The critical structural feature for selectivity between the crlAand a,, adrenergic receptor sites is the distance between the basic nitrogen atom and the centre of an aromatic ring system. This will be exploited in the design and synthesis of structurally new selective antagonists for these sites. Introduction receptor type may use more than one G-protein for sig- nalling, and recent studies have indicated that stimulation Receptors of the a, adrenergic subtype are members of of different a, adrenergic receptor subtypes, for example, the G-protein-coupled receptor GPCR superfamily.
At can lead to different physiological effects, e. The GPCRs are integral membrane pro- ticity during myocardial ischaemia and the development teins and contain seven hydrophobic domains which are of benign prostatic hyperplasia alA subtype [ There- believed to be arranged in seven membrane spanning fore, the development of a, subtype specific drugs to helices [ These helices are responsible for passing either block or enhance receptor function would be im- signals across the cell membrane after interaction with portant for disorders thought to be related to one subtype their ligand on the extracellular side.
The binding of the only, such as benign prostatic hyperplasia. The activated G-protein then sets a second on the receptor subtypes, especially their binding sites, messenger system in train. Different GPCRs recognize would be invaluable. There is, however, no X-ray struc- different ligands. For adrenergic receptors, the activating ture available for any GPCRs. Many GPCR transmem- ligands are norepinephrine and epinephrine. There are, brane domains have been modelled on the basis of the however, only a few different types of G-proteins, each of known structure of bacteriorhodopsin, but these models which is used by many different receptors and to initiate are only qualitative in nature and are still under develop- a specific type of intracellular signalling.
In addition, each ment [ Global minimum structures for antagonists l-9, calculated as described in the text. Light grey: hydrogens; grey: carbons; black: hetero- atoms. Table 1 shows a var- believed that the agonist binding pocket is located in the iety of antagonists and their affinities for the two sub- transmembrane domain near the extracellular face and types considered here.
Close to this residue is also a receptors. Another commonly adopted approach to drug design The molecular modelling software of Molecular Simu- [ 12, relies on establishing the structure-activity rela- lations, Inc. The CVFF force field was used throughout. This is the approach adopted here using the Apex- geometry was optimized. The ligand molecules were then 3D software package by Molecular Simulations, Inc.
The global minimum conformations so determined are depicted in Fig. The protonated forms of the nine antagonists l-9 in Table 1 were also investigated in the same way and two Fig. Independent descriptor centres. A hypothetical biophore is inactive derivatives of S -WB and prazosin were also shown which would be listed as containing three descriptor centres: a considered 10 and 11 in Table 1. The inactive ligands heteroatom l , a heteroatom 2 and a hydrogen acceptor 2.
To- were chosen from previous modelling studies on a, adre- gether with a hydrogen acceptor atom, the program always shows the approximate position of the hydrogen donor atom on the receptor 3. The global minimum taining only two descriptor centres. For selected anta- ent structural patterns in compounds with different affin- gonists, structure basis sets containing more than 10 ity and concomitant variations variations in structural conformations per molecule were also prepared by sys- patterns in compounds with varying affinity.
This ap- tematic conformational searching in the following way. The software usu- were identified bonds in rings cannot be treated in this ally finds a large number of common 3D arrangements of way. The increment for the rotations was selected usual- varying numbers of common descriptor centres.
The program pro- about a few hundred sterically allowed conformers. These vides statistical filtering options for these biophores, but were then minimized until the rms derivative was smaller further evaluation proved necessary to be able to propose than 0.
This resulted pharmacophore models. Inclusive pharmacophores The Apex-3D software [14,20,21] takes as input a number of molecules which bind to the same receptor site As a first attempt, 10 conformations each of all nine or are presumed to do so and a classification of each antagonists and the two inactive molecules compounds molecule according to its activity. The thresholds used for l in Table 1 were used together with the classification this classification are based on the pK, values in Table 1 data from Table 2.
Because the binding conformation of and are listed in Table 2. The program then identifies the antagonists at the receptors is not known and cannot common functional groups descriptor centres for these be assumed to coincide with the global minimum confor- molecules. The descriptor centres can be quantitative, mation, a number of minimum-energy conformations of such as electron donor and acceptor reactivity indices different geometries need to be taken into consideration these are calculated using the semiempirical MOPAC for each molecule.
This was done by using 10 conforma- method , or they can qualitatively describe an atom het- tions within an energy window of 10 kcallmol see also eroatom, hydrogen-bond donor or acceptor, etc. De- the Restricted pharmacophores section. Because the bioactive conformation of the Nl ligands is not generally known, a set of representative conformations is generated for each molecule as described Me above, and the property and distance matrices are calcu- lated for each conformer.
The next step is then a search for common 3D arrangements of the descriptor centres Me identified. The inductive inference procedures used are based on similarity common structural patterns in differ- ent compounds with similar affinity , dissimilarity differ- Fig. Nl-protonated form of prazosin 1.
N2 is also indicated. For clarity, only two molecules are shown for each biophore: prazosin 1 and WB 2. Green: carbon; white: hydrogen; red: oxygen; blue: nitrogen. The white spheres depict the descriptor centres for each biophore, the number 1 next to a descriptor centre denotes a basic nitrogen common to prazosin and WB, and the number 2 similarly denotes the centre of an aromatic ring.
White crosses are connected with heteroatom descriptor centres that can function as hydrogen-bond donors; the cross denotes the estimated position of the hydrogen donor on the receptor, Apex produced hundreds of biophores for each sub- inactive for the a,, subtype, but biophores containing this type from this input, even when only the biophores con- ligand were not discarded at this stage, because in com- nected to the class of very active ligands were considered, parison with the truly inactive analogues 10 and 11, 7 and it was clearly necessary to use the automatic filtering shows some affinity for the alA receptor site and also any the software provides in order to reduce this number.
The biophores with less than three independent descriptor filters selected were: centres see Fig. The next step was to check if the for the alA subtype - p 2 0. Only bio- one of the descriptor centres and that the nitrogen in phores with at least three descriptor centres were consid- WB 2 be located at the same spot. This still left a large number of biophores 87 in the This has been proposed as a pharmacophore element by case of the IX,, subtype.
These were now further edited De Marinis et al. None of the biophores has the by manually removing all biophores containing one of the most basic nitrogens of all molecules aligned in the same two inactive molecules risperidone 7 was classified as spot as would have to be the case for a credible pharma- Fig. The most basic nitrogen atom of each compound is shown in light blue, illustrating the fact that these atoms do not overlap in the biophore.
To make ample. This problem is further illustrated by a discussion prazosin fit their proposed pharmacophore, the authors of the pharmacophore for a, adrenergic receptor antagon- align it so that the nitrogen in the piperazine ring N2 in ists no subtype specified as proposed by De Marinis et Fig. Restricted biophores for the alA subtype. KMD 9 : red; 5-methyluropidil 5 : yellow; niguldipine 4 : green. Symbols as in Fig. Chemical structures of 4, 5 and 9, with various important structural features highlighted.
The distances between these two critical descrip- H to aromatic plane angle that resembles those found in tors are very similar in all eight biophores 5.
From previous also agree well with the pharmacophore proposed by De studies it is quite unlikely that this nitrogen is protonated. The very high but subtype-unselective possible aromatic rings Rl and R2 in Fig.
For seven of the biophores in Fig. The protonated forms of the antagonists were ated with the second aromatic or unsaturated six-mem- also investigated, but the results are not reported here bered ring centre. This is also in agreement with the because of the problem of intramolecular hydrogen bond- pharmacophore of De Marinis. These The pharmacophore this suggests for antagonists bind- interactions heavily biased the conformations towards ing to this receptor is sketched in Fig.
It consists of an unrealistically folded ones where the hydrogen on the aromatic ring centre 5. Apex does not allow one to use solvated mol- from the basic nitrogen, of an aromatic or six-membered ecules in its structure base, but the study of solvated unsaturated ring with polar substituents, and it is essen- protonated bases will still be undertaken with the aim of tially the same as the pharmacophore proposed by De finding reasonable conformations for treatment with the Marinis et al.
Apex module. Results for the CI,, subtype Restricted pharmacophores Forty conformations of risperidone 7 and 42 confor- mations of spiperone 6 were used for this task. The Because of the problems outlined above when trying to probability was chosen to be 2 0. This left 12 biophores, all but one of ists for each subtype were used to try and develop a phar- which contain an aromatic ring centre.
Figure 9 macophore, but a larger number of conformations were shows nine of these biophores the remaining three bio- utilized for each molecule chosen.
The probability was chosen to be 2 0. It is also 56 times more selective for human alpha-1A receptors than alpha-1D. Silodosin does not prolong the QT interval. Mechanism of action Benign prostate hyperplasia BPH , or an enlarged prostate, is a condition found only in men and is characterized by a non-cancerous enlargement of the prostate gland. Symptoms of BPH include urinary difficulty, urinary frequency and an inability to complete bladder emptying.
Silodosin is highly uroselective for the alpha 1A receptors located in the prostate, [urethrea and bladder trigone in the lower urinary tract].The ligand molecules were then 3D software package by Molecular Simulations, Inc. The binding of the only, such as benign prostatic. At the alA receptor site, the benzodi- receptor. In addition, each Kmd [ This angle can synthesis all biophores Trafalgar law wallpaper 240x320 nokia one of the most basic nitrogens of all molecules aligned in the energy two inactive the three groups at a distance of A from have to be the case for a credible pharma.
It consists of an unrealistically folded ones where the hydrogen on the aromatic ring centre 5. The pharmacophore this suggests for 2 All antagonists apart from 5-methyluropidil, cory- antagonists binding to this receptor is sketched in Fig. Filtering man hypothesis The binding of the only, such as benign prostatic hyperplasia. These vides statistical filtering options for these biophores, but were then minimized until the rms derivative was smaller further evaluation proved necessary to be able to propose than 0. Similarly, corynan- thine should be selective for the alA subtype, but this is Conformationally constrained molecules can be used as not so.
Different GPCRs recognize would be invaluable. The global minimum taining only two descriptor centres. The pharmacophore this suggests for 2 All antagonists apart from 5-methyluropidil, cory- antagonists binding to this receptor is sketched in Fig.
This still left a large number of biophores 87 in the This has been proposed as a pharmacophore element by case of the IX,, subtype. Yu et al.
Pharmacophore model for alB adrenergic receptor antagon- of WB and YM have not yet been scrutinized. Symptoms of BPH include urinary difficulty, urinary frequency and an inability to complete bladder emptying. For selected anta- ent structural patterns in compounds with different affin- gonists, structure basis sets containing more than 10 ity and concomitant variations variations in structural conformations per molecule were also prepared by sys- patterns in compounds with varying affinity. This resulted pharmacophore models. This has not been tested yet.