Hydrogen Bonding & Molecular Design

Peter Kenny(pwk.pub.2008@gmail.com)

Hydrogen Bonding

Interactions between drug molecules in crystal lattice

(Solubility, melting point polymorphism, crystallinity)

Interactions between drug and water molecules

(Solubility, distribution, permeability, potency, toxicity,

efflux, metabolism)

Interactions between drug molecules & (anti)target(s) (Potency, toxicity, efflux , metabolism, distribution)

Hydrogen Bonding in Drug Discovery & Development

Interactions between water molecules

(Hydrophobic effect)

Neglect of hydrogen bond strength:A recurring theme in medicinal chemistry

• Rule of 5• Rule of 3• Scoring functions for virtual screening • Polar surface area (PSA) as molecular descriptor• Relationships between binding thermodynamic and

buried polar/non-polar surface area

Measuring hydrogen bond strength

Acceptors

Donors

pKHB logKb

logKa

OH F OH NO2 (CH3CCl3)(CCl4)

Taft et al , JACS 1969, 91, 4801-4808Laurence & Berthelot, Perspect. Drug. Discov. Des. 2000, 18, 39-60.

Abraham et al, JCS Perkin Trans 2 1989, 1355-1375

N

O

Me

(CH3CCl3)

Abraham et al, JCS Perkin Trans 2 1989, 1355-1375

logKb: Heteroaromatic nitrogen

Azines

pKa 5.22 9.70 2.24 1.23 0.65

logKb 2.52 3.54 2.53 1.67 1.46

Azoles

pKa 7.25 2.09 0.80 -2.03

logKb 3.68 2.22 1.67 1.06

Abraham et al, JCS Perkin Trans 2 1989, 1355-1375

N N

N

NN

N

N

N

N

N

N

NN

N

O

NO

Modelling Hydrogen Bonding

Calculate energy of complex

• Need to know complexation partner

• Need to generate multiple 3D models of complex

• BSSE• Relevance to physiological

media?

Calculate molecular electrostatic properties

• No explicit reference to complexation partner

• More appropriate to general parameterisation

Minimised electrostatic potential has been show n to be an effective predictor of hydrogen bond basicity

Plot of V/kJmol-1 against r/Å for pyridine on lone pair axis show ing electrostatic potential minimum 1.2Å from nitrogen

-300

-200

-100

0

V

0 1 2 3 4 5

r

Electrostatic potential as function of position for acceptor

V/k

Jmo

l-1

r/Å

N

r/Å

r

Comparison of Vmin and pKa as predictors of logKb

logK

b

Vmin/(Hartree/electron) pKa

Heteroaromatic nitrogen in five and six-membered rings

Kenny JCS Perkin Trans 2 1994, 199-202

1.01 1.16 0.94 0.40 0.06

2.63 1.53 2.50 1.89 1.82

2.39

Predicted logKb 2.64 1.68 2.51 1.90 2.50

Measured logKb 2.38 1.98 2.36 2.17 1.99

NN

N

Ph

SN

N

S

Me

NN

N

Ph

N NN

MeN

N

NN

Non-equivalent acceptors provide validation set

Kenny JCS Perkin Trans 2 1994, 199-202

r

Donors: The Va(r) descriptor

Calculate electrostatic potential (V) at this point

Va(r) as predictor of logKa

Sensitivity to distance from donor hydrogen

V a /(Hartree/electron) V a /(Hartree/electron)

logK

a

logK

a

r = 0.55 Å r = 1.20 Å

R2 = 0.65RMSE = 0.43

R2 = 0.93RMSE = 0.20

Kenny, JCIM, 2009, 49, 1234-1244

Fluorine: A weak hydrogen bond acceptor

-0.122 -0.113 -0.071

-0.038

-0.054

-0.086-0.091

-0.072

-0.104 -0.093

Hydrogen bonding of esters

Toulmin et al, J. Med. Chem. 2008, 51, 3720-3730

-0.316

-0.315

-0.296

-0.295

Bioisosterism: Carboxylate & tetrazole

Kenny, JCIM, 2009, 49, 1234-1244

-0.262

-0.261

-0.268

-0.268

G

Do1

Do2

Ac1 Ac2

A

Do1 Do2

Ac1

Kenny, JCIM, 2009, 49, 1234-1244

DNA Base Isosteres: Acceptor & Donor Definitions

Watson-Crick Donor & Acceptor Electrostatic Potentials for Adenine Isosteres

N

N

N

N

NH2

N N

N

N

NH2

N

N

N

NH2

Vm

in (

Ac1

)

Va (Do1)Kenny, JCIM, 2009, 49, 1234-1244

NN

N

NH2

N

N

N

N

NH2

N

NN

N

NH2

IsostereDuplex stability

Donor (Do1) Donor (Do2)Acceptor

(Ac1)

reference 0.335 0.331 -0.086

- 0.329 0.325 -0.089

+ 0.340 0.336 -0.079

pred + 0.341 0.348 Minimum not located

N

N

N

NH

O

NH2

O

N

NH

NH2N

O

N

NH

NH2

N

N

NH2

NH N

N

N

O

Guanine bioisosteres & DNA duplex stability

Kenny, JCIM, 2009, 49, 1234-1244

+

+

+

Ternary complex

K1

K2

Quantifying the effect of complex formation

Kenny, JCIM, 2009, 49, 1234-1244

HO

H HO

H HO

H

H

OH

N

H

O

Effect of complex formation on predicted hydrogen bond acidity of water

1.2

(~ Alcohol)

2.0

(~ Phenol)

2.8

(~ 4-CF3Phenol)

Kenny, JCIM, 2009, 49, 1234-1244

Polarity

NClogP ≤ 5 Acc ≤10; Don ≤5

An alternative view of the Rule of 5

Octanol/Water Alkane/Water

OH OH

OHOH

Octanol/water is not the only partitioning system

logPalk: Experimental challenges

• Many polar solutes are poorly soluble in alkane solvents

• Self-association – Masks polarity– Limits concentration at which measurements can be made.– Need to vary concentration to demonstrate that it is not an

issue

logPoct = 2.1

logPalk = 1.9

DlogP = 0.2

OMe

OH

N

O

logPoct = 1.5

logPalk = -0.8

DlogP = 2.3

logPoct = 2.5

logPalk = -1.8

DlogP = 4.3

Differences in octanol/water and alkane/water logP values reflect hydrogen bonding between solute and octanol

Toulmin et al, J. Med. Chem. 2008, 51, 3720-3730

N

O

NO O

DlogP = 0.5

PSA = 48 Å2

DlogP = 4.3

PSA = 22 Å2

PSA is not predictive of hydrogen bond strength

Toulmin et al, J. Med. Chem. 2008, 51, 3720-3730

N N

N

N

N

NN

N

O

O

MeN

N

O

Me

Me

N

S

O O O

S

O

N

1.0 1.1 0.8 1.3 1.7

0.8 1.5

Measured values of DlogP

Toulmin et al, J. Med. Chem. 2008, 51, 3720-3730

1.6 1.1

DlogP (corrected)

Vmin/(Hartree/electron)

DlogP (corrected)

Vmin/(Hartree/electron)

N or ether OCarbonyl O

Prediction of contribution of acceptors to DlogP

Toulmin et al, J. Med. Chem. 2008, 51, 3720-3730

DlogP = DlogP0 x exp(-kVmin)

logPhxd logPoct

log(

Cb

rain/C

blo

od)

DlogP

Prediction of blood/brain partitioning

R2 = 0.66 RMSE = 0.54

R2 = 0.82 RMSE = 0.39

R2 = 0.88 RMSE = 0.32

Toulmin et al, J. Med. Chem. 2008, 51, 3720-3730

What is a hydrogen bond worth?Cathepsin L inhibitors

Asaad et al, Bioorg. Med. Chem. Lett. 2009 , 19, 4280-4283http://dx.doi.org/doi:10.1016/j.bmcl.2009.05.071

Bethel et al, Bioorg. Med. Chem. Lett. 2009 , 19, 4622-4625 http://dx.doi.org/doi:10.1016/j.bmcl.2009.06.090

pIC50 values:CatL: 8.7 CatB: 5.4 CatL2: 7.4 CatS: 6.7

pIC50 values:CatL: 7.9 CatB: 5.4 CatL2: 6.4 CatS: 6.1

NHN

O

NH

O

R

O

N NNH2

pIC50 values: CatL: 8.7 CatB: 5.2 CatL2: 7.0 CatS: 7.2

Cl

NHN

O

NH

O

R

R

O

NH

O

NHN

N

R

O

NH

O

NHN

NN

Cathepsin S pIC50

Cathepsin L2 pIC50

Cathepsin LpIC50

6.9 6.1 6.6

6.2 < 5 5.5

8.1 7.2 5.9

Cathepsin inhibition pIC50 values

Bethel et al, Bioorg. Med. Chem. Lett. 2009 , 19, 4622-4625

Hydrogen Bonding in the Design Context

• Biomolecular recognition occurs in buffered aqueous media– Binding of ligand to protein can be viewed as ‘exchange’

reaction– Balanced hydrogen bonding characteristics required for

optimal interaction– Non-local effects can be important

• Multiple contacts between protein and ligand – Intermolecular hydrogen bonds in ligand-protein complex

are likely to be of less ideal geometry than hydrogen bonds between protein or ligand and water (Molecular Complexity)

Some questions to finish(mainly of interest to computational chemists)

• How relevant are van der Waals radii to hydrogen bonding?

• How physical are atom-centered charges?• Why ignore the points where electrostatic potential is

most predictive of hydrogen bond strength when fitting atomic charges to electrostatic potential?

• Is it possible to identify those atom types for which polarisability treatment would be of most benefit?

Selected references

• Abraham (1993) Scales of Hydrogen-bonding: Their Construction and Application to Physicochemical and Biochemical Processes. Chem. Soc. Rev. 22, 73-83. http://dx.doi.org/10.1039/CS9932200073

• Abraham et al (1989) Hydrogen bonding. Part 9. Solute proton-donor and proton-acceptor scales for use in drug design. J. Chem. Soc. Perkin Trans. 2, 1989, 1355-1375. http://dx.doi.org/10.1039/P29890001355

• Laurence and Berthelot (2000) Observations on the strength of hydrogen bonding. Perspect. Drug. Discov. Des. 18, 39-60. http://dx.doi.org/10.1023/A:1008743229409

• Laurence et al (2009): The pKBHX Database: Toward a Better Understanding of Hydrogen-Bond Basicity for Medicinal Chemists. J. Med. Chem. 52, 4073-4086. http://dx.doi.org/10.1021/jm801331y

• Kenny (2009) Hydrogen Bonding, Electrostatic Potential and Molecular Design . J. Chem. Inf. Model. 2009, 49, 1234-1244. http://dx.doi.org/10.1021/ci9000234

• Kenny (1994) Prediction of hydrogen bond basicity from computed molecular electrostatic potential properties. J. Chem. Soc. Perkin Trans. 2 1994, 199-202. http://dx.doi.org/10.1039/P29940000199

• Toulmin, Wood & Kenny (2008) Toward Prediction of Alkane/Water Partition Coefficients . J. Med. Chem. 51, 3720-3730. http://dx.doi.org/10.1021/jm701549s