Novel synthesis of physovenine and physostigmine analogs

Tetrahedron Letters 57 (2016) 3046–3049

Contents lists available at ScienceDirect

Tetrahedron Letters journal homepage: www.elsevier.com/locate/tetlet

Novel synthesis of physovenine and physostigmine analogs C. H. Wang a, S. Alluri a, G. Nikogosyan a, C. DeCarlo a, C. Monteiro a, G. Mabagos a, H. H. Feng a, A. R. White a, M. Bartolini b, V. Andrisano c, L. K. Zhang d, A. K. Ganguly a,⇑ a

Department of Biomedical Engineering, Chemistry and Biological Sciences Stevens Institute of Technology, Hoboken, NJ 07030, USA Department of Pharmacy and Biotechnology, University of Bologna, via Belmeloro 6, 40126 Bologna, Italy c Department for Life Quality Studies, University of Bologna, Corso d0 Augusto, 237, 47921 Rimini, Italy d Merck Research Laboratories, 2015 Galloping Hill Road, Kenilworth, NJ 07033-1300, USA b

a r t i c l e

i n f o

Article history: Received 28 April 2016 Revised 29 May 2016 Accepted 1 June 2016 Available online 3 June 2016

a b s t r a c t This Letter describes a versatile synthetic approach to prepare physovenine and physostigmine analogs. A series of analogs were synthesized and evaluated for cholinesterase inhibition activities, including human acetylcholinesterase (AChE) and butyrylcholinesterase (BuChE) from human serum. Ó 2016 Elsevier Ltd. All rights reserved.

Keywords: Radical reactions AChE inhibitors Physovenine and physostigmine analogs

Introduction

Present study

()-Physovenine and ()-physostigmine (Fig. 1) are known to be acetylcholinesterase (AChE) inhibitors.1a–c These alkaloids have been used in the past for the treatment of myasthenia gravis and glaucoma and more recently AChE inhibitors such as Donepezil2 and Rivastigmine3 have found use for the treatment of Alzheimer’s patients. The main challenge for the synthesis of physovenine and physostigmine analogs is the construction of quaternary carbon center. There are reported syntheses in the literature4a–e toward achieving this goal. Our approach for the synthesis of physovenine and physostigmine analogs is different and novel. The present synthesis allows us to make structurally diverse compounds for establishing structure activity relationship. The key step in our synthesis is the formation of the spiro-oxindiole ring using radical cyclization process followed by further elaboration yielding the crucial alcohol intermediate (A). In a previous publication5 we have reported the synthesis of a novel class of progesterone receptor antagonists using the intermediate (A) and in this publication we have used the same intermediate (A) to synthesize novel physostigmine and physovenine analogs represented by structure (D, wherein X = N and X = O respectively). These novel analogs were synthesized through the intermediacy of aryl radicals (B) and (C).

Substituted c-butyrolactones required for the synthesis of physostigmine and physovenine analogs were prepared as follows. Treatment6 of c-butyrolactone 1a with sodium methoxide and aromatic aldehydes gave substituted butyrolactones 2a–d (Scheme 1). However, the reaction of 1a with acetaldehyde yielded an unexpected product 2e involving the consumption of two equivalents of acetaldehyde. For the synthesis7 of 2f and 2g we treated the corresponding ketone and aldehyde with NaH and diethyl(2-oxotetrahydrofuran-3-yl)phosphonate 1c, which in turn was derived from a-bromo-c-butyrolactone 1b. Compound 1a when treated with benzyl bromide yielded 2i. It should be noted that the reaction of 1a with 2-bromopropane yielded 2h, as a result of self-condensation of c-butyrolactone 1a. Treatment of compounds 2a–h with 2-bromo-4-methoxyaniline and trimethylaluminum in toluene gave amides 3a–3h (Scheme 2), respectively. These derivatives were acetylated using acetyl chloride and pyridine in DCM to protect the hydroxyl group and then N-alkylated using Cs2CO3 and alkyl halides in DMF yielding the desired precursors 5a–5h for reductive radical cyclization. Following the above procedure radical precursors 8a–c used for oxidative cyclization were prepared from 2i–j. Compounds 5a–5g upon treatment with AIBN and TBTH in toluene solution yielded both the exo and endo cyclization derived products through the intermediacy of 9 yielding compounds 10a–g and 11a–e, respectively (Scheme 3). Surprisingly, two atropisomers (10h and 10i) were isolated from the radical reaction of 5h. High resolution mass

⇑ Corresponding author. Tel.: +1 617 4843806. E-mail address: [email protected] (A.K. Ganguly). http://dx.doi.org/10.1016/j.tetlet.2016.06.005 0040-4039/Ó 2016 Elsevier Ltd. All rights reserved.

3047

C. H. Wang et al. / Tetrahedron Letters 57 (2016) 3046–3049

N H O

O N H

O

N H

O

H

N

O

O

N O

(-)Physostigmine

(-)Physovenine

R2 N

R2 N

Donepezil

R1

OH

(A)

H3CO

R1

R4 R

(B)

R3

R1

R4

H3CO

R2 N H X

O

O R3

N

O

Rivastigmine

R2 N

O

H

O

N

N

O

O

R

(C)

R1 (D)

R

Figure 1. Examples of AChE inhibitors.

O

O

R1

O

R1

(1). Diisopropylamine, n-BuLi, THF, -10oC

R

NaOMe, MeOH, 0oC

R

O O

(2). 2-Bromopropane THF, -78oC

O

and H11 (Fig. 2). Treatment of 8a–c with TBTH and AIBN produced the desired spirooxindoles (10a, 10j–k), involving the rearrangement of 12 to the more stable radical 13. Hydrolysis of 10a–b with 2 N NaOH in methanol yielded the corresponding alcohols 14a and 14b, which were then converted to aldehydes 15a and 15b, respectively using the Dess–Martin oxidation (Scheme 4). Condensation of methyl amine with 15a followed by reduction with LiAlH4 afforded 16a. O-demethylation of 16a using BBr3 yielded the phenol 17a. Similarly, 17b was obtained from 14b. Treatment of the phenols 17a–b with NaH and substituted isocyanates yielded the desired physostigmine analogs 18a–c. Treatment of compounds 10a–k and 14a–d with LiAlH4 in THF under reflux yielded the compounds with physovenine core structures 19a–j. O-demethylation of compounds 19a–j followed by reaction with substituted isocyanates furnished the desired physovenine analogs 21a–g. The biological activities of the physostigmines (18a–c) and physovenines (21a–g) analogs are summarized in Table 1. It is evident from the IC50 values reported in Table 1 that in the physostigmine series the benzyl substituted analogs 18a, 18b, and 18c were inactive against hAChE but active against hBuChE. In the physovenine series benzyl substituted compounds 21a, 21b, and 21c were selective hBuChE inhibitors with inhibitory potencies in the submicromolar range. The most active against hBuChE being 21a with an IC50 value of 70 nM and selectivity around 58 folds. In the alkyl substituted derivatives, 21d, 21e, 21f, and 21g were slightly selective toward hAChE. Comparing compounds 21f and

O

2h (25.0%)

O 1a

2a: R=C6H5, R1=H (72.6%) 2b: R=4-Cl-C6H4, R1=H (89.6%) 2c: R=4-CH3-C6H4, R1=H (88.7%) 2d: R=4-CH3O-C6H4, R1=H (40.0%) 2e: R=CH=CHCH3, R1=H (9.4%)

(1). Diisopropylamine, n-BuLi, THF, -10oC (2). benzyl bromide THF, -78oC

O O 2i (13.0%)

O O

O O

P(OEt)3 reflux 140oC

Br 1b

O

R1

O P O O

O

R

NaH

R1

O

THF

1c

R

2f: R=CH2CH(CH3)2, R1=H (27.0%) 2g: R=R1=CH3 (70.5%)

Scheme 1. Synthesis of substituted c-butyrolactones.

spectroscopy established identical molecular compositions of both the above compounds. NMR spectra of 10h and 10i indicated that H10 in these compounds appear at d 4.14 and d 4.03 respectively. The quaternary carbon C9 in 10h appeared at d 54.55, whereas C9 in 10i appeared at d 54.22. Both spiro-oxindoles show identical correlations in COSY and HMBC. For example in HMBC, C9 shows correlations with H7, H15, and H11, whereas C1 with H10, H14, and H17. In addition, 10h showed long range NOE between H7 and H10, whereas 10i showed the correlations of H7 with both H10

H N

Br

Br NH2

2a-2h Me3Al toluene

O CH3

O R1

O CH3

pyridine, DCM

R HO

O O CH3

O

H3 C

R1 R

Me3Al toluene

R1

DMF

R

O

H N

O

R1

pyridine, DCM

R

O CH3

HO H3 C

2i: R=C6H5, R1=H 2j: R=R1=H

H N

Br O

6a: R=C6H5, R1=H (70.2%) 6b: R=R1=H (75.4%)

R1 O

Cs2CO3 R 1 R 2X DMF

O

7a: R=C6H5, R1=H (97.5%) 7b: R=R1=H (86.2%)

Scheme 2. Preparation of radical precursors.

R O

R2 N

Br O

R

O

O

5a: R=C6H5, R1=H, R2=CH3 (97.2%) 5b: R=4-Cl-C6H4, R1=H, R2=CH3 (97.7%) 5c: R=4-CH3-C6H4, R1=H, R2=CH3 (95.7%) 5d: R=4-CH3O-C6H4, R1=H, R2=CH3 (97.0%) 5e: R=CH=CHCH3, R1=H, R2=CH3 (90.5%) 5f: R=CH2CH(CH3)2, R1=H, R2=CH3 (95.4%) 5g: R=R1=CH3, R2=CH3 (97.4%) 5h: R, R1= 2-tetrahydrofuran, R2=CH3 (94.2%)

acetyl chloride O CH3

O CH3 H3 C

4a: R=C6H5, R1=H (84.3%) 4b: R=4-Cl-C6H4, R1=H (90.8%) 4c: R=4-CH3-C6H4, R1=H (98.7%) 4d: R=4-CH3O-C6H4, R1=H (90.9%) 4e: R=CH=CHCH3, R1=H (85.0%) 4f: R=CH2CH(CH3)2, R1=H (82.5%) 4g: R=R1=CH3 (88.5%) 4h: R, R1= 2-tetrahydrofuran (91.4%)

Br

Br O

O CH3

R2 N

Br

Cs2CO3 R2X

acetyl chloride

3a: R=C6H5, R1=H (85.7%) 3b: R=4-Cl-C6H4, R1=H (88.6%) 3c: R=4-CH3-C6H4, R1=H (97.3%) 3d: R=4-CH3O-C6H4, R1=H (79.8%) 3e: R=CH=CHCH3, R1=H (41.4%) 3f: R=CH2CH(CH3)2, R1=H (57.1%) 3g: R=R1=CH3 (71.2%) 3h: R, R1= 2-tetrahydrofuran (85.2%)

NH2

H N

Br

O CH3 H3 C

O R1 R

O O

8a: R=C6H5, R1=H, R2=CH3 (97.6%) 8b: R=R1=H, R2=CH2CH3 (95.5%) 8c: R=R1=H, R2=CH3 (92.5%)

3048

C. H. Wang et al. / Tetrahedron Letters 57 (2016) 3046–3049

Br

R2 N

R2 N

O

AIBN R1 TBTH

O CH3

R

O O

H3 C

R1

O CH3

toluene

R2 N

O H3C

9

R2 N

O CH3 H3C

O

R

O

R2 N

AIBN TBTH

H3 C

O

O CH3

R2 N

O

O

O

H3C

8a: R=C6H5, R1=H, R2=CH3 8b: R=R1=H, R2=CH2CH3 8c: R=R1=H, R2=CH3

O CH3

R

O

O

O R1

O

R

11a: R=C6H5, R1=H, R2=CH3 (11.2%) 11b: R=4-Cl-C6H4, R1=H, R2=CH3 (12.9%) 11c: R=4-CH3-C6H4, R1=H, R2=CH3 (28.4%) 11d: R=4-CH3O-C6H4, R1=H, R2=CH3 (18.9%) 11e R=CH2CH(CH3)2, R1=H, R2=CH3 (24.8%)

R2 N

H3C

H3 C

O O O

O R1

O

CH3

R

10a: R=C6H5, R1=H, R2=CH3 (36.4%) 10j: R=R1=H, R2=CH2CH3 (43.5%) 10k: R=R1=H, R2=CH3 (46.2%)

13

O O

CH3 O

R1

O

R

O

12

R2 N

O

R

1, 5 transfer

toluene

H3C

O

CH3

10a: R=C6H5, R1=H, R2=CH3 (60.9%) 10b: R=4-Cl-C6H4, R1=H, R2=CH3 (68.8%) 10c: R=4-CH3-C6H4, R1=H, R2=CH3 (60.2%) 10d: R=4-CH3O-C6H4, R1=H, R2=CH3 (61.9%) 10e1: R=CHCH2CH3, R1=H, R2=CH3 (24.7%) 10e2 :R=CH=CHCH3, R1=H, R2=CH3 (18.3%) 10f: R=CH2CH(CH3)2, R1=H, R2=CH3 (30.1%) 10g: R=R1=CH3, R2=CH3 (82.2%) 10h: R, R1= 2-tetrahydrofuran, R2=CH3 (33.9%) 10i: R, R1= 2-tetrahydrofuran, R2=CH3 (31.1%)

O

H3C

O R1

5a-h

Br

O

R

O

R2 N

O

R2 N

CH3

NiCl2 6H2O, NaBH4

R

H3C

CH3OH, THF (1:1)

10e1: R=CH=CHCH3, R1=H, R2=CH3 10e2: R=CHCH2CH3, R1=H, R2=CH3

O OH

O R1

R

10e=CH2CH2CH3, R1=H, R2=CH3 (93.0%) Scheme 3. Radical cyclizations.

10h

10i

Figure 2. NOE correlations of 10h and 10i.

21g (at their maximum inhibition, according to the inhibition kinetics) it is apparent that 21g was five-fold more active than the aromatic carbamate 21f against hAChE and two-fold more

active against hBuChE. In general physovenine and physostigmine analogs, 18a–c and 21a–e with alkyl substituents were more active against hAChE than the arylalkyl substituents at the quaternary carbon center 3a. Indeed, it is evident that the size of the substituent at 3a has a strong effect on the inhibitory activity against hAChE. For physovenine analogs, the inhibitory activity toward hAChE decreased by about 2.8 times upon substitution of a methyl group with a butyl group (compare 21d with 21e), and 13 times upon introduction of a benzyl group (compare 21a with 21e), while the introduction of a p-chloro-benzyl substituent led to an inactive derivative (21b). This trend can be likely ascribed to the narrow AChE’s gorge which might not allow to accommodate physovenine derivatives with bulky substituents at the quaternary carbon center 3a. Conversely, due to the larger dimensions of the BuChE catalytic gorge (approximately 340 Å3 larger than the corresponding portion of the hAChE gorge) (Zha et al.), BuChE could easily accommodate analogs 21a–c. The evident effect of the size of the substituent at 3a on the inhibitory potency toward hAChE also determined a change of the selectivity profile. In fact while physovenine analogs 21e–g (R = R1 = H) and 21d (R = CH2CH2CH3; R1 = H) turned out to be selective hAChE inhibitors, analogs 21a–c with bulky aryl alkyl substituent at 3a were highly selective BuChE inhibitors. In conclusion, the most potent among the alkyl substituted compounds was 21g with an IC50 value of 53 nM against hAChE and selectivity of 18 fold in activities between the two enzymes. We believe the present work helps to establish the trend in SAR among physostigmine and physovenine analogs for activity against hAChE and hBuChE.

3049

C. H. Wang et al. / Tetrahedron Letters 57 (2016) 3046–3049 R2 N

2N NaOH 10a-k

H3 C

DCM:CH3OH 9:1

H3C

O R

BBr3 DCM

20a: R=C6H5, R1=H, R2=CH3 (78.9%) 20b: R=4-Cl-C6H4, R1=H, R2=CH3 (90.8%) 20c: R=4-CH3-C6H4, R1=H, R2=CH3 (64.5%) 20d: R=CH2CH2CH3, R1=H, R2=CH3 (94.6%) 20e: R=R1=CH3, R2=CH3 (66.6%) 20f: R=R1=H, R2=CH3 (86.0%) 20g: R=R1=H, R2=CH2CH3 (93.3%) HO

TEA, MgSO4, THF

O O

O

19a: R=C6H5, R1=H, R2=CH3 (70.4%) 19b: R=4-Cl-C6H4, R1=H, R2=CH3 (83.9%) O 19c: R=4-CH3-C6H4, R1=H, R2=CH3 (86.2%) 19d: R=4-CH3O-C6H4, R1=H, R2=CH3 (94.5%) R3 N O H 19e: R=CH2CH2CH3, R1=H, R2=CH3 (86.7%) 19f: R=R1=CH3, R2=CH3 (92.6%) 19g: R, R1= 2-tetrahydrofuran, R2=CH3 (95.1%) 19h: R, R1= 2-tetrahydrofuran, R2=CH3 (96.9%) 19i: R=R1=H, R2=CH2CH3 (59.2%) 19j: R=R1=H, R2=CH3 (66.0%)

R2 N H O

NaH R3 N C O

H3C

(2) LiAlH4 reflux

R3

R

N H

R2 N H N R1

N

R1 R 16a: R=C6H5, R1=H, R2=CH3 (50.2%) 16b: R=4-Cl-C6H4, R1=H, R2=CH3 (39.7%)

CH3 R N C O 3

THF R

DCM

BBr3

R2 N H

NaH

18a-c

CH3

O

N

CH3

HO R1

R

17a: R=C6H5, R1=H, R2=CH3 (80.6%) 17b: R=4-Cl-C6H4, R1=H, R2=CH3 (60.4%)

R2 N H O

O

THF R1

R2 N H

(1) CH3NH2 HCl

R1 R R 14a: R=C6H5, R1=H, R2=CH3 (90.4%) 15a: R=C6H5, R1=H, R2=CH3 (94.4%) 14b: R=4-Cl-C6H4, R1=H, R2=CH3 (85.2%) 14c: R=4-CH3-C6H4, R1=H, R2=CH3 (79.5%) 15b: R=4-Cl-C6H4, R1=H, R2=CH3 (92.5%) 14d: R=4-CH3O-C6H4, R1=H, R2=CH3 (90.6%) 14e: R=R1=H, R2=CH3 (70.7%) 14f: R=R1=H, R2=CH2CH3 (89.5%)

R2 N H O R1

H3 C

DCM

R1

LiAlH4 reflux THF

LiAlH4 THF

OH

O

R2 N

DMP

O

O R1

R

21a-g

Scheme 4. Synthesis of physovenine and physostigmine analogs.

Table 1 hAChE and hBuChE of the analogs

R2 N H 3a X

O R3

N H

O R1

Yield (%)

X

R

R1

R2

R3

IC50a hAChEb (lM) ± SEM

IC50a hBuChEb (lM) ± SEM

75.0 63.7 41.6 62.9 43.1 48.8 15.0 59.1 48.7 54.6

N-CH3 N-CH3 N-CH3 O O O O O O O

C6H5 C6H5 4-Cl-C6H4 C6H5 4-Cl-C6H4 4-CH3-C6H4 CH2CH2CH3 H H H

H H H H H H H H H H

CH3 CH3 CH3 CH3 CH3 CH3 CH3 CH3 C2H5 C2H5

C(CH3)3 3-Cl-4-CH3-C6H3 3-Cl-4-CH3-C6H3 3-Cl-4-CH3-C6H3 3-Cl-4-CH3-C6H3 3-Cl-4-CH3-C6H3 3-Cl-4-CH3-C6H3 3-Cl-4-CH3-C6H3 3-Cl-4-CH3-C6H3 Cyclohexane

Not active >30c Not active 4.04 ± 0.45 Not active 9.55 ± 0.51 0.638 ± 0.049 0.226 ± 0.022 0.361 ± 0.025 0.182 ± 0.010 0.053 ± 0.005d

4.35 ± 0.24 7.17 ± 0.54 13.7 ± 1.3 0.070 ± 0.007 0.173 ± 0.019 0.231 ± 0.018 1.20 ± 0.20 0.985 ± 0.125 1.87 ± 0.39 2.74 ± 0.11 0.959 ± 0.052e

R

18a 18b 18c 21a 21b 21c 21d 21e 21f 21g

a IC50 values represent the concentration of inhibitor required to decrease enzyme activity by 50% and are the mean of two independent measurements, each performed in duplicate. According to the inhibition kinetics, IC50 values were determined following a previously developed protocol (Zha8 et al) based on Ellman’s method (Ellman9 et al.) and a standard incubation time of 20 min, if not otherwise specified. b Human recombinant AChE and BuChE from human serum were used. c % inhibition at 30 lM = 27.2 ± 4.6. d IC50 value determined after a 120 min incubation period, according to the slower inhibition kinetics of 21g against hAChE. e IC50 value determined after a 60 min incubation period, according to the slower inhibition kinetics of 21g against hBuChE.

Acknowledgment We would like to thank Stevens Institute of Technology for the award of undergraduate fellowships to G.N., C.D., C.M., and G.M.

5.

References and notes

6. 7.

1. (a) Zhao, B.; Moochhala, S.; Tham, S. J. Chromatogr. B 2004, 183–192; (b) Crieg, N.; Pei, X. F.; Soncrant, T. T.; Ingram, D. K.; Brossi, A. Med. Res. Rev. 1995, 15, 3– 31; (c) Sano, M.; Bell, K.; Marder, K.; Stricks, L. Clin. Pharm. 1993, 61. 2. Nightingale, S. JAMA 1997, 277, 10. 3. Dhillon, S. Drugs 2011, 71, 1209–1231. 4. (a) Matsuura, T.; Overman, L.; Poon, D. J. Am. Chem. Soc. 1998, 120, 6500–6503; (b) Ashimori, A.; Matsuura, T.; Overman, L.; Poon, D. J. Org. Chem. 1993, 58,

8.

9.

6949–6951; (c) Trost, B.; Zhang, Y. J. Am. Chem. Soc. 2006, 128, 4590–4591; (d) Morales-Ríos, M.; Santos-Sánchez, N.; Joseph-Nathan, P. J. Nat. Prod. 2002, 65, 136–141; (e) Mukai, C.; Yoshida, T.; Sorimachi, M.; Odani, A. Org. Lett. 2006, 8, 83–86. Alluri, S.; Feng, H.; Livings, M.; Samp, L.; Biswas, D.; Lam, T.; Lobkovsky, E.; Ganguly, A. K. Tetrahedron Lett. 2011, 52, 3945–3948. Zimmer, H.; Rothe, J.; Holbert, J. J. Org. Chem. 1960, 25, 1234–1235. Murphy, J.; Rasheed, F.; Roome, S.; Scott, K.; Lewis, N. J. Chem. Soc., Perkin Trans. 1 1998, 2331–2340. Zha, X.; Lamba, D.; Zhang, L.; Lou, Y.; Xu, C.; Kang, D.; Chen, L.; Xu, Y.; De Simone, A.; Samez, S.; Pesaresi, A.; Stojan, J.; Lopez, M. G.; Egea, J.; Andrisano, V.; Bartolini, M. J. Med. Chem. 2016, 59, 114–131. Ellman, G. L.; Courtney, D. K.; Andres, V., Jr.; Feather-Stone, R. M. Biochem. Pharmacol. 1961, 7, 88–95.