Probing the interaction between cHAVc3 peptide and the EC1 domain of E-cadherin using NMR and molecular dynamics simulations

The goal of this work is to probe the interaction between cyclic cHAVc3 peptide and the EC1 domain of human E-cadherin protein. Cyclic cHAVc3 peptide (cyclo(1,6)Ac-CSHAVC-NH₂) binds to the EC1 domain as shown by chemical shift perturbations in the 2D ¹H,-¹⁵N-HSQC NMR spectrum. The molecular dynamics (MD) simulations of the EC1 domain showed folding of the C-terminal tail region into the main head region of the EC1 domain. For cHAVc3 peptide, replica exchange molecular dynamics (REMD) simulations generated five structural clusters of cHAVc3 peptide. Representative structures of cHAVc3 and the EC1 structure from MD simulations were used in molecular docking experiments with NMR constraints to determine the binding site of the peptide on EC1. The results suggest that cHAVc3 binds to EC1 around residues Y36, S37, I38, I53, F77, S78, H79, and I94. The dissociation constants (K_d values) of cHAVc3 peptide to EC1 were estimated using the NMR chemical shifts data and the estimated K_ds are in the range of .5 × 10^?5–7.0 × 10^?5 M.     

Insight into the structural requirements of aminopyrimidine derivatives for good potency against both purified enzyme and whole cells of M. tuberculosis: combination of HQSAR,CoMSIA, and MD simulation studies

The Mycobacterium tuberculosis protein kinase B (PknB) is critical for growth and survival of M. tuberculosis within the host. The series of aminopyrimidine derivatives show impressive activity against PknB (IC₅₀ < .5 μM). However, most of them show weak or no cellular activity against M. tuberculosis (MIC > 63 μM). Consequently, the key structural features related to activity against of both PknB and M. tuberculosis need to be investigated. Here, two- and three-dimensional quantitative structure–activity relationship (2D and 3D QSAR) analyses combined with molecular dynamics (MD) simulations were employed with the aim to evaluate these key structural features of aminopyrimidine derivatives. Hologram quantitative structure–activity relationship (HQSAR) and CoMSIA models constructed from IC₅₀ and MIC values of aminopyrimidine compounds could establish the structural requirements for better activity against of both PknB and M. tuberculosis. The NH linker and the R₁ substituent of the template compound are not only crucial for the biological activity against PknB but also for the biological activity against M. tuberculosis. Moreover, the results obtained from MD simulations show that these moieties are the key fragments for binding of aminopyrimidine compounds in PknB. The combination of QSAR analysis and MD simulations helps us to provide a structural concept that could guide future design of PknB inhibitors with improved potency against both the purified enzyme and whole M. tuberculosis cells.     

A novel identification approach for discovery of 5-HydroxyTriptamine 2A antagonists: combination of 2D/3D similarity screening,molecular docking and molecular dynamics

5-HydroxyTriptamine 2A antagonists are potential targets for treatment of various cerebrovascular and cardiovascular disorders. In this study, we have developed and performed a unique screening pipeline for filtering ZINC database compounds on the basis of similarities to known antagonists to determine novel small molecule antagonists of 5-HydroxyTriptamine 2A. The screening pipeline is based on 2D similarity, 3D dissimilarity and a combination of 2D/3D similarity. The shortlisted compounds were docked to a 5-HydroxyTriptamine 2A homology-based model, and complexes with low binding energies (287 complexes) were selected for molecular dynamics (MD) simulations in a lipid bilayer. The MD simulations of the shortlisted compounds in complex with 5-HydroxyTriptamine 2A confirmed the stability of the complexes and revealed novel interaction insights. The receptor residues S239, N343, S242, S159, Y370 and D155 predominantly participate in hydrogen bonding. π–π stacking is observed in F339, F340, F234, W151 and W336, whereas hydrophobic interactions are observed amongst V156, F339, F234, V362, V366, F340, V235, I152 and W151. The known and potential antagonists shortlisted by us have similar overlapping molecular interaction patterns. The 287 potential 5-HydroxyTriptamine 2A antagonists may be experimentally verified.     

Nmr and molecular dynamics study of four carbocyclic muramyl dipeptide analogues

We have performed a conformational analysis of the carbocyclic muramyl dipeptide analogues (1′R, 2′R)- and (1′S, 2′S)-N-[2-(2′-acetamidocyclohexyIoxy)acetyl]-L Ala-D -iGln (-D -Glu) utilizing ¹H-nmr spectroscopy and nuclear Overhauser effect restrained molecular dynamics. Intramolecular H bonding for all four diastereoisomers is suggested by the Ala-NH temperature coefficients. Distance restraints were obtained by NOE spectroscopy and rotating frame NOE spectroscopy experiments. Structures with low potential energy and high agreement with NOE data were sought by restrained molecular dynamics. The ring configuration was found to induce conformational preferences. The β-like turn characterized by the intramolecular C₁₀ H-bond Ala-NH-acetamido-CO is preferred with the (1′S, 2′S), but appears to be less stable with (1′R, 2′R), diastereoisomers. Calculations show the double β-turn proposed for muramyl dipeptide [S. Fermandjian, B. Perly, M. Level, and P. Lefrancier (1987) Carbohydrate Research, Vol. 162, pp. 23–32] to have higher potential energy. © 1993 John Wiley & Sons, Inc.     

Investigation on the binding mode of benzothiophene analogues as potent factor IXa (FIXa) inhibitors in thrombosis by CoMFA,docking and molecular dynamic studies

Recently, benzothiophenes attract much attention of interest due to its possible inhibitory activity targeting FIXa, a blood coagulation factor that is essential for the amplification or consolidation phase of blood coagulation. To explore this inhibitory mechanism, three-dimensional quantitative structure–activity relationship (3D-QSAR), molecular docking and molecular dynamics (MD) studies on a series of 84 benzothiophene analogues, for the first time, were performed. As a result, a highly predictive CoMFA model was developed with the q²?=?0.52, r²?=?0.97 and r²_pred?=?0.81, respectively. The CoMFA contour maps, the docking analysis, as well as the MD simulation results are all in a good agreement, proving the reliability and robustness of the model. These models and the information, we hoped, would be helpful in screening and development of novel drugs against thrombosis prior to synthesis.     

Relative mobility and activity of leaf malate dehydrogenase in flax(Linum usitatissimum) genotrophs and genotypes

Malate dehydrogenase (MDH) band relative mobility (R _m) and activity were examined in leaf extracts of Durrant's flax genotrophs, L and S, and flax genotypes, R and M. MDH activity in leaves from just below the inflorescence was higher in the two smaller, sparsely branched plant types, S and M, than in the larger, more branched plant types, L and R. The MDH electrophoretic banding pattern in flax leaf extracts consisted of three major anionic bands, MDH-1, MDH-2, and MDH-3. NoR _m differences were detected between corresponding isozymes of genotypes R and M. For the genotrophs, however, all three bands of S migrated faster than the corresponding bands of L. Codominance was absent in F₁ hybrids; SR _m was dominant for MDH-2 and MDH-3 and LR _m was dominant for MDH-1. The observations suggest that MDHR _m in L and S may be controlled by a modifier locus (or loci). Previous studies indicate that a modifier locus may also control heritable genotrophic differences in peroxidase (PER) and acid phosphates (AP)R _m. The three enzyme systems are compared.The financial assistance of the Natural Sciences and Engineering Research Council of Canada is acknowledged with thanks.     

Scrutiny of the mechanism of small molecule inhibitor preventing conformational transition of amyloid-β42 monomer: insights from molecular dynamics simulations

Alzheimer’s disease (AD) is a progressive neurodegenerative disorder that is characterized by loss of intellectual functioning of brain and memory loss. According to amyloid cascade hypothesis, aggregation of amyloid-β₄₂ (Aβ₄₂) peptide can generate toxic oligomers and their accumulation in the brain is responsible for the onset of AD. In spite of carrying out a large number of experimental studies on inhibition of Aβ₄₂ aggregation by small molecules, the detailed inhibitory mechanism remains elusive. In the present study, comparable molecular dynamics (MD) simulations were performed to elucidate the inhibitory mechanism of a sulfonamide inhibitor C1 (2,5-dichloro-N-(4-piperidinophenyl)-3-thiophenesulfonamide), reported for its in vitro and in vivo anti-aggregation activity against Aβ₄₂. MD simulations reveal that C1 stabilizes native α-helix conformation of Aβ₄₂ by interacting with key residues in the central helix region (13–26) with hydrogen bonds and π–π interactions. C1 lowers the solvent-accessible surface area of the central hydrophobic core (CHC), KLVFF (16–20), that confirms burial of hydrophobic residues leading to the dominance of helical conformation in the CHC region. The binding free energy analysis with MM–PBSA demonstrates that Ala2, Phe4, Tyr10, Gln15, Lys16, Leu17, Val18, Phe19, Phe20, Glu22, and Met35 contribute maximum to binding free energy (?43.1 kcal/mol) between C1 and Aβ₄₂ monomer. Overall, MD simulations reveal that C1 inhibits Aβ₄₂ aggregation by stabilizing native helical conformation and inhibiting the formation of aggregation-prone β-sheet conformation. The present results will shed light on the underlying inhibitory mechanism of small molecules that show potential in vitro anti-aggregation activity against Aβ₄₂.     

Inhibitor binding studies of Mycobacterium tuberculosis MraY (Rv2156c): Insights from molecular modeling,docking, and simulation studies

Abstract

Tuberculosis (TB) is a contagious disease caused by Mycobacterium tuberculosis (M.tb) or tubercule bacillus, and H37Rv is the most studied clinical strain. The recent development of resistance to existing drugs is a global health-care challenge to control and cure TB. Hence, there is a critical need to discover new drug targets in M.tb. The members of peptidoglycan biosynthesis pathway are attractive target proteins for antibacterial drug development. We have performed in silico analysis of M.tb MraY (Rv2156c) integral membrane protein and constructed the three-dimensional (3D) structure model of M.tb MraY based on homology modeling method. The validated model was complexed with antibiotic muraymycin D2 (MD2) and was used to generate structure-based pharmacophore model (e-pharmacophore). High-throughput virtual screening (HTVS) of Asinex database and molecular docking of hits was performed to identify the potential inhibitors based on their mode of interactions with the key residues involved in M.tb MraY–MD2 binding. The validation of these molecules was performed using molecular dynamics (MD) simulations for two best identified hit molecules complexed with M.tb MraY in the lipid bilayer, dipalmitoylphosphatidyl-choline (DPPC) membrane. The results indicated the stability of the complexes formed and retained non-bonding interactions similar to MD2. These findings may help in the design of new inhibitors to M.tb MraY involved in peptidoglycan biosynthesis.     

Molecular docking and 3D-QSAR-based virtual screening of flavonoids as potential aromatase inhibitors against estrogen-dependent breast cancer

Aromatase, catalyzing final step of estrogen biosynthesis, is considered a key target for the development of drug against estrogen-dependent breast cancer (EDBC). Identification and development of naturally occurring compounds, such as flavonoids, as drugs against EDBC is in demand due to their lesser toxicity when compared to those of synthetic ones. Thus, a three-dimensional quantitative structure–activity relationship, using comparative molecular field analysis (CoMFA) was done on a series of 45 flavonoids against human aromatase. A significant cross-validated correlation coefficient (q²) of 0.827 was obtained. The best predictive CoMFA model explaining the biological activity of the training and test sets with correlation coefficient values (r²) of 0.916 and 0.710, respectively, when used for virtual screening of a flavanoids database following molecular docking revealed a flavanone namely, 7-hydroxyflavanone beta-D-glucopyranoside showing highest predicted activity of 1.09?μM. In comparison to a well-established inhibitor of aromatase, namely 7-hydroxyflavanone (IC₅₀: 3.8?μM), the derivative identified in the present study, namely 7-hydroxyflavanone beta-D-glucopyranoside exhibited about 3.5 folds higher inhibitory activity against aromatase. The result of virtual screening was further validated using molecular dynamics (MD) simulation analysis. Thus, a 25 ns MD simulation analysis revealed high stability and effective binding of 7-hydroxyflavanone beta-D-glucopyranoside within the active site of aromatase. To the best of our knowledge, this is the first report of CoMFA-based QSAR model for virtual screening of flavonoids as inhibitors of aromatase.     

Dopamine D1 receptor and serotonin 5-HT1A receptor agonist effects of the natural product (–)-stepholidine: molecular modelling and dynamics simulations

In this study, by homology modelling and molecular dynamics (MD) simulation, models of l-stepholidine (l-SPD) activating the 5-HT_1A and D₁ receptors were constructed. In 100-ns MD simulations, the D₁ and 5-HT_1A receptors were activated by the partial agonist l-SPD, conforming with the global toggle switch activation model and the sequential activation model. The residues Y^7.53 and Y^5.58 swing significantly between different transmembrane (TM) domains after activation. Similarities between D₁ and 5-HT_1A included (1) the outward motion of TM-5; (2) the ionic lock was independent of the tilt of TM-6 and (3) there was an apparent bending of TM-6, and the ring of l-SPD formed strong π–π interactions with residue W^6.48. Differences between the two included the following: (1) in 5-HT_1A, l-SPD formed a hydrogen bond with Ala172^5.46 of TM-5, and the intracellular end of TM-5 moved outward slowly; that hydrogen bond did not form with the D₁ receptor; (2) l-SPD formed stronger interactions with D^3.32 and W^6.48 in the D₁ receptor than in the 5-HT_1A receptor and (3) the hydrogen bonding network was somewhat different in SPD-5-HT_1A and SPD-D₁ receptors. We propose the interaction between l-SPD and D^3.32 or/and W^6.48 is the original driving force during the whole activation process.     

Molecular dynamics simulation of the Al2O3 film structure during atomic layer deposition

Based on an understanding of atomic layer deposition (ALD) from prior experimental and computational results, all-atom molecular dynamics (MD) simulations are used to model the Al₂O₃ film structure and composition during ALD processing. By separating the large time-scale surface reactions from the small time-scale structural relaxation, we have focused on the growth dynamics of amorphous Al₂O₃ films at the atomic scale. The simulations are able to reproduce some important properties and growth mechanisms of Al₂O₃ ALD films, and hence provide a bridge between atomic-level information and experimental measurements. Information about the evolution of the microscopic structures of the Al₂O₃ films is generated, and the influence of operation parameters on the Al₂O₃ ALD process. The simulations predict a strong influence of the initial surface composition and process temperature on the surface roughness, growth rate and growth mode of the deposited films.     

Molecular dynamics simulation of d-Benzedrine transmitting through molecular channels within D3R

Dex-Benzedrine (known as d-Benzedrine or SAT) acts in dopamine receptors of central nerve cell system. In clinic, SAT is used to treat a variety of diseases; meanwhile, it has dependence and addiction. In order to investigate the pharmacology and addiction mechanisms of SAT as a medicine, in this paper, we have studied the structure of D₃R complex protein with SAT, and based on which, using potential mean force with umbrella samplings and the simulated phospholipid bilayer membrane (or POPC bilayer membrane), the molecular dynamics simulation was performed to obtain free energy changes upon the trajectories for SAT moving along the molecular channels within D₃R. The free energy change for SAT transmitting toward the outside of cell along the functional molecular channel within D₃R is 83.5 kJ mol^?1. The change of free energy for SAT to permeate into the POPC bilayer membrane along the protective molecular channel within D₃R is 87.7 kJ mol^?1. Our previous work gave that the free energy for Levo-Benzedrine (RAT) transmitting toward the outside of cell along the functional molecular channel within D₃R is 91.4 kJ mol^?1, while it is 117.7 kJ mol^?1 for RAT to permeate into the POPC bilayer membrane along the protective molecular channel within D₃R. The values of free energy suggest that SAT relatively prefers likely to pass through the functional molecular channel within D₃R for increasing the release of dopamine molecules resulting in a variety of functional effects for SAT. The obtained results show that the pharmacology and addiction mechanisms of SAT as a drug are closely related to the molecular dynamics and mechanism for SAT transmitting along molecular channels within D₃R.     

Studies of H4R antagonists using 3D-QSAR,molecular docking and molecular dynamics

Three-dimensional quantitative structure–activity relationship studies were performed on a series of 88 histamine receptor 4 (H4R) antagonists in an attempt to elucidate the 3D structural features required for activity. Several in silico modeling approaches, including comparative molecular field analysis (CoMFA), comparative similarity indices analysis (CoMSIA), molecular docking, and molecular dynamics (MD), were carried out. The results show that both the ligand-based CoMFA model (Q ² = 0.548, R _ncv² = 0.870, R _pre² = 0.879, SEE = 0.410, SEP = 0.386) and the CoMSIA model (Q ² = 0.526, R _ncv² =0.866, R _pre² = 0.848, SEE = 0.416, SEP = 0.413) are acceptable, as they show good predictive capabilities. Furthermore, a combined analysis incorporating CoMFA, CoMSIA contour maps and MD results shows that (1) compounds with bulky or hydrophobic substituents at positions 4–6 in ring A (R2 substituent), positively charged or hydrogen-bonding (HB) donor groups in the R1 substituent, and hydrophilic or HB acceptor groups in ring C show enhanced biological activities, and (2) the key amino acids in the binding pocket are TRP67, LEU71, ASP94, TYR95, PHE263 and GLN266. To our best knowledge, this work is the first to report the 3D-QSAR modeling of these H4R antagonists. The conclusions of this work may lead to a better understanding of the mechanism of antagonism and aid in the design of new, more potent H4R antagonists.     

Investigation of the selectivity of one type of small-molecule inhibitor for three Nav channel isoforms based on the method of computer simulation

Voltage-gated sodium (Na_v) channels play a pivotal role for the changes in membrane potential and belong to large membrane proteins that compose four voltage sensor domains (VSD1–4). In this study, we describe the binding mode and selectivity of one of the aryl sulfonamide sodium channel inhibitors, PF-04856264, for the VSD4s in Na_v1.4, Na_v1.5 and Na_v1.7, respectively, through molecular dynamics simulation and enhanced post-dynamics analyses. Our results show that there are three binding site regions (BSR1–3) in the combination of the ligand and receptors, of which BSR1 and BSR3 contribute to the selectivity and affinity of the ligand to the receptor. What’s more, the 39th residue (Y39 in VSD4^hNav1.4/ VSD4^hNav1.7 and A39 in VSD4^hNav1.5) and N42 in BSR1, the 84th residue (L84 in VSD4^hNav1.4, T84 in VSD4^hNav1.5, and M84 in VSD4^hNav1.7) in BSR2 and the conserved positive charged residues in BSR3 have major contributions to the interaction between the ligand and receptor. Further analysis reveals that if the 39th residue has a benzene ring structure, the connection of BSR1 and the ligand would be much stronger through π-stacking interaction. On the other hand, the strength and number of the hydrogen bonds formed by the ligand and the conserved arginines on S4 determine the contribution of BSR3 to the total free binding energy. We anticipate this study pave the way for the design of more effective and safe treatment for pain that selectively target Na_v1.7.     

Exploring the structural determinants of novel xanthine derivatives as A2B adenosine receptor antagonists: a computational study

Adenosine is a ubiquitous endogenous nucleoside that controls numerous physiological functions via interacting with its specific G-coupled receptors. Activation of adenosine receptors (AdoRs), particularly A_2B AdoRs promotes the release of inflammatory cytokines; reduces vascular permeabilization and induces angiogenesis, thereby making A_2B AdoR becomes a potentially pharmacological target for drug development. Presently, for investigating the structural determinants of 164 xanthine derivatives as A_2B AdoR antagonists, we performed an in silico study integrating with 3D-QSAR, docking and molecular dynamics (MD) simulation. The obtained optimal model shows strong predictability (Q²?=?0.647, R²_ncv?=?0.955, and R²_pred?=?0.848). Additionally, to explore the binding mode of the ligand with A_2B AdoR and to understand their binding mechanism, docking analysis, MD simulations (20?ns), and the calculation of binding free energy were also carried out. Finally, the structural determinants of these xanthine derivatives were identified and a total of 20 novel A_2B AdoR antagonists with improved potency were computationally designed, and their synthetic feasibility and selectivity were also evaluated. The information derived from the present study offers a better appreciation for exploring the interaction mechanism of the ligand with A_2B AdoR, which could be helpful for designing novel potent A_2B AdoR antagonists.

Communicated by Ramaswamy H. Sarma     

Selectivity and activation of dopamine D3R from molecular dynamics

D₃ receptor, a member of dopamine (DA) D₂-like receptor family, which belongs to class A of G-protein coupled receptors (GPCRs), has been reported to play a critical role in neuropsychiatric disorders. Recently, the crystal structure of human dopamine D3 receptor was reported, which facilitates structure-based drug discovery of D3R significantly. We dock D3R-selective compounds into the crystal structure of D3R and homology structure of D2R. Then we perform 20?ns molecular dynamics (MD) of the receptor with selective compounds bound in explicit lipid and water. Our docking and MD results indicate the important residues related to the selectivity of D3R. Specifically, residue Thr^7.39 in D3R may contribute to the high selectivity of R-22 with D3R. Meanwhile, the 4-carbon linker and phenylpiperazine of R-22 improve the binding affinity and the selectivity with D3R. We also dock the agonists, including dopamine, into D3R and perform MD. Our molecular dynamics results of D3R with agonist bound show strong conformational changes from TM5, TM6, and TM7, outward movement of intracellular part of TM6, fluctuation of “ionic lock” motif and conformational change of Tyr^7.53, which is consistent with recent crystal structures of active GPCRs and illustrates the dynamical process during activation. Our results reveal the mechanism of selectivity and activation for D3R, which is important for developing high selective antagonists and agonists for D3R.     

Enantiomeric Specificity of Biologically Significant Cu(II) and Zn(II) Chromone Complexes Towards DNA

Novel chiral Schiff base ligands (R)/(S)‐2‐amino‐3‐(((1‐hydroxypropan‐2‐yl)imino)methyl)‐4H‐chromen‐4‐one (L₁ and L₂) derived from 2‐amino‐3‐formylchromone and (R/S)‐2‐amino‐1‐propanol and their Cu(II)/Zn(II) complexes ( R1 , S1 , R2 , and S2 ) were synthesized. The complexes were characterized by elemental analysis, infrared (IR), hydrogen (¹H) and carbon (¹³C) nuclear magnetic resonance (NMR), electrospray ionization‐mass spectra (ESI‐MS), and molar conductance measurements. The DNA binding studies of the complexes with calf thymus were carried out by employing different biophysical methods and molecular docking studies that revealed that complexes R1 and S1 prefers the guanine–cytosine‐rich region, whereas R2 and S2 prefers the adenine–thymine residues in the major groove of DNA. The relative trend in K_b values followed the order R1 S1 R2 S2 . This observation together with the findings of circular dichroic and fluorescence studies revealed maximal potential of (R)‐enantiomeric form of complexes to bind DNA. Furthermore, the absorption studies with mononucleotides were also monitored to examine the base‐specific interactions of the complexes that revealed a higher propensity of Cu(II) complexes for guanosine‐5′‐monophosphate disodium salt, whereas Zn(II) complexes preferentially bind to thymidine‐5′‐monophosphate disodium salt. The cleavage activity of R1 and R2 with pBR322 plasmid DNA was examined by gel electrophoresis that revealed that they are good DNA cleavage agents; nevertheless, R1 proved to show better DNA cleavage ability. Topoisomerase II inhibitory activity of complex R1 revealed that the complex inhibits topoisomerase II catalytic activity at a very low concentration (25 μM). Furthermore, in vitro antitumor activity of complexes R1 and S1 were screened against human carcinoma cell lines of different histological origin. Chirality 24:977–986, 2012. © 2012 Wiley Periodicals, Inc.     

All-atom molecular dynamics simulations reveal significant differences in interaction between antimycin and conserved amino acid residues in bovine and bacterial bc1 complexes

Antimycin A is the most frequently used specific and powerful inhibitor of the mitochondrial respiratory chain. We used all-atom molecular dynamics (MD) simulations to study the dynamic aspects of the interaction of antimycin A with the Q_i site of the bacterial and bovine bc₁ complexes embedded in a membrane. The MD simulations revealed considerable conformational flexibility of antimycin and significant mobility of antimycin, as a whole, inside the Q_i pocket. We conclude that many of the differences in antimycin binding observed in high-resolution x-ray structures may have a dynamic origin and result from fluctuations of protein and antimycin between multiple conformational states of similar energy separated by low activation barriers, as well as from the mobility of antimycin within the Q_i pocket. The MD simulations also revealed a significant difference in interaction between antimycin and conserved amino acid residues in bovine and bacterial bc₁ complexes. The strong hydrogen bond between antimycin and conserved Asp-228 (bovine numeration) was observed to be frequently broken in the bacterial bc₁ complex and only rarely in the bovine bc₁ complex. In addition, the distances between antimycin and conserved His-201 and Lys-227 were consistently larger in the bacterial bc₁ complex. The observed differences could be responsible for a weaker interaction of antimycin with the bacterial bc₁ complex.