Proseminar der Bioinformatik

1
Proseminar der Bioinformatik

• Alejandro Pironti
• Betreuer Thomas Betz
• Drug Transport Group
• 31.05.2006

2
Polar Molecular Surface Properties Predict the
Intestinal Absorption of Drugs in Humans

• By Katrin Palm, Patric Sternberg, Kristina
  Luthman and Per Artursson (1997)

3
Drug Bioavailability

• Drug bioavailability rate and extent a
  therapeutically active drug reaches systemic
  circulation and is available at site of action

4
Absorbed fraction (FA)

• Intravenously FA 100
• Orally FA decreases due to incomplete
  absorption. Bioavailabilty may decrease due to
  presystemic metabolism

5
Orally administrable drugs

• Preferred by the consumer for their convenience
• Cost of drug development USD 800 million in
  2001

6
Processes Featuring Gastrointestinal Absorption

• Passive cell diffusion
• Cell-membrane and fat solubility of drugs
• Active drug transport
• Phagocytosis / Pinocytosis
• Presystemic Metabolism
• Disintegration and dissolution of tablets

7
Passive cell diffusion

• Most drugs are absorbed from the intestine by
  passive cell diffusion
• Most drugs are absorbed in the small intestine

8
Small intestine

• Duodenum
• Jejunum
• Ileum

http//training.seer.cancer.gov/module_anatomy/uni
t10_3_dige_region4_intestine.html
9
Jejunum

• Inner wall of jejunum covered with Villi

http//www.vivo.colostate.edu/hbooks/pathphys/dige
stion/smallgut/anatomy.html
10
Passive transport
http//www.vivo.colostate.edu/hbooks/pathphys/dige
stion/smallgut/transport.html
11
Lipinskis Rule of Five

• Number of hydrogen-bond donors 5
• Number of hydrogen-bond acceptors 10
• Molecular weight 500
• LogP 5

12
Ghoses Extension of Lipinskis Work

• LogP ? -0.4, 5.6
• Molar refractivity ? 40, 130
• Molecular weight ? 160, 480
• Total number of atoms ? 20,70

13
Hydrogen Bonds

• The strength of hydrogen bonds depends on donor
  and acceptor
• OH N 29 kJ/mol
• OH O 20 kJ/mol
• NH N 12 kJ/mol
• NH O 8 kJ/mol

http//www.lbl.gov/MicroWorlds/Kevlar/KevlarClue4.
html
14
Partition Coefficient
Organic solvent 1-octanol
15
Molar Refractivity (MR)

• Size and polarizability of a molecule

n refractive index, M molecular weight, ?
density
16
Summary of the article

• Theoretical method for predicting drug absorption
• Strong sigmoidal relationship between the
  absorbed fraction and dynamic polar molecular
  surface

17
Polar Molecular Surface Area

• Area occupied by nitrogen and oxygen atoms, and
  hydrogen atoms attached to these heteroatoms
• Depends on conformation

Sulfasalazine
18
Dynamic Polar Molecular Surface Area (PSAd)

• Statistical average
• Each low energy conformation is weighted by its
  probability of existence

19
PSAd Calculation

• 1) Determination of 3D structures
• 2) Surface calculation
• 3) Statistical average of conformations

20
Determination of 3D Structures

• Conformational analysis with aid of MM2
• Molecular mechanics force fields

21
Determination of 3D Structures

• Calculate probability of existence

• Set of low energy conformers for each drug
  molecule with respective probabilities of
  existence

22
Surface Calculation

• Free surface area results from fusion of
  van-der-Waals radii
• Gauß-Bonnet theorem

23
Surface Area Calculation

• Divide surface area into polar and non-polar
• Set of low energy conformers for each drug
  molecule with respective probabilities of
  existence
• Molecular volume, polar and non-polar surface
  area for each conformation of each drug compound

24
Statistical Average of Conformers

• Dynamic total, polar and non-polar surface areas.
  Dynamic volume.
• Take data generated by previous two steps and
  calculate statistical average
• Each low energy conformation weighted by its
  probability of existence

25
Criteria for the Selection of Model Drugs

• Drug absorption 0.3 .. 100
• Lipophilicity CLogP -8.09 .. 4.50
• Charge , -, o, -
• Number of hydrogen bond donors and acceptors
• Passive transport

26
Results
27
Results

• The sigmodial equation

was fitted to the data
28
FA-PSAd Relationship

• RMSE 9.2 r2 0.94

29
FA-PSAd Relationship

• PSAd lt 63 Å2 ? FA gt 90
• PSAd gt 139 Å2 ? FA lt 10

30
Other Relationships

• PSAd-FA weaker relationship. RMSE 12.7, r2
  0.89
• NPSAd-FA no correlation observed

31
Comparison with other theoretical methods

• Ht RMSE 13.9, r2 0.87
• Ha RMSE 14.1, r2 0.87
• Hd RMSE 24.4, r2 0.60
• CLogP RMSE 31.6, r2 0.34

32
Disregards

• Differences in the strengths of hydrogen bonds
  are not taken into consideration
• Other parameters are expected to influence
  absorption e.g. charge, molecular size
• Compound selection not representative for all
  available and theoretically possible molecules

33
Disregards

• Energy minimizations performed in vacuum.
  Disregard?
• Gas phase, water, chloroform the simulation
  environment has little influence on PSAd

34
Advantages towards hydrogen- bonds counting
methods

• PSAd is a better predictor than Ht or Ha since
  intramolecular hydrogen bond formation is taken
  into account

35
Further Research Motivated by this Study

• David E. Clark
• 74 compounds with absorption data from a study by
  Wessel et al.
• 67 were classified correctly by PSAd gt 139 Å2
  criterion

36
Further Research Motivated by this Study

• David E. Clark
• 24 compounds with absorption data from a study by
  Kansy et al.
• 21 compounds classified correctly
• PSAd gt 139 Å2 criterion is rough!!!

37
Further Research Motivated by this Study

• P. Ertl et al. methodology to calculate TPSA
  (topological polar surface area)
• TPSA can be calculated 2-3 orders of magnitude
  faster than PSAd while correlating highly with
  the latter
• The calculation of TPSA does not require the
  determination of the 3D structure of the
  compounds (input SMILES strings)

38
Bibliography

• K. Palm, P. Stenberg, K. Luthman, and P.
  Artursson. Pharm. Res. 14.5 (1997).
• K. Palm, K. Luthman, A.-L. Ungell, G. Strandlund,
  and P. Artursson. J. Pharm. Sci. 8532-39 (1996).
• Wessel, M.D. et al. J. Chem. Inf. Comput. Sci.
  38726-735 (1998).
• P. Ertl, B. Rohde, P. Selzer. J. Med. Chem.
  43.20 3714-3717 (2000)
• D.E. Clark. J. Pharm. Sci. 8.8807-814 (1999)
• http//www.chemie.uni-konstanz.de/www/fb/prghlp/mm
  ads_13.htmlRef94
• http//www.iainm.demon.co.uk/spring99/davclark.pdf
  
• http//en.wikipedia.org
• http//www.radcliffe-oxford.com/books/samplechapte
  r/6916/Chap1.pdf
• http//mathworld.wolfram.com/
• Slides of the following lectures
• Computer-Aided Drug Design by Andreas Kämper
  (2005) MPII Saarbrücken
• Modelling Drug Transport with Bioinformatics
  Methods by Dirk Neumann (2005) University of
  Saarland