This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.
Estrogens are steroid hormone that are involved in a wide range of physiological functions including proliferation and growth in the reproductive tract, mammary gland, development of endometrium and uterus in females and spermatogenesis, spermiogenesis in males (Hess et al., 1997). Estrogens generally function by binding to specific receptors which are called estrogen receptors. Estrogen receptors are the members of nuclear receptor family and their functions are based on two domains, namely N-terminal DNA binding domain and C-terminal ligand-binding domain (LBD). The DNA-binding domain has a constitutive activity, whereas the hormone triggers action by binding to a ligand binding domain which is a hydrophobic pocket, and binding with specific amino acids for correct function. In addition to endogenous hormone substances such as estrogen and androgen, endocrine disrupting chemicals can also elicit a variety of adverse effects in humans and wild life (Mathur et al., 2008).
Endocrine disruptors (EDs) are the compounds which obstruct with endocrine system, adversely affecting hormone balance or disrupting normal functions of reproductive systems, developmental, and metabolic dysfunction in humans, animals, and plants. This class of EDs consists of both naturally occurring and man-made compounds, which are structurally similar to estrogen, and can therefore bind to the estrogen receptor (ER) and interfere with the actions of this endogenous steroid hormone (Daston et al., 2003). The World Health Organization (WHO), Environmental Protection Agency (EPA), and the World Wild Life Fund have reported endocrine disruptors as a critical environmental and health issue in the 21st Century (Daston et al., 2003; Hoyer 2001).
Currently, there are only a few in vitro and in vivo methods for determining whether a chemical is an endocrine disruptor or not (Kolpman and Chakravarti 2003). Most of the in vitro and in vivo data are resulting from assays which measure estrogenic and androgenic activity, and less is known for progestrogenic effects. Although endocrine disruption is a global issue for human health, little is known of its impact and significance, and strategies for detecting offending chemicals are not well described. Rat model is being used as a model system to study human health effects since it has close homolog of many human proteins. It is used to see the effects of endocrine disruptors and its toxic effects in vivo as well as in vitro. To address this endocrine problem, a list of well known endocrine disruptors was used in this study to find the rat estrogen receptor binding assay. These assays are more predictive of human and ecological effects to identify endocrine disruptors (ICCVAM, 2003). The focus of the present study is to explore the binding mode of the endocrine disruptors on Rat ERα and ERβ. Since the X-ray crystal structure of rat ERβ is available, in the present work, we built a homology model for rat ERα using multi-template modeling in MODELLER 9V7. Pharmacophoric features and molecular docking studies of the endocrine disruptors with Rat ERα and ERβ were performed using Phase and Glide modules in SCHR-DINGER suite 2009. In order to compare amino acid variation in ligand binding domain of ERα and ERβ of various species, we carried out sequence alignment for the ERα and ERβ sequences.
Materials and methods
Sequence alignment and Homology modeling of estrogen receptor alpha
The protein sequence of rat ERα (GenPept: NP_036821) and rat ERβ (GenPept: AAB97424) was retrieved from the NCBI (http://www.ncbi.nlm.nih.gov/). Multiple sequence alignment was done for rat ER α and ER β with ER α and ER β sequences of other species like rainbow trout (Marchand-Geneste et al., 2006), medaka (Cui et al., 2009), zebra fish (Costache et al., 2005) and human (accession number given in Table1) and sequence variation among the estrogen binding domain was identified using Discovery studio-2.1, respectively (Table1).
In the absence of x-ray crystal structure, homology model was built for rat ERα (NP_036821.1) by using multi-template modeling method of MODELLER 9v7. Homology modeling is currently the accepted method for the prediction of 3D structure of protein. Initially, best PDB templates for the sequence were identified using online tool NCBI BLASTP. Experimentally determined structures with high homology to the ERα were used for multi-template homology modeling studies (PDB ID: 2OCF, 1A52, 2P15 and 3ERD). Aligned results were inspected and adjusted manually to minimize the number of gaps and insertions. All the templates are having more than 30 % amino acid sequence identity with ESα. Therefore this alignment is suitable for construction of a reliable 3D model for ERα (Tramontano 1998).
2.2 Refining and evaluating the quality of the model
Modeled rat ERα was process to assess the quality of the three dimensional structure. We used PROCHECK (Laskowski et al., 1996), Errat (Colovos and Yeates 1993), Verify3D (Eisenberg et al., 1997) and WHAT-IF (Vriend and Sander 1993) to assess the quality of the modeled ER α structure.
ProSA-web: ProSA is a tool (Sippl 1993) widely used to check 3D models of protein structures for potential errors. Its application includes error recognition in experimentally determined structures (Banci et al., 2006; Llorca et al., 2006; Teilum et al., 2006) theoretical models (Ginalski 2006; Mansfeld et al., 2006; Panteri et al., 2006; Petrey and Honig 2005) and engineered proteins (Beissenhirtz et al., 2006; Wiederstein and Sippl 2005).
The overall quality score calculated for a specific input structure is displayed in a plot that shows the scores of all experimentally determined protein chains currently available in the PDB (Berman et al., 2000). This feature relates the score of a specific model to the scores computed from all experimental structures deposited in PDB. Problematic parts of a model are identified by a plot of local quality scores and the same scores are mapped on a display of the 3D structure using color codes.
2.3 Dataset and Common pharmacophore generation
Three dimensional structures of the compounds co-crystallized with ERα and ER β were extracted from the corresponding PDB files and chosen as a training set for the common pharmacophore generation. The molecular structures and their activities (IC50) values of the dataset compounds were represented in Table 2. Three dimensional structures of the compounds were prepared using LigPrep (version 2.3, Schrödinger, LLC, New York, NY, 2009) of Maestro 9.0 and a common pharmacophore was generated for ERα and ERβ co-crystallized compounds, respectively. To select and validate the best pharmacophoric feature, interaction and ligplot results of all known PDB files (PDBsum database) were compared.
2.4 Database searching
Two dimensional structures of the endocrine disruptors were retrieved from PubChem compound database (http://pubchem.ncbi.nlm.nih.gov/) and it used as test set in the study. Then all possible three dimensional structures for the compounds were generated using LigPrep utility and saved as ligand library. Then this ligand library was screened using the validated pharmacophore individually for ERα and ERβ. Ligands passing through the screening were then subjected to rigid body docking with ERα and ERβ.
2.5 Docking studies
2.5.1 Protein structure preparation
The crystal structure of Rat ERβ co-crystallized with 17-epiestriol (PDB ID: 2J7Y, resolution-1.80 Å) was retrieved from PDB. Then, the three dimensional structures of retrieved rat ERβ and modeled rat ERα were prepared using Maestro 9.0 protein preparation wizard; waters deleted, bond orders assigned, and hydrogen added appropriately. Next, the orientation of hydroxyl groups, amide groups of Asn and Gln, and charge states of his residues were optimized. Finally, a restrained minimization of the protein structure was performed using the default constraint 0.3Å RMSD and the OPLS 2005 force field.
2.5.2 Grid preparation
Prepared protein structures were used for generating glide scoring grids for the subsequent docking calculations. For ERβ, a grid box size of 20x20x20 was centered on the centroid of the crystallographic epi-estriol bound on estrogen binding region to confine the centroid of the docked ligand. For the modeled ERα structure, the scoring grid was generated using grid size of 20x20x20 and ligand range was defined using corresponding gate keeper residues. During the grid generation, hydrogen bond constraints at gate keeper residues Glu 353, Arg394, His 524, Leu346 and Gly521 (in case of ERβ, Glu305, Arg 346, His475, Leu339) were also included during the grid generation.
2.5.3 Docking calculations
A standard precision rigid body docking was performed using the pharmacophore screened endocrine disruptors. Upon completion of docking, the best docked complexes were selected based on the glide scoring function, interactions and the binding energy. The close interaction of the ligands with active site residue of estrogen receptor was analyzed by generating ball and stick model with secondary structure representation of docked complexes using Maestro molecular modeling graphical interface.
3. Results and Discussion
In this study, the substrate specificity for ERα and ERβ has been described based on interaction studies and variation in the gate keeper residues of estrogen binding domains.
3.1 Construction and evaluation of quality of the model
Sequence alignment of Rat ERα with its templates show about 30% sequence identity in the effector domain region (Fig.1), which suggests that the most important part of the sequence is conserved. Totally five models were constructed. The backbone root mean-square deviation (RMSD) between templates and model was 0.201, 0.927, 5.458 and 5.465 which indicates close homology of the model (Fig.2) reliable for further docking study.
Over all quality factor for non bonded atomic interaction is called Errat and high scores means higher quality. The normal range is >50 for a high quality model (Colovos and Yeates 1993). In this study, Errat score is 81.98, which is well within the range for high quality model. The Errat score of the templates are 1A52-95.23, 2OCF-92.79, 2P15-98.70 and 3ERD-95.38. Thus the backbone confirmation and non-bonded interactions of the homology model is within normal range.
VERIFY-3D used to evaluate the quality of protein structures for each residue. Using this scoring function, if more than 80 % of the residue has a score of >0.2 then the protein structure is considered high quality. ERα have 77.59 % of the residue has a score of >0.2 (Fig.3) (Eisenberg et al., 1997).
WHAT-IF is used to check the normality of the local environment of amino acid (Vriend and Sander 1993). The quality of the distribution of atom types is determined around amino fragments in WHAT-IF evaluation. In the present study the z-score is -1.91 which is above -5.0. In this result none of the score for each residue is lower than -5.0 as depicted. Therefore, the WHAT-IF evaluation also indicates that the homology model structure is very reasonable.
The stereochemical quality of the model is the distribution of the main chain torsion angles phi and psi which may be examined in a Ramachandran plot (Fig.4). The plot shows that 94.4% of the amino acids are in a phi-psi distribution consistent with right-handed α-helices. The remaining residues fall into the random or beta configuration geometries are very short segments and are primarily in the loop regions of the protein. In this study, no amino acid residues are outside the generously allowed region. The result showed that spot distribution for the modeled structure was similar to the standard X-ray structure of ERα.
ProSA-web: The interaction energy per residue was calculated using the ProSA-web. The z-score indicates overall model quality and measures the deviation of the total energy of the structure with respect to an energy distribution derived from random conformations (Sippl 1993, 1995). In the present study the z-score is -7.56 which is within the range characteristic for native proteins indicate that the structure is error free structure. The plot can be used to check whether the z-score of the protein in question is within the range of scores typically found for proteins of similar size belonging to one of these groups (Fig. 5a).
Local model quality: The energy plot shows the local model quality by plotting energies as a function of amino acid sequence position i. The homology model of ERα showed negative values which confirm that the modeled structure is reliable and good quality structure. In general, positive values correspond to problematic or erroneous parts of a model. Local model quality plot of single residue energies usually contains large fluctuations and is of limited value for model evaluation. Therefore the plot is soothed by calculating the average energy over each 40 residue fragment (Fig. 5b). ProSA-web visualizes the 3D structure of the protein using the molecule viewer Jmol. Residues with unusually high energies stand out by colour from the rest of the structure (Fig. 5c).
The sequence alignment of ERα and ERβ of various species with rat is shown in Table 1. Sequence alignment shows the gate keeper and other residues around the region of estrogen binding domain are identical and conserved (Fig. 6 and 7). Since ligand binding domain is identical, the binding affinity and binding mode for estrogen and other endocrine disruptors with one another are conserved. The percentage of sequence similarity and identity for rat ER with other species, and identity and variation in amino acid sequence for estrogen binding domains are given in Table 1.
3.2 Common pharmacophore generation
The comparative analysis of the reported crystal structure of human ERα-ligand PDB complexes, reveals that the key residues involve in ligand binding are Glu305, Met343, Arg346, Thr347, Glu353, Leu354, Trp383, Leu387, Arg394, Phe404, Met421, His475, His524 and Leu525. Among the key residues, His475 and Arg346 in the estrogen binding pocket act as hydrogen bond donor whereas Glu305 acts as a hydrogen bond acceptor with ligands and all other residues engaged in hydrophobic contacts. Similarly, key residues for the ligand binding in the rat ERβ are Arg30, Met250, Leu256, Glu260, Met291, Leu294, Met295, Gly427, His430 and Leu431, and the key residues, His430 and Arg30 in the estrogen binding pocket act as hydrogen bond donor whereas Glu260 acts as a hydrogen bond acceptor with ligands and all other residues engaged in hydrophobic contacts. Thus, we grouped the co-crystallized ligands based on interaction and we identified possible common pharmacophores model for each group, reports pharmacophore model of ER binding ligands contains one hydrogen acceptor, one hydrogen bond donor and one aromatic feature. Using this pharmacophoric hypothesis, the entire dataset of endocrine disrupting compounds were screened.
3.3 Docking study
Molecular docking was performed to the selected endocrine disrupting chemicals from the ICVAAM list and docked with ligand binding domains of predicted model of ERα and crystal structure of ERβ. In the present study 17β estradiol is used as control and it has the docking score of -9.98 and -10.59 with ERα and ERβ respectively (Table 3). The binding free energy of methoxychlor and its metabolites (mono, di and tri hydroxy methoxychlor) are -7.06, -8.48, -8.87, -8.39 kcal/mol respectively with ER α (Fig.8) and -6.45, -7.40, -8.77, -9.35 respectively when docked with ER β (Fig 9). The amino acid residues of ERα and ERβ such as His, Arg and Glu are the only residues involved in the hydrogen bond formation with all the above mentioned compounds and this result is consistent with the crystallographic data of the complex structures. In animal experiments, methoxychlor has been proved to be characterized as proestrogen and its metabolites are more estrogenic in nature (Gaido et al., 2000). The results of docking experiment displays that methoxychlor is proestrogenic and it has the less affinity towards estrogen receptors among the metabolites. We also found several other endocrine disruptors such as meso-hexesterol, diethylstilbesterol and coumestrol had the best docking score with esterogen receptors. Both ERα and ERβ belong to the member of same nuclear receptor family, sharing around 40% amino acid identity in the LBDs and this might be the reason that both receptors show the similar binding affinity with different endocrine disruptors.
In this work, we have studied amino acid residue variation in the ligand binding domain of various species ERα and ERβ by sequence alignment. Moreover, we also identified the valid pharmacophoric feature, binding modes and the interaction details of the endocrine disruptors with rat ERα and ERβ by means of pharmacophore and molecular docking study. The sequence alignment reveals that amino acids in the ligand binding domain are more identical and the interaction of the ligand with the ERα and ERβ are the same. The docking study result shows that all the endocrine disruptors bind to the pocket in same orientation with same kind of interaction. Hence, by comparing the binding energies and glide scores, we can confirm that the binding modes of the each endocrine disruptors.