The Study of Topological Modeling on Lipoxygenase Inhibitor by QSAR

Sherendra K. Sahu¹, R. R. Dwivedi¹, Skand K. Mishra¹, Neeta Singh², Sunita Gupta³

¹Dept. of Biotechnology & Botany, Govt. New Science College, Rewa (M.P.)

²Dept. of Botany, Govt Girls P.G. College , Rewa (M.P.)

³Dept. of Chemistry, APSU, Rewa (M.P.)

*Corresponding Author E-mail: sherendra@gmail.com

ABSTRACT:

A Lipoxygenase inhibitor is a drug which slows down or stops the action of the lipoxygenase enzyme. More precisely, the term is almost always used to describe an inhibitor of the arachidonate 5-lipoxygenase enzyme, which transforms EFAs into leukotrienes. Examples include-Azelastine diethylcarbamazine ,nordihydroguaiaretic acid ,zileuton, In this study we have investigated that the oxidation of various substrates (linoleic acid, methyl linoleate, phosphatidylcholine, isolated LDL, and human plasma) by the arachidonate 15-lipoxygenases from rabbit reticulocytes and soybeans aiming at elucidating the effects of substrate, lipoxygenase and reaction milieu on the contribution and mechanism of random oxidation and also the effect of antioxidant.The complete descriptors data set of all compounds were considered as independent variable and biological activity as dependent variable www. ncss. com software was used to generate QSAR models by step wise multiple linear regression analysis statistical measures used were n–number of compounds in regression, in correlation coefficients F-test (Fischer's value for statistical significance). SE–standard error of estimations and correlation matrix to show correlation. This is clearly predict that our proposed tetra parametric model in most appropriate model for modeling inhibition activity log1/IC₅₀ , for lipoxygenase inhibitor set of 41 compounds.Finally I obtained my results, that is suggesting that proposed combination QSAR models could be useful in predicting the Lypoxygenase inhibiting activity of Arachidonate 5-Lypoxygenase enzyme.

KEYWORDS: QSAR, Topological Designing, Lipoxygenase inhibitor

INTRODUCTION:

The biosynthetic Cascade of Arachidonic acid has been the object of artiness research. Arachidonic acid liberated from phospholipids by various. Stonily can be metabolized by the cyclooxygenase (COX) pathway to prostaglandins (PGs) and thromboxane A₂ or by lipoxygenase. (LOX) pathways to hydroperoxyicosatetraenoic acids. (HPETE), hydroxyeicosatetranoic acids (HETE) and leukotrienes (LTS). Lipoxygenases (LOXs) are a family of cytosolic enzymes widely distributed in nature.

They are monomeric protoins that contain a "non-hene" iron per molecule in the active site as high-spin Fe(II) in the native state, and high spin Fe(III) in the activated state^2,4,9. Arachidonic acid, their main. substrate in mammals, can be cleared from phospholipids leading to the formation of lysophospholipids¹¹. Lipoxygenases as dioxygenases recognize the 1, 4 pentadiene structure of polyunsaturated fatty acids and catalyze their oxygenation to corresponding lipid hydroperoxides. They differ in their specificity for placing the hydroperoxy group so 5-lipoxygenase (5-LOX) inserts oxygen on position 5 of arachidonic acid, 12-lipoxygenase (12-LOX) on position 12 and 15-lipoxygenase (15-LOX) on position 15^10,14.

An overproduction of these products can cause disturbances in the metabolic reactions and are involved in some metabolic diseases and pathologies. These effects have been linked to immunological and radiation disorders, tumors, toxicoses hypodinamy coronary and angiological pathologies (vasospasm, thrombosis arteriosclerosis) .

The major products of 5-LOX, laukotrienes (LT's), are a family of important biologically active molecules.LTB₄ is patent chemotactic agent and inflammatory mediator¹ and the peptidoleukotrienes LTC₄ and LTD₄ are powerful spasmogens in vascular and bronchial tissues⁷. Elevated levels of LT_S are associated with a number of inflammatory conditions including asthma, psoriasis, ulcerative colitis, and rheumatoid arthritis and indeed LTs, have been recovered from variations pathological issues. Therefore, patent in linters of this enzyme are candidate drugs for the treatment of these disease^5,8. These inhibitors can be broadly classified into two main categories, first competitive lipid substrate inhibitors and second. redox–type inhibitors, which act by chelation or reduction of the Fe (III) of the active enzyme or by reaction with the fatty acids radical intermediate produced during the catalytic step¹².

Many thousands of compounds have been screened as LOX irihibitors uri industrial laboratores and a large number of active componds with nonel structure are undergoing climical trials. This evaluation provides data sets suitable for qualitative structure activity relationship (QSAR). The laboratory tests utilized in identifying Lipo oxygenase inhibitors are human. Granulocytes, rat basophilic leukemia cells (RBL-1) and human whole blood assay (HWBL).

Finally we have estimated model and compared with observed log 1/IC₅₀ and recorded the value of residue i.e. difference between observed and calculated of both log 1/IC₅₀ activity. Such results are presented in table – IV-6.

MATERIAL AND METHOD:

All the calculated descriptors (7 descriptors calculated by manually. The complete descriptors data set of all compounds was considered as independent variable and biological activity as dependent variable www.ncss.com software was used to generate QSAR models by step wise multiple linear regression analysis statistical measures used were n–number of compounds in regression, in correlation coefficients F-test (Fischer's value for statistical significance). SE–standard error and correlation matrix show correlation among the parameters.

The data presented in Table – IV-3 show that: -

W excellent correlated with P₂ and P₃ such correlation found to be linear positive and W it is also give excellent linear correlation with Id and IP₃ but such correlations are found to be linearly negative. That is Id and IP₃ decreased with increase in the magnitude of Winer index.

RESULT AND DISCUSSION:

These results indicate that multi parametric model will be suitable for modeling inhabitation activity log 1/IC₅₀.

It is interesting to record that here also looking to the sample size (i.e. 41 compounds) only mono bi, tri, to tetra parametric regressions were attempted. These regressions are shown in Table-IV-4. Based on R- values we observed that mono parametric model based on P₂ is the most appropriate among all mono parametric models. The regression models based on the c_eq is found under.

log 1/IC₅₀ = -0.0379 (±0.1103) c_eq +8.256

n = 41, R= 0.4825, R² = 0.2328, R²A = 0.2132

SE = 0.8104, F = 11.8357, Q= 0.5953 Q^{2 =}0.3543

Model No. (IV-5-1)

In above and subsequent correlation equations is the number of data point F is the F-ratio-between the variances of calculated and observed activities.

The above equation is not statistically significant one because R² value is too low and SE value is to high. Step wise regression the bi-parametric regression containing Id and IP₈ are correlating parameters is founds to good. Now use have considered the best equation containing two parameters is model:

log 1/IC₅₀ = 0.2817 (±0.007887) Id + 1.130 (±0.4068) IP₈ + 7.850

n = 41, R = 0.5965, R²A = 0.3219, R² = 0.3558

8E = 0.7003, F = 10.4962, Q =0.8518 Q² = 0.7256 Model (IV-5-2)

This equation is also statistically not significant one because R₂ value is too low and SE value is too high.

Further for obtaining excellent results under step wise regression we obtain same tri tetra to penta parametric regressions.

Out of these regression containing Vw, Id, IP₆ and IP₈ gave statistically good model this model is found as below.-

log 1/IC₅₀ = 0.3202 (I0.2409) Vw + 0.03474 (±0.01049)

+ 0.9435 (± 0.5769) IP₆ + 0.7797 ( ±0.4314) IP₈

n= 41, R = 0.6510, R²= 0.4238, R²A = 0.598

SE = 0.7021 + = 6.6207 Q=0.9272 Q²=0.8596 Model = (IV-5-3)

The above equation is statistically significant best model on the basis of R value and R² value.

It is worth recording that quality of regression expression can't judged on the basis of R-value in addition to R value, one has to consider the effect due to standard error of estimation also, so in order to find out relative correlation potential of the proposed models for modeling log 1/IC₅₀ . We have calculated quality factor.

The Q value as presented in table-IV-4 show that is highest for the tri parametric model based on the combination of Vw, Id, and IP₆. As compared to the Q value for tetra parametric model based on Vw, Id, IP₆ and IP₈. This Show that the quality of tetra parametric model is likely increase by the addition of IP₆ index.

	Log1/ic50	W	P₂	P₃	P2-P3	_eq	Vw	Id
Log1/ic50	1
W	-0.4668	1
P2	-0.4825	0.9006	1
P3	-0.4527	0.8821	0.9829	1
P2-P3	0.3155	-0.637	-0.7498	-0.8056	1
eq	-0.3263	0.3201	0.2561	0.2903	-0.2273	1
V.w.	-0.1116	0.5109	0.4513	0.4252	-0.3004	-0.225	1
Id	0.4743	-0.847	-0.804	-0.7541	0.4753	-0.1644	0.6541	1
Ip1	-0.0162	0.2075	0.3018	0.2416	-0.1369	-0.3236	0.102	-0.1477
Ip2	-0.111	-0.048	0.2346	0.2632	-0.2468	0.139	-0.1767	-0.0489
Ip3	0.0332	-0.895	-0.0337	0.015	-0.0476	0.1344	-0.1244	0.139
Ip4	0.1737	0.2906	0.181	0.22	-0.2193	0.1945	0.0465	-0.1752
Ip5	0.2786	-0.121	-0.1276	-0.1253	0.1241	-0.245	0.1361	0.0903
Ip6	0.3655	-0.148	-0.1732	-0.1717	0.1729	-0.2117	0.069	0.0978
Ip7	0.2655	-0.004	-0.0878	-0.0929	0.1035	-0.1108	0.1817	-0.044
Ip8	0.3736	-0.113	-0.1957	-0.21	0.2117	-0.2954	0.2169	0.0252
Ip9	0.2907	-0.25	-0.2484	-0.244	0.2272	-0.1848	-0.3667	0.3422

	IP1	IP2	IP3	IP4	IP5	IP6	IP7	IP8	IP9
Log1/ic50
W
P2
P3
P2-P3
eq
V.w.
Id
Ip1	1
Ip2	-0.025	1
Ip3	-0.025	-0.025	1
Ip4	-0.25	-0.025	-0.025	1
Ip5	-0.025	-0.025	-0.025	-0.025	1
Ip6	-0.036	-0.0358	-0.0358	-0.0358	-0.0358	1
Ip7	-0.025	-0.025	-0.025	-0.025	-0.025	-0.0358	1
Ip8	-0.052	-0.052	-0.052	-0.052	0.4809	0.3071	0.4809	1
Ip9	-0.025	-0.025	-0.025	-0.025	-0.025	-0.0358	-0.025	0.052	1

Now out of the four model tetra parametric model was selected as the best model on the basis of highest value (Q² and R² value). The values give in parentheses are 70% confidence intervals of the regression coefficient model IV-5-3. Could explain 49.5% predict 24.28% of the variances of the interaction activity data. It is interesting to record that all four; statistically significant regression expressions are given in table-IV-5.

Finally, in order to confirm our best model for modeling of inhibition activity we have estimate log 1/IC₅₀ values using equation IV-5-4.

The values so obtained are recorded and compares with observed log 1/IC₅₀ , values in table IV-6. The further more, for obtaining results in favor the proposed model we have determined the difference (Residue) between observed that only in case of model (IV-5-4) gave smaller value for residue. This is finally confirms that our proposed tetra parametric model in most appropriate model for modeling inhibition activity log 1/IC₅₀ , for lipoxygenase inhibitor set of 41 compounds.

CONCLUSION:

On the basis of above observation our most appropriated model for modeling inhibition activity.

log 1/IC₅₀ = 0.3202 (±0.2409)Vw + 0.03474 (±0.01049)Id + 0.9435 (±0.5769)

IP₆ + 0.7797 (±0.4314) IP₈ + 7.543

n=41, R=0.6510, R²= 0.4238, R²A 0.3598,

SE= 0.7021, F= 6.6207 Q = 0.9272 Q² = 0.85969

The confirmations of result led to fallowing conclusion:-

1- Inhibition activity of lipoxygenase inhibiter modeled by using topological indices.

2- Good mode for modeling inhibition activity is tetra parametric model in involving Vw Id, IP₆ and IP₈ index:

3- Positive sign of Vw, Id, IP₆ and IP₈ indicate that they favored the inhibition activity.

Thus when bulky group displaced form the series of lipoxygenase inhibitor and (CH₂)₄ OH or (CH₂)₅ OH on and (CH₂)₅ substitute at R positions.

Hence the log 1/IC₅₀ inhibitor activity is also function of size, shape and presence of above mentioned substituted group.

REFERENCES:

1. A. G. Gilman, T. W. Rall, A. S. Nies, and P. Taylor (Ed.) The Pharmacological Basis of Therapeutics, McGraw –Hill International Editions, Medical Series, 1992, pp. 600–617.

2. C. W. Stallings, B. A. Kroa, R. T. Carroll, A. L. Mertzer, and M. O. Funk, Crystallisation and Preliminary X–ray Characterization of a Soybean Seed Lipoxygenase, J. Mol. Biol. 1990, 211, 685–687.

3. Hansen P.J. and Jurs P.C., Chemical applications of graph theory, J. Chem Edu.,65,574-580,1988.

4. J. C. Boyington and L. M. Amzel, Crystallization and Preliminary X–ray Analysis of Soybean Lipoxygenase–1, a “non–heme” Iron–Containing Dioxygenase, J. Biol. Chem. 1990, 165, 12771–12773.

5. K. Muller, 5–Lipoxygenase and 12–Lipoxygenase: Attractive Targets for the Development of Novel Antipsoriatic drugs, Arch. Pharm. 1994, 327, 3–19.

6. Lucic B.,Nikolic S., Trinajstic N., Juretic D.and Juric A., A novel QSPR approach to physico-chemical properties of the alpha-amino acids, Croatica Chem. Acta.68,435-450,1985.

7. M. O’Donnell and A. Welton, in: Therapeutic Approaches in Inflammatory Disease, Eds. A. J. Lewis and N. S. Doherty, Elsevier, New York, 1989, p. 169.

8. P. J. Piper, L. G. Letts, M. N. Samhoun, J. R. Tippins, and M. A. Palmer, Advances in Prostaglandin, Thromboxane and Leukotriene Research, Vol 9, Raven Press, New York, 1982, p.169.

9. P. M. Hilbers, A. Rossi, A. Finazzi–Agro, G. A. Veldink, and F. G. J. Vliegenthart, The Primary Structure of a Lipoxygenase from the Shoots of Etiolated Lentil Seedings Derived from its cDNA, Biochim. Biophys. Acta 1994, 1211, 239–242.

10. P. Nihn, A. Buge, T. Kohler, H. Lettau, and R. Schneider Trends bei der Entwicklung von Lipoxygenase– Hemmern, Pharmazie 1991, 46, 81–88.

11. P. Nuhn and K. Koch, Hemmer der Phospholipase A2, Pharmazie 1993, 48, 494–508.

12. R. D. Abeysinghe, P. J. Roberts, Ch. E. Cooper, K. H. MacLean, R. C. Hider, and J. B. Porter, The Enviroment of the Lipoxygenase Iron Binding Site Explored with Novel Hydroxy–Pyridinone Iron Chelators, J. Biol. Chem. 1996, 271, 7965–7972.

13. Rouvray D.H., Predicting chemistry from topology, Sci.Amer, 254, 40-47, 1986.

14. S. Yamamoto, Mammalian Lipoxygenases–Molecular Structures and Functions, Biochim. Biophys. Acta 1992, 1128, 117–128.

15. Seybold P.G, May M. and Bagal U.A., Molecular structure-property relationship, J.Chem. Edu.,64,575-581, 1987.

16. The Annals of Pharmacotherapy: Vol. 30, No. 7, pp. 858-864. 1996 Harvey Whitney Books Company.

Received on 02.02.2014 Modified on 10.03.2014

Research J. Science and Tech. 6(1): Jan.-Mar. 2014; Page 56-59