- Email: [email protected]
Substituted piperazines as nootropic agents: 2- or 3-phenyl derivatives structurally related to the cognition-enhancer DM235
Bioorganic & Medicinal Chemistry Letters 25 (2015) 1700–1704
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters journal homepage: www.elsevier.com/locate/bmcl
Substituted piperazines as nootropic agents: 2- or 3-phenyl derivatives structurally related to the cognition-enhancer DM235 Luca Guandalini a, Maria Vittoria Martino a, Lorenzo Di Cesare Mannelli b, Gianluca Bartolucci a, Fabrizio Melani a, Ruchi Malik a, , Silvia Dei a, Elisa Floriddia a, Dina Manetti a, Francesca Orlandi a, Elisabetta Teodori a, Carla Ghelardini b, Maria Novella Romanelli a,⇑ a Department of Neuroscience, Psychology, Drug Research and Child’s Health, University of Florence, Section of Pharmaceutical and Nutraceutical Sciences, Via Ugo Schiff 6, 50019 Sesto Fiorentino, Italy b Department of Neuroscience, Psychology, Drug Research and Child’s Health, University of Florence, Section of Pharmacology, Viale Pieraccini 6, 50100 Florence, Italy
a r t i c l e
i n f o
Article history: Received 10 December 2014 Revised 27 February 2015 Accepted 2 March 2015 Available online 12 March 2015 Keywords: Sunifiram Cognition-enhancers Piracetam-like compounds Mouse passive-avoidance test Piperazines
a b s t r a c t A series of 2-phenyl- or 3-phenyl piperazines, structurally related to DM235 and DM232, two potent nootropic agents, have been prepared and tested in the mouse passive-avoidance test, to assess their ability to revert scopolamine-induced amnesia. Although the newly synthesized molecules were less potent than the parent compounds, some useful information has been obtained from structure–activity relationships. A small but significant enantioselectivity has been found for the most potent compound 5a. Ó 2015 Elsevier Ltd. All rights reserved.
Cognitive processes are a set of physiological activities, including attention, learning, memory, reasoning, which are impaired in several psychiatric disorders, such as Alzheimer’s disease (AD), or in age-related conditions such as Mild Cognitive Impairment.1 Several cellular mechanisms have been found to modulate cognition, offering different strategies to overcome cognitive dysfunction. Recently, Froestl et al. have reviewed cognitive enhancers according to their molecular targets,2–4 including in their survey also compounds whose mechanism of action has not been fully elucidated, such as piracetam, 1 and 2.2 We have previously described the cognition-enhancing activity of 1 and 2 (respectively, DM232, Unifiram, and DM235, Sunifiram, Chart 1) which potently reversed scopolamine-induced amnesia in different behavioral tests in rodents.5 Due to the presence of a 2oxopyrrolidine ring, DM232 can be formally related to piracetam; however, the cognition-enhancing potency of DM232 and some of its analogues is 3- 4 orders of magnitude higher than that of the lead. In addition, the pyrrolidinone ring of 1 can be opened, ⇑ Corresponding author. Tel.: +39 055 4573691; fax: +39 055 4573671. E-mail address: novella.romanelli@unifi.it (M.N. Romanelli). Present address: Department of Pharmacy, School of Chemical Sciences and Pharmacy, Central University of Rajasthan, NH-8, Bandar Sindri, Kishangarh, Dist., Ajmer 305801, Rajasthan, India.
http://dx.doi.org/10.1016/j.bmcl.2015.03.009 0960-894X/Ó 2015 Elsevier Ltd. All rights reserved.
to give 2, without loss of activity,6 while this structural feature is pivotal for the activity of piracetam analogues.7 Anyway, DM232 and DM235, as well as piracetam, do not show affinity for the most common neurotransmitter receptors and ion channels, thus making the discovery of the molecular target difficult. Evidence has been found that the cognition-enhancing activity of 1 and 2 involves modulation of cholinergic and glutamatergic transmission.8–10 In an attempt to find new substances which could help to derive structure–activity relationships in this class of compounds, and possibly to elucidate their mechanism of action, we designed new analogs, structurally related to DM235 (2), carrying a substituent on the piperazine ring (Chart 1, general formula I). In a previous work we prepared and tested 2- and 3-methylpiperazines 3a–d,6 whose difference in activity (Table 1) could be related to the position of the methyl group: when it was close to the aliphatic amide moiety, the compound was inactive or less active than when it was close to the aromatic amide or sulfonamide. This result was explained with a steric effect of the methyl ring precluding the correct orientation of the aliphatic amide moiety. Although compounds 3a–d were less potent than the leads 1 and 2, and also than 4 (DM194,6 Chart 1), the 4-fluorobenzenesulfonamide analog of 2, we reasoned that other substituents, with different electronic features with respect to a methyl group, could establish with the
1701
L. Guandalini et al. / Bioorg. Med. Chem. Lett. 25 (2015) 1700–1704
O
O
N
R2 N
Et
2 DM235 (MED 0.001 mg/kg)
N O
Ph
F O
R piracetam (R = H): MED 30 mg/kg Phenylpiracetam (R = Ph) NH 2
N
Me
5a: n = 0 6a: n = 1
I
R 1, R 2 = COEt, COPh, COMe, SO2C 6 H5 F. R 3 = Me (3), Ph (5), CH2 Ph (6)
H N
N
( ) nPh
7: n = 0 8: n = 1
4 DM194 (MED 0.01 mg/kg) F
MED: Minimal Effective Dose Chart 1.
biological target more productive interactions, possibly restoring some potency. Therefore, the piperazine ring of leads 2 and 4 has been decorated with substituents with different size and electronic properties (R3, Chart 1, general formula I). In this work, we report the synthesis and biological evaluation of the phenyl and phenylmethyl derivatives 5 and 6, and of the enantiomers of the most potent derivative 5a. Interestingly, the introduction of a phenyl ring on piracetam gave phenylpiracetam (Chart 1), endowed with interesting nootropic properties (reviewed in Ref. 11). 2-Phenylpiperazine 7 and 2-phenylmethylpiperazine 8 were prepared, respectively, according to Webster12 and Zindell13 and treated sequentially with the suitable acyl or sulfonyl chloride (Scheme 1) to give monoamides 9a–d and 10a,b, and diamides 5a–d and 6a,b. To obtain the enantiomers of 5a, the synthetic pathway (Scheme 2, described for the R enantiomer) started from the
O ( ) nPh
N COPh
Ph N
5b: n = 0 6b: n = 1 b
a
N H
N
O2 S
O
N R1
Et N
N 1 DM232 (MED 0.001 mg/kg)
N O 2S
O
O
R3
c H N
( ) nPh
N R1 n = 0 9a : R 1 = COPh 9b: R1 = COEt 9c: R1 = COMe 9d: R1 = SO2 C6 H 4F n = 1 10a:R1 = COPh 10b: R 1 = COEt
( ) nPh F
N COEt O2 S N
d
e
N COMe
COMe N Ph N
Ph 5c
5d
O2 S F
Scheme 1. Reagents: (a) R1Cl, Et3N; (b) EtCOCl, Et3N; (c) PhCOCl, Et3N; (d) 4FC6H4SO2Cl, Et3N; (e) MeCOCl, Et3N.
commercially available chiral phenylglycine: therefore, (R)-11 was prepared from (R)-phenylglycine according to Refs. 14,15, and treated with diethyl oxalate to give (R)-12. Reduction with LiAlH4 gave (R)-7; subsequent reaction with benzoyl chloride gave (R)-9a, and treatment with propionyl chloride yielded (R)-5a. In the same way (S)-5a was obtained starting from (S)-11. Enantioselective HPLC analysis on (R)-9a and (S)-9a showed a ee of 89% and 88%, respectively (Fig. 1). The compounds were tested in the passive-avoidance test of Jarvik and Kopp,16 using a slightly modified procedure applied in our previous work.17 The substances were dissolved in DMSO/saline (1:10) solution and tested s.c. in a 1:10 dilution sequence; the results are expressed as the Minimal Effective Dose (MED, mg/kg). Compounds were considered inactive if they did not show activity
Table 1 Minimal Effective Dose (MED) of the compounds against scopolamine-induced amnesia in the mouse passive avoidance test, in comparison with reference substances 2–4
R R1 N
N
R2
Treatment
R
R1
R2
c Log Pb
MED (mg/kg)
Training session
Retention session
D
Saline S 9a + S 9b + S 9c + S 9d + S 5a + S (S)-5a + S (R)-5a + S 5b + S 5c + S 5d + S 6a + S 6b + S 2 + Sa 3a + Sa 3b + Sa 3c + Sa 3d + Sa 4 + Sa Piracetam + Sa
— — Ph Ph Ph Ph Ph Ph Ph Ph Ph Ph CH2Ph CH2Ph H CH3 CH3 CH3 CH3 H —
— — COPh COEt COMe SO2C6H4F COPh COPh COPh COEt COMe SO2C6H4F COPh COEt COEt COPh COEt COMe SO2C6H4F SO2C6H4F —
— — H H H H COEt COEt COEt COPh SO2C6H4F COMe COEt COPh COPh COEt COPh SO2C6H4F COMe COMe —
— — 2.71 1.72 1.26 2.18 3.47 3.47 3.47 3.47 2.49 2.49 3.72 3.72 1.95 2.28 2.28 1.29 1.29 0.97 1.18
— — >10 >10 >10 >10 0.1 1.0 0.1 1.0 10 >10 1.0 1.0 0.001 >10 0.1 0.1 1.0 0.01 30
15.5 ± 2.6 14.1 ± 2.1 — — — — 17.7 ± 2.5 13.2 ± 3.0 14.5 ± 2.1 15.2 ± 2.4 13.9 ± 2.2 — 18.4 ± 2.3 18.2 ± 3.0 20.5 ± 3.4 — 21.0 ± 5.3 16.6 ± 4.1 12.5 ± 3.9 19.8 ± 4.1 15.2 ± 3.5
104.8 ± 7.5 47.2 ± 8.3 — — — — 60.2 ± 9.6^ 75.3 ± 9.1⁄ 61.2 ± 8.7^ 72.9 ± 9.5⁄ 79.5 ± 11.6⁄ — 76.3 ± 8.8^ 89.3 ± 9.5⁄ 91.5 ± 8.0⁄ — 81.2 ± 9.6^ 99.2 ± 8.5⁄ 96.4 ± 10.1⁄ 89.0 ± 18.3⁄ 97.6. ± 9.1⁄
89.3 33.1 — — — — 42.5 62.1 46.7 57.7 65.6 — 57.9 71.1 71.0 — 60.2 82.6 83.9 69.2 82.4
All compounds were dissolved in saline containing 10% DMSO and injected s.c. 20 min before the training session. Each value represent the mean of 8–16 mice. Scopolamine (S, 1.5 mg/kg ip) was injected immediately after punishment. ^P <0.05, ⁄P <0.01 in comparison with scopolamine-treated mice. a From Ref. 6. b Calculated on the web by means of Osiris Property Explorer (http://www.organic-chemistry.org/prog/peo/).
1702
L. Guandalini et al. / Bioorg. Med. Chem. Lett. 25 (2015) 1700–1704
H 2N
Ph
HO
O
a
H2N
Ph
O
H N
O
N H
Ph
b
NH 2 (R)- 11
(R)- 12 O
H N c
H N
Ph
Ph e
d N H
N O
(R)- 7
Ph
(R)- 9a
Et N
Ph
N COPh (R)- 5a
Scheme 2. Reagents: (a) Refs. 13,14; (b) diethyl oxalate; (c) LiAlH4, THF; (d) PhCOCl, Et3N; (e) EtCOCl, Et3N. In the same way (S)-5a was obtained starting from (S)-11.
up to the dose of 10 mg/kg, which is three and four orders of magnitude higher than the MED of the lead compounds 4 and 2, respectively. The ability of the compounds 5a–d, 6a,b, and 9a–d to revert scopolamine-induced amnesia is reported in Table 1, expressed as minimal effective dose (MED); compounds 2, 3a–d, 4 and piracetam are shown for comparison. It is necessary to keep in mind, when discussing structure–activity relationships, that the compounds can be only tested in vivo, where the biological activity is the consequence of both the pharmacokinetic and pharmacodynamic properties, which may be differently affected by structural modifications. Unfortunately, the lack of knowledge of the biological target does not allow in vitro tests, where pharmacokinetic factors are largely reduced. As a preliminary PK profiling, the stability of the compounds was assessed in plasma and in phosphate buffer solution to identify hydrolytically labile sites: no appreciable degradation was observed after 2 h incubation, indicating a good stability toward chemical or enzymatic hydrolysis. In addition, as an estimation of blood-brain barrier permeability by passive diffusion, c Log P values were calculated (Table 1); they are within the range for centrally-acting compounds, but there is no correlation with activity. The stability toward microsomal enzymes, the interaction with plasma protein or with transporters has not been assessed.
Concerning the analogs of compound 2, a 2-phenyl group on the piperazine ring, close to the propionyl residue, seems well tolerated (5a, MED 0.1 mg/kg), differently to what happen when a methyl group is in the same position, since 3a is inactive up to the highest tested dose. On the contrary, a methyl group close to the benzoyl residue (position 3) is better tolerated than a phenyl group, since 3b (MED 0.1 mg/kg) is ten times more active than its phenyl analog 5b. In order to see if an increase of the size and flexibility of the substituent could affect activity, the phenylmethyl derivatives 6a and 6b were also tested: both compounds revert scopolamine-induced amnesia with the same minimal effective dose (1.0 mg/kg) but with some loss of activity, 6a being less active that 5a. It seems that a benzyl group on the piperazine ring it is equally tolerated in both positions. As far as the fluorobenzenesulfonyl derivatives are concerned, a methyl group on the piperazine ring is much better tolerated than a phenyl group: as a matter of fact, while 3c and 3d show MED of, respectively, 0.1 and 1.0 mg/kg, compounds 5c is active only at 10 mg/kg, while 5d is not able to revert scopolamine-induced amnesia up to the tested dose. The higher steric hindrance of the phenyl group, compared to methyl, can be taken into account to explain the lower activity of compounds 5c and 5d, with respect to 4 or 3c,d; for this reason the phenylmethyl analogues have not been prepared. It is worth noticing that the different structure activity relationships derived for the analogs of compounds 2 and 4 suggest that the two series of derivatives may adopt different binding modes. Amines 9a–d were also tested, since their lipophilicity could suggest a good blood–brain barrier permeation, but their inactivity indicates, once again, that the removal of one amide moiety reduces activity. This happened also for benzoyl- and propionylpiperazine (13 and 14, Chart 2), which were two orders of magnitude less potent than the parent compound 2.18 This reduction can be due to a decrease in affinity, but also to the different physicochemical characteristics of monoamides: a basic secondary amine could lead to less favorable pharmacokinetic properties. While 13 and 14 still show a good potency, for amines 9a–d, unfavorable pharmacodynamic and/or pharmacokinetic properties could require doses too high and outside our limit. Since the difference in activity could be due to the conformational properties of the compounds, we investigated their conformational preference. The 1H NMR spectra (see Supplementary material) show that for compound 9d only a single conformer is
Figure 1. Enantioselective HPLC analysis of rac-9a and enantiomeric excess of (S)-9a and (R)-9a. Retention times are shown on the peaks.
L. Guandalini et al. / Bioorg. Med. Chem. Lett. 25 (2015) 1700–1704
H N
H N
N
N
O N
O
Ph
13 MED 0.1 mg/kg ClogP 1.19
O
N R 15a (R = COEt): MED > 10 mg/kg ClogP 1.66 15b (R = SO 2C 6 H4 F): MED 0.1 mg/kg ClogP 2.12
Et
14 MED 0.1 mg/kg ClogP 0.20
Chart 2.
present, while 9b, 9c and 5c are present as mixture of two conformers coming from the different orientation of the acetyl or propionyl group; the analysis of the coupling constants for the hydrogen atoms in position 2 and 3 clearly shows that the phenyl ring adopts an equatorial position for monoamides 9b–d, and an axial position for 5c. The poor resolution in the 1H NMR spectra of 9a, 5a,b and 5d precludes a similar analysis, but the presence of conformers is confirmed from the 13C NMR spectra, where the peaks related to the piperazine carbon atoms are doubled and show different height and width; only the propionyl group of 5b seems to adopt different orientations (two peaks for the CH2). Since NMR spectra were not conclusive, we performed some theoretical calculations. Molecular dynamic simulations (3 ns at 300 K) were run for compounds 5a– d; every 100 ps, one conformation was taken and minimized. The axial or equatorial position of the phenyl ring is expressed as a function of the internal angle between the piperazine and the phenyl rings, defined as reported in Figure 2: values close to 180° are indicative of an equatorial arrangement, while values close to 100° are indicative of an axial position. The most part of the minimized conformers of 5c show the phenyl ring in an axial position (Fig. 2), confirming the 1H NMR spectrum for this compound; a similar behavior is found also for 5d (see Supplementary material). On the contrary, a clear preference is not found for 5a (Fig. 2) nor for 5b (see Supplementary material), since the conformations showing an axial or equatorial position for the phenyl group are almost equally populated.
1703
Given these contrasting results, we thought it useful to compare 5a–d to 15a,b, which represent rigid analogs of 2 and 4.19 The phenyl ring of 15a,b, linked to the vicinal acyl group, adopts an arrangement similar to the equatorial position in compounds 5a– d. The inactivity of 15a suggests that for compounds 5a and 5b, the analogs of 2, the preferred orientation of the phenyl ring in position 2 or 3, respectively, is axial. On the contrary, the good potency of compound 15b suggests that for the sulfonyl derivatives 5c and 5d the optimal orientation of the phenyl ring is equatorial, and the lower activity of 5c and 5d may be explained by the low preference for these molecules to adopt such arrangement. Once again, this reasoning suggests a different binding mode for 5a,b and 5c,d. The enantioselectivity of compound 5a, the most potent within the new compounds, was also studied (Table 1): only the (R)-enantiomer show the same MED as the racemate (0.1 mg/kg), the MED of (S)-5a being higher. In previous work we have studied the enantioselectivity in this class of nootropics; among the tested compounds16,20,21 only for 1 a difference in the potency of the enantiomers was observed, the R enantiomer being 3 times more active than the S-one in the passive-avoidance test. Notably, for both 1 and 5a, the eutomers (R enantiomers) show the same spatial arrangement of groups on the chiral center. This low enantioselectivity may reflect the characteristic of the interaction with the biological target, but may be due, as well, to pharmacokinetic factors. Unfortunately, as pointed out already, in vivo experiments do not allow to clarify this issue. In conclusion, new analogues of DM235 have been prepared and tested; even if none of them reaches the outstanding potency of the lead, some compounds (i.e., 5a and its R-enantiomer) show interesting potency, their MED being below 1 mg/kg and 300 times lower than that of piracetam. The results of the passive-avoidance test show that a substituent on the piperazine ring can introduce interesting properties into the molecules and open the way to the design of new analogues. The new compounds allowed to derive further structure–activity relationships, which hopefully will help to elucidate their mechanism of action. Acknowledgments
3.5
This work was supported by grants from MUR (PRIN 2009, 2009ESXPT2_002) and from the University of Florence (ex 60%). We thank Prof. Fulvio Gualtieri for helpful discussion.
∆E Kcal/mol
3
O
2.5
Et 1'
N
2
2 5
1.5
N
1
5a
O
Ph
Supplementary data (synthetic and analytical procedures, spectroscopic characterization of the compounds, theoretical calculations and biological tests) associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.bmcl. 2015.03.009.
0.5 0 80
100
120
140
160
7
ArSO2 N
∆E Kcal/mol
6 5
1'
References and notes
3 6
4
5c
N
3
O
2
Me
1 0 80
100
120
140
Supplementary data
160
Angle τ Figure 2. Distribution of the minimized conformations of 5a (top) and 5c (bottom) as a function of the internal angle s, formed by C5–C2–C10 for 5a, and by C6–C3–C10 for 5c: values close to 180° are indicative of an equatorial position, values close to 100° are indicative of an axial arrangement of the phenyl ring. Similar graphs for 5b and 5d are reported in the Supplementary material.
1. Millan, M. J.; Agid, Y.; Brüne, M.; Bullmore, E. T.; Carter, C. S.; Clayton, N. S.; Connor, R.; Davis, S.; Deakin, B.; DeRubeis, R. J.; Dubois, B.; Geyer, M. A.; Goodwin, G. M.; Gorwood, P.; Jay, T. M.; Joëls, M.; Mansuy, I. M.; MeyerLindenberg, A.; Murphy, D.; Rolls, E.; Saletu, B.; Spedding, M.; Sweeney, J.; Whittington, M.; Young, L. J. Nat. Rev. Drug Disc. 2012, 11, 141. 2. Froestl, W.; Muhs, A.; Pfeifer, A. J. Alzheimers Dis. 2012, 32, 793. 3. Froestl, W.; Muhs, A.; Pfeifer, A. J. Alzheimers Dis. 2013, 33, 547. 4. Froestl, W.; Muhs, A.; Pfeifer, A. J. Alzheimers Dis. 2013, 34, 1. 5. Romanelli, M. N.; Galeotti, N.; Ghelardini, C.; Manetti, D.; Martini, E.; Gualtieri, F. CNS Drug Rev. 2006, 12, 39. 6. Manetti, D.; Ghelardini, C.; Bartolini, A.; Dei, S.; Galeotti, N.; Gualtieri, F.; Romanelli, M. N.; Teodori, E. J. Med. Chem. 2000, 43, 4499. 7. Scapecchi, S.; Martelli, C.; Ghelardini, C.; Guandalini, L.; Martini, E.; Gualtieri, F. Farmaco 2003, 58, 715.
1704
L. Guandalini et al. / Bioorg. Med. Chem. Lett. 25 (2015) 1700–1704
8. Manetti, D.; Ghelardini, C.; Bartolini, A.; Bellucci, C.; Dei, S.; Galeotti, N.; Gualtieri, F.; Romanelli, M. N.; Scapecchi, S.; Teodori, E. J. Med. Chem. 2000, 43, 1969. 9. Galeotti, N.; Ghelardini, C.; Pittaluga, A.; Pugliese, A. M.; Bartolini, A.; Manetti, D.; Romanelli, M. N.; Gualtieri, F. Naunyn-Schmiedeberg’s Arch. Pharmacol. 2003, 368, 538. 10. Moriguchi, S.; Tanaka, T.; Narahashi, T.; Fukunaga, K. Hippocampus 2013, 23, 942. 11. Malykh, A. G.; Sadaie, M. R. Drugs 2010, 70, 287. 12. Webster, S.P.; Seckl, J.R.; Walker, B.R.; Ward, P.; Pallin, T.D.; Dyke, H.J.; Perrior, T.R. WO2009112845A1. 13. Zindell, R.; Riether, D.; Thomson, D.; Hickey, E.R.; Mushi, I.; Ermann, M. WO2007070760A2. 14. Belokon, Y. N.; Pritula, L. K.; Tararov, V. I.; Bakhmutov, V. I.; Struchkov, I. T.; Timofeeva, T. V.; Belikov, V. M. J. Chem. Soc., Dalton Trans. 1990, 1867. 15. Malkov, A. V.; Gouriou, L.; Lloyd-Jones, G.; Stary, C. I.; Langer, V.; Spoor, P.; Vinader, V.; Kocˇovsky, P. Chem. Eur. J. 2006, 12, 6910.
16. Jarvik, M. E.; Kopp, R. Psychol. Rep. 1967, 21, 221. 17. Guandalini, L.; Martini, E.; Di Cesare Mannelli, L.; Dei, S.; Manetti, D.; Scapecchi, S.; Teodori, E.; Ghelardini, C.; Romanelli, M. N. Bioorg. Med. Chem. Lett. 2012, 22, 1936. 18. Scapecchi, S.; Martini, E.; Manetti, D.; Ghelardini, C.; Martelli, C.; Dei, S.; Galeotti, N.; Guandalini, L.; Romanelli, M. N.; Teodori, E. Bioorg. Med. Chem. 2004, 12, 71. 19. Martini, E.; Norcini, M.; Ghelardini, C.; Manetti, D.; Dei, S.; Guandalini, L.; Melchiorre, M.; Pagella, S.; Scapecchi, S.; Teodori, E.; Romanelli, M. N. Bioorg. Med. Chem. 2008, 16, 10034. 20. Martini, E.; Di Cesare Mannelli, L.; Bartolucci, G.; Bertucci, C.; Dei, S.; Ghelardini, C.; Guandalini, L.; Manetti, D.; Scapecchi, S.; Teodori, E.; Romanelli, M. N. J. Med. Chem. 2011, 54, 2512. 21. Martini, E.; Ghelardini, C.; Bertucci, C.; Dei, S.; Gualtieri, F.; Guandalini, L.; Manetti, D.; Scapecchi, S.; Teodori, E.; Romanelli, M. N. Med. Chem. 2005, 1, 473.
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters journal homepage: www.elsevier.com/locate/bmcl
Substituted piperazines as nootropic agents: 2- or 3-phenyl derivatives structurally related to the cognition-enhancer DM235 Luca Guandalini a, Maria Vittoria Martino a, Lorenzo Di Cesare Mannelli b, Gianluca Bartolucci a, Fabrizio Melani a, Ruchi Malik a, , Silvia Dei a, Elisa Floriddia a, Dina Manetti a, Francesca Orlandi a, Elisabetta Teodori a, Carla Ghelardini b, Maria Novella Romanelli a,⇑ a Department of Neuroscience, Psychology, Drug Research and Child’s Health, University of Florence, Section of Pharmaceutical and Nutraceutical Sciences, Via Ugo Schiff 6, 50019 Sesto Fiorentino, Italy b Department of Neuroscience, Psychology, Drug Research and Child’s Health, University of Florence, Section of Pharmacology, Viale Pieraccini 6, 50100 Florence, Italy
a r t i c l e
i n f o
Article history: Received 10 December 2014 Revised 27 February 2015 Accepted 2 March 2015 Available online 12 March 2015 Keywords: Sunifiram Cognition-enhancers Piracetam-like compounds Mouse passive-avoidance test Piperazines
a b s t r a c t A series of 2-phenyl- or 3-phenyl piperazines, structurally related to DM235 and DM232, two potent nootropic agents, have been prepared and tested in the mouse passive-avoidance test, to assess their ability to revert scopolamine-induced amnesia. Although the newly synthesized molecules were less potent than the parent compounds, some useful information has been obtained from structure–activity relationships. A small but significant enantioselectivity has been found for the most potent compound 5a. Ó 2015 Elsevier Ltd. All rights reserved.
Cognitive processes are a set of physiological activities, including attention, learning, memory, reasoning, which are impaired in several psychiatric disorders, such as Alzheimer’s disease (AD), or in age-related conditions such as Mild Cognitive Impairment.1 Several cellular mechanisms have been found to modulate cognition, offering different strategies to overcome cognitive dysfunction. Recently, Froestl et al. have reviewed cognitive enhancers according to their molecular targets,2–4 including in their survey also compounds whose mechanism of action has not been fully elucidated, such as piracetam, 1 and 2.2 We have previously described the cognition-enhancing activity of 1 and 2 (respectively, DM232, Unifiram, and DM235, Sunifiram, Chart 1) which potently reversed scopolamine-induced amnesia in different behavioral tests in rodents.5 Due to the presence of a 2oxopyrrolidine ring, DM232 can be formally related to piracetam; however, the cognition-enhancing potency of DM232 and some of its analogues is 3- 4 orders of magnitude higher than that of the lead. In addition, the pyrrolidinone ring of 1 can be opened, ⇑ Corresponding author. Tel.: +39 055 4573691; fax: +39 055 4573671. E-mail address: novella.romanelli@unifi.it (M.N. Romanelli). Present address: Department of Pharmacy, School of Chemical Sciences and Pharmacy, Central University of Rajasthan, NH-8, Bandar Sindri, Kishangarh, Dist., Ajmer 305801, Rajasthan, India.
http://dx.doi.org/10.1016/j.bmcl.2015.03.009 0960-894X/Ó 2015 Elsevier Ltd. All rights reserved.
to give 2, without loss of activity,6 while this structural feature is pivotal for the activity of piracetam analogues.7 Anyway, DM232 and DM235, as well as piracetam, do not show affinity for the most common neurotransmitter receptors and ion channels, thus making the discovery of the molecular target difficult. Evidence has been found that the cognition-enhancing activity of 1 and 2 involves modulation of cholinergic and glutamatergic transmission.8–10 In an attempt to find new substances which could help to derive structure–activity relationships in this class of compounds, and possibly to elucidate their mechanism of action, we designed new analogs, structurally related to DM235 (2), carrying a substituent on the piperazine ring (Chart 1, general formula I). In a previous work we prepared and tested 2- and 3-methylpiperazines 3a–d,6 whose difference in activity (Table 1) could be related to the position of the methyl group: when it was close to the aliphatic amide moiety, the compound was inactive or less active than when it was close to the aromatic amide or sulfonamide. This result was explained with a steric effect of the methyl ring precluding the correct orientation of the aliphatic amide moiety. Although compounds 3a–d were less potent than the leads 1 and 2, and also than 4 (DM194,6 Chart 1), the 4-fluorobenzenesulfonamide analog of 2, we reasoned that other substituents, with different electronic features with respect to a methyl group, could establish with the
1701
L. Guandalini et al. / Bioorg. Med. Chem. Lett. 25 (2015) 1700–1704
O
O
N
R2 N
Et
2 DM235 (MED 0.001 mg/kg)
N O
Ph
F O
R piracetam (R = H): MED 30 mg/kg Phenylpiracetam (R = Ph) NH 2
N
Me
5a: n = 0 6a: n = 1
I
R 1, R 2 = COEt, COPh, COMe, SO2C 6 H5 F. R 3 = Me (3), Ph (5), CH2 Ph (6)
H N
N
( ) nPh
7: n = 0 8: n = 1
4 DM194 (MED 0.01 mg/kg) F
MED: Minimal Effective Dose Chart 1.
biological target more productive interactions, possibly restoring some potency. Therefore, the piperazine ring of leads 2 and 4 has been decorated with substituents with different size and electronic properties (R3, Chart 1, general formula I). In this work, we report the synthesis and biological evaluation of the phenyl and phenylmethyl derivatives 5 and 6, and of the enantiomers of the most potent derivative 5a. Interestingly, the introduction of a phenyl ring on piracetam gave phenylpiracetam (Chart 1), endowed with interesting nootropic properties (reviewed in Ref. 11). 2-Phenylpiperazine 7 and 2-phenylmethylpiperazine 8 were prepared, respectively, according to Webster12 and Zindell13 and treated sequentially with the suitable acyl or sulfonyl chloride (Scheme 1) to give monoamides 9a–d and 10a,b, and diamides 5a–d and 6a,b. To obtain the enantiomers of 5a, the synthetic pathway (Scheme 2, described for the R enantiomer) started from the
O ( ) nPh
N COPh
Ph N
5b: n = 0 6b: n = 1 b
a
N H
N
O2 S
O
N R1
Et N
N 1 DM232 (MED 0.001 mg/kg)
N O 2S
O
O
R3
c H N
( ) nPh
N R1 n = 0 9a : R 1 = COPh 9b: R1 = COEt 9c: R1 = COMe 9d: R1 = SO2 C6 H 4F n = 1 10a:R1 = COPh 10b: R 1 = COEt
( ) nPh F
N COEt O2 S N
d
e
N COMe
COMe N Ph N
Ph 5c
5d
O2 S F
Scheme 1. Reagents: (a) R1Cl, Et3N; (b) EtCOCl, Et3N; (c) PhCOCl, Et3N; (d) 4FC6H4SO2Cl, Et3N; (e) MeCOCl, Et3N.
commercially available chiral phenylglycine: therefore, (R)-11 was prepared from (R)-phenylglycine according to Refs. 14,15, and treated with diethyl oxalate to give (R)-12. Reduction with LiAlH4 gave (R)-7; subsequent reaction with benzoyl chloride gave (R)-9a, and treatment with propionyl chloride yielded (R)-5a. In the same way (S)-5a was obtained starting from (S)-11. Enantioselective HPLC analysis on (R)-9a and (S)-9a showed a ee of 89% and 88%, respectively (Fig. 1). The compounds were tested in the passive-avoidance test of Jarvik and Kopp,16 using a slightly modified procedure applied in our previous work.17 The substances were dissolved in DMSO/saline (1:10) solution and tested s.c. in a 1:10 dilution sequence; the results are expressed as the Minimal Effective Dose (MED, mg/kg). Compounds were considered inactive if they did not show activity
Table 1 Minimal Effective Dose (MED) of the compounds against scopolamine-induced amnesia in the mouse passive avoidance test, in comparison with reference substances 2–4
R R1 N
N
R2
Treatment
R
R1
R2
c Log Pb
MED (mg/kg)
Training session
Retention session
D
Saline S 9a + S 9b + S 9c + S 9d + S 5a + S (S)-5a + S (R)-5a + S 5b + S 5c + S 5d + S 6a + S 6b + S 2 + Sa 3a + Sa 3b + Sa 3c + Sa 3d + Sa 4 + Sa Piracetam + Sa
— — Ph Ph Ph Ph Ph Ph Ph Ph Ph Ph CH2Ph CH2Ph H CH3 CH3 CH3 CH3 H —
— — COPh COEt COMe SO2C6H4F COPh COPh COPh COEt COMe SO2C6H4F COPh COEt COEt COPh COEt COMe SO2C6H4F SO2C6H4F —
— — H H H H COEt COEt COEt COPh SO2C6H4F COMe COEt COPh COPh COEt COPh SO2C6H4F COMe COMe —
— — 2.71 1.72 1.26 2.18 3.47 3.47 3.47 3.47 2.49 2.49 3.72 3.72 1.95 2.28 2.28 1.29 1.29 0.97 1.18
— — >10 >10 >10 >10 0.1 1.0 0.1 1.0 10 >10 1.0 1.0 0.001 >10 0.1 0.1 1.0 0.01 30
15.5 ± 2.6 14.1 ± 2.1 — — — — 17.7 ± 2.5 13.2 ± 3.0 14.5 ± 2.1 15.2 ± 2.4 13.9 ± 2.2 — 18.4 ± 2.3 18.2 ± 3.0 20.5 ± 3.4 — 21.0 ± 5.3 16.6 ± 4.1 12.5 ± 3.9 19.8 ± 4.1 15.2 ± 3.5
104.8 ± 7.5 47.2 ± 8.3 — — — — 60.2 ± 9.6^ 75.3 ± 9.1⁄ 61.2 ± 8.7^ 72.9 ± 9.5⁄ 79.5 ± 11.6⁄ — 76.3 ± 8.8^ 89.3 ± 9.5⁄ 91.5 ± 8.0⁄ — 81.2 ± 9.6^ 99.2 ± 8.5⁄ 96.4 ± 10.1⁄ 89.0 ± 18.3⁄ 97.6. ± 9.1⁄
89.3 33.1 — — — — 42.5 62.1 46.7 57.7 65.6 — 57.9 71.1 71.0 — 60.2 82.6 83.9 69.2 82.4
All compounds were dissolved in saline containing 10% DMSO and injected s.c. 20 min before the training session. Each value represent the mean of 8–16 mice. Scopolamine (S, 1.5 mg/kg ip) was injected immediately after punishment. ^P <0.05, ⁄P <0.01 in comparison with scopolamine-treated mice. a From Ref. 6. b Calculated on the web by means of Osiris Property Explorer (http://www.organic-chemistry.org/prog/peo/).
1702
L. Guandalini et al. / Bioorg. Med. Chem. Lett. 25 (2015) 1700–1704
H 2N
Ph
HO
O
a
H2N
Ph
O
H N
O
N H
Ph
b
NH 2 (R)- 11
(R)- 12 O
H N c
H N
Ph
Ph e
d N H
N O
(R)- 7
Ph
(R)- 9a
Et N
Ph
N COPh (R)- 5a
Scheme 2. Reagents: (a) Refs. 13,14; (b) diethyl oxalate; (c) LiAlH4, THF; (d) PhCOCl, Et3N; (e) EtCOCl, Et3N. In the same way (S)-5a was obtained starting from (S)-11.
up to the dose of 10 mg/kg, which is three and four orders of magnitude higher than the MED of the lead compounds 4 and 2, respectively. The ability of the compounds 5a–d, 6a,b, and 9a–d to revert scopolamine-induced amnesia is reported in Table 1, expressed as minimal effective dose (MED); compounds 2, 3a–d, 4 and piracetam are shown for comparison. It is necessary to keep in mind, when discussing structure–activity relationships, that the compounds can be only tested in vivo, where the biological activity is the consequence of both the pharmacokinetic and pharmacodynamic properties, which may be differently affected by structural modifications. Unfortunately, the lack of knowledge of the biological target does not allow in vitro tests, where pharmacokinetic factors are largely reduced. As a preliminary PK profiling, the stability of the compounds was assessed in plasma and in phosphate buffer solution to identify hydrolytically labile sites: no appreciable degradation was observed after 2 h incubation, indicating a good stability toward chemical or enzymatic hydrolysis. In addition, as an estimation of blood-brain barrier permeability by passive diffusion, c Log P values were calculated (Table 1); they are within the range for centrally-acting compounds, but there is no correlation with activity. The stability toward microsomal enzymes, the interaction with plasma protein or with transporters has not been assessed.
Concerning the analogs of compound 2, a 2-phenyl group on the piperazine ring, close to the propionyl residue, seems well tolerated (5a, MED 0.1 mg/kg), differently to what happen when a methyl group is in the same position, since 3a is inactive up to the highest tested dose. On the contrary, a methyl group close to the benzoyl residue (position 3) is better tolerated than a phenyl group, since 3b (MED 0.1 mg/kg) is ten times more active than its phenyl analog 5b. In order to see if an increase of the size and flexibility of the substituent could affect activity, the phenylmethyl derivatives 6a and 6b were also tested: both compounds revert scopolamine-induced amnesia with the same minimal effective dose (1.0 mg/kg) but with some loss of activity, 6a being less active that 5a. It seems that a benzyl group on the piperazine ring it is equally tolerated in both positions. As far as the fluorobenzenesulfonyl derivatives are concerned, a methyl group on the piperazine ring is much better tolerated than a phenyl group: as a matter of fact, while 3c and 3d show MED of, respectively, 0.1 and 1.0 mg/kg, compounds 5c is active only at 10 mg/kg, while 5d is not able to revert scopolamine-induced amnesia up to the tested dose. The higher steric hindrance of the phenyl group, compared to methyl, can be taken into account to explain the lower activity of compounds 5c and 5d, with respect to 4 or 3c,d; for this reason the phenylmethyl analogues have not been prepared. It is worth noticing that the different structure activity relationships derived for the analogs of compounds 2 and 4 suggest that the two series of derivatives may adopt different binding modes. Amines 9a–d were also tested, since their lipophilicity could suggest a good blood–brain barrier permeation, but their inactivity indicates, once again, that the removal of one amide moiety reduces activity. This happened also for benzoyl- and propionylpiperazine (13 and 14, Chart 2), which were two orders of magnitude less potent than the parent compound 2.18 This reduction can be due to a decrease in affinity, but also to the different physicochemical characteristics of monoamides: a basic secondary amine could lead to less favorable pharmacokinetic properties. While 13 and 14 still show a good potency, for amines 9a–d, unfavorable pharmacodynamic and/or pharmacokinetic properties could require doses too high and outside our limit. Since the difference in activity could be due to the conformational properties of the compounds, we investigated their conformational preference. The 1H NMR spectra (see Supplementary material) show that for compound 9d only a single conformer is
Figure 1. Enantioselective HPLC analysis of rac-9a and enantiomeric excess of (S)-9a and (R)-9a. Retention times are shown on the peaks.
L. Guandalini et al. / Bioorg. Med. Chem. Lett. 25 (2015) 1700–1704
H N
H N
N
N
O N
O
Ph
13 MED 0.1 mg/kg ClogP 1.19
O
N R 15a (R = COEt): MED > 10 mg/kg ClogP 1.66 15b (R = SO 2C 6 H4 F): MED 0.1 mg/kg ClogP 2.12
Et
14 MED 0.1 mg/kg ClogP 0.20
Chart 2.
present, while 9b, 9c and 5c are present as mixture of two conformers coming from the different orientation of the acetyl or propionyl group; the analysis of the coupling constants for the hydrogen atoms in position 2 and 3 clearly shows that the phenyl ring adopts an equatorial position for monoamides 9b–d, and an axial position for 5c. The poor resolution in the 1H NMR spectra of 9a, 5a,b and 5d precludes a similar analysis, but the presence of conformers is confirmed from the 13C NMR spectra, where the peaks related to the piperazine carbon atoms are doubled and show different height and width; only the propionyl group of 5b seems to adopt different orientations (two peaks for the CH2). Since NMR spectra were not conclusive, we performed some theoretical calculations. Molecular dynamic simulations (3 ns at 300 K) were run for compounds 5a– d; every 100 ps, one conformation was taken and minimized. The axial or equatorial position of the phenyl ring is expressed as a function of the internal angle between the piperazine and the phenyl rings, defined as reported in Figure 2: values close to 180° are indicative of an equatorial arrangement, while values close to 100° are indicative of an axial position. The most part of the minimized conformers of 5c show the phenyl ring in an axial position (Fig. 2), confirming the 1H NMR spectrum for this compound; a similar behavior is found also for 5d (see Supplementary material). On the contrary, a clear preference is not found for 5a (Fig. 2) nor for 5b (see Supplementary material), since the conformations showing an axial or equatorial position for the phenyl group are almost equally populated.
1703
Given these contrasting results, we thought it useful to compare 5a–d to 15a,b, which represent rigid analogs of 2 and 4.19 The phenyl ring of 15a,b, linked to the vicinal acyl group, adopts an arrangement similar to the equatorial position in compounds 5a– d. The inactivity of 15a suggests that for compounds 5a and 5b, the analogs of 2, the preferred orientation of the phenyl ring in position 2 or 3, respectively, is axial. On the contrary, the good potency of compound 15b suggests that for the sulfonyl derivatives 5c and 5d the optimal orientation of the phenyl ring is equatorial, and the lower activity of 5c and 5d may be explained by the low preference for these molecules to adopt such arrangement. Once again, this reasoning suggests a different binding mode for 5a,b and 5c,d. The enantioselectivity of compound 5a, the most potent within the new compounds, was also studied (Table 1): only the (R)-enantiomer show the same MED as the racemate (0.1 mg/kg), the MED of (S)-5a being higher. In previous work we have studied the enantioselectivity in this class of nootropics; among the tested compounds16,20,21 only for 1 a difference in the potency of the enantiomers was observed, the R enantiomer being 3 times more active than the S-one in the passive-avoidance test. Notably, for both 1 and 5a, the eutomers (R enantiomers) show the same spatial arrangement of groups on the chiral center. This low enantioselectivity may reflect the characteristic of the interaction with the biological target, but may be due, as well, to pharmacokinetic factors. Unfortunately, as pointed out already, in vivo experiments do not allow to clarify this issue. In conclusion, new analogues of DM235 have been prepared and tested; even if none of them reaches the outstanding potency of the lead, some compounds (i.e., 5a and its R-enantiomer) show interesting potency, their MED being below 1 mg/kg and 300 times lower than that of piracetam. The results of the passive-avoidance test show that a substituent on the piperazine ring can introduce interesting properties into the molecules and open the way to the design of new analogues. The new compounds allowed to derive further structure–activity relationships, which hopefully will help to elucidate their mechanism of action. Acknowledgments
3.5
This work was supported by grants from MUR (PRIN 2009, 2009ESXPT2_002) and from the University of Florence (ex 60%). We thank Prof. Fulvio Gualtieri for helpful discussion.
∆E Kcal/mol
3
O
2.5
Et 1'
N
2
2 5
1.5
N
1
5a
O
Ph
Supplementary data (synthetic and analytical procedures, spectroscopic characterization of the compounds, theoretical calculations and biological tests) associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.bmcl. 2015.03.009.
0.5 0 80
100
120
140
160
7
ArSO2 N
∆E Kcal/mol
6 5
1'
References and notes
3 6
4
5c
N
3
O
2
Me
1 0 80
100
120
140
Supplementary data
160
Angle τ Figure 2. Distribution of the minimized conformations of 5a (top) and 5c (bottom) as a function of the internal angle s, formed by C5–C2–C10 for 5a, and by C6–C3–C10 for 5c: values close to 180° are indicative of an equatorial position, values close to 100° are indicative of an axial arrangement of the phenyl ring. Similar graphs for 5b and 5d are reported in the Supplementary material.
1. Millan, M. J.; Agid, Y.; Brüne, M.; Bullmore, E. T.; Carter, C. S.; Clayton, N. S.; Connor, R.; Davis, S.; Deakin, B.; DeRubeis, R. J.; Dubois, B.; Geyer, M. A.; Goodwin, G. M.; Gorwood, P.; Jay, T. M.; Joëls, M.; Mansuy, I. M.; MeyerLindenberg, A.; Murphy, D.; Rolls, E.; Saletu, B.; Spedding, M.; Sweeney, J.; Whittington, M.; Young, L. J. Nat. Rev. Drug Disc. 2012, 11, 141. 2. Froestl, W.; Muhs, A.; Pfeifer, A. J. Alzheimers Dis. 2012, 32, 793. 3. Froestl, W.; Muhs, A.; Pfeifer, A. J. Alzheimers Dis. 2013, 33, 547. 4. Froestl, W.; Muhs, A.; Pfeifer, A. J. Alzheimers Dis. 2013, 34, 1. 5. Romanelli, M. N.; Galeotti, N.; Ghelardini, C.; Manetti, D.; Martini, E.; Gualtieri, F. CNS Drug Rev. 2006, 12, 39. 6. Manetti, D.; Ghelardini, C.; Bartolini, A.; Dei, S.; Galeotti, N.; Gualtieri, F.; Romanelli, M. N.; Teodori, E. J. Med. Chem. 2000, 43, 4499. 7. Scapecchi, S.; Martelli, C.; Ghelardini, C.; Guandalini, L.; Martini, E.; Gualtieri, F. Farmaco 2003, 58, 715.
1704
L. Guandalini et al. / Bioorg. Med. Chem. Lett. 25 (2015) 1700–1704
8. Manetti, D.; Ghelardini, C.; Bartolini, A.; Bellucci, C.; Dei, S.; Galeotti, N.; Gualtieri, F.; Romanelli, M. N.; Scapecchi, S.; Teodori, E. J. Med. Chem. 2000, 43, 1969. 9. Galeotti, N.; Ghelardini, C.; Pittaluga, A.; Pugliese, A. M.; Bartolini, A.; Manetti, D.; Romanelli, M. N.; Gualtieri, F. Naunyn-Schmiedeberg’s Arch. Pharmacol. 2003, 368, 538. 10. Moriguchi, S.; Tanaka, T.; Narahashi, T.; Fukunaga, K. Hippocampus 2013, 23, 942. 11. Malykh, A. G.; Sadaie, M. R. Drugs 2010, 70, 287. 12. Webster, S.P.; Seckl, J.R.; Walker, B.R.; Ward, P.; Pallin, T.D.; Dyke, H.J.; Perrior, T.R. WO2009112845A1. 13. Zindell, R.; Riether, D.; Thomson, D.; Hickey, E.R.; Mushi, I.; Ermann, M. WO2007070760A2. 14. Belokon, Y. N.; Pritula, L. K.; Tararov, V. I.; Bakhmutov, V. I.; Struchkov, I. T.; Timofeeva, T. V.; Belikov, V. M. J. Chem. Soc., Dalton Trans. 1990, 1867. 15. Malkov, A. V.; Gouriou, L.; Lloyd-Jones, G.; Stary, C. I.; Langer, V.; Spoor, P.; Vinader, V.; Kocˇovsky, P. Chem. Eur. J. 2006, 12, 6910.
16. Jarvik, M. E.; Kopp, R. Psychol. Rep. 1967, 21, 221. 17. Guandalini, L.; Martini, E.; Di Cesare Mannelli, L.; Dei, S.; Manetti, D.; Scapecchi, S.; Teodori, E.; Ghelardini, C.; Romanelli, M. N. Bioorg. Med. Chem. Lett. 2012, 22, 1936. 18. Scapecchi, S.; Martini, E.; Manetti, D.; Ghelardini, C.; Martelli, C.; Dei, S.; Galeotti, N.; Guandalini, L.; Romanelli, M. N.; Teodori, E. Bioorg. Med. Chem. 2004, 12, 71. 19. Martini, E.; Norcini, M.; Ghelardini, C.; Manetti, D.; Dei, S.; Guandalini, L.; Melchiorre, M.; Pagella, S.; Scapecchi, S.; Teodori, E.; Romanelli, M. N. Bioorg. Med. Chem. 2008, 16, 10034. 20. Martini, E.; Di Cesare Mannelli, L.; Bartolucci, G.; Bertucci, C.; Dei, S.; Ghelardini, C.; Guandalini, L.; Manetti, D.; Scapecchi, S.; Teodori, E.; Romanelli, M. N. J. Med. Chem. 2011, 54, 2512. 21. Martini, E.; Ghelardini, C.; Bertucci, C.; Dei, S.; Gualtieri, F.; Guandalini, L.; Manetti, D.; Scapecchi, S.; Teodori, E.; Romanelli, M. N. Med. Chem. 2005, 1, 473.