[¹^]The emerging variety of non-viral targeting devices will

enable the direct use of our enzyme-prodrug system to test and

(ESI) as ion sources. GC–MS analyses were carried out on an Agilent

6890N Network GC system equipped with a 5973 Network Mass Selective

Detector (both Agilent Technologies, Santa Clara, USA) and a DB-5ms

column (length = 30 m, diameter = 0.25 mm, film = 0.25 μm; Agilent

Technologies). A carrier gas (helium) flow of 1 mL min⁻¹was used. The

injection volume was 1 μl with a split ratio of 41.7:1 at an injector

temperature of Tinjector = 250 °C. The temperature of the ion source was

T_i_o_n_s_o_u_r_c_e= 230 °C. Following temperature program (column oven) was

applied: 0-3 min: 60 °C; 3-14 min: linear increase to 280 °C; 14-19 min:

280 °C. HPLC analysis was performed to determine the purity of all final

compounds on an Agilent Technologies 1260 Infinity system using UV

detection at 210 nm and a Phenomenex Kinetex 5u XB-C18 100 Å 250 x

4.60 mm column. Eluent A was water containing 0.05% TFA and eluent B

was acetonitrile containing 0.05% TFA. Linear gradient conditions were as

follows: 0-4 min: A/B (90:10); 4-29 min: linear increase to 100% of B;

29−31 min: 100% B; 31–40 min: A/B (90:10). All final compounds

displayed a chemical purity of > 95% at the wavelength of 210 nm. The

HPLC analysis of isomer distributions were performed on an Agilent

establish these vehicles, such as liposomes and polymeric

_n_a_n_o_p_a_r_t_i_c_l_e_s_._[84,85] _B_o_t_h_l_i_t_e_r_a_t_u_r_e_d_a_t_a[86] and our studies in this

Technologies HP 1100 chromatography system equipped with a

photodiode array detector measuring an UV absorbance at 210 nm. The

different methods (chiral-M1 – chiral-M7) used are described in the

supplemental (S2-S3). The stereochemical descriptors R and S are

complemented with (*) to show their interchangeability, e.g. that R*S*

represents both RS and SR isomers.

manuscript with THP1 cells indicate that there is enough NADH

to achieve activation intracellularly. In contrast, the dependency

on the cofactor NADH may limit the use of approaches that

directly target the NTR protein (e.g. as an antibody conjugate)

towards the surface of cancer cells, as NADH has a very short

half-life in plasma. In this case, the addition of more stable NADH

analogs, such as reduced nicotinic acid riboside, together with the

targeted NTR should enable its use also for approaches that rely

on extracellular delivery.^[⁸⁷^]Our luminescent NTR probe can be

used as a primary and simple tool to evaluate cell-selectivity and

specificity in such new applications. Overall, these promising

results encourage further application of our enzyme-prodrug

system to target LSD1 or other epigenetic or signal transduction

inhibitors specifically in cancer cells where their function is

aberrant, thereby minimising off-target toxicity to non-tumour cells.

General procedure for reductive amination. The procedure described

_b_y_S_c_h_u_l_z_-_F_i_n_c_k_e_e_t_a_l_._w_a_s_a_p_p_l_i_e_d_.[62]

trans-N-((2-Methoxypyridin-3-yl) methyl)-2-phenylcyclopropan-1-amine 1a.

Colourless oil; yield 55% (450 mg, 1.77 mmol). R

¹H NMR (400 MHz, DMSO-d ): δ=8.03 (dd, ³JH,H=5.0 Hz, ⁴JH,H=1.9 Hz, 1H),

.67–7.62 (m, 1H), 7.24–7.18 (m, 2H), 7.13–7.07 (m, 1H), 7.01–6.97 (m,

H), 6.93 (dd, ³JH,H=7.2 Hz, ³

H,H=5.0 Hz, 1H), 3.81 (s, 3H), 3.73 (s, 2H),

.80 (br s,1H), 2.23 (ddd, ³Jcis=7.1, ³

trans=3.1 Hz, 1H), 1.81

f

=0.18 (CH/EtOAc 7:3);

6

7

2

J

trans=4.1, ³

J

(

ddd, ³Jcis=9.1, ³Jtrans=5.7, ³Jtrans=3.1 Hz, 1H), 1.01 (ddd, ³Jcis=9.1,

H,H=4.8, ³Jtrans=4.1 Hz, 1H), 0.93 ppm (ddd, ³Jcis=7.1, ³Jtrans=5.7,

2

J

2_JH,H=4.8 Hz, 1H); C NMR (101 MHz, DMSO-d

37.5, 128.5, 125.8, 125.5, 123.3, 117.2, 53.4, 47.0, 42.1, 25.0, 17.3 ppm;

LRMS (APCI): m/z (%) 255.2 (100) [M+H]⁺; HPLC retention time

13

6

): δ=161.5, 144.9, 142.9,

1

5

4.096 min, 97.5%; HPLC (chiral-M3) retention time 4.923 min (49.4%),

.399 min (50.6%).

Experimental Section

General Procedures. The reactions were carried out in glassware under

inert (nitrogen) atmosphere. Used reagents and solvents were purchased

from commercial sources and used without further purification. Reactions

were monitored by thin-layer chromatography (TLC) performed with Merck

alumina plates coated with silica gel 60 F254 (layer thickness: 0.2 mm) and

analysed under UV light (254 nm) or revealed using ninhydrin as a staining

agent for primary and secondary amines. Yields were not optimised. Flash

_c_o_l_u_m_n_c_h_r_o_m_a_t_o_g_r_a_p_h_y_w_a_s_p_e_r_f_o_r_m_e_d_o_n_a_B_i_o_t_a_g_e® Isolera Prime/One

purification system using pre-packed silica gel columns (40-60 µM) from

Biotage (SNAP) or Telos and the purifications were followed by TLC. NMR

(

1S*,2S*,3R*)-N-((2-Methoxypyridin-3-yl) methyl)-2-methyl-3-phenylcyclo

propan-1-amine 2a. Colourless oil; yield 39% (193 mg, 0.719 mmol).

R

3

=0.35 (CH/EtOAc 7:3); ¹H NMR (400 MHz, DMSO-d

): δ=8.03 (dd,

H,H=5.0 Hz, ⁴JH,H=1.9 Hz, 1H), 7.65 (dd, ³JH,H=7.3 Hz, ⁴JH,H=1.9 Hz, 1H),

f

6

J

7

3

.22–7.18 (m, 2H), 7.11–7.06 (m, 1H), 6.97–6.92 (m, 3H), 3.81 (s, 3H),

.73 (m, 2H), 2.66–2.61 (br s, 1H), 2.30 (dd, Jcis=6.9, ³Jtrans=3.4 Hz, 1H),

3

trans=4.9, ³Jtrans=3.4 Hz, 1H), 1.24–1.14 ppm (m, 4H); ¹³C NMR

1.44 (dd, J

(

1

101 MHz, DMSO-d

23.4, 117.2, 53.4, 47.4, 46.0, 31.7, 23.8, 12.2 ppm; LRMS (APCI) m/z

6

): δ=161.5, 145.0, 143.4, 137.6, 128.5, 125.7, 125.3,

+

(100) 269.4 [M+H] ; HPLC retention time 14.466 min, 95.4%; HPLC

spectroscopy and mass spectrometry were used for product identification.

₁_H_-_a_n_d13C-NMR spectral data were recorded on a Bruker Advance II+

(chiral-M7) retention time 5.922 min (49.0%), 6.208 min (51.0%).

6 3

400 MHz spectrometer using as solvents DMSO-d and CDCl . Chemical

shifts (δ) are reported in ppm and referenced to a residual solvent peak,

coupling constant (J) are expressed in Hz and multiplicity abbreviations

are as follows: s = singlet, d = doublet, dd = doublet of doublets, ddd =

doublet of doublets of doublets, dt = doublet of triplets, t = triplet,

q = quartet, br = broad signal, m = multiplet and st-m = stereoisomerism-

derived multiplet. The H and ¹³C assignment of new compounds resulted

from 2D experiments and was numbered according to the IUPAC system

and abbreviated as follows: B = benzyl, Cy = cyclopropyl, P = phenyl, Py

General procedure for the synthesis of prodrugs 1b – 1g and negative

controls 2c and 2f. For the activation of amine 1a or 2a, triphosgene

(0.5 eq) was dissolved in CH

bath. The addition of pyridine (2.2 eq) afforded a yellow suspension to

which a solution of the amine (1.0 eq) in CH Cl (1 mL per 0.6 mmol) was

2 2

Cl (1 mL per 0.3 mmol) at 0 °C using an ice-

2

1

added over 15 min. The resulting reddish solution was stirred for 20 min

on ice and was then allowed to adjust to rt. After completion of the reaction,

the reaction mixture was treated with 1 M HCl aq. and extracted with CH

2 2

Cl .

=

pyridinyl, I = imidazolyl and T = thiophenyl. ¹³C signals marked with (*)

The combined organic layers were washed with brine, dried over Na SO ,

2

4

are only detected in HSQC and HMBC spectra. Mass spectra were

recorded on an Advion expression CMS mass spectrometer (LRMS: low-

resolution MS) and on a Thermo Scientific Exactive mass spectrometer

filtered and concentrated in vacuo to yield the carbamoyl chloride as a

yellow oil. For following carbamate formation, the respective carbamoyl

chloride (1 eq) was immediately dissolved in CH

and pyridine (1.5 eq) was added. The corresponding alcohol (1.0 eq) was

dissolved in CH Cl (0.25 mL/0.1 mmol) and treated with NaH

2 2

Cl (0.5 mL/0.1 mmol)

(

HRMS) using ASAP^®(Atmospheric Solids Analysis Probe; aka APCI:

Atmospheric Pressure Chemical Ionization) and electrospray ionization

2

1

2

ChemBioChem

10.1002/cbic.202000138

FULL PAPER

+

(

60% dispersion in mineral oil, 1.0 eq). For compounds 1f, 1g and 2f, THF

(m); LRMS (APCI) m/z (%) 502.2 (100) [M+H] ; HRMS (ESI): m/z calcd for

+

was additionally used as solvent. In the case that the alcohol was still

present after 16 h, more NaH was added until gas development stopped.

After consumption of starting material, the reaction mixture was quenched

C H O N F : 502.1584 [M+H] ; found: 502.1581; HPLC retention time

28.260 min, 97.7%; HPLC (chiral-M2) retention time 11.562 min (7.5%),

12.804 min (7.6%), 15.002 min (22.6%), 16.410 min (62.4%).

25 23 5 3 3

by addition of 1 M HCl aq. An appropriate amount of CH

the organic layer was washed with 1 M HCl aq. and brine (2 ×), dried over

Na SO , filtered and evaporated. For subsequent column chromatography,

the composition of the mobile phase was adjusted to the compound

properties. Synthesis of prodrug 1e differed, starting with activation of the

alcohol 8 to carbonate 11 and subsequent coupling with 1a, as described

below.

2 2

Cl was added and

(

5-Nitrothiophen-2-yl)methyl((2-methoxypyridin-3-yl)methyl)((1S*,2R*)-2-

phenylcyclo propyl) carbamate 1e. Carbonate 11 (78 mg, 0.241 mmol,

.5 eq) was dissolved in DMF (0.5 mL) and added dropwise to a solution

of amine 1a (41 mg, 0.160 mmol, 1.0 eq) and DIPEA (41 µL, 0.241 mmol,

.5 eq) in DMF (1.0 mL). The colour of the solution turned from yellow to

green. After 17 h, the reaction mixture was diluted with EtOAc (15 mL) and

washed with brine (3 × 10 mL). The organic layer was dried over Na SO

2

4

1

2

4

,

4

-Nitrobenzyl((2-methoxypyridin-3-yl)methyl)((1S*,2R*)-2-phenylcyclo

filtered and concentrated in vacuo. Purification via column

chromatography (10 to 30% EtOAc in CH over 10 CV on a biotage system,

followed by a manual column using CH/THF 9:2) afforded 1e as an orange

propyl)carbamate 1b. Colourless oil; yield 66% (107 mg, 0.247 mmol).

R

1

f

=0.68 (CH/EtOAc 1:1); H NMR (400 MHz, DMSO-d

6

): δ=8.20–8.10 (m,

2

H; B-3 and B-5), 8.09–8.05 (m, 1H; Py-6), 7.61–7.42 (m, 3H; B-2, B-6 and

Py-4), 7.27–7.17 (m, 2H; P-3 and P-5), 7.17–7.11 (m, 1H; P-4), 7.11–6.98

m, 2H; P-2 and P-6), 6.98–6.91 (m, 1H; Py-5), 5.35–5.17 (st-m, 2H; OCH ),

), 2.92–2.73 (m, 1H; Cy-1),

.34–2.25 (m, 1H; Cy-2), 1.38–1.28 (m, 1H; Cy-3b), 1.27–1.18 ppm (m,

H; Cy-3a); ¹³C NMR (101 MHz, DMSO-d

): δ=160.7 (Py-2), 156.2 (C=O),

oil; yield 47% (33 mg, 0.075 mmol). R

(500 MHz, DMSO-d

, 70 °C): δ=8.04 (dd, ³JH,H=5.0 Hz, ⁴JH,H=1.8 Hz, 1H;

=0.17 (CH/THF 9:2); ¹H NMR

f

6

(

2

Py-6), 7.95 (d, ³JH,H=4.2 Hz, 1H; T-4), 7.48 (dd, ³JH,H=7.2 Hz, ⁴JH,H=1.8 Hz,

1H; Py-4), 7.22–7.17 (m, 3H; P-3, P-5 and T-3), 7.14–7.10 (m, 1H; P-4),

4

2

1

2

2 3

.56–4.40 (st-m, 2H; NCH ), 3.83 (s, 3H; CH

7.06–7.03 (m, 2H; P-2 and P-6), 6.90 (dd, ³

Py-5), 5.35–5.29 (st-m, 2H; OCH

), 4.51 (d, ²JH,H=16.5 Hz, 1H; NCH

4.43 (d, ²JH,H=16.5 Hz, 1H; NCH ), 2.74 (ddd, ³Jcis=7.5,

), 3.83 (s, 3H; CH

J

H,H=7.2, ³

J

H,H=5.0 Hz, 1H;

),

6

2

a b

H

46.8 (B-4), 145.2 (Py-6), 144.5 (B-1), 140.6 (P-1), 136.4 (Py-4), 128.1 (B-

and B-6), 128.0 (P-3 and P-5), 125.9 (P-2 and P-6), 125.7 (P-4), 123.3

H

a b

3

³Jtrans=4.4, ³Jtrans=3.5 Hz, 1H; Cy-1), 2.24 (ddd, ³Jcis=9.8, ³Jtrans=6.5,

³Jtrans=3.5 Hz, 1H; Cy-2), 1.32 (ddd, Jcis=9.8, ²JH,H=6.0, ³Jtrans=4.4 Hz, 1H;

3

(

B-3 and B-5), 120.0 (Py-3), 116.9 (Py-5), 65.4 (OCH

NCH

2 3

), 53.1 (CH ), 45.8

), 40.1* (Cy-1), 25.9 (Cy-2), 16.4 ppm (Cy-3); LRMS (ESI) m/z (%)

Cy-3b), 1.20 ppm (ddd, ³Jcis=7.5, ³

J JH,H=6.0 Hz, 1H; Cy-3a);

trans=6.5, ²

2

+

¹³C NMR (126 MHz, DMSO-d

434.1 (43) [M+H] , 456.1 (100) [M+Na] ; HRMS (ESI): m/z calcd for

6

, 70 °C): δ=160.6 (Py-2), 155.6 (C=O), 150.6

+

C

H O N

24 24 5 3

: 434.1710 [M+H] ; found: 434.1715; HPLC retention time

(T-5), 147.6 (T-2), 145.0 (Py-6), 140.1 (P-1), 136.4 (Py-4), 128.9 (T-4),

127.7 (P-3 and P-5), 126.8 (T-3), 125.8 (P-2 and P-6), 125.4 (P-4), 119.6

2

1

6.282 min, 97.3%; HPLC (chiral-M3) retention time 15.352 min (49.7%),

6.283 min (50.3%).

(Py-3), 116.5 (Py-5), 61.1 (OCH

2

), 52.6 (CH

3

), 45.6 (NCH

2

), 38.8 (Cy-1),

+

25.3 (Cy-2), 15.9 ppm (Cy-3); LRMS (APCI): m/z (%) 440.1 (100) [M+H] ;

+

HRMS (ESI): m/z calcd for C22

40.1272; HPLC retention time 25.816 min, 98.4%; HPLC (chiral-M5)

H O N S : 440.1275 [M+H] ; found:

22 5 3

2-Fluoro-4-nitrobenzyl((2-methoxypyridin-3-yl)methyl)((1S*,2R*)-2-

4

phenylcyclopropyl) carbamate 1c. Colourless oil; yield 43% (32 mg,

.071 mmol). R ):

=0.39 (CH/EtOAc 7:3); ¹H NMR (400 MHz, DMSO-d

δ=8.11 (dd, ³JH,H=9.9 Hz, ⁴ H,H=2.1 Hz, 1H; B-5), 8.07 (dd, ³

H,H=5.0 Hz,

⁴JH,H=1.6 Hz, 1H; Py-6), 8.05–7.95 (m, 1H; B-3), 7.65–7.57 (m, 1H; B-6),

.55–7.45 (m, 1H; Py-4), 7.30–7.12 (m, 3H; P-3, P-4 and P-5), 7.12–6.91

m, 3H; P-2, P-6 and Py-5), 5.33–5.20 (st-m, 2H; OCH ), 4.54–4.38 (st-m,

), 2.84–2.71 (m, 1H; Cy-1), 2.26 (ddd, ³Jcis=9.8,

trans=3.3 Hz, 1H; Cy-2), 1.38–1.29 (m, 1H; Cy-3b), 1.26–

.17 ppm (m, 1H; Cy-3a); ¹³C NMR (101 MHz, DMSO-d

): δ=161.2 (Py-2),

59.7 (d, ¹JC,F=250.6 Hz; B-2), 156.6 (C=O), 148.5 (d, ³JC,F=9.1 Hz; B-4),

45.7 (Py-6), 140.9 (P-1), 136.9 (Py-4), 131.9 (d, ²

C,F=14.7 Hz; B-1),

31.2 (d, ³

C,F=4.5 Hz; B-6), 128.5 (P-3 and P-5), 126.3 (P-2 and P-6),

26.1 (P-4), 120.4 (Py-3), 119.9 (d, JC,F=3.4 Hz; B-5), 117.4 (Py-5), 111.6

retention time 9.538 min (49.7%), 10.544 (50.3%).

0

f

6

J

(1-Methyl-2-nitro-1H-imidazol-5-yl)methyl((2-methoxypyridin-3-yl)methyl)

7

(

((1S*,2R*)-2-phenylcyclopropyl)carbamate 1f. Yellowish solid; yield 50%

=0.42 (CH/EtOAc 3:7); ¹H NMR (400 MHz,

f

2

(83 mg, 0.190 mmol). R

DMSO-d

, 50 °C): δ=8.06 (dd, ³JH,H=4.9 Hz, ⁴JH,H=1.6 Hz, 1H; Py-6), 7.51–

2

J

H; NCH

2

), 3.82 (s, 3H; CH

3

6

3

trans=6.5, ³

J

7.46 (m, 1H; Py-4), 7.23 (s, 1H; I-4), 7.18–7.08 (m, 3H; P-3, P-4 and P-5),

2

1

6

7.00–6.89 (m, 3H; P-2, P-6 and Py-5), 5.26 (d, JH,H=13.0 Hz, 1H; OCH

a

H

b

),

), 3.65 (s, 1H;

5.16 (d, ²JH,H=13.0 Hz, 1H; OCH

4.40 (d, ²

H,H=16.5 Hz, 1H; NCH

NCH

b a b

H H

), 4.50 (d, ²JH,H=16.5 Hz, 1H; NCH

a

J

a b

H

), 3.85 (s, 3H; OCH

3

J

3

), 2.73 (ddd, Jcis=7.6, ³Jtrans=4.5, Jtrans=3.5 Hz, 1H; Cy-1), 2.21 (ddd,

4

³Jcis=9.9, ³Jtrans=5.9, ³Jtrans=3.5 Hz, 1H; Cy-2), 1.31 (ddd, ³Jcis=9.9,

²JH,H=5.9, ³ trans=4.5 Hz, 1H; Cy-3b), 1.22–1.17 ppm (m, 1H; Cy-3a); ¹³C

NMR (101 MHz, DMSO-d , 50 °C): δ=161.3 (Py-2), 156.3 (C=O), 146.5

(

2

–

d, ²JC,F=26.6 Hz; B-3), 60.7 (OCH

2

), 53.5 (CH

3

), 46.3 (NCH

2

), 39.6* (Cy-1),

6

): δ= -114.03

J

6.5 (Cy-2), 16.7 ppm (Cy-3); ¹⁹F NMR (376 MHz, DMSO-d

6

+

-114.13 ppm (m); LRMS (APCI): m/z 452.2 (100) [M+H] ; HRMS (APCI):

(I-2), 145.7 (Py-6), 140.9 (P-1), 137.0 (Py-4), 133.7 (I-5), 129.0 (I-4), 128.3

(P-3 and P-5), 126.2 (P-2 and P-6), 126.1 (P4), 120.5 (Py-3), 117.3 (Py-5),

+

23 5 3

m/z calcd for C24H O N F : 452.1616 [M+H] ; found: 452.1617; HPLC

retention time 26.375 min, 97.8%; HPLC (chiral-M3) retention time

56.6 (OCH

2

), 53.5 (OCH

3

), 46.3 (NCH

2

), 39.7* (Cy-1), 34.2 (NCH

3

), 26.4

+

15.612 min (48.7%), 16.491 min (51.3%).

(Cy-2), 16.3 ppm (Cy-3); LRMS (APCI): m/z (%) 438.3 (100) [M+H] ;

24 5 5

H O N

⁺: 438.1772 [M+H]⁺; found:

HRMS (ESI): m/z calcd for C22

38.1769; HPLC retention time 22.003 min, 96.5%; HPLC (chiral-M6)

4

2,2,2-Trifluoro-1-(4-nitrophenyl)ethyl((2-methoxypyridin-3-yl)methyl)

retention time 14.767 min (50.3%), 18.637 min (49.7%).

(

(1S*,2R*)-2-phenyl cyclopropyl) carbamate 1d. Yellowish oil; yield 48%

61 mg, 0.122 mmol). R

=0.46 (CH/EtOAc 7:3); ¹H NMR (400 MHz,

, 50 °C): δ=8.26 (d, ³

f

DMSO-d

6

J

H,H=8.8 Hz, 1H; B-3 or B-5), 8.20 (d,

(1-Methyl-5-nitro-1H-imidazol-2-yl)methyl((2-methoxypyridin-3-yl)methyl)

³JH,H=8.6 Hz, 1H; B-3 or B-5), 8.08 (d, JH,H=4.9 Hz, 1H; Py-6), 7.78–7.74

3

((1S*,2R*)-2-phenylcyclopropyl)carbamate 1g. Reddish oil; yield 52%

(

m, 1H; B-2 or B-6), 7.68 (d, ³JH,H=8.6 Hz, 1H; B2 or B-6), 7.49–7.47 (m,

(55 mg, 0.125 mmol). R

DMSO-d ): δ=8.10–8.05 (m, 1H; Py-6), 7.58–7.49 (m, 1H; Py-4), 7.26–

=0.44 (CH/EtOAc 3:7); ¹H NMR (400 MHz,

f

1

7

H; Py-4), 7.35–7.23 (m, 2H; P-3 and P-5), 7.23–7.15 (m, 1H; P-4), 7.15–

.08 (m, 2H; P-2 and P-6), 6.95–6.90 (m, 1H; Py-5), 6.62–6.56 (st-m, 1H;

6

7.02 (m, 4H; I-4, P-3, P-4 and P-5), 7.00–6.82 (m, 3H; P-2, P-6 and Py-5),

2

CHCF

2

1

3

), 4.71–4.38 (st-m, 2H; NCH

.84 (m, 1H; Cy-1), 2.40–2.24 (m, 1H; Cy-2), 1.44–1.37 (m, 1H; Cy-3b),

.36–1.17 ppm (m, 1H; Cy-3a); ¹³C NMR (101 MHz, DMSO-d

, 50 °C):

2 3

), 3.83–3.77 (st-m, 3H; CH ), 2.97–

5.26 (d, JH,H=13.3 Hz, 1H; OCH

a

H

b

), 5.18 (d, JH,H=13.3 Hz, 1H; OCH

a

H

b

),

2

4.49 (d, JH,H=16.5 Hz, 1H; NCH

a

H

b

), 4.38 (d, JH,H=16.5 Hz, 1H; NCH

a

H

b

6

3.84 (s, 3H; OCH

3

), 3.60 (s, 3H; NCH

3

), 2.74–2.70 (m, 1H; Cy-1), 2.21

3

δ=160.7 (Py-2), 153.8 (C=O), 148.2 (B-4), 145.4 (Py-6), 140.1 (st-m; P-1),

(ddd, Jcis=10.1, ³Jtrans=6.1, ³Jtrans=3.4 Hz, 1H; Cy-2), 1.31 (ddd, ³Jcis=10.1,

2

H,H=5.8, ³

J

1

7

2

38.2–138.1 (st-m; B-1), 136.8–136.7 (st-m; Py-4), 129.0 (B-2 and B-6),

27.9 (P-3 and P-5), 125.8 (P-2 and P-6), 125.7 (P-4), 123.4 (B-3 or B-5),

J

trans=4.8 Hz, 1H; Cy-3b), 1.24–1.17 ppm (m, 1H; Cy-3a);

): δ=161.1 (Py-2), 156.4 (C=O), 148.1 (I-5),

13

C NMR (101 MHz, DMSO-d

6

23.3 (B-3 or B-5), 122.8 (q, ¹JC,F=280 Hz; CF

), 119.3 (Py-3), 116.7 (Py-5),

), 46.1 (NCH ), 39.1 (Cy-1), 28.8 (Cy-3),

5.2 ppm (Cy-2); ¹⁹F NMR (376 MHz, DMSO-d

): δ= -74.66 – -75.06 ppm

145.6 (Py-6), 140.9 (P-1), 139.8 (I-2), 136.9 (Py-4), 132.1 (I-4), 128.2 (P-3

and P-5), 126.1 (P-2, P-4 and P-6), 120.4 (Py-3), 117.4 (Py-5), 59.4

(OCH ), 53.6 (OCH ), 46.3 (NCH ), 39.7* (Cy-1), 33.6 (NCH ), 26.9 (Cy-2),

2 3 2 3

3

1.8–70.8 (m; CHCF

3

), 52.8 (CH

3

2

6

1

3

ChemBioChem

10.1002/cbic.202000138

FULL PAPER

+

1

6.1 ppm (Cy-3); LRMS (APCI): m/z (%) 438.2 (100) [M+H] ; HRMS (ESI):

(OCH

3

), 47.2 (Fmoc-C9), 46.0 (NCH

2

), 39.1* (Cy-3), 26.7 (Cy-1), 25.5 ppm

+

m/z calcd for C22

H O N

24 5 5

: 438.1772 [M+H] ; found: 438.1770; HPLC

retention time 22.338 min, 98.1%; HPLC (chiral-M1) retention time

(Cy-1); LRMS (APCI): m/z (%) 477.2 (100) [M+H] ; HRMS (ESI): m/z calcd

for C31H N O : 477.2173 [M+H] ; found: 477.2177; HPLC retention time

+

29 2 3

20.942 min (49.8%), 21.680 min (50.2%).

29.668 min, 99.6%; HPLC (M-FLD) retention time 11.990 min, 100%.

2

-Fluoro-4-nitrobenzyl((2-methoxypyridin-3-yl)methyl)((1S*,2S*,3R*)-2-

Synthesis of luminescent probes 4 and 5. Probe 4 was synthesised

using standard Mitsunobu conditions as described and synthesis of probe

5 was performed following the procedures adopted from Mustafa et al.^[⁷³^]

methyl-3-phenylcyclo propyl)carbamate 2c. Colourless oil; yield 41%

1

(

62 mg, 0.133 mmol). R

f

=0.47 (CH/EtOAc 7:3); H NMR (400 MHz, DMSO-

d

1

2

6

1

2

1

6

) δ=8.19–8.07 (m, 2H; B-3 and B-5), 8.01 (dd, ³JH,H=5.0 Hz, ⁴JH,H=1.9 Hz,

H; Py-6), 7.82–7.67 (m, 1H; B-6), 7.57–7.46 (m, 1H; Py-4), 7.21–7.14 (m,

H; P-3 and P-5), 7.13–7.07 (m, 1H; P-4), 7.01–6.91 (m, 2H; P-2 and P-6),

.90–6.83 (m, 1H; Py-5), 5.37–5.18 (m, 2H; OCH

6-((5-Nitrothiophen-2-yl)methoxy)benzo[d]thiazole-2-carbonitrile 4.

2-cyano-6-hydroxybenzothiazole (107 mg, 0.608 mmol, 1.0 eq), alcohol 8

), 4.61 (d, ²JH,H=16.0 Hz,

), 3.71 (s, 3H; OCH ),

trans=4.0 Hz, 1H; Cy-3),

); ¹³C NMR

2

3

(101 mg, 0.638 mmol, 1.05 eq) and PPh (167 mg, 0.638 mmol, 1.05 eq)

H; NCH

H

a b

), 4.41 (d, ²

J

H,H=16.0 Hz, 1H; NCH

a

H

b

3

were dissolved in dry DMF (200 µL) in a sonication bath. During sonication,

DIAD (125 µL, 0.638 mmol, 1.05 eq) was added dropwise over 3 min. After

.85–2.80 (m, 1H; Cy-1), 1.94 (dd, ³Jtrans=6.1, ³

J

3

.53–1.44 (m, 1H; Cy-2), 1.13 ppm (d, JH,H=6.1 Hz, 3H; CH

101 MHz, DMSO-d

6

): δ=161.4 (Py-2), 159.8 (d, ¹JC,F=251.0 Hz; B-2),

3

15 min sonication, the red suspension was dissolved with cyclohexane

(

1

(

500 µL) and directly loaded on a silica column. After purification via silica

56.8* (C=O), 148.5 (d, ³JC,F=9.1 Hz; B-4), 145.8 (Py-6), 141.2 (P-1), 137.9

chromatography (15% to 50% EtOAc in CH over 10 CV), the product-

containing fractions were concentrated in vacuo. Crystallization from

EtOAc/MeOH/CH 1:1:1 afforded a brown solid; yield 2.7% (5.22 mg,

Py-4), 131.9 (d, ²JC,F =14.5 Hz; B-1), 131.3 (d, ³JC,F=4.5 Hz; B-6), 128.5

P-3 and P-5), 126.2 (P-2 and P-6), 126.0 (P-4), 120.1 (Py-3 and B-5),

1

17.2 (Py-5), 111.6 (d, ²

6.9 (NCH

J

C,F=27.2 Hz; B-3), 60.7 (OCH

), 43.9* (Cy-1), 30.8 (Cy-3), 24.9 (Cy-2), 13.1 ppm (CHCH

F NMR (376 MHz, DMSO-d ): δ= -113.95 – -114.03 ppm (m); LRMS

APCI): m/z (%) 466.6 (100) [M+H]⁺; HRMS (ESI): m/z calcd for

2

), 53.5 (OCH

3

),

₌₀_.₆₂₍_C_H_/_E_t_O_A_c₁_:₁₎_;1H NMR (400 MHz, DMSO-d

H,H=9.1 Hz, 1H; H-4), 8.10 (d, 3_JH,H=4.1 Hz, 1H; T-4), 8.04 (d,

0

.016 mmol). R

f

6

) δ

4

2

3

);

_.₂₁₍_d_,3

8

J

19

6

4

J

H,H=2.6 Hz, 1H; H-7), 7.45 (dd, JH,H₌₉_.₁_,4_JH,H=2.6 Hz, 1H; H-5), 7.39 (d,

3

(

3_JH,H=4.1 Hz, 1H; T-3), 5.55 ppm (s, 2H; OCH

DMSO-d ): δ=158.3 (C-6 or T-5), 151.2 (C-6 or T-C-5), 148.2 (C-3a), 147.1

T-2), 137.8 (C-2), 134.9 (C-7a), 130.2 (T-4), 127.8 (T-3), 126.0 (C-4),

₎_;13C NMR (101 MHz,

2

+

C

25 25 5 3

H O N F : 466.1773 [M+H] ; found: 466.1770; HPLC retention time

6

2

1

7.298 min, 96.5%. HPLC (chiral-M3) retention time 10.729 min (49.5%),

1.379 min (50.5%).

(

1

19.1 (C-5), 113.9 (CN), 106.6 (C-7), 65.4 ppm (OCH

2

) ; LRMS (APCI):

+Cl^-:

: 315.9856 [M–

+

m/z (%) 318.0 (100) [M+H] ; HRMS (ESI): m/z calcd for C13

351.9623 [M+Cl] ; found: 351.9619; calcd for C13

H] ; found: 315.9852; HPLC retention time 24.385 min, 97.3%.

H

7

O

3

N

3

S

2

–

S ^-

(

1-Methyl-2-nitro-1H-imidazol-5-yl)methyl((2-methoxypyridin-3-yl)methyl)

(1S*,2S*,3R*)-2-methyl-3-phenylcyclopropyl)carbamate 2f. Yellowish

H O N

6 3 3

2

–

1

foam; yield 50% (74 mg, 0.164 mmol). R

f

=0.62 (CH/EtOAc 3:7); H NMR

(

400 MHz, DMSO-d

6

): δ=8.05–7.98 (m, 1H; Py-6), 7.60–7.38 (m, 1H; Py-4),

.34–7.23 (m, 1H; I-4), 7.21–7.05 (m, 3H; P-3, P-4 and P-5), 6.97–6.79 (m,

H; P-2, P-6 and Py-5), 5.31–5.19 (st-m, 2H; OCH

), 4.55 (d, ²JH,H=16.0 Hz,

), 3.81–3.68 (m, 6H;

2

-Cyanobenzo[d]thiazol-6-yl ((5-nitrothiophen-2-yl)methyl) ethane-1,2-

diylbis(methyl carbamate) 5. Boc-protected 12 (178 mg, 0.456 mmol,

.0 eq) was dissolved in CH Cl (1 mL), cooled in an ice bath and treated

with thioanisol (269 µL, 2.28 mmol, 5.0 eq), previously diluted with CH Cl

1 mL). TFA (873 µL, 11.4 mmol, 25 eq) was added dropwise over 5 min

7

3

1

2

1

2

H; NCH H

a b

), 4.42 (d, ²

J

H,H=16.0 Hz, 1H; NCH

a

H

b

2

), 2.78 (dd, Jcis=7.3, ³Jtrans=4.0 Hz, 1H; Cy-1), 1.96–1.87

3

OCH

3

and NCH

3

(

(m, 1H; Cy-3), 1.52–1.41 (m, 1H; Cy-2), 1.18–1.09 (m, 3H; CH

(101 MHz, DMSO-d ): δ=161.3 (Py-2), 156.8 (C=O), 146.4 (I-2), 145.8

(Py-6), 141.3 (P-1), 137.8 (Py-4), 133.8 (I-5), 129.0 (I-4), 128.4 (P-3 and

); ¹³C NMR

3

while stirred on ice. After 20 min, the ice bath was removed and the slightly

yellow mixture was stirred for additional 2 h. The excess TFA was removed

by co-evaporation with toluene to yield an orange oil as twofold TFA salt

6

P-5), 126.1 (P-2 and P-6), 126.0 (P-4), 120.1 (Py-3), 117.2 (Py-5), 56.5

OCH ), 53.5 (OCH ), 47.0 (NCH ), 44.0 (Cy-1), 34.4 (NCH ), 31.3 (Cy-3),

4.7 (Cy-2), 13.2 ppm (CHCH ); LRMS (APCI): m/z (%) 452.6 (100)

(

1

(

2

242 mg, 0.468 mmol, 103%) that was used directly for the coupling with

1. Therefore, 11 (443 mg, 1.37 mmol, 3.0 eq) was dissolved in THF

2 mL), cooled on an ice bath and treated with DIPEA (170 µL, 1.00 mmol,

.2 eq). The previously deprotected amine was dissolved in THF (2 mL)

(

2

3

2

3

+

[

M+H] ; HRMS (ESI) m/z calcd for C23

H O N

26 5 5

: 452.1928 [M+H] ; found:

452.1927; HPLC retention time 22.894 min, 98.2%; HPLC (chiral-M4)

and added dropwise to the activated alcohol. After 15 min, the ice bath

was removed and the orange solution was stirred for 21 h at rt. The THF

was evaporated and the crude was purified via silica chromatography

(20% to 65% EtOAc in CH over 10 CV) to afford a yellow oil that was further

purified via semi-preparative HPLC (column: Phenomenex Synergi 10u

Hydro-RP 80Å 250 x 15.00 mm; flow rate: 2.5 mL/min; eluent A was water

retention time 10.423 min (50.0%), 13.395 min (50.0%).

Synthesis of fluorophore 3. (9H-Fluoren-9-yl)methyl ((2-methoxypyridin-

3

(

5

-yl)methyl)((1S*,2R*)-2-phenylcyclopropyl) carbamate 3. Amine 1a

86 mg, 0.34 mmol, 1.0 eq) was dissolved in acetonitrile (6 mL) and

% NaHCO3 aq. (2 mL). Fmoc chloride (86 mg, 0.34 mmol, 1.0 eq) was

containing 0.05% TFA and eluent B was acetonitrile containing

added and the solution was stirred for 1 h at rt. Acetonitrile was removed

under reduced pressure and the aqueous layer was extracted with EtOAc.

0.05% TFA; Linear gradient conditions were as follows: 0–4 min: A/B

(30:70); 4−29 min: linear increase to 90% of B; 29−31 min: 90% B; 31–

The organic layer was washed with brine, dried over Na

purified via silica chromatography (10% to 15% EtOAc in CH over 10 CV)

to afford 3 as a colourless solid; yield 89.4% (144 mg, 0.30 mmol). R =0.49

CH/EtOAc 7:3); ¹H NMR (400 MHz, DMSO-d ): δ=8.04 (dd, ³

H,H=5.0,

H,H=1.9 Hz, 1H; Py-6), 7.88–7.81 (m, 2H; Fmoc-H-2 and H-7), 7.68–7.43

m, 2H; Fmoc-H-3 and H-6), 7.43–7.32 (m, 2H; Fmoc-H-1 and H-8), 7.29–

.15 (5H, m, P-3, P-5, Py-4, Fmoc-H-4 and H-5), 7.15–6.94 (3H, m, P-2,

P-4 and P-6), 6.93–6.84 (m, 1H; Py-5), 4.62–4.40 (m, 2H; OCH ), 4.40–

), 4.25–4.18 (m, 1H; Fmoc-H-9), 3.78 (s, 3H; OCH ),

.70–2.64 (m, 1H; Cy-1), 2.21–2.15 (m, 1H; Cy-2), 1.28–0.96 ppm (m, 2H;

): δ=161.0 (Py-2), 156.8 (C=O), 145.5

Py-6), 144.3 (Fmoc-C-8a or C-9a), 144.1 (Fmoc-C-8a or C-9a), 141.23

2

SO

4

, filtered and

40 min: A/B (30:70)) to yield a colourless oil; yield 1% (1.5 mg,

1

0.003 mmol). R

f

=0.15 (CH/EtOAc 1:1); H NMR (400 MHz, DMSO-d

6

):

f

δ=8.25–8.16 (m, 1H; H-4), 8.07–7.99 (m, 1H; H-7), 7.97–7.82 (m, 1H; T-4),

7.40–7.31 (m, 1H; H-5), 7.29–7.18 (m, 1H; T-3), 5.32–5.25 (st-m, 2H;

(

6

J

4

J

OCH

2

), 3.66–3.47 (m, 4H; (CH

2

)

2

), 3.11–2.90 ppm (m, 6H; NCH

3

);

1

³C NMR (101 MHz, DMSO-d

(

7

6

): δ=155.8* (CH

2

OC=O), 153.9* (P-OC=O),

151.1 (C-6), 150.8* (T-5), 149.1* (C-3a), 148.9 (T-2), 137.0 (C-2), 136.1

(C-7a), 129.5–129.3 (st-m; T-4), 127.8–127.5 (st-m; T-3), 124.9 (C-4),

123.3 (C-5), 115.7–115.4 (st-m; C-7), 113.3 (CN), 61.1 (OCH ), 46.7–45.8

2

4

.25 (m, 2H; NCH

2

3

2

(st-m; CH

115.2 (100) [N(CH

(ESI): m/z calcd for C19

2

CH

2

), 34.7 (NCH

3

), 34.6 ppm (NCH

3

); LRMS (APCI): m/z (%)

Cy-3); ¹³C NMR (101 MHz, DMSO-d

)CH CH

N(CH

)C=O+H] , 476.4 (50) [M+H] ; HRMS

+

6

3

2

3

+

(

18

H N

O S

5 6 2

: 476.0693 [M+H] ; found: 476.0690;

Fmoc-C-4a or C-4b), 141.16 (Fmoc-C-4a or C-4b), 141.1 (P-1), 136.5

Py-4), 128.6 (P-3 and P-5), 128.0 (Fmoc-C-1 or C-8), 127.9 (Fmoc-C-1 or

HPLC retention time 22.239 min, 97.4%.

C-8), 127.43 (Fmoc-C-4 or C-5), 127.37 (Fmoc-C-4 or C-5), 126.2 (P-2,

P-4 and P-6), 125.3 (Fmoc-C-3 and C-6), 120.6 (Py-3), 120.45 (Fmoc-C-2

Synthesis of nitroaromatic alcohols. For the synthesis of alcohols 6-8,

the corresponding acid or aldehyde was reduced as described below. 7

_w_a_s_p_r_e_p_a_r_e_d_a_c_c_o_r_d_i_n_g_t_o_p_a_t_e_n_t_l_i_t_e_r_a_t_u_r_e_.[64] A five-step synthesis

or C-7), 120.40 (Fmoc-C-2 or C-7), 117.4 (Py-5), 67.1 (OCH

2

), 53.5

1

4

ChemBioChem

10.1002/cbic.202000138

FULL PAPER

1

scheme for 9 has been previously described by O’Connor et al that was

slightly modified as described in the supplemental (Scheme S1).^[⁶⁵^]Alcohol

(CH

2

Cl

2

/MeOH 95:5); H NMR (400 MHz, DMSO-d

6

): δ=8.02 (s, 1H), 5.69

(t, JH,H=5.3 Hz, 1H), 4.57 (d, ³JH,H=5.3 Hz, 2H), 3.91 ppm (s, 3H); ¹³C NMR

3

10 was synthesised by microwave-assisted addition of paraformaldehyde

6

(101 MHz, DMSO-d ): δ=152.3, 139.2, 131.5, 56.0, 33.3 ppm; GC-MS

to 1-methyl-5-nitroimidazole.

retention time 10.200 min; m/z (%) 157 (100) [M].

(

2-Fluoro-4-nitrophenyl)methanol 6. 2-Fluoro-4-nitrobenzoic acid (592 mg,

PAINS analysis. Prodrugs 1b – 1g and negative controls 2c and 2f were

3

1

.20 mmol, 1.0 eq) was dissolved in THF and PyBOP (1.83 g, 3.52 mmol,

.1 eq) and DIPEA (653 µL, 3.84 mmol, 1.2 eq) were added to form an

tested for known classes of assay interference compounds with the

publicly

available

online

tool

“False

Positive

Remover”

[

88]

active ester. The resulting yellow solution was stirred 1 h at rt and then

NaBH (133 mg, 3.52 mmol, 1.1 eq) was added. The yellow colour

(www.cbligand.org/PAINS/login.php).

were flagged as PAINS.

None of the tested compounds

4

disappeared and after 23 h, the solution was concentrated in vacuo. The

residue was dissolved in EtOAc (40 mL) and treated with 1M HCl aq.

Biological Evaluation.

(

1

20 mL). After 15 min of sonication, the organic layer was washed with

M HCl aq. (2 × 15 mL), sat. NaHCO3 aq. (2 × 15 mL), and brine (10 mL),

dried over Na SO , filtered and evaporated yielding the alcohol as a white

solid; yield 75% (408 mg, 2.39 mmol). R

=0.28 (CH/EtOAc 7:3); ¹H NMR

400 MHz, DMSO-d

): δ=8.12 (ddd, ³JH,H=8.5, ⁴JH,H=2.2, ⁵JH,H=0.6 Hz, 1H),

H,H=10.0, ⁴

H,H=2.2 Hz, 1H), 7.79–7.73 (m, 1H), 5.62 (t,

H,H=5.7 Hz, 1H), 4.66 ppm (d, ³JH,H=5.7 Hz, 2H); ¹³C NMR (101 MHz,

Peroxidase based LSD1 assay. Determination of enzyme activity and

inhibition was performed in an established HRP-coupled assay system

based on the Amplex Red protocol from Invitrogen (BPS Bioscience). The

2

4

f

assay was conducted in

a

white OptiPlate-384 microtiter plate

(

8

3

6

(

PerkinElmer) using a 45 mM HEPES buffer at pH 8.5 containing 40 mM

.06 (dd, ³

J

NaCl. For pretesting experiments, 0.01% Tween20 was added to the buffer

in order to avoid precipitation at higher compound concentrations. LSD1

enzyme (8 µL; final concentration 0.045 µg/µL; expressed in Sf9 cells as

_p_u_b_l_i_s_h_e_d_e_l_s_e_w_h_e_r_e_[[89]]), was incubated with inhibitor solutions of varying

concentration (2 µL in DMSO; final DMSO concentration 10%) for 20 min

at rt. Demethylation reaction was initiated by the addition of

H3K4(me2)aa1-20 (10 µL; final concentration 20 µM; sequence:

ARTK(me2)QTARKSTGGKAPRKQL; from Peptide Specialty Laboratories

GmbH). As control for 0% LSD1 activity, buffer solution was added instead

of peptide solution, whereas for 100% reference value, DMSO was used

without inhibitor. After the plate was incubated for 60 min at rt, the Amplex

Red reagent/ Horseradish Peroxidase (HRP) mixture (20 µL; final

1

3

DMSO-d

6

): δ=159.0 (d, JC,F=249 Hz), 147.6 (d, JC,F=9.02 Hz), 137.9 (d,

²JC,F=15.4 Hz), 129.7 (d, JC,F=5.54 Hz), 120.0 (d, ⁴JC,F=3.07 Hz), 110.9 (d,

3

²JC,F=27.0 Hz), 56.9 ppm (d, ³

DMSO-d ): δ= -116.02 -116.09 ppm (m); GC-MS: retention time

.716 min; m/z (%) 125 (100) [M – NO ], 171 (50) [M].

J

C,F=3.87 Hz); ¹⁹F NMR (376 MHz,

6

–

9

2

,2,2-Trifluoro-1-(4-nitrophenyl)ethan-1-ol 7. Yellow solid; yield 94%

=0.38 (CH/EtOAc 7:3); ¹H NMR (400 MHz,

): δ=8.38–8.21 (m, 2H), 7.80 (d, ³JH,H=8.7 Hz, 2H), 7.21 (d,

):

δ=148.6, 140.3, 128.4, 125.0, 123.6, 122.2, 71.8 ppm (q, ¹JC,F=32.4 Hz);

¹⁹F NMR (376 MHz, DMSO-d ): δ= -76.55 ppm (d, ³JH,F=7.9 Hz); GC-MS:

retention time 9.712 min; m/z (%) 152 (100) [M – CF ], 221 (15) [M].

(1.35 g, 6.10 mmol). R

f

DMSO-d

6

³JH,H=5.7 Hz, 1H), 5.51–5.39 ppm (m, 1H); ¹³C NMR (101 MHz, DMSO-d

6

™

concentration 50 μM Amplex Red reagent (Ampliflu Red, Sigma-Aldrich)

6

and 1 U/mL HRP (Sigma-Aldrich, P8125) in reaction buffer) were added.

Immediately after addition, fluorescence intensity of the forming product

resorufin was measured at λex=510 nm and λem=615 nm on a POLARstar

Optima microplate reader (BMG Labtech, Germany). Values were blank‐

corrected. Inhibition in [%] is in comparison to compound‐free DMSO

control and no‐substrate negative control. Inhibition curves were analysed

by sigmoidal curve fitting using OriginPro 2018b and IC50 values are given

as mean±SD from two independent experiments.

3

(

5-Nitrothiophen-2-yl)methanol 8. To an ice-cold solution of

5

1

-nitrothiophene-2-carboxaldehyde (1.41 g, 8.96 mmol, 1.0 eq) in

,2-dichloroethane (15 mL) was added sodium triacetoxyborohydride

(

2.28 g, 10.8 mmol, 1.2 eq). After 20 min, the ice bath was removed and

after 2.5 h stirring at rt, additional sodium triacetoxyborohydride (980 mg,

.62 mmol, 0.5 eq) was added. To drive the reaction to completion, NaBH

150 mg, 3.97 mmol, 0.44 eq) was added after 20 h and the suspension

4

(

4

In vitro analysis of prodrug activation and fragmentation. In order to

assess the activation of the prodrugs by NTR, the NADH oxidation to NAD⁺

during the enzymatic reduction is measured. The assay was performed in

a white OptiPlate-384 microtiter plate (PerkinElmer) using a 50 mM KPi

buffer at pH 8.0. The NTR (2 µL; final concentration 100 nM; NfsB(N)his

expressed in E. coli and purified adapted from Haas et al.^[⁹⁰^]) was added

to a solution of compound (8 µL; final concentration 60 µM and 5% DMSO).

After pre-incubation for 5 min at 37 °C, the kinetic measurement was

started by the addition of NADH (10 µL; final concentration 500 µM). The

decrease of fluorescence intensity was measured at λex=330 nm and

was stirred for 30 min. To quench the reaction, 5% NaHCO3 aq. (20 mL)

was added to the reaction mixture cooled in an ice bath. The mixture was

extracted with CH

washed with brine, dried over Na

2

Cl

2

(3 × 15 mL), the organic layers were combined,

SO , filtered and concentrated to afford

2

4

a brown oil that was used without further purification; yield 96% (1.37 g,

8

.64 mmol). R

f

=0.42 (CH/EtOAc 1:1); ¹H NMR (400 MHz, DMSO-d

6

):

δ=8.03 (d, JH,H=4.2 Hz, 1H), 7.08 (dt, JH,H=4.2 Hz, ⁴JH,H=1.1 Hz, 1H), 6.04

3

(

t, ³JH,H=5.8 Hz, 1H), 4.73 ppm (dd, ³JH,H=5.8 Hz, ⁴

J

H,H=1.1 Hz, 2H).

): δ=158.4, 149.4, 130.6, 123.6, 59.1 ppm;

GC-MS: retention time 10.344 min; m/z (%) 113 (100) [M-NO ], 159 (67)

1

³C NMR (101 MHz, DMSO-d

6

2

λem=460 nm on a POLARstar Optima microplate reader (BMG Labtech,

[

M].

Germany) at 37 °C and at 37 s intervals. Changes in fluorescence were

converted to changes in NADH concentration using a calibration curve for

NADH. The values were corrected by the spontaneous NADH oxidation

evaluated with the DMSO control and by NADH oxidation by the 1a

scaffold. It was assumed that two equivalents of NADH are used in the

(

1-Methyl-2-nitro-1H-imidazol-5-yl)methanol 9. Yellowish solid; yield 24%

64 mg, 0.41 mmol) over five steps. R

400 MHz, DMSO-d

_H₎_,₄_.₅₅₍_d_d_,3

1

f

=0.50 (CH

2

Cl

2

/MeOH 9:1); H NMR

₎_:_δ₌₇_.₁₂₍_d_,4_JH,H=0.6 Hz, 1H), 5.50 (t, JH,H=5.4 Hz,

H,H₌₅_.₄_,4

JH,H=0.6 Hz, 2H), 3.92 ppm (s, 3H); GC-MS:

3

6

2

reduction (ArNO to ArNHOH) and hence 120 µM NADH consumption is

1

J

equated with 100% activation of prodrugs. The activation of the prodrugs

in [%] is given as mean±SD from two independent experiments and was

calculated after 15 min of NTR reaction.

retention time 10.978 min; m/z (%) 157 (100) [M].

(

1

1-Methyl-5-nitro-1H-imidazol-2-yl)methanol

134 mg, 4.46 mmol, 2.0 eq) was suspended in H

-methyl-5-nitroimidazole (283 mg, 2.23 mmol, 1.0 eq) was added to the

10.

Paraformaldehyde

2

O (2.0 mL) and

In the following experiment, the fragmentation of activated prodrugs is

quantified by derivatisation of the formed LSD1 inhibitor 1a with FMOC-Cl

and subsequent HPLC analysis. After the kinetic measurement, an ice-

cold mixture of ACN/NaHCO3 aq. 250 mM pH 9 (4:1; 20 µL) was added to

each well and the plate was centrifuged for 1 min at 170 rcf. 40 µL of the

reaction mixtures were transferred to reaction tubes and the NTR was fully

inactivated at 60 °C for 10 min in a digital dry bath (Labnet). After the tubes

were centrifuged for 5 min at 17949 rcf, 20 µL were taken from the

microwave glass tube. The reaction was performed in a CEM Discover

Microwave (2.5 h, 30 W, 130 °C, followed by 4 h, 60 W, 140 °C). Further

addition of paraformaldehyde (1.7 eq and 2.7 eq) and subsequent

microwave reaction (each time 2.6 h, 60 W, 140 °C) resulted in the

formation of the product. The reaction mixture was concentrated in vacuo

and purified by column chromatography (3 to 6% MeOH in CH

2

Cl

2

over

10 CV) to afford a white solid; yield 13% (45 mg, 0.29 mmol). R

f

=0.26

1

5

ChemBioChem

10.1002/cbic.202000138

FULL PAPER

supernatant and transferred to a freshly prepared FMOC-Cl solution in

ACN (3 µL; final concentration 250 µM) that was stored on ice. The

reaction mixture was mixed and incubated for 30 min at rt. To convert the

remaining FMOC-Cl after incubation, an aqueous solution of glycine (1 µL;

final concentration 250 µM) was added, the tubes were mixed again. After

incubation for 20 min at rt, 20 µL of the solutions were transferred in

Rotilabo^®sample vials (0.1 mL, Roth) and analysed on an Agilent

Technologies 1260 Infinity system and a Phenomenex Kinetex 5u XB-C18

20 μl of Turbofect transfection reagent before 20 min of incubation at rt.

After the addition of 7 mL fresh medium onto the 70% confluent (in 60 cm²

plates) packaging cells, the transfection mixture was added dropwise.

Retroviral Infection and Selection. 24 h after transfection, the medium

was removed and 8 mL fresh medium appropriate for the cells aimed to

infect was added. 12 h afterwards, the first harvest of the virus and

simultaneously the first round of infection was performed. For this, the

medium was collected and filtered through the Acrodisc® 0,45 μm Syringe

Filter (PALL Life Sciences). The virus-producing cells were gently covered

in 8 mL fresh medium accordingly to the step before. The harvested and

filtered virus was added to the THP-1 cells: 200.000 cells per well were

seeded in 1 mL and covered with 2 mL virus. Infections were performed in

100Å 250 x 4.60 mm column. The fluorescence of the formed fluorophore

3

(final concentration 0-25 µM) was measured at ex=265 nm and

λ

λem=310 nm. Eluent A was water containing 0.05% TFA and eluent B was

acetonitrile containing 0.05% TFA. Linear gradient conditions were as

follows: 0–10 min: linear increase to 100% of B; 10–12 min: 100% B; 12–

15 min: A/B (50:50). This method is designated as M-FLD in the compound

6-well plates. After the addition of polybrene (Sigma-Aldrich; 4 μg/mL), the

characterisation part. Signals were corrected by values obtained by

performing the same experiment without NTR. The calibration curve was

generated by performing the described experimental procedure both with

amine 1a and with 3, resulting in nearly identical calibration curves with

cells were centrifuged for 15 min at 1500 rpm and 30 °C. This procedure

was repeated twice every 12 h so that the spin infection was performed

three times in total. Infection efficiency was tested via FACS analysis 48 h

after the third infection round (ca. 70% GFP ). Successfully infected cells

were selected via the puromycin resistance cassette on the vector

+

2

R = 1.000 (from 0.1 – 25 µM). The calibration curve of derivatised 1a was

used to calculate the concentration of fragmented prodrugs and the

obtained values were corrected by the factor “25 µM (theoretical

concentration of positive control 1a) / [calculated concentration of 1a]”. The

conversion of the prodrugs to 1a in [%] is in comparison to positive control

(

1μg/mL). The selection process was monitored via FACS analysis until

more than 99% of the cells were GFP-positive.

Luminescent measurement of NTR activity in THP1 cells. Different cell

numbers were spread as triplicates (50 µL each) in white, sterile Cellstar

9

1a as 100% control and is given as mean±SD from two independent

experiments.

6 plates (Greiner Bio). Compounds were added to a final concentration

of 25 µM in 0.5% DMSO and incubated for 60 min at 37 °C and 5% CO

2

.

Analysis of prodrug stability. The fragmentation of prodrugs in KPi

buffer and cell medium was analysed using a similar derivatisation protocol

as for the in vitro prodrug fragmentation. The prodrugs were diluted to a

After incubation, Luciferin Detection Reagent (Promega, V8921) was

complemented with 10 mM d-cysteine and 0.05% IGEPAL CA-630 and

added in equal volume (50 µL) to the cells. After incubation for 20 min at

rt, the luminescent signal was measured at an EnVision 2102 multilabel

plate reader (PerkinElmer). The blank signals from DMSO controls were

subtracted and the signals from NTR probes were analysed in correlation

to the positive signal at the respective cell numbers. The resulting

conversion rates were calculated as mean±SD from two independent

experiments.

5

0 µM solution with either sterilised 50 mM KPi buffer pH 8.0 or RPMI1640

medium supplemented with 10% (v/v) FCS, mM L-glutamine,

% Penicillin/Streptomycin with a final DMSO concentration of 10%. The

solutions were transferred to a sterile 96-well TC-Plate (Suspension, F,

Sarstedt), covered with gas-permeable sealing foil and kept under CO

atmosphere (5%) at 37 °C. After 0 h, 24 h, 3 days and 8 days, 20 µL were

removed and treated with aqueous NaHCO solution (5 µL; 250 mM,

pH 9.0) and freshly prepared FMOC-Cl in ACN (21 µL; final concentration

50 µM for KPi buffer or 1 mM for medium). After incubation for 30 min at

2

1

2

3

Colony Forming Unit assay. MethoCult H4230 (StemCell Technologies)

aliquots, prepared as described by the manufacturer, were supplemented

with FCS, IMDM (Iscove's Modified Dulbecco's Medium; StemCell

Technologies) and vehicle or inhibitor solution to a final concentration of

2

rt, aqueous glycine solution was added (2 µL; final concentration 250 µM

or 1 mM) and the mixed probe was incubated for additional 20 min at rt.

The chromatographic analysis was performed as described for the

fragmentation assay. The calibration curve of derivatised 1a was also

_.₅_%_D_M_S_O_._C_e_l_l_s_w_e_r_e_a_d_d_e_d_t_o₁₀₀_c_e_l_l_s_·_w_e_l_l-1 and aliquoted in 1.1 mL

0

aliquots in duplicates in TC dishes (O35 mm; Sarstedt). TC dishes were

placed together with a third dish (O35 mm; Greiner bio-one) containing

sterile water in a TC dish (O100 mm; Sarstedt) and incubated for 10 days

2

generated in medium (R = 0.999) and used to calculate the concentration

of fragmented prodrugs in medium. The obtained values were corrected

by the factor “20.83 µM (theoretical concentration of positive control 1a) /

in a 5% CO

microscope (Zeiss PrimoVert). Z

(x-min(x)/max(x)-min(x)) for each assay. For bystander effect

2

atmosphere at 37 °C. Colonies were counted by a light

[

calculated concentration of 1a]”.

i

-values were calculated with the equation

z

i

=

i

Cell Culture. THP1-NTR⁺cells were generated within this project and

derive from the THP1 cell line (RRID:CVCL_0006), which was a kind gift

of Prof. Lübbert from the University Hospital, Freiburg. Both cell lines were

cultivated in RPMI1640 medium supplemented with 10% (v/v) FCS,

+

wt

experiments, the two cell lines THP1-NTR and THP1 were mixed (50%,

7

+

5% THP1-NTR cells) before being added to the MethoCult media.

Flow cytometry. Cells were treated with compounds to a final DMSO

concentration of 0.1% for 72 h and stained with the antibody APC Mouse

Anti-Human CD86 (BD Bioscience, cat: 555660, lot: 8018951) and 7-AAD

2

mM L-glutamine, 1% Penicillin/Streptomycin at 37 °C in a humidified

atmosphere with 5% CO

2

.

(

BD Bioscience). Measurement was done using a CyAn (Beckmann

Coulter) with 10000 events per sample. Z-values were calculated with the

equation z = (x-min(x)/max(x)-min(x)), values from DMSO treated cells

Cell viability assay. Cells were diluted to 7∙10 cells∙mL^-¹and mixed with

compounds to a final DMSO concentration of 0.5% and seeded at 100 µL

in 96-well plates in triplicates. After 72 h incubation, the CellTiter 96®

AQueous Non-Radioactive Cell Proliferation Assay from Promega was

performed according to the manufacturer’s instructions. Assay plates were

measured at 492 nm on a POLARstar Optima microplate reader (BMG

Labtech, Germany).

4

i

were set as min(x) and values from cells treated with 500 nM 1a were set

as max(x). For bystander effect experiments, the two cell lines THP1-NTR

_a_n_d_T_H_P₁wt _w_e_r_e_m_i_x_e_d₍₂₅_%_,₅₀_%_,₇₅_%_T_H_P₁_-_N_T_R+ cells) before

compounds were added.

+

5

-1

GSH assay. Cells were diluted to 7·10 cells·mL , mixed with compounds

to a final DMSO concentration of 0.5% and incubated for 1 h. Cell lysates

were obtained by repeating freeze/thaw cycles and centrifugation for

Transfection. The retroviral plasmid was cloned with the nfsb template

pGL4.26-SS-352, which was a gift from James Collins (Addgene plasmid

#

68791).^[⁹¹^]For the production of retroviruses, plasmid DNA was

20 min at 4 °C with 16060 rcf. Protein concentration of different samples

introduced into packaging HEK 293GP cells using Turbofect (Thermo

Fisher Scientific). 10 μg DNA (8 μg NTR plasmid + 2 μg VSV-G) were

mixed with 1000 μl of Opti-MEM I Reduced Serum Medium (Gibco) and

was normalised after the Pierce BCA Protein Assay Kit (Thermo Scientific),

performed according to the manufacturer’s instructions. GSH-levels were

quantified with Ellman’s reagent (5,5'-dithiobis-(2-nitrobenzoic acid)) as

1

6

ChemBioChem

10.1002/cbic.202000138

FULL PAPER

described by Hulsman et al.^[⁷⁹^]Absorbance was measured at 430 nm on

a POLARstar Optima microplate reader (BMG Labtech, Germany).

Searle, C. J. Wrighton, E. I. Hyde, J. Biol. Chem. 2005, 280, 13256–

1

3264.

S. Zenno, H. Koike, M. Tanokura, K. Saigo, J. Biochem. 1996, 120,

36–744.

[

20]

21]

22]

23]

24]

7

Acknowledgements

A. B. Mauger, H. H. Somani, F. Friedlos, R. J. Knox, P. J. Burke, J.

Med. Chem. 1994, 37, 3452–3458.

This study was supported by the Deutsche Forschungs-

gemeinschaft (DFG, German Research Foundation) under

Germany's Excellence Strategy (CIBSS – EXC-2189 – Project ID

M. P. Hay, W. R. Wilson, W. A. Denny, Bioorg. Med. Chem. Lett.

1

995, 5, 2829–2834.

M. Tercel, W. A. Denny, W. R. Wilson, Bioorg. Med. Chem. Lett.

996, 6, 2741–2744.

3

9

90939984). M.J and R.S. received funding from the DFG (SFB

92). R.S. was further supported by grants from the European

1

M. Lee, J. E. Simpson, S. Woo, C. Kaenzig, G. M. Anlezark, E. Eno-

Research Council (AdGrant 322844) and the DFG (CRC 850,

CRC 746, and SCHU688/12-1).

Amooquaye, P. J. Burkeb, Bioorg. Med. Chem. Lett. 1997, 7, 1065–

1070.

[

25]

26]

27]

28]

M. P. Hay, G. J. Atwell, W. R. Wilson, S. M. Pullen, W. A. Denny, J.

Med. Chem. 2003, 46, 2456–2466.

Keywords: chemistry • drug design • inhibitors • luminescence

medicinal • oxidoreductases • prodrugs

M. P. Hay, R. F. Anderson, D. M. Ferry, W. R. Wilson, W. A. Denny,

J. Med. Chem. 2003, 46, 5533–5545.

[

1]

2]

L. F. Tietze, K. Schmuck, Curr. Pharm. Des. 2011, 17, 3527–3547.

Y. M. Choi-Sledeski, C. G. Wermuth, Pract. Med. Chem. Fourth Ed.

M. P. Hay, B. M. Sykes, W. A. Denny, W. R. Wilson, Bioorg. Med.

Chem. Lett. 1999, 9, 2237–2242.

2015, 657–696.

T. D. Gruber, C. Krishnamurthy, J. B. Grimm, M. R. Tadross, L. M.

Wysocki, Z. J. Gartner, L. D. Lavis, ACS Chem. Biol. 2018, 13,

[

3]

4]

K. B. Daniel, E. D. Sullivan, Y. Chen, J. C. Chan, P. A. Jennings, C.

A. Fierke, S. M. Cohen, J. Med. Chem. 2015, 58, 4812–4821.

R. Delatouche, I. Denis, M. Grinda, F. El Bahhaj, E. Baucher, F.

Collette, V. Héroguez, M. Grégoire, C. Blanquart, P. Bertrand, Eur.

J. Pharm. Biopharm. 2013, 85, 862–872.

2888–2896.

[

29]

30]

31]

32]

T. Jiang, P. Kumar, W. Huang, W. Kao, A. O. Thompson, F. M.

Camarda, S. T. Laughlin, ChemBioChem 2019, 20, 2222–2226.

H. P. Mohammad, O. Barbash, C. L. Creasy, Nat. Med. 2019, 25,

[

5]

6]

E. D. D. Calder, A. Skwarska, I. N. Mistry, D. Sneddon, S. J.

Conway, E. M. Hammond, ChemRxiv Prepr. 2019, DOI

https://doi.org/10.26434/chemrxiv.9963503.v1.

4

03–418.

S. Xiao Pfister, A. Ashworth, Nat. Rev. Drug Discov. 2017, 16, 241–

63.

2

M. Engel, Y. S. Gee, D. Cross, A. Maccarone, B. Heng, A. Hulme,

G. Smith, G. J. Guillemin, B. W. Stringer, C. J. T. Hyland, et al., J.

Neurochem. 2019, 149, 535–550.

T. Somervaille, O. Salamero, P. Montesinos, C. Willekens, J. A.

Perez Simon, A. Pigneux, C. Recher, R. Popat, C. Molinero, C.

Mascaro, et al., Blood 2016, 128, 4060.

[

7]

S. Zheng, S. Guo, Q. Zhong, C. Zhang, J. Liu, L. Yang, Q. Zhang,

G. Wang, ACS Med. Chem. Lett. 2018, 9, 149–154.

[

33]

34]

C. Stresemann, F. Lyko, Int. J. Cancer 2008, 123, 8–13.

Y. Shi, F. Lan, C. Matson, P. Mulligan, J. R. Whetstine, P. A. Cole,

R. A. Casero, Y. Shi, Cell 2004, 119, 941–953.

8]

Y. Ota, Y. Itoh, A. Kaise, K. Ohta, Y. Endo, M. Masuda, Y. Sowa, T.

Sakai, T. Suzuki, Angew. Chem. Int. Ed. 2016, 55, 16115–16118.

Y. Ota, A. Nakamura, E. E. Elboray, Y. Itoh, T. Suzuki, Chem.

Pharm. Bull. 2019, 67, 192–195.

[

35]

36]

37]

E. Metzger, M. Wissmann, N. Yin, J. M. Müller, R. Schneider, A. H.

F. M. Peters, T. Günther, R. Buettner, R. Schüle, Nature 2005, 437,

9]

436–439.

10]

11]

B. Rubio-Ruiz, J. T. Weiss, A. Unciti-Broceta, J. Med. Chem. 2016,

J. Huang, R. Sengupta, A. B. Espejo, M. G. Lee, J. A. Dorsey, M.

Richter, S. Opravil, R. Shiekhattar, M. T. Bedford, T. Jenuwein, et

al., Nature 2007, 449, 105–108.

59, 9974–9980.

A. M. Pérez-López, B. Rubio-Ruiz, V. Sebastián, L. Hamilton, C.

Adam, T. L. Bray, S. Irusta, P. M. Brennan, G. Lloyd-Jones, D.

Sieger, et al., Angew. Chem. Int. Ed. 2017, 56, 12548–12552.

M. Jung, Mini-Rev. Med. Chem. 2001, 1, 399–407.

J. Wang, S. Hevi, J. K. Kurash, H. Lei, F. Gay, J. Bajko, H. Su, W.

Sun, H. Chang, G. Xu, et al., Nat. Genet. 2009, 41, 125–129.

H. Kontaki, I. Talianidis, Mol. Cell 2010, 39, 152–160.

[

12]

13]

[

38]

39]

Y. Singh, M. Palombo, P. J. Sinko, Curr. Med. Chem. 2008, 15,

A. Adamo, B. Sesé, S. Boue, J. Castaño, I. Paramonov, M. J.

Barrero, J. C. Izpisua Belmonte, Nat. Cell Biol. 2011, 13, 652–659.

C. T. Foster, O. M. Dovey, L. Lezina, J. L. Luo, T. W. Gant, N.

Barlev, A. Bradley, S. M. Cowley, Mol. Cell. Biol. 2010, 30, 4851–

1802–1826.

[

14]

15]

S. Eda, I. Nasibullin, K. Vong, N. Kudo, M. Yoshida, A.

Kurbangalieva, K. Tanaka, Nat. Catal. 2019, 2, 780–792.

K. Tsubokura, K. K. H. Vong, A. R. Pradipta, A. Ogura, S. Urano, T.

Tahara, S. Nozaki, H. Onoe, Y. Nakao, R. Sibgatullina, et al.,

Angew. Chem. Int. Ed. 2017, 56, 3579–3584.

[

40]

4

863.

S. Saleque, J. Kim, H. M. Rooke, S. H. Orkin, Mol. Cell 2007, 27,

62–572.

[

41]

42]

43]

44]

45]

5

[

16]

E. M. Williams, R. F. Little, A. M. Mowday, M. H. Rich, J. V. E.

Chan-Hyams, J. N. Copp, J. B. Smaill, A. V. Patterson, D. F.

Ackerley, Biochem. J. 2015, 471, 131–153.

M. A. Kerenyi, Z. Shao, Y.-J. Hsu, G. Guo, S. Luc, K. O’Brien, Y.

Fujiwara, C. Peng, M. Nguyen, S. H. Orkin, Elife 2013, 2, e00633.

J. Wu, L. Hu, Y. Du, F. Kong, Y. Pan, Onco. Targets. Ther. 2015, 8,

[

17]

18]

19]

G. M. Anlezark, R. G. Melton, R. F. Sherwood, B. Coles, F. Friedlos,

R. J. Knox, Biochem. Pharmacol. 1992, 44, 2289–2295.

N. P. Michael, J. K. Brehm, G. M. Anlezark, N. P. Minton, FEMS

Microbiol. Lett. 1994, 124, 195–202.

2565.

T. Maes, C. Mascaró, A. Ortega, S. Lunardi, F. Ciceri, T. C. P

Somervaille, C. Buesa, Epigenomics 2015, 7, 609–626.

J. P. McGrath, K. E. Williamson, S. Balasubramanian, S. Odate, S.

P. R. Race, A. L. Lovering, R. M. Green, A. Ossor, S. A. White, P. F.

1

7

ChemBioChem

10.1002/cbic.202000138

FULL PAPER

Arora, C. Hatton, T. M. Edwards, T. O’Brien, S. Magnuson, D.

[71]

W. Zhou, D. Leippe, S. Duellman, M. Sobol, J. Vidugiriene, M.

O’Brien, J. W. Shultz, J. J. Kimball, C. Dibernardo, L. Moothart, et

al., ChemBioChem 2014, 15, 670–675.

Stokoe, et al., Cancer Res. 2016, 76, 1975–1988.

[

46]

47]

48]

Y. Ishikawa, K. Gamo, M. Yabuki, S. Takagi, K. Toyoshima, K.

Nakayama, A. Nakayama, M. Morimoto, H. Miyashita, R. Dairiki, et

al., Mol. Cancer Ther. 2017, 16, 273–284.

[72]

[73]

Z. Li, X. Gao, W. Shi, X. Li, H. Ma, Chem. Commun. 2013, 49, 5859.

D. Mustafa, D. Ma, W. Zhou, P. Meisenheimer, J. J. Cali, Bioconjug.

Chem. 2016, 27, 87–101.

M. E. Vinyard, C. Su, A. P. Siegenfeld, A. L. Waterbury, A. M.

Freedy, P. M. Gosavi, Y. Park, E. E. Kwan, B. D. Senzer, J. G.

Doench, et al., Nat. Chem. Biol. 2019, 15, 529–539.

[74]

A. Maiques-Diaz, G. J. Spencer, J. T. Lynch, F. Ciceri, E. L.

Williams, F. M. R. Amaral, D. H. Wiseman, W. J. Harris, Y. Li, S.

Sahoo, et al., Cell Rep. 2018, 22, 3641–3659.

S. Takagi, Y. Ishikawa, A. Mizutani, S. Iwasaki, S. Matsumoto, Y.

Kamada, T. Nomura, K. Nakamura, Cancer Res. 2017, 77, 4652–

[75]

[76]

J. T. Lynch, M. J. Cockerill, J. R. Hitchin, D. H. Wiseman, T. C. P.

Somervaille, Anal. Biochem. 2013, 442, 104–106.

4662.

[

49]

50]

51]

T. E. McAllister, K. S. England, R. J. Hopkinson, P. E. Brennan, A.

Kawamura, C. J. Schofield, J. Med. Chem. 2016, 59, 1308–1329.

M. Yang, J. C. Culhane, L. M. Szewczuk, P. Jalili, H. L. Ball, M.

Machius, P. A. Cole, H. Yu, Biochemistry 2007, 46, 8058–8065.

S. Mimasu, T. Sengoku, S. Fukuzawa, T. Umehara, S. Yokoyama,

Biochem. Biophys. Res. Commun. 2008, 366, 15–22.

A. K. Ghosh, M. Brindisi, J. Med. Chem. 2015, 58, 2895–2940.

A. Alouane, R. Labruère, T. Le Saux, F. Schmidt, L. Jullien, Angew.

Chem. Int. Ed. 2015, 54, 7492–509.

T. Maes, C. Mascaró, I. Tirapu, A. Estiarte, F. Ciceri, S. Lunardi, N.

Guibourt, A. Perdones, M. M. P. Lufino, T. C. P. Somervaille, et al.,

Cancer Cell 2018, 0, DOI 10.1016/j.ccell.2018.02.002.

K. N. Smitheman, T. M. Severson, S. R. Rajapurkar, M. T. McCabe,

N. Karpinich, J. Foley, M. B. Pappalardi, A. Hughes, W. Halsey, E.

Thomas, et al., Haematologica 2019, 104, 1156–1167.

H. P. Mohammad, K. N. Smitheman, C. D. Kamat, D. Soong, K. E.

Federowicz, G. S. Van Aller, J. L. Schneck, J. D. Carson, Y. Liu, M.

Butticello, et al., Cancer Cell 2015, 28, 57–69.

[77]

[78]

[79]

[

52]

53]

[

54]

55]

56]

57]

58]

59]

60]

61]

62]

S. Papot, I. Tranoy, F. Tillequin, J. Florent, J. Gesson, Curr. Med.

Chem. Agents 2002, 2, 155–185.

N. Hulsman, J. P. Medema, C. Bos, A. Jongejan, R. Leurs, M. J.

Smit, I. J. P. De Esch, D. Richel, M. Wijtmans, J. Med. Chem. 2007,

50, 2424–2431.

N. Guibourt, A. Ortega Munoz, J. Castro-Palomino Laria,

WO2010084160, 2010.

[80]

[81]

Y. Higuchi, J. Cell. Mol. Med. 2004, 8, 455–464.

M. P. Hay, W. R. Wilson, W. A. Denny, Bioorg. Med. Chem. 2005,

M. P. Hay, B. M. Sykes, W. A. Denny, C. J. O’Connor, J. Chem.

Soc. Perk. T. 1 1999, 2759–2770.

13, 4043–4055.

P. L. Carl, P. K. Chakravarty, J. A. Katzenellenbogen, J. Med.

Chem. 1981, 24, 479–480.

[82]

[83]

[84]

H. Nivinskas, R. L. Koder, Ž. Anusevičius, J. Šarlauskas, A.-F.

Miller, N. Čenas, Arch. Biochem. Biophys. 2001, 385, 170–178.

L. R. Bérubé, S. Farah, R. A. McClelland, A. M. Rauth, Int. J.

Radiat. Oncol. 1992, 22, 817–820.

M. P. Hay, W. R. Wilson, W. a. Denny, Tetrahedron 2000, 56, 645–

6

57.

K. Sinhababu, D. R. Thakker, Adv. Drug Deliv. Rev. 1996, 19, 241–

73.

J. A. Roacho-Perez, H. L. Gallardo-Blanco, M. Sanchez-

Dominguez, P. Garcia-Casillas, C. Chapa-Gonzalez, C. N.

Sanchez-Dominguez, Mol. Med. Rep. 2018, 17, 1413–1420.

R. Rai, S. Alwani, I. Badea, Polymers (Basel). 2019, 11, 745.

J. Da Veiga Moreira, M. Hamraz, M. Abolhassani, E. Bigan, S.

Pérès, L. Paulevé, M. L. Nogueira, J. M. Steyaert, L. Schwartz,

Metabolites 2016, 6, 33.

2

R. F. Alvaro, P. G. Wislocki, G. T. Miwa, A. Y. H. Li, Chem. Biol.

Interact. 1992, 82, 21–30.

[85]

[86]

R. F. Borch, J. Liu, J. P. Schmidt, J. T. Marakovits, C. Joswig, J. J.

Gipp, R. T. Mulcah§, J. Med. Chem. 2000, 43, 2258–2265.

J. Schulz-Fincke, M. Hau, J. Barth, D. Robaa, D. Willmann, A.

Kürner, J. Haas, G. Greve, T. Haydn, S. Fulda, et al., Eur. J. Med.

Chem. 2018, 144, 52–67.

[87]

R. J. Knox, F. Friedlos, M. Jarman, L. C. Davies, P. Goddard, G. M.

Anlezark, R. G. Melton, R. F. Sherwood, Biochem. Pharmacol.

1995, 49, 1641–1647.

[

63]

C. Binda, S. Valente, M. Romanenghi, S. Pilotto, R. Cirilli, A.

Karytinos, G. Ciossani, O. A. Botrugno, F. Forneris, M. Tardugno, et

al., J. Am. Chem. Soc. 2010, 132, 6827–6833.

[88]

[89]

J. B. Baell, G. A. Holloway, J. Med. Chem. 2010, 53, 2719–2740.

D. Willmann, S. Lim, S. Wetzel, E. Metzger, A. Jandausch, W. Wilk,

M. Jung, I. Forne, A. Imhof, A. Janzer, et al., Int. J. Cancer 2012,

131, 2704–2709.

[

64]

65]

V. L. Narayanan, E. A. Sausville, G. Kaur, R. K. Varma,

WO9943636 (A2), 1999.

L. J. O’Connor, C. Cazares-Körner, J. Saha, C. N. G. Evans, M. R.

L. Stratford, E. M. Hammond, S. J. Conway, Nat. Protoc. 2016, 11,

[90]

[91]

J. Haas, M. A. Schätzle, S. M. Husain, J. Schulz-Fincke, M. Jung,

W. Hummel, M. Müller, S. Lüdeke, Chem. Commun. 2016, 52,

5198–5201.

781–794.

[

66]

S. Hameed P, N. Bharatham, N. Katagihallimath, S. Sharma, R.

Nandishaiah, A. P. Shanbhag, T. Thomas, R. Narjari, M. Sarma, P.

Bhowmik, et al., Sci. Rep. 2018, 8, 1–18.

S. Slomovic, J. J. Collins, Nat. Methods 2015, 12, 1085–1090.

[

67]

68]

B. Puente, M. A. Garcia, E. Hernandez, M. A. Bregante, S. Perez, L.

Pablo, E. Prieto, J. Chromatogr. Sci. 2011, 49, 745–752.

Y. Yang, A. Voak, S. R. Wilkinson, L. Hu, Bioorg. Med. Chem. Lett.

2012, 22, 6583–6586.

[

69]

70]

R. B. P. Elmes, Chem. Commun. 2016, 52, 8935–8956.

P. Feng, H. Zhang, Q. Deng, W. Liu, L. Yang, G. Li, G. Chen, L. Du,

B. Ke, M. Li, Anal. Chem. 2016, 88, 5610–5614.

1

8

ChemBioChem

10.1002/cbic.202000138

FULL PAPER

Entry for the Table of Contents

Site-specific targeting: Nitroaromatic prodrugs were activated by the bacterial Nitroreductase NfsB (NTR) to selectively inhibit Lysine-

specific demethylase 1 (LSD1) in transfected leukemic cells. The effects of LSD1 inhibition and cytotoxicity of the side products were

distinguished by negative controls. This system may be applied in targeting approaches to reach site-specific LSD1 inhibition.

Institute and/or researcher Twitter usernames: @chemepi (Manfred Jung), @cibss (Cluster of Excellence)

1

9

Article Doi

Nitroreductase-Mediated Release of Inhibitors of Lysine-Specific Demethylase 1 (LSD1) from Prodrugs in Transfected Acute Myeloid Leukaemia Cells

DOI: 10.1002/cbic.202000138

Source and publish data:

Authors:

Article abstract of DOI:10.1002/cbic.202000138

Full text of DOI:10.1002/cbic.202000138

Products guided by the article

R&D Labs maybe for 660432-43-9

Relevant to this article

Hot Product