50 years of hurdles and hope in anxiolytic drug discovery

Abstract

Anxiety disorders are the most prevalent group of psychiatric diseases, and have high personal and societal costs. The search for novel pharmacological treatments for these conditions is driven by the growing medical need to improve on the effectiveness and the side effect profile of existing drugs. A huge volume of data has been generated by anxiolytic drug discovery studies, which has led to the progression of numerous new molecules into clinical trials. However, the clinical outcome of these efforts has been disappointing, as promising results with novel agents in rodent studies have very rarely translated into effectiveness in humans. Here, we analyse the major trends from preclinical studies over the past 50 years conducted in the search for new drugs beyond those that target the prototypical anxiety-associated GABA (γ-aminobutyric acid)-benzodiazepine system, which have focused most intensively on the serotonin, neuropeptide, glutamate and endocannabinoid systems. We highlight various key issues that may have hampered progress in the field, and offer recommendations for how anxiolytic drug discovery can be more effective in the future.

Figure 1. Fifty-year trends in preclinical anxiolytic drug discovery

The values represent the number of experiments investigating the anxiety-related effects of targeting the 5-hydroxytryptamine (5-HT; also known as serotonin), neuropeptide, glutamate and endocannabinoid systems between 1960 and 2012. The graph shows that the volume of research steadily increased from the 1980s onwards, peaking at the end of the 1990s, and has remained relatively constant up to now. More than half of the experiments focused on the 5-HT system, but neuropeptide drugs have also been a major focus of anxiolytic drug discovery, accounting for about one-third of all experiments. Over the past decade, the field has seen a rise in studies focusing on the glutamate and endocannabinoid systems. In this figure, an experiment refers to one drug (single or multiple dosing) that is tested in one assay or model. For more information on each experiment, including the drug, preclinical model, results and references, see Supplementary information S1 (box).

Figure 2. The ten most commonly used tests in anxiolytic drug discovery

The values represent the number of experiments performed with each test between 1960 and 2012. The elevated plus-maze test, the light/dark test and the open-field test have been a mainstay of anxiolytic drug discovery research for many years. They assay anxiety-like behaviour by generating a conflict between a drive to approach novel areas and, simultaneously, to avoid potential threat therein. They have clear intuitive appeal, are inexpensive to construct, and ostensibly quick and easy to run. The term ‘conflict-based test’ is also often used to describe measures of behaviour suppression by mild electric shock. This group includes the Vogel conflict and Geller-Seifter tests, which measure anxiolytic-like activity as the maintenance of a behavioural response (for example, licking or bar pressing) despite the receipt of a shock. Another set of fear-based tests involves variations on classical Pavlovian fear conditioning. Here, an animal learns to associate a context or specific environmental stimulus (for example, a light or a sound) with electric shock to produce a conditioned fear response that can be quantified in various ways (for example, freezing, escaping, avoidance or startle). Although the elevated plus-maze test, the light/dark test and the open-field test continue to be very popular, conflict-based tests — which were part of many drug discovery programmes in the 1980s and 1990s — are less frequently used today, perhaps because they require animals to be trained over multiple days and are more labour-intensive and time-consuming than the approach-avoidance tests.

Figure 3. Anxiety-related effects of drugs targeting the 5-HT, neuropeptide, glutamate and endocannabinoid systems

Findings from experiments conducted between 1960 and 2012 are shown as the percentage of experiments that showed anxiolytic-like, anxiogenic-like and inactive effects. The number of experiments reporting anxiolytic-like effects is shown on the graph. This figure shows that although compounds modulating the 5-hydroxytryptamine (5-HT), corticotropin-releasing factor (CRF), cholecystokinin (CCK), endocannabinoid and tachykinin systems have shown variable effects, compounds acting at several glutamatergic receptors (that is, metabotropic glutamate receptor 2 (mGluR2) and mGluR5), compounds targeting neuropeptide Y (NPY) and compounds that block melanin-concentrating hormone (MCH) receptors have all produced relatively consistent anxiolytic-like effects. CB₁, cannabinoid 1; FAAH, fatty acid amide hydrolase; MPEP, 2-methyl-6-(phenylethynyl)pyridine; NK₁, neurokinin 1; NMDA, N-methyl-D-aspartate; NPS, neuropeptide S; SSRI, selective serotonin reuptake inhibitor.

Figure 4. Experiments in animal models that investigated the effects of drugs modulating neuropeptide systems in models of anxiety disorders from 1960 to 2012

Seventeen different peptide systems have been suggested to have a role in the modulation of anxiety behaviours. This graph shows that, among them, corticotropin-releasing factor (CRF), the tachykinins and cholecystokinin (CCK) have been a major focus of anxiolytic drug discovery, accounting for about one-third of all experiments. MCH, melanin-concentrating hormone; NPS, neuropeptide S; NPY, neuropeptide Y; OFQ, orphanin FQ/nociceptin; TRH, thyrotropin-releasing hormone.

Figure 5. Fifty-year trends in the species, strain, sex and chronicity of drug treatment in anxiolytic drug discovery studies

The values represent the absolute numbers and percentages of experiments performed with different species (part a), strains (parts b, c, d) and sexes (part e), regardless of whether these involved acute or chronic treatment (part f), between 1960 and 2012. Rats represented the species of choice for anxiety tests, but mice have been extensively used as well. In addition, the majority of studies have used male subjects (part e) rather than females, and tested the effects of drugs following acute treatment (part f) rather than chronic treatment.

Figure 6. Recommendations for improving anxiolytic drug discovery

The figure details simple and actionable, rather than idealized, suggestions and points to keep in mind. Although the early stages of the anxiolytic discovery process require high-throughput tests, these have generally limited predictive validity. Later-stage profiling using behavioural models with increased translatability potential could confirm or reject the initial findings, thereby increasing the probability of having selected the drug candidate with the highest anxiolytic potential. 5-HT, 5-hydroxytryptamine.