![Identifying CNS drugs requires unique considerations beyond efficacy BIOAVAILABILITY – drug available in the body to act at target Inability to reach target in sufficient amounts during appropriate time window LIMITS opportunity for efficacy – BBB, metabolism, efflux Caveat: Bioavailability DOES NOT guarantee drug efficacy STARTING POINT: How does an oral drug get into the CNS? Quantification LogBB = comparison of brain, plasma concentrations Relative bioavailability %F = [AUC po ] / [AUC iv ] Molecular properties influence how drugs are absorbed, how they are distributed, how they interact with transporters and metabolizing enzymes Absorption Metabolism Tissue Distribution Time [Drug]](https://image.slidesharecdn.com/session1part2-120227111146-phpapp01/85/Session-1-part-2-1-320.jpg)
![Synthetic Chemistry Essentials for Biologists February 2012 Heather Behanna, PhD Biotechnology Research Associate [email_address] (312) 768-1795](https://image.slidesharecdn.com/session1part2-120227111146-phpapp01/85/Session-1-part-2-16-320.jpg)
Session 1 part 2
- 1. Identifying CNS drugs requires unique considerations beyond efficacy BIOAVAILABILITY – drug available in the body to act at target Inability to reach target in sufficient amounts during appropriate time window LIMITS opportunity for efficacy – BBB, metabolism, efflux Caveat: Bioavailability DOES NOT guarantee drug efficacy STARTING POINT: How does an oral drug get into the CNS? Quantification LogBB = comparison of brain, plasma concentrations Relative bioavailability %F = [AUC po ] / [AUC iv ] Molecular properties influence how drugs are absorbed, how they are distributed, how they interact with transporters and metabolizing enzymes Absorption Metabolism Tissue Distribution Time [Drug]
- 2. Case study: Antihistamine CNS bioavailability changes impact adverse events First-generation antihistamines characterized by sedative side effects Undesirable feature!! Second-generation antihistamines lack drowsiness properties Better safety index DIPHENHYDRAMINE FEXOFENADINE Brain penetrant Avoids penetrating CNS Antihistamines lacking sedative properties tend to possess limited CNS bioavailability compared to antihistamines with drowsiness Obradovic T et al. (2007) Pharm Res, 24 , 318-327. Avoids P-glycoprotein efflux P-glycoprotein substrate
- 3. Case study: CYP2D6 metabolism alters bioavailability, impacts safety/efficacy CYP2D6 - major isoform involved in CNS drug metabolism! Genetic polymorphisms affect CYP2D6 expression, function CYP2D6 phenotype correlates with disease progression in breast cancer Morphine toxicity risk with UM phenotype; Poor efficacy with PM phenotype CODEINE MORPHINE CYP2D6 “ Ultra-rapid” metabolizer phenotype “ Poor” metabolizer phenotype Increased CYP2D6 function Decreased CYP2D6 function TAMOXIFEN 4-HYDROXY TAMOXIFEN CYP2D6 “ Ultra-rapid” metabolizer phenotype “ Poor” metabolizer phenotype Increased CYP2D6 function Decreased CYP2D6 function
- 4. Bioavailability…it’s a big deal! So, what can you do to find compounds that are bioavailable? Hint: you don’t need to do in vivo testing just yet..
- 5. Molecular Properties 101: Physical properties influence how drugs interact with the body Solubility, lipophilicity, size impact ADME outcomes Absorption : Will the drug penetrate across the GI tract to the circulatory system? Distribution : Will the drug remain soluble in the blood? Will it remain bound to plasma proteins? Metabolism : Will the drug be chemically modified by CYPs? How much will be available to get to the target? Excretion : How will the body eliminate the drug? *Modifying one property has consequences on others Figure modified from van de Waterbeemd H. (2009) Chem Biodiv, 6 , 1760-1766. SOLUBILITY Charge Ionization Dissolution LIPOPHILICITY SIZE H-Bonding Shape Amphiphilicity Charge Distribution LogP MW PSA
- 6. Improving the odds: Using properties guidelines can increase bioavailability odds “ Rule of 5” - Christopher Lipinski Poor absorption/permeation MORE LIKELY if: >5 Hydrogen bond donor atoms (HBD) MW > 500 LogP > 5 N + O > 10 1990s: analyses used to identify ways to improve attrition due to poor bioavailability Today = Smarter screening platforms CAVEAT: The Ro5 is NOT CNS specific! Gleevec (imatinib) LogP 2.89 MW 493.6 PSA 86.28 HBD = 8 N + O = 8 Norvir (ritonavir) LogP 2.33 MW 720.6 PSA 202.26 HBD = 11 N + O = 11
- 7. CNS drug discovery properties analysis What molecular properties are most relevant to CNS? LogP – lipophilicity, solubility in octanol/H 2 O MW – size PSA – polar surface area (N’s, O’s) How do I calculate these? Experimental pION* www.pion.com CEREP www.cerep.fr Protocols – “ home grown ” In silico – calculate estimated values derived from real structures ACD/Labs* Schroedinger ChemAxon* *Discounts available for academics DISCOVERY TIP: Prior to purchasing or screening libraries – look at the property landscape. How much is CNS relevant?
- 8. CNS drug discovery properties analysis – what are “ good ” values? CNS drugs occupy a more restricted molecular properties space Properties guidelines also depend on development status (hit versus lead versus drug) Rees et al. (2004) Nat Rev Drug Discov, 3, 660-672. Lipinski CA et al. (2001) Adv Drug Deliv Rev, 46, 3-26. CNS Drugs LogP < 4 MW < 400 PSA < 80 Chico et al. (2009) Nature Rev Drug Discov , 8, 892-909. Fragments LogP < 3 MW < 300 PSA < 90 Oral Drugs LogP < 5 MW < 500 PSA < 140
- 9. Case Study: CNS properties analysis identifies guidelines Properties were computed using ACD Labs (v.11). Data shown are mean±SEM. Student’s t-test used to compare mean values with CNS means. *, p <0.05; ***, p <0.001. Chico et al. (2009) Nature Rev Drug Discov , 8, 892-909. PSA discriminates CNS+ better than LogP Pgp+ compounds possess higher LogP, MW than Pgp- compounds
- 10. Case study: Properties guidelines help prioritize CNS drug discovery efforts Simple properties filters helped prioritize the top 6% of candidates! <100 compounds were synthesized from start lead clinical candidate. Wing et al. (2006) Curr Alz Res, 3, 205-214. Chico et al. (2009) Drug Metab Dispos, 37, 2204-11. Chico et al. (2009) Nature Rev Drug Discov, 8 , 892-909. 5 amines + 18 alkyl/aromatic groups = 1700+ possibilities PSA <80Å 2 MW <400 LogP < 4 (80%) (80%) (80%)
- 11. Case study: Overlapping properties analyses focuses discovery efforts Most property analyses focus on one outcome or endpoint… … but CNS bioavailability involves multiple outcomes (penetration, metabolism for example). CNS+/CYP2D6- = good! CNS+/CYP2D6+ = bad! Future direction of the field – perform properties analysis on multiple outcomes and “overlap” results Query: where are we most likely to find compounds that are both CNS+ AND CYP2D6-? Approach: Superimpose properties to find “hotspots” associated with CNS+/CYP2D6- candidates Chico et al. (2009) Drug Metab Dispos, 37 , 2204-11. Chico et al. (2009) Nature Rev Drug Discov, 8 , 892-909.
- 12. Find the “sweet spot” of CNS+/CYP2D6- using overlapping analyses CNS+/CYP2D6+ Avoid this region CNS+/CYP2D6- Minimized risk of CYP2D6 involvement, but still have CNS+ CNS+ PSA ≤ 80Å 2 LogP ≤ 4 MW ≤ 400 Database summary statistics:
- 13. Multidimensional properties analyses helps refine “CNS” space Wager et al. (2010) ACS Chem Neurosci, 1 , 420-434. Wager et al. (2010) ACS Chem Neurosci, 1, 435-449 Analyzing properties associated with multiple ADME features helps identify more restrictive guidelines, increases probability of finding CNS+ compounds.
- 14. Takeaways – how can I use properties guidelines in my discovery efforts? Library screening/selection Properties can help you focus screening on most “CNS”-relevant members. Some libraries are more CNS friendly than others. Hit-to-lead refinement It is easier to add than subtract later! Start low – expect to increase as you proceed Applying guidelines allows chemists to budget their selections Guidelines are guidelines – NOT rules Don’t get tripped up by numbers. Rationale trumps rules!! Resources Experimental pION www.pion.com CEREP www.cerep.fr In silico ACD/Labs Schroedinger ChemAxon CNS LogP < 4 MW < 400 PSA < 80 Fragments LogP < 3 MW < 300 PSA < 90 Oral Drugs LogP < 5 MW < 500 PSA < 140
- 15. Thank you for your time
- 16. Synthetic Chemistry Essentials for Biologists February 2012 Heather Behanna, PhD Biotechnology Research Associate [email_address] (312) 768-1795
- 19. An overview of the drug discovery process Nature Review Drug Discovery,8, 892 2009.
- 20. The Drug Discovery Chemist Synthetic chemistry- How to make things Medicinal chemistry- What makes a drug Pattern recognition and recall
- 21. Pattern recognition and recall TNT Salinsporamide – clinical trials for cancer Point of covalent attachment to proteins Azo-blue
- 22. Chemical space versus drug-like space Lipinski, C and Hopkins A, Nature , 2004 , 432(16) 855. Nature Biotechnology 24, 805 - 815 (2006)
- 23. Scaffolds for drug design Core structures (scaffolds) tend to be heterocycles Rings (that can be involved in stacking and hydrophobic interactions Heteroatoms (non-carbon atoms) for potential hydrogen bonding interactions Heterocylces can interact with proteins through both hydrogen bonds and hydrophobic factors Scaffolds must have synthetic “handles” Accessible chemistry
- 24. Properties of scaffolds Some scaffold changes or substitutions will drastically affect activity Privileged scaffolds Viagra Levitra No serotonergic and dopaminergic activity Strong M1 receptor ligand The scaffolds of some drugs can be modified without changing the mechanism of action Might show changes of ADME properties
- 25. An overview of the drug discovery process Looking for a starting point – either binding or weak activity that can then be optimized Obtainment of a Hit
- 26. How to get a hit? High throughput screening Screen a library for activity against a target or phenotype Traditional assays Adaption of patented compounds or natural products Test for some activity and against others Fragment screening Screen for binding to a target (may not have activity) Biophysical methods
- 27. High throughput screening (HTS) Advantages: Ability to screen hundreds of thousands of compounds in weeks Automated systems Novel in-house libraries Disadvantages Limited to chemical space in the library Lead to discovery of “red flag” compounds Generally larger than “optimal” leads
- 28. HTS pitfalls - Bad Hits and Frequent Hitters J Chem Inf Model. 2007 Jul-Aug;47(4):1319-27. Pattern recognition and recall Compounds that are potent in HTS are not necessarily Hits!
- 29. Adaption of natural products Genistein – natural product shown to have promise for: Cancer (topoisomerase inhibitor) Cystic Fibrosis (CFTR corrector) Anthelmintic (inhibits glycolysis) Tumor metastisis (MEK4) US 2010/0137425 A1
- 30. Fragment based approach Fragments consist of Low MW Low LogP High ligand efficiency (binding energy per atom) Combination of hydrophobic and H-bonding properties Fragments are screened for binding to a target Expanded to gain efficacy Structure assisted Nature Reviews Drug Discovery 3, 660-672 (2004) Curr Top Med Chem 7, 1600-1629 (2007); Current Topics in Medicinal Chemistry, 5, 751-762 (2005)
- 31. How can we do that?
- 32. Hit criteria Regardless of how a hit is generated, it must pass certain criteria Show potency in cell assays Precursor to a drug, not just a ligand! Show potential chemical handles for structure modification Possess certain ADME properties Quality of the library will strongly influence the chance of finding drug-like suitable hits Fragment libraries tend to have better properties as hits than HTS libraries Library properties should be considered Interdisciplinary teams are best for hit evaluation Not all active compounds are worth pursuing as a drug Certain compounds come with “red flags”
- 33. An overview of the drug discovery process “ Hit to Lead” Nature Review Drug Discovery,8, 892 2009.