Advertisement

Comparative docking and ADMET study of some curcumin derivatives and herbal congeners targeting β-amyloid

  • Dev Bukhsh SinghEmail author
  • Manish Kumar Gupta
  • Rajesh Kumar Kesharwani
  • Krishna Misra
Original Article

Abstract

Alzheimer’s disease (AD) is caused by the accumulation of beta-amyloid (β-A) in the brain that forms amyloid plaque. β-A is an oligo-peptide consisting of 39–42 amino acid residues, produced by proteolytic cleavage of amyloid precursor protein by secretase enzymes. Evolutionary trace and protein family analysis of β-A indicates that C-terminal residues of β-A are highly conserved and hydrophobic in nature. Prevalence of hydrophobic residues at C-terminal of β-A promote aggregation, and also provide the stability to β-A plaque due to hydrophobic–hydrophobic interaction between residues of β-A. In this work, designing, evaluation and screening of potent inhibitors for β-A formation were studied. Curcumin and some of its herbal congeners were taken into consideration, to evaluate their anti-alzheimeric property against β-A as well as with β-A fibrils. Most of these herbal compounds were found potent inhibitors of β-A aggregation than known drugs for AD. The binding affinity of cassumunins A and B was compared with that of curcumin and it was found that cassumunins A and B may cause more potent inhibition of β-A than curcumin. Principal descriptors as well as absorption, distribution, metabolism, excretion and toxicity (ADMET) properties for these compounds were predicted, and were found satisfactory.

Keywords

Alzheimer’s Disease Beta-amyloid Cassumunin Curcumin ADMET Descriptors 

1 Introduction

AD is a neuro-degenerative disease, which causes degeneration of nerve cell due to accumulation of β-A in brain tissues and creates problem by interrupting transmission of the brain’s signals. A person with Alzheimer’s disease has problems with memory, judgment, and thinking (Fiala 2007). Alois Alzheimer observed a relationship between cognitive loss and the presence of plaque of β-A in the brain of AD patients.

The main cause of AD is accumulation of a protein called β-A formed by enzymatic cleavage of amyloid precursor protein (APP). Aggregation of β-A in human brain results in tangled brain tissue and formation of dense areas called plaques. A major component of amyloid plaques found in AD patients is ~4 kDa β-A peptide (Cam and Bu 2006). Peptides formed by proteases are composed of 39–42 residues, but β-A with 42 residues forms major composition of plaque in the brain of AD patients (Kung et al. 2003). The annular structures of beta-amyloid (17–42 residues) are significant in β-A fibril formation (Zheng et al. 2008). Hydrophobic residues present at C-terminal position of β-A 42, are sufficient to promote fibril formation (Kim and Hecht 2006). β-A undergoes conformational changes and forms β-A fibrils. These β-A fibrils are toxic for the brain and cause loss of memory leading to AD. Statistical models were developed for detection of neurological disease in rats through locomotion analysis (Tang et al. 2012). Locomotion analysis, as an important division of behavioral tests, has been widely used in diagnostic decision making, comparative biomechanics, biometric identification and forensics (Tang and Su 2012)

One way to cure AD is to inhibit the pathway that leads to β-A production in the brain. Curcumin is a yellow color pigment present in the rhizome of the curry spice turmeric (Curcuma longa) plant which occurs along with minor amounts of its mono and dimethoxy derivatives (Fig. 1). Curcumin is a polyphenol compound known for its antitumor, antioxidant, anti-amyloid, anti-inflammatory and a plethora of other therapeutic properties (Ruby et al. 1995; Ono et al. 2004; Srimal and Dhawan 1973). Curcumin is herbal in origin so it has negligible risk of toxicity and side effects. It was observed that aggregation of β-A protein was inhibited in mice that were fed curcumin compared with control mice. It was observed that curcumin can inhibit formation of new plaques, as well as break up or disaggregate existing plaques in Alzheimer’s brain (Yang et al. 2005).
Fig. 1

Curcumin: R=R′=OCH3, R″=H; Bisdemethoxy curcumin: R=R′=R″=H

Certain amyloid-binding molecules, such as Congo red (CR), chrysamine G (CG) and thioflavin S (TS) have been shown to bind with amyloid plaques with high affinity and they can also prevent the formation of β-A fibrils. CR, CG and TS were not found suitable for treatment of AD because they are not able to cross the blood brain barrier (BBB) (Lee 2002). Therefore, for a compound to serve as a drug specifically for Alzheimer, it should not only qualify the ADMET rule for drugs but also cross BBB. Because of lipophilic nature, curcumin crosses the BBB and inhibits the aggregation of β-A as well as causes the disruption of β-A fibrils (Mishra and Palanivelu 2008).

Curcumin reduces the concentration of insoluble β-A, soluble β-A and amyloid plaque by 43–50 %. However, level of amyloid precursor protein (APP) in the membrane fraction remains same (Lim et al. 2001). It has also been reported that curcumin has potent anti-amyloidogenic property (Ringman et al. 2005). β-A can be used as a potential target for anti-alzheimeric drugs. Curcumin-derived oxazoles and pyrazoles were found to be potent inhibitors of gamma-secretase (Narlawar et al. 2007). It was also reported that NSAIDS (ibuprofen and naproxen drugs) and curcumin prevent AD in animal models and in vitro (Cole et al. 2004) which might be due to its role in anti-aggregation of β-A (Agdeppa et al. 2003).

The numbers of individuals affected by AD are increasing rapidly, which suggests urgent need for development of effective treatment strategy for AD patients. β-A can serve as a potential target for inhibition of β-A aggregation and curcumin has been reported to be a potent inhibitor for β-A aggregation. The combination of these two factors motivated us for in silico study of curcumin and its herbal congeners in terms of binding and docking energies which is likely to help in designing a new synthetic analog of curcumin and its herbal congeners as a potent inhibitor of β-A aggregation. Most of compounds taken for study are natural derivative or congeners of curcumin which strongly motivates to use it as a lead molecule for enhancing its biological activities. Principal descriptor and ADMET properties of potent inhibitors are also predicted for assessing efficacy, inhibition capability, absorption, distribution, metabolism and toxicity response of drug after its administration in body.

2 Materials and methods

2.1 Target and cavity prediction

3D structure files of Alzheimer’s β-A (PDB file: 1IYT) and β-A fibrils (PDB file: 2BEG) were retrieved from Protein Database (PDB) (http://www.rcsb.org. Both 1IYT and 2BEG contain 10 models of NMR-determined structures (Lührs et al. 2005; Crescenzi et al. 2002).

CASTp (Computed Atlas of Surface Topography of proteins) was used for prediction of pockets and cavities present in β-A. CASTp is a web server that aims to provide a detailed quantitative characterization of interior cavities and surface pockets of proteins, which are prominent concave regions of proteins that are frequently associated with binding events (Dundas et al. 2006). Input to the CASTp is four letter PDB file. Probe radius 1.4 Å (default value) was used for cavity prediction of β-A. Only values of probe radius between 0.0 and 10.0 Å are accepted for cavity prediction. The whole methodology used in this study is shown in Fig. 2.
Fig. 2

Flow diagram of methodology used for study

2.2 Evolutionary study

Evolutionary Trace method was used to find out binding surfaces of β-A that are common in protein families. It provides an evolutionary perspective for determining the functional or structural role of each residue (amino acid) in protein structure. The evolutionary trace report maker is a server developed by Lichtarge Computational Biology lab (Mihalek et al. 2006). It is the easiest way to find out information about functional residues present in β-A. It takes a Protein Data Bank identifier or a UniProt accession number as it’s only input.

Protein family database analysis (Pfam) was done to identify the conserved domain and hydrophobic residues of β-A. Pfam is a database of protein domain families (Punta et al. 2012). Pfam is a collection of multiple sequence alignments for each family and is based on hidden markov models (HMMs) profile for finding domains in query sequences.

2.3 Retrieval and preparation of ligand molecules

Physico-chemical properties and 2-D structure of curcumin and its herbal analogs were retrieved from PubChem database of NCBI (National Center for Biotechnology Information) (http://pubchem.ncbi.nlm.nih.gov). Chemical formula, SMILES (simplified molecular input line entry system) notation, molecular weight, logP (partition coefficient), hydrogen bond donor and hydrogen bond acceptor information for these compounds is stored in PubChem database. Compound Id (CID) can be used to retrieve more detailed information about these compounds. CORINA version 3.1 is an on-line server that automatically generates three dimensional (3D) atomic coordinates from the constitution of a molecule. Input to the CORINA server is SMILES notation of molecule (http://www.molecularnetworks.com). ACD/ChemSketch was used to draw the structure of herbal analogs (http://www.acdlabs.com). It is a chemical drawing and graphics software package from ACD/Labs developed to draw molecules, reactions, schematic diagrams and calculate chemical properties. 3D structure of cyclocurcumin (Adhikary et al. 2011) was designed using this tool. Two dimensional structure of all the chemical compounds were represented using Chemdraw tool (Mills 2006).

2.4 Docking

Molegro Virtual Docker (MVD) 2007.2.0.0 (Thomsen and Christensen 2006) was used for flexible docking study. MVD requires a 3D structure of both protein and ligand. Binding mode of ligand is determined by iteratively evaluating a number of ligand conformations and estimating the energy of their interactions with target molecules (Thomsen and Christensen 2006). Candidates with the best conformational and energetic results were selected. MVD was used to calculate the interaction energies between ligands and macromolecular systems from 3D structures of protein and ligands. MolDock score, an adaptation of the differential evolution (DE) algorithm, was used for energy calculation (Thomsen and Christensen 2006). MolDock score energy, E score, is defined by Eq. (1), where E inter is the ligand–protein interaction energy and E intra is the internal energy of the ligand. E inter is calculated according to Eq. (2).
$$ E_{\text{score}} = E_{\text{inter}} + E_{\text{intra}} $$
(1)
$$ E_{\text{inter}} = \mathop \sum \limits_{{i {\text{ligand}}}} \mathop \sum \limits_{{j {\text{protein}}}} \left[ {E_{\text{PLP}} \left( {\gamma_{ij} } \right) + 332.0\frac{{q_{i} q_{j} }}{{4r_{ij}^{2} }}} \right] $$
(2)
The E PLP term is a ‘‘piecewise linear potential’’ (Yang and Chen 2004) using two different parameters, one for the approximation of the steric term such as van der Waals between atoms and another for the potential for hydrogen bonds; it describes the electrostatic interactions between charged atoms (Thomsen and Christensen 2006). E intra is defined by Eq. (3).
$$ E_{\text{intra}} = \mathop \sum \limits_{{i {\text{ligand}}}} \mathop \sum \limits_{{j {\text{protein}}}} \left[ {E_{\text{PLP}} \left( {\gamma_{ij} } \right)} \right] + \mathop \sum \limits_{{{\text{flexible}}\;{\text{bonds}}}} A\left[ {1 - \cos \left( {m\theta - \theta } \right)} \right] + E_{\text{clash}} $$
(3)

The first term in Eq. (3) calculates the total energies involving pairs of atoms of the ligand, except those linked by two bonds. The second term stands for torsional energy, where θ is the torsional angle of the bond. The average of torsional energy bond contributions is used if several torsions have to be determined. The term, E clash, defines a penalty of 1,000 kcal/mol if the distance between two heavy atoms (more than two bonds apart) is smaller than 2.0 Å, ignoring infeasible ligand conformations (Thomsen and Christensen 2006).

2.5 ADMET prediction

Schrodinger software (QikProp v3.3) was used for calculation of principal descriptors and prediction of ADMET. The BOSS program and OPLS-AA force field were used to perform Monte Carlo statistical mechanics simulations on organic solutes in periodic boxes of explicit water molecule (Jorgensen 1998). This process resulted in configurational averages for a number of descriptors, including hydrogen bond counts and solvent-accessible surface area (SASA). Correlation of these descriptors to experimentally determined properties was obtained, and then algorithms that mimic the full Monte Carlo simulations and produce comparable results were developed. Structural data file (SDF) format of compounds was used as input to QikProp. Output of QikProp is consisting of a number of principal descriptors and ADMET predictions.

3 Results and discussion

3.1 Binding site prediction

Binding sites and active sites of proteins are commonly related with structural pockets and cavities. CASTp server uses the weighted Delaunay triangulation and the alpha complex for shape measurements. It helps in identification and measurements of surface accessible regions (pockets) and interior inaccessible cavities for β-A. PBD file of Alzheimer’s β-A i.e. 1IYT was uploaded to CASTp server. Cavity predictions by CASTp have shown the presence of three pockets 1, 2 and 3. Diagrammatic representation of each pocket present in β-A and their corresponding amino acid is given in Fig. 3. Pockets are empty concavities on surface of macromolecules such as receptor, enzyme, etc., into which solvent (probe sphere 1.4 Å) can gain contact.
Fig. 3

Pockets 1, 2 and 3 predicted in β-A are shown by cyan, blue and green color, respectively. Amino acid residues belonging to each pocket are also shown by respective color (color figure online)

Area and volume of solvent accessible surface (SA) and molecular surface (MS) for each pocket of β-A is listed below in Table 1. Area, volume and length measurement for pocket 3 is greater than pockets 1 and 2. Amino acid residue belonging to pocket 3 may be associated with binding. As in 85 % cases, the pocket with largest area and volume is related to binding site. The binding pockets were also identified using the MVD tool during docking. The cavity with largest size and volume were specified as docking site.
Table 1

Area and volume of solvent accessible surface (sa) and molecular surface (ms) for each pocket of β-A

S. no.

Pocket number

Area_sa (Å2)

Area_ms (Å2)

Vol_sa (Å3)

Vol_ms (Å3)

1.

Pocket 1

1.989

22.31

0.125

14.60

2.

Pocket 2

2.424

24.03

0.247

16.58

3.

Pocket 3

2.819

61.11

0.127

35.76

Solvent accessible surface area for pocket 3 is larger among all pockets. Similarly, molecular surface area and volume of pocket 3 is larger as compared to pockets 1 and 2. Solvent accessible surface area and molecular surface area for pocket 1 is smaller than other pockets as displayed in Fig. 4.
Fig. 4

a Area (Area_sa) and volume (Vol_sa) of solvent accessible surface in angstrom unit. b Area (Area_ms) and volume (Vol_ms) of molecular surface for three pocket of β-A

3.2 Structural conservation of β-A

For chain A of β-A, the alignment of β-A (1iytA.msf) with 13 sequences were done. The alignment was downloaded from the HSSP database, and fragments shorter than 75 % of the query as well as duplicate sequences were removed. Furthermore, 45 % of residues have shown as conserved in this alignment. The alignment consists of 84 % eukaryotic (84 % vertebrata) sequences. Each residue in β-A sequence is colored according to its estimated importance. Motivation behind this study is to find out the significance of β-A residues in its aggregation into β-A fibrils.

Trace results are commonly expressed in terms of coverage: residue is important if its coverage is small (Fig. 5a). Importance of residues in β-A is shown by bright red and yellow color which indicates that residues falling in these regions are more important, and conserved during course of evolution (Fig. 5b). Top 45 % of all residues, clustered in two clusters C1 followed by C2 are the largest clusters. Residues belonging to each cluster are listed in Table 2. C1 is largest cluster having 16 residues with their position in chain A of β-A. Only two residues at position 17 and 19 belong to second cluster, i.e., C2. Evolutionary trace analysis reveals that the residues present in C-terminal position (26–42) in β-A, are highly conserved during course of evolution and functionally important.
Fig. 5

a Residues in β-A (1–42) colored by their relative importance, residues shown in dark color are more important. b Coloring scheme used to show relative importance of residues in β-A (color figure online)

Table 2

Clustering of top ranking residue in β-A, chain-A

S. no.

Cluster

Size

Member residue position in β-A

1.

C1

16

26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 40, 41, 42

2.

C2

2

17, 19

Protein family study shows that residues present at C-terminal position of β-A are conserved throughout family and hydrophobic stretches are conserved and functionally important. Pfam studies have shown that C-terminal residues of beta-amyloid are highly conserved and hydrophobic in nature. These hydrophobic residues at C-terminal of β-A may promote aggregation of β-A. Conservation, quality and consensus of residues in protein family search of β-A is shown in Fig. 6.
Fig. 6

Hydrophobic residue (red color) present in chain A of β-A, conservation, quality and consensus of residues in β-A obtained by Pfam analysis (color figure online)

3.3 Comparative docking study of known drugs, curcumin derivatives and herbal congeners with β-A and β-A fibrils

There are two known commercial drugs of AD targeting β-A, Tramiprosate and AZD-103 or cyclohexanehexol (Santa-Maria et al. 2007; McLaurin et al. 2006). Tramiprosate is used to block the aggregation of β-A into plaque. In silico binding analysis of these drugs with β-A was done by MVD to evaluate their binding affinity. Accuracy of different docking tools has been evaluated by docking flexible ligands to 77 protein targets and it has been found that MVD was able to identify the correct binding mode of ligands in 87 % of the complexes, as compared to Glide (82 %), Surflex (75 %), FlexX (58 %) and GOLD (78 %) (Thomsen and Christensen 2006). Real binding site of curcumin for β-A has not been reported so far. Inhibition of β-A and β-A fibrils by curcumin has been studied and it was found that curcumin causes inhibition of β-A (IC50 = 0.8 μM) as well as disaggregation of β-A fibrils (IC50 = 1 μM) (Yang et al. 2005). MVD also determines volume of possible binding site before docking and cavity with largest volume is selected as docking point.

Population of Indian subcontinent has very low risk of AD as compared to other continent of the world (Boersma et al. 1998; Brayne et al. 1998; Azzimondi et al. 1998; Shaji et al. 1996; Yamada et al. 1999; Woo et al. 1998; Lin et al. 1998; Fillenbaum et al. 1998; Herrera et al. 1998; Lopes et al. 2007). This may be due to regular consumption of curry spice turmeric and other herbal congeners by Indian population. Worldwide prevalence (percentage) of dementia or AD has been shown in Fig. 7. Indians have low risk of AD as compared to other countries of the world; this may be due to regular consumption of turmeric and other herbal compounds in diet.
Fig. 7

World wide prevalence of dementia or AD, from 1996 to 1999

Curcumin is a well-known herbal compound that shows better binding with β-A and prevents its aggregation to form amyloid plaque and also dissolves already formed β-A fibrils (Yang et al. 2005). For abrogation of toxicity, the complete disruption of β-A fibril formation is not required (Ghanta et al. 1996). Curcumin derivatives and other herbal congeners of curcumin were evaluated for their binding affinity with β-A and β-A fibrils. Our aim was to screen those compounds that show better binding with β-A and β-A fibrils. Curcumin derivatives taken for study are as follows: curcumin, cyclocurcumin, bisdemethoxy curcumin, curcumin dimethyl ether, allyl curcumin, curcumin bis-acetate, [18F]fluoropropyl-substituted curcumin. Similarly, curcumin congeners such as piperine, caffeic Acid, cassumunin A, cassumunin B, chlorogenic acid, dehydrozingerone, dibenzoylmethane, ferulic acid, chalconeonoid and yakuchinone-A were also studied. A list of all the compounds and their binding energies with β-A and β-A fibrils are given in Table 3.
Table 3

Binding energies of commercial drugs for alzheimer, curcumin, its derivatives and herbal congeners with β-A

S. no.

Compounds

PubChem CID

Binding energies with β-A (kcal/mol)

Binding energies with β-A fibrils (kcal/mol)

Total energy

H-bond score

v.d.w. (LJ 12-6)

Steric score

Total energy

H-bond score

v.d.w. (LJ 12-6)

Steric score

1

Tramiprosate

1,646

−37.36

−0.122

−9.99

−36.07

−71.99

−6.08

−15.17

−64.62

2

AZD-103

892

−19.54

−14.35

−3.4

−39.00

−48.08

−13.34

−11.83

−68.54

3

Curcumin

969,516

−73.11

−0.511

−15.59

−84.82

−129.63

−1.61

108.09

−147.63

4

Bisdemethoxy curcumin

5,315,472

−67.40

0

−24.68

−77.01

−159.42

−4.60

−30.80

−166.23

5

Curcumin dimethyl ether

6,477,182

−72.08

−1.60

−13.08

−93.26

−126.87

−2.55

195.68

−151.89

6

Allyl curcumin

16,727,530

−78.05

−2.5

4.18

−75.86

−14.54

0

1,085.62

−47.75

7

Curcumin bis-acetate

6,441,419

−75.88

−2.44

−5.37

−65.51

−30.81

−2.99

916.54

−60.48

8

[18F]Fluoropropyl-substituted curcumin

11,947,775

−86.65

−2.43

−20.52

−105.3

−82.44

−2.5

620.23

−101.69

9

Piperine

638,024

−78.19

−4.21

−19.01

−84.09

−125.53

0

13.57

−134.60

10

Caffeic acid

689,043

−65.50

−9.579

−14.90

−63.94

−95.02

−7.60

12.44

−94.50

11

Cassumunin A

10,460,395

−92.55

−0.853

−18.65

−69.06

44.24

0

2,412.12

21.23

12

Cassumunin B

10,054,109

−91.58

−3.59

−22.06

−75.51

86.16

0.27

2,742.56

44.52

13

Chlorogenic acid

1,794,427

−67.03

−8.53

−24.67

−88.17

−104.08

−7.77

66.26

−129.82

14

Cyclocurcumin

N/A

−74.01

−6.27

−3.10

−85.93

−90.99

−0.59

429.55

−113.73

15

Dehydrozingerone

14,121

−66.02

−1.63

−20.55

−73.02

−101.96

−4.43

−25.75

−106.41

16

Dibenzoylmethane

8,433

−75.00

0.796

−14.34

−77.18

−73.42

0

536.14

−83.78

17

Ferulic acid

445,858

−61.25

−2.32

−19.83

−71.80

−93.87

−5.36

−17.25

−102.81

18

Chalcone (benzalacetophenone)

637,760

−72.25

−2.5

−20.61

−81.45

−117.08

−2.5

−32.57

−125.54

19

Yakuchinone-A

133,145

−64.76

0

−2.45

−77.51

−104.25

0

72.73

−136.22

Curcumin and its derivatives have shown very good binding affinity for β-A, and can potentially inhibit the clumping of β-A. Herbal congeners of curcumin such as cassumunin A, cassumunin B, piperine, dibenzoyalmethane and cyclocurcumin have shown very high affinity to bind with β-A than curcumin. This study indicates that there are other herbal compounds than curcumin, which can prevent the formation of β-A fibrils from β-A. Curcumin and its other herbal congeners have shown better binding affinity with β-A (−61.25 to −92.55 kcal/mol) than known drugs of AD such as tramiprosate (−37.36 kcal/mol).

Cassumunins A and B are phenylbutanoids obtained from tropical ginger, Zingiber cassumunar (Nakamura et al. 2009). Cassumunins A and B were found to have high in silico binding affinity for β-A. Cassumunin A has shown very high binding affinity (bonding energy: −92.55 kcal/mol) for β-A. Cassumunin B has also shown excellent binding affinity (bonding energy: −91.58 kcal/mol) equivalent to cassumunin A. Cassumunin A was found to be involved in strong binding interaction with residues His13, Lys16, Leu17, Phe20 and Ala21 of β-A. A single hydrogen bonding interaction of cassumunin A was observed with residue His13 of β-A (bond energy: −0.853 kcal/mol and bond length: 3.240). Binding interaction of cassumunin B is similar to cassumunin A, as it utilizes the same residues His13, Lys16, Leu17 and Phe20 of β-A for docking. Hydrogen bonding interaction was more significant in cassumunin B than cassumunin A, due to the presence of three hydrogen bonds. Out of three hydrogen bonds in case of cassumunin B, two contributed by residue His13 (bond energy: −0.632 and −0.556 kcal/mol) and one by residue Lys16 (−2.404 kcal/mol) of β-A. Docked view of cassumunins A and B with β-A and its hydrogen bonding patterns are shown in Fig. 8. Coloring schemes used to represent the interaction are given as follows: hydrogen atoms (H): white, carbon atoms (C): light dark, oxygen atoms (O): red, nitrogen atoms (N): blue, single bonds: single black line, double bonds: two red line and hydrogen bonding: bright green dash line.
Fig. 8

a Docked view of cassumunin A with residues His13, Lys16, Leu17, Phe20, Ala21 of β-A and hydrogen bonding interaction with His13. b Docked view of cassumunin B with residues His13, Lys16, Leu17, Phe20 of β-A and hydrogen bonding interaction with His13 and Lys16 (color figure online)

The anti-oxidative effects of cassumunins A and B were compared with those of curcumin and it was found that cassumunins A and B are more anti-oxidant than curcumin (Nagano et al. 1997). Cassumunins A and B have shown stronger protective activity than curcumin against oxidative cell death induced by hydrogen peroxide in a rat (Masuda et al. 1998). Docking study also suggests that cassumunins A and B may be more potent inhibitors of β-A aggregation than curcumin. It is interesting to note that chemical structure of curcumin is included in those of cassumunins A and B.

Docking study of all the compounds with β-A fibrils have shown that bisdemethoxy curcumin (−159.42 kcal/mol), curcumin (−129.63 kcal/mol), curcumin dimethyl ether (−126.87 kcal/mol), piperine (−125.53 kcal/mol), chalcone (−117.08 kcal/mol) and chlorogenic acid (−104.08 kcal/mol) have very high binding affinity with β-A fibrils than β-A. But cassumunins A and B were not showing favorable interaction with β-A fibrils. Docking studies reveal that these compounds can prevent the aggregation of β-A, and may also have capability to dissolve already existing β-A fibrils.

Bisdemethoxy curcumin has very favorable binding with β-A fibrils, and has two hydrogen bonding interaction. Bisdemethoxy curcumin has shown interaction with all the five chains of β-A fibrils (Fig. 9). Chains and amino acid residues involved in interaction are as follows: A (Leu17, Val18, and Phe19), B (Leu17, Val18, and Phe19), C (Leu17, Val18, Phe19, and Val40), D (Leu17, Phe19, Gly38, and Val40) and E (Leu17and Phe19). Leu17 residue of chain C (energy: −2.3978 kcal/mol) and chain D (energy: −2.211 kcal/mol) has been involved in hydrogen bonding with bisdemethoxycurcumin. Docking energies for most of the compounds with β-A fibrils were found very favorable as compared to β-A (Fig. 10).
Fig. 9

Binding interaction of bisdemethoxycurcumin with β-A fibrils along with contribution of residue Leu17 of chain C and D in hydrogen bonding

Fig. 10

Comparative binding of commercial drugs, curcumin, its derivatives and herbal congeners with β-A and β-A fibrils

3.4 Principal descriptors and ADMET analysis

Curcumin and other herbal compounds have shown good preventive role in AD. ADMET properties of curcumin, cassumunins A and B, piperine, dibenzoyalmethane, allyl curcumin and curcumin bis-acetate were predicted, and compared with the ADMET of two known drugs, tramiprosate and AZD-103. The principal descriptors and ADMET properties were calculted by QikProp software. Principal descriptors for inhibitors (Fig. 11) and its normal range in 95 % of drugs are given in Table 4. Following 12 principal descriptors are included in the study: molecular weight (MW), total solvent accessible surface area (SASA), hydrophobic SASA (FOSA), hydrophilic SASA (FISA), carbon Pi SASA (PISA), weakly polar SASA (WPSA), molecular volume (MV), van der Waals polar SA (PSA), no. of rotatable bonds (rotatB), donor-hydrogen bonds (donorHB), acceptor-hydrogen bonds (accptHB) and globularity (Glob).
Fig. 11

Values of principal descriptors for potent inhibitors

Table 4

Principal descriptors for compounds and its normal range in 95 % of drugs

S. no

Principal descriptors

Tramiprosate

AZD-103

Cassumunin A

Cassumunin B

Piperine

Dibenzoyalmethane

Allyl curcumin

Curcumin

Curcumin bis-acetate

Range 95 % of drugs

1

MW

180.157

139.169

558.627

588.653

285.342

224.259

448.515

368.385

452.460

130/725

2

SASA

346.594

320.035

1,009.203

1,045.666

553.283

488.147

882.607

692.604

824.628

300/1,000

3

FOSA

96.032

117.942

525.395

612.952

331.614

20.039

465.788

252.250

402.107

0/750

4

FISA

250.562

199.819

136.108

133.109

41.712

78.396

81.503

156.272

167.174

7/330

5

PISA

0.000

0.000

347.700

299.606

179.957

389.712

335.316

284.082

255.347

0/450

6

WPSA

0.000

2.274

0.000

0.000

0.000

0.000

0.000

0.000

0.000

0/175

7

MV

551.206

475.955

1,808.927

1,888.111

955.706

805.529

1,543.099

1,199.485

1,456.188

500/2,000

8

PSA

126.862

88.971

116.540

123.333

47.983

52.525

80.449

107.172

137.532

7/200

9

RotatB

6.000

5.000

17.000

18.000

5.000

4.000

16.000

12.000

12.000

0/15

10

DonorHB

6.000

3.000

2.000

2.000

0.000

0.000

0.000

2.000

0.000

0/6

11

AccptHB

10.200

5.000

8.500

9.250

4.500

4.000

7.000

7.000

10.500

2/20

12

Glob

0.938

0.921

0.711

0.707

0.848

0.858

0.732

0.788

0.753

0.75/0.95

MW for cassumunins A and B was found above 500, and violating the Lipinski’s rule of five. There are many compounds having the MW greater than 500, but still have medicinal importance and used as drug. Anti-oxidant property of cassumunins A and B is well reported in the literature (Nagano et al. 1997; Masuda et al. 1998). Total solvent accessible surface area (SASA in Å2) was calculated using a probe with a 1.4 Å radius. Normal range of total SASA in 95 % of drugs varies from 300 to 1,000 Å2. SASA values for cassumunin A (1,009.203 Å2) and cassumunin B (1,045.666 Å2) has been found higher than the normal range. This may be due to its larger size as compared to other compounds. Similarly, normal range of MV varies from 500 to 2,000 Å3, which has been found very low (475.955 Å3) in drug AZD-103. MV for most of the compounds has been found within satisfactory range. PSA is formed by van der Waals surface area of polar nitrogen and oxygen atoms. It predicts the passive transport of molecules through membranes and is related with the bioavailability of drug. For a molecule to have good oral bioavailability, PSA should be less than 140 Å2 (Ertl et al. 2000). Globularity descriptor (4πr2)/(SASA), where r is the radius of a sphere with a volume equal to the molecular volume. Globularity is 1.0 for a spherical molecule. Normal range of Glob for most of the drugs is 0.75–0.95. Glob values for cassumunins A and B were found within defined normal limit which has been predicted as 0.711 and 0.707, respectively. In majority of cases, the values of principal descriptors were observed within normal range for drugs and satisfactory.

A large number of ADMET properties such as polarizability in cubic angstroms (Polrz), logP for hexadecane/gas (logP C16), logP for octanol/gas (logP oct), logP for water/gas (logP w), logP for octanol/water (logP o/w), logS for aqueous solubility (logS), logS-conformation independent (CIlogS), logK hsa serum protein binding (logK hsa), log BB for brain/blood (logBB), number of primary metabolites (Metab), predicted central nervous system activity (CNS), HERG K+ channel blockage: log IC50 (logHERG), apparent Caco-2 permeability in nm/s (PCaco), apparent MDCK permeability in nm/s (PMDCK), QP logK p for skin permeability (logK p), maximum transdermal transport rate (Jm), Lipinski rule of 5 violations (rule of 5), Jorgensen rule of 3 violations (rule of 3) and percentage human oral absorption in GI (pHOA) were calculated. Predicted ADMET properties for inhibitors (Fig. 12) and normal range in 95 % of drugs are given in Table 5.
Fig. 12

Values of different ADMET properties for potent inhibitors

Table 5

Predicted ADMET properties for compounds and its normal range in 95 % of drugs

S. no.

Properties prediction

Tramiprosate

AZD-103

Cassumunin A

Cassumunin B

Piperine

Dibenzoyalmethan

Allyl curcumin

Curcumin

Curcumin bis-acetate

Range 95 % of drugs

1

Polrz

12.053

9.683

58.644 M

60.712 M

30.601 M

27.237 M

48.529 M

36.857 M

46.856 M

13/70

2

logP C16

6.752

4.716

19.772 M

20.239 M

8.630 M

8.906 M

16.036 M

12.965 M

15.040 M

4/18

3

logP oct

18.113

10.299

27.032 M

28.083 M

12.077 M

11.503

19.178 M

18.655 M

21.057 M

8/35

4

logP w

19.768

10.065

12.990 M

13.249 M

5.888 M

7.136

8.044 M

11.558 M

12.273 M

4/45

5

logP o/w

−2.412

−2.994

6.351

6.493

3.261

2.557

5.782

2.997

2.824

−2.0/6.5

6

logS

−1.264

0.195

−8.511

−8.706

−3.532

−2.740

−6.673

−4.349

−4.393

−6.5/0.5

7

CIlogS

−0.306

0.430

−8.189

−8.496

−3.556

−2.699

−6.070

−4.616

−4.531

−6.5/0.5

8

logK hsa

−0.747

−1.092

1.021

1.034

−0.024

−0.283

0.546

−0.032

−0.473

−1.5/1.5

9

log BB

−1.697

−0.904

−2.254

−2.299

−0.124

−0.388

−1.416

−1.836

−2.123

−3.0/1.2

10

Metab

6

2

9

10

0

1

7

5

3

1.0/8.0

11

CNS

±

±

−2 inactive, +2 active

12

logH ERG

−2.535

−1.726

−7.859

−7.710

−4.849

−5.693

−7.473

−6.313

−6.739

Concern below −5

13

PCaco

41

7

507

541

3,984

1,788

1,671

326

257

<25 poor, >500 great

14

PMDCK

15

3

237

254

2,204

927

861

147

114

<25 poor, >500 great

15

logK p

−5.561

−6.785

−1.171

−1.189

−1.174

−1.207

−0.304

−2.247

−2.549

−8.0/−1.0 Kp cm h−1

16

Jm

0.027

0.036

0.000

0.000

5.615

26.382

0.047

0.093

0.052

μg cm−2 h−1

17

Rule of 5

1

0

2

2

0

0

1

0

0

max. 4

18

Rule of 3

0

1

2

2

0

0

2

0

0

max. 3

19

pHOA

29

26

87

88

100

100

100

89

87

<25 % is poor

Normal range of polarizability varies from 13 to 70 M, which was predicted satisfactory and high in cassumunin A (58.644 M) and cassumunin B (60.712 M). Normal range of logP for octanol/water varies from −2.0 to 6.5 (95 % drugs), and it has been found negatively high in case of tramiprosate (−2.412) and AZD-103 (−2.994). Normal range of log BB for brain/blood varies from −3.0 to 1.2 which was favorable for all the compounds except dibenzoyalmethane (−0.388). IC50 represents the concentration of a drug that is required for 50 % inhibition in vitro (Pajeva et al. 2009). LogHERG values for 95 % drugs are concern if it is less than −5. LogHERG was predicted high in case of tramiprosate (−2.535) and AZD-103 (−1.726). CNS has been predicted on −2 (inactive) to +2 (active) scale. Inacitve CNS response was predicted for all the compounds, except for piperine and dibenzoyalmethane. Prediction of human oral absorbtion is based on quantitaive multiple linear regression model. PCaco and PMDCK descriptors were used for prediction of non-active transport. PCcao (3,984) and PMDCK (2,204) values were found very high for piperine, which is considered excellent in case greater than 500. PCcao and PMDCK values were found not satisfactory for tramiprosate and AZD-103.

Tramiprosate, AZD-103, cassumunin A, cassumunin B and allyl curcumin were not satisfying the Lipinski rule’s of five and Jorgensen’s rule of three (logs > −5.7, PCaco >22 nm/s, primary metabolite <7). Compounds that satisfy the Lipinski rule’s of five are considered drug-like (Lipinski et al. 2001). Compounds with no violations of Jorgensen’s rule of three are more likely to be orally available. Percentage human oral absorption (pHOA) of inhibitors in gastro intestine may varies by ±20 %). All the compounds have shown very good and satisfactory human oral absorption as in case of majority of drugs. Known drugs of AD (tramiprosate and AZD-103) were found to have very low oral absorption (29 and 26 %, respectively) as compared to other potential lead molecules. Curcumin derivative and other herbal analogs can serve as a good qualitatve model for human oral absorption. ADMET properties are very important for a molecule to serve as a drug (Wang 2009).

Docking energy for most of the inhibitors was found favorable for β-A fibrils rather than β-A, which indicates that these compounds can get stuck into β-A fibril due to positive interaction. Therefore, it does not seem meaningful to draw a concluding remark from this study that these compounds can promote disaggregation of already formed β-A fibrils or aggregates. Cassumunins A and B show very favorable bonding interaction with β-A and very unfavorable interaction with β-A fibrils as compared to others compounds. This study indicates that these two inhibitors can specially target β-A rather than β-A fibrils. Specificity of cassumunins for β-A can be further utilized in development of more potent inhibitors for β-A aggregation. Cassumunins and other herbal congeners are natural and have low risk of toxicity. This indicates that curcumin and other herbal compounds bind with β-A presents in brain, and inhibit the aggregation of β-A into β-A fibrils. So these compounds can have preventive role in AD rather than curative.

This study also projects some new ways for further development and screening of new inhibitors for a specific drug target. The comparative analysis can be carried out to evaluate the performance of these tools and softwares (Singh and Chandra 2012). Metabolic modeling and simulation analysis can be used to determine the more potent targets related with AD pathways which can suppress the progression of neurodegenerative disorder (Gupta et al. 2012). Overall, the work is based on in silico study which could be further verified by experimental study. This kind of approach could also be applied to other pathogenic and metabolic diseases.

4 Conclusion

Curcumin, its derivatives and some of its herbal congeners may be more potent inhibitors of β-A aggregation than known drug tramiprosate. Natural herbal inhibitors have revealed exceptionally high binding affinity for β-A, and can prevent aggregation of β-A into amyloid plaques. Majority of curcumin derivatives and herbal congeners have good and satisfactory range of principal descriptors and ADMET values. Being herbal in nature, these compounds can be used as lead molecules to design potent inhibitors for β-A aggregation. This study also suggests that consumption of herbal spices may reduce the risk of AD. Inhibitory effect of these compounds to prevent β-A aggregation can be quantified by chemical kinetics and potential of these compounds to disaggregate already formed β-A fibrils or aggregates can also be evaluated. Further structural optimization of these herbal compounds could open the way for improved modalities for cure of AD.

References

  1. ACD/ChemSketch Freeware, version 8.0 (2006) Advanced Chemistry Development, Inc., Toronto, ON, CanadaGoogle Scholar
  2. Adhikary R, Barnes CA, Trampel RL, Wallace SJ, Kee TW, Petrich JW (2011) Photoinduced trans-to-cis isomerization of cyclocurcumin. J Phys Chem B 115(36):10707–10714CrossRefGoogle Scholar
  3. Agdeppa ED, Kepe V, Petri A, Satyamurthy N, Liu J, Huang SC, Small GW, Cole GM, Barrio JR (2003) In vitro detection of (S)-naproxen and ibuprofen binding to plaques in the Alzheimer’s brain using the positron emission tomography molecular imaging probe 2-(1-[6-[(2-[(18)F]fluoroethyl)(methyl)amino]-2-naphthyl]ethylidene)malononitrile. Neuroscience 117(3):723–730CrossRefGoogle Scholar
  4. Azzimondi G, D’Alessandro R, Pandolfo G, Feruglio FS (1998) Comparative study of the prevalence of dementia in two Sicilian communities with different psychosocial backgrounds. Neuroepidemiology 17(4):199–209CrossRefGoogle Scholar
  5. Boersma F, Eefsting JA, van den Brink W, Koeter M, van Tilburg W (1998) Prevalence of dementia in a rural Netherlands population and the influence of DSM-III-R and CAMDEX criteria for the prevalence of mild and more severe forms. J Clin Epidemiol 51(3):189–197CrossRefGoogle Scholar
  6. Brayne C, Nickson J, McCraken C, Gill C, Johnson AL (1998) Cognitive function and dementia in six areas of England and Wales: distribution of MMSE and prevalence of GMS organicity level in the MRC CFA Study. Psychol Med 28(2):319–335CrossRefGoogle Scholar
  7. Cam JA, Bu G (2006) Modulation of beta-amyloid precursor protein trafficking and processing by the low density lipoprotein receptor family. Mol Neurodegener 1:8CrossRefGoogle Scholar
  8. Cole GM, Morihara T, Lim GP, Yang F, Begum A, Frautschy SA (2004) NSAID and antioxidant prevention of Alzheimer’s disease: lessons from in vitro and animal models. Ann N Y Acad Sci 1035:68–84CrossRefGoogle Scholar
  9. Crescenzi O, Tomaselli S, Guerrini R, Salvadori S, D’Ursi AM, Temussi PA, Picone D (2002) Solution structure of the Alzheimer amyloid beta-peptide (1–42) in an apolar microenvironment. Similarity with a virus fusion domain. Eur J Biochem 269(22):5642–5648CrossRefGoogle Scholar
  10. Dundas J, Ouyang Z, Tseng J, Binkowski A, Turpaz Y, Liang J (2006) CASTp: computed atlas of surface topography of proteins with structural and topographical mapping of functionally annotated residues. Nucleic Acids Res 34(Web Server issue):W116–W118Google Scholar
  11. Ertl P, Rohde B, Selzer P (2000) Fast calculation of molecular polar surface area as a sum of fragment-based contributions and its application to the prediction of drug transport properties. J Med Chem 43(20):3714–3717CrossRefGoogle Scholar
  12. Fiala JC (2007) Mechanism of amyloid plaque pathogenesis. Acta Neuropathol 114:551–571CrossRefGoogle Scholar
  13. Fillenbaum GG, Heyman A, Huber MS, Woodbury MA, Leiss J, Schmader KE, Bohannon A, Trapp-Moen B (1998) The prevalence and 3-year incidence of dementia in older black and white community residents. J Clin Epidemiol 51(7):587–595CrossRefGoogle Scholar
  14. Ghanta J, Shen CL, Kiessling LL, Murphy RM (1996) A strategy for designing inhibitors of beta-amyloid toxicity. J Biol Chem 271(47):29525–29528CrossRefGoogle Scholar
  15. Gupta MK, Singh DB, Rath SK, Misra K (2012) Metabolic modeling and simulation analysis of thyroid disorder pathway. J Comput Sci Syst Biol 5(2):52–61CrossRefGoogle Scholar
  16. Herrera E Jr, Caramelli P, Nitrini R (1998) Estudo epidemiológico populacional de demência na cidade de Catanduva—Estado de São Paulo, Brasil. Rev Psiq Clín (São Paulo) 25:70–73Google Scholar
  17. Jorgensen WL (1998) BOSS—Biochemical and Organic Simulation System. In: Schleyer PvR (ed) The Encyclopedia of Computational Chemistry, vol 5. Wiley, Athens, pp 3281–3285Google Scholar
  18. Kim W, Hecht MH (2006) Generic hydrophobic residues are sufficient to promote aggregation of the Alzheimer’s Abeta42 peptide. Proc Natl Acad Sci 103(43):15824–15829CrossRefGoogle Scholar
  19. Kung MP, Skovronsky DM, Hou C, Zhuang ZP, Gur TL, Zhang B, Trojanowski JQ, Lee VM, Kung HF (2003) Detection of amyloid plaques by radioligands for Abeta40 and Abeta42: potential imaging agents in Alzheimer’s patients. J Mol Neurosci 20(1):15–24CrossRefGoogle Scholar
  20. Lee VM (2002) Amyloid binding ligands as Alzheimer’s disease therapies. Neurobiol Aging 23(6):1039–1042CrossRefGoogle Scholar
  21. Lim GP, Chu T, Yang F, Beech W, Frautschy SA, Cole GM (2001) The curry spice curcumin reduces oxidative damage and amyloid pathology in an Alzheimer transgenic mouse. J Neurosci 21(21):8370–8377Google Scholar
  22. Lin RT, Lai CL, Tai CT, Liu CK, Yen YY, Howng SL (1998) Prevalence and subtypes of dementia in southern Taiwan: impact of age, sex, education, and urbanization. J Neurol Sci 160(1):67–75CrossRefGoogle Scholar
  23. Lipinski CA, Lombardo F, Dominy BW, Fenney PJ (2001) Experimental and computational approaches to estimate solubility and permeability in drug discovery and development settings. Adv Drug Deliv Rev 46:3–26CrossRefGoogle Scholar
  24. Lopes MA, Hototian SR, Reis GC, Elkis H, Bottino CMC et al (2007) Systemic review of dementia prevalence—1994 to 2000. Dement Neuropsychol 1(3):230–240Google Scholar
  25. Lührs T, Ritter C, Adrian M, Riek-Loher D, Bohrmann B, Döbeli H, Schubert D, Riek R (2005) 3D structure of Alzheimer’s amyloid-beta (1–42) fibrils. Proc Natl Acad Sci 102(48):17342–17347CrossRefGoogle Scholar
  26. Masuda T, Matsumura H, Oyama Y, Takeda Y, Jitoe A, Kida A, Hidaka K (1998) Synthesis of (±)-cassumunins A and B, new curcuminoid antioxidants having protective activity of the living cell against oxidative damage. J Nat Prod 61(5):609–613CrossRefGoogle Scholar
  27. McLaurin J, Kierstead ME, Brown ME, Hawkes CA, Lambermon MH, Phinney AL, Darabie AA, Cousins JE, French JE, Lan MF, Chen F, Wong SS, Mount HT, Fraser PE, Westaway D, St. George-Hyslop P (2006) Cyclohexanehexol inhibitors of Abeta aggregation prevent and reverse Alzheimer phenotype in a mouse model. Nat Med 12(7):801–808CrossRefGoogle Scholar
  28. Mihalek I, Res I, Lichtarge O (2006) Evolutionary trace report_maker: a new type of service for comparative analysis of proteins. Bioinformatics 22(13):1656–1657CrossRefGoogle Scholar
  29. Mills N (2006) ChemDraw Ultra 10.0. J Am Chem Soc 128(41):13649–13650CrossRefGoogle Scholar
  30. Mishra S, Palanivelu K (2008) The effect of curcumin (turmeric) on Alzheimer’s disease: an overview. Ann Indian Acad Neurol 11(1):13–19CrossRefGoogle Scholar
  31. Nagano T, Oyama Y, Kajita N, Chikahisa L, Nakata M, Okazaki E, Masuda T (1997) New curcuminoids isolated from Zingiber cassumunar protect cells suffering from oxidative stress: a flow-cytometric study using rat thymocytes and H2O2. Jpn J Pharmacol 75(4):363–370CrossRefGoogle Scholar
  32. Nakamura S, Iwami J, Matsuda H, Wakayama H, Pongpiriyadacha Y, Yoshikawa M (2009) Structures of new phenylbutanoids and nitric oxide production inhibitors from the rhizomes of Zingiber cassumunar. Chem Pharm Bull (Tokyo) 57(11):1267–1272CrossRefGoogle Scholar
  33. Narlawar R, Baumann K, Schubenel R, Schmidt B (2007) Curcumin derivatives inhibit or modulate beta-amyloid precursor protein metabolism. Neurodegener Dis 4(2–3):88–93CrossRefGoogle Scholar
  34. Ono K, Hasegawa K, Naiki H, Yamada M (2004) Curcumin has potent anti-amyloidogenic effects for Alzheimer’s beta-amyloid fibrils in vitro. J Neurosci Res 75(6):742–750CrossRefGoogle Scholar
  35. Pajeva IK, Globisch C, Wiese M (2009) Combined pharmacophore modeling, docking, and 3D QSAR studies of ABCB1 and ABCC1 transporter inhibitors. ChemMedChem 4(11):1883–1896CrossRefGoogle Scholar
  36. Punta M, Coggill PC, Eberhardt RY, Mistry J, Tate J, Boursnell C, Pang N, Forslund K, Ceric G, Clements J, Heger A, Holm L, Sonnhammer EL, Eddy SR, Bateman A, Finn RD (2012) The Pfam protein families database. Nucleic Acids Res 40(Database issue):D290–D301Google Scholar
  37. QikProp, version 3.3 (2010) Schrodinger, LLC, New YorkGoogle Scholar
  38. Ringman JM, Frautschy SA, Cole GM, Masterman DL, Cummings JL (2005) A potential role of the curry spice curcumin in Alzheimer’s disease. Curr Alzheimer Res 2(2):131–136CrossRefGoogle Scholar
  39. Ruby AJ, Kuttan G, Babu KD, Rajasekharan KN, Kuttan R (1995) Anti-tumour and antioxidant activity of natural curcuminoids. Cancer Lett 94(1):79–83CrossRefGoogle Scholar
  40. Santa-Maria I, Hernández F, Del Rio J, Moreno FJ, Avila J (2007) Tramiprosate, a drug of potential interest for the treatment of Alzheimer’s disease, promotes an abnormal aggregation of tau. Mol Neurodegener 2:17CrossRefGoogle Scholar
  41. Shaji S, Promodu K, Abraham T, Roy KJ, Verghese A (1996) An epidemiological study of dementia in a rural community in Kerala, Índia. Br J Psychiatry 168(6):745–749CrossRefGoogle Scholar
  42. Singh TR, Chandra S (2012) Linear B cell epitope prediction for epitope vaccine design against meningococcal disease and their computational validations through physicochemical properties. Netw Model Anal Health Inform Bioinform 1:153–159CrossRefGoogle Scholar
  43. Srimal RC, Dhawan BN (1973) Pharmacology of diferuloylmethane (curcumin), a non-steroidal anti-inflammatory agent. J Pharm Pharmacol 25(6):447–452CrossRefGoogle Scholar
  44. Tang W, Su D (2012) Locomotion analysis and its applications in neurological disorders detection: state-of-art review. Netw Model Anal Health Inform Bioinform. doi: 10.1007/s13721-012-0020-8 Google Scholar
  45. Tang W, Yarowsky P, Tasch U (2012) Detecting ALS and Parkinson’s disease in rats through locomotion analysis. Netw Model Anal Health Inform Bioinform 1:63–68CrossRefGoogle Scholar
  46. Thomsen R, Christensen MH (2006) MolDock: a new technique for high-accuracy molecular docking. J Med Chem 49(11):3315–3321CrossRefGoogle Scholar
  47. Wang J (2009) Comprehensive assessment of ADMET risks in drug discovery. Curr Pharm 15(19):2195–2219CrossRefGoogle Scholar
  48. Woo JI, Lee JH, Yoo KY, Kim CY, Kim YI, Shin YS (1998) Prevalence estimation of dementia in a rural area of Korea. J Am Geriatr Soc 46(8):983–987Google Scholar
  49. Yamada M, Sasaki H, Mimori Y, Kasagi F, Sudoh S, Ikeda J, Hosoda Y, Nakamura S, Kodama K (1999) Prevalence and risks of dementia in the Japanese population: RERF’S adult health study Hiroshima subjects. J Am Geriatr Soc 47(2):189–195Google Scholar
  50. Yang J, Chen C (2004) GEMDOCK: a generic evolutionary method for molecular docking. Proteins 55:288–304CrossRefGoogle Scholar
  51. Yang F, Lim GP, Begum AN, Ubeda OJ, Simmons MR, Ambegaokar SS, Chen PP, Kayed R, Glabe CG, Frautschy SA, Cole GM (2005) Curcumin inhibits formation of amyloid beta oligomers and fibrils, binds plaques, and reduces amyloid in vivo. J Biol Chem 280(7):5892–5901CrossRefGoogle Scholar
  52. Zheng J, Jang H, Ma B, Nussinov R (2008) Annular structures as intermediates in fibril formation of Alzheimer Abeta17–42. J Phys Chem B 112(22):6856–6865CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Wien 2012

Authors and Affiliations

  • Dev Bukhsh Singh
    • 1
    Email author
  • Manish Kumar Gupta
    • 2
  • Rajesh Kumar Kesharwani
    • 3
  • Krishna Misra
    • 3
  1. 1.Department of BiotechnologyInstitute of Biosciences and Biotechnology, Chhatrapati Shahu Ji Maharaj UniversityKanpurIndia
  2. 2.Department of BioinformaticsUniversity Institute of Engineering and Technology, Chhatrapati Shahu Ji Maharaj UniversityKanpurIndia
  3. 3.Division of Applied Science and Indo-Russian Center for Biotechnology [IRCB]Indian Institute of Information TechnologyAllahabadIndia

Personalised recommendations