Annals of West University of Timisoara

Annals of West University of Timisoara Series Chemistry 19 (4) (2010) 59-64

PRELIMINARY FUJITA – BAN STUDY FOR GSK-3 INHIBITION BY INDIRUBIN DERIVATIVES L umi ni t a C ri ş a n a , L . Kur un c zi b a

Institute of Chemistry of Romanian Academy, 24 Mihai Viteazul Bvd., 300223

Timisoara, ROMANIA. b

University of Medicine and Pharmacy “Victor Babes”, Faculty of Pharmacy, 2 Eftimie

Murgu Avenue, 300041, Timisoara, ROMANIA Received: 27 November 2010

Modified 30 November 2010

Accepted 3 December 2010

SUMMARY Glycogen synthase kinase 3 (GSK-3) is involved in a variety of physiological processes and its inhibitors have been evaluated as promising drug candidates. Indirubin analogues showed favorable inhibitory activity on GSK-3β target, which is closely related to the nature and position of substituents. For this reason in the present study we carried out a preliminary study on indirubin derivatives as GSK-3 inhibitors using a Fujita Ban analysis, a simple QSAR technique that relates directly the structural features to biological activity. A statistically significant structure - activity correlation was obtained, suggested by several parameters including: squared correlation coefficient, R2=0.982, Fischer test, F = 77.780 and standard error of estimate, see = 0.146.

Keywords: indirubin, GSK-3, QSAR, Fujita –Ban

INTRODUCTION

Glycogen synthase kinase-3 (GSK-3) is a serine-threonine kinase encoded by two isoforms in mammals, termed GSK-3α and GSK-3β [1]. GSK-3 was originally identified as one of the five protein kinases that phosphorylate the rate-limiting enzyme of glycogen synthesis, glycogen synthase (GS)[1]. Since 1998 there have been many reports of GSK-3 inhibitors [2]. In this work we consider indirubin derivatives which were reported as

59

CRIŞAN L., KURUNCZI L.

selectively active against GSK-3 [3,4]. Indirubin, a red colored 3,2′-bisindole isomer, is a component of Indigo naturalis and is an active ingredient used in traditional Chinese medicine for the treatment of chronic diseases such as leukemia [5]. Indirubin and its derivatives (indigo type compounds) may have important implications in the development of therapies for a number of diseases such as ischemia–reperfusion, Alzheimer's disease, cancer and type 2 diabetes, as well as in stem cell therapy [5]. In the present work we have applied a Fujita Ban QSAR (Quantitative Structure – Activity Relationships) approach on a series of 51 indirubin derivatives[4,5] with known biological activity based on large structural variation in the indirubin data set (nature and size of the substituents) which is appropriate for our investigation. Our aim is to develop a robust model that suggests a quantitative 2D structure activity relationship of indirubin derivatives in terms of their affinity to GSK-3. MATERIALS

AND

METHODS

The Fujita-Ban modification of the Free - Wilson approach was applied for 51 compounds (Figure 1, Table I) in order to find out a 2D QSAR model. A Fujita-Ban analysis was carried out using a regressional technique in order to detect the contribution of the substituents to the activity of the current molecules. The 2D QSAR methods are usually faster and prone to computerize. 3'

5' 6'

O

N H O

5 6

N1 H

Figure1. The template of indirubin analogues

A Fujita-Ban analysis is appropriate if the series meets the following conditions: (i) all the compounds investigated must have the same scaffold, otherwise the different frame should be considered as an additional parameter; (ii) at least two positions of substitution in molecules to be varied; (iii) substituents in different positions of the molecule should not influence each other. The Fujita Ban equation is given by (equation 1):

Ai   

P

b

'

jk

X ijk

(1)

j 1

where, μ denotes biological activity of the H substituted common scaffold; i is the number

60

PRELIMINARY FUJITA –BAN STUDY FOR GSK-3 INHIBITION BY INDIRUBIN DERIVATIVES

of compound; j represents the substitution position; k is the placeholder index; Xijk = 1 when Rk is present in position j, and 0 otherwise; b’jk denotes the regression coefficients. The positive sign of b’jk coefficients are considered beneficial and the negative sign of these coefficients are considered detrimental. Table I. Structure of bis-indoles used in the current study No

pIC50

R1

R3'

R5

R5'

R6

001

6.000

H

O

H

H

H

R6' H

002

7.657

H

NOH

H

H

H

H

003

6.699

H

NOAc

H

H

H

H

004

6.824

H

NOCH3

H

H

H

H

005

5.347

H

O

H

H

Br

Br

006

6.921

H

NOH

H

H

Br

Br

007

4.657

H

O

H

H

H

Br

008

6.468

H

NOH

H

H

H

Br

009

7.347

H

O

H

H

Br

H

010

8.301

H

NOH

H

H

Br

H

011

8.000

H

NOAc

H

H

Br

H

012

7.523

H

NOCH3

H

H

Br

H

013

6.854

H

O

H

H

Cl

H

014

7.699

H

NOH

H

H

Cl

H

015

7.769

H

NOAc

H

H

Cl

H

016

7.259

H

O

H

H

I

H

017

8.000

H

NOH

H

H

I

H

018

7.886

H

NOAc

H

H

I

H

019

6.619

H

O

H

H

CH=CH2

H H

020

7.222

H

NOH

H

H

CH=CH2

021

7.187

H

NOAc

H

H

CH=CH2

H

022

6.187

H

O

H

H

F

H

023

6.886

H

NOH

H

H

F

H

024

7.046

H

NOAc

H

H

F

H

025

7.602

H

O

CH3

H

Br

H

026

8.222

H

NOH

CH3

H

Br

H

027

8.155

H

NOAc

CH3

H

Br

H

028

7.523

H

O

Cl

H

Cl

H

029

8.398

H

NOH

Cl

H

Cl

H

030

8.398

H

NOAc

Cl

H

Cl

H

61


031

7.000

H

O

NO2

H

Br

H

032

8.155

H

NOH

NO2

H

Br

H

033

8.222

H

NOH

NO2

H

Br

H

034

8.046

H

NOH

I

H

H

H

035

7.481

H

O

SO2-NH-C2H4-OH

H

H

H

036

7.398

H

O

SO2NH2

H

H

H

037

7.377

H

O

NO2

H

H

H

038

7.301

H

O

Cl

H

H

H

039

7.259

H

O

Br

H

H

H

040

7.208

H

O

CH3

H

H

H

041

7.167

H

O

I

H

H

H

042

7.108

H

O

F

H

H

H

043

7.097

H

NOH

SO3H

H

H

H

044

6.959

H

O

SO2-NHCH3

H

H

H

045

6.745

H

O

SO3-N(CH3)2

H

H

H

046

6.602

H

O

Br

Br

H

H

047

6.553

H

O

SO3H

H

H

H

048

6.456

H

O

H

Br

H

H

049

6.398

H

O

SO2-N(C2H4OH)2

H

H

H

050

5.398

H

O

SO3H

Br

H

H

051

3.699

Phenyl

O

H

H

H

H

RESULTS

AND

DISCUSSION

The Fujita Ban calculations were performed with the Statistica package [6] in order to find out the contribution of the substituents to the activity of the molecules investigated. The model was constructed using 24 variables for different substituents at positions 1, 3’, 5, 5’, 6 and 6’ of the indirubin skeleton and biological activities of the compounds. To determine how the substituents contribute to the biological activity, a multiple regression procedure was carried out. In the first step of Fujita Ban analysis, an initial model was constructed for the whole X matrix (N=51 compounds/rows and X=25 (24 variables and 1 biological activity/columns). In order to improve model M1 (the first model obtained) the casewise plot of outliers was studied. Finally model M4, was obtained gradually eliminating the outliers (on first step 001, 003, 037 were eliminated, in the second step 007 was removed and in the final step 005 was excluded). For model M4 excellent statistical parameters were obtained for the remaining set of 46 compounds: R2=0.982, F = 77.780 and see

62

PRELIMINARY FUJITA –BAN STUDY FOR GSK-3 INHIBITION BY INDIRUBIN DERIVATIVES

= 0.146 for a domain of approximately 3 log units (only 2 compounds are situated in the 5.4 - 3.7 activity range). The Fujita–Ban contribution of each substituent is presented in Table II. Only 16 of the 24 structural variables were found to be significant for 95% probability level (p=10.95 confidence level). Table II. Fujita–Ban contribution of substituents* R

Intercept

Beta

Std.Err. of Beta

B 7.751

R1 R3’

R5

R5’

R6

R6’

Std.Err of B 0.059

Effect

130.130

plevel 0.000

d

t(26)

*

Phenyl

-0.574

0.027

-3.276

0.157

-20.892

0.000

=O

-0.465

0.031

-0.775

0.052

-14.862

0.000

d d

- NOCH3

-0.191

0.028

-0.782

0.116

-6.734

0.000

- SO3H - SO2N(C2H4OH)2 -Cl - SO2-NH-C2H4OH - SO2NH2

-0.192

0.029

-0.647

0.101

-6.427

0.000

-Br -I

d d

-0.101

0.027

-0.577

0.157

-3.682

0.001

0.183

0.030

0.542

0.089

6.025

0.000

0.088

0.027

0.506

0.157

3.227

0.003

0.074

0.027

0.422

0.157

2.694

0.012

b

0.076

0.029

0.311

0.122

2.546

0.017

b

0.059

0.028

0.243

0.116

2.104

0.045

b

b b

– Br

-0.211

0.031

-0.711

0.104

-6.857

0.000

d

–F

-0.233

0.029

-0.786

0.099

-7.877

0.000

d

- CH=CH2

-0.143

0.029

-0.482

0.099

-4.838

0.000

d

-I

0.066

0.029

0.223

0.099

2.233

0.034

b

- Br

0.209

0.038

0.410

0.075

5.421

0.000

b

- Br

-0.309

0.028

-1.261

0.116

-10.854

0.000

d

d= detrimental effect of substituent; b= beneficial effect of substituent

The outcomes provided by the Model M4 suggest an evident beneficial effect of the substituents R6-Br, R5-Cl, R5-SO2-NH-C2H4-OH, R5-SO2NH2, R5-Br, R6-I, R5-I, i.e. are positively correlated with GSK-3 inhibitory activity, while the substituents R1- Phenyl, R3’-O, R5-SO3H, R6’-Br, R6-F, R5’-Br, R6-CH=CH2, R3’-NOCH3, R5-SO2-N(C2H4OH)2 are negatively correlated with GSK-3 inhibitory activity.

63


CONCLUSION A number of reliable statistical data have been obtained, particularly interpreting the regression coefficients and their signs obtained by the best Fujita-Ban QSAR equation applied to a series of indirubin-type GSK-3 inhibitors. Thus, we have identified different functional groups grafted to the indirubin scaffold showing beneficial effects (positive regression coefficients) and others displaying detrimental effects (negative regression coefficients). Our analysis provided key information regarding ligand–GSK-3 interactions which we believe that will help medicinal chemists to design GSK-3 inhibitors with favorable pharmacological properties, using bio-isosteric replacements for the beneficial substituents, and modifying the detrimental ones. The Fujita-Ban method usually is recommended in a preliminary phase of a project. Nevertheless the method gives valuable suggestions about the contribution of each substituent upon the biological activity. ACKNOWLEDGMENT

We are grateful to Dr. Simona Funar-Timofei for the access to STATISTICA 7.1 software. REFERENCES 1. 2. 3.

4.

5.

6.

64

Ali A., Hoeflich K. P., Woodgett J. R., “Glycogen Synthase Kinase-3: Properties, Functions, and Regulation”, Chem. Rev., 101 (2001) 2527-2540. Sivaprakasam P., Xie A., Doerksen R.J., “Probing the physicochemical and structural requirements for glycogen synthase kinase-3a inhibition: 2D-QSAR for 3-anilino-4phenylmaleimides”, Bioorgan. Med. Chem., 14 (2006) 8210–8218. Snyder G. L., Greengard P., Biernati J., Wui Y-Z., Mandelkowi E-M., Eisenbrand G., Meijer L., “Indirubins Inhibit Glycogen Synthase Kinase-3b and CDK5/P25, Two Protein Kinases Involved in Abnormal Tau Phosphorylation in Alzheimer’s Disease”, J. Biol. Chem., 276(1) (2001) 251– 260. Polychronopoulos P., Magiatis P., Skaltsounis A-L., Myrianthopoulos V., Mikros E., Tarricone A., Musacchio A., S. Roe S.M, Pear L., Leost M., Greengard P, Meijer L., “Structural Basis for the Synthesis of Indirubins as Potent and Selective Inhibitors of Glycogen Synthase Kinase-3 and Cyclin-Dependent Kinases”, J. Med. Chem., 47 (2004) 935-946. Varela A.T., Gomes A.P., Simões A.M., Teodoro J.S., Duarte F.V., Rolo A.P., Palmeira C.M., “Indirubin-3′ -oxime impairs mitochondrial oxidative phosphorylation and prevents mitochondrial permeability transition induction”, Toxicol. Appl. Pharm., 233 (2008) 179–185. STATISTICA 7.1, Tulsa, StatSoft Inc., OK, USA.