- Email: [email protected]
Synthesis and biological evaluation of strained unusual amino acid containing tetrapeptides as potential antidepressant agents
Bioorganic Chemistry 63 (2015) 53–57
Contents lists available at ScienceDirect
Bioorganic Chemistry journal homepage: www.elsevier.com/locate/bioorg
Synthesis and biological evaluation of strained unusual amino acid containing tetrapeptides as potential antidepressant agents Prathama S. Mainkar a,⇑, Sumana Chakravarty b,⇑, Takkallapally Srujana c, Libi Anandi Vishwanathan c, Santosh Kumar Prajapti c, Karisetty Bhanu Chandra b, Lenin Veeraval b, Bathini Nagendra Babu c,⇑ a b c
Division of Natural Products Chemistry, CSIR-Indian Institute of Chemical Technology, Tarnaka, Hyderabad 500007, India Centre for Chemical Biology, CSIR-Indian Institute of Chemical Technology, Tarnaka, Hyderabad 500007, India Department of Medicinal Chemistry, National Institute of Pharmaceutical Education and Research (NIPER), Hyderabad, Balanagar, Hyderabad 500037, India
a r t i c l e
i n f o
Article history: Received 30 June 2015 Revised 24 September 2015 Accepted 26 September 2015 Available online 30 September 2015 Keywords: NMDA receptor GRIN2B Norbornane amino acid Neurotrophic
a b s t r a c t Strained unusual amino acid derived tetrapeptides were synthesized as mimics of GLYX-13, a clinical candidate for neuroprotective and anti-depressant properties, were studied. The synthesized compounds were screened for neurite growth and anti-depressant properties in vitro and in vivo respectively comparing with the parent GLYX-13 compound. Neurite growth property was assessed by neurite length and anti-depressant property by percentage of immobility in forced swim test, a behavioural assay. Mechanistic insights about protein–ligand interactions were obtained using molecular docking study. Based on the in vitro and in vivo screening data and molecular docking study, a new analogue of GLYX-13, Compound 11a has been found to be as good as the parent compound in all respects. Ó 2015 Elsevier Inc. All rights reserved.
1. Introduction Many neuroactive peptides are employed by the brain and nervous system of mammals which help in specialized signalling with central nervous system (CNS). A careful study of the signalling pathways has led to the identification of some specific receptors modulated by these neuroactive peptides which opened up an important therapeutic area in CNS disorders. The N-methyl-Daspartate receptor (NMDAR) is one such receptor which is reported to be involved in various disorders such as cognitive function associated with ageing, ischemia, stroke, head trauma, spinal cord injury, epilepsy, depression, schizophrenia etc. [1]. NMDARs are widely distributed in brain and spinal cord; higher densities are found in cortex and hippocampus. NMDAR ion channels are one of the postsynaptic targets of glutamate-mediated synaptic transmission in mammalian brain. For opening, NMDAR channels require binding of two different types of agonists-glutamate and glycine [2]. Many glutamate receptors are found in CNS but strychnine-insensitive site is unique to NMDAR [3]. Drugs, which are capable of blocking NMDAR activation and glutamate binding site antagonists, create some strong side-effects. As a result, attention has shifted to the glycine site agonists for altering the NMDAR, which is a potentially important target for development of cogni⇑ Corresponding authors. E-mail address: [email protected] (B.N. Babu). http://dx.doi.org/10.1016/j.bioorg.2015.09.007 0045-2068/Ó 2015 Elsevier Inc. All rights reserved.
tive enhancing therapeutics and anticonvulsant and neuroprotective agents. The glycine site of NMDAR complex is a co-agonist site with affinity for glycine and D-serine. It is an allosteric site and hence has been explored for identifying and developing safer drugs. The advantage of partial NMDAR agonists is that they block excessive NMDA function and also potentiate NMDAR in case of abnormally depressed NMDA-mediated neuro-transmission. D-Cycloserine [4] and glycine prodrug, milacemide [5] have shown cognitiveenhancing properties in vivo but result in desensitization with chronic administration. Monoclonal antibody B6B21 has been shown to act as a partial agonist at the glycine site of NMDAR. The hyper variable region of the light chain of B6B21 was cloned and sequenced [6]. Based on this sequence information several peptides were synthesized and screened using rat hippocampal membrane preparations to measure [3H]MK-801 binding in presence of 7-chlorokynuric acid, a glycine site-specific competitive inhibitor of NMDAR. Peptides that increased the binding of [3H]MK-801, an open channel blocker, in dose dependent manner were named Glyxins [7]. Glyxins NT-1 to NT-18 were synthesized which comprised of peptides containing amino acids up to 26. Of these, GLYX-13 (NT-13) [22] (1, Fig. 1) was found to be the best. GLYX-13 is a tetrapeptide (Thr-ProPro-Thr) [23] and is known to readily cross the blood–brain barrier. Clinical data showed that it has robust anti-depressant activity [8], the effect is immediate (within 20 min) and long lasting (at least
54
P.S. Mainkar et al. / Bioorganic Chemistry 63 (2015) 53–57
Fig. 1. Structure of parent compound GLYX-13.
for 2 weeks after single dose) with no CNS-related side-effects [9]. It was also found to enhance learning (cognitive function) [10], help in autism treatment and increase neuroprotection in cases of stroke [11]. GLYX-13 was able to modulate NMDAR in a glycine-like fashion when examined pharmacologically and electro-physiologically. Based on the biological screening data of Glyxins, the optimal length of the peptides was found to be four. We have been working in the area of synthesis of peptides and foldamers consisting of unusual strained b-amino acids and their structural investigations [12]. In addition, the incorporation of a strained b-amino acid derived from sugar into a marine peptide natural product, azumamide, for enhanced activity [13] and many peptides with unusual amino acids have been reported [13]. Inspired by this observation, GLYX-13 skeleton was chosen as a starting point in the synthesis of b-amino acid analogues of GLYX-13, 1, to understand the SAR. A first look at 1 revealed the presence of turninducing Pro–Pro dimer as the central core. This prompted us to initiate incorporation of cis-furan sugar amino acid, 2 [12] and cis-exo-b-norbornene amino acid, 3 and 4 [12] as substitutes of central Pro–Pro core of 1 (Fig. 2). These unusual amino acids have already been proven to be turn inducers as observed by us earlier [12]. The terminal amino acids in 1 are threonine and are expected to participate in hydrogen-bonding. Based on this observation we initiated incorporation of turn-inducing amino acid pairs as core and substituted the terminal amino acids, threonine, with 2-aminobenzoic acid. A preliminary molecular modelling study data indicated that insertion of sugar amino acids will not be able to induce the required turn for the expected activity as they exhibit formation of helical structures [12,13]. Thus the focus then shifted to inserting norbornene amino acid which is a known turn inducer [12]. To determine the impact of stereochemistry on the neuroprotective activity both the isomers of norbornene amino acid viz., (2R,3S)-cis-exo-b-norbornene amino acid, 3 and (2S,3R)-cis-exo-bnorbornene amino acid, 4, were used.
2. Result and discussion 2.1. Chemistry The work on the analogue synthesis began with retention of the proline–proline core with modification in the terminal amino acids. Thus a set of four analogues were synthesized incorporating glycine, 5, 4-aminobenzoic acid 6, b-alanine 7 and 2-aminobenzoic acid 8, following usual peptide synthesis protocols (Fig. 3). Biological screening of these compounds did not yield encouraging results (data not shown). Therefore further attempts of analogue synthesis were shifted to modifying the turn-inducing core.
Fig. 2. cis-Furan sugar amino acid and cis-exo-norbornane amino acid.
Enantiopure norbornene amino acid (both enantiomers exo(2S,3R) and exo-(2R,3S) individually) were used to prepare the analogues 11a, 11b, 13a and 13b [12] N- and C-Protected dimers (9a and 9b) were prepared from both 2S,3R- and 2R,3S-cis-exo-bnorbornene amino acid, 3 and 4, independently using our earlier protocol [12]. These dimers are used to synthesize the proposed analogues 11a, 11b, 13a and 13b by attaching the appropriately protected monomer units of L-threonine and 2-aminobenzoic acid. The trimers were prepared via classical peptide forming HOBt– EDCI protocol which involves coupling of amine of the dimer (readily obtained from dimer 9a and 9b by treating with trifluoroacetic acid) and N-Boc protected L-threonine N-Boc-2-aminobenzoic acid respectively. Similarly, tetramers were prepared from trimer acids coupled with methyl esters of L-threonine and 2-aminobenzoic acid (Scheme 1). All the analogues along with the parent compound 1 were evaluated for their biological efficacy. 2.2. Biological evaluation GLYX-13 (1), a compound approved by FDA in Fast Track designation on 3rd March 2014 as an adjunctive therapy in treatment of resistant major depressive disorder. This molecule completed phase IIb clinical trials and phase III trials will start in 2015. A major problem associated with 1 is its bioavailability and hence it is administered via intraperitoneal route to undergo first-pass effect which is bypassed in intravenous route. As the molecule has shown great promise, analogues are planned which are less prone to enzymatic degradation and increase its bioavailability. The GLYX-13 (1) and its analogues 11a, 11b, 13a and 13b were preliminarily screened in vitro on Neuro-2a cell line (ATCC CCL-131) [14–16] for neurotrophic activity at 0.01, 0.1 and 1.0 lM concentrations and a concurrent MTT cell cytotoxic assay was carried out to substantiate the compounds concentrations are non-toxic (Supplementary Fig. 1 and Supplementary Table 1). All the analogues except 11b showed noticeable neurotrophic activity at the lowest 0.01 lM concentration (Fig. 4 and Supplementary Table 2). But, the BDNF gene expression did not show consistent changes in analogous with the neurite length (Supplementary Fig. 4). This data suggests that the neurotrophic action of 1, an NMDA receptor agonist, and its analogues might not be directly via activation of BDNF gene. Further the analogues were screened in vivo for their anti-depressant activity and were compared with 1 using a behavioural assay, Forced Swim Test [17,18] (Fig. 5). This assay indicated that compounds 11a and 11b had activity similar to 1. With this indication, we explored compatibility of both 11a and 11b to NMDA receptor which could be mediated through GRIN2B or NR2B receptor as 1, based on the literature precedence [9]. Neuro-2a cells were treated with 1.0, 0.1, 0.01 and 0.001 lM concentrations of 11a, 11b, 1 and DMSO. All the three has shown increased GRIN2B gene expression levels at 0.01 lM concentration when compared to other. Further GRIN2B gene expression of 1, 11a and 11b at 0.01 lM were compared and 1 and 11a induced a significant up regulation compared with DMSO treated (Fig. 6). The data suggests that 11a and 11b work on the same pathway as 1, viz., as NMDA receptor agonist and 11a induced more GRIN2B gene expression than others. After analysing both in vitro and in vivo results GLYX-13 analogue 11a seems to be the best lead molecule to consider for further development as a novel antidepressant as this molecule showed significant neurotrophic activity at 0.01 lM concentration as well with the noticeable improvement in immobility time in forced swim test as compared to vehicle treatment (Figs. 4 and 5). The present study also suggests the compound 11a might be working as NMDA receptor agonist like parent compound, GLYX13, in a better way at mRNA gene expression levels. A protein level work would have further substantiated the mRNA results. It also
P.S. Mainkar et al. / Bioorganic Chemistry 63 (2015) 53–57
55
Fig. 3. Analogues of 1 with terminal amino acids modified.
Scheme 1. Synthesis of analogues with cis-exo-b-norbornene amino acid.
reports that the neurotrophic property of the analogues is not directly through BDNF pathway. To conclude about the mechanism of action, more specific molecular level studies need to be performed in future. Especially, it would be interesting to find interlink between neurotrophic nature and anti-depressant properties of the analogues. 2.3. Molecular docking study In order to gain insight about protein–ligand interactions of newly synthesized analogues in the active site of NMDA receptor, molecular docking was performed for compounds 5–8, 11a, 11b, 13a and 13b. Coordinates of NMDA receptor were obtained from
the Protein Data Bank (PDB ID: 1Y1Z) [19,20]. Geometry of the molecules was optimized by using the optimized potentials for liquid simulations-2005 (OPLS-2005) force field. The docking simulations in the active site of NMDA receptor was done with aid of Glide docking tool of Schrödinger suite [21]. For the purpose of docking protocol validation, ligand bound to the crystal structure was removed from the binding pocket and docked back into the same binding site resulted in the same binding interactions. Compounds 5–8, 11a, 11b, 13a, 13b and GLYX13 (reference compound) were docked flexibly into the crystallographic structure of the NMDA receptor using Docking Glide Extra Precision (XP), version 6.0 (Schrödinger, 2013) [21].
56
P.S. Mainkar et al. / Bioorganic Chemistry 63 (2015) 53–57
Fig. 4. Bar representation of the neurite outgrowth of differentiated Neuro-2a cells by GLYX-13 and its analogues at 0.01 lM concentration in comparison with DMSO and NGF (A). Bright field images showing optimum changes in neurite outgrowth of differentiated Neuro-2a cells treated with analogues (B).
Fig. 5. The new set of GLYX-13 analogues and their antidepressant efficacies in Forced swim test (FST).
Docking result showed that the ligands bound in same site formed from Gly249, Arg187, Arg247, Pro140, Phe246, Tyr143, Phe137, Ile127, Pro140, Ala132, Gln144 and Asn 129 amino acids (Fig. 7). Among the synthesized compounds, the calculated docking score for the compound 11a and 11b were 3.9 and 4.5 kcal/mol respectively, which are in agreement with experimental assay. Compound 11a and 11b showed hydrogen bonding similar to GLYX-13 with Arg187, Arg 247 and Gly249. Also, compound 11a displayed hydrogen bonding with Asn129 and Phe137. Beside this, docking pose of compound 11a and 11b showed hydrophobic interaction with Ile127, Phe246, Ala132 and Pro140 in addition to Vander Waals interactions with other residue of enzyme. All these interactions together contribute to the hypothesis that the compounds bind efficiently to the active site of NMDA receptor, thus ensuing antidepressant activity as in case of GLYX-13.
Fig. 6. Fold change of GRIN2B gene expression in Neuro-2a cells treated with GLYX-13, 11a and 11b at concentrations 1.0–0.001 lM with DMSO treated cells as control (A, B and C). Logarithmic representation of fold change of GRIN2B gene expression in Neuro-2a cells treated with compounds GLYX-13, 11a and 11b at 0.01 lM concentration with DMSO treated cells as control (B). Here ⁄p < 0.05 and ⁄⁄⁄p < 0.001, n = 5–6.
P.S. Mainkar et al. / Bioorganic Chemistry 63 (2015) 53–57
57
Fig. 7. Top view of 3-D image showing interactions of compound 11a, 11b and GLYX-13 with NMDA receptor (PDB ID: 1Y1Z).
Moreover, the results for the compounds 5–8, 13a and 13b were less favourable than the calculated docking score of the GLYX-13, which was 4.2 kcal/mol (Supplementary Table 5).
3. Conclusion In summary, we have designed, synthesized, and identified novel analogues of GLYX-13, a compound in clinical trials for treatment of depression and cognitive disorders. Based on very encouraging results from the in vitro and in vivo studies and molecular docking study, compound 11a has been found to be as good as the parent compound in all respects. This study also establishes the fact that proline–proline core can be replaced with turn inducing b-amino acids which are less prone to enzymatic degradation. Acknowledgments Authors thank National Institute of Pharmaceutical Education and Research (NIPER)-Hyderabad; and CSIR-Indian Institute of Chemical Technology-Hyderabad for providing facilities and Department of Pharmaceuticals (DoP); Government of India (GoI) for financial support. Authors like to thank Dr Srivari Chandrasekhar for his suggestions and extended support. BN likes to acknowledge Dr K. Srinivas for helpful discussion in molecular docking study. SC acknowledges the support of CSC0111/SMiLE for biological assays.
Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.bioorg.2015.09. 007.
References [1] M. Hoffman, Science 254 (5033) (1991) 801–802. [2] A.P. Kozikowski, W. Tuckmantel, I.J. Reynolds, J.T. Wroblewski, J. Med. Chem. 33 (6) (1990) 1561–1571. [3] P.L. Wood, IDrugs 8 (3) (2005) 229–235. [4] W.F. Hood, R.P. Compton, J.B. Monahan, Neurosci. Lett. 98 (1) (1989) 91–95. [5] G.E. Handelmann, M.E. Nevins, L.L. Mueller, S.M. Arnolde, A.A. Cordi, Pharmacol. Biochem. Behav. 34 (4) (1989) 823–828. [6] P.K. Stanton, J.M. Sarvey, J.R. Moskal, Proc. Natl. Acad. Sci. USA 84 (6) (1987) 1684–1688. [7] X.L. Zhang, J.A. Sullivan, J.R. Moskal, P.K. Stanton, Neuropharmacology 55 (7) (2008) 1238–1250. [8] J. Burgdorf, X.L. Zhang, C. Weiss, E. Matthews, J.F. Disterhoft, P.K. Stanton, J.R. Moskal, Neurobiol. Aging 32 (4) (2011) 698–706. [9] J. Burgdorf, X.L. Zhang, K.L. Nicholson, R.L. Balster, J.D. Leander, P.K. Stanton, A. L. Gross, R.A. Kroes, J.R. Moskal, Neuropsychopharmacology 38 (5) (2013) 729– 742. [10] P.K. Stanton, P.E. Potter, J. Aguilar, M. Decandia, J.R. Moskal, Neuroreport 20 (13) (2009) 1193–1197. [11] J.R. Moskal, A.G. Kuo, C. Weiss, P.L. Wood, A. O’Connor Hanson, S. Kelso, R.B. Harris, J.F. Disterhoft, Neuropharmacology 49 (7) (2005) 1077–1087. [12] S. Chandrasekhar, M.S. Reddy, B.N. Babu, B. Jagadeesh, A. Prabhakar, B. Jagannadh, J. Am. Chem. Soc. 127 (27) (2005) 9664–9665. [13] S. Chandrasekhar, C.L. Rao, M. Seenaiah, P. Naresh, B. Jagadeesh, D. Manjeera, A. Sarkar, M.P. Bhadra, J. Org. Chem. 74 (1) (2009) 401–404. [14] Z. Khaing, D. Kang, A.M. Camelio, C.E. Schmidt, D. Siegel, Bioorg. Med. Chem. Lett. 21 (16) (2011) 4808–4812. [15] G. Mehta, R. Samineni, P. Srihari, R.G. Reddy, S. Chakravarty, Org. Biomol. Chem. 10 (34) (2012) 6830–6833. [16] R.G.R.V.P. Kumar, D.D. Vo, S. Chakravarty, S. Chandrasekhar, R. Grée, Bioorg. Med. Chem. Lett. 22 (3) (2012) 1439–1444. [17] Q. LaPlant, S. Chakravarty, V. Vialou, S. Mukherjee, J.W. Koo, G. Kalahasti, K.R. Bradbury, S.V. Taylor, I. Maze, A. Kumar, A. Graham, S.G. Birnbaum, V. Krishnan, H.T. Truong, R.L. Neve, E.J. Nestler, S.J. Russo, Biol. Psychiat. 65 (10) (2009) 874–880. [18] R.D. Porsolt, A. Bertin, M. Jalfre, Arch. Int. Pharmacodyn. Ther. 229 (2) (1977) 327–336. [19] J. Burgdorf, R.A. Kroes, C. Weiss, M.M. Oh, J.F. Disterhoft, S.M. Brudzynski, J. Panksepp, J.R. Moskal, Neuroscience 192 (2011) 515–523. [20] A. Inanobe, H. Furukawa, E. Gouaux, Neuron 47 (1) (2005) 71–84. [21] Glide, version 6.0, Schrödinger, LLC, New York, NY, 2013. [22] Refer patent US 5,763,393. [23] Refer patent WO 2001044089.
Contents lists available at ScienceDirect
Bioorganic Chemistry journal homepage: www.elsevier.com/locate/bioorg
Synthesis and biological evaluation of strained unusual amino acid containing tetrapeptides as potential antidepressant agents Prathama S. Mainkar a,⇑, Sumana Chakravarty b,⇑, Takkallapally Srujana c, Libi Anandi Vishwanathan c, Santosh Kumar Prajapti c, Karisetty Bhanu Chandra b, Lenin Veeraval b, Bathini Nagendra Babu c,⇑ a b c
Division of Natural Products Chemistry, CSIR-Indian Institute of Chemical Technology, Tarnaka, Hyderabad 500007, India Centre for Chemical Biology, CSIR-Indian Institute of Chemical Technology, Tarnaka, Hyderabad 500007, India Department of Medicinal Chemistry, National Institute of Pharmaceutical Education and Research (NIPER), Hyderabad, Balanagar, Hyderabad 500037, India
a r t i c l e
i n f o
Article history: Received 30 June 2015 Revised 24 September 2015 Accepted 26 September 2015 Available online 30 September 2015 Keywords: NMDA receptor GRIN2B Norbornane amino acid Neurotrophic
a b s t r a c t Strained unusual amino acid derived tetrapeptides were synthesized as mimics of GLYX-13, a clinical candidate for neuroprotective and anti-depressant properties, were studied. The synthesized compounds were screened for neurite growth and anti-depressant properties in vitro and in vivo respectively comparing with the parent GLYX-13 compound. Neurite growth property was assessed by neurite length and anti-depressant property by percentage of immobility in forced swim test, a behavioural assay. Mechanistic insights about protein–ligand interactions were obtained using molecular docking study. Based on the in vitro and in vivo screening data and molecular docking study, a new analogue of GLYX-13, Compound 11a has been found to be as good as the parent compound in all respects. Ó 2015 Elsevier Inc. All rights reserved.
1. Introduction Many neuroactive peptides are employed by the brain and nervous system of mammals which help in specialized signalling with central nervous system (CNS). A careful study of the signalling pathways has led to the identification of some specific receptors modulated by these neuroactive peptides which opened up an important therapeutic area in CNS disorders. The N-methyl-Daspartate receptor (NMDAR) is one such receptor which is reported to be involved in various disorders such as cognitive function associated with ageing, ischemia, stroke, head trauma, spinal cord injury, epilepsy, depression, schizophrenia etc. [1]. NMDARs are widely distributed in brain and spinal cord; higher densities are found in cortex and hippocampus. NMDAR ion channels are one of the postsynaptic targets of glutamate-mediated synaptic transmission in mammalian brain. For opening, NMDAR channels require binding of two different types of agonists-glutamate and glycine [2]. Many glutamate receptors are found in CNS but strychnine-insensitive site is unique to NMDAR [3]. Drugs, which are capable of blocking NMDAR activation and glutamate binding site antagonists, create some strong side-effects. As a result, attention has shifted to the glycine site agonists for altering the NMDAR, which is a potentially important target for development of cogni⇑ Corresponding authors. E-mail address: [email protected] (B.N. Babu). http://dx.doi.org/10.1016/j.bioorg.2015.09.007 0045-2068/Ó 2015 Elsevier Inc. All rights reserved.
tive enhancing therapeutics and anticonvulsant and neuroprotective agents. The glycine site of NMDAR complex is a co-agonist site with affinity for glycine and D-serine. It is an allosteric site and hence has been explored for identifying and developing safer drugs. The advantage of partial NMDAR agonists is that they block excessive NMDA function and also potentiate NMDAR in case of abnormally depressed NMDA-mediated neuro-transmission. D-Cycloserine [4] and glycine prodrug, milacemide [5] have shown cognitiveenhancing properties in vivo but result in desensitization with chronic administration. Monoclonal antibody B6B21 has been shown to act as a partial agonist at the glycine site of NMDAR. The hyper variable region of the light chain of B6B21 was cloned and sequenced [6]. Based on this sequence information several peptides were synthesized and screened using rat hippocampal membrane preparations to measure [3H]MK-801 binding in presence of 7-chlorokynuric acid, a glycine site-specific competitive inhibitor of NMDAR. Peptides that increased the binding of [3H]MK-801, an open channel blocker, in dose dependent manner were named Glyxins [7]. Glyxins NT-1 to NT-18 were synthesized which comprised of peptides containing amino acids up to 26. Of these, GLYX-13 (NT-13) [22] (1, Fig. 1) was found to be the best. GLYX-13 is a tetrapeptide (Thr-ProPro-Thr) [23] and is known to readily cross the blood–brain barrier. Clinical data showed that it has robust anti-depressant activity [8], the effect is immediate (within 20 min) and long lasting (at least
54
P.S. Mainkar et al. / Bioorganic Chemistry 63 (2015) 53–57
Fig. 1. Structure of parent compound GLYX-13.
for 2 weeks after single dose) with no CNS-related side-effects [9]. It was also found to enhance learning (cognitive function) [10], help in autism treatment and increase neuroprotection in cases of stroke [11]. GLYX-13 was able to modulate NMDAR in a glycine-like fashion when examined pharmacologically and electro-physiologically. Based on the biological screening data of Glyxins, the optimal length of the peptides was found to be four. We have been working in the area of synthesis of peptides and foldamers consisting of unusual strained b-amino acids and their structural investigations [12]. In addition, the incorporation of a strained b-amino acid derived from sugar into a marine peptide natural product, azumamide, for enhanced activity [13] and many peptides with unusual amino acids have been reported [13]. Inspired by this observation, GLYX-13 skeleton was chosen as a starting point in the synthesis of b-amino acid analogues of GLYX-13, 1, to understand the SAR. A first look at 1 revealed the presence of turninducing Pro–Pro dimer as the central core. This prompted us to initiate incorporation of cis-furan sugar amino acid, 2 [12] and cis-exo-b-norbornene amino acid, 3 and 4 [12] as substitutes of central Pro–Pro core of 1 (Fig. 2). These unusual amino acids have already been proven to be turn inducers as observed by us earlier [12]. The terminal amino acids in 1 are threonine and are expected to participate in hydrogen-bonding. Based on this observation we initiated incorporation of turn-inducing amino acid pairs as core and substituted the terminal amino acids, threonine, with 2-aminobenzoic acid. A preliminary molecular modelling study data indicated that insertion of sugar amino acids will not be able to induce the required turn for the expected activity as they exhibit formation of helical structures [12,13]. Thus the focus then shifted to inserting norbornene amino acid which is a known turn inducer [12]. To determine the impact of stereochemistry on the neuroprotective activity both the isomers of norbornene amino acid viz., (2R,3S)-cis-exo-b-norbornene amino acid, 3 and (2S,3R)-cis-exo-bnorbornene amino acid, 4, were used.
2. Result and discussion 2.1. Chemistry The work on the analogue synthesis began with retention of the proline–proline core with modification in the terminal amino acids. Thus a set of four analogues were synthesized incorporating glycine, 5, 4-aminobenzoic acid 6, b-alanine 7 and 2-aminobenzoic acid 8, following usual peptide synthesis protocols (Fig. 3). Biological screening of these compounds did not yield encouraging results (data not shown). Therefore further attempts of analogue synthesis were shifted to modifying the turn-inducing core.
Fig. 2. cis-Furan sugar amino acid and cis-exo-norbornane amino acid.
Enantiopure norbornene amino acid (both enantiomers exo(2S,3R) and exo-(2R,3S) individually) were used to prepare the analogues 11a, 11b, 13a and 13b [12] N- and C-Protected dimers (9a and 9b) were prepared from both 2S,3R- and 2R,3S-cis-exo-bnorbornene amino acid, 3 and 4, independently using our earlier protocol [12]. These dimers are used to synthesize the proposed analogues 11a, 11b, 13a and 13b by attaching the appropriately protected monomer units of L-threonine and 2-aminobenzoic acid. The trimers were prepared via classical peptide forming HOBt– EDCI protocol which involves coupling of amine of the dimer (readily obtained from dimer 9a and 9b by treating with trifluoroacetic acid) and N-Boc protected L-threonine N-Boc-2-aminobenzoic acid respectively. Similarly, tetramers were prepared from trimer acids coupled with methyl esters of L-threonine and 2-aminobenzoic acid (Scheme 1). All the analogues along with the parent compound 1 were evaluated for their biological efficacy. 2.2. Biological evaluation GLYX-13 (1), a compound approved by FDA in Fast Track designation on 3rd March 2014 as an adjunctive therapy in treatment of resistant major depressive disorder. This molecule completed phase IIb clinical trials and phase III trials will start in 2015. A major problem associated with 1 is its bioavailability and hence it is administered via intraperitoneal route to undergo first-pass effect which is bypassed in intravenous route. As the molecule has shown great promise, analogues are planned which are less prone to enzymatic degradation and increase its bioavailability. The GLYX-13 (1) and its analogues 11a, 11b, 13a and 13b were preliminarily screened in vitro on Neuro-2a cell line (ATCC CCL-131) [14–16] for neurotrophic activity at 0.01, 0.1 and 1.0 lM concentrations and a concurrent MTT cell cytotoxic assay was carried out to substantiate the compounds concentrations are non-toxic (Supplementary Fig. 1 and Supplementary Table 1). All the analogues except 11b showed noticeable neurotrophic activity at the lowest 0.01 lM concentration (Fig. 4 and Supplementary Table 2). But, the BDNF gene expression did not show consistent changes in analogous with the neurite length (Supplementary Fig. 4). This data suggests that the neurotrophic action of 1, an NMDA receptor agonist, and its analogues might not be directly via activation of BDNF gene. Further the analogues were screened in vivo for their anti-depressant activity and were compared with 1 using a behavioural assay, Forced Swim Test [17,18] (Fig. 5). This assay indicated that compounds 11a and 11b had activity similar to 1. With this indication, we explored compatibility of both 11a and 11b to NMDA receptor which could be mediated through GRIN2B or NR2B receptor as 1, based on the literature precedence [9]. Neuro-2a cells were treated with 1.0, 0.1, 0.01 and 0.001 lM concentrations of 11a, 11b, 1 and DMSO. All the three has shown increased GRIN2B gene expression levels at 0.01 lM concentration when compared to other. Further GRIN2B gene expression of 1, 11a and 11b at 0.01 lM were compared and 1 and 11a induced a significant up regulation compared with DMSO treated (Fig. 6). The data suggests that 11a and 11b work on the same pathway as 1, viz., as NMDA receptor agonist and 11a induced more GRIN2B gene expression than others. After analysing both in vitro and in vivo results GLYX-13 analogue 11a seems to be the best lead molecule to consider for further development as a novel antidepressant as this molecule showed significant neurotrophic activity at 0.01 lM concentration as well with the noticeable improvement in immobility time in forced swim test as compared to vehicle treatment (Figs. 4 and 5). The present study also suggests the compound 11a might be working as NMDA receptor agonist like parent compound, GLYX13, in a better way at mRNA gene expression levels. A protein level work would have further substantiated the mRNA results. It also
P.S. Mainkar et al. / Bioorganic Chemistry 63 (2015) 53–57
55
Fig. 3. Analogues of 1 with terminal amino acids modified.
Scheme 1. Synthesis of analogues with cis-exo-b-norbornene amino acid.
reports that the neurotrophic property of the analogues is not directly through BDNF pathway. To conclude about the mechanism of action, more specific molecular level studies need to be performed in future. Especially, it would be interesting to find interlink between neurotrophic nature and anti-depressant properties of the analogues. 2.3. Molecular docking study In order to gain insight about protein–ligand interactions of newly synthesized analogues in the active site of NMDA receptor, molecular docking was performed for compounds 5–8, 11a, 11b, 13a and 13b. Coordinates of NMDA receptor were obtained from
the Protein Data Bank (PDB ID: 1Y1Z) [19,20]. Geometry of the molecules was optimized by using the optimized potentials for liquid simulations-2005 (OPLS-2005) force field. The docking simulations in the active site of NMDA receptor was done with aid of Glide docking tool of Schrödinger suite [21]. For the purpose of docking protocol validation, ligand bound to the crystal structure was removed from the binding pocket and docked back into the same binding site resulted in the same binding interactions. Compounds 5–8, 11a, 11b, 13a, 13b and GLYX13 (reference compound) were docked flexibly into the crystallographic structure of the NMDA receptor using Docking Glide Extra Precision (XP), version 6.0 (Schrödinger, 2013) [21].
56
P.S. Mainkar et al. / Bioorganic Chemistry 63 (2015) 53–57
Fig. 4. Bar representation of the neurite outgrowth of differentiated Neuro-2a cells by GLYX-13 and its analogues at 0.01 lM concentration in comparison with DMSO and NGF (A). Bright field images showing optimum changes in neurite outgrowth of differentiated Neuro-2a cells treated with analogues (B).
Fig. 5. The new set of GLYX-13 analogues and their antidepressant efficacies in Forced swim test (FST).
Docking result showed that the ligands bound in same site formed from Gly249, Arg187, Arg247, Pro140, Phe246, Tyr143, Phe137, Ile127, Pro140, Ala132, Gln144 and Asn 129 amino acids (Fig. 7). Among the synthesized compounds, the calculated docking score for the compound 11a and 11b were 3.9 and 4.5 kcal/mol respectively, which are in agreement with experimental assay. Compound 11a and 11b showed hydrogen bonding similar to GLYX-13 with Arg187, Arg 247 and Gly249. Also, compound 11a displayed hydrogen bonding with Asn129 and Phe137. Beside this, docking pose of compound 11a and 11b showed hydrophobic interaction with Ile127, Phe246, Ala132 and Pro140 in addition to Vander Waals interactions with other residue of enzyme. All these interactions together contribute to the hypothesis that the compounds bind efficiently to the active site of NMDA receptor, thus ensuing antidepressant activity as in case of GLYX-13.
Fig. 6. Fold change of GRIN2B gene expression in Neuro-2a cells treated with GLYX-13, 11a and 11b at concentrations 1.0–0.001 lM with DMSO treated cells as control (A, B and C). Logarithmic representation of fold change of GRIN2B gene expression in Neuro-2a cells treated with compounds GLYX-13, 11a and 11b at 0.01 lM concentration with DMSO treated cells as control (B). Here ⁄p < 0.05 and ⁄⁄⁄p < 0.001, n = 5–6.
P.S. Mainkar et al. / Bioorganic Chemistry 63 (2015) 53–57
57
Fig. 7. Top view of 3-D image showing interactions of compound 11a, 11b and GLYX-13 with NMDA receptor (PDB ID: 1Y1Z).
Moreover, the results for the compounds 5–8, 13a and 13b were less favourable than the calculated docking score of the GLYX-13, which was 4.2 kcal/mol (Supplementary Table 5).
3. Conclusion In summary, we have designed, synthesized, and identified novel analogues of GLYX-13, a compound in clinical trials for treatment of depression and cognitive disorders. Based on very encouraging results from the in vitro and in vivo studies and molecular docking study, compound 11a has been found to be as good as the parent compound in all respects. This study also establishes the fact that proline–proline core can be replaced with turn inducing b-amino acids which are less prone to enzymatic degradation. Acknowledgments Authors thank National Institute of Pharmaceutical Education and Research (NIPER)-Hyderabad; and CSIR-Indian Institute of Chemical Technology-Hyderabad for providing facilities and Department of Pharmaceuticals (DoP); Government of India (GoI) for financial support. Authors like to thank Dr Srivari Chandrasekhar for his suggestions and extended support. BN likes to acknowledge Dr K. Srinivas for helpful discussion in molecular docking study. SC acknowledges the support of CSC0111/SMiLE for biological assays.
Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.bioorg.2015.09. 007.
References [1] M. Hoffman, Science 254 (5033) (1991) 801–802. [2] A.P. Kozikowski, W. Tuckmantel, I.J. Reynolds, J.T. Wroblewski, J. Med. Chem. 33 (6) (1990) 1561–1571. [3] P.L. Wood, IDrugs 8 (3) (2005) 229–235. [4] W.F. Hood, R.P. Compton, J.B. Monahan, Neurosci. Lett. 98 (1) (1989) 91–95. [5] G.E. Handelmann, M.E. Nevins, L.L. Mueller, S.M. Arnolde, A.A. Cordi, Pharmacol. Biochem. Behav. 34 (4) (1989) 823–828. [6] P.K. Stanton, J.M. Sarvey, J.R. Moskal, Proc. Natl. Acad. Sci. USA 84 (6) (1987) 1684–1688. [7] X.L. Zhang, J.A. Sullivan, J.R. Moskal, P.K. Stanton, Neuropharmacology 55 (7) (2008) 1238–1250. [8] J. Burgdorf, X.L. Zhang, C. Weiss, E. Matthews, J.F. Disterhoft, P.K. Stanton, J.R. Moskal, Neurobiol. Aging 32 (4) (2011) 698–706. [9] J. Burgdorf, X.L. Zhang, K.L. Nicholson, R.L. Balster, J.D. Leander, P.K. Stanton, A. L. Gross, R.A. Kroes, J.R. Moskal, Neuropsychopharmacology 38 (5) (2013) 729– 742. [10] P.K. Stanton, P.E. Potter, J. Aguilar, M. Decandia, J.R. Moskal, Neuroreport 20 (13) (2009) 1193–1197. [11] J.R. Moskal, A.G. Kuo, C. Weiss, P.L. Wood, A. O’Connor Hanson, S. Kelso, R.B. Harris, J.F. Disterhoft, Neuropharmacology 49 (7) (2005) 1077–1087. [12] S. Chandrasekhar, M.S. Reddy, B.N. Babu, B. Jagadeesh, A. Prabhakar, B. Jagannadh, J. Am. Chem. Soc. 127 (27) (2005) 9664–9665. [13] S. Chandrasekhar, C.L. Rao, M. Seenaiah, P. Naresh, B. Jagadeesh, D. Manjeera, A. Sarkar, M.P. Bhadra, J. Org. Chem. 74 (1) (2009) 401–404. [14] Z. Khaing, D. Kang, A.M. Camelio, C.E. Schmidt, D. Siegel, Bioorg. Med. Chem. Lett. 21 (16) (2011) 4808–4812. [15] G. Mehta, R. Samineni, P. Srihari, R.G. Reddy, S. Chakravarty, Org. Biomol. Chem. 10 (34) (2012) 6830–6833. [16] R.G.R.V.P. Kumar, D.D. Vo, S. Chakravarty, S. Chandrasekhar, R. Grée, Bioorg. Med. Chem. Lett. 22 (3) (2012) 1439–1444. [17] Q. LaPlant, S. Chakravarty, V. Vialou, S. Mukherjee, J.W. Koo, G. Kalahasti, K.R. Bradbury, S.V. Taylor, I. Maze, A. Kumar, A. Graham, S.G. Birnbaum, V. Krishnan, H.T. Truong, R.L. Neve, E.J. Nestler, S.J. Russo, Biol. Psychiat. 65 (10) (2009) 874–880. [18] R.D. Porsolt, A. Bertin, M. Jalfre, Arch. Int. Pharmacodyn. Ther. 229 (2) (1977) 327–336. [19] J. Burgdorf, R.A. Kroes, C. Weiss, M.M. Oh, J.F. Disterhoft, S.M. Brudzynski, J. Panksepp, J.R. Moskal, Neuroscience 192 (2011) 515–523. [20] A. Inanobe, H. Furukawa, E. Gouaux, Neuron 47 (1) (2005) 71–84. [21] Glide, version 6.0, Schrödinger, LLC, New York, NY, 2013. [22] Refer patent US 5,763,393. [23] Refer patent WO 2001044089.