High throughput physicochemical profiling for drug discovery

Applications of high-throughput ADME in drug discovery

Assessment of physicochemical and pharmacological properties is now conducted at very early stages of drug discovery for the purpose of accelerating the conversion of hits and leads into qualified development candidates. In particular, in vitro absorption, distribution, metabolism and elimination (ADME) assays and in vivo drug metabolism pharmacokinetic (DMPK) studies are being conducted throughout the discovery process, from hit generation through to lead optimization, with the goal of reducing the attrition rate of these potential drug candidates as they progress through development. Because the continuing trend in drug discovery has been to access ADME information earlier and earlier in the discovery process, the need has arisen within the analytical community to introduce faster and better analytical methods to enhance the 'developability' of drug leads. Strategies for streamlined ADME assessment of drug candidates in discovery and pre-clinical development are presented within.

Pharmaceutical evaluation of early development candidates “the 100 mg-approach”

Early development candidates are often selected for pre-clinical and clinical development based primarily on pharmacological and toxicological data. In order to choose the best compounds from a biopharmaceutical point of view, physicochemical parameters such as solubility, dissolution rate, hygroscopicity, lipophilicity, pKa, stability, polymorphism and particle characteristics need to be evaluated as early as possible and above all with the highest accuracy. However, the low amounts of drug substance available in early development often compromise data quality, and therefore, hamper an early pharmaceutical assessment. This article summarises the Aventis approach on early pharmaceutical compound profiling with the aim of providing a high quality assessment requiring not more than 100 mg of drug substance. In particular, the evaluation criteria, process and miniaturised analytical technology that can be applied for this purpose are discussed.

Validation and long-term assessment of an approach for the high throughput determination of lipophilicity (logPOW) values using multiplexed, absorbance-based capillary electrophoresis

A critical evaluation of the use of 96-capillary multiplexed microemulsion electrokinetic chromatography (MMEEKC) for the indirect determination of octanol-water partition coefficients (log POW values) for a wide range of structurally different compounds is presented. The various components of the microemulsion solution were evaluated and optimized for use in a multiplexed capillary format. A six-component calibration mixture and 23 different solutes (n = 4 each) were analyzed simultaneously, providing a throughput of up to 46 samples/h, which translates to greater than a 20-fold improvement over existing indirect log POW methods. Agreement to within +/-0.5 log P units of literature values was obtained for 51 of the 54 tested neutral and basic (uncharged) solutes. A linear free energy relationship (LFER) analysis performed on the MMEEKC system supports its use as a viable and effective model of the classical shake-flask method for log POW determinations. Moreover, a standard deviation of 0.1 or less log P units was obtained for 35 of 36 solutes analyzed repeatedly over an 8-month time period, documenting the long-term effectiveness of the analysis format. Critical comparisons between the proposed MMEEKC method and existing separation methods for the indirect determination of log POW values are also made. Overall, the results indicate that 96-capillary MMEEKC can serve as a high throughput, cost effective and robust approach and as a valid model for log POW determinations.

Prediction of Human Volume of Distribution Values for Neutral and Basic Drugs. 2. Extended Data Set and Leave-Class-Out Statistics

We present an extension and confirmation of our previously published method (J. Med. Chem. 2002, 45, 2867-2876) for the prediction of volume of distribution (VD) in humans for neutral and basic compounds. It is based on two experimentally determined physicochemical parameters, ElogD(7.4) and f(i(7.4)), the latter being the fraction of compound ionized at pH 7.4, and on the fraction of free drug in plasma (fu). By regressing the fraction unbound in tissues, fut, vs the above parameters, we demonstrate the ruggedness of the method in predicting VD through the Oie-Tozer equation, via the use of several testing approaches. A comparison is also presented between several methods based on animal pharmacokinetic data, using the same set of proprietary compounds, and it lends further support for the use of this method, as opposed to methods that require the gathering of pharmacokinetic data in laboratory animals. The reduction in the use of animals and the overall faster and cheaper accessibility of the parameters used make this method highly attractive for prospectively predicting the VD of new chemical entities in humans.

Evaluation of the immortalised mouse brain capillary endothelial cell line, b.End3, as an in vitro blood–brain barrier model for drug uptake and transport studies

Well-characterised cell lines represent important tools for the study of endogenous solute or xenobiotic transport. A brain microvascular cell line, b.End3, isolated from mice transformed with the Polyoma virus middle T-antigen is available commercially. Here we report the characterisation of some features of b.End3 of relevance to its use in blood-brain barrier transport investigations. The b.End3 cells displayed a distinctive spindle-like squamous morphology in culture. Clathrin coated pits and numerous uncoated intracellular vesicles were evident within the cells, as was the expression of the vesicle-associated proteins, clathrin, caveolin-1, flotillin and dynamin II. In the presence of C6 astroglial co-culture b.End3 monolayers achieved a maximal transendothelial electrical resistance of 130 Omega cm2, but lacked real discrimination with respect to the permeation of transcellular and paracellular probes, e.g. permeability coefficients (x 10(-6) cm s(-1)) for propranolol of approximately 23 vs. 16 for sucrose. RT-PCR analysis confirmed the presence within the b.End3 cells of mRNA transcripts for the following transporters: GLUT-1; MCT 1 and 2; OAT1; Oatp1; mdr 1a and 1b; MRP 1 and 5; beta-alanine, system L and system y+L amino acid carriers; the nucleoside transporters cNT1 and 2, eNT1 and 2, and the tight junctional elements, ZO-1, JAM, occludin, claudin-1 and -5. The b.End3 cells actively accumulated D-glucose in a sodium-independent manner with characteristics consistant with that of GLUT-1. Functionality for P-glycoprotein efflux was evident as assessed by a rhodamine-123 accumulation and retention assay. The system L LAT1/4F2hc amino acid transporter was examined through uptake of L-leucine and L-phenylalanine and provided Km and Vmax values of approximately 16 microM and 350-480 pmol/mg protein/10 min, respectively; the affinity of transport for these substrates being weaker, approximately threefold, when the b.End3 cells were grown in the presence of C6 astroglial factors. Although the b.End3 cells appear unsuitable for transendothelial permeability assessments they display characteristics that would allow their worthwhile use in studies addressing blood-brain barrier transport mechanisms.

ADME-Tox in drug discovery: integration of experimental and computational technologies

Over the past ten years, in vitro experimental tools to characterize ADME-Tox profiles of compounds have been applied in early stages of the drug discovery process to increase the success rate of discovery programmes and to progress better candidates into drug development. Application of in silico ADME-Tox models has further enhanced discovery support, enabling virtual screening of compounds and thus, application of ADME-Tox at every stage of the discovery process. Ultimately, effective and efficient ADME-Tox support of discovery will depend on a complementary and synergistic use of experimental and in silico ADME-Tox.

Estimation of octanol-water partition coefficients for neutral and weakly acidic compounds by microemulsion electrokinetic chromatography using dynamically coated capillary columns

Microemulsion electrokinetic chromatography (MEEKC) using dynamically coated capillary columns is shown to be suitable for estimating the octanol-water partition coefficient (log P) for neutral and weakly acidic compounds at pH 3. The solvation parameter model is used to demonstrate that the retention properties of sodium dodecyl sulfate (1.4% w/v), n-butanol (8% v/v) and n-heptane (1.2% v/v) microemulsion are strongly correlated with the octanol-water partition system. For compounds of varied structure and log P values from 0.3 to 5.15, the correlation model is able to estimate log P to better than 0.25 log units. The dynamically coated columns consisting of a bilayer of poly(vinylsulfonate) adsorbed on top of polybrene provide a suitable electroosmotic flow at pH 3 without interfering in the retention properties of the microemulsion. For automated measurements the microemulsion run buffer should be replenished after 10 runs to maintain a stable cycle time and the coated columns replaced after 40-70 runs, depending on sample properties.

Pharmaceutical profiling method for lipophilicity and integrity using liquid chromatography-mass spectrometry

A method is described for the simultaneous profiling of sample lipophilicity, integrity, and purity. The method is rapid and is applicable to high throughput profiling of pharmaceutical properties in drug discovery. A short Polaris C(18) column is used with a rapid, wide-polarity mobile phase gradient, UV detection, and MS analysis. The lipophilicity of each component is estimated from a calibration curve using six drug or organic compounds and plotting their respective measured retention time versus LogD(7.4) (literature). The correlation of LogD(7.4) (literature) to LogD(7.4) (HPLC) for 60 structurally diverse drugs has a correlation coefficient r(2) of 0.89. The method is applicable to compounds with MW>200 and retention time>1.5 min for rapid, initial pharmaceutical profiling in drug discovery.

Structure-activity studies of a series of dipyrazolo[3,4-b:3',4'-d]pyridin-3-ones binding to the immune regulatory protein B7.1

The interaction of co-stimulatory molecules on T cells with B7 molecules on antigen presenting cells plays an important role in the activation of naive T cells. Consequently, agents that disrupt these interactions should have applications in treatment of transplant rejection as well as autoimmune diseases. To this end, specific small molecule inhibitors of human B7.1 were identified and characterized. Herein, we report the identification of potent small molecule inhibitors of the B7.1-CD28 interaction. In a high-throughput screen we identified several leads that prevented the interaction of B7.1 with CD28 with activities in the nanomolar to low micromolar range. One of these, the dihydrodipyrazolopyridinone 1, was subsequently shown to bind the V-like domain of human B7.1 at equimolar stoichiometry. With this as a starting point, we report here the synthesis and initial in vitro structure-activity relationships of a series of these compounds.

General pharmaceutics--the new physical pharmacy

General Pharmaceutics is proposed as the broad study of the biopharmaceutical and physical chemical properties of each potential drug substance. When the first quality bulk lot is delivered, usually the first GMP bulk lot, an extensive profiling of the potential drug substance should commence. This profile should include solid-state characterization as well as thorough analyses of solubility, stability, and absorption properties of the drug substance that could affect the development of a viable medicine. As a result of these studies, a number of initial specifications could be developed: the preferred polymorphic or crystalline form identified, the preferred particle size to optimize absorption/development, and an initial biopharmaceutics classification with a dose limit to identify those cases in which the formulation can be expected to improve absorption and exposure. The broad topic of General Pharmaceutics is discussed in this Minireview including many advances in technology in this field as well as the rationale behind the proposed initial specifications.

Profiling drug-like properties in discovery research

Measurement and application of compound properties for candidate selection and optimization is an emerging trend. Property-based design supplements successful activity-based strategies to produce drug-like candidates. High-throughput screening hits are evaluated for integrity and aggregation to ensure quality leads. Solubility data assures accurate activity assays and predicts absorbance. Cellular and artificial membrane permeability assays indicate compound penetration through membranes in cells, intestine and blood-brain barrier. Lipophilicity and pK(a) provide fundamental structure design elements. Stability in liver, plasma and buffer evaluates compound lifetime. Drug-drug interaction is predicted using CYP inhibition assays. Drug-like properties are vital to successful drug candidates and enhance drug discovery.

Prediction of human intestinal permeability using artificial membrane permeability

The purpose of the present study was to examine a correlation between the human intestinal permeability (P(eff)) and the bio-mimetic artificial membrane permeability corrected by the paracellular pathway model based on the Renkin function (P(PAMPA-PP-RF)) and to construct a prediction scheme. The effect of the unstirred water layer was incorporated to the prediction scheme. Eighteen P(eff) values of passively absorbed drugs were employed for the analysis. The correlation coefficient (CC) between the predicted and observed logP(eff) was 0.91. P(eff) of furosemide, hydrochlorothiazide and creatinine were underestimated by P(PAMPA-PP-RF). When these compounds were excluded, CC was 0.97. Without the correction for the paracellular pathway, P(eff) of small, cationic and hydrophilic compounds were underestimated. Therefore, P(PAMPA-PP-RF) was found to be an adequate in vitro surrogate for P(eff).

Solution and solid state properties of a set of procaine and procainamide derivatives

A set of potential Class III antiarrhythmic agents of structure p-HOOC-R-CO-NH-C(6)H(4)-CO-X-C(2)H(5)-N(C(2)H(5))(2) were isolated as crystalline solids of the amide and ester derivatives, I: succinylprocainamide (X=-NH-, R=-C(2)H(4)-); II: succinylprocaine (X=-O-, R=-C(2)H(4)-); III: maleylprocainamide (X=-NH-, R=-C(2)H(2)-) and IV: maleylprocaine (X=-O-, R=-C(2)H(2)-). Although compounds I-IV exhibit similar solution properties (i.e. acid-base speciation, with zwitterionic (+-) to neutral (00) form ratios higher than 10(4)), aqueous solubility of -NH- derivatives is significantly higher than that of -O- derivatives and also, solvent effects on solubility (i.e. the change of water by ethanol) is clearly different in both series. Solution and solid-state properties of I-IV were characterized to account for the observed differences. Results indicate that procainamide derivatives I and III crystallizes as (+-)(s) but procaine derivatives II and IV as (00)(s). Besides, I is anhydrous but II-IV are hydrates. Aqueous solubility and solvent effect on solubility are controlled by the intrinsic solubility of the species (+-) in I and III and (00) in II and IV. The rise of hydrophilicity of species (00) due to the structural change from -O- to -NH- would determine the change in the structure of the precipitating crystals from (00)(s) to (+-)(s). Solid structure (zwitterionic or neutral), as well as composition (anhydrous or hydrated) may be recognized as the main factors in determining the rank of aqueous solubility of the set: (+-)>(+-.H(2)O)>(00.H(2)O).

An intravenous formulation decision tree for discovery compound formulation development

Discovery and pre-clinical animal efficacy assessment formulation development efforts are challenged by limited compound availability and stringent timelines. The implementation and use of a systematic discovery formulation scheme can facilitate this important process. We observed that nearly 85% of Pfizer, Ann Arbor discovery compounds (n>300) submitted for discovery and pre-clinical injectable formulation development in the year 2000 could be formulated by pH adjustment, cosolvent addition, or a combination of the two approaches. Based on the vehicle data generated by this laboratory, a discovery formulation decision tree, that utilizes the solubilization approaches described above, is proposed. The proposed decision tree can be adapted and modified by pharmaceutical scientists to conform to best practices put forth by their institutions for discovery animal studies requiring injectable dosage forms.

High throughput artificial membrane permeability assay for blood-brain barrier

The recent advances in high throughput screening for biological activities and combinatorial chemistry have greatly expanded the number of drug candidates. Rapid screening for BBB penetration potential early in drug discovery programs provides important information for compound selection and guidance of synthesis for desirable CNS properties. In this paper, we discuss a modification of the parallel artificial membrane permeation assay (PAMPA) for the prediction of blood-brain barrier penetration (PAMPA-BBB). The assay was developed with 30 structurally diverse commercial drugs and validated with 14 Wyeth Research compounds. The PAMPA-BBB assay has the advantages of: predicting passive blood-brain barrier penetration with high success, high throughput, low cost, and reproducibility.

Rapid screening of pKa values of pharmaceuticals by pressure-assisted capillary electrophoresis combined with short-end injection

A method applying pressure-assisted capillary electrophoresis combined with short-end injection has been developed for the rapid screening of the pKa values of pharmaceuticals. The electrophoretic separation is performed on a short capillary length with short-end injection under an applied pressure, and the effective mobility is measured in a series of 10 different buffers with constant ionic strength (I = 0.05). The application of pressure not only reduces migration times, particularly in lower pH buffers, but also improves the repeatability of effective mobility measurements. The influence of pressure on the effective mobility was investigated at various pH values. It was observed for the first time that an increase in pressure resulted in a slight decrease in the effective mobility when the pH was above the pKa for acidic analytes, whereas an increased effective mobility with increasing pressures was observed when the pH was below the pKa. However, the observed effective mobility shift by the applied pressure did not significantly affect the determined pKa values. The determined pKa values were in good agreement with published data. Furthermore, a stacking condition was applied to increase the sensitivity, and a concentration down to 2 microM could readily be detected with UV detection using a 50 microm I.D. capillary. This technique is particularly suitable for measurement of pKa values for compounds with poor aqueous solubility. The method also omits the commonly used preconditioning steps with sodium hydroxide and water. The exclusion of excessive preconditioning steps and the use of pressure reduces the total cycling analysis time, and makes it possible to determine the pKa in less than 40 min per compound without loss of accuracy.

Prediction of volume of distribution values in humans for neutral and basic drugs using physicochemical measurements and plasma protein binding data

We present a method for the prediction of volume of distribution in humans, for neutral and basic compounds. It is based on two experimentally determined physicochemical parameters, ElogD(7.4) and f(i(7.4)), the latter being the fraction of compound ionized at pH 7.4 and on the fraction of free drug in plasma (f(u)). The fraction unbound in tissues (f(ut)), determined via a regression analysis from 64 compounds using the parameters described, is then used to predict VD(ss) via the Oie-Tozer equation. Accuracy of this method was determined using a test set of 14 compounds, and it was demonstrated that human VD(ss) values could be predicted, on average, within or very close to 2-fold of the actual value. The present method is as accurate as reported methods based on animal pharmacokinetic data, using a similar set of compounds, and ranges between 1.62 and 2.20 as mean-fold error. This method has the advantage of being amenable to automation, and therefore fast throughput, it is compound and resources sparing, and it offers a rationale for the reduction of the use of animals in pharmacokinetic studies. A discussion of the potential errors that may be encountered, including errors in the determination of f(u), is offered, and the caveats about the use of computed vs experimentally determined logD and pK(a) values are addressed.