QSAR Studies of Novel 1- and 8-Substituted-3-Furfuryl Xanthines: An Adenosine Receptor Antagonist

 

Prarthana V Rewatkar¹ and Ganesh R Kokil^2*

¹Dept. of Pharma. Chemistry, Bharati Vidyappeth, Poona College of Pharmacy, Erandawane, Pune, India -411038

²Dept. of Pharma. Chemistry, Sinhgad Institute of Pharmaceutical Sciences, Kusgaon(Bk.), Lonavala, India-410401

*Corresponding Author E-mail: ganesh.kokil@gmail.com

ABSTRACT:

Adenosine modulates many important physiological functions, that affect the cardiovascular, renal, immune and central nervous systems. The receptors modulate the activity of adenylate cyclase either by stimulation (A_2A and A_2B) or inhibition (A1 and A3) of the activity. Quantitative structure activity relationship (QSAR) study for a series of 1- and 8- substituted xanthine derivatives as adenosine receptor (A_2A and A_2B) antagonist was performed. The QSAR models were developed using series of compounds against A_2A and A_2B antagonistic activity. The statistical quality of QSAR models was assessed by statistical parameters r², r²cv (cross validated r²) and r² pred (predictive r²).

 

 

 

INTRODUCTION:

Adenosine is an important metabolite and a building block for many biologically relevant molecules. Most abundantly, it contributes the purine base adenine and a ribose to ATP which as an energy-providing compound occurs in millimolar concentrations in every cell. Such high concentrations of ATP are the basis for functionally significant levels of adenosine to occur in all cells and in the extracellular space. Adenosine concentrations vary widely in tissues and body fluids as it is formed both intra- and extracellularly, the known diverse effects of adenosine on mitogenesis may be related to changes in mitogen-activated protein kinases. It undergoes metabolism by adenosine deaminase and adenosine kinase, and may be transported through the plasma membrane with equilibrative and concentrative types of transporter proteins.^{1- 3}

 

The endogenous nucleoside adenosine binds to and activates four different subtypes of G protein-coupled receptors (GPCR): adenosine A₁, A_2A, A_2B, and A₃ receptors. The different subtypes can be distinguished pharmacologically and differ also in their coupling to second messenger systems.^{4, 5}

 

Adenosine A1 and A3 receptors inhibit adenylyl cyclase and stimulate phospholipase Cβ via activation of the pertussis toxin-sensitive G proteins Gi and/or Go. Adenosine A_2A and A_2B receptors are positively coupled to adenylyl cyclase, but also may activate alternative signaling pathways.⁶

 

Adenosine administration by inhalation elicits concentration-related bronchoconstriction in subjects with asthma and chronic obstructive pulmonary disease (COPD). The mechanisms of adenosine-induced bronchoconstriction appear to involve a selective interaction with activated mast cells with subsequent release of preformed and newly-formed mediators. Further evidence linking adenosine signaling to asthma and COPD comes from the finding that many cell types that play important roles in the exacerbation of these conditions express adenosine receptors and demonstrate relevant effects through stimulation of these receptors.^7,8 Therefore, blockade of these receptors may be a valuable approach to the treatment of asthma and chronic obstructive pulmonary disease. Promising adenosine-receptor targets for novel therapeutics of asthma and chronic obstructive pulmonary disease have recently been identified in a number of inflammatory cell types, including mast cells, eosinophils⁹, lymphocytes¹⁰, neutrophils¹¹, and macrophages¹². The recent characterization of the A_2B receptors indicates the human lung mast cell as one of the most strategic cellular targets.¹³

 

Over two hundred 1-, 3-, 8-, and 9-substituted-9-deazaxanthines were prepared and evaluated for their binding affinity at the recombinant human adenosine receptors, in particular at the hA_2B and hA_2A subtypes. Several ligands endowed with sub-micromolar to low nanomolar binding affinity at hA_2B receptors, good selectivity over hA_2A and hA₃, but a relatively poor selectivity over hA₁ were obtained.^{14- 16}

 

Recently, Balo et al. et al. studied 1- and 8- substituted furylxanthine analouges as a new class of potent adenosine receptor antagonsist¹⁷. Urged by the need to develop novel potent adenosine receptor antagonsist, we applied the linear free energy related (LFER) approach of Hansch on furylxanthine analogues to rationalize the physicochemical properties before designing and developing new effective antagonist¹⁸. QSAR studies provide deeper insight into the mechanism of action of compounds that ultimately becomes of great importance in modification of the structure of compounds. In addition, QSAR also provides quantitative models, which permits prediction of activity of compounds prior to the synthesis.

 

The  main  objective  of present work was to provide  some  useful  information  by  QSAR  analysis  and  design  new  specific antagonist  of  adenosine receptor with the  hope  that  these  molecules  may  be  further  explored  as  powerful therapeutic agents.

 

COMPUTATIONAL METHODS:

Data sets and biological activity:

Dataset:

A Dataset of 27 and 25 molecules belonging to furylxanthine derivatives as adenosine receptor (hA_2A and hA_2B) antagonist were taken from the literature.  Different QSAR models were generated for this series. Various furylxanthine derivatives used in the present QSAR study and observed biological activities against hA_2A and hA_2B have been presented in Table 1 and 2.

 

Biological Activity:

The affinity (pKi) values of the 3- furfurylxanthine derivatives, at cloned human adenosine receptors expressed in HeLa cells (hA_2A) and HEK-293 cells (hA_2B) receptors was used as dependent variable¹⁹. Since some compound exhibited insignificant /no inhibition, such compounds were excluded from the present study.

 

Molecular Modeling

The three-dimensional structures of the 1- and 8- substituted furfurylxanthine derivatives were constructed using Chemdraw Ultra 8.0 running on an Intel Pentium IV 2.80 GHz Processor / Microsoft Win XP Home Edition platform. All molecules were built using Chemdraw module and subjected to energy minimization using molecular mechanics (MM2). The minimization is continued until the root mean square (RMS) gradient value reaches a value smaller than 0.001 kcal/mol Å. The lowest energy structure of the compounds in the series was used to calculate physicochemical properties using the ‘Analyze’ option of the Chem3D²⁰ and Dragon 5 evaluation version.

(Table 1 and 2)

 

The physicochemical properties calculated include thermodynamic, steric and electronic descriptors. Molar refractivity, Ellipsoidal Volume, logP ,Connolly accessible area (CAA), Connolly molecular area (CMA), Connolly solvent excluded volume (CSEV), molecular weight, principal moments of inertia–x component (PMI–X), principal moments of inertia–y component (PMI–Y), and principal moments of inertia–z component (PMI–Z), dipole moment (DM), highest occupied molecular orbital energy (HOMO), lowest unoccupied molecular orbital energy (LUMO) were used in the present study.

 

Multiple linear regression analysis²¹ method was used to generate QSAR models employing quickstat software. To check predictive power of the models, cross validation was done by leave one out procedure Following statistical parameters were considered to compare the generated QSAR models: correlation coefficient r, r ² , r ²cv, Standard deviation (S), F–test and internal predictive power by cross validated coefficient (r² cv)²².

 

RESULTS AND DISCUSSION:

The correlation between calculated descriptors as independent variable and the affinity (pKi) values of the furfurylxanthine derivatives, at cloned human adenosine receptors expressed in HeLa cells (hA_2A)  and HEK-293 cells (hA_2B)  receptor as response variable was calculated using multiple linear regression analysis. Only those parameters having Inter-correlation below 0.5 were considered to select the best model. The best model obtained is given below along with its statistical measures.

 

For hA_2A receptor

Y = -0.028879778*X1 + 0.00017821023*X2 + 0.00029372287*X3 + 0.13235475

r = 0.754; r² = 0.569; r²_cv = 0.989; S= 0.00536; F-Value = 10.119

X1: Balaban Topological index, X2: Surface Area, X3: Quadpole YY

 

The model exhibits good internal predictivity as established by the cross validation r² value (0.989) of the model and also good external predictivity indicated by r² (0.569). The absence of any serious multi-collinearity between the descriptors present in the model was confirmed by the calculation of correlation matrix which shows that the descriptors balaban-topological index, surface area, quadpole YY not inter-correlated which is shown in Table 3.

 

The descriptors in the best model indicate effect of, balaban-topological index, surface area and quadpole YY on biological activity. Surface area and quadpole YY are positively correlated with biological activity means more the surface area and quadpole YY more will be the activity. While balaban-topological index is negatively correlated with biological activity means less will be the value more potent will be the compound.

 

 

Table 1: Chemical structures and binding affinities at hA_2A of 1,8-disubstituted 3-furfurylxanthine derivatives

 

 

Compound	R1	R2	R3	1/ ahA2A	Balaban-topological index	surface area	Quadpole YY
9d	Ethyl	Thiophen-2-yl	-H	0.157978	1.59376	364.13	-8.376
9e	Propyl	Phenyl	-H	0.153374	1.69577	406.083	-9.755
9f	Propyl	Furan-2-yl	-H	0.15674	1.63706	380.646	-9.242
9g	Propyl	Thiophen-2-yl	-H	0.15456	1.63706	385.79	-7.178
9h	Isobutyl	Phenyl	-H	0.158479	1.74908	431.823	-9.652
9j	Pentyl	Phenyl	-H	0.165017	1.75326	449.386	13.698
9l	Cyclopropylmethyl	Thiophen-2-yl	-H	0.154799	1.43516	394.038	-9.687
9n	Prop-2-ynyl	Thiophen-2-yl	-H	0.15015	1.56296	369.728	-9.824
9o	Prop-2-ynyl	2,6-Difluorophenyl	-H	0.157233	1.62693	401.785	-3.556
9w	2-(Methylthio)ethyl	Thiophen-2-yl	-H	0.146628	1.62875	407.117	-25.9
9x	2-(Methylthio)ethyl	2,6-Difluorophenyl	-H	0.149254	1.68922	439.386	-19.666
9y	2-(Ethylthio)ethyl	Phenyl	-H	0.15625	1.70675	453.622	-27.679
10a	Methyl	Furan-2-yl	-CH3	0.146413	1.63546	354.684	-11.467
10b	Ethyl	Phenyl	-CH3	0.141844	1.75478	405.071	-9.462
10c	Ethyl	Furan-2-yl	-CH3	0.143266	1.68848	376.603	-12.396
10d	Ethyl	Thiophen-2-yl	-CH3	0.141243	1.68848	384.549	-7.709
10e	Propyl	Phenyl	-CH3	0.158479	1.79333	426.626	-8.473
10f	Propyl	Furan-2-yl	-CH3	0.15361	1.72916	397.621	5.317
10g	Propyl	Thiophen-2-yl	-CH3	0.151057	1.72916	405.568	1.217
10h	Cyclopropylmethyl	Thiophen-2-yl	-CH3	0.158479	1.5098	414.066	0.684
10i	Cyclopropylmethyl	2,6-Difluorophenyl	-CH3	0.174216	1.55849	446.131	2.533
10j	Allyl	Thiophen-2-yl	-CH3	0.15361	1.69032	397.006	-6.858
10k	Allyl	2,6-Difluorophenyl	-CH3	0.155763	1.7574	428.931	4.53
10l	2-Methoxyethyl	Phenyl	-CH3	0.165563	1.786	436.849	-15.943
10m	2-Methoxyethyl	Thiophen-2-yl	-CH3	0.162338	1.72026	416.036	6.647
10n	2-Methoxyethyl	2,6-Difluorophenyl	-CH3	0.158479	1.786	446.635	0.863
10o	2-(Ethylthio)ethyl	Thiophen-2-yl	-CH3	0.165289	1.7354	448.198	12.913

^aBinding affinity is expressed as pKi

 

Table 2: Chemical structures and binding affinities at hA_2B of 1, 8-disubstituted 3-furfuryl-7-methylxanthine derivatives

 

^a Binding affinity is expressed as pKi

 

Compound	R1	R2	R3	1/ahA2B	Total Dipole Moment	Total Lipole Moment	Heat of Formation
9a	Methyl	Furan-2-yl	-H	0.156986	6.12741	7.76954	37.874
9b	Ethyl	Phenyl	-H	0.167224	5.19158	8.96323	30.5634
9c	Ethyl	Furan-2-yl	-H	0.159236	5.3536	6.55883	15.5437
9d	Ethyl	Thiophen-2-yl	-H	0.15528	6.12731	6.34408	31.0974
9e	Propyl	Phenyl	-H	0.15674	5.31705	4.95206	8.76891
9n	Prop-2-ynyl	Thiophen-2-yl	-H	0.15748	6.10836	5.3194	63.2157
9o	Prop-2-ynyl	2,6 Difluorophenyl	-H	0.157729	5.81032	2.43134	-29.2684
9p	Allyl	Thiophen-2-yl	-H	0.171233	4.95299	11.1989	-2.61584
9w	2-(Methylthio)ethyl	Thiophen-2-yl	-H	0.176056	5.38919	7.60827	21.2922
9x	2-(Methylthio)ethyl	2,6-Difluorophenyl	-H	0.169205	6.13914	6.8833	66.6557
10a	Methyl	Furan-2-yl	-CH3	0.133511	5.04635	8.18953	-94.7516
10b	Ethyl	Phenyl	-CH3	0.126743	5.19165	7.14926	23.786
10c	Ethyl	Furan-2-yl	-CH3	0.127226	5.75003	3.79773	-25.8338
10d	Ethyl	Thiophen-2-yl	-CH3	0.123001	4.10161	12.513	-9.94459
10e	Propyl	Phenyl	-CH3	0.13587	6.87934	8.43894	33.3461
10f	Propyl	Furan-2-yl	-CH3	0.129534	8.95496	8.15549	32.3369
10g	Propyl	Thiophen-2-yl	-CH3	0.127714	8.0723	7.17697	19.1787
10h	Cyclopropylmethyl	Thiophen-2-yl	-CH3	0.134228	8.2368	7.71593	14.1307
10i	Cyclopropylmethyl	2,6-Difluorophenyl	-CH3	0.136426	9.08843	7.3388	96.8605
10j	Allyl	Thiophen-2-yl	-CH3	0.130378	8.67671	5.06462	3.74705
10k	Allyl	2,6-Difluorophenyl	-CH3	0.130378	8.21766	6.95182	8.37172
10l	2-Methoxyethyl	Phenyl	-CH3	0.138696	8.09294	7.65467	26.2661
10m	2-Methoxyethyl	Thiophen-2-yl	-CH3	0.132275	9.48801	6.28916	-59.0115
10n	2-Methoxyethyl	2,6-Difluorophenyl	-CH3	0.134953	9.07738	5.44168	57.6789
10o	2-(Ethylthio)ethyl	Thiophen-2-yl	-CH3	0.141844	9.57254	9.88142	34.0837

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 3: Correlation matrix of the descriptors used in developing QSAR model

	Balaban Topological index	Surface Area	Quadpole YY
Balaban Topological index	1	0.44668	0.14813
Surface Area	0.44668	1	0.20487
Quadpole YY	0.14813	0.20487	1

 

For hA_2B receptor

Y = 0.0092859147*X1 + 0.0015690546*X2 - 7.2047224e-005*X3 + 0.07077425

r = 0.930; r² = 0.866; r²_cv = 0.918; S= 0.00642; F-Value = 45.0781

X1: Total Dipole Moment, X2: Total Lipole, X3: Heat of Formation

 

The model also exhibits good internal predictivity as established by the cross validation r² value (0.918) of the model and also good external predictivity indicated by r² (0.866). The absence of any serious multi-collinearity between the descriptors present in the model was confirmed by the calculation of correlation matrix which shows that the descriptors Total Dipole Moment, Total Lipole, Heat of Formation not inter-correlated which is shown in Table 4.

 

Table 4: Correlation matrix of the descriptors used in developing QSAR model

	Total Dipole Moment	Total Lipole	Heat of Formation
Total Dipole Moment	1	-0.13724	0.25128
Total Lipole	-0.13724	1	0.047827
Heat of Formation	0.25128	0.047827	1

 

The descriptors in the best model indicate effect of, total dipole moment, total lipole and heat of formation on biological activity. Total dipole moment and total lipole are positively correlated with biological activity means more the total dipole moment and total lipole, more will be the activity. While other heat of formation is negatively correlated with biological activity means less will be the value more potent will be the compound.

 

CONCLUSIONS:

We have developed predictive QSAR models for furfurylxanthine derivatives having adenosine receptor (hA_2A and hA_2B) antagonist activity. The results obtained for the present series of derivatives showed good correlation with adenosine receptor (hA_2A and hA_2B) antagonist activity of furfurylxanthine derivatives. The prediction power of the QSAR model was tested by LOO method which gives a good internal predictivity. The results of the QSAR study indicate characteristic influence  and these result will help in providing structural insights for designing  new  specific furfurylxanthine inhibitors  of  adenosine receptor (hA_2A and hA_2B) with  the  hope  that  these  molecules  may  be  further  explored.

 

REFERENCES:

1.       Arch JRS, Newsholme EA. The control of the metabolism and the hormonal role of adenosine. Essays Biochem. 1978; 14: 82–123.

2.       Ali S, Mustafa SJ, Metzger WJ. Adenosine receptor-mediated bronchoconstriction and bronchial hyperresponsiveness in allergic rabbit model. Am J Physiol. 1994; 266: L271–L277.

3.       Bucchioni E, Csoma Z, Allegra L, Chung KF, Barnes PJ, Kharitonov SA. Adenosine 5-monophosphate increases levels of leukotrienes in breath condensate in asthma. Respir Med.  2004; 98: 651–655.

4.       Fredholm BB, Abbracchio MP, Burnstock G, Daly JW, Harden TK, Jacobson KA, Leff P and Williams M. Nomenclature and classification of purinoceptors. Pharmacol Rev. 1994; 46:143–156.

5.       Fredholm BB, Arslan G, Kull B, Kontny E, Svenningsson P. Adenosine (P1) receptor signalling. Drug Dev Res.1996; 39: 262–268.

6.       Feoktistov I, Murray JJ, Biaggioni I. Positive modulation of intracellular Ca21 levels by adenosine A2b receptors, prostacyclin, and prostaglandin E1 via a cholera toxin-sensitive mechanism in human erythroleukemia cells. Mol Pharmacol. 1994; 45: 1160–1167.

7.       Polosa R, Holgate ST. Adenosine bronchoprovocation: a promising marker of allergic inflammation in asthma? Thorax.  1997;  52: 919–923.

8.       Fozard JR, Hannon JP. Adenosine receptor ligand: Potential as therapeutic agents in asthma and COPD. Pulm Pharmacol Ther. 1999; 12: 111–114.

9.       Walker BA, Jacobson MA, Knight DA. Adenosine A3 receptor expression and function in eosinophils. Am J Respir Cell Mol Biol. 1997; 16: 531– 537.

10.     Huang S, Apasov S, Koshiba M, Sitkovsky M. Role of A2a extracellular adenosine receptor-mediated signaling in adenosine-mediated inhibition of T-cell activation and expansion. Blood.  1997; 90: 1600– 1610.

11.     Cronstein BN. Adenosine regulation of neutrophil function and inhibition of inflammation via adenosine receptors. In: Jacobson KA, Jarvis MF, eds. Purinergic Approaches in Experimental Therapeutics. New York, Wiley-Liss, 1997; pp. 285–293.

12.     Eppell BA, Newell AM, Brown EJ. Adenosine receptors are expressed during differentiation of monocytes to macrophages in vitro: implications for regulation of phagocytosis. J Immunol. 1989; 143: 4141–4145

13.     Olah ME, Stiles GL. Adenosine receptor subtypes: characterization and therapeutic regulation. Annu Rev Pharmacol Toxicol. 1995; 35: 581.

14.     Francis JE, Cash WD, Psychoyos S, Ghai G, Wenk P, Friedmann RC, Atkins C, Warren V, Furness P, Hyun JL, Stone GA, Desai M, Williams M. J Med Chem. 1988; 31: 1014.

15.     Moro S, Gao ZG, Jacobson KA, Spalluto G. Progress in the pursuit of therapeutic adenosine receptor antagonists. Med Res Rev. 2006; 26: 131.

16.     Akkari R, Burbiel JC, Hockemeyer J, Müller CE. Curr Top Med Chem. 2006; 6: 1375.

17.     Balo MC, Brea J, Caamano O, Fernandez F. Synthesis and pharmacological evaluation of novel 1- and 8-substituted- 3-furfuryl xanthines as adenosine receptor antagonists. Bioorg Med Chem. 2009; 17: 6755–6760

18.     Hansch C, Sammes P G and Taylor J B, in: Comprehensive Medicinal Chemistry, 4^th Ed., Pergamon Press, Great Britain, 1990, 497.

19.     Karki RG, Kulkarni VM. Three-dimensional quantitative structure-activity relationship (3D-QSAR) of 3-aryloxazolidin-2-one antibacterials. Bioorg Med Chem. 2001; 9: 3153.

20.     Choudhary G, Karthikeyan C, Moorthy N S, Sharma S K , Trivedi P. QSAR Analysis of Some Cytotoxic Thiadiazino-acridines.Inter Elec J Mol Design. 2005; 4: 793.

21.     Kulkarni SS, Gediya LK, Kulkarni VM. Three-dimensional quantitative structure activity relationships (3-D-QSAR) of antihyperglycemic agents. Bioorg & Med Chem. 1999; 7: 1475.

22.     Wagh NK, Deokar HS, Juvale DC, Kadam SS, Kulkarni VM. QSAR analysis of N-myristoyltransferase inhibitors: antifungal activity of benzofurans. Indian J. Biochem. Biophys. 2006; 43: 360.

 

 

Received on 11.02.2010        Modified on 09.03.2010

Accepted on 30.03.2010        © AJRC All right reserved