THE EFFECT OF THE MOLECULAR PROPERTIES OF CALCIUM CHANNEL BLOCKERS ON THEIR ELIMINATION ROUTE

Calcium channel blockers (CCBs) are among the most widely used drugs in cardiovascular medicine. In this study, nine CCBs (amlodipine, felodipine, isradipine, nicardipine, nifedipine, nimodipine, nisoldipine, verapamil and diltiazem) were investigated to assess the relationship between their molecular properties and elimination data obtained from literature. The descriptors of the molecular properties of CCBs were calculated using three software packages. The relationship between computed molecular properties and elimination data collected from relevant literature, initially investigated with simple linear regression analysis, showed poor correlation (R2 <0.25). Application of molecular weight or volume data as additional independent variable, multiple linear regression (MLR) revealed better correlations (R2 ~ 0.38) between CCB renal and fecal elimination data and their lipophilicity. Excluding nimodipine from the calculations resulted in more acceptable correlations. The best correlations were established after computed lipophilicity descriptor and molecular weight were applied (R2 = 0.66 with acceptable probability value).


INTRODUCTION
Calcium channel blockers (CCBs), also known as calcium antagonists, are commonly prescribed drugs for the treatment of various conditions of the heart and blood vessels, such as hypertension, angina pectoris, supraventricular dysrhythmias, after myocardial infarction and in migraine prophylaxis.They support vasodilator activity and reduce blood pressure by decreasing calcium influx into vascular smooth muscle cells (Stella et al., 2007;Lemke and Williams, 2008;Sweetman, 2009;Beale and Block, 2011;Mofat et al., 2011).
Most CCBs, with the exception of amlodipine (64-90%) and nifedipine (45-70%), have variable oral bioavailability because of extensive first-pass metabolism.Their half-life is relatively short -usually shorter than 12 hours, with the exception of amlodipine which, although extensively metabolized, has half-life of 35-50 hours.According to the available literature data, CCBs have a dual route of elimination, renal and fecal.The lowest values for renal and highest values for fecal elimination of CCBs was found for diltiazem, while for nimodipine renal elimination is the major route of elimination.CCBs combine well with antihypertensive drugs that block the rennin-angiotensin system, ACE inhibitors and angiotensin receptor blockers (Stella et al., 2007;Lemke and Williams, 2008;Sweetman, 2009;Beale and Block, 2011;Mofat et al., 2011).
The clinical success of drugs mostly depends on their absorption, distribution, metabolism or route of elimination (Di and Kernsy, 2003).Lipophilicity is one of the most important molecular properties that influences these values, but a number of other molecular properties, such as molecular weight (Mw), molecular volume (Vol), polar surface area (PSA) and solubility data (logS), also play an important role in drug absorption, penetration into tissues, degree of distribution, degree of plasma protein binding and route of elimination (Hartman and Schmitt, 2004;Remko et al.,, 2006;Remko, 2007;Zhao, 2001).According to the available literature, a number of authors investigated antihypertensive drugs including those belonging to the CCB group, their design and synthesis (Christiaans et al., 1993;Kalavagunta et al., 2014), as well as pharmacokinetics, pharmacodynamics and efficacy (Pozo and Baeyens, 1986;Sepehr-Ara et al., 2011;Ian Whyte et al., 2012;Mayama, 2014).Their acidity, lipophilicity, solubility or absorption were evaluated together with large groups of different drugs based on their molecular structure with the application of computer programs (Remko et al., 2006;Remko, 2007;Zhao, 2001).
The aim of this study was to compare the different molecular properties of nine CCBs (amlodipine, felodipine, isradipine, nicardipine, nifedipine, nimodipine, nisoldipine, verapamil and diltiazem) and their elimination data.1).
CCB aqueous solubility data (logS) were calculated using the software Virtual Computational Chemistry Laboratory and the software package Molinspiration Depiction Software (Molinspiration Cheminfirmatics) was used for the calculation of electronic descriptors: polar surface area (PSA); constitutional parameter (molecular weight (Mw)); geometric descriptor (volume value (Vol)).Chemdraw ultra 12.0 was used for the calculation of lipophilicity parameters, ClogP values.All calculated molecular descriptors are presented in Table 1.The elimination data of the investigated CCBs (Table 1) were obtained from the relevant literature (Lemke and Williams, 2008).Microsoft Excel 2003 and Origin 7.0 PRO (Origin Lab Corporation, USA) were used for statistical analysis.

RESULTS
According to the available data from literature, CCBs have a dual route of elimination, (renal and fecal) and renal elimination has been shown as the major one.The lowest values for renal and highest for fecal elimination were found for diltiazem (35% and 65%, respectively).For nimodipine, renal elimination is the major route of elimination (Table 1).
The five CCB molecular descriptors (PSA, Mw, Vol, logP, logS) were calculated using different software packages (Table 1).The correlations between CCB elimination data obtained from the relevant literature and the calculated molecular descriptors (PSA, Mw, Vol, logP, logS) were initially investigated using simple linear regression, providing very poor correlation with coefficients (R 2 ) of around 0.25.
In the next stage of the study, the relationships between CCB renal and fecal elimination data and two different CCB molecular descriptors were investigated using multiple linear regression (MLR).The relationship between CCB fecal as well as renal elimination data and lipophilicity with the application of molecular weight or volume data as an additional independent variable provided similar correlations, with coefficients (R 2 ) around 0.38.The predicted values of CCB renal and fecal elimination data are presented in Table 2.
Finally, all relationships were investigated after nimodipine exclusion, since it was noticed that However, for relationships between renal elimination data collected from the relevant literature and those predicted, applying computed lipophilicity descriptor (ClogP) and molecular weight (Mw) or molecular volume (Vol) as independent variables, higher correlations were established.The better correlation, obtained using MLR analysis applying ClogP and Mw as independent variables (Eq. 1) is presented in Fig. 1.The established correlation can be considered as good, as has been proposed previously by Asuero et al. (2006), with acceptable P values, due to the limited number of compounds.
The correlation that was found between CCB renal elimination data and their in silico molecular descriptors, the lipophilicity parameter (ClogP) and the constitutional parameter (molecular weight (Mw)), confirmed descriptor calculation as the highthroughput screening technique (HTS) for evaluation of elimination of the selected compounds.

DISCUSSION
In this research nine CCBs (amlodipine, felodipine, isradipine, nicardipine, nifedipine, nimodipine, nisoldipine, verapamil and diltiazem) were studied in order to evaluate the correlation between their renal and fecal elimination data obtained from relevant literature and calculated molecular descriptors.The main aim was to establish a high throughput approach using simple or multiple linear regression analysis capable of predicting elimination data of the selected CCBs.Several CCB molecular descriptors were calculated using three different software packages.All calculated descriptors, electronic descriptor -PSA, constitutional parameter -Mw, geometric descriptor -Vol, lipophilicity descriptors − ClogP values, as well as aqueous solubility data -logS, play an important role in drug absorption, distribution, metabolism and route of elimination.
The molecules with high lipophilicity show a higher degree of absorption, better penetration into tissues and distribution compared to less lipophilic molecules with similar properties.Drug absorption as well as duration of action or efficiency and elimination is highly affected by its lipophilicity, solubility, molecular size and other molecular properties (Lipinski, 2000;Ghose el al. 1999).The weakly lipophilic drugs are mostly eliminated by urine, while the highly lipophilic ones usually exhibit a high degree of fecal elimination.
According to the available literature, CCB pharmacokinetics, pharmacodynamics and efficacy were investigated by a number of authors (Pozo and Baeyens, 1986;Sepehr-Ara et al., 2011;Ian Whyte et al., 2012;Mayama, 2014).Also, various authors suggested several assays that could be employed in studies of the elimination of different drugs (Hellstern et al., 1990;Kullak-Ublick and Becker, 2003;Verho et al., 1995;Martin et al., 2003).However, most of these methods have certain limitations and a new approach for a fast, reliable and cost-effective evaluation of the CCB route of elimination should be developed.Since the route and degree of elimination of drugs may affect their duration of action and activity, the application of computed molecular descriptors in the prediction of drug elimination are of great importance, especially for the newly synthesized drugs.
In the present study, the correlations between calculated CCB molecular descriptors and their elimination data obtained from relevant literature were examined.In the first step of the investigation the correlations between CCBs elimination data obtained from relevant literature and the calculated descriptors were investigated using simple linear regression.Fecal elimination data and CCBs molecular descriptors − Vol, Mw and logS, showed very low correlations (R 2 <0.10), while only the electronic descriptor (polar surface area, PSA) and lipophilicity descriptor (ClogP) provided correlation with coefficients (R 2 ) of around 0.25.In the next stage of the study, the relationship between CCB renal as well as fecal elimination data and two different CCB molecular descriptors were investigated using multiple linear regression.ClogP was chosen as the first independent variable since it showed the best correlations with CCB elimination data (renal as well as fecal), while values of Mw, Vol and logS were chosen as possible second independent variables.Values of electronic descriptor (PSA) couldn't be used as the second independent variable, since in correlation with ClogP, R 2 was 0.39.The relationship between CCB fecal elimination data and lipophilicity with application of logS as the second independent variable provided very poor correlation (R 2 < 0.25).Correlations between CCB fecal elimination data and lipophilicity with the application of molecular weight or volume data as additional independent variables provided the same coefficients (R 2 = 0.39), while correlations with R 2 = 0.37 were established for renal elimination data.However, among all investigated CCBs, only for nimodipine was renal elimination appointed as the major route of elimination, without absolute values of fecal elimination data.Therefore, the investigated relationships were recalculated after excluding nimodipine and significantly higher correlations were found.The best correlation was established between renal elimination data collected from relevant literature and those predicted applying the computed ClogP and Mw (R 2 = 0.66).

CONCLUSION
The correlation between nine CCB elimination data and five different molecular descriptors were investigated.The applicability of calculated molecular descriptors, especially the lipophilicity descriptor, ClogP and molecular weight (Mw) in CCB elimination evaluation was established.The proposed methodology confirmed that lipophilicity, together with other molecular properties, is essential for drug elimination and could assist in the in vitro approach to assessing CCB elimination.

Fig. 1 .
Fig. 1.The relationship between CCB renal elimination data collected from the literature (A) and predicted (B) using ClogP and Mw values.Numbers denote CCBs -calcium channel blockers (Table1).

Table 1 .
Data of CCBs renal and fecal elimination data collected from relevant literature and calculated molecular descriptors

Table 2 .
CCB renal and fecal elimination data predicted from ClogP and Mw ( a ) and ClogP and Vol values ( b ).