Indian Journal of Biochemistry & Biophysics Vol. 44, December 2007, pp. 458-469

.Biological activity predictions, crystallographic comparison and hydrogen bonding analysis of cholane derivatives Rajnikant*, Dinesh and Bhavnaish Chand Condensed Matter Physics Group, Post-Graduate Department of Physics, University of Jammu, Jammu Tawi-180006, India

Received 13 November 2006; revised 27 July 2007

A total of eighteen molecules of cholane derivatives (I-XVIII) (a series of ) have been included to predict their pharmacological effects, specific mechanisms of action, known toxicities, drug-likeness, etc, by using the statistics of multilevel neighbourhoods of atoms (MNA) descriptors for active and inactive fragments. The biological activity spectra for substances have been correlated on SAR base (structure-activity relationships data and knowledge base), which provides the different Pa (possibility of activity) and Pi (possibility of inactivity). Most of the probable activities have been characterized by Pa and Pi values, which depict that all the molecules have high value of teratogen activity. The Lipinski’s thumb rule predicts that all the cholane derivatives have stronger preponderance for “cancer-like-drug” molecules and some of their related analogous have entered in the ANCI (American National Cancer Institute) database. Some selected bond distances and bond angles of interest have been taken into account and deviation of bond distances/bond angles, vis-a-vis the substitutional group and X–H…A intra/intermolecular hydrogen bonds has been discussed in detail. X–H…A intra and intermolecular hydrogen bonds in the molecules have been described with the standard distance and angle cut-off criteria. D–θ and d–θ scatter plots for intra- and intermolecular interactions are presented for better understanding of packing interactions existing among these derivatives. There exists only one C–H…O intramolecular bifurcated hydrogen bond, while high tendency of intermolecular bifurcated hydrogen bonds based on a defined O–H…O has been observed, in which O atom acts as a prototype donor as well as acceptor. The frequency of occurrence of C–H…O hydrogen bonds is predominant (i.e. 85.7%) in intramolecular interactions, whereas in intermolecular interactions, frequency of occurrence for O–H…O interactions is 62.9%. Solvent-solute/solute-solvent interactions have also been investigated to understand more complicated processes that occur for biomolecules in aqueous solutions. The number of hydrogen donors in each derivative is less than 5, except for molecule XVIII and which has 91.3% of drug-likeness, instead of observed range of 96.5-99.3%.

Keywords: Cholane, X-ray diffraction, Biological activity, Intra and intermolecular hydrogen bonds, Bifurcated hydrogen bonds, Solvent/solute interaction, Lipinski’s rule

The basic cholane molecule contains a carbon skeleton or nucleus consisting of four-ring structure of which three are six-membered cyclohexane rings, one is five-membered cyclopentane ring and a side chain of five carbon atoms located at C17 position of the nucleus. A representative illustration of cholane molecule is presented in Fig. 11. Generally, the C24 compounds (bile acids) are hydroxy derivatives of cholanic acid and they have a 3α- hydroxyl group (with one exception), while other Fig. 1—Basic cholane molecule (C24) with standard atomic numbering scheme hydroxyl groups may be present at C6, C7, C12 and occasionally at other carbon atoms. Bile acids play a mammals, and chenodeoxycholic acid are key role in regulation of cholesterol metabolism the principal products. Before excretion into the bile, including synthesis, absorption, esterification, these acids conjugate with either glycine or taurine to excretion and catabolism to bile acids2. They are produce the bile salts3 which enter the small intestine produced in the liver from cholesterol and in most and facilitate lipid absorption4,5. Bile acids are largely reabsorbed from the intestine6,7 and pass back to the ______liver in the enterohepatic circulation. Bile acid level in *Corresponding author the enterohepatic circulation regulates the rate of bile E-mail: [email protected] 8 Tel/Fax: 0191-2432051 acid synthesis . RAJNIKANT et al.: BIOLOGICAL ACTIVITY PREDICTIONS 459

Deoxycholic acid (DCA) is a bile acid having the conditions. By using a qualitative representation of unusual property of forming canal complexs with a biological activity, it is possible to compare and mix wide variety of organic compounds, including alipha- activity data for obtaining robust quantitative models. tic and aromatic hydrocarbons, fatty acids, alcohols, The crystal structures can be described by descriptors esters, ethers, phenols, azo-dyes, alkaloids, camphor, Multilevel Neighbourhoods of Atoms (MNA)39. MNA methyl-orange, β-carotene and cholesterol9-11. The descriptors are based on a molecular structure sodium salt of DCA can associate with several description according to the valences and partial biological systems as a particular class of histones charges of connected atoms (including hydrogen from chromatin12, phospholipids13, proteins14,15 and atoms). They are generated as a recursively defined sequential polypeptides16. Cholic acid (CA), another sequence: important bile acid has proved particularly useful • Zero-level MNA descriptor for each atom is the as engineering component for supramolecular label A of the atom itself; chemistry17. In the crystalline state, it permits optical • Any next-level MNA descriptor for the atom is the resolution of racemic lactones18 and demonstrates sub-structure notation 19 interecalation phenomena for organic molecules . A(D1D2 Di ), where Di is the previous-level MNA descriptor for ith immediate neighbour of the In view of the relationship of various parent steroid atom A. The neighbor descriptors D D D are hydrocarbons viz., , cholane, , 1 2 i arranged in a unique manner, e.g., in lexicographic , , we got interested in the said order. The atom label A may include not only the series of steroids which should essentially have (i) atomic type but also any additional information about some biological activity, (ii) molecular and crystal the atom. Iterative procedure for MNA descriptor structure, (iii) molecular geometry and conforma- generation can be continued to cover the first, second, tional parameters, (iv) non-planar conformations of etc. neighbourhoods of each atom. In this manner, individual ring systems and (v) hydrogen bonding structure of any molecule is represented as a interactions. The work reported in this paper is a part set of MNA descriptors. The MNA descriptors of our on-going research on the steroids20-23. represent various structure property relationships In the present paper, a series of 18 cholane including biological activities40, mutagenecity and derivatives have been undertaken for study from the carcinogenecity41, boiling point39, drug-likeness42, etc. literature available through on-line Cambridge The three-dimensional (3D) coordinate data of all structural database (U.K.). A survey of this database the compounds have been selected as an input for the has been made to ascertain the existence of number of PASS software43 to predict the structure-activity structures reported on the cholane derivatives. We relationship. On the basis of x, y, z coordinates, have considered the structural data of only those molecular structures have been drawn and compounds, whose crystal data have been collected accordingly biological activity predictions have been by using single crystal X-ray diffractometer and a determined on the statistics of MNA descriptors for comparative study has been evolved vis-a-vis each active and inactive fragments. The biological activity other. The chemical structures of molecules I–XVIII spectra for substances have been correlated on SAR are shown in Fig. 2. The chemical name and base (structure-activity relationships data and published references of each molecule are presented knowledge base), which provides different P in Table 124-38. a (possibility of activity) and Pi (possibility of inactivity) values. Based on statistics of MNA Methodology descriptors for active and inactive compounds, two Biological activity predictions probabilities: Pa and Pi⎯probability of the compound Biological activity spectrum provides the rationale being active and inactive respectively, have been for predicting biological activity types for different calculated for each activity. Influence of these compounds. Within this concept, the biological descriptors can be positive or negative (if they are activity is considered as an intrinsic property of the found in compounds with or without particular compound, depending only on its structure. Any activity respectively) or even neutral. The Pa and Pi component of this spectrum of a given compound is values for the molecules (I-XVIII) are given in assumed to be detectable under suitable experimental Table 2. 460 INDIAN J. BIOCHEM. BIOPHYS., VOL. 44, DECEMBER 2007

Fig. 2—Chemical structures of molecules 1–XVIII RAJNIKANT et al.: BIOLOGICAL ACTIVITY PREDICTIONS 461

Table 1—Chemical name, formula and molecular weight of molecules (I–XVIII)

Molec Chemical name Chemical formula Molecular Reference ule no. weight (amu)

I Complex between 3α, 12α-dihydroxy-5β-cholan- C24H40O4 C2H5OH.H2O 456 24 24-oic acid (DCA), ethanol and water (3:2:1) II 3β-hydroxy-22-(4-methyl-1-pyrrolin-2-yl)-23-dinor- C27H41NO3 427 25 5α-cholane-4, 22-dione III Methyl 3α, 12α-dihydroxy-5β-cholan-24-oate-methanol C25H42O4.1/2CH3OH 422 26 IV 3α, 7α, 12α-trihydroxy-5β-cholan-24-oic acid C24H40O5 408 27 V 7α-hydroxy-4, 4, 8 trimethyl-21, 23-epoxy-24-nor- C26H32O4 418 28 5α, 13α, 17β-1, 14, 20, 22-tetraene-3, 16-dione VI Structure of methyl cholate with methanol (1:1) C25H42O5.CH3OH 454 29 VII Structure of methyl cholate with 2-propanol (1:1) C25H42O5 . C3H8O 482 29 VIII 23, 24-dinor-5α-cholan-12-one C22H36O 316 30 IX (+)-3-oxo-5α-cholan-24-oic acid C24H38O3 374 31 X Methyl 4β-bromo-7α-cathloxy-3-oxo-5β-5β-cholan- C23H43BrO6 495 32 22-oic acid. XI (+)-23, 24-dinor-3α, 9α-epoxy-11-oxo-androstadien 3-one C20H28O2.H2O 318 33 monohydrate XII (-)-3, 6-dioxo-5β-cholanic acid C24H36O.1/2 H2O 397 34 XIII 4, 4-dimethyl-23-phenyl-24-nor-5α-chola-8, 14-diene- C20H30O2 302 35 3β-ol XIV 4, 4-dimethyl-22-phenyl -23, 24-dinor-5α-chola-8, C30H42O 418 35 14-diene-3β-ol XV 4, 4-dimethyl-23-phenyl-22-oxa-24-nor-5α-chola-8, C30H42O2.1/2H2O 443 35 14- diene-3β-ol water XVI (20S)-methyl-3-oxochola-1, 4, 22-trien-24-oate C25H34O3 398 36 XVII 3β7α, 12α-triformloxy-24-nor-5β-chol-22-ene C26H38O6 446 37 XVIII (-3), 3, 7-dioxo-5β-cholanic acid C24H36O4 388 38

Table 2—Pa and Pi values for molecules (I-XVIII)

Molecule Teratogen Embryotoxic Carcinogenic Nitric oxide Drug-likeness Potassium no. Pa > Pi Pa > Pi Pa > Pi agonist channel Pa > Pi activator Pa > Pi

I 0.493>0.039 0.455>0.059 0.336>0.068 0.315>0.134 0.986 – II 0.482>0.043 0.403>0.072 0.297>0.167 – 0.992 0.166>0.164 III 0.496>0.038 0.427>0.065 0.250>0.133 0.368>0.071 0.993 – IV 0.486>0.042 0.461>0.058 0.332>0.070 0.259>0.256 0.987 – V 0.464>0.050 0.357>0.090 0.321>0.078 – 0.993 – VI 0.464>0.080 0.357>0.090 0.321>0.078 – 0.993 – VII 0.385>0.087 0.301>0.124 0.273>0.114 0.055>0.036 0.988 – VIII 0.425>0.068 0.429>0.065 0.283>0.105 0.406>0.047 – – IX 0.467>0.049 0.422>0.067 0.323>0.077 0.333>0.110 0.966 – X 0.355>0.104 0.262>0.171 0.305>0.089 0.278>0.207 0.782 – XI – – – 0.304>0.153 0.965 0.169>0.151 XII 0.479>0.044 0.380>0.045 0.401>0.073 0.323>0.123 0.965 – XIII 0.427>0.067 0.282>0.142 0.215>0.167 – 0.990 – XIV 0.424>0.064 0.292>0.146 0.218>0.169 – – – XV 0.420>0.070 0.308>0.087 0.300>0.125 – – – XVI 0.319>0.161 0.332>0.091 0.298>0.117 – 0.972 – XVII 0.289>0.150 0.285>0.139 – – 0.967 – XVIII 0.347>0.102 0.291>0.117 0.313>0.271 – 0.913 –

–, Absence of biological activity

462 INDIAN J. BIOCHEM. BIOPHYS., VOL. 44, DECEMBER 2007

Hydrogen bonding through solvent-solute/solute-solvent interactions, Based on the earlier studies on hydrogen (iii) know preference of linearity for C–H… 44-47 bonding , we have carried out a detailed analysis of O/O–H…O/C–H…Br hydrogen bonds, (iv) make a C–H…O, O–H…O, C–H…Br, etc. interactions in small compendium of hydrogen bonding on cholane series of steroids (molecules I-XVIII). The a comparative graphical scale, and (v) under- study has been carried out to (i) know whether intra or stand relationship between number of intermole- intermolecular C–H…O/O–H…O/C–H…Br bonding cular hydrogen bonds and drug likeness. Com- is dominant in this class of steroids and to reason out parative data of intra and intermolecular hydro- predominance of one type of interaction over another, gen bonds of the type C–H…O, O–H…O and (ii) examine role of hydrogen bonding in molecules C–H…Br in molecules (I-XVIII) are presented existing in aqueous solutions and crystal packing in Table 3.

Table 3—Geometry of C–H…O, O–H…O and C-H…Br intra and intermolecular interactions

Intramolecular hydrogen bonds Molecule X–H…A H…A (Å) X…A (Å) X–H…A (°) [No of donors and d D θ acceptors]

X C4 – H4…O4 2.300 3.002 128 Donors = 3 C6 – H6A…Br 2.740 3.285 116 Acceptors = 3 C27 – H27B…O5 2.240 2.686 107

XVI C4 – H4A…O7 2.350 3.029 126 Donors = 4 C12 – H12…O121 2.370 2.750 102 Acceptors = 3 C14 – H14…O12 2.530 2.929 104 C17 – H17…O12 2.430 2.890 108

Intermolecular hydrogen bonds I O27 – H(O27)…O25 1.770 2.590 154 Donors = 4 O25 – H(O25)…O28 1.820 2.790 160 Acceptors = 4 O26 – H(O26)... O(E) 1.790 2.710 158 O(E) – H(OE)... O26 1.780 2.690 162 [ II O3– H(O3)... O27 2.100 3.178 170 Donors = 1 Acceptors = 1

III O26 – H(O26)...O25´ 1.780 2.777 174 Donors = 3 O25´– H´(O25´)...O(M) 1.790 2.801 176 Acceptors = 4 O(M) – H(M)...O25 1.750 2.721 162 O25 – H(O25)...O26´ 1.810 2.812 171 O26´– H´(O26´)... O27´ 1.780 2.778 177

IV O28 – H(O28)…O26 1.680 2.634 172 Donors = 4 O26– H(O26)...O25 1.810 2.723 174 Acceptors = 4 O25– H(O25)…O29 1.880 2.774 171 O29 – H(O9)…O27 1.980 2.856 168

V O2 – H(O2)…O3 2.000 2.977 177 Donors = 2 C20–H20…O1 2.530 3.281 137 Acceptors = 2

VI O25 – H(O25)…O26 1.780 2.777 169 Donors = 4 O26 – H(O26)…O(M) 1.723 2.692 174 Acceptors = 4 O(M) – H(M)…O29 1.792 2.749 145 O29 – H(O29)…O25 1.830 2.847 165

VII O25 – H(O25)…O26 1.820 2.817 165 Donors = 4 O26 – H(O26)…O(P) 1.700 2.682 176 Acceptors = 4 O(P) – H(P)…O29 1.760 2.727 162 O29 – H(O29)…O25 1.830 2.913 173

IX O3 – H(O3)…O1 1.980 2.712 155 Donors = 1 Acceptors = 1

X C2 – H2B…O5 2.510 3.383 150 Donors = 4 C17 – H17…O1 2.450 3.382 158 Acceptors = 3 C23 – H23A…O1 2.600 3.493 153 C25 – H25A…O6 2.500 3.253 135 Contd RAJNIKANT et al.: BIOLOGICAL ACTIVITY PREDICTIONS 463

Intermolecular hydrogen bonds Molecule X–H…A H…A (Å) X…A (Å) X–H…A (°) [No of donors d D θ and acceptors]

XI O3 – H(O3)…O4 ´ 1.920 2.718 164 Donors = 4 O3´– H´(O3´)…O4 1.900 2.701 167 Acceptors = 4 C14´ – H14´B…O3 2.650 3.589 161 C8 – H8A…O1´ 2.610 3.366 134 C19´ – H19´B…O2´ 2.550 3.433 152 C14 – H14A…O3´ 2.650 3.601 165

XII O4 – H(O4)…O4 1.840 2.643 165 Donors = 7 O3 – H(O3)…O3 1.920 2.716 165 Acceptors = 5 O(W) – H1(OW)…O2 2.150 2.891 153 C4 – H4B…O(W) 2.670 3.600 160 C2 – H2A…O(W) 2.710 3.670 171 C5 – H5A…O1 2.670 3.574 154 C1 – H1B…O2 2.650 3.576 160

XIII O3 – HO3(2)…O3´ 2.060 2.890 143 Donors = 1 O3´– HO3´(1)…O3 1.960 2.890 164 Acceptors = 1

XIV O3 – HO3(1)…O3´ 2.010 2.855 146 Donors = 1 O3 – HO3(2)…O3´ 1.810 2.749 167 Acceptors = 1 O3´– HO3´(1)…O3 1.960 2.855 155 O3´– HO3´(2)…O3 1.850 2.749 154

XV O3´– HO3´(1)…O3´ 2.010 2.882 150 Donors = 2 O3´– HO3´(2)…O3 2.220 2.816 119 Acceptors = 2 O3 – HO3(1)…O(W) 2.010 2.831 143 O3 – HO3(1)…O(W) 2.010 2.831 143 O3 – HO3(2)…O3´ 2.000 2.816 142 O(W) – H2(OW)…O3 2.150 2.831 133

XVI C7-H7B…O1 2.220 2.882 154 Donors = 1 Acceptors = 1

XVII C31– H31…O31 2.490 3.396 166 Donors = 1 Acceptors = 1

XVIII C4-H4A…O2´´ 2.570 3.369 137 Donors = 9 C6-H6A...O4 2.620 3.345 130 Acceptors = 4 C11-H11A...O2 2.670 3.495 141 C16-H16A...O3´´ 2.600 3.459 145 C22-H22A...O3´´ 2.670 3.527 145 C2´´-H2’’A...O4´´ 2.650 3.297 123 C11´-H11´A...O2´ 2.580 3.095 113 C19´-H19´A...O1´´ 2.540 3.434 152 C6´´-H6’’B...O4´´ 2.660 3.223 117 C14´´-H14´´A...O2 2.640 3.593 161 O4-H(O4)...O1´ 1.870 2.755 180 O4´-H´(O4)...O3´´ 1.810 2.649 178 O4´´-H4´´(O4)...O3´ 1.820 2.659 178

´indicates second crystallographically independent molecule; ´´, third crystallographically independent molecule; (M), methanol molecule; (P), propanol molecule; (E), ethanol molecule; (W), water molecule; O(M), oxygen atom of the methanol molecule; O(P), oxygen atom of the propanol molecule; O(E), oxygen atom of the ethanol molecule; HO3(1) and HO3(2), hydrogen atom of the OH group attached at C3 position is disordered having two occupancies

Results and Discussion VIII, IX, whereas other molecules predict almost Biological-activity predictions neutral or low value of positive embryotoxic activity. A comparison of possible activities as given in The Pa > Pi values show that carcinogenic activity is Table 2 shows that all the molecules, except X and positive for molecule XII only. Similarly, molecule XVII possess high possibility of positive teratogen VIII has possibility of nitric oxide agonist activity. activity. Molecules X and XVII show almost neutral Only molecules II and XI act as neutral potassium teratogen activity. The high possibility of positive channel activator which may be attributed to the embryotoxic activity lies in molecules I, II, III, IV, orthodox position of oxygen atom O4 in molecule XI, 464 INDIAN J. BIOCHEM. BIOPHYS., VOL. 44, DECEMBER 2007

which are involved in intermolecular hydrogen depending upon whether C2–C3/C3–C4/C3–O is a bonding (Table 3). Thus, it depicts that unusual single or double bond. The bond distance C2(sp3)– substitution with cholane nucleus may change the C3(sp3) with substitution at C3 lies in the range 1.496- biological activity of the molecule. All the molecules 1.545 Å (average value 1.512 Å). The bond distance have shown high probability of drug-likeness, except C2(sp3)–C3(sp3) in molecules I (1.496 Å), VII (1.495 molecule X (0.782). This has been further authenti- Å) and XI' (1.496 Å) are significantly shorter than the cated by using Lipinski’s rule of five (Lipinski’s rule standard value of 1.533 Å49,50. The bond distance of thumb)48. Lipinski’s rule is used to evaluate drug- C3(sp3)–C4(sp3) with substitution at C3 position lies likeness or to determine, if a chemical compound with in the range 1.498-1.550 Å (average value 1.522 Å) a certain pharmacological or biological activity has and this bond distance in molecule XI (1.498 Å) is properties that would make it a likely active drug in smaller than the standard value (1.533 Å). humans. It predicts that the compounds undertaken The deviation of bond distances C2(sp3)–C3(sp3) for this study are important for drug development. and C3(sp3)–C4(sp3) in the said molecules could be

Selected bond distances and angles due to the effect of hydroxyl group located at C3 Most of the molecules undertaken in the present position which invariably is involved in O–H…O study have substituents (hydroxy, keto, acetoxy, etc.) intermolecular interactions. The C3(sp3)–O bond at C3 position of steroid nucleus. Therefore, it is of distance in molecules having substitution at C3 lies in interest to investigate C2–C3, C3–C4 and C3–O bond the range 1.419-1.468 Å (average value 1.442 Å), distances and the C2–C3–C4 bond angle (Table 4). whereas the bond distance C3(sp2)=O ranges from The substitution at C3 position of ring A of steroidal 1.204-1.219 Å (average value 1.214 Å). The C3(sp3)– nucleus causes visible change in the bond distances, O bond distance in molecules I (1.447 Å), III (1.447 Table 4—Some selected bond distances C2–C3, C3–C4 and C3–O (Å) and C2–C3–C4 bond angle (°) for molecules I–XVIII

Molecule no. Bond distance C2–C3 Bond distance C3–C4 Bond Distance C3–O Bond angle C2–C3–C4 sp3–sp3 sp3–sp2 sp2–sp2 sp3–sp3 sp3–sp2 C(sp3)–O C(sp2)=O C3(sp3) C3(sp2) sp2–sp3 sp2–sp3

I 1.496 - - 1.543 - 1.447 - 108.0 - II 1.512 - - - 1.512 1.419 - 109.2 - III 1.523 - - 1.508 - 1.447 - 109.5 - III´ 1.545 - - 1.508 - 1.443 - 112.1 - IV 1.519 - - 1.513 - 1.443 - 109.7 - V - - 1.468 - 1.522 - 1.215 - 118.6 VI 1.510 - - 1.506 - 1.436 - 110.7 - VII 1.495 - - 1.515 - 1.444 - 109.5 - VIII 1.541 - - 1.518 - - - 111.6 - IX - 1.492 - - 1.493 - 1.204 - 116.5 X - 1.497 - - 1.517 - 1.212 - 114.5 XI 1.507 - - 1.498 - 1.458 - 111.1 - XI´ 1.496 - - 1.513 - 1.453 - 111.5 - XII - 1.488 - - 1.507 - 1.214 - 115.1 XIII 1.507 - - 1.513 - 1.438 - 114.1 - XIII´ 1.501 - - 1.526 - 1.434 - 113.3 - XIV 1.528 - - 1.550 - 1.440 - 112.5 - XIV´ 1.527 - - 1.542 - 1.442 - 112.2 - XV 1.512 - - 1.545 - 1.437 - 113.6 - XV´ 1.506 - - 1.532 - 1.449 - 113.0 - XVI - - 1.455 - - 1.439 1.210 - 116.7 XVII 1.507 - - 1.512 - 1.468 - 110.9 - XVIII - 1.497 - - 1.507 - 1.219 - 114.6 XVIII´ - 1.496 - - 1.498 - 1.219 - 114.7 XVIII´´ - 1.499 - - 1.504 - 1.216 - 114.8

´, indicates second crystallographically independent molecule; ´´, third crystallographically independent molecule; -, absence of a particular type of bond/angle RAJNIKANT et al.: BIOLOGICAL ACTIVITY PREDICTIONS 465

Å), VII (1.445 Å), XI (1.458 Å), XI' (1.453 Å), XV' (1.449 Å) and XVII (1.468 Å) shows significant deviation from the standard value (1.426 Å)50 . The deviation of this bond distance in molecules I, III, VII and XV' could be due to involvement of OH group in O–H…O intermolecular interactions, whereas involvement of epoxy group between C3 and C9 in O–H…O intermolecular interactions causes deviation 3 Fig. 3—Frequency of occurrence of (a): C–H…O and C–H…Br of bond distance C3(sp )–O in molecules XI and XI'. intramolecular hydrogen bonds; (b): C–H…O and O–H…O The involvement of formoloxy group in C–H…O intermolecular hydrogen bonds; and (c): C–H…O, O–H…O and intermolecular interaction causes deviation in the said C–H…Br intra/intermolecular hydrogen bonds bond distance in molecule XVII. The substitution of a group at C3 position also (frequency of occurrence 57.1%), whereas in other causes a significant change in the value of bond angle series of steroids (cholestane and pregnane), C2–C3–C4 in ring A, depending upon whether C3 is predominance of C–H…O interactions has been (sp3) or (sp2) hybridized. The bond angle C2–C3–C4 observed23,51. with C3(sp3) in molecules with substituent at C3 position varies from 108.0-114.1 [average value d–θ and D–θ Scatter plots 111.3 ]. The bond angle C2–C3–C4 with C3(sp3) in The main structural feature distinguishing hydrogen bond from other non-covalent interactions molecules XIII (114.1°), XIII' (113.3°) and XV 52 (113.6°) shows significant deviation from the is the preference for linearity , which can be tetrahedral value of 109.46°. The deviation in C2–C3– analyzed in a better way by drawing the d–θ and C4 bond angle in these molecules causes O–H…O D–θ scatter plots, where ‘d’ represents distance intermolecular interactions and is probably due to the between hydrogen atom ‘H’ and acceptor atom ‘A’, presence of OH group located at C3 position. The ‘D’ distance between donor atom ‘X’ and acceptor value of bond angle C2–C3–C4 with C3(sp2) in atom ‘A’ and ‘θ’ angular disposition between ‘X’ and molecules with substitutent at C3 position varies from ‘A’ at ‘H’. The plots include all contacts found in 114.5 to 118.6° [average value 116.2°]. The bond molecules I-XVIII with d <2.75 Å and D <3.68 Å at angle C2–C3–C4 with C3(sp2) in molecules X any occurring angle. The graphical projection of (114.5°), XII (115.1°), XVIII (114.6°), XVIII' d(H…A) against θ(X-H…A) and D(X…A) against (114.7°) and XVIII'' (114.8°) shows some deviation θ(X–H…A) i.e. d–θ and D–θ scatter plots have been from the value of 120° for (sp2) type of hybridization. made for intramolecular hydrogen bonds as shown in The presence of a keto group deviates the bond angle Fig. 4a and b, respectively. The d–θ and D–θ scatter C2–C3–C4 significantly and could be due to the plots have also been made for intermolecular occurrence of C–H…O/O–H…O intermolecular hydrogen bonds (Fig. 5a and b, respectively). interactions. In case of intermolecular hydrogen bonds, density of spots [for d(H…A) = 2.50-2.68 Å and D(X…A) = Hydrogen bonding 3.4-3.6 Å] is predominant in the θ(X–H…A) range Based on the values as given in Table 3, frequency ~150-165° for C–H…O hydrogen bonds. For of occurrence of various types of intra and O–H…O type of hydrogen bond, density of spots is intermolecular hydrogen bonds has been analyzed by greater in the d(H…A) Å range 1.75-2.0 Å and drawing the histogram charts (Fig. 3). From Fig. 3, it D(X…A) Å range 2.68-2.90 Å for θ(X–H…A)° in the is observed that (i) out of total number of range ~155-180°. There are densely populated intramolecular hydrogen bonds formed, frequency of clusters of data points at short distances and fairly occurrence of C–H…O and C–H…Br intramolecular linear angles and each point in these clusters hydrogen bonds is 85.7% and 14.3%, respectively, (ii) represents a hydrogen bond. The cluster of frequency of occurrence of C–H…O and O–H…O O-H…O data points around d(1.8 Å) and intermolecular hydrogen bonds is 37.1% and 62.9%, θ(164-177°), whereas C-H...O points cluster around respectively, and (iii) on basis of both intra as well as d(2.65 Å) and θ(145-165°). These plots indicate that intermolecular hydrogen bonds, it may be concluded the angular characteristics of all kinds of hydrogen that O–H…O hydrogen bonding is predominant 466 INDIAN J. BIOCHEM. BIOPHYS., VOL. 44, DECEMBER 2007

Fig. 4—(a): d–θ scatter plot for intramolecular C–H…O and C– H…Br hydrogen bonds; and(b): D–θ scatter plot for intramolecular C–H…O and C–H…Br hydrogen bonds

Fig. 6—(a): Bifurcated donor atom (X); (b) Bifurcated acceptor atom (A); (c) Bifurcated donor/acceptor atom (A); and (d): Distribution of bifurcated H-bond angle (Ψ) in bifurcated hydrogen bonds

a pronounced softness of the angle at H, donors may interact simultaneously with more than one acceptor (Fig. 6a). In Fig. 6a, X atom is referred to as Fig. 5—(a): d–θ scatter plot for intermolecular C–H…O and O– bifurcated donor atom and bond as bifurcated H…O hydrogen bonds; and (b): D–θ scatter plot for hydrogen bond or three-center hydrogen bond intermolecular C–H…O and O–H…O hydrogen bonds X–H…(A1, A2). Sometimes, acceptor may interact simultaneously with more than one donor (Fig. 6b). In bonds are related. Fig. 6b, A atom is referred to as a bifurcated acceptor From analysis of nearest neighbour contacts of atom and bond as bifurcated hydrogen bond each H-atom bonded to C or O atom in intra as well (X1, X2)–H...A. It may happen sometimes that a atom as intermolecular interactions, the following may act as donor as well as acceptor simultaneously observations have been made: (i) from frequency of (Fig. 6c). Here atom A is referred to as a bifurcated occurrence of contacts from H(C) to O and Br vis-a- donor/acceptor. Bifurcated hydrogen bonds are vis their crystal structures, it is observed that H(C) characterized by distances d1, d2 and angles θ1 + θ2. atoms have a statistical preference for contacts to ‘Br’ In a multi-furcated hydrogen bond, a donor forms rather than ‘O’ atoms in molecules 1–XVIII, (ii) hydrogen bonds with more than one acceptor. Multi- comparison of contacts from H(O) to O, N, etc. with furcated hydrogen bonding requires a high density of stoichiometry of crystal structures suggests the H(O) acceptors. Over 25% of hydrogen bonds in atoms have a statistical preference for contacts to ‘O’ carbohydrates are multi-furcated and this fraction is rather than any other atom, (iii) almost all C–H…O even higher in amino acids45,52,54. Proteins also contacts have distance d(H…O) less than 2.7 Å and contain multi-furcated hydrogen bonds in large based on the criterion that van der Waals distance numbers. Bifurcated or three-centre hydrogen bonds should be <2.7 Å, it has been regarded as a certain X–H…(A1, A2) are very common in O–H…O and indication of hydrogen bonding, and (iv) geometrical N–H…O hydrogen bonded structures55. Bifurcated characteristics of C–H…O contacts are not similar to hydrogen bonds are also observed in C–H…O/N those of O–H…O hydrogen bonds as C–H…O patterns. In the present study, most of the bifurcated contacts are less linear than the O–H…O contacts. It hydrogen bonds are found in O–H…O pattern, in is mostly through a consideration of the effects of which O acts as a donor as well as acceptor. C–H…O and O–H…O hydrogen bonds on the crystal packing that a relaxation of the distance criteria Intramolecular bifurcated hydrogen bond exists becomes necessary50,53. only in molecule XVII. In XVII, oxygen atom O12 of formyloxy group attached at C12 position is involved Bifurcated hydrogen bonds in two intramolecular C–H…O hydrogen bonds Due to long-range character of hydrogen bond and [C14–H14…O12; C17–H17…O12 with bifurcated RAJNIKANT et al.: BIOLOGICAL ACTIVITY PREDICTIONS 467

angle of 212°]. In molecule I, oxygen atom O25 …O(M) with bifurcated angle of 319° in molecule attached at C3 position is involved in bifurcated VI] and [O(P)–H(P)…O29; O26–H(O26)…O(P) with hydrogen bond and acts as a donor as well as bifurcated angle of 338° in molecule VII]. acceptor, forming two intermolecular O–H…O H- In molecule X, oxygen atom of keto group bonds [O25–H(O25)…O28; O27–H(O27)…O25 with located at C3 position acts as a bifurcated acceptor bifurcated angle of 225°]. In molecule III, asymmetric [C17–H17…O1; C23–H23A…O1 with bifurcated unit has two independent molecules and one methanol angle of 311°]. In molecule XV, oxygen atom O3 of molecule. The oxygen atom O(M) of methanol the OH group located at C3 position acts as a molecule acts as a donor as well as acceptor, forming bifurcated hydrogen donor [O3–HO3(1)….O(W); two intermolecular O–H…O H-bonds [O(M)– O3–HO3(1)….O(W) with bifurcated angle of 286°]. H(M)…O25; O25'–H'(O25')…O(M) with bifurcated In molecule XVIII, there are three steroidal molecules angle of 338°]. In the same molecule, oxygen atoms in the asymmetric unit. Here, oxygen atom O3 (of the O25 and O25' of hydroxyl group located at C3 and third independent molecule) of keto group located at C3' positions are involved in bifurcated hydrogen C24 position acts as bifurcated acceptor forming two bond in which O25 and O25' act as donor as well as intermolecular C–H...O hydrogen bonds with atoms acceptor [O25–H(O25)…O26'; O(M)–H(M)…O25 C16 and C22, respectively [C16–H16A…O3''; C22– and O25'–H' (O25)…O(M); O26–H(O26)…O25' with H22A…O3'' with bifurcated angle of 290°]. In the bifurcated angles of 333° and 350°, respectively]. same molecule, oxygen atom O4 of hydroxy group The oxygen atom O26' of hydroxyl group located acts as bifurcated acceptor, forming two intermole- at C12' position in the second independent molecule cular C–H...O hydrogen bonds [C1''–H1''A…O4''; III' is involved in bifurcated hydrogen bond in which C2''–H2''A…O4'' with bifurcated angle of 229°]. atom O26' acts as a donor as well as acceptor [O26'– In all the above-mentioned bifurcated hydrogen H'(O26')…O27'; O25–H(O25)…O26' with bifurcated bonds, the atom O acts as a prototype acceptor as well angle of 348°]. In molecule IV, oxygen atoms O26, as donor. Therefore, it may be concluded that in the O25 and O29 of the hydroxyl groups located at C12, cholane derivatives, existence of bifurcated hydrogen C3 and C7 positions, respectively are involved in bonds in O–H…O pattern is most predominant. The bifurcated hydrogen bonds in which O26, O25 and tendency towards linearity for three center hydrogen bonds stems from the reason of the electrostatic O29 act as donors as well as acceptors [O26– 56 55 H(O26)…O25; O28–H(O28)…O26 with bifurcated nature of all hydrogen bonds . Jeffery and Saenger angle of 346°], [O25–H(O25)…O29; O26– characterized the bifurcated hydrogen bonds by H(O26)…O25 with bifurcated angle of 345°] and distances and angles. In the present case, bifurcated [O29–H(O29)… O27; O25–H(O25)…O29 with H-bond angle (Ψ = θ1 + θ2) has been calculated for bifurcated angle of 339°]. each bifurcated donor and acceptor atom and disrtibution of Ψ angles in crystal structures of In molecules VI and VII, oxygen atoms O26 and cholane derivatives is shown in Fig. 6d. A detailed O29 of hydroxyl groups located at C12 and C7 analysis of Fig. 6d explicitly indicates that only one positions, respectively are involved in bifurcated intramolecular bifurcated H-bond exists which falls in hydrogen bonds in which O26 and O29 act as donors the Ψ range of 200-220°, two intermolecular H-bonds as well as acceptors [O26–H(O26)…O(M); O25– fall in the Ψ range of 220-240°, whereas all other H(O25)…O26 with bifurcated angle of 343° in intermolecular bifurcated H-bonds fall in the range of molecule VI], [O29–H(O29)…O25; O(M) – 280-360°. The maximum tendency of occurrence of H(M)…O29 with bifurcated angle of 310° in intermolecular bifurcated H-bonds is observed in the molecule VI], [O26–H(O26)…O(P); O25– Ψ range of 340-360° (31.5%). H(O25)…O26 with bifurcated angle of 341° in molecule VII], [O29–H(O29)… O25; O(P)– Solvent/solute interactions H(P)…O29 with bifurcated angle of 335° in molecule The effects of solvent on the properties of organic VII]. The oxygen atoms O(M) and O(P) of methanol and biological molecules have been successfully and propanol in molecules VI and VII, respectively described by using different and complementary are involved in bifurcated hydrogen bonds in which theoretical models57-59. In this direction, investigation O(M) and O(P) act as donors as well as acceptors carried out60 on solvation mechanism and the specific simultaneously. [O(M)–H(M)…O29; O26–H(O26)- role of solute-solvent interactions could be used as a 468 INDIAN J. BIOCHEM. BIOPHYS., VOL. 44, DECEMBER 2007

tool for supramolecular structures61,62. The specific biomolecules. In addition, receptors for 3-keto C–H…O hydrogen bonds between solute and solvent cholane derivatives have a conserved glutamine play an important role in solid-state chemistry63. It has which allows accommodating 3-hydroxy groups of been suggested that C–H…O hydrogen bonds are the aromatic ring-A steroid and its analog67. present in the liquid along with the stronger O–H…O Superpositions of 3-keto and 3-hydroxy steroids using bonds and both hydrogen bonds play an important computer programs like SEAL68 indicate that these role in stabilizing the extended structures64. This study steroids may bind to globulin in different modes. is likely to be of importance, because it strongly These results clearly indicate the importance of suggests that common feature in structures of two hydrogen bonding potentials in molecular- condensed phases is the simultaneous presence of superposition programs. O–H…O and C–H…O hydrogen bonds65. In solu- In drug design, ligands are also recognized by tions, anions and cations interact with one another via properties apart from molecular structure, hence weak hydrogen bonds in a highly-ordered manner. similar interactions with functional groups of the This self-organization results in a definite supramole- binding site have similar affinities, even if their cular identity in aqueous solution66. chemical structures are quite different69. In this paper, The presence of solvent in some cholane deriva- most of the probable activities are characterized by Pa tives leads to the formation of solute-solvent/solvent- and Pi values which depict that all the molecules have solute intermolecular interactions. The solute-solvent high value of teratogen activity and also have stronger interactions viz., [O26–H(O26)…O29(E)] for preponderance for “cancer-like-drug” molecules. molecule I, [O25′–H′(O25′)…O(M)] for molecule III, According to Lipinski’s rule (a good permeation or [O26–H(O26)…O(M)] for molecule VI, [O26– absorption is more likely when there are less than 5H- H(O26)…O (M)] for molecule VII, [C4– bond donors, 10H-bond acceptors, qualifying H4B…O(W); C2–H2B…O(W)] for molecule XII and molecular weight range is 160-480, etc), most of [O3–H3O(1)…O(W); O3–H3O(2) …O(W)] for cholane derivatives are drug-like molecules which molecule XV have been observed. The solvent-solute could be helpful for drug design as well as in the interactions viz, [O29(E)–H(O29)…O26], [O(M)– development of advances in the field of biophysics. H(M)…O25], [O(M)–H(M)….O29], [O(P)–H(P)… O29], [O(W)–H1(OW)…O2], [O(W)–H2(OW)…O3] Acknowledgement have been observed for molecules I, III, VI, VII, XII The author (Rajnikant) is grateful to Science and and XV, respectively. In all the above descri- Engineering Research Council of the Department of bed solute-solvent/solvent-solute interactions, only Science and Technology, Govt. of India for funding O–H…O type of intermolecular hydrogen bonds are under a sponsored Research project (No. SR/S2/CMP- involved. 47/2003).

Concluding remarks References DNA and proteins are held together in their defined 1 Makin (1975) Biochemistry of Steroid Hormones, Ist edn, 3-D structures by hydrogen bonds. The hydrogen Chapt. 1, pp. 5, Blackwell Scientific Publications, Oxford bonding also affects membrane transport as well as 2 Wilson J D (1972) Arch Inter Med 130, 493 distribution of drugs within the biological system. In 3 Makin (1975) Biochemistry of Steroid Hormones, Ist edn, the present study, an attempt has been made to Chapt. 2, pp. 42, Blackwell Scientific Publications, Oxford 4 Holt P R (1972) Arch Inter Med 130, 574 understand influence of hydrogen bonding on ligand 5 Hoffman A F & Mekjian H S (1972) In: The Bile Acids (Nair affinities. Steroid-binding proteins have high speci- P P & Kritchevsky D, eds), Vol. 2, pp. 103, Plenum Press, ficity of drugs and transporters and all steroid NewYork receptors have a conserved arginine in the binding site 6 Dietschy J M (1968) J Lipid Res 19, 297 67 7 Lack L & Weiner I M (1973) In: The Bile Acids (Nair P P & for the steroid ring A . The crystallographic analysis Kritchevsky D, eds), Vol. 2, pp. 33, Plenum Press, New York comprising some geometrical and structural features 8 Small D M, Dowling R H & Redinger (1972) Arch Inter Med for cholane derivatives has shown that substituents are 130, 533 mostly located at the C3 position of the steroid 9 Sobotaka H (1934) Chem Rev 15, 311-375 nucleus. These substituents are linked by intra or 10 Feiser L F & Feiser M (1959) Steroids Chemistry, pp. 115- 118, Reinhold, New York intermolecular hydrogen bonds, which help in 11 Herndon W C (1967) J Chem Educ 44, 724-728 understanding the stacking interactions in RAJNIKANT et al.: BIOLOGICAL ACTIVITY PREDICTIONS 469

12 Hadler S C, Smart J E & Bonner J (1971) Biochim Biophy Press, Moscow Acta 236, 253-258 41 Suchkov A P, Filimonov D A, Stepanchikova A V & 13 Phillippot J (1971) Biochim Biophy Acta 225, 201-213 Poroikov V V (2001) Environ Res, 12, 327-344 14 Smart J E & Bonner J (1971) J Mol Biol 58, 651-659 42 Anzali S, Barnickel, G, Cezanne B & Krug M (2001) J Med 15 Visser L, Robinson N C & Tanford C (1975) Biochemistry Chem 44, 2432-2437 14, 1194-1199 43 Poroikov V V, Filimonov D A, Ihlenfeldt W D, Gloriozova T 16 Corsi E, D’Alagni M & Giglio E (1976) Polymer 20, A, Lagunin A A, Borodina Y V, Stepanchikova A V & 259-260 Nicklaus M C (2003) J Chem Inf Comput Sci 43, 228. 17 Davis A P (1993) Chem Soc Rev 22, 243-253 44 Steiner T (1996) Cryst Rev 6, 1-57 18 Miki K, Kasai N, Shibakami M, Takemoto K & Miyata M 45 Jeffery G A (1997) An introduction to Hydrogen Bonding, (1991) J Chem Soc Chem Commun 14, 450-460 pp. 400-408, Oxford University Press, New York 19 Miyata M, Shibakami M, Chirachanchai S Takemoto K, 46 Steiner T (1998) Acta Cryst B54, 456-463 Kasai N & Miki K (1990) Nature (London) 343, 446-447 47 Desiraju G R & Steiner T (1999) The Weak Hydrogen Bond 20 Singh A, Gupta V K, Rajnikant & Goswami K N (1994) in Structural Chemistry and Biology, IUCr/Oxford Cryst Res Technol 29, 837-842 University Press 21 Singh A, Gupta V K, Rajnikant, Goswami K N, Gupta B D 48 Lipinski C A, Lombardo F, Dominy B W & Feeney P J & Banerjee S K (1996) Mol Mat 6, 53-60 (1997) Adv Drug Delivery Rev 23, 3-25 22 Rajnikant, Dinesh, Anshu & Mousmi (2004) Cryst Res 49 Sutton L E (1965). Tables of Interatomic Distances and Technol 39, 353-358 Configuration in Molecules and Ions, Special Publication 23 Rajnikant, Dinesh & Bhavnaish Chand (2006) Acta Cryst no.18, The Chemical Society, London A62, 136-145 50 Allen F H, Kennard O, Watson D G, Bramer L, Orpen A G 24 Candeloro S, Sanctis De, Coiro V M, Giglio E, Pagliuca S, & Taylor R (1987) J Chem Soc Perkin Trans 2, S1-19 Pauel N V & Quagliata C (1978) Acta Cryst B34, 1982-1983 51 Rajnikant, Dinesh & Bhavnaish Chand (2006) Acta Cryst 25 Usubilliga A, Zabel, V & Watson W H (1982) Acta Cryst B62, 90-101 B38, 966-969 52 Steiner T (2002) Angew Chem Int Ed Eng 41, 48-50 26 Miki M, Masui A, Kasai N, Miyata M, Goonewardena W, 53 Desiraju G R (1991) Acc Chem Res 24, 270-276 Shibakami M & Takemoto K (1989) Acta Cryst C45, 79-83 54 Preibner R, Egner U & Saenger (1991) WFEBS Lett 288, 27 Miki M, Kasai N, Shibakami M, Chirachanchai S, Takemoto 192-196 K & Miyata M (1990) Acta Cryst C46, 2442-2445 55 Jeffery G A & Saenger W (1991) Hydrogen Bonding in 28 Das K, Sinha U C, Mayelvaganan T, Bhat S V & Tavale S S Biological Structures, pp. 64-65, Springer-Verlag, Berlin (1991) Acta Cryst C47, 1426-1429 56 Desiraju G R & Steiner T (1999) The Weak Hydrogen Bond 29 Miki M, Masui A, Kasai N, Goonewardena W, Shibakami M in Structural Chemistry and Biology, pp. 66-67, IUCr/Oxford & Takemoto K & Miyata M (1992) Acta Cryst C48, 503-507 University Press 30 Decanniere K, Maes D, Lisgarten J N, Zegers I & Beisemans 57 Kollman P (1993) Chem Rev 93, 2395-2417 M (1993) Acta Cryst C49, 1824-1826 58 Cramer C J & Truhlar D G (1999) Chem Rev 99, 2161-2200 31 Brunskill A P J, Lalancette R A & Thompson H W (1999) 59 Baldridge K K, Jonas V & Bain A D (2000) J Chem Phys Acta Cryst C55, 419-422 113, 7519-7529 32 Armas H N D, Blaton N M, Peters O M, Ranter C J D, 60 Allen M P & Tildesley D J (1987) Computer Simulation of Hernandez R P, Manchado F C & Gil R P (1999) Acta Cryst Liquid, pp. 93-94, Oxford University Press, New York C55, 1677-1680 33 Thompson H W, Brunskill A P J & Lalancette R A (1999) 61 Coutinho K, Canuto S & Zerner M C (2000) J Chem Phys Acta Cryst C55, 1677-1680 112, 9874-9880 34 Thompson H W, Lalancette R A & Brunskill A P J (2001) 62 Canuto S, Coutinho K & Trzesniak D (2002) Adv Quantum Acta Cryst C57, 632-633 Chem 41, 161-167 35 Boer D R, Kooijman H, Groen M, Louw J V Kelder J & 63 Desiraju G R & Steiner T (1999) The Weak Hydrogen Bond Kroon J (2001) Acta Cryst C57, 1177-1181 in Structural Chemistry and Biology, pp. 116-117, IUCr/ 36 Linker M, Schiirmann M, Preut H & Kreiser W (2001) Acta Oxford University Press Cryst C57, 574-574 64 Jedlovszky P & Turi L (1997) J Phys Chem B101, 37 Andrade L C R, Paixao J A, Almedia M J M D, Silva E J T 5429-5436 D, Melo M L S & Roleira F M F (2004) Acta Cryst C60, 65 Tezuka T, Nakagawa M, Yokoi K, Nagawa Y, Yamagaki T 82-83 & Nakanishi H (1997) Tetrahedron Lett 38, 4223-4226 38 Kikolski E M, Lalancette R A & Thompson H W (1999) 66 Davidson M G & Lamb S (1997) Polyhedron 16, 4393-4395 Acta Cryst C62, 394-396 67 Williams S P & Sigler P B (1998) Nature (London) 393, 39 Filimonov D A, Poroikov V, Borodina, Yu & Gloriozova T 392-304 (1999) J Chem Inf Comput Sci 39, 666-670 68 Klebe G, Mietzner T & Weber F (1999) J Comput–Aided 40 Poroikov V V & Filimonov D A (2001) Computer-assisted Mol Design 13, 35-47 Predictions of Biological Activity in Search for and 69 Chen J M, Xu S L, Wawrzak Z, Basarab G S & Jordan D B Optimization of New drugs, Vol. 1, pp. 149-154, Iridium (1998) Biochemistry 17735-17743