Electronic Structure ammett Correlations Quantitative Structure Activity Relationships (QSAR) Quantitative Structure Activity Relationships (QSAR) We have already seen that by changing groups in a drug it may increase or decrease the activity. These changes may also effect drug distribution, metabolism, etc. owever, in order to alter such properties in a logical and predictable fashion, we must consider a more quantitative method that addresses these changes on a molecular level. Understanding structure-activity relationships can be done by trial and error for each particular drug or chemical system under consideration. owever, a more logical and generalized approach may present ignorant reinventions and would be useful in predicting outcomes. Ideally, QSAR will provide a method to quantify a linear relationship between physiochemical properties of a drug and its biological activity. What properties would be of interest? 1
Quantitative Structure Activity Relationships (QSAR) Inductive and Field Effects Resonance Effects Lipophilic Effects Steric Effects Inductive Effects Proximity effects associated with activating and deactivating groups. assically these occur through bonds and will be discussed further in relation to LFER s. 2 N 2 N - + + δδδ+ δ + δ - δδδδ + δδ + CF 3 2
Field Effects Field effects are similar to inductive effects, but do not need to act through bonds. C 2 pk a C 3 5.79 C 3 5.08 5.07 CF 3 4.69 N 2 4.43 Resonance Effects Also known as mesomeric effects, these electronic interactions occur through π-bonds. Br C 3 C 3 "A" S N 1 Br messed up! C 3 "B" C 3 C 3 3
Steric Effects Effect of group size on selectivity and reactivity. R ΔG = -x (value depends on R group) R Linear free energy relationships allow a correlation of substituents with a reaction rate, biological activity, pk a, etc. To help us understand the magnitude of the sensitivity of reaction to changing substituents, we need a reference reaction. This is precisely what ammett set out to accomplish. ammett Equation (electronic effects) relates reaction rates with acid dissociation constants. Consider the following dissociation of a substituted benzoic acid. 2 25 o C, rt K a - + + K a = dissociation constant K a() = dissociation constant when = 4
Let s consider a few different types of groups compared to If = Deactivating group (EWG), the carboxylate ion is stabilized more than when =. As a result, K a increases (compared to when = ) If = Activating group (EDG), the carboxylate ion is less stabilized more than when =. As a result, K a decreases (compared to when = ) Thus we can make measurements of K a for many different substituents and then develop the following definition log K a /K a() = σ When =, K a = K a() and log K a /K a() = σ = 0 When = EWG, K a > K a() and σ > 0 (+ value) When = EDG, K a < K a() and σ < 0 (- value) Therefore, σ depends upon the electronic properties of the substituent and its position. Sigma (σ) is also known as a substituent constant. When the substituent is in the meta position on the aromatic ring, the effect is mostly inductive. owever, when the substituent is in the para position, both inductive and resonance effects are contributing. Therefore, σ para and σ meta are generally not the same for the same substituent. Σ ortho are more difficult to measure and simply quantify due to steric effects on the reactive portion of the molecule. 5
2 N K a C 2 3.45 x 10-4 (x=n 2 ) σ = log (3.45 x 10-4 / 6.46 x 10-5 ) = 0.73 C 2 6.46 x 10-5 (x=) C 3 C 2 3.38 x 10-5 (x=c 3 ) σ = log (3.38 x 10-5 / 6.46 x 10-5 ) = -0.28 Group N 2 CF 3 N 4 + CN N 2-0.38 0.13-0.92-0.37 C 3-0.28 0.10-0.78 0.05-0.26 C 3-0.14-0.06-0.31-0.10-0.17 Ph F Br I C C()Me N(C 3 ) 3 + σ para -0.57 0 0.05 0.15 0.24 0.26 0.28 0.44 0.47 0.53 0.60 0.70 0.81 0.82 σ meta 0 0.05 0.34 0.37 0.37 0.34 0.35 0.36 0.46 0.86 0.62 0.71 0.88 0-0.18-0.07 0.11 0.15 0.14 0.42 0.61 0.66 0.79 0.41 σ meta + - -0.81-0.47-2.30-0.82 NMe 2-0.63-0.10-1.70-0.12-0.09 σ para + -1.30-0.16 0 0 0.35 0.40 0.41 0.36 0.32 0.66 0.73 0.36 σ - para -0.15 0 0.02-0.03 0.19 0.25 0.27 0.77 0.84 0.65-0.56 1.00 1.27 0.77 6
Now that we have a reference reaction let s compare it to several different systems. K a + C 2 K a C 2 + C 2 C 2 K a C 2 C 2 + C 2 C 2 C 2 K a C 2 C 2 C 2 + C=CC 2 K a C=CC 2 + If we plot log (K a /K a() for these dissociation reactions verses σ values obtained from our reference reaction, we should see similar trends but perhaps to a lesser extent as the -group is now in different relationships to the business portion of the molecule. rate(k) -log(k/k o )... x x x x x x x x C 2 C 2 C 2 C 2 C 2 C 2-0.4-0.2 0 +0.2 +0.4 +0.6 +0.8 σ 7
From these plots we also get more information. We now can compare the slopes for different reactions and draw more conclusions. First we must define the slope = ρ (rho) For our reference reaction, rho = 1. This is because we would be plotting log K a /K a() for the substituted benzoic acids. Therefore rho becomes a measure of a reactions sensitivity to electronic effects. Log K a /K a() = ρσ ρ is highly dependent on the substituent distance from the reaction center. It may be considered as an effect of insulation. The slope intensity and direction of the line also can reveal changes or indications of mechanism. In biological systems the ρ value may show where binding [to the molecule] occurs. If ρ = (+); the reaction is accelerated by electron withdrawing substituents. If ρ = (-); the reaction is accelerated by electron donating substituents. So how does this effect activity of our drug? Consider that an amine may be protonated in a receptor and must be so to form electrostatic interactions with the receptor. Thus if you want to maximize the interactions of your drug with the receptor, you can now logically select which -group would be beneficial. N 3 K a N 2 + + Should one select a substituent with a σ that is positive or negative? Would the sign of ρ be positive or negative? 8
C 2 C 3 k - - + Et rate(k) -log k ρ = (+) -0.4-0.2 0 +0.2 +0.4 +0.6 +0.8 σ ammett also considered reaction rates as well as dissociation constants. Specifically, he examined the alkaline hydrolysis of substituted ethyl benzoates. C 3 C 2 P C 3 C 2 _ R C 3 C 2 P C 3 C 2 + - biomolecule C 3 C 2 C 3 C 2 P biomolecule 9
8 m-n(c 3 ) 3 p-n 2-3.0 m-n 2 log fly brain cholinesterase (I 50 ) 7 6 5 4 3 m-n(c 3 ) 2 m-t-bu p-sc 3 p- p-t-bu p-c 3 m-c 3 p-c p-c log K hyd -3.5-4.0-4.5-5.0-5.5 m-n 2 p- m-t-bu m-c 3 p-n 2 p-c 3 m-n(c 3 ) 2-0.2 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 ammett's σ constants -0.2 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 ammett's σ constants Base ydrolysis NR - 2 ρ = +1.36 + NR Acid ydrolysis NR + 2 ρ = -0.483 + NR Step 1 + (RDS) Step 3 NR NR 2 Step 2 - + NR 10
Remember some groups act by resonance. These include: Electron withdrawers by resonance: C 2, CN, C()R, S()R, S 2 R Electron donors by resonance: R, SR, (=CI, Br, I),, N 2, NR 2 2 N C 2 - + N N Normal σ values do not account for a direct resonance interaction between the substituent and a reaction center. Because of the unique effects imparted by resonance on these systems, other sigma s (σ) evolved to aid in the interpretation of particular reaction mechanisms. σ p - = Sigma para minus 2 N 2 N - N σ p - (N 2 ) = 1.27 Whereas σ p (N 2 ) = 0.81 C C 3 Use σ p YDRLYSIS Z C 3 Z =, N, S Use σ p - σ p + = Sigma para plus C 3 C 3 Solvolysis C 3 C 3 e.g. 3 C C 3 C 3 σ + p (C 3 ) = -0.65 σ + p (N 2 ) = -1.11 σ p (C 3 ) = -0.27 σ p (N 2 ) = -0.66 11
Additive Nature of Sigma (σ) Values Substituent effects of σ m and σ p are additive. - E 2CB ρ = +0.58 C N R 2 + C + N R -1.4-1.6 rate(k) -log(k/k o ) -1.8-2.0-2.2-2.4-0.4-0.2 0 +0.2 +0.4 +0.6 +0.8 σ 12
C N R + k 1 C N R 1 2 2 C N R k 2 3 6 C C + 5 N R C N R 4 1. When is a super donator, intermediate 2 is highly stabilized or k 1 becomes extremely fast and k 2 becomes extremely slow (or the rate determining step; RDS) by comparison. The overall reaction is retarded because k 2 is the attack by an electron rich nucleophile (water) at the benzylic position and electron donors would oppose such interaction. 2. When is a super withdrawing group, k 1 becomes slow as intermediate 2 formation is destabilized thereby retarding reaction. owever, strong withdrawers enhance k 2. At some point k 2 does not become rate limiting and the process is controlled solely by k 1. 3. At medium electronic substituent effects everything is balanced and the highest rate is achieved. Steric Effects: The Taft Equation (Es) In the 1950 s, Taft proposed that could be extended to include steric facots. The reference reaction that he utilized was the acid-catalyzed hydrolysis of substituted methyl acetates. k C 3 2, + + C 3 The steric contribution Es was defined by the following equation which should seem similar to that for sigma values for electronic effects discussed previously. Es = log (k -CMe / k Me-CMe ) The reference substituent was chosen to be = C 3 (Me). In this case any group larger than C 3 should give a negative (-) value and a group smaller than C 3 (i.e., = ) gives a positive value. 13
Steric Effects: The Taft Equation (Es) FC 2 C 2 Me + - FC 2 C 2 Me + + tbuc 2 C 2 Me + - tbuc 2 C 2 Me + + FC 2 C 2- + Me FC 2 C 2 + Me tbuc 2 C 2- + Me tbuc 2 C 2 + Me 3 C C 3 C 3 C 3 - C 3 3 C C 3 C 3 3 C C 3 C 3 SP 2 SP 3 Tetrahedral SP 2 Steric Effects: The Taft Equation (Es) Group Es Group Es +1.24 Cyclo C 7 13-1.10 Me 0 (Me)(Ph)C 2-1.19 Et -0.07 tbu -1.54 C 2-0.19 tbuc 2-1.74 n-pr -0.36 (Ph) 2 C -1.76 n-bu -0.39 (Br) 2 C -1.86 n-pentyl -0.40 (Et) 2 C -1.98 PhC 2-0.38 3 C -2.06 Cyclo-C 5 9-0.51 (ibu) 2 C -2.47 Cyclo-C 6 11-0.79 (Br) 3 C -2.43 i-bu -0.93 (Me) 2 (tbu)c -3.9 (Me)(Et)C 2-1.13 (Et) 3 C -3.8 14
ydrophobicity Constant (π) Sometimes it may be useful to be able to predict the lipophilicity (or hydrophobicity) of a compound similar to steric and electronic effects. ansch derived substituent constants for the contribution of atoms and groups on lipophilicity or more specifically, partition coefficients (P). Why is the partition coefficient important? ydrophobicity may effect a drug s ability to cross cell membranes or receptor binding interactions. ydrophobicity Constant (π) ow is P related to biological activity? f we let C = concentration of drug needed to achieve a desired pharmacological effect, we can express biological activity as 1/C. As a result, a low C would have great biological activity. In general, we find that an increase in hydrophobicity causes an increase in biological activity. This is most likely due to the fact that drugs need to pass across membrane barriers to reach their site of action. Thus, we should expect a linear correlation of biological response verses hydrophobicity. owever, we observe a parabolic relationship. -log activity [conc] -0.4-0.2 0 0.2 0.4 0.6 0.8 15
ydrophobicity Constant (π) Why do we see such a relationship? There are 3 major reasons to account for this relationship 1. Drugs that are too hydrophobic are not soluble in the aqueous phase. 2. Drugs then get trapped in fat deposits and never reach their site of action. 3. ydrophobic drugs may be more susceptible to metabolic degradation. So, back to ansch s hydrophobicity constants (π). ow could it be useful? Well first we need to define this constant. ansch defined the hydrophobicity constant (π) by the following familiar equation: log (P / P ) = π P is the partition coefficient of a compound containing substituent P is the partition coefficient for the parent compound or unsubstituted drug Substituents that increase the hydrophobicity of a compound will have positive (π) values. Substituents that decrease the hydrophobicity of a compound will have negative (π) values. ydrophobicity Constant (π) ow do we compute Log P? Because of the additive nature of π a Log P value can be computed by the fragmentation summation method using partition coefficients obtained from standard tables (see ansch Chem. Rev. 1971, 71, 525). For Example: π C 2 = log P nitroethane - log P nitromethane = 0.18 - (-0.33) = 0.51 Keep in mind that the effect of substituents on the hydrophobicity will be different depending on the carbon backbone to which they are attached (aromatic verses aliphatic) 16
ydrophobicity Constant (π) C 3 C C 2 C 2 N C 3 Diphenylhydramine log P = 2 π + Ph π + π C C π 2 C 2 + + -0.2 π NMe2 = 2(2.13) + 0.50-0.73 + 0.50-0.95-0.2 = 3.38 (experimental log P = 3.27) The Craig Plot We ve discussed electronic, steric, and hydrophobic effects and perhaps it would be useful to combine and display this information. Such a presentation Is called a Craig Plot. ften you will see values for π and σ plotted together as In the following graph. 17
The Craig Plot The advantage of such a plot is that one may quickly identify substituents with similar electronic effects (σ) but vary in hydrophobic (π) effects. Conversely, substituents with similar (π) values and differing (σ) values can be just as easily identified. Thus, if one is looking for a substituent to cause a specific change in a property of a drug, a logical choice can be made. It should be noted that other Craig Plots may be generated for P, Es, etc. 18