QSAR studies of benzofuran/benzothiophene biphenyl derivatives as inhibitors of PTPase-1B

Abstract

Objectives:

Insulin resistance is associated with a defect in protein tyrosine phosphorylation in the insulin signal transduction cascade. The PTPase enzyme dephosphorylates the active form of the insulin receptor and thus attenuates its tyrosine kinase activity, therefore, the need for a potent PTPase inhibitor exists, with the intention of which the QSAR was performed.

Materials and Methods:

Quantitative structure-activity relationship (QSAR) has been established on a series of 106 compounds considering 27 variables, for novel biphenyl analogs, using the SYSTAT (Version 7.0) software, for their protein tyrosine phosphatase (PTPase-1B) inhibitor activity, in order to understand the essential structural requirement for binding with the receptor.

Results:

Among several regression models, one per series was selected on the basis of a high correlation coefficient (r, 0.86), least standard deviation (s, 0.234), and a high value of significance for the maximum number of subjects (n, 101).

Conclusions:

The influence of the different physicochemical parameters of the substituents in various positions has been discussed by generating the best QSAR model using multiple regression analysis, and the information thus obtained from the present study can be used to design and predict more potent molecules as PTPase-1B inhibitors, prior to their synthesis.

Diabetes has been recognized as a genetic disorder, in which the glucose metabolism is altered. The ability of insulin to bring about such a dramatic reversal in the symptoms of diabetes, with a return to a near normal life expectancy led the medical community to conclude that the problems of etiology and treatment had been resolved, but these conclusions were premature. While insulin did return control of the blood glucose level and did offset the development of ketoacidosis, it did not appear to rectify all the metabolic defects.[1]

It is thus evident that insulin therapy, as currently practiced, is not the panacea for diabetes mellitus. This realization has promoted a great deal of research toward the development of more effective ways of treating the disease, which has led to the discovery of various hypoglycemic agents, for example, Sulfonylurea, Biguanide, and more recently, glitazones.

It is now well established that insulin resistance can result from a defect in the insulin receptor signaling system at a site post binding of insulin to its receptor.[2] Insulin resistance is associated with a defect in protein tyrosine phosphorylation in the insulin signal transduction cascade.[3] This defect in tyrosine phosphorylation in the insulin-responsive muscle, tissue, liver, and adipose, causes a reduction in the metabolic actions of insulin and hyperglycemia. This decrease in protein tyrosine phosphorylation in the cells does not appear to be due to an inherent problem in the insulin receptor (IR) tyrosine kinase, but instead is caused by an elevation in the protein phosphate activity.

The PTPase enzyme dephosphorylates the active form of the insulin receptor and thus attenuates its tyrosine kinase activity.[4]

Various research articles published on compounds possessing the PTPase inhibition activity, such as, phosphotyrosine bioisostere,[5] phosphonic acid derivative,[6] vanadium compounds,[7] benzofuran / benzothiophene derivative,[8] 11-aryl benzo[b] naptho (2,3d) furan and 11-arylbenzo[b]naptho (2,3d) thiophene,[9] and azolidinedione.[10] From overviewing all these various classes of compounds, we can suggest the general structure of PTPase 1B inhibitors as given in Figure 1.

Figure 1.

Chemical data of substituted biphenyls

Insulin resistance is thus one of the obstacles that we confront while undergoing therapy for diabetes mellitus. A number of PTPase inhibitors have been designed and studied to overcome this problem, to gain an insight into the structural and molecular requirements influencing the PTPase-1B inhibition activity; we herein describe the QSAR analysis of a set of structurally different compounds of PTPase inhibitors, for which it is conceivable to make an assumption that they interact with the enzyme.

Materials and Methods

In silico studies

Malamas et al.[8] reported seven series of compounds based on benzofuran/benzothiophene biphenyl moiety. We had performed the QSAR analysis of all these series having 138 compounds, out of which only 106 compounds could be subjected to 2-D QSAR analysis, because of the non-availability of physicochemical substituent values and exact IC₅₀ values for some substituted compounds. The 2D QSAR study was carried out in the following steps:

Calculations of physicochemical constants

The values for the physicochemical constants for various substituents were determined from the literature.[11] The determined parameters for a series included, the Hansch constant (π), Molar Refractivity (η), Sigma / Hammet constant (σ), Field effect (F), and the Indicator value (I).

To get the generated model we had clubbed all the series together for the sake of simplicity. We had designated different rings and positions as shown in the chemical structures, as (U, V, W, X) and (a, b, c, d, e, f), respectively. To simplify and make all the series collinear to each other the use of the indicator variable had been included. Thus, the compounds of the seven series were designated as follows:

Series I: As to x of V ring [Figure 2], we had assigned this position as [a], which was either oxygen or sulfur, so we had considered O=1 and S=0 as the indicator variable, V_aI.

Figure 2.

Structure of benzofuran and benzothiophene biphenyls

R¹ substitution was assigned as [b], so different physicochemical parameters were designated on the basis of ring and position, as πV_b, σV_b, and ηV_b.

R² substitution position was [c], so the presence of an aceto moiety had been considered as an indicator variable with value 1 and the others as 0, and hence this position was considered as X_cI (where X-ring, c-position, I- indicator). The parameters for the substituents on this aceto moiety had been designated as πX_cI, σX_cI, ηX_cI, FX_cI. Due to the substitution on the aceto moiety, there occurred the presence of a chiral center, due to which most of the compounds were enantiomers. Therefore, an indicator variable X_cEI, (R=1, S=1, dl=0) was included, where the X-ring, c-position, E-enantiomer, I-indicator variable, and the value of X_cEI= 1, -1, 0 depended on the optical rotation.

Series II: Similar to the first series, V_aI was considered as an indicator variable [Figure 3].

Figure 3.

Structure of 2-benzyl benzofuran and benzothiophene Biphenyls

R¹ position of the second series was considered as [d], so different parameters for the R² position had been designated as πX_d, σX_d, ηX_d, FX_d.

R² position was [e], so parameters of R² position were πX_e, σX_e, ηX_e, FX_e.

R³ position was designated as X _cI , X_cEI, πX_cI, σX_cI, ηX_cI, FX_cI.

Series III: Benzofuran attached to biphenyl through the X linkage [Figure 4], which was designated as [f], so the corresponding parameters were πf, σf, ηf, Ff.

Figure 4.

Structure of 2-butyl benzofuran biphenyls

Similarly the R² position was [d] and the parameters were πX_d, σX_d, ηX_d, FX_d

Position [e] was πX_e, σX_e, ηX_e, FX_e

Position [c] was X_cI.

Series IV: R¹ position [Figure 5] was [c], so the earlier assigned R¹ position was an indicator variable, as X_cI,, and its substitutions were as πX_cI, σX_cI, ηX_cI, FX_cI

Figure 5.

Structure of Substituted oxazole Biphenyls

R² and R³ position parameters were πX_d, σX_d, ηX_d, FX_d and πX_e, σX_e, ηX_e, FX_erespectively.

As the isoxazole ring attached to either at third or fourth position, the indicator variable position was assigned as VW_pI (position of attachment of V and W rings as indicator parameters at the fourth position as 1, and at the third position as 0).

Series V: R¹ position [Figure 6] was earlier designated as X_cI, πX_cI, σX_cI, ηX_cI, FX_cI

Figure 6.

Structure of 2-Butyl Benzofuran Naphthalenes

R² position was [e] here and thus parameter R² position was termed as πX_e, σX_e, ηX_e, FX_e

The X group linkage between benzofuran and the naphthalene ring position was designated as [f], so the parameter of the X group became πf, σf, ηf, Ff.

Series VI: As R¹, on the phenyl sulfono ring [Figure 7] was attached to position [c] of ring X, the R¹ position was assigned to X_c3R¹ (X-ring, c-position, 3-third position of the phenyl ring, and R as the indicator).

Figure 7.

Structure of 2-Benzyl Benzofuran / Benzothiophne Sulfono Biphenyls

Wherever R¹ was -OH, X_c3R¹ =1, For -COOH, X_c3R¹ =0

Similarly R² was X_c4R²

Wherever R² was COOH, there X_c4R² =1,

Wherever R² was OH, there X_c4R² =0

R³ was [d] position and thus the parameter of R³ position was designated as πX_d, σX_d, ηX_d, FX_d

R⁴ was [e] position and thus the parameters were πX_e, σX_e, ηX_e, FX_e .

As x was in a V ring, which was assigned position [a], the indicator variable V_aI

was considered.

V_aI =1, when X=O

=0, when X=S

Series VII: Similar to the sixth series except that R⁵ was taken as V_aI [Figure 8] wherever the benzofuran ring comes

Figure 8.

Structure of Sulfono Biphenyls

V_aI =1, when R⁵ was benzofuran

V_aI =0, when R⁵ was either Benzothiophene or other.

However, when it was considered to club all 26 variables of the series, in the fifth series the biphenyl ring was replaced by a naphthalene ring, and the presence of the biphenyl ring was considered as variable I_wx

I_wx =1, where the W, X rings are present as biphenyl

=0, where the W, X rings are present as naphthalene

Thus for all 106 compounds, the value of the substituent constant for all 27 variables was obtained from literature.

Determination of the correlation matrix

The correlation matrix for all the 27 variables along with the biological activity was determined using the program ‘SYSTAT’ (version 7.0).[12] The most significant parameters of the PTPase inhibiting activity were chosen on the basis of their correlation ship and Interco relationship.

Multiple regression analysis

It was performed by using the program ‘SYSTAT’ for the PTPase inhibiting activity, that is, -Log_IC50 was considered as a dependent variable and all the 27 physicochemical parameters were considered as independent variables.

Results and Discussion

The correlation matrix [Table 1] reveals the dependency of various parameters on the biological activity and among themselves. Those which have a high value of correlation ship for biological activity and the least Inter-corelationship between themselves are considered for multiple regression analysis. The multiple regression analysis is generated following the QSAR model:

Table 1.

Correlation matrix between physicochemical parameter and biological activity

-Log_IC50 =0.020(± 0.004) ηV_b + 0.297 (± 0.041) πX_d - 0.057(± 0.18) ηf

+ 0.546(± 0.113) X_c3R¹ + 0.069(± 0.105)

n=101, r=0.86, s=0.234, F=42.697

The derived QSAR equation showed a very good r (correlation coefficient) value of 0.86 with 76.93% of the variation in biological activity being explained by the equation. This was associated with a low value of standard error of estimate, s, of 0.23. The equation was found to be highly statistically significant, with an F-test value of 42.697 and the critical F-test value at 99.95% confidence limit being 4.82.

Five compounds were considered as outliers because of high residual and leverage values. Perusal of the above-mentioned equation indicated that molar refractivity at the V_bposition and the hydrophobic substituent at the X_d position contributed positively to the biological activity, while any substitution between biphenyl and benzofuran / benzothiophene, that is, at the f position had a detrimental effect on the biological activity of the compounds. The model also suggested that substitution on the aceto moiety also had a positive contribution for biological activity, as suggested by the QSAR model. These findings were in agreement with the pharmacophore model, as reported earlier, which suggested that the benzofuran / benzothiophene ring should be attached directly to biphenyl, which should have some phenyl moiety attached to it.

A different set of descriptors may give a different correlation with activity and this must be considered while interpreting the equation. The QSAR equation generated presently for Benzofuran Biphenyl is in agreement with the literature reports for ligand requirements based on the Structure-Activity Relationships (SAR). Furthermore, this equation provides a mathematical tool for designing compounds with better activity. Thus from the QSAR model, appropriate variations of the substituent at various positions can be effected to obtain an effective molecule, which may be remarkable to experimentally confirm the predictive power of the QSAR model by actual synthesis and its biological evaluation of the most potent theoretical compound.

Conclusion

The regression model shows that the molar refractivity and hydrophobic substituent at the V_b and X_d positions, respectively, are the significant descriptors responsible for describing the downregulation of PTPase 1B. Besides this, any substitution between biphenyl and benzofuran / benzothiophene has a detrimental effect on the biological activity, which is also an important parameter to be taken into consideration while designing new inhibitors belonging to the above said class of compounds. The QSAR model is statistically and chemically sound and explains more than 95% variance in experimental activity. Finally, it can be concluded that the study presented here will play an important role in understanding the relationship of the physicochemical parameters with the structure and biological activity of the PTPase 1B inhibitor and will help in choosing a suitable substituent for obtaining the active compound with maximum potency.