Competition binding

Radioligands are fairly expensive and only very few of them are specific enough for the purpose of receptor identification. Fortunately, radiolabelling of a drug is not strictly required for determining its affinity for a given receptor. This parameter can indeed be determined on basis of the drug's ability to compete with a (specific) radioligand for binding to that receptor. These competition binding experiments are now widely used by pharmacologists as a screening tool to evaluate the affinity of newly synthesized compounds (or of natural substances) for one or more receptors of interest. This approach has several advantages over the measurement of physiological responses. First, the same experimental set-up can be used to investigate the affinity of a drug for different receptors, whereas physiological responses may be very diverse and, hence, need to be monitored by different techniques. Second, the affinity of a drug for a specific receptor can be determined without ambiguity, whereas physiological responses are remote events, which may be triggered by different receptors or even be modulated at steps intermediate between receptor-stimulation and the final response.

It is important to note that the terms 'competition binding' and 'competitor' (for the non-radioactive substance) are commonly utilized irrespective of whether the 'competitor' is truly competitive or not. This semantic problem merits proper attention.

For competition binding experiments, constant amounts of membrane suspension are incubated with a fixed concentration of radioligand and increasing concentrations of the non-radioactive substance to be tested (the competitor), after which binding of the radioligand is measured. Binding of the radioligand is expressed as a function of the free concentration of the competitor by a competition binding plot (Figure 36). Competitor concentrations may span several orders in magnitude, so they are often expressed on a logarithmic scale. In the simple situation (in which the competitor is truly 'competitive') the radioligand and the competitor bind in a reversible fashion to the same site of the receptor. The radioligand (L)-receptor (R) and the competitor

Ic50 Curve
Figure 36 Competition binding curve (100% binding is binding in the absence of competitor) and determination of the competitor's Kj from the IC50.

(I)-receptor (R) interactions can be expressed as reversible bimolecular reactions: i.e.

The equilibrium dissociation constants for these interactions are denoted as KD for the radioligand and K (with i instead of D, to avoid confusion) for the competitor. The relationship between the amount of radioligand binding (B) and the competitor concentration (i.e. the competition binding plot) obeys the following equation:

B = Bc0ntrol - Bc0ntrol/(1 + K X (1 + [L]/Kd)/[I]) (8)

where control binding (Bcontrol) represents radioligand binding in the absence of the competitor.

An interesting situation occurs when the competitor has decreased control binding by 50% (i.e. when B = Bcontrol/2). This situation occurs when the concentration of competitor (usually denoted as IC50) is equal to KI X (1 + [L]/KD). The competitor's Kj can thus be calculated from the experimental IC50 value by the following equation (Cheng and Prusoff, 1973):

[L] is known and KD is obtained from saturation binding experiments. Please note that Ki is a constant, but that the IC50 value is dependent on the concentration and the KD of the radioligand used (Figure 37). Accordingly:

• Ki values represent affinity constants. They are the only valid parameters when comparing data from competition binding experiments which have not been performed under strictly the same conditions (e.g. with receptors from different

Figure 37 Effect of the [L]/KD ratio of the radioligand on the competition curve of a drug with K = 0.1 |iM.

cell types, different radioligands). Ki values of different competitors may be compared with each other, e.g. to give a rank order of drug affinities. Yet it should be kept in mind that their value depends on experimental factors such as the incubation temperature and buffer composition.

• IC50 values may only be compared to one another when they are obtained under strictly identical conditions; i.e. when the same source of receptors (membranes or cells of the same origin) and the same concentration of the same radioligand are used for all the competition binding experiments. This provides information about the rank order of drug affinities and about drug affinity ratios without the need to know the proper Ki value of each drug.

• IC50 values approximate Ki values when the radioligand concentration is well below its KD for the receptor. This is often the case in high-throughput screening.

A practical example of the utility of competition binding curves for finding out whether a radioligand truly binds to the desired receptor is shown in Figures 38 and 39. In these experiments, various unlabelled drugs compete with the tritiated ^-adrenergic antagonist [3H]-dihydroalprenolol for binding to turkey erythrocyte membranes. The experiments were performed under identical conditions (i.e. the radioligand concentration: 10 nM) so that the IC50 values of the curves can be compared with each other. The affinity of the agonists decreases as: (—)-isoproterenol > (—)-noradrenaline ^ (—)-adrenaline.

The non-selective a-adrenergic antagonist phentolamine has only very low affinity and no competition can be demonstrated for the non-bioactive compounds catechol

Competitor concentration (Log M)

Figure 38 Competition binding curves for pradrenergic receptors in turkey erythrocyte membranes. Reprinted from Biochemical and Biophysical Research Communications, 86, Bottari S., Vauquelin G., Durieu O., Klutchko C. and Strosberg A.D., The beta-adrenergic receptor of turkey erythrocyte membranes: conformational modification by beta-adrenergic agonists, 1311-1318. Copyright (1979), with permission from Elsevier.

Competitor concentration (Log M)

Figure 38 Competition binding curves for pradrenergic receptors in turkey erythrocyte membranes. Reprinted from Biochemical and Biophysical Research Communications, 86, Bottari S., Vauquelin G., Durieu O., Klutchko C. and Strosberg A.D., The beta-adrenergic receptor of turkey erythrocyte membranes: conformational modification by beta-adrenergic agonists, 1311-1318. Copyright (1979), with permission from Elsevier.

10 -9 -8 -7 -6 -5 -4 -3 Competitor concentration (Log M)

Figure 39 Stereoselective competition binding for pa-adrenergic receptors (Strosberg, 1987).

10 -9 -8 -7 -6 -5 -4 -3 Competitor concentration (Log M)

Figure 39 Stereoselective competition binding for pa-adrenergic receptors (Strosberg, 1987).

and 3,4-dihydroxy phenylglycol. These characteristics fit with those obtained for (3r adrenergic receptors by functional studies. Moreover, in agreement with the known stereoselectivity of ^-adrenergic receptors for antagonists such as propranolol and agonists such as adrenaline, the dextrorotary isomers display lower affinity than the levorotary isomers. In this respect, adrenaline and noradrenaline possess an asymmetric carbon atom in the ethanolamine side chain. They can thus exist either as levorotary (prefix 'l' or '(—)') or as dextrorotary (prefix 'd' or '(+)') stereoisomers. Only the levorotary form of adrenaline is produced and released into the bloodstream. It had already been observed in 1926 that this natural messenger is about 10 times more active than its (synthetic) (+)-isomer in raising blood pressure. This difference was explained in 1933 by the three-point attachment model of Easson and Stedmann (Figure 40): i.e.

interaction

Figure 40 Three-point attachment model of Easson and Stedmann: three bonds for (—^noradrenaline and only two bonds for (+)-noradrenaline.

interaction

Receptor

Figure 40 Three-point attachment model of Easson and Stedmann: three bonds for (—^noradrenaline and only two bonds for (+)-noradrenaline.

Figure 41 Sigmoidal, shallow and biphasic competition binding curves.

three of the four groups linked to the assymetric carbon of (-)-adrenaline (aromatic ring, OH and amino group) are involved in the interaction with the receptor whereas the OH group of (+)-adrenaline has the 'wrong' orientation. Such 'stereoselectivity' is very common for receptors.

When the radioligand and the competitor bind in a reversible fashion to a single population of non-interacting receptors (i.e. in the simple situation), the competition binding curve should have a steep sigmoidal shape (with 11, 50 and 89% decrease in radioligand binding when the competitor concentration is 1/10, equal or 10 times its IC50 value) (Figure 41A). The Hill coefficient of the competitor 'nHj' can be calculated from competition binding plots (Equation 10). In the case of a sigmoidal plot, nHi equals 1.

where Bo and BI are the binding of the radioligand in the absence and presence of competitor (I), respectively.

Radioligands may possess the same affinity for two (or more) receptors, receptor subtypes or even receptor subpopulations. When such different receptors co-exist in the same cells or membrane preparation, they will not be discriminated from each other by the radioligand. Indeed, the saturation binding curves appear as if the radioligand binds to a single class of non-co-operative sites. However, these different receptors (subtypes, subpopulations) may possess different affinities for certain unlabelled drugs, so that they can be detected and discriminated from each other by competition binding experiments with these drugs. In such cases, the nHi values of such curves will be less than one. There are two situations:

• First, the competitor displays a large (>1000-fold) difference in affinity for the different receptors, subtypes or subpopulations (Figure 41C). In this situation, the competition binding curve will be biphasic (i.e. with a plateau) and the parameters of each component (% of binding, IC50) are easy to measure. This is the case for [3H]-rauwolscine, which binds with the same affinity to a2-adrenergic receptors and serotonergic receptors of the 5-HT1A- type (Figure 42). Serotonin possesses

5-HT concentration (Log M)

Figure 42 Competition binding curve of serotonin (5-HT) for a2-adrenergic and 5-HT1A serotonergic receptors in membranes from human frontal cortex. The radioligand, [3H]rauwolscine, binds with equal affinity to both receptors and both receptors are present in human frontal cortex. Reprinted from Journal of Neurochemistry, 58, De Vos H. Czerwiec E. De Backer J.-P. De Potter W. and Vauquelin G., Regional distribution of alpha2A and alpha2B adrenoceptor subtypes in postmortem human brain, 1555-1560. Copyright Blackwell Publishing.

5-HT concentration (Log M)

Figure 42 Competition binding curve of serotonin (5-HT) for a2-adrenergic and 5-HT1A serotonergic receptors in membranes from human frontal cortex. The radioligand, [3H]rauwolscine, binds with equal affinity to both receptors and both receptors are present in human frontal cortex. Reprinted from Journal of Neurochemistry, 58, De Vos H. Czerwiec E. De Backer J.-P. De Potter W. and Vauquelin G., Regional distribution of alpha2A and alpha2B adrenoceptor subtypes in postmortem human brain, 1555-1560. Copyright Blackwell Publishing.

a much higher affinity for its own receptor than for the a2-adrenergic receptors and can be used to distinguish both receptors from each other in, for example, human frontal cortex membranes. At low concentrations it will first occupy the 5-HT1A receptors and only when its concentration gets high enough will it start to occupy the a2-adrenergic receptors. 5-HT1A receptors represent 40% of the binding and a2-adrenergic receptors 60%. The Kj values of serotonin for these receptors can be calculated from the IC50 values according to the equation of Cheng and Prusoff.

• Second, the competitor only possesses a limited (<100 times) difference in affinities for the different receptors, subtypes or subpopulations (Figure 41B). Such competition curves are shallow (nHi < 1) but, since both components are not separated by a distinct plateau, it is necessary to calculate the competition binding parameters of each component (% of binding, IC50) by computer-assisted analysis. Analysis of shallow competition curves is illustrated by Figure 43 for a2A- and a2B-adrenergic receptors. A radioligand such as the antagonist [3H]-RX821002 is unable to discriminate between them, but certain antagonists such as prazosin possesses a relatively weak selectivity for the a2B receptors. For the nucleus caudatus, the competition curve is quite shallow (nHi = 0.48). This indicates that a2A and a2B receptors are both present. The simplest way to describe such curve is to give its nHi and IC50. However, since Kis refer to individual competitor-receptor interactions, it is not possible to calculate any Ki from this IC50. Computer-assisted analysis is necessary to calculate the proportion of a2A and a2B receptors and their IC50 (and Ki) for prazosin.

Prazosin concentration {Log M)

Figure 43 Competition binding curve of prazosin (a2B- subtype- selective antagonist) for a2-adrenergic receptors in membranes from different human brain regions. Reprinted from European Journal of Pharmacology, 207, De Vos, H., Vauquelin, G., De Keyser, J., De Backer, J.-P. and Van Liefde, I., [3H]rauwolscine behaves as an agonist for the 5-HT1A receptors in human frontal cortex membranes, 1-8. Copyright (1991), with permission from Elsevier.

Prazosin concentration {Log M)

Figure 43 Competition binding curve of prazosin (a2B- subtype- selective antagonist) for a2-adrenergic receptors in membranes from different human brain regions. Reprinted from European Journal of Pharmacology, 207, De Vos, H., Vauquelin, G., De Keyser, J., De Backer, J.-P. and Van Liefde, I., [3H]rauwolscine behaves as an agonist for the 5-HT1A receptors in human frontal cortex membranes, 1-8. Copyright (1991), with permission from Elsevier.

• The curve is also shallow for the putamen but it is steep (nHi = 1.01) for the cortex. For this brain area, the curve can be analyzed according to a single-site model and the high Ki of prazosin indicates that only a2A receptors are present in this brain region.

An interesting situation occurs for G protein-linked receptors (GPCRs) in broken cell preparations. These can often be divided into two subpopulations with different affinities for agonists, but with the same affinities for antagonists. This heterogeneity towards agonists is not related to differences in primary amino acid sequence, but rather to their capability to undergo functional coupling to the G proteins (see Section 4.9). The receptor population that undergoes functional coupling to G proteins (coupling-prone receptors) displays high agonist affinity. The receptor population that is unable to couple (non-coupled receptors) displays low agonist affinity. The competition curves will therefore depend on the nature of the radioligand and competitor used (Figure 44):

• If the radioligand is an antagonist, it will regard the receptors as a single class of non- co-operative sites. Antagonist competition binding curves will be steep, but agonist competition binding curves will be shallow.

• If the radioligand is an agonist, it will preferentially label the coupling-prone receptors (especially at low concentrations). Hence, the non-coupled receptors may not be detected in these assays.

Figure 44 Effect of GTP and related guanine nucleotides on competition binding curves.

Figure 44 Effect of GTP and related guanine nucleotides on competition binding curves.

Shallow competition binding curves for agonists may thus reflect two distinct phenomena: the presence of different receptors or functional receptor heterogeneity. Fortunately, it is possible to distinguish between these two possibilities by using guanine nucleotides such as GTP. These compounds break up agonist-receptor-G protein complexes, so that the receptors return to the uncoupled, low agonist-affinity form (see Section 4.1). In practice, GTP is thus capable of producing a rightward shift and steepening of the agonist versus radiolabelled antagonist competition binding curve, at least if the high affinity is related to functional coupling of the receptor to the G proteins (Figure 44, Figure 157). When the radioligand is an agonist, its binding will be greatly reduced by GTP.

Was this article helpful?

0 0

Post a comment