-This presentation is meant to be a short tutorial on assays that can be reasonably implemented in academic drug discovery labs that will facilitate compound optimization in important dimensions besides potency and selectivity. -In particular, we will discuss testing of compounds to verify that they are suitable for in vivo efficacy testing in rodent disease models. In addition, there are certain basic toxicological assessments that can be performed that will help determine whether a compound has the potential to be an IND candidate.-It is assumed that the academic laboratory will have in place adequate potency assays for the target of choice, as well as assays for highly related targets (e.g., if studying a serotonin receptor, functional or binding assays to related serotonin receptors). -It is also assumed that the laboratory has a source of compounds, either through collaboration or an internal chemistry effort.-The aspects of compound characterization and optimization that I will cover in this presentation are ADME, in vitro safety toxicology assays, and rodent tolerability studies.
-Before launching into ADME and toxicology assays, it is important to establish the objectives of a drug discovery project. While “drug discovery” implies that the goal is to develop a drug suitable for human testing, in many instances the objectives are more modest, with a goal of validating that a target is truly druggable or that a modulation of a target has a desired response in a disease model (i.e., an in vivo probe compound)-An in vivo probe compound will require less characterization than a true drug candidate.-Critical aspects of an in vivo probe compound are listed.-A potential IND candidate requires more rigorous analysis, as listed.
Compound insolubility can greatly complicate both in vitro and in vivo data interpretation. It is important that compounds are soluble in the buffer systems used for in vitro assays. Similarly, compound solubility affects in vivo activity, whether delivered i.v., i.p. or p.o.-A rule of thumb is that you want compound solubility to be at least 60 ug/ml (120 uM for a MW=500 compound). You can sometimes work with compounds with slightly less solubility.-How one determines and defines solubility is somewhat dependent on the assay format.-For our lab and many drug discovery labs, solubility is determined by dilution into aqueous systems from DMSO stocks. These are typically kinetic measurements (i.e., measurements are made before equilibrium is assured).
-Kinetic solubility assays are a reasonable first approximation of compound solubility, and there are several assay methods that typically follow loss of compound absorbance after filtration of aggregates or light scattering resulting from aggregate formation.-We utilize a light scattering method using a UV-vis plate reader. We monitor multiple visible wavelengths that are longer than the absorbance wavelength of most compounds.-Absorbance is monitored as a function of compound concentration after dilution from a DMSO stock (final DMSO is 1%). The solubility can be estimated as shown on the graph, where a compound with limited solubility is depicted.
-This is not an advertisement for any particular brand, but more broadly for the valuable ADME information that can be obtained via LC-MS/MS (tandem quadrupole).-If a drug discovery lab cannot afford an LC/MS, then they should seek access to a lab that does.
-An LC-MS/MS systems allows for determination of -Compound PK -Compound brain penetration (critical for CNS active compounds) -Approximation of compound non-specific plasma protein binding and brain protein/lipid binding. These determinations allow for estimation of plasma and brain free drug levels. -Estimation of human metabolism using human liver microsome preparation, including identification of CYP450 enzyme systems involved in compound metabolism.
-On this slide I’ve provided an example PK analysis of one of our compounds.-This would be typical of a PK analysis of a compound that we have already determined to be brain-penetrant based on an evaluation of brain and plasma drug levels in 3 mice 1 hour after i.p. injection.-In this full PK analysis, the compound was administered to groups of mice both by i.v. and p.o. Plasma and brain compound levels were determined at multiple times after administration.-An integration of the total area under the curves reveals that the brain-to-plasma drug levels exceed one after i.v. and p.o. dosing. Moreover, the plasma AUC values from the oral and i.v. dosing allows for determination of the oral bioavailability.-In addition, compound clearance can be ascertained from the AUC and administered dose. The clearance is a measure of how rapidly a compound is eliminated/converted by the animal, and since most compounds are metabolized by the liver relating the compound clearance to HBF gives a sense of how metabolically stable a compound is.
-If PK indicates that brain levels are near or greater than plasma levels, then the compound is brain penetrant. However, B/P ratios significantly less than one do not always indicate an absence of compound equilibration across the BBB.-It is free (i.e., non-plasma protein bound) compound that can equilibrate across the BBB. It is possible for free compound to be fully BBB permeable and yet have a total compound B/P ratio <1 if the non-specific binding is greater in the plasma than in the brain. -If a total compound B/P of, for example, 0.3 is obtained, it is relatively straightforward to obtain an approximation of the fraction unbound in plasma and brain via equilibrium dialysis.
-Here are some example equilibrium dialysis data with the same compound that was examined for its PK on the earlier slide.-The basics of equilibrium dialysis are quite simple, and systems are commercially available. The basic idea is depicted above.-Compound amount in the two chambers is readily determined by LC-MS/MS. This particular compound shows very high PPB and BrainP/L binding.-Substituting these values into the equation above results in a B/Pfree ratio of nearly one, suggesting full equilibrium across the BBB. We generally find most of our brain-penetrant compounds show B/Pfree ratios of 0.9-1.1.
-Before a compound progresses into efficacy testing in animal models of disease, such as transgenic mice that recapitulate some disease feature, it is important to establish that the compound can be safely tolerated by the animal.-I am not talking about the rigorous testing that is required to support human clinical testing, which will be discussed by Dr. Sagartz. Rather, I’m referring to relatively simple tolerability testing that most academic labs with vivarium access can conduct.-All of our efficacy testing for neurodegenerative disease is done in transgenic mouse models. We will usually perform two types of tolerability testing before preceding to efficacy testing: MTD and repeated dose tolerability.
-Although the PK properties of a compound in research animals can be readily assessed, it is obviously not as simple to determine the PK properties in human absent clinical testing. As discussed in the introduction, information on human metabolism is not essential if the objective is only to identify an in vivo probe compound.-Human metabolism of a compound can be approximated through the utilization of mixed human liver microsome preparations. Microsomes are ER vesicles obtained after low speed centrifugation of homogenzied liver preps, and contain the CYP450 enzymes involved in the metabolism of most small molecule drugs.-The compound metabolic rate can be determined by mixing compound with microsomes. As depicted in the graph, there is a time-dependent loss of the added compound with time, as determined with LC-MS/MS. The slope of this compound decay can be used to determine the intrinsic clearance, which can be compared to the human hepatic blood flow to get a sense of relative compound stability.-The CYP450 isozymes involved in compound metabolism can also be identified from microsomal metabolism studies. A reduction of compound metabolism upon addition of specific CYP450 inhibitor identifies that CYP450 isozyme as involved in compound metabolism.
-If the objective is to identify a compound potentially suitable for human testing (i.e., IND candidate), a large number of safety and toxicology parameters will have to be assessed. Although it is not practical for an academic group to undertake all of these studies, particularly GLP studies, certain safety analyses can be conducted relatively inexpensively that will help determine whether a compound is worth considering as a potential drug candidate.-One relatively simple set of analyses that can be performed inexpensively is CYP450 inhibition profiling. Commercial kits are available for this, although it can also be done using baculozomes (insect cells microsomes) and LC/MS.-Most drugs are metabolized by a relatively small number of CYP450 isozymes. It is undesirable to significantly inhibit these as it can result in altered metabolism of any other drugs that a patient is taking (i.e., DDI). -The extent to which CYP450 inhibition will prove problematic depends on the strength of interaction the compound with the CYP, and the relative concentration of the compound in the blood at efficacy doses. This is difficult to know without human data in hand, although animal efficacy data coupled with PK can give some guidance.-We will flag compound if there is significant CYP inhibition at 10 uM, recognizing that this may not be a knockout depending on required blood levels.
-Another human safety assay that most labs can run is hERG binding.-hERG is a cardiac potassium channel to which many drugs bind. Inhibition of hERG can result in long QT intervals and potentially fatal long QT syndrome.-We use a commercial FP assay to assess compound hERG binding. Data suggest that the results from hERG binding assays correlate fairly well with results from the more definitive patch-clamp analyses.-As with CYP450 inhibition, the extent to which a compound interaction with hERG is problematic will depend on the required drug concentration in blood and the binding constant to hERG. The FDA is quite conservative on hERG binding, and thus there needs to be an appreciable safety window.
Session 2 part 2
Overview• Discuss systems and assays that can be reasonably implemented by academic groups for CNS drug discovery.• Assumes existing target-specific potency and selectivity assays (e.g., related receptors or enzymes).• Discussion topics: – ADME (Absorption, Distribution, Metabolism and Excretion) • Solubility • Pharmacokinetics • Metabolism – Toxicology • In Vitro assays • Rodent tolerability studies
Probe Compound vs. Drug Candidate• The extent of compound characterization will be dictated by whether the molecule is being developed as a research “probe” for POC in animal models or as a potential IND candidate. – Probe Compound Requirements: • Adequate ADME properties for valid in vivo assessment (solubility, clearance, half-life) • Demonstration of adequate (free) drug levels in brain. • Tolerability in animal species of choice (mouse/rat) at projected efficacy doses. • Evidence that compound metabolism doesn’t change upon repeated dosing. – IND Candidate Requirements: all of the above, plus • In vitro safety pharmacology, including hERG and human CYP450 inhibition profiling. • Human microsome studies to determine predicted clearance, CYP metabolism, and ideally preliminary metabolite identification. • Ultimately, IND-enabling studies, including GLP respiratory, CV, and safety toxicology in two species.
Compound Solubility• Compound solubility is affected by many factors (salt forms, pH, buffer systems, etc.) – Typical objective in academic drug discovery is to ensure compounds are sufficiently soluble in assay buffer systems to allow interpretable results.• For animal studies, need sufficient solubility to allow adequate dosing.• Generally want solubility >60 μg/ml (Lipinski et al., Adv. Drug Disc. Rev. 23:3-25).• Two basic types of solubility determinations – DMSO stock dilutions: measure compound precipitation (typically a kinetic measurement) – Solid compound: measure compound dissolution (typically an equilibrium measurement)• We typically conduct kinetic solubility measurements of compounds dissolved in DMSO.
Compound Solubility• Multiple simple kinetic solubility methods exist (e.g., see Pan et al., J. Pharm. Sci. 90:521-29 and Hoelke et al., Anal. Chem. 81:3165-72).• One method accessible to most labs is solubility determination based on light scatter of precipitates using a UV-Vis plate reader. 0.45 Sum Abs 550,600,650,700 0.40 quercetin 1 0.35 quercetin 2 0.30 0.25 quercetin 3 nm 0.20 0.15 0.10 0.05 0.00 -0.05 -6 -5 -4 -3 log conc (M)
ADME & LC-MS/MS• Key aspects of ADME are enabled with a LC- MS/MS – Compound plasma pharmacokinetics, including clearance and half-life determinations. – Compound brain penetration. ~98% of compounds do not equilibrate across the BBB. – Estimation of free drug levels in plasma and brain through equilibrium dialysis studies. – Approximation of human metabolism using liver microsomes, including identification of major CYP450 isozymes involved in compound metabolism.
Compound BBB Penetration• A B/P ratio of ≥1 is indicative of full BBB penetration, but compounds with B/P <1 might still equilibrate across the BBB.• It is free (unbound) compound that crosses the BBB. At true equilibrium, B/Pfree =1, where B/Pfree = B/P x fu(brain))/fu(plasma)• The unbound fraction in plasma and brain can be approximated by equilibrium dialysis.
Projected Dosing Effect of Unbound Fraction Dose = [(Cl x Ct x T)/F]/fu Where Cl = Clearance (L/h) Assuming: Ct = Target drug level Ct = 50 nM T = Dosing interval (hours) Drinking interval = 1 hour F = Bioavailability Mouse Weight = 25 g Compound 51362 51397 CL,brain (mL/hr) 66 50 CL, plasma (mL/hr) 108 78.6 fu,brain 0.011 0.043 fu,plasma 0.019 0.069 F 0.64 0.63Predicted Daily Dose (mg/kg) 160 30
Rodent Tolerability Studies• Dr. John Sagartz will provide a more detailed overview of animal toxicology.• Academic research centers are typically equipped to conduct rodent efficacy studies and non-GLP preliminary rodent toxicological assessments.• Our laboratory will typically conduct two types of mouse tolerability studies for novel lead compounds that appear to have appropriate potency and ADME properties. – Tier 1: Maximum Tolerated Dose (acute) – Tier 2: Repeated Dose Tolerability
Rodent Tolerability Studies• MTD Design – Normal mice (n=4) are dosed via oral gavage (0.5% methylcellulose) starting at 1-3 mg/kg, with 3X dose-escalation – Dosed every two days until two or more mice show signs of intolerance (altered locomotor activity, sedation, ataxia, hypo- or hypertonia, salivation or excitation).• 1-Month Tolerability – Normal mice (n=6/dose) are dosed at 0.1x-, 0.3x- and 1x-MTD via oral gavage for 2 weeks or 1 month. – Assessments include • Behavioral observations • Body weights • Organ weights at study completion • Complete blood counts at study completion • Plasma and brain compound levels at study completion (compared to separate group receiving drug for 3 days)
Human Liver Microsomes• Drug metabolism: – Phase I (oxidation) - CYP450 isozymes and FMOs – Phase II (conjugation) - UDP-glycosyltranferases, glutathione transferases and sulfotransferases.• Human metabolic rate can be estimated through use of human liver microsomes (contain CYP450s, FMOs and UGTs). – ~80% of drugs metabolized by CYP450s.• Compound-metabolizing CYP450 isozymes can be indentified with liver microsomes. – Significant CYP2D6 metabolism is a flag due to allelic variation in humans (5-10% of Caucasians are poor 2D6 metabolizers).• Compound metabolites can be identified with microsomes and LC/MS-MS. CNDR-51362/hLM with CYP3A4 10.8 Inhibitor ln(Peak Area) 10.6 10.4 y = -0.004x + 10.74 10.2 10 y = -0.013x + 10.65 9.8 0 20 40 60 51362 + hLM 51362 + KETO + hLM Minutes Clint = ke (slope) X (mg microsomes/g liver) x (g liver/kg BW) x (1/ mg/ml microsomal protein) Human HBF = 20 ml/min/kg Compound CL = 2.15 ml/min/kg (11% HBF)
In Vitro Safety/Toxicology CYP450 Inhibition• CYP450 inhibition profiling can be assessed via – LC-MS/MS (measure inhibition of known CYP450 substrates using baculozomes). – Commercial assay kits (e.g. Invitrogen Vivid fluorescent kits).• Although there are >50 CYP450 isozymes, most drugs are metabolized by just 7 isozymes (3A4/5, 2C9, 2D6, 2C19, 1A2, 2C8 and 2E1).• Significant inhibition of major CYP450 isozymes can result in drug-drug interactions.• FDA guidance suggests that in vivo CYP450 inhibition studies are needed if plasma Cmax is >10% of CYP450 Ki.• We flag compounds if CYP450 inhibition is >75% at 10 μM.
In Vitro Safety/Toxicology• At least 11 drugs have been withdrawn from the U.S. as a result of their likely hERG cardiac channel inhibition (long QT intervals).• Compound interaction with the hERG channel can be readily estimated with commercial ligand binding kits (e.g., Invitrogen Predictor hERG FP assay kit).• hERG Binding assays have generally good correlation with more definitive patch-clamp analyses. Want ~100-fold window between effective plasma drug concentrations and hERG Ki/IC50. We flag compounds that show >50% inhibition at 10 μM.
Conclusions• It is important that compounds intended for use in animal efficacy models undergo sufficient characterization to ensure adequate exposure at doses that are well tolerated.• Academic labs can also perform preliminary safety pharmacological assessments at relatively little expense which will provide important information about whether a compound (or compound series) has drug candidate potential.• Academic labs should seek CRO assistance in planning and conducting IND-supporting studies.
Acknowledgements CNDR Drug Discovery CNDR Drug Discovery(Biology/Pharmacology) (Chemistry) Andrea Asimoglou Amos B. Smith, III Jenna Carroll Carlo Ballatore Alex Crowe Francesco Piscitelli Julia Durante Longchuan Huang Edward Hyde Michiyo Iba Michael James CNDR Directors Katie Robinson Virginia Lee Laurel Vana John Trojanowski Sharon Xie Mandy Yao Drug Discovery Bin Zhang Director Kurt BrundenFunding Sources: NIA/NIH, Astra-Zeneca, Bristol-Myers Squibb, AHAF CART, Marian S. Ware AD Foundation, Nathan Bilger Alzheimer Drug Discovery Initiative
IND-Enabling Safety Evaluation John E. Sagartz, DVM, PhD, DACVP President, Seventh Wave Laboratories LLC
“What is there that is not poison? All things are poison and nothing (is) without poison. Solely the dose determines that a thing isPhilippus Aureolus Theophrastus Bombastus von Hohenheim not a poison.” Paracelsus (1493 - 1541)
“The Poison Squad” „Hygenic Table Studies‟ initiated December 20, 1902 US Congress appropriated funds “Whether preservatives should ever be used or not, if so, what preservatives and in what quantities?” Teams of 12 ate food containing larger doses of „preservatives‟ such as borax, sulfuric acid and formaldehyde Preceded pure Food and Drugs Act (1906) and 1938 Food Drug and Cosmetic Act
Nonclinical Safety Assessment• International Conference on Harmonization (ICH) Guidelines – Harmonization of expectations between Europe, Japan, and US – Expectation for evaluation of • Genetic toxicity: ICH S2 • Safety pharmacology: ICH S7A,B • General toxicology with exposure assessment: ICH M3(R2), ICH S3
ICH S2: Genetic Toxicity• Genetic toxicology studies include in vitro and in vivo systems• Evaluate the potential to induce mutations and chromosomal damage – bacterial mutation – cytogenetics – mammalian gene mutation
Genetic Toxicology• “Standard Battery” for Genotoxicity – A test for gene mutation in bacteria • Ames assay (bacterial reverse mutation assay) • Identifies comparatively subtle effects on chromosomes (point mutations, substitutions, frame shifts) • Tester strains of E. coli and Salmonella typhimurium – An in vitro test with cytogenetic evaluation of chromosomal damage with mammalian cells (Chromosomal aberrations assay) – An in vivo test for chromosomal damage using rodent hematopoietic cells