The majority of the information described in this paper pertains to research directed towards the discovery of neuromuscular blocking agent drugs. The methods and techniques are equally applicable to other classes of drugs.
Drug discovery has benefited from increases in scientific and medical knowledge and several ‘defined’ periods have been identified:
- • In the 1940s and 1950s, the majority of drug discovery was serendipitous or the result of investigation of the products of chemical synthesis.
- • In the 1970s and 1980s, emphasis shifted with increased developments in receptor biochemistry and the knowledge that drugs act at specific receptors and mediate their activity.
- • To date, drug discovery is driven by advances in molecular biology, particularly by genomics and bioinformatics.
It is well known that poison on the arrow-tips used by South American Indians contained substances, now known to be ‘curares’ that could paralyse animals. In 1850, Bernard demonstrated that the observed paralysis was mediated via the neuromuscular junction, although the precise mechanism was at that time unknown . It was not until early in the 20th century following the purification of curare that King elucidated the chemical structure of tubocurarine . It was described as a bisquaternary compound with two distinct l-benzylisoquinolinium moieties united by ether links (Figure 1 a). Unfortunately, subsequent research in the 1970s  revealed that the structure originally described  was that of a related curariform substance, chondrocurine, as the interionium distance of the nitrogen atoms was incorrect and tubocurarine was found to be monoquaternary.
In the 1960s, an alkaloid with curariform properties, malouétine, was isolated and its structure also determined . Two features were of particular interest: its steroidal structure and the finding that it had an interionium distance comparable with that of tubocurarine. The two compounds depicted in Figure 1, illustrate the relation between the molecular structure and the interionium distance characteristic of neuromuscular blocking drugs. The figure shows two different structures, a benzylisoquinolinium, d-tubocurarine (Figure 1 a) and a steroid, pancuronium (Figure 1 b), with a similar interionium distance, as indicated by the arrow.
The structural similarities between the steroidal curariform alkaloid and tubocurarine captured the attention of medicinal chemists and initiated the process of synthesizing molecules with nitrogen atoms at similar distances. Subtle modifications to the parent structure (e.g. the replacement of an ethyl group for a methyl group) and analyses of the resulting molecules were undertaken in an attempt to improve the ‘active molecule’. By studying the activities of such a series of molecules, chemists obtained an insight into the structure activity relation of a molecule and could optimize it accordingly. Based on this insight, pharmaceutical chemists were able to synthesize the muscle relaxant pancuronium bromide .
During the past 35 years, further work on the structure activity relations of these steroidal compounds and of benzylisoquinoliniums has resulted in the formation of thousands of compounds that are candidates as potential muscle relaxants. From pancuronium, four related products have reached the market; vecuronium [6,7], pipecuronium , rocuronium , and more recently rapacuronium (Figure 2) . Four clinical products have also resulted from the benzylisoquinolinium muscle relaxant d-tubocurarine: atracurium , mivacurium , doxacurium  and cis-atracurium .
The role of the receptor
The majority of pharmacological responses (contractile, haemodynamic or secretory events) are mediated by receptors. Receptors are protein molecules that recognize either natural ligands or drugs designed to mimic their effects. Recognition of the ligand or drug by the receptor determines the specificity of the response. From the work of Claude Bernard  it was established that the curariform drugs act by affecting physiological processes mediated via receptors at the neuromuscular junction. In the case of the neuromuscular junction, the receptor is the nicotinic acetylcholine receptor. Developments in the fields of protein chemistry and molecular biology led to the complete chemical characterization of the nicotinic acetylcholine receptor . This finding was significant and heralded the beginning of a plethora of receptor research in the 1980s.
Molecular biology and the Human Genome Project
Receptors remain central to modern drug discovery, although the technology is now driven heavily by molecular biology. The Human Genome Project has played, and will continue to play, a pivotal role. It was first anticipated that the human genome would contain between 40 and 100 000 genes. We now know that the human genome has only 40 000 genes  and each of these is a potential drug target. If one considers the drugs that are currently available, they act at approximately 500 different gene targets, including receptors, enzymes and growth factors. Therefore, only a fraction of the potential drug targets are being exploited, which implies that there is a huge opportunity for new drug discovery.
The search for a new drug or medicine starts with the identification of a relevant target or receptor. General medical knowledge, disease aetiology and resources, such as the Human Genome Project, can aid this process. However, a fundamental question in this search is the relevance of a particular target, i.e. target validation. As previously indicated, very few of the 40 000 genes of the human genome can be associated with a specific function. One specific example of target validation is the use of Genechip® array technology. Genechips are essentially small slides of glass containing millions of tiny spots of DNA. The nucleic acid (target) to be analysed is isolated, amplified and labelled with a fluorescent ‘reporter’. They can be used to compare extracts of healthy and diseased tissue with respect to possible differences in gene expression, i.e. the amount of light emitted by the fluorescent reporter (Figure 3). In a particular disease state, certain genes may be expressed to a greater or lesser extent than in the healthy situation, and those genes may be relevant targets for drug discovery. Another use for gene chips could be the creation of individual genetic profiles, indicating an individual's predisposition to a particular disease. This was further corroborated by President Clinton, who in his ‘State of the Union’ address in 1998, suggested that within a decade, gene chips would act as road maps for the prevention of illness throughout a lifetime.
Due to the potential volume of work involved, there is a delicate balance between the choice of viable, relevant targets and the extensive costs to develop a new drug.
The initial or ‘recognition’ stage of a drug interaction with a receptor or target can be studied in an assay; the most simple of which is a traditional binding assay using a suitable radioactive or fluorescent ligand. However, sometimes it is more important to look at the function that the drug executes on the receptor/target. In such instances, functional assays are employed. An example of this sort of assay is the coupling of the luciferase gene to its particular receptor in such a manner that when the receptor is activated, luciferase is transcribed and synthesized by the cell. Its function is subsequently measured by the amount of light produced .
In the past, assays were conducted in test tubes and were a long, drawn out process. Nowadays, the volume of potential drugs to be tested is so enormous that such practice is both physically and economically impossible. The process has become fully automated, permitting the use of smaller samples of test compound and test target and ultimately allowing for high-throughput screening. Through such a process, it may be possible to identify 10 potentially interesting compounds for further development out of several million initially evaluated. These are called ‘hits’, and are the basis for further optimization by structure activity relation or subsequent development. At the end of the optimization process, one hopes to end up with a molecule with properties that are good enough to allow clinical testing, i.e. it is a clinical candidate. Obviously, prior to this, several preclinical studies must be conducted, particularly those of toxicology and pharmaceutical formulation. In reality, the process is more complex, as it also investigates more detailed parameters including potential side-effects, solubility and stability.
The pharmaceutical industry remains committed to the development of innovative products; for example, drugs with a new mechanistic approach for the management of neuromuscular transmission. Future projects will focus on these and other pharmacotherapeutic areas of anaesthesiology, including the development of new analgesics and new reversal agents.
Preliminary investigations using chemical chelation as a mechanism for the reversal of neuromuscular block are very promising. In monkeys, it is possible to reverse a block of 3 × ED90 dose of rocuronium to a train-of-four recovery ratio of 0.9 within 2 min. The chelator encapsulates the rocuronium molecule to form a high-affinity complex. Rocuronium is then unable to bind to its target receptor, the nicotinic acetylcholine receptor. Thus, its neuromuscular blocking action is reversed. More importantly, it has a very clean cardiovascular side-effect profile. Such a concept could revolutionize anaesthetic practice.
Compared with drug discovery in the middle of the 20th century, we now have the capability to produce and evaluate the activities of several thousand compounds within a short period of time. However, very few drugs that are discovered based on this type of technology will reach or prove their effectiveness in the market. We should certainly not be complacent in relying on this type of technology to produce the drugs of the future. The chance that one drug will succeed remains serendipitous.
The author acknowledges the work of the research group of Organon Laboratories, Newhouse, in particular Drs T. Bom, A. Muir and D. Rees.
1 Bernard C. Action de curare et de nicotine sur le système nereux et sur les systèmes musculaires. C R So Biol
1850; 2: 195.
2 King H. Curare alkaloids. Part 1. Tubocurarine. J Chem Soc
1935; 2: 1381–1389.
3 Everett AJ, Lowe LA, Wilkinson S. Revision of the structures of (+) -tubocurarine chloride and (+) -chondrocurine. J Chem Soc
1970; D: 1020–1021.
4 Huu-Lainé FK, Pinto-Scognamiglio W. Activité curarisante du dichlorure de 3 β
bistrimethylammonium 5 α
-prégane (malouétine) et de ses stéréoisomères. Arch Int Pharmacodyn Ther
1964; 147: 209–219.
5 Baird WLM, Reid AM. The neuromuscular blocking properties of a new steroid compound, pancuronium bromide. Br J Anaesth
1967; 39: 775–780.
6 Savage DS, Sleigh T, Carlyle I. The emergence of ORG NC 45, 1-[(2 β
0–3−17–bis (acetyloxy)−2–(1–piperidinyl)–androstan−16–yl]−1–methylpiperidinium bromide, from the pancuronium series. Br J Anaesth
1980; 52: 3S–9S.
7 Marshall IG, Agoston S, Booij LHDJ, Durant NN, Foldes FF. Pharmacology of ORG NC 45 compared with other non-depolarizing neuromuscular blocking drugs. Br J Anaesth
1980; 52 (Suppl. 1): 11S–19S.
8 Boros M, Szenohradsky J, Kertész A et al.
Clinical experiences with pipecuronium bromide. Acta Chir Hung
1983; 24: 207–214.
9 Booij LHDJ, Knape HTA. The neuromuscular blocking effect of Org 9426. A new intermediately-acting steroidal non-depolarising muscle relaxant in man. Anaesthesia
1991; 46: 341–343.
10 Wierda JMKH, van den Broek L, Proost JH, Verbaan BW, Hennis PJ. Time course of action and endotracheal intubating conditions of Org 9487, a new short-acting steroidal muscle relaxant; a comparison with succinylcholine. Anesth Analg
1993; 77: 579–584.
11 Payne JP, Hughes R. Evaluation of atracurium in anaesthetized man. Br J Anaesth
1981; 53: 45–54.
12 Savarese JJ, Ali HH, Basta SJ et al.
The clinical neuromuscular pharmacology of mivacurium chloride (BW B1090U): a short-acting nondepolarizing ester neuromuscular blocking drug. Anesthesiology
1988; 68: 723–732.
13 Basta SJ, Savarese JJ, Ali HH et al.
Clinical pharmacology of doxacurium chloride, a new long acting nondepolarizing muscle relaxant. Anesthesiology
1988; 69: 478–486.
14 Belmont MR, Lien CA, Quessy S et al.
The clinical neuromuscular pharmacology of 51W89 in patients receiving nitrous oxide/opioid/barbiturate anesthesia. Anesthesiology
1995; 82: 1139–1145.
15 Numa S, Noda M, Takahashi H et al.
Molecular structure of the acetylcholine receptor. Cold Spring Harb Symp Quant Biol
1983; 48: 57–69.
16 Multiple papers and multiple authors. The Human Genome. Nature
2001; 409: 813–958.
17 Nie Z, Mei Y, Ford M, Rybak L et al.
Oxidative stress increases A1 adenosine receptor expression by activating nuclear factor kappa B. Mol Pharmacol
1998; 53: 663–669.