in silico admet · pharmacokinetic properties for the selection of the e ective and bioavailable...

Hindawi Publishing CorporationISRN Structural BiologyVolume 2013 Article ID 373516 12 pageshttpdxdoiorg1011552013373516

Research Article2D-QSAR Docking Studies and In Silico ADMETPrediction of Polyphenolic Acetates as Substrates for ProteinAcetyltransferase Function of Glutamine Synthetase ofMycobacterium tuberculosis

Prija Ponnan12 Shikhar Gupta3 Madhu Chopra4 Rashmi Tandon12

Anil S Baghel1 Garima Gupta1 Ashok K Prasad2 Ramesh C Rastogi2

Mridula Bose1 and Hanumantharao G Raj1

1 Department of Biochemistry and Microbiology V P Chest Institute University of Delhi Delhi 110 007 New Delhi India2Department of Chemistry University of Delhi Delhi 110 007 New Delhi India3 Department for Pharmacoinformatics National Institute of Pharmaceutical Education and Research SAS Nagar MohaliPunjab 160062 India

4Dr BR Ambedakar Centre for Biomedical Research University of Delhi Delhi 110 007 India

Correspondence should be addressed to Shikhar Gupta shiksungmailcom

Received 30 November 2012 Accepted 20 December 2012

Academic Editors M Espinoza-Fonseca and D D Leonidas

Copyright copy 2013 Prija Ponnan et al This is an open access article distributed under the Creative Commons Attribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

A novel transacetylase (TAase) function of glutamine synthetase (GS) in bacterial species such as Mycobacterium smegmatis andMycobacterium tuberculosisH37Rv was established by us termed as mycobacterial TAase (MTAase) Several polyphenolic acetates(PAs) were found to be substrates for MTAase by inhibiting certain receptor proteins such as glutathione S-transferase by way ofacetylation The present work describes the descriptor-based 2D-QSAR studies developed for a series of PA synthesized by us andevaluated for MTAase and antimycobacterial activity using stepwise multiple linear regression method with the kinetic constantsand the minimum inhibitory constant (MIC) as the dependent variables to address the fact that TAase activity was leading to theantimycobacterial activity Further blind docking methods using AutoDock were carried out to study the interaction of potent PAwith the crystal structure of M tuberculosis GS PAs were predicted to bind M tuberculosis GS on the protein surface away fromthe known active site of GS Subsequent focussedrefined docking of potent PA with GS showed that the 120576-amino group of Lys4of GS formed a cation-120587 interaction with the benzene ring of PA Also ADMET-related descriptors were calculated to predict thepharmacokinetic properties for the selection of the effective and bioavailable compounds

1 Introduction

Our laboratory is credited for the discovery of novel TAasewhich catalyzes the possible transfer of acetyl group fromPA to certain functional proteins such as GST cytochromeP-450 reductase and nitric oxide synthase (NOS) leadingto their functional modifications [1ndash3] An assay procedurewas developed utilizing the inhibition of cytosolic GSTbrought about by TAase-catalyzed acetylation by PA Boththe substrates namely the target protein GST and the acetyl

group donor PAs were found to take part in the TAase-catalyzed bimolecular reaction [2] This assay procedure wasutilized to purify TAase from tissues like human placentaand rat liver and characterized as calreticulin a calcium-binding ER luminal protein [4 5]The acetylation of receptorproteins such as GST and NOS at 120576-amino group lysineresidues was established by immunoblotting using acetylatedlysine antibody andmass spectrometry [6 7] Recently TAasewas identified and established by us in bacterial speciessuch as Mycobacterium smegmatis [8] and Mycobacterium

2 ISRN Structural Biology

tuberculosis (Mtb) H37Rv [9] as glutamine synthetase (GS)Glutamine synthetase catalyzes the conversion of glutamateto glutamine in the presence of ammonium ion with simul-taneous hydrolysis of ATP which is used as the energy sourceand plays an essential role in bacterial nitrogen metabolism[10 11] Several PAs including acetoxycoumarins in generalwere found to be the substrates for mycobacterial TAase(MTAase) The specificities of various acetoxycoumarinstowards MTAase were determined by their ability to inhibitGST irreversibly and their kinetic constants (119870

119898and 119881max)

were determined [9] Several inhibitors are known for GSand most of them are analogues of glutamate and replacethis substrate in the active site of the enzyme Among theknown inhibitors methionine sulfoximine (MSO) and 2-amino-4-(hydroxymethyl-phosphoryl) butanoic acid (phos-phinothricin) are the well-established inhibitors of GS [1213] During the examination of the role of GS inhibitor onMTAase function of GS it was observed that MSO failedto inhibit MTAase-catalysed reaction indicating that theTAase activity of MTAase is independent of the catalyticactivity of GS [8 9] Electron microscopic studies carriedout by us have shown cell wall attacking properties of thesecompounds in M smegmatis [8] and M tuberculosis [14]The cell wall of Mycobacterium species is responsible formaintaining the cell integrity and thus is considered to be apotential drug target owing to its crucial role in cell survivaland viability If a compound is found to affect the cell wallor its biosynthesis in any manner it is bound to bring aboutthe inhibition of bacterial growth The PAs referred to inthe present study have been reported earlier [14] to possessldquocell-wall attackingrdquo characteristic that is these moleculeshave been found to bring about changes in cell morphologyranging from indentations in the wall to complete rupturingof cell wall along with extrusion of cytoplasmic material insome cases and complete disintegrationdisappearance of thewall in others [14]Moreover these pronounced changes wererecorded when the bacteria were grown in the presence ofsublethal doses of the test molecules These observations ledus to believe that these compounds may serve as potentialdrug candidates and therefore these were further exploredto determine their drug likeness and also establish a structureactivity relationship [14] The present work describes thedescriptor-based QSAR studies developed for a series of ace-toxycoumarins synthesized by us and evaluated forTAase andantimycobacterial activity Also ADMET-related descriptorswere calculated to predict the pharmacokinetic properties forthe selection of the effective and bioavailable compoundsFurther docking studies were done to analyze the interactionof the potent acetoxycoumarins with the crystal structure ofM tuberculosis GS

2 Methodology

21 2119863minusQSAR Analysis

211 Data Set and Methodology Compounds 1ndash14 were syn-thesized and characterized following the published syntheticprocedures [1ndash3 9 15] The PAs were screened for theirantimycobacterial activity and TAase activity that are listed

in Table 1 according to our published methods [8] utilizingMtbGS as the target protein Biological activity data reportedas MIC values for the antimycobacterial activity and kineticconstants (119870

119898and 119881max) for TAase activity (Table 1) were

first converted to -logMIC onmolar basis and log (119881max119870119898)respectively and were used as the dependent variables to getthe linear relationship in the QSAR models

Hyperchem-8 program [16] was used to build the struc-tures and perform geometry optimizations of the com-pounds The lowest energy conformations of the compoundswere determined first byminimizing the structures bymolec-ular mechanics method using MM+ force field followed bysemiempirical self-consistent field molecular orbital (SCFMO) theory (parametric model 3 (PM3) method withinthe restricted Hartree-Fock (RHF) formalism) Conjugategradientmethod (Polak-Ribiere algorithm)with SCF conver-gency set to 0001 kcalmol was considered in the geometryoptimization stage of calculations Frequency calculationshave been performed to confirm all stationary points

212 Molecular Descriptors for QSAR Analysis TSAR 33software package (Accelrys San Diego CA USA) wasemployed to calculate descriptors for entire molecule andthe defined substituents Substituents were defined for allPA a single hydrogen atom also served as a substituent(Table 1) TSAR includes various physicochemical topolog-ical and electrostatic descriptors molecular surface area andvolume molecular mass moments of inertia (moment 12 3 (size length)) ellipsoidal volume Verloop parametersDipole moments (total bond and x y z components)Lipole moments (total bond and x y z components)topological indices (Wiener Randic and Balaban indices)molecular connectivity indices (Chi ChiV indices) of atomsbonds path cluster and pathclusterMolecular shape indices(Kappa KAlpha indices) Electrotopological state indicesLogP Atom counts (CNS amp H) Ring count (aromaticand aliphatic) and Group count (methyl hydroxyl ethyl)Electrostatic properties like Total energy Electronic energyNuclear repulsion energy Accessible surface area Atomiccharge Mean polarizability Heat of formation HOMOand LUMO eigenvalues Ionization potential Total dipolePolarizability and Dipole components Pairwise correlationanalysis of the descriptors was performed and the intercor-related descriptors (gt06) were discarded depending on theirindividual correlation with the biological activity

213 Stepwise Multiple Regression In an effort to investigatethe role of structural parameters which appears to influencethe observed activities of reported compounds stepwisemultiple linear regressions were performed using TSAR 33software TSAR uses a two-way stepping algorithm to selectvariables for the regression equation At each step partial 119865values are calculated for each variable as an estimate of theirpotential contribution to the model The partial 119865 values arecompared with the 119865-to- Leave and 119865-to-Enter settings Theoverall 119865 statistic for a model is

119865 =

explained mean squareresidual mean square

(1)

ISRN Structural Biology 3

Table 1 Structures of PA used in the 2D-QSAR analysis with corresponding TAase and antimycobacterial activities

O O

R2

R3

R1

1

2

34567

8

Compound R1 R2 R3Antimycobacterial activity TAase activityMIC minuslog MIC

119870119898

119881max log(119881max 119870119898)1 H NHCOC5H11 H 2 569897 220 45 36892102 H NHCOC4H9 H 2 569897 210 51 36146493lowast H NHCOC3H7 H 2 569897 205 54 36585414 H NHCOCH3 H 3 5522879 151 105 31877515 H NHCOC2H5 H 3 5522879 110 130 29274496 H OCOCH3 OCOCH3 12 4920819 100 142 28477127 H SCOCH3 H 14 4853872 Nonenzymatic8 H OCOC2H5 OCOC2H5 14 4853872 152 98 31606549 C10H21 OCOCH3 OCOCH3 20 469897 105 125 290309010 C6H13 OCOCH3 OCOCH3 30 4522879 110 130 292744911lowast C10H21 OCOCH3 H 40 439794 160 95 322639612lowast H OCOC3H7 OCOC3H7 50 430103 198 60 327135913lowast H OCOCH3 H 60 4221849 148 115 310956414 H OH OH 80 409691 Not a substratelowastTest setValues are mean of three observations in triplicate with variation less than 5

Partial 119865 values are an estimation of the sequential con-tribution towards the 119865 statistic for the final model 119865-to-Leave forward and backward stepping algorithms cangive regression equations that use different variables Thisis caused by collinearity or multicollinearity of variables inthe data set and may indicate instability in the model Ina forward stepping process once a variable has entered themodel it cannot leave If 119865-to-Leave is set to zero a forwardstepping process is used At each step the partial 119865values ofall variables outside the model are calculated If any variablehas a value greater than 119865-to-Enter the variable with thehighest partial 119865 value is added to the model The processis continued until no more variables qualify to enter themodel or the required number of steps has been reached In abackward stepping process all variables are used in the initialmodel (overriding any choice of starting variables) Once avariable has left the model it may not reenter If 119865-to-Enteris set to zero a backward stepping process is used At eachstep the partial 119865 values of all variables inside the model arecalculated If any variable has a value less than 119865-to-Leavethe variable with the lowest partial 119865 value is removed fromthe model The process is continued until no more variablesqualify to leave the model or the required number of stepshas been reached

The default values for ldquosteppingrdquo that is 119865-to-Enterand 119865-to-Leave were set to 4 and 35 respectively Thewhole dataset was randomly divided into test set (includingcompounds 3 11 12 and 13) and remaining compounds astraining set Statistical quality of the regression models wasjudged based on parameters such as correlation coefficient

(119903) squared correlation coefficient (1199032) standard error ofestimate (119904) and fisher test value (119865-value) A compound wasconsidered as an outlier when the residual value exceeded15 times the standard error of estimate in an equationFurther the predictive ability of the model was quantifiedinternally by determining cross-validated 1199032 by leave-one-out(LOO) method (q2LOO) and the predictive residual sum ofsquares (PRESS) Predictive ability of the generated modelwas validated by using the external test set by determiningexternal set cross validation 119903

2 (1199022ext) determination coeffi-cient between observed and predicted values with (1199032pred) andwithout intercept (1199032

0) slopes 119896 and 1198961015840 of regressions through

the origin of predicted versus observed and observed versuspredicted intensities respectively Models were considered tohave high predictive ability [17 18] if 1199022ext gt 05 1199032pred gt 06both 1199032

0and 11990310158402

0had to be close to each other such that (1199032predminus

1199032

0)1199032

pred lt 01 or (1199032

pred minus 11990310158402

0)1199032

pred lt 01 and the corres-ponding slopes should follow the criteria 085 le 119896 le 115 or085 le 119896

1015840

le 115 [17 18]

214 ADMET Prediction for Acetoxycoumarins Absorp-tion distribution metabolism elimination and toxicity(ADMET) properties were predicted using ADMET descrip-tors in Discovery Studio 21 (Accelrys San Diego CA USA)The module uses six mathematical models to quantitativelypredict properties by a set of ruleskeys (Table 2) thatspecify threshold ADMET characteristics for the chemicalstructure of the molecules based on the available drug


Table 2 ADMET descriptors and their ruleskeys

ADMET absorption level (human intestinal absorption)Level Description0 Good absorption1 Moderate absorption2 Low absorption3 Very low absorption

ADMET aqueous solubility levelLevel Value Description

0 log (molar solubility)lt minus80 Extremely low

1 minus80 lt log (molarsolubility) lt minus60 No very low but possible

2 minus60 lt log (molarsolubility) lt minus40 Yes low

3 minus40 lt log (molarsolubility) lt minus20 Yes good

4 minus20 lt log (molarsolubility) lt 00 Yes optimal

5 00 lt log (molar solubility) No too soluble

6 minus1000Warning molecules withone or more unknown

AlogP98 typesADMET (blood brain barrier penetration level) BBB

Level Description0 Very High1 High2 Medium3 Low4 Undefined

5Warning molecules withone or more unknownAlogP calculation

ADMET CYP2D6Predictedclass Value

0 Noninhibitor1 Inhibitor

ADMET hepatotoxicityPredictedclass Value

0 Nontoxic1 ToxicADMET (plasma protein binding level) PPBLevel Description0 Binding is lt901 Binding is ge902 Binding is ge95

information ADMET absorption predicts human intesti-nal absorption (HIA) after oral administration The modelwas developed using 199 compounds in the training setbased on the calculations AlogP (ADMET AlogP98) and 2D

polar surface area (PSA 2D) The absorption levels of HIAmodel are defined by 95 and 99 confidence ellipses inthe ADMET PSA 2D ADMET AlogP98 plane [19] Theseellipses describe the regionswherewell-absorbed compoundsare expected to be found The upper limit of PSA 2D valuefor the 95 confidence ellipsoid is at 13162 while the upperlimit of PSA 2D value for the 99 confidence ellipsoid isat 14812 ADMET aqueous solubility predicts the solubilityof each compound in water at 25∘C The model is basedon genetic partial least squares method on a training set of784 compounds with experimentally measured solubilities[20] ADMET blood brain barrier model predicts blood-brain penetration (blood brain barrier BBB) of a moleculeafter oral administration This model was derived from aquantitative linear regression model for the prediction ofblood-brain penetration as well as 95 and 99 confidenceellipses (analogous to that of HIA) in the ADMET PSA 2DADMET AlogP98 plane They were derived from over 800compounds that are known to enter the CNS after oraladministration [21] ADMET plasma protein binding modelpredicts whether a compound is likely to be highly boundto carrier proteins in the blood Predictions are basedon AlogP98 and 1D similarities to two sets of ldquomarkerrdquomolecules One set of markers is used to flag binding at a levelof 90 or greater and the other set is used to flag bindingat a level of 95 or greater Binding levels predicted bythe marker similarities are modified according to conditionson calculated logP [22] ADMET CYP2D6 binding predictscytochrome P450 2D6 enzyme inhibition using 2D chemicalstructure as input as well as a probability estimate for theprediction Predictions are based on a training set of 100compounds with known CYP2D6 inhibitions [23] ADMEThepatotoxicity predicts the potential human hepatotoxic-ity for a wide range of structurally diverse compoundsPredictions are based on an ensemble recursive partition-ing model of 382 training compounds known to exhibitliver toxicity (ie positive dose-dependent hepatocellularcholestatic neoplastic etc) or to trigger dose-related elevatedaminotransferase levels inmore than 10 percent of the humanpopulation [24]

215 Molecular Docking In order to corroborate the novelTAase function of Mtb GS it was thought importantto study the interaction of model PA 78-diacetoxy-4-methylcoumarin (DAMC) 7-acetoxy-4-methylcoumarin (7-AMC) and 7-NH-acetoxy-4-methylcoumarin (7-NH-AMC)with the structure of Mtb GS using computational dockingstudy In the absence of any known active site for theTAase activity of Mtb GS blind docking approach wasutilized wherein the entire protein surface is scanned for theprobable ligand binding sites for PA [25] For this purposeAutodock program was used [26] and PAs were dockedto the crystal structure of Mtb GS (PDB ID 2BVC) [27]in two steps Firstly a grid field of 60 A cube with gridpoints separated by 1 A centered at the middle of the proteinwas considered using AUTOGRID The final binding modeconformation was determined by focusedrefined dockingwhere the binding site determined with blind docking wassubjected to more detailed calculations by considering the


Table 3 Descriptors included in the best model obtained for antimycobacterial and TAase activity

Descriptor Coefficienta Jackknife SEb Covariance SEc119905-valued 119905-probabilitye

X1 Balabantopological index(Substituent 2)

025917 012484 0050123 51706 00020731

Antimycobacterialactivity

X2 Number of N atoms(Substituent 2)

084199 010326 007821 10766 37971119890 minus 005

X3 quadrupoleXX component(whole molecule)

0064479 0028036 0032179 20037 0091947

C constant 40866 043577

MTAase activity

X1 balabantopological index(Substituent 2)

013387 0018757 0027883 48012 00007223

C constant 28493 0045981aThe regressions coefficient for each variable in the QSAR equations bAn estimate of the standard error on each regression coefficient derived from a jackknife method on the final regression model cAn estimate of the standard error on each regression coefficient derived from covariance matrix dMeasures thesignificance of each variable included in the final modelestatistical significance for 119905 values

grid field of 60 A cube and the grid points were separated by0375 A centered on the best scored conformation obtained inthe first step Polar hydrogens and partial charges for proteinsand ligands were added using the Kollman United atom andGasteiger charges respectively using AUTODOCKTOOLS[28] An automated molecular docking was performed usingthe hybrid genetic algorithm-local search (GA-LS) Defaultparameters were used for the number of generations energyevaluations and docking runs which were set to 100025000000 and 256 respectively The docking energy repre-sents the sum of the intermolecular energy and the internalenergy of the ligand while the free-binding energy is thesum of the intermolecular energy and the torsional-freeenergy [29]

3 Results and Discussion

31 QSAR Analysis In an attempt to determine the roleof structural features of PA which appears to influencethe antimycobacterial activity by its acyl group donatingability mediated by TAase QSAR models was generatedThe inhibitory activity of PA determined in terms of MICvalues were taken as minus log MIC and the logarithmic valueof catalytic efficiency of PA (log(119881max119870119898)) to donate acetylgroup to receptor proteinmediated by TAase were used as thedependent values in the QSAR study (Table 1) As indicatedin Table 1 only 12 PAs were considered for TAase activ-ity compounds 7 being a nonenzymatic substrate wherebythis compound is capable of acetylating receptor proteinsindependent of acetyltransferase and compound 14 whichis the dihydroxy analogue of compound 6 The compoundpossesses hydroxyl group at C-7 and C-8 position andlacks acetyl group substituent and thus is a nonsubstratefor the protein acetyltransferase activity Hence these two

compounds (compounds 7 and 14) were thus excluded fromthe QSAR model generation of TAase activity

The QSAR model with high statistical significanceobtained for antimycobacterial activity can be representedby the following equation and the descriptors are detailed inTable 3

minus log MIC = 017540908 lowast X1 + 10271472 lowast X2

+ 010474976 lowast X3 + 4107533

(2)

119904 = 018 119865 = 4194 119903 = 096 1199032

= 093

1199022

LOO = 077 PRESS = 104

High predictive power of this model is demonstrated inFigure 1(a) and the histogram for residual is shown inFigure 1(b)

The obtained correlation equation was screened by usingtest set Figures 2(a) and 2(b) illustrate the predictive abilityof the QSAR where the statistical parameters 1199032pred = 09571199022

ext = 088 (1199032pred minus 1199032

0)1199032

pred = 0071(1199032pred minus 1199031015840

0

2

)1199032

pred lt

0031 119896 = 1026 1198961015840 = 097 were within the limits [17 18]The stepwise regression resulted in the following statis-

tically significant monoparametric model for TAase activityand the details of the descriptor are provided in Table 3

log (119881max119870119898) = 013387173 lowast X1 + 28492985 (3)

119904 = 0173 119865 = 2305 119903 = 0835

1199032

= 0697 1199022

LOO = 0609 PRESS = 0387

The plot of the calculated versus predicted log(119881max119870119898) ispresented in Figure 3(a) and the histogram for residual isshown in Figure 3(b)


Table4ADMET

predictio

nof

PAs

ADMET

absorptio

nlevel

ADMET

AlogP

98

ADMET

unkn

own

AlogP

98

ADMET

PSA

2D

ADMET

BBB

level

ADMET

BBB

ADMET

solubility

ADMET

solubility

level

ADMET

hepatotoxicity

ADMET

hepato-

toxicity

prob

ability

ADMET

CYP2

D6

ADMET

CYP2

D6

prob

ability

ADMET

PPBlevel

10

0345

0119

4minus10

64

00019

00455

0

20

0594

05949

3minus091

minus10

54

0006

60

0029

03

013

390

5056

3minus054

minus16

94

00052

00118

04

00328

06337

3minus10

6minus093

40

0052

00029

05

0278

08924

3minus071

minus272

30

0052

00455

06

04605

08924

4minus372

30

006

60

0435

1

70

5149

05949

10496

minus426

20

0052

0040

52

80

0051

08924

3minus15

5minus075

40

0059

00277

09

0minus001

07138

3minus12

9minus019

40

004

60

0029

010

2minus13

30

1309

40014

50

0086

00247

0

111

minus097

11342

4minus074

40

0052

00277

0

120

2269

08924

3minus087

minus233

30

0152

00366

013

10213

01309

4minus113

40

0039

00386

0

140

1357

08924

3minus115

minus17

24

0006

60

0316

0


442444648

552545658

6

4 42 44 46 48 5 52 54 56 58 6

Training setTest set

Pred

icte

dminuslog

(MIC

)

Calculated minus log(MIC)

(a)

001020304

1 2 3 4 5 6 7 8 9 10 11 12 13 14

Resid

ual v

alue

s

minus04minus03minus02minus01

(b)

Figure 1 (a) Graph of calculated versus predicted minus logMIC activi-ties fromQSARmodel (b) Histogram of residuals of calculated andpredicted minus logMIC activities PA in the training set

The model also followed the criteria for the predictiveability of the QSAR (Figures 4(a) and 4(b)) and the statisticalparameters 1199032pred = 0978 1199022ext = 0603 (1199032pred minus 119903

2

0)1199032

pred =

0078 (1199032pred minus 1199031015840

0

2

)1199032

pred lt 0091 119896 = 0971198961015840 = 102 werewithin the limits [17 18]

The descriptors based on the model used in the presentstudy are indicated in Table 3 It is observed that all thedescriptors have positive contribution to the antimycobacte-rial activityThe obtainedQSARmodel for antimycobacterialactivity demonstrates the significance of Balaban index forsubstituent 2 of PAThe descriptor Balaban index is a type oftopological index that represents extended connectivity andis a good descriptor for the shape of themolecules [31] All thetopological indices used are calculated from the hydrogen-suppressedmolecular graphs Balaban index can be describedas the average distance sum connectivity Balaban index 119869 ofa connected molecular graph 119866 can be defined as

119869 (119866) =

119864

120583 + 1

sum

edges(119889119904119894119889119904119895)

minus12

(4)

where 119864 is the number of edges in 119866 and 120583 is the cyclomaticnumber of 119866 The cyclomatic number 120583 of a cyclic graph119866 is equal to the minimum number of edges that must beremoved before119866 becomes acyclic and 119889119904

119894(119894 = 1 2 119873119873

4

45

5

55

6

65

4 42 44 46 48 5 52 54 56 58 6

Predictedminuslog(M

IC)


1199100 = 10265119909

11990320 = 08893

119910 = 0811119909 + 1032

1199032pred = 0957

(a)

4424446485

525456586

4 42 44 46 48 5 52 54 56 58 6

Predicted minus log(MIC)Ca

lculatedminuslog(M

IC)

1199039984002 = 09576

119910 = 11803119909 minus 1018

11990399840020 = 09279

1199100 = 09732119909

(b)

Figure 2 Regression plot between (a) calculated versus predictedvalues (minus logMIC) The dotted line indicates the regression linethrough origin (for equation 119910

0= 10265119909 with intercept = 0) and

the solid line indicates the regression lines for equation 119910 = 0811119909+

1032 (with intercept = 1032) and (b) predicted versus calculatedvalues (log 119881max119870119898) for compounds from test set justifying thepredictive ability of QSAR model The dotted line indicates theregression line through origin (for equation 119910

0= 09732119909 with

intercept = 0) and the solid line indicates the regression lines forequation 119910 = 11803119909 minus 1018 (with intercept = not1018)

is the number of vertices in119866) is a distance sumThe distancesum 119889119904

119894 for a vertex 119894 represents the sum of all entries in the

corresponding row (or column) of the distance matrix119863

119889119904119894=

119873

sum

119895=1

119863119894119895 (5)

The direct relationship between Balaban index of substituentat 2nd position (C-7 position of coumarin ring) and ndashlogMIC (see (2) Table 3) indicates that a bigger size and highbranching of substituent 2 increase the antimycobacterialactivity Balaban index has been successfully used to studythe antibacterial activity of sulfa drugs [32] Similarly thepositive correlation coefficient for number of nitrogen atomsat substituent 2 shows the significance of N-acyl substitutionat 2nd position in PA (see (2) Table 3) The presence ofthis descriptor in high magnitude in (2) demonstrates thedominating role of N-acyl substituted PA in antimycobac-terial activity The equation also expresses the significanceof quadrupole XX component (whole molecule) for theantimycobacterial activity It characterizes molecular chargedistribution in PA However only Balaban topological index


25

27

29

31

33

35

37

39

25 27 29 31 33 35 37 39 41

Training set

Test set

(a)

0

01

02

1 2 3 4 5 6 7 8 9 10 11 12 13

Res

idu

al v

alu

es

minus04

minus03

minus02

minus01

minus05

(b)

Figure 3 (a) Graph of calculated versus predicted log(119881max119870119898)

activities from QSAR model (b) Histogram of residuals of calcu-lated and predicted log(119881max119870119898) activities PA in the training set

for the substituent 2 of acetoxycoumarins showed significantcorrelation with the TAase activity (Table 3) Thus PA withhigh degree of bonding linearity with groups that increasemolecular weight was found to possess TAase activity EarlierBasak et al have indicated a predominant role of topologicalsteric parameters such as connectivity indices and informa-tion theoretic topological indices in determining the ratesof the enzymatic N-acetylation reaction [33] Further thesignificance of the descriptor Balaban topological index atsubstituent 2 could be understood in the way that PA withlong-chain acyl group could be a good substrate for MTAaseactivity This can be correlated with our recent investigationsthat led to the conclusion that PA with higher acyl groupsubstituent at C-7 position (other than acetyl group) such 7-propoxycoumarin was capable of transferring propoxy groupto the receptor proteins [34]HenceMTAase could be viewedas accommodating PAwith long chain acyl group in its activesiteOther acetyltransferases such as histone acetyltransferasewas found capable of accommodating higher chain CoAs(such as propionyl CoA and butyryl CoA) without sterichindrance [35]These observations give a tacit explanation for

3313233343536

3 31 32 33 34 35 36 37

1199032pred = 0978

119910 = 0761119909 + 0714

11990320 = 09006

1199100 = 09759119909

Predictedlog(119881

max119870119898)

Calculated log(119881max 119870119898)

(a)

3

31

32

33

34

35

36

37

3 31 32 33 34 35 36

(b)

Figure 4 Regression plot between (a) calculated versus predictedvalues (log 119881max119870119898) The dotted line indicates the regression linethrough origin (for equation 119910




0= 10245119909 with

intercept = 0) and the solid line indicates the regression lines forequation119910 = 125119909 minus 0846 (with intercept = not0846)

the monoparametric model (3) for TAase activity Further-more it is important to note the occurrence of an overlappingdescriptor (Balaban topological index at substituent 2) fromour two QSAR models clearly indicates that TAase activitymediated by GS utilizing PA as acetoxy group donor wasleading to the antimycobacterial activity of PA

32 Binding Studies Blind docking calculationwas employedto identify potential binding sites of PA on the GS structureThe 2D-QSAR model developed by us showed the impor-tance of substituent 2 (C-7 position of PA) for the MTAaseactivity hence we have considered 7-NH-AMC (4) DAMC(6) and 7-AMC (13) as the model PA for the docking studyThe resulting protein-ligand conformations for the model PAwere found to be located on the surface region of the proteinaway from the known active site of Mtb GS Figure 5 showsthe representative binding modes of the best docked confor-mations for the three PA in the putative active site of Mtb GSAn important finding is that in all the docking poses obtainedfor DAMC 7-AMC and 7-NH-AMC a cation-120587 interaction isobserved between 120576-NH

3group of Lys4 and aromatic ring of

coumarin (Figure 5) DAMC is found to form an additional


Lys4

Ala78

Arg79Leu12

Asp8

(a)

Lys4

Asp8

Leu12

Lys4

AAsp8

Leu12

(b)

Lys4

Asp8

Leu12

Ala78

(c)

Figure 5 Cation-120587 interaction (represented as yellow cone) between side chain of Lys4 of Mtb GS carrying net positive charge and aromaticrings of PA (a) Simultaneous formation of H-bond (represented as green dotted line) is observed between 120576-NH2 group of Lys4 of MtbGS and O-atom at C-7 position of DAMC (b) interaction of 7-AMC with crystal structure of Mtb GS (c) interaction of 7-NH-AMC withthe crystal structure of Mtb GS Cation-120587 interaction occurs when the distance between a positively ionisable atom and the centroid of anaromatic ring is equal to or less than 40 A and the angle between the normal vector of the plane and the vector between the ionisable atomand the centroid is equal to or greater than 45∘ and less than 90∘ [30] All the three interactions are in the permissible limits of the cation-120587interaction (as labeled in the figure)

H-bond between oxygen atom of C-7 acetyl group and 120576-NH3group of Lys4 (Figure 5(a))The cation-120587 interaction is a

non-covalent interaction of a positively charged cationwith120587electrons of an aromatic group Experimental and ab initiocalculations indicated that this interaction is influenced byelectrostatic forces between the monopole (cation) and thelarge quadrupole moment of the aromatic ring (120587-system)[30 36] Cation-120587 interactions involving the aromatic ringsof ligand and amino acids with a net positive charge (Arg orLys) have been reported to rationalize specific drug-receptorinteractions [37ndash39] Localization of ammonium-binding sitein the crystal structure of GS from Salmonella typhimurium(PDB ID 2GLS) has implicated a cation-120587 bonding betweenthe Tyr179 and ammonium ion [40] It is evident from theresults that PAs interact with Mtb GS by way of cation-120587interaction and such type of interaction may be conducivefor the transfer of acetyl group to the receptor protein byMtbGS The observation that quadrupolar XX moment is oneof the descriptor in the 2D-QSAR model very well validatethe cation-120587 interaction predicted by docking analysis for theMtb GS-PA interaction

33 ADMET Prediction Most of drug failures at early andlate pipeline occur due to undesired pharmacokinetics andtoxicity problems If these issues could be addressed earlyit would be extremely advantageous for the drug discoveryprocess In viewof these the use of in silicomethods to predictADMET properties is intended as a first step in this directionto analyze the novel chemical entities to prevent wasting timeon lead candidates that would be toxic or metabolized by thebody into an inactive form and unable to cross membranesand the results of such analysis are herein reported in Table 4together with a biplot (Figure 6) and discussed The phar-macokinetic profile of all the molecules under investigationwas predicted by means of six precalculated ADMETmodelsprovided by the Discovery Studio 21 program The biplotshows the two analogous 95 and 99 confidence ellipsescorresponding to HIA and BBB models PSA was shown tohave an inverse relationship (with percent human intestinalabsorption and thus cell wall permeability [41] Though arelationship of PSA to permeability has been demonstratedthe models usually do not take into account the effects ofother descriptors The fluid mosaic model of cell membrane


6

4

2

0

minus2

minus50 minus25 0 25 50 75 100 125 150

ADMET_PSA_2D

AD

ME

T_

Alo

gP

98

ADMET_AlogP98

ADMET_AlogP98 versus ADMET_PSA_2D

119

1012

8

614

12

354

713

Absorption-95

Absorption-99

BBB-95

BBB-99

Figure 6 Prediction of drug absorption for various PA consideredfor anti-mycobacterial activity Discovery Studio 21 (Accelrys SanDiego CA) ADMET Descriptors 2D polar surface area (PSA 2D)in A2 for each compound is plotted against their correspondingcalculated atom-type partition coefficient (ALogP98) The areaencompassed by the ellipse is a prediction of good absorption withno violation of ADMET properties On the basis of Egan et al[19] absorption model the 95 and 99 confidence limit ellipsescorresponding to the Blood Brain Barrier (BBB) and IntestinalAbsorption models are indicated

suggests that themembrane phospholipid bilayer is capable ofhydrophobic and hydrophilic interactions hence lipophilic-ity is also considered as a pivotal property for drug designLipophilicity could be assessed as the log of the partitioncoefficient between n-octanol andwater (log P)Though log Pis generally used to estimate a compoundrsquos lipophilicity thefact that log P is a ratio raises a concern about the use oflog P to estimate hydrophilicity and hydrophobicity Thusthe information of H-bonding characteristics as obtained bycalculating PSA could be taken into consideration along withlogP calculation [19] Therefore a model with descriptorsAlogP98 and PSA 2Dwith a bi-plot comprising 95 and 99confidence ellipseswas considered for the accurate predictionfor the cell permeability of compounds The 95 confidenceellipse represents the region of chemical space where we canexpect to find well-absorbed compounds (ge90) 95 out of100 times Whereas 99 is a confidence ellipse represents theregion of chemical space with compounds having excellentabsorption through cell membrane According to the modelfor a compound to have an optimum cell permeability shouldfollow the criteria (PSA lt 140 A2 and AlogP98 lt 5) [19] Allthe compounds showed polar surface area (PSA) lt 140 A2Considering the AlogP98 criteria all PAs had AlogP98 valuelt5 except compound 7 that has also in turn violated the 99and 95 confidence ellipse for both HIA and BBB (Figure 6)Table 4 shows that majority of the compounds have low orundefined values for BBB penetration levels (levels 3 and 4as mentioned in Table 2) with the exception of compound7 having high value and compound 18 having medium BBBpenetration level The aqueous solubility plays a critical role

in the bioavailability of the candidate drugs and with theexception of compound 7 having low aqueous solubility level(level 2) as referred in Table 2 all other PAs are having goodor optimal aqueous solubility levels Further all compoundshave been predicted to have hepatotoxicity level of 0 Themodel was developed from available literature data of 382compounds known to exhibit liver toxicity (ie positivedose-dependent hepatocellular cholestatic neoplastic etc)or trigger dose-related elevated aminotransferase levels inmore than 10 of the human population [24] The modelclassifies compounds either as ldquotoxicrdquo or ldquonontoxicrdquo andprovides a confidence level indicator of the likelihood of themodels predictive accuracy (Table 2) Our results indicatethat all PA are nontoxic to liver (level 0 Table 2) and thus theyexperience significant first-pass effect Similarly all ligandsare satisfactory with respect to CYP2D6 liver (with referenceto Table 2) suggesting that PA are noninhibitors of CYP2D6(Table 4) This indicates that all PAs are well metabolizedin Phase-I metabolism Finally the ADMET plasma proteinbinding property prediction denotes that all of 14 PAs withan exception of compounds 6 and 7 have binding ge90 andge95 respectively (refer to Table 2) clearly suggesting thatmost PAs have good bioavailability and are not likely to behighly bound to carrier proteins in the blood An interestingobservation was that the dihydroxy analogue of PA that is78-dihydroxy-4-methylcoumarin (DHMC) (compound 14)which is the deacetylated product of MTAase activity wasalso found to pass the entire ADMET test This observa-tion denotes that even by product of MTAase reaction isnontoxic

4 Conclusion

We have made an effort to develop QSAR models using thekinetic constants and the MIC values to address the fact thatTAase activity was leading to the antimycobacterial activityThe study indicated that Balaban index at C-7 position of PAwas the only contributing descriptor forMTAase activityTheBalaban index number of nitrogen atomatC-7 position of PAand quadrupole XX component (whole molecule) showeda good contribution to the antimycobacterial activity Ourobservation of an overlapping descriptor (Balaban topolog-ical index at substituent 2) from our two QSAR models thusclearly indicates that TAase activity mediated by GS utilizingPA as acetoxy group donor was leading to the antimycobacte-rial activity of PA Furthermajority of PAs were found to havefavorable ADMET characteristics ADMET studies provedthat PA can be developed as a potential antimycobacterialdrug The deacetylated product of TAase activity DHMCwas also found to pass the entire ADMET test An importantfinding is that in all the docking poses obtained for potent PAa cation-120587 interaction is observed between 120576-NH

3group of

Lys4 and aromatic ring of coumarin DAMC is found to forman additional H-bond between oxygen atom of C-7 acetylgroup and 120576-NH3 group of Lys4 Cation-120587 interactions resultessentially from a quadrupolar electrostatic interaction Theresults of QSAR and docking studies validated each other andprovided insight into the structural requirements for PA andMtb GS interaction


Abbreviations

MTAase Mycobacterial TAasePA Polyphenolic acetatesGS Calreticulin glutamine synthetaseDAMC 78-Diacetoxy-4-methylcoumarin7-AMC 7-acetoxy-4-methylcoumarin7-NH-AMC 7-NH-acetoxy-4-methylcoumarinQSAR Quantitative structure activity

relationshipADMET Absorption distribution metabolism

elimination toxicityPSA Polar surface area

Acknowledgments

The financial assistance of the Department of BiotechnologyGovt of New Delhi India is gratefully acknowledged Thisresearch was partially supported by grants from the Ministryof Chemicals and Fertilizers Government of India India

References

[1] H G Raj V S Parmar S C Jain et al ldquoMechanism ofbiochemical action of substituted 4-methylbenzopyran-2-onesPart 4 hyperbolic activation of rat liver microsomal nadph-cytochrome C reductase by the novel acetylator 78-diacetoxy-4-methylcoumarinrdquo Bioorganic amp Medicinal Chemistry vol 7no 2 pp 369ndash373 1999

[2] H G Raj V S Parmar S C Jain et al ldquoMechanismof biochemical action of substituted 4-methylbenzopyran-2-ones Part 7 assay and characterization of 78-diacetoxy-4-methylcoumarinprotein transacetylase from rat liver micro-somes based on the irreversible inhibition of cytosolic glu-tathione S-Transferaserdquo Bioorganic amp Medicinal Chemistry vol8 no 7 pp 1707ndash1712 2000

[3] P Khurana R Kumari P Vohra et al ldquoAcetoxy drug proteintransacetylase catalyzed activation of human platelet nitricoxide synthase by polyphenolic peracetatesrdquo Bioorganic ampMedicinal Chemistry vol 14 pp 575ndash583 2006

[4] H G Raj R Kumari S Bansal et al ldquoNovel function ofcalreticulin characterization of calreticulin as a transacetylase-mediating protein acetylator independent of acetyl CoA usingpolyphenolic acetates rdquo Pure and Applied Chemistry vol 78 pp985ndash992 2006

[5] Seema R Kumari G Gupta et al ldquoCharacterization of proteintransacetylase from human placenta as a signaling moleculecalreticulin using polyphenolic peracetates as the acetyl groupdonorsrdquo Cell Biochemistry and Biophysics vol 47 pp 53ndash642007

[6] E Kohli M Gaspari H G Raj et al ldquoAcetoxy drug pro-tein transacetylase of buffalo livermdashcharacterization and massspectrometry of the acetylated protein productrdquo Biochimica EtBiophysica Acta vol 1698 pp 55ndash66 2004

[7] S Bansal M Gaspari H G Raj et al ldquoCalreticulin transacety-lase mediates the acetylation of nitric oxide synthase bypolyphenolic acetaterdquo Applied Biochemistry and Biotechnologyvol 144 pp 37ndash45 2008

[8] G Gupta A S Baghel S Bansal et al ldquoEstablishment ofglutamine synthetase ofMycobacterium smegmatis as a proteinacetyltransferase utilizing polyphenolic acetates as the acetyl

group donorsrdquo Journal of Biochemistry vol 144 no 6 pp 709ndash715 2008

[9] A S Baghel R Tandon G Gupta et al ldquoCharacterization ofprotein acyltransferase function of recombinant purified GlnA1from Mycobacterium tuberculosis a moon lighting propertyrdquoMicrobiological Research vol 166 pp 662ndash672 2011

[10] G RHirschfieldMMcNeil and P J Brennan ldquoPeptidoglycan-associated polypeptides ofMycobacterium tuberculosisrdquo Journalof Bacteriology vol 172 no 2 pp 1005ndash1013 1990

[11] G Harth D L Clemens M A Horwitz et al ldquoGlutaminesynthetase of Mycobacterium tuberculosis extracellular releaseand characterization of its enzymatic activityrdquo Proceedings of theNational Academy of Sciences of theUnited States of America vol91 pp 9342ndash9346 1994

[12] O W Griffith and A Meister ldquoDifferential inhibition of glu-tamine and 120574-glutamylcysteine synthetases by 120572-alkyl analogsof methionine sulfoximine that induce convulsionsrdquo Journal ofBiological Chemistry vol 253 no 7 pp 2333ndash2338 1978

[13] B Lejczak H Starzemska and P Mastalerz ldquoInhibition of ratliver glutamine synthetase by phosphonic analogues of glutamicacidrdquo Experientia vol 37 no 5 pp 461ndash462 1981

[14] R Tandon P Ponnan N Aggarwal et al ldquoCharacterizationof 7-amino-4-methylcoumarin as an effective antitubercularagent structure-activity relationshipsrdquo Journal of AntimicrobialChemotherapy vol 66 pp 2543ndash2555 2011

[15] A Kathuria A Gupta N Priya et al ldquoSpecificities of cal-reticulin transacetylase to acetoxy derivatives of 3-alkyl-4-methylcoumarins effect on the activation of nitric oxide syn-thaserdquo Bioorganic ampMedicinal Chemistry vol 17 pp 1550ndash15562009

[16] Hyperchem Release8 Windows Molecular Modelling SystemHypercube Inc and Autodesk Inc Developed by HypercubeInc

[17] A Golbraikh and A Tropsha ldquoBeware of q2rdquo Journal ofMolecular Graphics and Modelling vol 20 no 4 pp 269ndash2762002

[18] A Tropsha PGramatica andVKGombar ldquoThe importance ofbeing earnest validation is the absolute essential for successfulapplication and interpretation of QSPR modelsrdquo QSAR andCombinatorial Science vol 22 no 1 pp 69ndash77 2003

[19] W J Egan K M Merz and J J Baldwin ldquoPrediction of drugabsorption using multivariate statisticsrdquo Journal of MedicinalChemistry vol 43 no 21 pp 3867ndash3877 2000

[20] A Cheng and KMMerz ldquoPrediction of aqueous solubility of adiverse set of compounds using quantitative structure-propertyrelationshipsrdquo Journal ofMedicinal Chemistry vol 46 no 17 pp3572ndash3580 2003

[21] W J Egan and G Lauri ldquoPrediction of intestinal permeabilityrdquoAdvanced Drug Delivery Reviews vol 54 no 3 pp 273ndash2892002

[22] S L Dixon and K M Merz ldquoOne-dimensional molecularrepresentations and similarity calculations methodology andvalidationrdquo Journal of Medicinal Chemistry vol 44 no 23 pp3795ndash3809 2001

[23] R G Susnow and S L Dixon ldquoUse of robust classificationtechniques for the prediction of human cytochrome P450 2D6inhibitionrdquo Journal of Chemical Information and ComputerSciences vol 43 pp 1308ndash1315 2003

[24] A Cheng and S L Dixon ldquoIn silico models for the predictionof dose-dependent humanhepatotoxicityrdquo Journal of Computer-Aided Molecular Design vol 17 no 12 pp 811ndash823 2003


[25] C Hetenyi and D Spoelvander ldquoEfficient docking of peptidesto proteins without prior knowledge of the binding siterdquo ProteinScience vol 11 pp 1729ndash1737 2002

[26] G M Morris D S Goodsell R S Halliday et al ldquoAutomateddocking using a Lamarckian genetic algorithm and an empiricalbinding free energy functionrdquo Journal of Computational Chem-istry vol 19 no 14 pp 1639ndash1662 1998

[27] W W Krajewski A T Jones S L Mowbray et al ldquoStructureofMycobacterium tuberculosis glutamine synthetase in complexwith a transition-state mimic provides functional insightsrdquoProceedings of the National Academy of Sciences of the UnitedStates of America vol 102 pp 10499ndash10504 2005

[28] M F Sanner B S Duncan C J Carrillo et al ldquoProteinmorpho-sis a mechanical model for protein conformational changesrdquo inProceedings of the Pacific Symposium in Biocomputing (PSB rsquo99)pp 401ndash412 Big Island Hawaii USA 1999

[29] T J A Ewing and I D Kuntz ldquoCritical evaluation of searchalgorithms for automated molecular docking and databasescreeningrdquo Journal of Computational Chemistry vol 18 no 9pp 1175ndash1189 1997

[30] D A Dougherty ldquoCation-120587 interactions in chemistry andbiology a new view of benzene Phe Tyr and Trprdquo Science vol271 no 5246 pp 163ndash168 1996

[31] A T Balaban ldquoHighly discriminating distance-based topologi-cal indexrdquo Chemical Physics Letters vol 89 pp 399ndash404 1982

[32] D Mandloi S Joshi P V Khadikar et al ldquoQSAR study on theantibacterial activity of some sulfa drugs building blockers ofMannich basesrdquo Bioorganic amp Medicinal Chemistry Letters vol15 pp 405ndash411 2005

[33] S C Basak D P Gieschen D K Harriss and V R MagnusonldquoPhysicochemical and topological correlates of the enzymaticacetyltransfer reactionrdquo Journal of Pharmaceutical Sciences vol72 no 8 pp 934ndash937 1983

[34] P Singh P Ponnan S Krishnan et al ldquoProtein acyltransferasefunction of purified calreticulin Part 1 characterization ofpropionylation of protein utilizing propoxycoumarin as thepropionyl group donorrdquo Journal of Biochemistry vol 147 no 5pp 625ndash632 2010

[35] Y Chen R Sprung Y Tang et al ldquoLysine propionylationand butyrylation are novel post-translational modifications inhistonesrdquo Molecular amp Cellular Proteomics vol 6 pp 812ndash8192007

[36] J HWilliams ldquoThemolecular electric quadrupolemoment andsolid-state architecturerdquo Accounts of Chemical Research vol 26pp 593ndash598 1993

[37] M Dennis J Giraudat F Kotzyba-Hibert et al ldquoAmino acids ofthe torpedomarmorata acetylcholine receptor120572 subunit labeledby a photoaffinity ligand for the acetylcholine binding siterdquoBiochemistry vol 27 no 7 pp 2346ndash2357 1988

[38] P D Leeson R Baker R W Carling et al ldquoAmino acidbioisosteres design of 2-quinolone derivatives as glycine-siteN-methyl-D-aspartate receptor antagonistsrdquo Bioorganic amp Medic-inal Chemistry Letters vol 3 pp 299ndash304 1993

[39] B Yang J Wright M E Eldefrawi S Pou and A DMacKerellldquoConformational aqueous solvation and pK(a) contributionsto the binding and activity of cocaine WIN 32065-2 and theWIN vinyl analogrdquo Journal of the American Chemical Societyvol 116 no 19 pp 8722ndash8732 1994

[40] S H Liaw I Kuo and D Eisenberg ldquoDiscovery of the ammon-ium substrate site on glutamine synthetase a third cationbinding siterdquo Protein Science vol 4 no 11 pp 2358ndash2365 1995

[41] K Palm P Stenberg K Luthman and P Artursson ldquoPolarmolecular surface properties predict the intestinal absorptionof drugs in humansrdquo Pharmaceutical Research vol 14 no 5 pp568ndash571 1997

Submit your manuscripts athttpwwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Anatomy Research International

PeptidesInternational Journal of


Hindawi Publishing Corporation httpwwwhindawicom

International Journal of

Volume 2014

Zoology


Molecular Biology International

GenomicsInternational Journal of


The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014


BioinformaticsAdvances in

Marine BiologyJournal of



Signal TransductionJournal of


BioMed Research International

Evolutionary BiologyInternational Journal of



Biochemistry Research International

ArchaeaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014


Genetics Research International


Advances in

Virolog y

Hindawi Publishing Corporationhttpwwwhindawicom

Nucleic AcidsJournal of

Volume 2014

Stem CellsInternational



Enzyme Research



Microbiology


tuberculosis (Mtb) H37Rv [9] as glutamine synthetase (GS)Glutamine synthetase catalyzes the conversion of glutamateto glutamine in the presence of ammonium ion with simul-taneous hydrolysis of ATP which is used as the energy sourceand plays an essential role in bacterial nitrogen metabolism[10 11] Several PAs including acetoxycoumarins in generalwere found to be the substrates for mycobacterial TAase(MTAase) The specificities of various acetoxycoumarinstowards MTAase were determined by their ability to inhibitGST irreversibly and their kinetic constants (119870

119898and 119881max)

were determined [9] Several inhibitors are known for GSand most of them are analogues of glutamate and replacethis substrate in the active site of the enzyme Among theknown inhibitors methionine sulfoximine (MSO) and 2-amino-4-(hydroxymethyl-phosphoryl) butanoic acid (phos-phinothricin) are the well-established inhibitors of GS [1213] During the examination of the role of GS inhibitor onMTAase function of GS it was observed that MSO failedto inhibit MTAase-catalysed reaction indicating that theTAase activity of MTAase is independent of the catalyticactivity of GS [8 9] Electron microscopic studies carriedout by us have shown cell wall attacking properties of thesecompounds in M smegmatis [8] and M tuberculosis [14]The cell wall of Mycobacterium species is responsible formaintaining the cell integrity and thus is considered to be apotential drug target owing to its crucial role in cell survivaland viability If a compound is found to affect the cell wallor its biosynthesis in any manner it is bound to bring aboutthe inhibition of bacterial growth The PAs referred to inthe present study have been reported earlier [14] to possessldquocell-wall attackingrdquo characteristic that is these moleculeshave been found to bring about changes in cell morphologyranging from indentations in the wall to complete rupturingof cell wall along with extrusion of cytoplasmic material insome cases and complete disintegrationdisappearance of thewall in others [14]Moreover these pronounced changes wererecorded when the bacteria were grown in the presence ofsublethal doses of the test molecules These observations ledus to believe that these compounds may serve as potentialdrug candidates and therefore these were further exploredto determine their drug likeness and also establish a structureactivity relationship [14] The present work describes thedescriptor-based QSAR studies developed for a series of ace-toxycoumarins synthesized by us and evaluated forTAase andantimycobacterial activity Also ADMET-related descriptorswere calculated to predict the pharmacokinetic properties forthe selection of the effective and bioavailable compoundsFurther docking studies were done to analyze the interactionof the potent acetoxycoumarins with the crystal structure ofM tuberculosis GS

2 Methodology

21 2119863minusQSAR Analysis

211 Data Set and Methodology Compounds 1ndash14 were syn-thesized and characterized following the published syntheticprocedures [1ndash3 9 15] The PAs were screened for theirantimycobacterial activity and TAase activity that are listed

in Table 1 according to our published methods [8] utilizingMtbGS as the target protein Biological activity data reportedas MIC values for the antimycobacterial activity and kineticconstants (119870

119898and 119881max) for TAase activity (Table 1) were

first converted to -logMIC onmolar basis and log (119881max119870119898)respectively and were used as the dependent variables to getthe linear relationship in the QSAR models

Hyperchem-8 program [16] was used to build the struc-tures and perform geometry optimizations of the com-pounds The lowest energy conformations of the compoundswere determined first byminimizing the structures bymolec-ular mechanics method using MM+ force field followed bysemiempirical self-consistent field molecular orbital (SCFMO) theory (parametric model 3 (PM3) method withinthe restricted Hartree-Fock (RHF) formalism) Conjugategradientmethod (Polak-Ribiere algorithm)with SCF conver-gency set to 0001 kcalmol was considered in the geometryoptimization stage of calculations Frequency calculationshave been performed to confirm all stationary points

212 Molecular Descriptors for QSAR Analysis TSAR 33software package (Accelrys San Diego CA USA) wasemployed to calculate descriptors for entire molecule andthe defined substituents Substituents were defined for allPA a single hydrogen atom also served as a substituent(Table 1) TSAR includes various physicochemical topolog-ical and electrostatic descriptors molecular surface area andvolume molecular mass moments of inertia (moment 12 3 (size length)) ellipsoidal volume Verloop parametersDipole moments (total bond and x y z components)Lipole moments (total bond and x y z components)topological indices (Wiener Randic and Balaban indices)molecular connectivity indices (Chi ChiV indices) of atomsbonds path cluster and pathclusterMolecular shape indices(Kappa KAlpha indices) Electrotopological state indicesLogP Atom counts (CNS amp H) Ring count (aromaticand aliphatic) and Group count (methyl hydroxyl ethyl)Electrostatic properties like Total energy Electronic energyNuclear repulsion energy Accessible surface area Atomiccharge Mean polarizability Heat of formation HOMOand LUMO eigenvalues Ionization potential Total dipolePolarizability and Dipole components Pairwise correlationanalysis of the descriptors was performed and the intercor-related descriptors (gt06) were discarded depending on theirindividual correlation with the biological activity

213 Stepwise Multiple Regression In an effort to investigatethe role of structural parameters which appears to influencethe observed activities of reported compounds stepwisemultiple linear regressions were performed using TSAR 33software TSAR uses a two-way stepping algorithm to selectvariables for the regression equation At each step partial 119865values are calculated for each variable as an estimate of theirpotential contribution to the model The partial 119865 values arecompared with the 119865-to- Leave and 119865-to-Enter settings Theoverall 119865 statistic for a model is

119865 =

explained mean squareresidual mean square

(1)



O O

R2

R3

R1

1

2

34567

8


119870119898








0and 11990310158402


1199032

0)1199032

pred lt 01 or (1199032

pred minus 11990310158402

0)1199032


1015840

le 115 [17 18]



























025917 012484 0050123 51706 00020731



084199 010326 007821 10766 37971119890 minus 005


0064479 0028036 0032179 20037 0091947

C constant 40866 043577

MTAase activity


013387 0018757 0027883 48012 00007223








+ 010474976 lowast X3 + 4107533

(2)

119904 = 018 119865 = 4194 119903 = 096 1199032

= 093

1199022

LOO = 077 PRESS = 104



ext = 088 (1199032pred minus 1199032

0)1199032

pred = 0071(1199032pred minus 1199031015840

0

2

)1199032

pred lt



log (119881max119870119898) = 013387173 lowast X1 + 28492985 (3)

119904 = 0173 119865 = 2305 119903 = 0835

1199032

= 0697 1199022

LOO = 0609 PRESS = 0387



Table4ADMET

predictio

nof

PAs

ADMET

absorptio

nlevel

ADMET

AlogP

98

ADMET

unkn

own

AlogP

98

ADMET

PSA

2D

ADMET

BBB

level

ADMET

BBB

ADMET

solubility

ADMET

solubility

level

ADMET

hepatotoxicity

ADMET

hepato-

toxicity

prob

ability

ADMET

CYP2

D6

ADMET

CYP2

D6

prob

ability

ADMET

PPBlevel

10

0345

0119

4minus10

64

00019

00455

0

20

0594

05949

3minus091

minus10

54

0006

60

0029

03

013

390

5056

3minus054

minus16

94

00052

00118

04

00328

06337

3minus10

6minus093

40

0052

00029

05

0278

08924

3minus071

minus272

30

0052

00455

06

04605

08924

4minus372

30

006

60

0435

1

70

5149

05949

10496

minus426

20

0052

0040

52

80

0051

08924

3minus15

5minus075

40

0059

00277

09

0minus001

07138

3minus12

9minus019

40

004

60

0029

010

2minus13

30

1309

40014

50

0086

00247

0

111

minus097

11342

4minus074

40

0052

00277

0

120

2269

08924

3minus087

minus233

30

0152

00366

013

10213

01309

4minus113

40

0039

00386

0

140

1357

08924

3minus115

minus17

24

0006

60

0316

0


442444648

552545658

6

4 42 44 46 48 5 52 54 56 58 6


Pred

icte

dminuslog

(MIC

)


(a)

001020304

1 2 3 4 5 6 7 8 9 10 11 12 13 14

Resid

ual v

alue

s


(b)



2

0)1199032

pred =

0078 (1199032pred minus 1199031015840

0

2

)1199032



119869 (119866) =

119864

120583 + 1

sum

edges(119889119904119894119889119904119895)

minus12

(4)


119894(119894 = 1 2 119873119873

4

45

5

55

6

65

4 42 44 46 48 5 52 54 56 58 6

Predictedminuslog(M

IC)


1199100 = 10265119909

11990320 = 08893

119910 = 0811119909 + 1032

1199032pred = 0957

(a)

4424446485

525456586

4 42 44 46 48 5 52 54 56 58 6


lculatedminuslog(M

IC)

1199039984002 = 09576

119910 = 11803119909 minus 1018

11990399840020 = 09279

1199100 = 09732119909

(b)





0= 09732119909 with





119889119904119894=

119873

sum

119895=1

119863119894119895 (5)



25

27

29

31

33

35

37

39

25 27 29 31 33 35 37 39 41

Training set

Test set

(a)

0

01

02

1 2 3 4 5 6 7 8 9 10 11 12 13

Res

idu

al v

alu

es

minus04

minus03

minus02

minus01

minus05

(b)




3313233343536

3 31 32 33 34 35 36 37

1199032pred = 0978

119910 = 0761119909 + 0714

11990320 = 09006

1199100 = 09759119909

Predictedlog(119881

max119870119898)


(a)

3

31

32

33

34

35

36

37

3 31 32 33 34 35 36

(b)





0= 10245119909 with







Lys4

Ala78

Arg79Leu12

Asp8

(a)

Lys4

Asp8

Leu12

Lys4

AAsp8

Leu12

(b)

Lys4

Asp8

Leu12

Ala78

(c)






6

4

2

0

minus2

minus50 minus25 0 25 50 75 100 125 150

ADMET_PSA_2D

AD

ME

T_

Alo

gP

98

ADMET_AlogP98


119

1012

8

614

12

354

713

Absorption-95

Absorption-99

BBB-95

BBB-99




4 Conclusion


3group of



Abbreviations




Acknowledgments


References



















































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology



O O

R2

R3

R1

1

2

34567

8


119870119898








0and 11990310158402


1199032

0)1199032

pred lt 01 or (1199032

pred minus 11990310158402

0)1199032


1015840

le 115 [17 18]



























025917 012484 0050123 51706 00020731



084199 010326 007821 10766 37971119890 minus 005


0064479 0028036 0032179 20037 0091947

C constant 40866 043577

MTAase activity


013387 0018757 0027883 48012 00007223








+ 010474976 lowast X3 + 4107533

(2)

119904 = 018 119865 = 4194 119903 = 096 1199032

= 093

1199022

LOO = 077 PRESS = 104



ext = 088 (1199032pred minus 1199032

0)1199032

pred = 0071(1199032pred minus 1199031015840

0

2

)1199032

pred lt



log (119881max119870119898) = 013387173 lowast X1 + 28492985 (3)

119904 = 0173 119865 = 2305 119903 = 0835

1199032

= 0697 1199022

LOO = 0609 PRESS = 0387



Table4ADMET

predictio

nof

PAs

ADMET

absorptio

nlevel

ADMET

AlogP

98

ADMET

unkn

own

AlogP

98

ADMET

PSA

2D

ADMET

BBB

level

ADMET

BBB

ADMET

solubility

ADMET

solubility

level

ADMET

hepatotoxicity

ADMET

hepato-

toxicity

prob

ability

ADMET

CYP2

D6

ADMET

CYP2

D6

prob

ability

ADMET

PPBlevel

10

0345

0119

4minus10

64

00019

00455

0

20

0594

05949

3minus091

minus10

54

0006

60

0029

03

013

390

5056

3minus054

minus16

94

00052

00118

04

00328

06337

3minus10

6minus093

40

0052

00029

05

0278

08924

3minus071

minus272

30

0052

00455

06

04605

08924

4minus372

30

006

60

0435

1

70

5149

05949

10496

minus426

20

0052

0040

52

80

0051

08924

3minus15

5minus075

40

0059

00277

09

0minus001

07138

3minus12

9minus019

40

004

60

0029

010

2minus13

30

1309

40014

50

0086

00247

0

111

minus097

11342

4minus074

40

0052

00277

0

120

2269

08924

3minus087

minus233

30

0152

00366

013

10213

01309

4minus113

40

0039

00386

0

140

1357

08924

3minus115

minus17

24

0006

60

0316

0


442444648

552545658

6

4 42 44 46 48 5 52 54 56 58 6


Pred

icte

dminuslog

(MIC

)


(a)

001020304

1 2 3 4 5 6 7 8 9 10 11 12 13 14

Resid

ual v

alue

s


(b)



2

0)1199032

pred =

0078 (1199032pred minus 1199031015840

0

2

)1199032



119869 (119866) =

119864

120583 + 1

sum

edges(119889119904119894119889119904119895)

minus12

(4)


119894(119894 = 1 2 119873119873

4

45

5

55

6

65

4 42 44 46 48 5 52 54 56 58 6

Predictedminuslog(M

IC)


1199100 = 10265119909

11990320 = 08893

119910 = 0811119909 + 1032

1199032pred = 0957

(a)

4424446485

525456586

4 42 44 46 48 5 52 54 56 58 6


lculatedminuslog(M

IC)

1199039984002 = 09576

119910 = 11803119909 minus 1018

11990399840020 = 09279

1199100 = 09732119909

(b)





0= 09732119909 with





119889119904119894=

119873

sum

119895=1

119863119894119895 (5)



25

27

29

31

33

35

37

39

25 27 29 31 33 35 37 39 41

Training set

Test set

(a)

0

01

02

1 2 3 4 5 6 7 8 9 10 11 12 13

Res

idu

al v

alu

es

minus04

minus03

minus02

minus01

minus05

(b)




3313233343536

3 31 32 33 34 35 36 37

1199032pred = 0978

119910 = 0761119909 + 0714

11990320 = 09006

1199100 = 09759119909

Predictedlog(119881

max119870119898)


(a)

3

31

32

33

34

35

36

37

3 31 32 33 34 35 36

(b)





0= 10245119909 with







Lys4

Ala78

Arg79Leu12

Asp8

(a)

Lys4

Asp8

Leu12

Lys4

AAsp8

Leu12

(b)

Lys4

Asp8

Leu12

Ala78

(c)






6

4

2

0

minus2

minus50 minus25 0 25 50 75 100 125 150

ADMET_PSA_2D

AD

ME

T_

Alo

gP

98

ADMET_AlogP98


119

1012

8

614

12

354

713

Absorption-95

Absorption-99

BBB-95

BBB-99




4 Conclusion


3group of



Abbreviations




Acknowledgments


References



















































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology


























025917 012484 0050123 51706 00020731



084199 010326 007821 10766 37971119890 minus 005


0064479 0028036 0032179 20037 0091947

C constant 40866 043577

MTAase activity


013387 0018757 0027883 48012 00007223








+ 010474976 lowast X3 + 4107533

(2)

119904 = 018 119865 = 4194 119903 = 096 1199032

= 093

1199022

LOO = 077 PRESS = 104



ext = 088 (1199032pred minus 1199032

0)1199032

pred = 0071(1199032pred minus 1199031015840

0

2

)1199032

pred lt



log (119881max119870119898) = 013387173 lowast X1 + 28492985 (3)

119904 = 0173 119865 = 2305 119903 = 0835

1199032

= 0697 1199022

LOO = 0609 PRESS = 0387



Table4ADMET

predictio

nof

PAs

ADMET

absorptio

nlevel

ADMET

AlogP

98

ADMET

unkn

own

AlogP

98

ADMET

PSA

2D

ADMET

BBB

level

ADMET

BBB

ADMET

solubility

ADMET

solubility

level

ADMET

hepatotoxicity

ADMET

hepato-

toxicity

prob

ability

ADMET

CYP2

D6

ADMET

CYP2

D6

prob

ability

ADMET

PPBlevel

10

0345

0119

4minus10

64

00019

00455

0

20

0594

05949

3minus091

minus10

54

0006

60

0029

03

013

390

5056

3minus054

minus16

94

00052

00118

04

00328

06337

3minus10

6minus093

40

0052

00029

05

0278

08924

3minus071

minus272

30

0052

00455

06

04605

08924

4minus372

30

006

60

0435

1

70

5149

05949

10496

minus426

20

0052

0040

52

80

0051

08924

3minus15

5minus075

40

0059

00277

09

0minus001

07138

3minus12

9minus019

40

004

60

0029

010

2minus13

30

1309

40014

50

0086

00247

0

111

minus097

11342

4minus074

40

0052

00277

0

120

2269

08924

3minus087

minus233

30

0152

00366

013

10213

01309

4minus113

40

0039

00386

0

140

1357

08924

3minus115

minus17

24

0006

60

0316

0


442444648

552545658

6

4 42 44 46 48 5 52 54 56 58 6


Pred

icte

dminuslog

(MIC

)


(a)

001020304

1 2 3 4 5 6 7 8 9 10 11 12 13 14

Resid

ual v

alue

s


(b)



2

0)1199032

pred =

0078 (1199032pred minus 1199031015840

0

2

)1199032



119869 (119866) =

119864

120583 + 1

sum

edges(119889119904119894119889119904119895)

minus12

(4)


119894(119894 = 1 2 119873119873

4

45

5

55

6

65

4 42 44 46 48 5 52 54 56 58 6

Predictedminuslog(M

IC)


1199100 = 10265119909

11990320 = 08893

119910 = 0811119909 + 1032

1199032pred = 0957

(a)

4424446485

525456586

4 42 44 46 48 5 52 54 56 58 6


lculatedminuslog(M

IC)

1199039984002 = 09576

119910 = 11803119909 minus 1018

11990399840020 = 09279

1199100 = 09732119909

(b)





0= 09732119909 with





119889119904119894=

119873

sum

119895=1

119863119894119895 (5)



25

27

29

31

33

35

37

39

25 27 29 31 33 35 37 39 41

Training set

Test set

(a)

0

01

02

1 2 3 4 5 6 7 8 9 10 11 12 13

Res

idu

al v

alu

es

minus04

minus03

minus02

minus01

minus05

(b)




3313233343536

3 31 32 33 34 35 36 37

1199032pred = 0978

119910 = 0761119909 + 0714

11990320 = 09006

1199100 = 09759119909

Predictedlog(119881

max119870119898)


(a)

3

31

32

33

34

35

36

37

3 31 32 33 34 35 36

(b)





0= 10245119909 with







Lys4

Ala78

Arg79Leu12

Asp8

(a)

Lys4

Asp8

Leu12

Lys4

AAsp8

Leu12

(b)

Lys4

Asp8

Leu12

Ala78

(c)






6

4

2

0

minus2

minus50 minus25 0 25 50 75 100 125 150

ADMET_PSA_2D

AD

ME

T_

Alo

gP

98

ADMET_AlogP98


119

1012

8

614

12

354

713

Absorption-95

Absorption-99

BBB-95

BBB-99




4 Conclusion


3group of



Abbreviations




Acknowledgments


References



















































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology


Table4ADMET

predictio

nof

PAs

ADMET

absorptio

nlevel

ADMET

AlogP

98

ADMET

unkn

own

AlogP

98

ADMET

PSA

2D

ADMET

BBB

level

ADMET

BBB

ADMET

solubility

ADMET

solubility

level

ADMET

hepatotoxicity

ADMET

hepato-

toxicity

prob

ability

ADMET

CYP2

D6

ADMET

CYP2

D6

prob

ability

ADMET

PPBlevel

10

0345

0119

4minus10

64

00019

00455

0

20

0594

05949

3minus091

minus10

54

0006

60

0029

03

013

390

5056

3minus054

minus16

94

00052

00118

04

00328

06337

3minus10

6minus093

40

0052

00029

05

0278

08924

3minus071

minus272

30

0052

00455

06

04605

08924

4minus372

30

006

60

0435

1

70

5149

05949

10496

minus426

20

0052

0040

52

80

0051

08924

3minus15

5minus075

40

0059

00277

09

0minus001

07138

3minus12

9minus019

40

004

60

0029

010

2minus13

30

1309

40014

50

0086

00247

0

111

minus097

11342

4minus074

40

0052

00277

0

120

2269

08924

3minus087

minus233

30

0152

00366

013

10213

01309

4minus113

40

0039

00386

0

140

1357

08924

3minus115

minus17

24

0006

60

0316

0


442444648

552545658

6

4 42 44 46 48 5 52 54 56 58 6


Pred

icte

dminuslog

(MIC

)


(a)

001020304

1 2 3 4 5 6 7 8 9 10 11 12 13 14

Resid

ual v

alue

s


(b)



2

0)1199032

pred =

0078 (1199032pred minus 1199031015840

0

2

)1199032



119869 (119866) =

119864

120583 + 1

sum

edges(119889119904119894119889119904119895)

minus12

(4)


119894(119894 = 1 2 119873119873

4

45

5

55

6

65

4 42 44 46 48 5 52 54 56 58 6

Predictedminuslog(M

IC)


1199100 = 10265119909

11990320 = 08893

119910 = 0811119909 + 1032

1199032pred = 0957

(a)

4424446485

525456586

4 42 44 46 48 5 52 54 56 58 6


lculatedminuslog(M

IC)

1199039984002 = 09576

119910 = 11803119909 minus 1018

11990399840020 = 09279

1199100 = 09732119909

(b)





0= 09732119909 with





119889119904119894=

119873

sum

119895=1

119863119894119895 (5)



25

27

29

31

33

35

37

39

25 27 29 31 33 35 37 39 41

Training set

Test set

(a)

0

01

02

1 2 3 4 5 6 7 8 9 10 11 12 13

Res

idu

al v

alu

es

minus04

minus03

minus02

minus01

minus05

(b)




3313233343536

3 31 32 33 34 35 36 37

1199032pred = 0978

119910 = 0761119909 + 0714

11990320 = 09006

1199100 = 09759119909

Predictedlog(119881

max119870119898)


(a)

3

31

32

33

34

35

36

37

3 31 32 33 34 35 36

(b)





0= 10245119909 with







Lys4

Ala78

Arg79Leu12

Asp8

(a)

Lys4

Asp8

Leu12

Lys4

AAsp8

Leu12

(b)

Lys4

Asp8

Leu12

Ala78

(c)






6

4

2

0

minus2

minus50 minus25 0 25 50 75 100 125 150

ADMET_PSA_2D

AD

ME

T_

Alo

gP

98

ADMET_AlogP98


119

1012

8

614

12

354

713

Absorption-95

Absorption-99

BBB-95

BBB-99




4 Conclusion


3group of



Abbreviations




Acknowledgments


References



















































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology


442444648

552545658

6

4 42 44 46 48 5 52 54 56 58 6


Pred

icte

dminuslog

(MIC

)


(a)

001020304

1 2 3 4 5 6 7 8 9 10 11 12 13 14

Resid

ual v

alue

s


(b)



2

0)1199032

pred =

0078 (1199032pred minus 1199031015840

0

2

)1199032



119869 (119866) =

119864

120583 + 1

sum

edges(119889119904119894119889119904119895)

minus12

(4)


119894(119894 = 1 2 119873119873

4

45

5

55

6

65

4 42 44 46 48 5 52 54 56 58 6

Predictedminuslog(M

IC)


1199100 = 10265119909

11990320 = 08893

119910 = 0811119909 + 1032

1199032pred = 0957

(a)

4424446485

525456586

4 42 44 46 48 5 52 54 56 58 6


lculatedminuslog(M

IC)

1199039984002 = 09576

119910 = 11803119909 minus 1018

11990399840020 = 09279

1199100 = 09732119909

(b)





0= 09732119909 with





119889119904119894=

119873

sum

119895=1

119863119894119895 (5)



25

27

29

31

33

35

37

39

25 27 29 31 33 35 37 39 41

Training set

Test set

(a)

0

01

02

1 2 3 4 5 6 7 8 9 10 11 12 13

Res

idu

al v

alu

es

minus04

minus03

minus02

minus01

minus05

(b)




3313233343536

3 31 32 33 34 35 36 37

1199032pred = 0978

119910 = 0761119909 + 0714

11990320 = 09006

1199100 = 09759119909

Predictedlog(119881

max119870119898)


(a)

3

31

32

33

34

35

36

37

3 31 32 33 34 35 36

(b)





0= 10245119909 with







Lys4

Ala78

Arg79Leu12

Asp8

(a)

Lys4

Asp8

Leu12

Lys4

AAsp8

Leu12

(b)

Lys4

Asp8

Leu12

Ala78

(c)






6

4

2

0

minus2

minus50 minus25 0 25 50 75 100 125 150

ADMET_PSA_2D

AD

ME

T_

Alo

gP

98

ADMET_AlogP98


119

1012

8

614

12

354

713

Absorption-95

Absorption-99

BBB-95

BBB-99




4 Conclusion


3group of



Abbreviations




Acknowledgments


References



















































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology


25

27

29

31

33

35

37

39

25 27 29 31 33 35 37 39 41

Training set

Test set

(a)

0

01

02

1 2 3 4 5 6 7 8 9 10 11 12 13

Res

idu

al v

alu

es

minus04

minus03

minus02

minus01

minus05

(b)




3313233343536

3 31 32 33 34 35 36 37

1199032pred = 0978

119910 = 0761119909 + 0714

11990320 = 09006

1199100 = 09759119909

Predictedlog(119881

max119870119898)


(a)

3

31

32

33

34

35

36

37

3 31 32 33 34 35 36

(b)





0= 10245119909 with







Lys4

Ala78

Arg79Leu12

Asp8

(a)

Lys4

Asp8

Leu12

Lys4

AAsp8

Leu12

(b)

Lys4

Asp8

Leu12

Ala78

(c)






6

4

2

0

minus2

minus50 minus25 0 25 50 75 100 125 150

ADMET_PSA_2D

AD

ME

T_

Alo

gP

98

ADMET_AlogP98


119

1012

8

614

12

354

713

Absorption-95

Absorption-99

BBB-95

BBB-99




4 Conclusion


3group of



Abbreviations




Acknowledgments


References



















































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology


Lys4

Ala78

Arg79Leu12

Asp8

(a)

Lys4

Asp8

Leu12

Lys4

AAsp8

Leu12

(b)

Lys4

Asp8

Leu12

Ala78

(c)






6

4

2

0

minus2

minus50 minus25 0 25 50 75 100 125 150

ADMET_PSA_2D

AD

ME

T_

Alo

gP

98

ADMET_AlogP98


119

1012

8

614

12

354

713

Absorption-95

Absorption-99

BBB-95

BBB-99




4 Conclusion


3group of



Abbreviations




Acknowledgments


References



















































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology


6

4

2

0

minus2

minus50 minus25 0 25 50 75 100 125 150

ADMET_PSA_2D

AD

ME

T_

Alo

gP

98

ADMET_AlogP98


119

1012

8

614

12

354

713

Absorption-95

Absorption-99

BBB-95

BBB-99




4 Conclusion


3group of



Abbreviations




Acknowledgments


References



















































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology


Abbreviations




Acknowledgments


References



















































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology


























Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology

in silico admet · pharmacokinetic properties for the selection of the e ective and bioavailable...

Documents