sign in sign up
<<
Home , Drug development, Drug discovery, Hit to lead

Hit to lead: 19/4/07 19:09 Page 9

Lead optimisation

‘Hit’ to ‘lead’ and ‘lead’ to ‘candidate’ optimisation using multi-parametric principles

As the numbers of potential target become available for pharmaceutical companies of all sizes, the drive towards more cost-effective discovery assumes greater significance. It is a truism to state that the biggest bottlenecks in the process are the ‘hit’ to ‘lead’ and ‘lead’ to ‘candidate’ phases of lead and candidate optimisation. One philosophical approach to the lead optimisation phase that is gaining wider acceptance as a methodology for improving this process is that of parallel optimisation. This article focuses on how to approach the issues of high throughput investigation of the physical and physicochemical properties such that data can be acquired in real time together with and selectivity.

he genomics and proteomics revolution of The process of taking a screening ‘hit’ to a ‘lead’ By Dr Anthony D. the last decade has given unprecedented to a ‘candidate’ is to a large degree dependent upon Baxter and Dr Peter Taccess to new target proteins to pharma- the definition of ‘hits’, ‘leads’ and ‘candidates’ in M. Lockey ceutical companies of all sizes, either through their each research organisation, and while there is own efforts or in collaboration with academics or broad agreement with the definitions there are still specialist biotech companies. This fact, combined some considerable differences – especially when with technological advances in high throughput one considers at which stage of the process a result screening (HTS), rapid assay development and is apposite. For the purposes of this article and high throughput parallel synthesis and compound using our in-house terminology a ‘hit’, ‘lead’ and husbanding has meant that for each new ‘candidate’ can be defined in Table 1. We believe target, a plethora of ‘hits’ can be reasonably that candidate compounds evolve over time with expected from an HTS campaign. Herein lies the various improvements in the indicated properties problem, for the next phase of is and we break down the classification of candidates the optimisation of the properties of these ‘hits’ into Bronze, Silver and Gold to reflect their relative such that an exploratory development ‘candidate’ value to an organisation. It should be noted that compound or compound series can be identified. these definitions are by no means fixed but are This phase represents a multi-dimensional techni- variable from project to project and compound cal challenge that will prove to be resistant to sim- series to compound series. The figures in brackets plification and automation. In short, the industry are average timelines for progression from one can identify ‘hits’ for many new targets with rea- stage to another and are again highly variable. sonable efficiency, but the next phase of lead opti- It is a truism to say that the industry has in the misation represents the new bottleneck for efficient past been too focused on the optimisation of drug discovery (Figure 1). potency to the exclusion of the optimisation of

Drug Discovery World Winter 2001/2 9 Hit to lead:Hit to lead 19/4/07 19:09 Page 11

Lead optimisation

Ta bl e 1

Suggested definitions of hits, leads and candidates and activities to be successfully completed to attain each of the project stages

HIT LEAD BRONZE SILVER GOLD CANDIDATE CANDIDATE CANDIDATE (6 MONTHS (12 MONTHS (18 MONTHS (24 MONTHS FROM HIT) FROM HIT) FROM HIT) FROM HIT)

G 1° screen established G 1-10 µM potency G 100 nM activity G 10 nM activity G <10 nM potency G Compounds screened achieved achieved achieved achieved G IC50s established G SAR established G Activity in screen with G Good selectivity (500- G Known mechanism of G Structures verified G Selectivity measured RSA/HSA 1,000 times) versus interaction with target G Minimum of three G Competitivity G Viable selectivity (30- related family targets protein independent chemical established 50x) versus related G Any problems of G Required selectivity series to evaluate G Mechanism of activity family target(s) specificity lost with (up to 1000x) vs G Positive PK (reversible/irreversible, G functional latest compounds (‘in related target(s) data kinetics) established activity in 2° screen(s) house’ testing on G Ideal specificity profile G Specificity of key G Solubility OK on lead relevant target) on main candidate(s) compound(s) from compound G Required solubility including HERG each lead series G Physicochemical characteristics on lead channel against selected properties compound(s) G Required solubility number of receptors/ (experimental) G In vitro PK data characteristics on enzymes G CYP 450 inhibition/ confirmed with in vivo candidate G Preliminary PK in vivo induction potential PK study data compound(s) (rodent) to establish known G In vitro functional G In vivo activity in benchmark for in vitro G Patent filed activity in 2° screen(s) functional model SAR increased wrt potency G In vivo safety studies G In vitro PK data good G Satisfactory in vitro (CNS/CVS profiles) predictor for in vivo G In vitro mutagenicity activity characteristics in (Ames test/ screening assay clastogenicity) G Patent completed G Patent published G Structurally distinct back-up class identified

other biological (selectivity), physical (solubility) ally be improved by reducing molecular weight, and physico-chemical (pharmacokinetic (PK) but potency by increasing molecular weight, related) based properties. As a result many com- medicinal chemists caught with optimising a sub- pounds have been elevated to candidate status nanomolar compound with all the features guaran- which in retrospect should not have been. Their teeing non-absorption, are caught in a cul-de-sac. consequent failure in pre-clinical or clinical stud- A typical project will see several promising series of ies raises the attrition rate of compounds pro- compounds failing at these latter hurdles and the gressing through the discovery pipeline with obvi- requirement to develop a subsequent series. Such ous cost and time penalties. ‘serial’ lead optimisation can therefore be lacking The traditional project management of ‘hit’ to in efficiency (Figure 2). ‘lead’ and lead optimisation has more recently been Consequently, an alternative strategy of simulta- that of performing the seductive optimisation of neously optimising all the properties necessary for potency, followed by an assessment of pharmaco- the drug at the outset is a rational one. Hence, the kinetic and toxicological parameters. Medicinal now commonly accepted ‘best practice’ for lead chemists then are caught in a trap of trying to opti- optimisation is that of ‘parallel’ optimisation of mise a feature like absorption, while maintaining potency, selectivity, physical property measurement high potency. A typical problem is that of potency and PK related properties (Figure 3). We refer to versus absorption – given that absorption can usu- this approach as ‘multi-parametric optimisation’.

Drug Discovery World Winter 2001/2 11 Hit to lead:Hit to lead 19/4/07 19:09 Page 12

Lead optimisation

Such an approach puts a more equal value on the Absorption data generated for all the above properties and the In recent times, the efforts of parallel synthesis net result of this method gives compounds that are chemists in producing large numbers of com- more rounded in terms of their overall characteris- pounds has been tempered by the suggestions of tics and more likely to pass the first pre-clinical physical chemists that the ‘drug-like’ nature (or and clinical investigations of their utility. The aim not) of the compounds is of equal value. The work is to improve the efficiency of the lead optimisation by Chris Lipinski1 and the wide acceptance of the process and reduce the attrition rate of the process ‘rule of 5’ is testament to the importance of using downstream of this position. appropriate design criteria for production of such However, it should be stated that multi-para- lead discovery libraries2. The rule of 5 (compounds metric optimisation is not without its own issues should not: a) exceed a molecular weight of 500, b) and problems. Firstly, while synthesis, analysis have a cLogP of less than 5, c) have more than 5 and screening can be geared up to be considered H-bond donors d) have more than 10 H-bond ‘high throughput’, this is not the case in the meas- acceptors is in large part a prediction that a com- urement of some of the PK-related properties pound satisfying its dictates will have a small such as S9 microsomal assays and Cytochrome enough polar surface area (PSA) to efficiently pas- P450 induction or of physical chemistry measure- sively diffuse across a cell membrane and thus be ments such as solubility. It is clearly important effectively absorbed across the gut surface mem- that such data can be sourced in synchrony with branes and into the blood stream. More recently, the other information acquired. Secondly, with we and others3 are routinely predicting such each compound in the lead optimisation process behaviour using experimental models of absorp- now generating many more data points from the tion. At its simplest, the use of immobilised artifi- different ‘assays’, many of which are duplicated, cial membranes (IAM) and the parallel artificial there is a real danger of information overload. membrane permeability assay (PAMPA) systems There is a requirement for software packages that can predict passive diffusion across cell mem- assimilate the data from this exercise and portray branes. Caco-2 cells, a human colon adenocarcino- the information in easy to digest and preferably ma cell line, are a closer approximation of the in graphical formats. Given that lead optimisation vivo situation. Caco-2 cell monolayers have been is, by its very nature, an iterative process where- used to examine compound absorption by model- by such information is generated, assimilated, ling the epithelial cell layer barrier and absorption interpreted and then translated into further action from the gut lumen to the blood stream. Typical then it is imperative that the process does not studies determine transport in the apical to baso- itself cause a bottleneck. lateral direction in vitro, although comparison Each of the following properties associated with with transport in the basolateral to apical direction in vitro PK (absorption; , solubility) is can provide indications of the involvement of now discussed with reference to their relative active drug efflux mechanisms, eg P-glycoprotein Figure 1 importance to the drug discovery process. (Pgp) activity.

The new bottleneck in drug discovery

Disease Target Screen Hit Lead Candidate

12 Drug Discovery World Winter 2001/2 Hit to lead:Hit to lead 19/4/07 19:09 Page 14

Lead optimisation

Serial lead optimisation strategy

Optimisation of potency/selectivity In vitro PK Candidate for Candidate for in vitro PK in vivo PK

Time

Figure 2

Multi-parametric optimisation strategy

Synthesis/analysis

Screening/testing

Lead In vivo candidate

In vitro PK

Phys. prop. meas.

Throughput Figure 3

14 Drug Discovery World Winter 2001/2 Hit to lead:Hit to lead 19/4/07 19:09 Page 15

Lead optimisation

Bovine brain microvascular endothelial cells throughput manner it is only practical to use tur- References (BBMEC) can be used in an analogous manner bidometric methods to measure kinetic solubility1 1 Lipinski, CA, Lombardo F, to predict Blood Brain Barrier (BBB) penetra- utilising robotic technologies. In this way, it is pos- Dominy, BW, Feeney, PJ. Adv. Rev. 2001, 46, 1. tion, as part of the optimisation of potential sible to determine aqueous solubilities with reason- 2 Baxter AD. Current Opinion CNS or in other projects where CNS pen- able accuracy and in a throughput comparable Chem. Biol. 1997,1, 79. etration is deleterious. with the above early ADMET screens. It is sug- 3 Balimane PV, Chong S, gested that an aqueous solubility of at least 1mg/ml Morrison RA. J. Pharmacol. Metabolic stability is required for an orally dosed candidate. Toxicol. Methods 2001, 44, 301-312. Metabolic stability is an important factor affecting In summary, the concept of multi-parametric the progression of potential lead compounds. In optimisation is now being embraced throughout the vitro metabolism studies, using liver S9 or micro- pharmaceutical and industries and it some subcellular fractions, are widely used in the is envisaged that this approach will yield positive optimisation of compound selection. The relative results by increasing the speed of candidate discov- stability of compounds can be measured as loss of ery while reducing the attrition rate. DDW parent compound by HPLC-MS analysis. With addition of the appropriate enzyme cofactors the phase2 conjugation pathways of a drug, eg glu- curonidation, can also be predicted.

Cytochrome P450 enzyme induction and inhibition The dependent group of enzymes are the principle enzymes involved in the oxidative metabolism of drugs. In response to cer- tain drugs CYP450 enzyme levels, particularly in the liver, may increase (are induced) several-fold to cope with the increased insult to the tissues. P450 enzyme induction is thus an early marker of the possible toxic side-effects of a drug. P450 expres- sion can be monitored by determining protein mass, enzyme activity or message levels. Argenta has developed a high throughput method to evalu- ate P450 induction by drugs as part of its drug optimisation process. Inhibition of the cytochrome P450 enzymes can also lead to clinical drug-drug interactions. P450 enzyme inhibition can be assessed by performing in vitro inhibition studies using recombinant P450 enzymes and high throughput 96-well plate fluoro- metric assays. Among the companies that offer Dr Anthony Baxter is CEO of Argenta Discovery such metabolism studies as a service are Cerep SA Ltd and was one of the founders of the company (France) and Cyprotex (UK). Lion Biosciences since its inception in 2000. Before Argenta, Dr (Germany) which acquired Trega’s Navicyte capa- Baxter was Chief Scientific Officer at Oxford bility also offers such assays in 96-well kit form. Asymmetry International Ltd and prior to that Alternatively inhibition of individual enzymes in worked in ‘big pharma’ (Ciba and Glaxo) as a liver fractions can be monitored by the use of selec- research manager and medicinal chemist. tive substrates, metabolism of which can be fol- lowed by LC-MS analysis. Dr Peter Lockey is Director of Biochemistry at Argenta. Prior to Argenta, Dr Lockey was head of Solubility the high throughput screening group at Aventis While not being strictly an in vitro pharmacokinet- UK, and has more than 14 years’ experience in the ic parameter, the aqueous solubility of compounds . He holds degrees in bio- is possibly the most underestimated and yet critical chemistry and in addition to a property that a potential possesses. PhD which focused on drug targeting in cancer In order to measure aqueous solubility in a high chemotherapy.

Drug Discovery World Winter 2001/2 15