Bioorganic & Medicinal Chemistry 22 (2014) 2123–2132

Contents lists available at ScienceDirect

Bioorganic & Medicinal Chemistry journal homepage: www.elsevier.com/locate/bmc

Novel nitroimidazole alkylsulfonamides as hypoxic cell radiosensitisers Muriel Bonnet, Cho Rong Hong, Yongchuan Gu, Robert F. Anderson, William R. Wilson, Frederik B. Pruijn, Jingli Wang, Kevin O. Hicks, Michael P. Hay ⇑ Auckland Cancer Society Research Centre, University of Auckland, Private Bag 92019, Auckland 1142, New Zealand

a r t i c l e

i n f o

Article history: Received 13 December 2013 Revised 7 February 2014 Accepted 18 February 2014 Available online 3 March 2014 Keywords: Nitroimidazole Radiosensitiser Sulfonamide Hypoxia

a b s t r a c t A novel class of nitroimidazole alkylsulfonamides have been prepared and evaluated as hypoxia-selective cytotoxins and radiosensitisers. The sulfonamide side chain markedly influences the physicochemical properties of the analogues: lowering aqueous solubility and raising the electron affinity of the nitroimidazole group. The addition of hydroxyl or basic amine groups increased aqueous solubility, with charged amine groups contributing to increased electron affinity. The analogues covered the range of electron affinity for effective radiosensitisation with one-electron reduction potentials ranging from 503 to 342 mV. Cytotoxicity under normoxia or anoxia against a panel of human tumour cell lines was determined using a proliferation assay. 2-Nitroimidazole sulfonamides displayed significant hypoxia-selective cytotoxicity (6 to 64-fold), while 4- and 5-nitroimidazole analogues did not display hypoxia-selective cytotoxicity. All analogues sensitised anoxic HCT-116 human colorectal cells to radiation at non-toxic concentrations. 2-Nitroimidazole analogues provided modest sensitisation due to the relatively low concentrations used while several 5-nitroimidazole analogues provided equivalent sensitisation to misonidazole and etanidazole at similar molar concentrations. Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction Fractionated radiotherapy (FRT) is one of the main treatments for cancer patients with over 50% of patients receiving FRT (typically 60–70 Gy delivered in 2–3 Gy fractions over 7–8 weeks), mostly with curative intent.1 FRT is a difficult regimen for both health service delivery and patient compliance and often normal tissue toxicity precludes delivery of sufficient doses of radiation to achieve local tumour control. Stereotactic body radiotherapy (SBRT) uses hypofractionated (1–5 doses) high dose (25–60 Gy total dose) radiation to treat tumours.2 This new approach leverages recent advances in the accuracy and precision of radiation delivery to allow dose intensification to small tumours while minimising the effects to adjacent normal tissue. Clinical trials using SBRT to treat various solid tumours have demonstrated comparable control, toxicity and efficacy profiles to FRT.2 The reduced treatment time and number of patient visits, combined with emerging poten-

Abbreviations: DCM, dichloromethane; E(1), one-electron reduction potential; FCS, foetal calf serum; FRT, fractionated radiotherapy; HCR, hypoxic cytotoxicity ratio; NCS, N-chlorosuccinimide; SAR, structure activity relationship; SBRT, stereotactic body radiotherapy. ⇑ Corresponding author. Tel.: +64 649 923 1190; fax: +64 949 373 7502. E-mail address: [email protected] (M.P. Hay). http://dx.doi.org/10.1016/j.bmc.2014.02.039 0968-0896/Ó 2014 Elsevier Ltd. All rights reserved.

tial to replace surgery with an outpatient procedure, indicates substantial health, social and economic advantages for SBRT and is driving increasing use of SBRT for treating cancer. However, evidence is emerging that SBRT accentuates the role of hypoxia in radioresistance.3,4 Hypoxia, which is a consequence of the inefficient vascularisation of tumours, contributes to altered tumour metabolism,5 invasion,6 and metastasis,7 and is associated with poor prognosis and resistance to therapeutic agents.8 Hypoxia is prevalent in a wide range of solid tumours9,10 and patients with hypoxic tumours have significantly poorer outcomes than those with lower levels of hypoxia.11,12 The significance of hypoxia in resistance to cytotoxic therapy has renewed interest in targeting these cells.13 Hypoxic cells are less sensitive to radiation-induced DNA breakage, because oxygen is not available to oxidise radiation-induced DNA radicals to generate strand breaks.14 One approach to increase radiation response in hypoxic tumours is to use nitroimidazole radiosensitisers.14 These radiosensitisers are relatively non-toxic molecules which react with radiation-induced DNA radicals and cause DNA strand breaks analogously to oxygen. They have to be present at the time of irradiation and are mechanistically distinct from hypoxia-selective cytotoxins which are enzymatically reduced under hypoxia to generate a cytotoxic moiety.13 Misonidazole (1) (Fig. 1) was extensively trialled with FRT in the clinic and despite indications of

2124

M. Bonnet et al. / Bioorg. Med. Chem. 22 (2014) 2123–2132

NO2

NO2 N

N

N

OMe

N

OH

NO2

H N

N

OH

O

N

N

NO2 4

3

2

N

N

OH

O

OH 1

O OH

Figure 1. Clinically evaluated nitroimidazole radiosensitisers.

clinical benefit in some trials, failed to provide significant improvement on radiotherapy alone, with delayed peripheral neuropathy limiting treatment.15 Attempts to design more polar analogues16 with increased systemic clearance to minimise neurotoxicity17 were only partially successful with etanidazole (2) failing to provide benefit in head & neck cancer.18,19 A similar approach with the more polar doranidazole (3)20 is currently being clinically evaluated. Although a recent meta-analysis has confirmed the clinical activity of nitroimidazole radiosensitisers,21 only nimorazole (4) is in clinical use.22 Nonetheless, the recent description of a hypoxic gene signature,23 that in a retrospective study identifies head & neck cancer patients with hypoxic tumours and predicts the benefit of nimorazole in only those patients, has provided clinical validation of the use of nitroimidazole radiosensitisers with radiotherapy.24 However, existing agents have no or limited intellectual property protection which restricts options for clinical development. The advent of SBRT, combined with new approaches to identify patients with hypoxic tumours, heralds a new opportunity for the use of novel nitroimidazole radiosensitisers. We have addressed this opportunity and identified a class of nitroimidazole that incorporates an alkylsulfonamide moiety as a key element, with the objective of developing improved, novel radiosensitisers for SBRT. Here, we report the synthesis of a pilot set of compounds and in vitro appraisal of their potential as hypoxia-selective radiosensitisers. We also evaluate their hypoxia-selective cytotoxicity in culture as a measure of their sensitivity to bioreductive activation.

and a strong modulating effect on the electron affinity of the nitroimidazole moiety. We initially explored a strategy previously used to prepare alkylaryl sulfonamides,25 generating chloromethylthioacetate in situ to alkylate 2-nitroimidazole (5), but this failed to provide the key sulfonyl chloride 9 in our hands (Scheme 1). To avoid the use of the volatile chloromethylthioacetate we attempted a stepwise strategy to prepare 9. Alkylation of 5 with bromochloromethane under a variety of conditions was explored; however formation of the dimer 8 predominated under most conditions. Optimum results were obtained using Cs2CO3 in the presence of an excess (20 equiv) of bromochloromethane giving chloride 6 which was condensed with potassium thioacetate to give thioester 7 in modest yield. Oxidation of 7 with NCS, gave the unstable sulfonyl chloride 9. Efforts to isolate 9 resulted in significant loss of material and so 9 was used directly in subsequent reactions. Reaction of crude sulfonyl chloride 9 with 2-methoxyethylamine gave sulfonamide 10 in low yield. In an effort to avoid the unstable intermediates 7 and 9 we prepared the sulfonate salt 11 and sulfonic acid 12. Attempts to prepare sulfonamides from either 11 or 12 with an array of coupling reagents (oxalyl chloride, EDCI, HOBt, HBTU, HATU) using multiple reaction conditions failed to provide positive results. We explored an alternative approach to 9 by condensing chloromethanesulfonyl chloride with 2-methoxyethylamine to give sulfonamide 13 which was used to alkylate 2-nitroimidazole. Although only providing a similar yield to the oxidative route, this approach was more direct and also provided access to the isomeric analogues 15 and 16, albeit in low yield. The direct oxidative approach provided low yields of 17, reflecting losses of material in the aqueous workup, however use of a protected ethanolamine 18 gave no overall improvement in yield of 17 (Scheme 2). An alternative alkylation strategy using the protected chlorosulfonamide 19 provided a modest improvement in the yield of 20, and subsequently 17.

2. Results 2.1. Chemistry Our synthetic strategy was directed towards nitroimidazoles bearing an alkyl sulfonamide as a moiety conferring both novelty

NO2

NO2

O

(a), (b) N

N

NH 5

NO2

(c) N

N

N

S

NO2

(e)

Me

N

7

(d) NO2

Cl N

6

+

SO 2Cl

N

NO2 N

N

9

N

8

(g)

(f)

NO2 N

N

R S O O

11 R = O- Na + 12 R = OH (f) Cl

SO2 Cl

Cl

H N S O O 13

(h) NO2

(i) OMe

N

N

N

N

H N S O O 10

(i) N

NH N

O2N

N

S O

14 O2N

H N

OMe

+

S

O

O

15

NO2

H N

OMe

OMe

O 16

Scheme 1. Reagents and conditions: (a) ClCH2Br, potassium thioacetate, MeCN; (b) iPr2NEt, NaI, MeCN; (c) ClCH2Br, Cs2CO3, DMF, 45%; (d) potassium thioacetate, DMF, 42%; (e) NCS, 2 M aq HCl, DCM; (f) MeOCH2CH2NH2, Et3N, DCM, 16% from 6; (g) Na2SO3, acetone/H2O; (h) DOWEX50 WX8, water, 87%; (i) 5 (18%) or 14 (7% and 3%), Cs2CO3, DMF.

2125

M. Bonnet et al. / Bioorg. Med. Chem. 22 (2014) 2123–2132

NO2 N

N

NO2 (a)

Cl

N

6

NO2

O

N

S

Me

(b)

N

NO2

N

7

SO 2Cl

(c) N

H N S O O

9

(f)

H2 N

OH

OTBDMS (e)

18

H N S O O

NO2 N

(g)

Cl

OH

17

(d)

H2 N

N

(h)

N

H N S O O

OTBDMS

20 OTBDMS 19

Scheme 2. Reagents and conditions: (a) potassium thioacetate, DMF; (b) NCS, 2 M aq HCl, DCM; (c) HOCH2CH2NH2, Et3N, DCM, 10% from 6; (d) TBDMSCl, imidazole, DCM, 82%; (e) 9, Et3N, DCM, 11%; (f) AcOH, THF, H2O, 73%; (g) ClCH2SO2Cl, Et3N, DCM, 88%; (h) 5, Cs2CO3, DMF, 35%.

Me N

Me OH

N

(a), (b) N

Me

S

N

O NO 2

NO 2

21

22 (c)

Me N

O

O S

N

Me

N H

OMe

(d)

N

24

NO 2

Me N

NO 2

S

N H

Cl

23

(f)

Me

O

O

N

S

NO 2 (e)

O

O

N

N

25

O

O

N N NO 2

S

N H

N

26

Scheme 3. Reagents and conditions: (a) MsCl, Et3N, DMAP, DCM; (b) potassium thioacetate, DMF, 70% from 21; (c) NCS, 2 M aq HCl, DCM; (d) MeOCH2CH2NH2, Et3N, DCM, 50% from 22; (e) 2-(pyrrolidin-1-yl)ethanamine, Et3N, DCM, 25% from 22; (f) 2-(piperidin-1-yl)ethanamine, Et3N, DCM, 68% from 22.

Metronidazole (21) provided a convenient starting point for the preparation of several 5-nitroimidazole analogues (Scheme 3). Mesylation and displacement with potassium thioacetate gave the thioester 22 in good yield. The sulfonyl chloride 23 was not isolated and converted directly to sulfonamides 24–26 in fair to good yields. 2.2. Physicochemical data Aqueous solubility, lipophilicity and electron affinity are three key physicochemical parameters that influence the efficacy of

nitroimidazole radiosensitisers. Typically, nitroimidazole radiosensitisers need to be present at millimolar concentrations to be effective sensitisers and so aqueous solubility is an important prerequisite. A pragmatic measurement of aqueous solubility was adopted; determining solubility by HPLC (see Section 4.1.1) for solutions of compounds diluted from a DMSO stock solution to 100 >100 6.5 23.5 >27 18.1 29.6 61.7 >51

Log Da 0.41 1.37 1.19 0.83 0.64 1.53 0.84 1.87 0.39

PSAb 93.1 113.0 127.4 127.4 127.4 138.3 127.4 121.4 121.4

E(1) mV

Reference E(1)d

c

389 388c 352 ± 10 503 ± 7 421 ± 8 342 ± 8 500± 8 475 ± 8 nd

BV MV MV MV TQ TQ

Calculated by ACDLab log D calculator v12.5; bACD/PhysChem v12; cRef.31; dReference compounds: BV, benzylviologen, MV, methylviologen, TQ, triquat.

M. Bonnet et al. / Bioorg. Med. Chem. 22 (2014) 2123–2132

5- and 4-isomers. The sulfonamide group provides a strong electron-withdrawing influence to raise the electron affinity, with the two 2-nitroimidazoles 10 and 17 having E(1) values raised by 36 and 46 mV, respectively, compared to etanidazole. Similarly, the 4-nitroimidazole 15 and the 5-nitroimidazole 16 have raised E(1) values compared to comparable nitroimidazoles (e.g., nimorazole (4) has an E(1) value of 457 mV, while the analogous 4-isomer has an E(1) of 554 mV).31 The 5-nitroimidazole 24 has a slightly reduced E(1) compared to metronidazole (21) (480 mV)31 reflecting the reduced electronic effect over the longer ethyl spacer unit, but a positively charged solubilising side chain (e.g., 25) offsets this to some extent.

100

HT29 H1299 HCT116 PC3 22RV1 10

HCR

2126

1

2.3. Biology 0.1

100 HT29 H1299 HCT116 PC3 22RV1

IC50 (mM)

10

1

2

10

15

16

17

24

25

26

Figure 3. Hypoxic selectivity as hypoxic cytotoxicity ratio [HCR = IC50(normoxic)/ IC50(anoxic)]. Cell lines: HT29 and HCT116 colorectal carcinomas, H1299 non-small cell lung carcinoma, 22Rv1 and PC3 prostate carcinomas. Error bars represent the SEM for 2–3 independent experiments.

ber.33 Since radiosensitisation by nitroimidazoles (or by oxygen) is a physicochemical process it is largely independent of cell type. HCT116 was chosen as a representative line which provides efficient colony formation. Total compound treatment time was 3 h, after which cells were trypsinized and plated out for colony formation. Colonies (>50 cells) were counted after 10 days, surviving fractions calculated and compared to control radiation survival curves produced in the same experiment without compound treatment. This approach is demonstrated for etanidazole (2) and compound 25 and indicates the derivation of the Sensitiser Enhancement Ratio (SER1% = radiation dose for 1% survival without/with compound) (Fig. 4). Compounds were tested at con-

100

10-1

Surviving Fraction

The cytotoxicity (as IC50: the compound concentration over a 4 h aerobic or anoxic exposure at 37 °C required to inhibit subsequent cell growth by 50%) was determined using a sulforhodamine B proliferation assay in a panel of human tumour cell lines (HT29 and HCT116 colorectal carcinomas, H1299 non-small cell lung carcinoma, 22Rv1 and PC3 prostate carcinomas) which reflect a range of one-electron reductase capacities.32 The anoxic IC50 values of the compounds spans a range from ca. 70 lM to ca. 24 mM (Fig.2) with the PC3 prostate cell line being the least sensitive under anoxia, presumably reflecting low one-electron reductase activity. Notably, the 2-nitroimidazoles 10 and 17 show almost a 10-fold increase in anoxic cytotoxic potency compared to misonidazole and etanidazole, reflecting their increased electron affinity. Similarly, the 5-nitroimidazoles 16 and 24–26 show increased anoxic cytotoxic potency compared to nimorazole (4) (IC50(anoxic) 9.8 ± 1.6 mM in HCT116 cells) and metronidazole (21) (IC50(anoxic) >16.7 mM in HCT116 cells), respectively. The compounds fall into two categories: those that have modest (ca. 6–64 fold) hypoxia-selective cytotoxicity [HCR = IC50(normoxic)/ IC50(anoxic)] and those that show little or no hypoxic selectivity (Fig. 3). The 2-nitroimidazole analogues 10 and 17 with higher reduction potentials (>380 mV) display hypoxic selectivity of ca. 10 to 20-fold and this is also seen for etanidazole and misonidazole. In contrast, nitroimidazoles with lower E(1) values (