KR101004924B1 - Method for predicting activation energy with effective atomic descriptors - Google Patents

Abstract

The present invention relates to a method for predicting the activation energy of phase I metabolism by the CYP450 enzyme using effective atomic descriptors. Specifically, the present invention provides a method for predicting the activation energy of the hydrogen separation reaction by cytochrome P450 enzyme using an equation including an effective atomic descriptor. The present invention provides a method for predicting the activation energy of tetrahedral intermediate formation of aromatic hydroxylation by cytochrome P450 enzymes using equations containing effective atomic descriptors. The activation energy prediction result using the method of the present invention showed a high correlation with the activation energy calculated quantum mechanically. In addition, it was possible to predict the experimental metabolic rate through the activation energy predicted using the method of the present invention. The present invention makes computer development much easier by rapidly predicting the activation energy of Phase I metabolism up to a practical level without any experimental parameters or quantum calculations. In addition, through the prediction of activation energy of the present invention, i) rate of metabolism, ii) regioselectivity of metabolism, iii) inhibition of metabolism, and iv) interaction between drugs.

Activation energy, atomic descriptor, prediction, metabolism, hydrogen separation reaction, aromatic hydroxylation reaction, tetrahedral intermediate.

Description

Method for predicting activation energy with effective atomic descriptors}

The present invention relates to a method for predicting the activation energy of phase I metabolism by the CYP450 enzyme using effective atomic descriptors.

Metabolic prediction of foreign substances is important in the early stages of drug development. In particular, the reaction rate and regioselectivity of phase I metabolism are very important as the main pharmacokinetic properties of the substance, which can predict metabolic toxicity.

These reaction rates and regioselectivity can be predicted from activation energies, but current methods for estimating activation energies rely on time-consuming quantum mechanical calculations or difficult experiments. For example, KR Korzekwa et al. (J. Am. Chem. Soc. 1990, 112, 7042) has been reported a method of predicting, by quantum-mechanical calculation of the activation energy of the hydrogen separation reaction, TS Dowers et al. (Drug Metab . Dispos. 2004, 32, 328. ) has bar reported a method of predicting, by quantum-mechanical calculation of the activation energy of the aromatic hydroxylation. However, because these quantum mechanical methods have to be calculated at different states of the molecule, the exact activation energy cannot be obtained due to the complexity of the conformational differences between these states.

Therefore, the present inventors have developed a new model that is fast and accurate to predict the activation energy of Phase I metabolism only with the characteristics of an external substrate.

It is an object of the present invention to provide a method for predicting the activation energy of phase I metabolism by the CYP450 enzyme using effective atomic descriptors.

It is also an object of the present invention to provide a method for predicting the relative rate of metabolism, metabolic regioselection, inhibition of metabolism, and drug-drug interactions through the activation energy predicted by the method.

In order to achieve the object of the present invention, the present invention provides a method for predicting the activation energy of the hydrogen separation reaction by the CYP450 enzyme by a formula containing an effective atomic descriptor. The method of the present invention is fast and accurate and does not require any quantum mechanical calculations or experiments. (See Fig. 1)

Hydrogen separation reaction by Cytochrome P450 enzyme can be expressed as follows.

In the above scheme, the circle with Fe-O represents the oxyferryl intermediate.

]를 이용하여 사이토크롬 P450 효소에 의한 수소분리반응의 활성화에너지를 예측하는 방법을 제공한다.The present invention provides atomic _descriptors [ δ _het ], [max ( δ _heavy )], [ μ _CH ], [

] Provides a method of predicting the activation energy of the hydrogen separation reaction by cytochrome P450 enzyme.

In the above formula,

Circles with Fe-O represent oxyferryl intermediates,

[ δ _het ] represents the net atomic charge of the heteroatom in the alpha neighbor of the reaction center;

[max ( δ _heavy )] represents the highest atomic charge of X ¹ , X ² , X ³ which is not hydrogen or helium;

[ μ _CH ] represents a bonding dipole of carbon-hydrogen bonds;

]는 원자 H, C, X¹, X², X³ 의 알짜 원자 편극도(polarizability)들의 합을 나타낸다.[

] Represents the sum of the net atomic polarizabilities of atoms H, C, X ¹ , X ² , X ³ .

The present invention describes the overall CYP 450 enzyme, and it is clear that all of the disclosed invention can be applied to human CYP 450 enzyme. Cytochrome P450 enzymes according to the present invention include, but are not limited to, CYP2E1, CYP3A4, CYP2B6, CYP2C8, CYP2C9, CYP1A1, CYP1A2, CYP2C19, CYP2D6, CYP1B1, or CYP2A6.

In the activation energy prediction method, all CH bonds to the target molecule can be regarded as a position where metabolism can occur in the target molecule, and when C is aliphatic carbon among all CH bonds of the target molecule, hydrogen It can be determined that the metabolic site where the separation reaction can occur.

Hydrogen separation reaction by CYP450 enzyme can determine the atomic type depending on the presence or absence of heteroatom in the alpha neighbor of the reaction center (C-H where the actual metabolism occurs).

When there are heteroatoms in the alpha neighborhood of the reaction center, the atomic _descriptors [ δ _het ] and [max ( δ _heavy )] may be calculated and activation energies may be predicted according to the following equation.

Equation 1-1

E _a ^{Habs _ (B)} = 25.94 + 1.88 * [ δ _het ] + 1.03 * [max ( δ _heavy )]

In the above equation,

E _a ^{Habs _ (B)} represents the activation energy required to separate hydrogens attached to carbons having heteroatoms in the alpha neighbors of the reaction center.

]를 계산하고, 하기 방정식에 따라 활성화에너지를 예측할 수 있다. If there are no heteroatoms in the alpha neighborhood of the reaction center, the atomic descriptors [ μ _CH ] and [

], And the activation energy can be predicted according to the following equation.

[Equation 1-2]

] E _a ^{Habs _ (A)} = 28.50-2.22 * [ μ _CH ] + 1.12 * [

]

In the above equation,

E _a ^{Habs _ (A)} represents the activation energy required to separate hydrogens attached to carbons that do not have heteroatoms in the alpha neighbors of the reaction center.

In order to achieve the object of the present invention, the present invention provides a method for predicting the activation energy of the aromatic hydroxylation reaction by the CYP450 enzyme by a formula including an effective atomic descriptor. The method of the present invention is fast and accurate and does not require any quantum mechanical calculations or experiments. (See Fig. 1)

The tetrahedral intermediate formation reaction of the aromatic hydroxylation reaction with the Cytochrome P450 enzyme can be expressed as follows.

In the above scheme, the circle with Fe-O represents the oxyferryl intermediate.

The present invention provides a method for predicting the activation energy of the hydrogen separation reaction by the cytochrome P450 enzyme using the atomic descriptors [ δ _H ], [average ( α _alpha )].

In the above formula,

Circles with Fe-O represent oxyferryl intermediates,

[ δ _H ] represents the net atomic charge of the reactive hydrogen;

[Mean (α _alpha)] represents a mean value of the polarization of neighboring carbon also (polarizability).

The present invention describes the overall CYP 450 enzyme, and it is clear that all of the disclosed invention can be applied to human CYP 450 enzyme. Cytochrome P450 enzymes according to the present invention include, but are not limited to, CYP2E1, CYP3A4, CYP2B6, CYP2C8, CYP2C9, CYP1A1, CYP1A2, CYP2C19, CYP2D6, CYP1B1, or CYP2A6.

In the activation energy prediction method, all CH bonds to the target molecule can be regarded as a position where metabolism can occur in the target molecule, and when C is an aromatic carbon among all CH bonds of the target molecule, aromatic It can be determined that the metabolic site where the hydroxylation reaction can occur.

The present invention can calculate the atomic descriptors [ δ _H ], [average ( α _alpha )] and predict the activation energy according to the following equation.

Equation 2-1

E _a ^{aro _o, p} = ^21.34-0.75 * [ δ _H ] -1.24 * [mean ( α _alpha )]

[Equation 2-2]

E _a ^{aro _m} = 22.14 -0.68 * [ δ _H ] -0.83 * [mean ( α _alpha )]

[Equation 2-3]

E _a ^{aro _0,2,3} = ^21.02-1.49 * [ δ _H ] -0.92 * [mean ( α _alpha )]

In the above equation,

_A E ^{_o aro, p} is ortho (ortho) / p (para), which has one substituent in the position represents the activation energy of the tetrahedral intermediate formed in benzene;

E is _a ^{aro _m} meth (meta) 4 tetrahedra of the benzene with a single substituent at position represents the activation energy of the intermediates formed;

^Aro E _a ^{_0,2, 3} are 0, 2, it represents the activation energy of the tetrahedral intermediate formed in benzene which has 3 substituents.

], [δ _H ], [평균(α _alpha )] 등을 포함하지만, 이것에 제한되는 것은 아니다.As used herein, the term "atomic descriptor" means a value defined to reflect the properties of the atom itself, the binding environment of the atom, and the like. For example, in the present invention, the "atomic _descriptor " is [ δ _het ], [max ( δ _heavy )], [ μ _CH ], [

], [ δ _H ], [average ( α _alpha )] and the like, but are not limited thereto.

In another aspect, the present invention provides a method for predicting the relative kinetics of metabolism using the Arrhenius equation for the activation energy predicted by the method.

In the above equation, k is the reaction rate constant, A is the frequency factor, E _a is the activation energy, R is the gas constant, T represents the absolute temperature.

The reason why this equation is designed is that the atomic fraction above the activation energy f = e ^{- Ea / RT} Because. That is, since only a molecule exceeding the activation energy can cause a reaction, the rate constant is determined at a rate exceeding the activation energy.

In another aspect, the present invention provides a method for predicting regioselectivity of metabolism through activation energy predicted by the method.

More specifically, the present invention predicts the relative rate of metabolism by activating the energy predicted by the above method through the Arrhenius equation, and the positional selectivity of metabolism through the equation of the following equation. provide a way to predict regioselectivity:

In the above schemes and equations, P is the relative probability of formation of any metabolite of all possible metabolites of the substrate, E is an enzyme, S is a substrate, ES is a complex of enzyme and substrate, [ ES] is the concentration of enzyme and substrate complex, and k is the rate constant.

That is, if the reaction rate of each atom in a molecule is obtained through the Arrhenius equation, metabolism occurs well as the reaction rate is low, and thus the position selectivity in the molecule can be determined. Higgins, L .; Korzekwa, K. R .; Rao, S .; Shou, M .; Jones, J. P., An assessment of the reaction energetics for cytochrome P450-mediated reactions. Arch. Biochem. Biophys. 2001, 385, 220-230.

In another aspect, the present invention provides a method for predicting the inhibition of metabolism through the activation energy predicted by the method. For example, if a substrate has a relatively high activation energy, the substrate will not be metabolized and the substrate will remain in the active site of the CYP450 enzyme. As a result, it can be expected that the probability of inhibition by the substrate is high.

In another aspect, the present invention provides a method for predicting drug-drug interaction through the activation energy predicted by the method.

Here, 'drug interaction' refers to the effects of two or more drugs used together, changes in the absorption of the digestive system, changes in the rate of decomposition of drugs in the liver, new or increased side effects, changes in drug activity. Etc. CYP2C9, the isoform of CYP, is one of the major enzymes involved in the first-stage metabolism of drugs. Inhibition of this enzyme can lead to undesirable drug-to-drug interactions or drug toxicity [Lin, J. H .; Lu, A. Y. H., Inhibition and induction of cytochrome P450 and the clinical implications. Clin. Pharmacokinet. 1998, 35 (5), 361-390]. In other words, when the activation energy of the substrate is relatively high, metabolism is inhibited, resulting in inhibition of the CYP450 enzyme, which may eventually cause undesirable drug-to-drug interactions.

The method of the present invention makes it easier to develop new drugs using computers by rapidly predicting the activation energy of Phase I metabolism up to the practical level without any experimental parameters or quantum calculations, It suggests new empirical approaches that can be easily applied to activation energies and allows the explanation of the physicochemical factors that determine activation energies. In addition, the prediction of activation energy using this method can induce i) relative speed of metabolism, ii) regioselectivity of metabolism, iii) inhibition of metabolism, and iv) drug-drug interactions, as well as v) metabolic toxicity.

Below, The invention is illustrated in detail by the following examples. However, the following examples are merely to illustrate the invention, but the content of the present invention is not limited by the following examples.

Example  1: Development of model to predict activation energy of hydrogen separation reaction

The activation energy of the hydrogen separation is a good measure of the regioselectivity of the hydroxylation and dehydration of aliphatic hydrocarbons in phase I metabolism.

Scheme 1

Hydrogenation by Cytochrome P450 enzyme (circles with Fe-O represent oxyferyl intermediate).

In order to model the reaction, the activation energy of 431 cases of 119 molecules was calculated using the Austin Model 1 (AM1) molecular orbital theory.

Here, the cases refer to the number of atoms. For example, if there are three molecules with three, four or seven atoms, it would be '14 cases of three molecules'. AM1 is a semi-empirical method for quantum computation of the electronic structure of molecules in computational chemistry, generalizing the modified neglect of differential diatomic overlap approximation (Dewar, MJS et al., Journal of the American Chemical Society , 1985, 107, 3902).

The list of calculated organic molecules is shown in Table 1 below.

These information were used to train and evaluate experimental equations. These cases include methyl, primary, secondary and tertiary carbons in various chemical environments.

The inventors have separated these cases into two forms depending on whether the broken carbon-hydrogen bond has atoms other than electrons that are electronically negative, ie heteroatoms.

The equations modeled by atomic descriptors through the correlation between the effective atomic descriptors and the quantum mechanically calculated E _a of the hydrogen separation reaction are shown in Tables 2 and 3 below.

Correlation between the effective atomic descriptors and the quantum-mechanical calculated E _a of hydrogen separation reactions (if no heteroatoms are in the alpha neighborhood) Atomic descriptor Training set equation R ^a RMSE ^b μ _C-H 0.88 0.63 E _a ^{Habs _ (A)} = 28.50-1.19 * [ μ _CH ]

0.67 1.00 E _a ^{Habs _ (A)} = 28.50-0.90 * [

]

R ^a is the correlation coefficient.

RMSE ^b is the root mean squared error.

Correlation between Effective Atomic Descriptors and Quantum Mechanically Calculated E _a in Hydrogen Separation Reactions (When Heteroatoms Are in Alpha Neighbors) Atomic descriptor Training set equation R ^a RMSE ^b δ _het 0.82 1.51 E _a ^{Habs _ (B)} = 25.94 + 2.14 * [ δ _het ] max ( δ _heavy ) 0.57 2.16 E _a ^{Habs _ (B)} = 25.94 + 1.51 * [max ( δ _heavy )]

R ^a is the correlation coefficient.

RMSE ^b is the root mean squared error.

As a result of the training of Tables 2 and 3, the inventors have two standardized effective atomic descriptors for each form, allowing linear equations to predict activation energies in various chemical environments (Equations 1-1 and 1- below). 2).

]는 전이상태의 안정성을 나타낸다. 본 발명에서 모든 전이상태는 진동수 분석을 통해 검증되었다. Among these effective atomic _descriptors [ δ _het ], [max ( δ _heavy )], [ μ _CH ] indicate the fragility of the carbon-hydrogen bond, [

] Indicates stability of the transition state. All transition states in the present invention were verified through frequency analysis.

Activation energy prediction flowchart using the model of the present invention is as shown in FIG.

Specifically, the model for predicting the activation energy of the hydrogen separation reaction developed in the present invention is as follows:

i) identifying the metabolic position of the molecule of interest;

ii) determining the response type of the molecule of interest;

iii) determining the atomic form according to the presence or absence of a heteroatom in an alpha neighbor of a hydrogen separation reaction center;

]를 계산하는 단계;iv) atomic descriptors in the presence of heteroatoms [ δ _het ], [max ( δ _heavy )], atomic groups in the absence of heteroatoms [ μ _CH ], [

Calculating;

v) normalizing the atomic descriptors; And

vi) predicting the activation energy according to the following equation.

In the above formula,

Circles with Fe-O represent oxyferryl intermediates,

Equation 1-1

E _a ^{Habs _ (B)} = 25.94 + 1.88 * [ δ _het ] + 1.03 * [max ( δ _heavy )]

R = 0.91, RMSE = 1.14, n = 62, P value <0.0001

[Equation 1-2]

] E _a ^{Habs _ (A)} = 28.50-2.22 * [ μ _CH ] + 1.12 * [

]

R = 0.95, RMSE = 0.43, n = 224, P value <0.0001

In the above equation,

는 반응중심의 알파 이웃에 헤테로원자(탄소가 아닌 원자)를 가지고 있는 탄소에 붙은 수소가 분리되는데 필요한 활성화 에너지를 나타내고;

Represents the activation energy required for the separation of hydrogens attached to carbons having heteroatoms (non-carbon atoms) in the alpha neighbor of the reaction center;

는 반응중심의 알파 이웃에 헤테로원자(탄소가 아닌 원자)를 가지고 있지 않은 탄소에 붙은 수소가 분리되는데 필요한 활성화 에너지를 나타내고;

Denotes the activation energy required for the separation of hydrogens attached to carbons which do not have heteroatoms (non-carbon atoms) in the alpha neighbor of the reaction center;

[ δ _het ] represents the net atomic charge of a heteroatom (a non-carbon atom) in the alpha neighborhood of the reaction center;

[max ( δ _heavy )] represents the highest atomic charge of X ¹ , X ² , X ³ which is not hydrogen or helium;

[ μ _CH ] represents a bonding dipole of carbon-hydrogen bonds;

]는 원자 H, C, X¹, X², X³ 의 알짜 원자 편극도(polarizability) 들의 합을 나타낸다. [

] Represents the sum of the net atomic polarizabilities of atoms H, C, X ¹ , X ² , X ³ .

In Equations 1-1 and 1-2, R is a correlation coefficient, RMSE is a root mean squared error, n is the number of atoms used during training, and a P value ) Is a value indicating the significance of the correlation coefficient.

In step i), all of the C-H bonds to the molecule of interest can be seen as locations where metabolism can occur in the molecule of interest.

In step ii), when C (carbon) is aliphatic carbon among all C-H bonds of the target molecule, it may be regarded as a metabolic position where hydrogen abstraction of the target molecule may occur.

In step iii), when there is a heteroatom in the alpha neighbor of the reaction center (CH where the actual metabolism occurs), [Equation 1-1] is used, and there is no heteroatom. Equation 1-2 is used for this.

In step v), normalization means, from a statistical point of view, to make the mean of the values of the atomic descriptors zero and the standard deviation one. That is, normalization is performed before prediction with the mean and standard deviation of the values of the atomic descriptors used in the training of the prediction model of the present invention.

As shown in Figure 3, the activation energy predicted using the model of the present invention showed a high correlation with the activation energy calculated quantum mechanically. 386 of the 430 cases are within chemical accuracy (1 kcal per mol). Some of the discrepancies are due to the interactions other than the carbon-hydrogen-oxygen reactions reacting in quantum calculations. The activation energy in the gaseous state of various molecules was calculated using Gaussian 03 (Rev. C.02, M. J. Frisch et al., Pittsburgh, Pennsylvania, USA, 2003).

Example  2: Verification of Activation Energy Predicted by Activation Energy Prediction Model of Hydrogen Separation Reaction

Using the predictive model of Example 1, the activation energies of the predicted hydrogen separation reactions for the following four molecules were verified by comparison with experimental values.

[a] # represents the atomic number of each molecule of formula (2).

[b] Activation energy predicted by the method of the invention.

[c] Metabolic rate induced by incorporating the predicted activation energy of [b] into the Arrhenius equation.

[d] in vitro experimental metabolic rate.

Experimental metabolic rate of each molecule in Table 4 is already known in the art, hexane (Ken-ichirou MOROHASHI, Hiroyuki SADANO, Yoshiie OKADA, Tsuneo OMURA. Position Specificity in n-Hexane Hydroxylation by two forms of Cytochrome P450 in Rat liver Microsomes. J. Biochem . 1983, 93, 413-419; Octane is described by Jeffrey P. Jones, Allan E. Rettie, William F. Trager. Intrinsic Isotope Effects Suggest That the Reaction Coordinate Symmetry for the Cytochrome P-450 Catalyzed Hydroxylation of Octane Is Isozyme Independent. J. Med . Chem . 1990, 33, 1242-1246; Ethylbenzene is described by Ronald E. White, John P. Miller, Leonard V. Favreau, Apares Bhattacharyya. Stereochemical Dynamics of Aliphatic Hydroxylation by Cytochrome P-450. J. AM. Chem . Soc . 1986, 108, 6024-6031; 1-chloromethyl-4-methyl-benzene is described by LeeAnn Higgins, Kenneth R. Korzekwa, Streedhara Rao, Magong Shou, and Jeffrey P. Jones. An Assessment of the Reaction Energetics for Cytochrome P450-Mediated Reactions. Arch. Biochem. Biophys. 2001, 385, 220-230] can confirm the experimental metabolic rate of each molecule.

As shown in Table 4, four molecules such as hexane, octane, ethylbenzene, and 1-chloromethyl-4-methyl-benzene were obtained. Comparing the metabolic rate ^[c] and the experimental metabolic rate ^[d] derived by substituting the activation energy of the hydrogen separation reaction predicted according to the method of the present invention into the Arrhenius equation, a similar trend can be confirmed. That is, the experimental metabolic rate can be predicted through the activation energy predicted according to the present invention.

Example  3: Development of a model for predicting activation energy of tetrahedral intermediate formation of aromatic hydroxylation

We modeled tetrahedral intermediate formation, which is a good measure of the regioselectivity of aromatic hydroxylation of phase I metabolism (Scheme 2, Formula 2).

Scheme 2

Tetrahedral intermediate formation reaction of aromatic hydroxylation with cytochrome P450 enzyme (circle with Fe-O represents oxyferyl intermediate).

In order to model the reaction, the activation model of 85 cases of 31 molecules of benzenecarbons in various chemical environments was calculated using the Austin Model 1 (AM1) molecular orbital theory.

Here, the cases refer to the number of atoms. For example, if there are three molecules with three, four or seven atoms, it would be '14 cases of three molecules'. AM1 is a semi-empirical method for quantum computation of the electronic structure of molecules in computational chemistry, generalizing the modified neglect of differential diatomic overlap approximation (Dewar, MJS et al., Journal of the American Chemical Society , 1985, 107, 3902).

The list of calculated organic molecules is shown in Table 5 below.

These information were used to train and evaluate experimental equations. These cases were divided into three forms: i) one substituent at the ortho / para position, ii) one substituent at the meta position, iii) zero , Having two or three substituents.

The equations modeled by the atomic descriptors through the correlation between the effective atomic descriptors and the quantum mechanically calculated E _a of the aromatic hydroxylation reaction are shown in Tables 6, 7, and 8.

Correlation between Effective Atomic Descriptor and Quantum Mechanically Calculated E _a of Aromatic Hydroxide Reactions (with Ortho Substituents) Atomic descriptor Training set equation
R ^a RMSE ^b δ _H 0.08 1.31 E _a ^{aro _o, p} = ^14.67 + 63.33 * [ δ _H ]

0.57 1.07 E _a ^{aro _o, p} = 61.60-26.53 * [

]

R ^a is the correlation coefficient.

RMSE ^b is the root mean squared error.

Correlation between Effective Atomic Descriptors and Quantum Mechanically Calculated E _a of Aromatic Hydroxide Reactions (With Meta-Substituents) Atomic descriptor Rue set equation
R ^a RMSE ^b δ _H 0.03 0.56 E _a ^{aro _m} = -12.61 + 333.87 * [ δ _H ]

0.50 0.49 E _a ^{aro _m} = 132.75-72.54 * [

]

R ^a is the correlation coefficient.

RMSE ^b is the root mean squared error.

Correlation between Effective Atomic Descriptors and Quantum Mechanically Calculated E _a of Aromatic Hydroxide Reactions (with One, Two, Three Substituents) Atomic descriptor Training set equation
R ^a RMSE ^b δ _H 0.69 0.95 E _a ^{aro _0,2,3} = 70.00-465.88 * [ δ _H ]

0.05 1.31 E _a ^{aro _0,2,3} = 17.65 + 2.21 * [

]

R ^a is the correlation coefficient.

RMSE ^b is the root mean squared error.

As a result of the training process of Tables 6 to 8, the inventors have two standardized effective atomic descriptors for each type to enable linear equations to predict activation energies in various chemical environments (Equations 2-1 and 2- below). 2 and 2-3).

From the tetrahedral effective atomic presenter used in the equation to predict the activation energy of the intermediate formed in the reaction [δ _H] is oxygenate species determining the proximity between (oxygenating species) and the substrate, and [mean (α _alpha)] is It is related to the stability of the transition state. In the present invention, a para-nitrosophenoxy radical (PNR) was used as an oxygenating species, and all transition states were verified through frequency analysis.

A flowchart of predicting the activation energy using the model developed in the present invention is shown in FIG. 2.

Specifically, the model for predicting the activation energy of tetrahedral intermediate formation of aromatic hydroxylation by the CYP450 enzyme developed in the present invention is as follows:

i) identifying the metabolic position of the molecule of interest;

ii) determining the response type of the molecule of interest;

iii) calculating the atomic descriptors [ δ _H ], [average ( α _alpha )] when it is determined in the step ii) that the aromatic hydroxide reaction;

iv) normalizing the atomic descriptors; And

v) predicting the activation energy according to the equation

In the above formula,

Circles with Fe-O represent oxyferryl intermediates.

Equation 2-1

E _a ^{aro _o, p} = 21.34 -0.75 * [ δ _H ] -1.24 * [mean ( α _alpha )]

R = 0.71, RMSE = 0.95, n = 16, P value = 0.009

[Equation 2-2]

E _a ^{aro _m} = 22.14 -0.68 * [ δ _H ] -0.83 * [mean ( α _alpha )]

R = 0.88, RMSE = 0.30, n = 8, P value = 0.026

[Equation 2-3]

E _a ^{aro _0,2,3} = 21.02 -1.49 * [ δ _H ] -0.92 * [mean ( α _alpha )]

R = 0.87, RMSE = 0.65, n = 33, P value <0.0001

In the above equation,

_A E ^{_o aro, p} is ortho (ortho) / p (para), which has one substituent in the position represents the activation energy of the tetrahedral intermediate formed in benzene;

E is _a ^{aro _m} meth (meta) 4 tetrahedra of the benzene with a single substituent at position represents the activation energy of the intermediates formed;

^Aro E _a ^{_0,2, 3} are 0, 2, represents the activation energy of the tetrahedral intermediate formed in benzene which has 3 substituents;

[ δ _H ] represents the net atomic charge of the reactive hydrogen;

[Mean (α _alpha)] represents a mean value of the polarization of neighboring carbon also (polarizability).

In Equations 2-1, 2-2, and 2-3, R is a correlation coefficient, RMSE is a root mean squared error, n is the number of atoms used during training, P P value is a value indicating the significance of the correlation coefficient.

As shown in Figure 4, the activation energy prediction results using the method of the present invention showed a high correlation with the activation energy calculated quantum mechanically. 70 of the 85 cases are within chemical accuracy (1 kcal per mol). Some discrepancies were caused by the model not considering the ortho, meta, and para effects when modeling benzene with 0, 2, or 3 substituents. The activation energy in the gaseous state of various molecules was calculated using Gaussian 03 (Rev. C.02, M. J. Frisch et al., Pittsburgh, Pennsylvania, USA, 2003).

Example  4: Verification of activation energy predicted by activation energy prediction model of tetrahedral intermediate formation of aromatic hydroxylation

Using the predictive model of Example 3, the activation energy of the tetrahedral intermediate formation reaction predicted for the next two molecules was verified by comparing with the experimental values.

[a] # represents the atomic number of each molecule of formula (4).

[b] Activation energy predicted by the method of the invention.

[c] Metabolic rate derived by putting the predicted activation energy of [b] into the Arrhenius equation.

[d] in vitro experimental metabolic rate.

The experimental metabolic rate of each molecule in Table 9 is already known in the art, and Methoxybenzene is described by Robert P. Hanzlik, Kerstin Hogberg, Charles M. Judson. Microsomal hydroxylation of specifically deuterated monosubstituted benzenes. Evidence for direct aromatic hydroxylation. Biochemstry . 1984, 23, 3048-3055; Chlorobenzene is described by HG Selander, DM Jerina, JW Daly. Metabolism of Chlorobenzene with Hepatic Microsomes and Solubilized Cytochrome P-450 Systems. Arch. Biochem . Biophys . 1975, 168, 309-321] can be found in the experimental metabolic rate of each molecule.

As shown in Table 9, two molecules, such as methoxybenzene and chlorobenzene, were derived by substituting the energetic energy of the hydrogen separation reaction predicted according to the method of the present invention into an Arrhenius equation. Comparing metabolic rate ^[c] and experimental metabolic rate ^[d] , a similar trend can be observed. That is, the experimental metabolic rate can be predicted through the activation energy predicted according to the present invention.

As such, those skilled in the art to which the present invention pertains will understand that the present invention may be embodied in other specific forms without changing to the technical spirit or essential features of the present invention. Therefore, the above-described embodiments are to be understood in all respects as illustrative and not restrictive. The scope of the invention is indicated by the following claims rather than the above description, and it should be construed that all changes or modifications derived from the meaning and scope of the claims and their equivalents are included in the scope of the invention. .

1 illustrates an empirical method for predicting the activation energy of phase I metabolism by CYP 450 enzyme using effective atomic descriptors.

2 is a flowchart illustrating activation energy using a model of the present invention.

Figure 3 shows the correlation between the activation energy (QM E _a ) and the activation energy (Predicted E _a ) predicted by the present invention with respect to the hydrogen separation reaction by the CYP450 enzyme.

Figure 4 shows the correlation between the activation energy (QM E _a ) and the activation energy (Predicted E _a ) predicted by the present invention with respect to the aromatic hydroxylation reaction by the CYP450 enzyme.

Claims (24)

When C is aliphatic carbon among all CH bonds of the target molecule having the following chemical formula, the atomic _descriptors [ δ _het ], [max ( δ _heavy )], [ μ _CH ], [

] To predict the activation energy by judging the metabolic position where hydrogen separation reaction by cytochrome P450 enzyme can occur:

In the above formula, Circles with Fe-O represent oxyferryl intermediates, [ δ _het ] represents the net atomic charge of the heteroatom in the alpha neighbor of the reaction center; [max ( δ _heavy )] represents the highest atomic charge of X ¹ , X ² , X ³ which is not hydrogen or helium; [ μ _CH ] represents a bonding dipole of carbon-hydrogen bonds; [

] Represents the sum of the net atomic polarizabilities of atoms H, C, X ¹ , X ² , X ³ . The method of claim 1, wherein the cytochrome P450 enzyme is any one selected from the group consisting of CYP2E1, CYP3A4, CYP2B6, CYP2C8, CYP2C9, CYP1A1, CYP1A2, CYP2C19, CYP2D6, CYP1B1, and CYP2A6. delete 2. The method of claim 1, wherein the atomic descriptors [ δ _het ] and [max ( δ _heavy )] are calculated when there are heteroatoms in the alpha neighbor of the reaction center. The atomic descriptors [ μ _CH ] and [when the alpha neighbor of the reaction center is absent.

]. The method of claim 4, wherein the activation energy is predicted according to the following equation: E _a ^{Habs _ (B)} = 25.94 + 1.88 * [ δ _het ] + 1.03 * [max ( δ _heavy )] In the above equation, E _a ^{Habs _ (B)} represents the activation energy required to separate hydrogens attached to carbons having heteroatoms in the alpha neighbors of the reaction center. The method of claim 5, wherein the activation energy is predicted according to the following equation: E _a ^{Habs _ (A)} = 28.50-2.22 * [ μ _CH ] + 1.12 * [

] In the above equation, E _a ^{Habs _ (A)} represents the activation energy required to separate hydrogens attached to carbons that do not have heteroatoms in the alpha neighbors of the reaction center. A method of predicting the relative rate of metabolism (k) of the activation energy predicted by the method of any one of claims 1, 2 and 4 to 7, through the following Arrhenius equation:

In the above equation, k is the reaction rate constant, A is the frequency factor, E _a is the activation energy, R is the gas constant, T represents the absolute temperature. A method for predicting the regioselectivity of metabolism through the activation energy predicted by the method of any one of claims 1, 2 and 4-7. 10. The metabolic regioselectivity of claim 9, wherein the predicted activation energy is predicted by the Arrhenius equation and the relative rate of metabolism, and the predicted relative rate of metabolism is determined through the equation below. How to predict:

In the above schemes and equations, P is the relative probability of formation of any metabolite of all possible metabolites of the substrate, E is an enzyme, S is a substrate, ES is a complex of enzyme and substrate, [ ES] is the concentration of enzyme and substrate complex, and k is the rate constant. A method for predicting the inhibition of metabolism through the activation energy predicted by the method of any one of claims 1, 2 and 4-7. A method for predicting drug-drug interactions through activation energy predicted by the method of any one of claims 1, 2 and 4-7. When C is aliphatic carbon among all CH bonds of the target molecule having the following formula, 4 is the aromatic hydroxylation reaction by cytochrome P450 enzyme using atomic descriptors [ δ _H ], [mean ( α _alpha )]. Method of predicting activation energy by judging the formation of tetrahedral intermediates as a metabolic site where hydrogen separation reaction can occur:

In the above formula, Circles with Fe-O represent oxyferryl intermediates, [ δ _H ] represents the net atomic charge of the reactive hydrogen; [Average ( _alpha )] represents an average value of polarizability of neighboring carbons. The method of claim 13, wherein the cytochrome P450 enzyme is any one selected from the group consisting of CYP2E1, CYP3A4, CYP2B6, CYP2C8, CYP2C9, CYP1A1, CYP1A2, CYP2C19, CYP2D6, CYP1B1, and CYP2A6. delete The method of claim 13, wherein the atomic descriptors [ δ _H ], [average ( α _alpha )] are calculated. The method of claim 16, wherein the activation energy is predicted according to the following equation: E _a ^{aro _o, p} = ^21.34-0.75 * [ δ _H ] -1.24 * [mean ( α _alpha )] In the above equation, _A E ^{_o aro, p} indicates the ortho (ortho) / p (para) have a single substituent at position there is activation of the tetrahedral intermediate formed in benzene energy. The method of claim 16, wherein the activation energy is predicted according to the following equation: E _a ^{aro _m} = 22.14-0.68 * [ δ _H ] -0.83 * [mean ( α _alpha )] In the above equation, E _a represents an activation energy of ^{aro _m} meth forming tetrahedral intermediates of benzene, which has one substituent in the (meta) position. The method of claim 16, wherein the activation energy is predicted according to the following equation: E _a ^{aro _0,2,3} = ^21.02-1.49 * [ δ _H ] -0.92 * [mean ( α _alpha )] In the above equation, E _a ^{aro _0,2,3} Represents the activation energy of the tetrahedral intermediate formation of benzene having 0, 2, 3 substituents. 20. A method of predicting the relative kinetics of metabolism of the activation energy predicted by the method of any one of claims 13, 14 and 16-19 by using the following Arrhenius equation:

In the above equation, k is the reaction rate constant, A is the frequency factor, E _a is the activation energy, R is the gas constant, T represents the absolute temperature. 20. A method for predicting regioselectivity of metabolism via activation energy predicted by the method of any one of claims 13, 14 and 16-19. 22. The metabolic regioselectivity of claim 21, wherein the predicted activation energy is predicted by the Arrhenius equation for the relative rate of metabolism, and the predicted relative rate of metabolism is determined through the equation of How to predict:

In the above schemes and equations, P is the relative probability of formation of any metabolite of all possible metabolites of the substrate, E is an enzyme, S is a substrate, ES is a complex of enzyme and substrate, [ ES] is the concentration of enzyme and substrate complex, k is the rate constant. 20. A method for predicting the inhibition of metabolism via the activation energy predicted by the method of any one of claims 13, 14 and 16-19. 20. A method for predicting drug-drug interactions via activation energy predicted by the method of any one of claims 13, 14 and 16-19.

Images