Molecular evolution of snake toxins: is the functional diversity of snake toxins associated with a mechanism of accelerated evolution?

Recent studies revealed that animal toxins with unrelated biological functions often possess a similar architecture. To tentatively understand the evolutionary mechanisms that may govern this principle of functional prodigality associated with a structural economy, two complementary approaches were considered. One of them consisted of investigating the rates of mutations that occur in cDNAs and/or genes that encode a variety of toxins with the same fold. This approach was largely adopted with phospholipases A2 from Viperidae and to a lesser extent with three-fingered toxins from Elapidae and Hydrophiidae. Another approach consisted of investigating how a given fold can accommodate distinct functional topographies. Thus, a number of topologies by which three-fingered toxins exert distinct functions were investigated either by making chemical modifications and/or mutational analyses or by studying the three-dimensional structure of toxin-target complexes. This review shows that, although the two approaches are different, they commonly indicate that most if not all the surface of a snake toxin fold undergoes natural engineering, which may be associated with an accelerated rate of evolution. The biochemical process by which this phenomenon occurs remains unknown.

Two-step models to predict binding affinity of chemicals to the human estrogen receptor alpha by three-dimensional quantitative structure-activity relationships (3D-QSARs) using receptor-ligand docking simulation.

Binding of chemicals to the estrogen receptor (ER) is known to be a key mode of action of endocrine disruption effects. In this study, combined quantitative structure-activity relationship (QSAR) models from discriminant and multilinear regression (MLR) analyses, termed a two-step model, were developed. These were used to predict the binding potency to human ERalpha of four chemical groups, namely alkylphenols, phthalates, diphenylethanes and benzophenones. These groups are considered to be important chemical classes of ER-binders. The descriptors investigated were calculated following the simulation of docking between the receptor and ligand. Discriminant analysis in the first step of a two-step model was applied to distinguish binders from non-binders. It had a concordance, following leave-one-out (LOO), of greater than 87% for all chemical classes. Binders were defined as chemicals whose IC50 was reliably measured in a competitive binding assay. The MLR analysis in the second step was performed for the quantitative prediction of the binding affinity of chemicals that were previously discriminated as binders. The q2 values for alkylphenols and diphenylethanes were 0.75 and 0.74, respectively. However good MLR relationships were not obtained for phthalates and benzophenones as the observed binding affinities of chemicals in these categories were weak and in a too narrow range.

Enhanced follicle-stimulating hormone activity of deglycosylated human chorionic gonadotropin in ovarian granulosa cells.

The receptor binding properties and biological actions of chemically deglycosylated asialo human CG (AHF-hCG) were studied in ovarian granulosa cells from diethylstilbestrol (DES)-treated immature rats. In ovarian homogenates from DES- and FSH-treated rats, the relative binding affinity of AHF-hCG was 2-fold higher than that of native hCG and 14-fold higher than that of ovine LH. When assayed for LH-like activity in granulosa cells from DES plus FSH-treated animals, the deglycosylated hormone behaved as a partial agonist in terms of cAMP formation, but fully stimulated progesterone production to the same extent as that elicited by LH. When added with LH to FSH-treated cells, AHF-hCG inhibited LH-stimulated cAMP formation by 70% but did not alter the elevated level of progesterone production. These findings are consistent with the presence of excess or spare LH receptors in the maturing granulosa cell. When added to freshly prepared granulosa cells which have minimal LH receptors, AHF-hCG decreased FSH-stimulated cAMP production by 20% and reduced progesterone production by 50% and increased cGMP formation by 100% during 48 h of culture. The ability of AHF-hCG to decrease the progesterone response to FSH suggests that no spare FSH receptors are present during granulosa cell differentiation. In contrast, native hCG did not alter FSH-induced cAMP or progesterone production but reduced the cGMP responses to FSH and choleragen. Whereas native hCG displayed negligible binding potency when compared with that of ovine FSH in competition with [125I]iodo-human FSH for ovarian receptors, AHF-hCG bound to FSH receptors with about 5% of the binding affinity of ovine FSH. In choleragen-treated granulosa cells, the increases in cAMP and progesterone synthesis were enhanced by addition of both hCG and AHF-hCG, and cGMP production was increased by AHF-hCG but slightly decreased by hCG. These results indicate that the enhanced LH receptor affinity caused by removal of the sugar moieties from hCG is accompanied by a relatively greater increase in FSH receptor affinity, and that deglycosylated hCG acts as a partial agonist with the ability to modify granulosa cell responses to both LH and FSH.

Structures of genes encoding phospholipase A2 inhibitors from the serum of Trimeresurus flavoviridis snake.

Inhibitors (PLIs) against snake venom gland phospholipases A2 (PLA2s) have been found in their sera. A cDNA encoding a PLI from Trimeresurus flavoviridis (Tf, habu snake, Crotalinae) serum, cPLI-A, was isolated from the Tf liver cDNA library and sequenced. Northern blot analysis with cPLI-A showed that PLIs are expressed only in liver. Genes for PLIs, gPLI-A and gPLI-B, were isolated from the Tf genomic DNA library and their nucleotide (nt) sequences were determined. The genes consisted of four exons and three introns, and exon 4 encoded the carbohydrate recognition domain (CRD)-like motif. Comparison of the nt sequences between gPLI-A and gPLI-B showed that these genes are highly homologous, including introns, except that exon 3 is rich in nonsynonymous nt substitutions which are almost four times as frequent as synonymous nt substitutions. This evolutionary feature of PLI genes is different from that of venom gland PLA2 isozyme genes in which nonsynonymous nt substitutions are spread over the entire mature protein-coding region.

Accelerated evolution of Trimeresurus okinavensis venom gland phospholipase A2 isozyme-encoding genes.

Three Trimeresurus okinavensis (To; himehabu snake, Crotalinae) venom gland phospholipase A2 (PLA2) isozymeencoding genes, gPLA2-o1, gPLA2-o2 and gPLA2-o3, were isolated from its genomic DNA library. The nucleotide (nt) sequence analysis revealed that two of the three genes (gPLA2-o2 and gPLA2-o3) occasionally have been converted to inactivated genes by introduction of one base insertion or substitution. It was confirmed from Southern blot analysis that the To haploid genome contains only three venom gland PLA2 isozyme genes herein isolated. Comparison of these genes showed that nonsynonymous nt substitutions have occurred more frequently than synonymous nt substitutions in the protein-coding regions, except for the signal-peptide coding domain, implying that To venom gland PLA2 isozyme genes have evolved via accelerated evolution. Such an evolutionary feature of To venom gland PLA2 isozyme genes proves the general universality of accelerated evolution previously drawn for venom gland PLA2 isozyme genes of other crotalinae snakes. The variability in the mature protein-coding regions of three To venom gland PLA2 isozyme genes appears to have been brought about by natural selection for point mutations.

Structures of genes encoding TATA box-binding proteins from Trimeresurus gramineus and T. flavoviridis snakes.

A cDNA encoding the Trimeresurus gramineus (Tg; green habu snake) TATA-box-binding protein (TgTBP) was cloned and sequenced. The cDNA encodes a 33-kDa protein with an extensive sequence similarity to those derived from other organisms, except for the N-terminal domain. Genes encoding TgTBP and Trimeresurus flavoviridis (Tf; habu snake) TBP (TfTBP) were isolated using a TgTBP cDNA and their nt sequences were determined. They are the first TBP genes entirely sequenced in higher animals. Both genes span over 15 kb and are constructed from eight exons and seven introns. Comparison of the loci of introns on the aligned amino-acid sequences of TBP from six organisms (Tg, Tf, mouse, Arabidopsis thaliana, Schizosaccharomyces pombe and Acanthamoeba castellanii) indicated that there are three highly conserved loci in the C-terminal domain.

A highly selective ligand for brain delta opiate receptors, a cyclopropyl(E)Phe(4)-enkephalin analog, suppresses mu receptor-mediated thermal analgesia by morphine.

[D-Ala(2)(2R,3S)-delta(E)Phe(4)Leu(5)]enkephalin (CP-OH) [delta denoting cyclopropyl; superscript E indicating the E-configuration about the cyclopropane ring], a highly selective opioid ligand for delta receptors in rat brain, but not for those in the mouse vas deferens, was examined for in vivo biological activities by intracerebroventricular administration. CP-OH (5-20 micrograms) showed no analgesic activity in the hot plate (51 degrees C) test using rats. However, it suppressed completely the analgesic effects of intraperitoneally administered morphine (3 mg/kg rat) in a dose-dependent manner. CP-OH showed no binding affinity for brain kappa receptors to which dynorphin, an opioid peptide that inhibits morphine analgesia, binds predominantly. These results suggest that, besides the conventional delta receptors which mediate analgesia, the rat brain contains another delta-like receptor which has a modulatory role to attenuate morphine-induced analgesia mediated through the mu receptors, and that this modulatory receptor does not exist in the mouse vas deferens.

delta EPhe4-enkephalin analogs. Delta receptors in rat brain are different from those in mouse vas deferens.

Conformationally restricted enkephalin analogs containing E-cyclopropylphenylalanine (delta EPhe), [D-Ala2, (2R,3S)-delta EPhe4,Leu5]enkephalin and its (2S,3R) isomer, were evaluated in receptor-binding assays using rat brain and in assays using muscle preparations. The (2S,3R) isomer was almost completely inactive in all assays. In contrast, the (2R,3S) isomer showed a very high affinity for the delta and a very weak affinity for the mu receptors in rat brain. The extent of delta affinity and the selectivity of this isomer were almost equal to those of [D-Pen2,D-Pen5]enkephalin. However, the (2R,3S) isomer was inactive in both the mouse vas deferens and guinea pig ileum assays, and showed no antagonistic activity in these tissues. These results indicate that the (2R,3S) isomer interacts with the delta receptors in rat brain, but not with those in the mouse vas deferens, and they suggest that the delta receptors in the central and peripheral nervous systems are different from each other.

Structural elements of Trimeresurus flavoviridis serum inhibitors for recognition of its venom phospholipase A2 isozymes.

Five inhibitors (PLI-I-V) against Trimeresurus flavoviridis (Tf, habu snake, Crotalinae) venom phospholipase A2 (PLA2) isozymes have been isolated from its serum. PLI-I, which is composed of two repeated three-finger motifs, and PLI-IV and PLI-V, which contain a sequence similar to the carbohydrate recognition domain (CRD) of C-type lectins, were expressed in the forms fused with glutathione S-transferase (GST). The resulting GST-PLIs showed ability to bind to three Tf venom PLA2 isozymes. The binding study with the truncated forms indicated that one of two three-finger motifs of PLI-I was able to bind to PLA2 isozymes. The N-terminal 37-amino acid fragment and the CRD-like domain of PLI-IV and PLI-V were bound to PLA2 isozymes. On the other hand, their C-terminal 12-amino acid segment also associated with PLA2 isozymes. When either of two units of a hydrophobic tripeptide in this sequence was replaced by trialanine, the binding was completely abolished, indicating that the C-terminal hydrophobic cores of PLI-IV and PLI-V were critically responsible for the binding to venom PLA2 isozymes.

Delta and mu opiate receptor probes: fluorescent enkephalins with high receptor affinity and specificity.

The fluorescent amino acid, L-1-pyrenylalanine (Pya) was incorporated into [D-Ala2,Leu5]enkephalin and its methyl ester at position 4 or 5. Pya-enkephalins showed strong fluorescent intensity and displayed high binding affinity for opiate receptors. Pya4-enkephalins showed high specificity for the mu receptors, while Pya5-enkephalins showed high specificity and selectivity for the delta receptors. Particularly, [D-Ala2,Pya5]enkephalin was as potent as the most utilized delta-specific ligand of [D-Ala2,D-Leu5]enkephalin (DADLE), and yet its delta-selectivity was about 5-times greater than that of DADLE. Thus, Pya-enkephalins per se can be utilized as a fluorescent probe or tracer for the opiate receptor-binding assays.

Enzyme inhibition by dipeptides containing 2,3-methanophenylalanine, a sterically constrained amino acid.

Both isomers of (E)-2,3-methanophenylalanine (delta EPhe), a sterically restricted amino acid, were incorporated into peptides in order to examine their possible enzyme inhibitory activity. Both (2R,3S)- and (2S,3R)- delta EPhe-Phe(or Leu)-OMe were found to inhibit effectively the hydrolysis of Ac-Tyr-OEt by chymotrypsin in a competitive manner. The ester groups of these dipeptides were quite resistant to chymotrypsin hydrolysis, and the delta EPhe-Phe peptide bond was also entirely stable. The inhibition constant (Ki) of the most potent dipeptide of H-(2R,3S)-delta EPhe-Phe-OMe was 0.16 mM at 25 degrees C. The inhibitory action of delta Phe-containing peptides was found to depend on the configuration of the delta Phe residue. The electrophilic nature of the cyclopropane ring which is conjugated with both the phenyl ring and the ester carbonyl group appears to be relevant to the inhibitory activity. Fully irreversible inactivation of chymotrypsin was achieved by its incubation with H-(2R,3S)- delta EPhe-Leu-OMe. An enzyme carboxylate group is thought to be responsible for nucleophilic attack on the cyclopropane ring leading to irreversible inactivation.

[4,4'-(Z)-dehydrophenylalanine]gramicidin S with stabilized bioactive conformation and strong antimicrobial activity.

Dehydrophenylalanine (delta Phe) was incorporated into an antibiotic peptide gramicidin S (GS) in place of D-Phe4,4' to prepare an unsaturated analog. Conformational analysis with 1H-NMR indicated that the unsaturated analog has much the same backbone conformation as that of natural gramicidin S as shown by NOE experiments. Studies on temperature dependences and on the chemical shift differences showed that the hydrogen bonds between Val-NH and Leu-CO in the unsaturated analog are strengthened by the incorporation of delta Phe4,4'. This resulted in the reinforcement of the beta-sheet structure which is the most important structural element for GS bioactivity. [delta Phe4,4']gramicidin S exhibited indeed very strong antimicrobial activities against Gram-positive bacteria as well as the natural peptide.

Accelerated evolution of crotalinae snake venom gland serine proteases.

Eight cDNAs encoding serine proteases isolated from Trimeresurus flavoviridis (habu snake) and T. gramineus (green habu snake) venom gland cDNA libraries showed that nonsynonymous nucleotide substitutions have accumulated in the mature protein-coding regions to cause amino acid changes. Southern blot analysis of T. flavoviridis genomic DNAs using two proper probes indicated that venom gland serine protease genes form a multigene family in the genome. These observations suggest that venom gland serine proteases have diversified their amino acid sequences in an accelerating manner. Since a similar feature has been previously discovered in crotalinae snake venom gland phospholipase A2 (PLA2) isozyme genes, accelerated evolution appears to be universal in plural isozyme families of crotalinae snake venom gland.

Receptor binding and biological activity of bivalent enkephalins.

Two series of dimeric enkephalin analogues were assayed for opioid activity in two isolated smooth muscle preparations: the guinea pig ileum (GPI) and the mouse vas deferens (MVD). Dimers have the general structure: X-(CH2)n-X, where X is H-Tyr-D-Ala-Gly-Phe-Leu-NH-(n = 0, 2, 4, 6, 8, 10, 12), for the first series of dimeric pentapeptide enkephalins (DPEn), and H-Tyr-D-Ala-Gly-Phe-NH-(n = 2, 4, 6, 8, 12), for the series of dimeric tetrapeptide enkephalins (DTEn). Comparison of biological activities with binding affinities revealed that: (1) the DPE series with n = 2-8 showed increased potency in the MVD assay relative to monomeric [D-Ala2, Leu5]enkephalinamide (DALEA); (2) there was an associated increase affinity for the delta receptor of rat brain or neuroblastoma-glioma hybrid cells. (however, the relative potencies were higher in the MVD assay then predicted on the basis of binding affinities); (3) the DTE series also showed an increase in delta receptor affinities and MVD potencies relative to DALEA, for n = 2-12; (4) for the DTE series, the increase in MVD activities was less than that expected on the basis of delta binding affinity; (5) for both the DPE and DTE series, activities in the GPI assay and mu-receptor affinities were highly correlated: as the length of the methylene bridge increased from 2 to 12, there was a progressive loss of activity in both assays, with a similar pattern for DPE and DTE. Two selected dimers and their corresponding monomers were also assayed for antinociceptive activity in vivo: results were consistent with GPI and mu-binding but not with MVD and delta-binding. Two alkylamide analogs of penta- and tetrapeptide monomers, representing the monomer with the attached spacer of the most active dimers, were also assayed in biological and binding assays. Comparison of these compounds with the corresponding dimers suggest that the changes in activities and selectivities induced by dimerization are not a spurious effect of the presence of an akylamide derivative of the carboxy terminal of enkephalin but rather may represent a specific effect due to the bivalent nature of the ligands.

Reversal of the antagonism of deglycosylated human chorionic gonadotropin by aggregation with wheat germ agglutinin.

Human chorionic gonadotropin (hCG) behaves as an antagonist upon chemical deglycosylation with hydrogen fluoride. In this study it was found that the alpha- and beta-subunits of deglycosylated hCG (DGhCG) still retained their ability to bind to wheat germ agglutinin (WGA) but not to concanavalin A. The antagonism of DGhCG against activation of adenylate cyclase and steroidogenesis was reversed, when WGA was bound to the N-linked carbohydrate moieties of hCG alpha- and beta-subunits followed by addition of purified Leydig cells. The complex formed by incubation at an approximately equimolar ratio induced the maximal reversal of antagonism. This reversal of antagonism was diminished by addition of N-acetylglucosamine. No increase of steric hindrance at the receptor sites was seen in binding studies of the [125I]DGhCG-WGA complex, indicating that the receptor-binding domain of hCG may not be adjacent to the carbohydrate moieties. Kinetic studies of the hormonal response showed that the DGhCG-WGA complex terminated cAMP accumulation after 30 min of incubation, but not testosterone production. Our results suggest that tetravalent WGA can also reverse the antagonism of DGhCG, as in bivalent antibodies to hCG beta.

Conformation of the beta subunit of deglycosylated human chorionic gonadotropin in the interaction at receptor sites.

The present study was performed to evaluate the relationship between the antagonist effect and the conformation of deglycosylated human chorionic gonadotropin (DGhCG) on the Leydig cell lutropin receptors. The maximum binding of [125I]DGhCG to anti-hCG beta antibody was decreased by 50%, while its binding profile to anti-hCG, anti-DGhCG and anti-hCG alpha antibodies remained unchanged, suggesting conformational changes in the beta subunit of DGhCG. However, the association of [125I]DGhCG to the binding sites of the receptors was much faster than that of [125I]hCG, and the ligand reached the binding equilibrium at 4 and 37 degrees C for 3 h and 15-30 min, respectively. Thus, the conformational changes in the beta subunit were not accompanied by loss of its receptor binding ability. The agonist properties of DGhCG which was bound to the receptors was fully restored by the addition of anti-hCG beta subunit antibody, while anti-hCG or anti-DGhCG restored only about 30% of the full agonist activity. This was probably due to a change of the conformation of the beta subunit to make it similar to that of the intact hormone. This restoration caused by anti-hCG beta was partially prevented by anti-hCG alpha. These facts indicate that some conformational change only in the beta subunit, not in the alpha subunit, of the deglycosylated hormone bound to the receptors is essential for the restoration of agonist properties.

Evidence for functional involvement of asparagine 67 in substrate recognition by snake venom phospholipases A2.

Site-directed mutagenesis studies of recombinant Trimeresurus flavoviridis venom gland phospholipase A2 (PLA2) showed that the Asn residue at position 67 takes part in recognition of the substrate 2-arachidonoyl sn-glycero-3-phosphocholine in both monomeric and micellar states. The amount of arachidonate released from phosphatidylcholine mixed micelles was reduced to 30% for N67D and N67K mutants, and to 70% for N67G mutant, but remained unchanged for N67S mutant. In contrast, for monomeric substrate, the activity was decreased to 40% for N67D and N67G and to 60% for N67K but remained unchanged for N67S. These results suggest that the properties of the side chain of residue 67 exert a significant influence on recognition of 2-arachidonoyl sn-glycero-3-phosphocholine.

Edge-to-face CH/pi interaction between ligand Phe-phenyl and receptor aromatic group in the thrombin receptor activation.

In the ligand/receptor interaction, the side chain phenyl group of phenylalanine (Phe) is involved in a so-called hydrophobic interaction, in which the Phe-phenyl group functions as a p element or merely as a hydrophobic element. The thrombin receptor-tethered ligand SFLLRNP consists of the Phe-2 residue essential for receptor activation. In order to explore the molecular characteristics of this Phe-2-phenyl group, a complete set of S/Phe/LLRNP peptides comprising six different difluorophenylalanine isomers [(F(2))Phe] was newly synthesized and assayed to evaluate their ability to induce the aggregation of human platelets. The assay results clarified several important structural elements to conclude that Phe-2-phenyl of S/Phe/LLRNP is in the edge-to-face CH/pi interaction with the receptor aromatic group, utilizing the Phe-phenyl edge along with adjacent benzene hydrogens at positions (2-3) or (5-6). It was also found that the fluorine atom at position 4 increases the acidity of the hydrogen mainly at its ortho position, resulting in a reinforcement of the CH/pi interaction and thus in an enhancement of biological activity. The H-->F replacement in the benzene ring was found to provide an effective structural examination to the Phe residue; i.e., to identify the hydrogens in the CH/pi interaction, and to strengthen the CH/pi interaction.

Lysine-49-phospholipases A2 from Trimeresurus flavoviridis venom are membrane-acting enzymes.

Basic proteins I and II (BP-I and BP-II) isolated from the venom of Trimeresurus flavoviridis (Habu snake) are isozymes of highly active Asp-49 phospholipase A2 (Asp-49-PLA2) and classified into the group Lys-49-PLA2. BP-II was found to elicit a strong contraction of guinea pig ileum, and this activity was inhibited completely by 1 microM indomethacin, an inhibitor of the arachidonate cascade. BP-II was inactive in the Ca(2+)-free medium, and p-bromphenacylated His-48-BP-II was also inactive. BP-II exhibited no binding affinity for the cells expressing PLA2 receptors. These results indicated that the contraction elicited by BP-II is due to the hydrolytic action of BP-II, liberating arachidonic acid from the ileum phospholipid biomembranes. In spite of its limited lipolytic activities (av. 0.9% of Asp-49-PLA2) against monomers and micelles of synthetic phospholipids, BP-II hydrolyzed considerably strongly the phospholipids in the artificial bilayer vesicles. Arachidonic acid released from liposomes of beta-arachidonoyl-gamma-stearoyl-L-alpha-phosphatidylcholine was determined by HPLC, and the activity of BP-II was estimated to be about 75% as compared to Asp-49-PLA2. Liposomes encapsulating carboxyfluorescein exhibited a strong dye-leakage induced by BP-II in a concentration-dependent manner, only in the Ca(2+)-containing buffer. The net result from all these observations was that BP-II, a Lys-49-PLA2, is an enzyme that hydrolyzes the membrane phospholipids. In contrast to BP-II, BP-I was found to be considerably weak in hydrolyzing membrane phospholipids, although its activities were distinct. BP-I and BP-II share a common sequence with the sole exception of Asp-67 (BP-I) and Asn-67 (BP-II) in the aligned sequences. This implies that the amino acid at position 67 of Lys-49-PLA2s is the residue required for discriminatory recognition of beta-arachidonoyl-phospholipid membranes.

Interaction mode of the phe-phenyl group of thrombin receptor-tethered ligand SFLLRNP in receptor activation.

Phenylalanine at position 2 of thrombin receptor-tethered ligand peptide (SFLLRNP) is crucially important for the activation of thrombin receptor. Its substitution by para-fluorophenylalanine [(p-F)Phe] enhanced several times the activity in human epithelial-like SH-EP cells [Nose et al. (1993) Biochem. Biophys. Res. Commun. 193, 694-699]. To clarify the interaction mode of Phe-2-phenyl in receptor activation, a series of analogs having chemical modifications on the benzene ring of Phe-2 were synthesized and examined for their ability to induce the aggregation of human platelets. When the fluorine atom was placed at the meta or ortho position, the resulting analogs exhibited considerably diminished activity (about 10-20% of para-derivative), indicating that the substitution is allowed only at the para position. The derivative with pentafluorophenylalanine was totally devoid of activity. These results suggested that Phe-2 requires hydrogen atom(s) on the benzene ring presumably for interaction with the receptor. No activity enhancement was observed for analogs with para-chloro-, bromo-, or iodophenylalanine, indicating the importance of the high electronegativity of fluorine to intensify the dipole of CH(s) remaining in the Phe-2-benzene ring. Inactivity of analogs having para-iodophenylalanine and homophenylalanine indicated the importance of the size of para substituents, and the placement of hydroxyl, nitro, and trifluoromethyl groups at the para position led to no activity. The interaction of Phe-2 of SFLLRNP appeared to be structurally restricted to a limited space in the receptor. The results suggested the presence of face-to-edge pi-pi interaction based upon the CH/pi interaction between the ligand Phe-2-phenyl group and the receptor aromatic group.