New Approaches to Drug Discovery

Gernot Langer, Langer, Gernot

The impressive advances in the generation and interpretation of functional omics data have greatly contributed to a better understanding of the (patho-)physiology of many biological systems and led to a massive increase in the number of specific targets and phenotypes to investigate in both basic and applied research. The obvious complexity revealed by these studies represents a major challenge to the research community and asks for improved target characterisation strategies with the help of reliable, high-quality assays. Thus, the use of living cells has become an integral part of many research activities because the cellular context more closely represents target-specific interrelations and activity patterns. Although still predominant, the use of traditional two-dimensional (2D) monolayer cell culture models has been gradually complemented by studies based on three-dimensional (3D) spheroid (Sutherland 1988) and other 3D tissue culture systems (Santos et al. 2012; Matsusaki et al. 2014) in an attempt to employ model systems more closely representing the microenvironment of cells in the body. Hence, quite a variety of state-of-the-art cell culture models are available for the generation of novel chemical probes or the identification of starting points for drug development in translational research and pharma drug discovery. In order to cope with these information-rich formats and their increasing technical complexity, cell-based assay development has become a scientific research topic in its own right and is used to ensure the provision of significant, reliable and high-quality data outlasting any discussions related to the current "irreproducibility epidemic" (Dolgin 2014; Prinz et al. 2011; Schatz 2014). At the same time the use of cells in microplate assay formats has become state of the art and greatly facilitates rigorous cell-based assay development by providing the researcher with the opportunity to address the multitude of factors affecting the actual assay results in a systematic fashion and a timely manner. This microplate-based assay development strategy should result in the setting up of more robust and reliable test systems that ensure and increase the confidence in the statistical significance of the actual data generated. And, although assay miniaturisation is essential in order to achieve this, most, if not all, cell-based assays can be easily reformatted and adapted to be used in this format in a straightforward manner. This synopsis aims at summarising valuable, general observations made when implementing a diverse set of functional cellular in vitro assays at Bayer Pharma AG without claiming to deeply review all of the literature available in each and every detail. In addition, phenotypic assays (Moffat et al. 2014) or label-free detection methods (Minor 2008) are not discussed. Although this essay tries to cover the most relevant technological developments in the field, it nevertheless may express personal preferences and peculiarities of the author's approach to state-of-the-art cell-based assay development. For additional reviews covering the actual field, see Wunder et al. (2008) and Michelini et al. (2010). Table 1 Cell culture media requirements and costs of consumables in standard cell culture vessels compared to higher-density microplates Dimensions Cell culture media requirements (fill quantity) Productivity Costs (consumables) Well number Layout Surface area μL per well mL per plate mL/void volume corrected (+20%) Fold increase in datapoint numbers (reference: 6-well format) Plates** (per plate) Tips*** (per tip) Average (mm(2)) Optimal Maximal Optimal Maximal Optimal Maximal From To From To From To 6 2 × 3 962 1,924 2,000 5,000 5.77 12.00 30.00 6.93 14.40 36.00 1 1.99 € -/- 24 4 × 6 194 388 500 3,000 4.66 12.00 72.00 5.59 14.40 86.40 4 2.33 € -/- 96 8 × 12 34 68 100 300 3.26 9.60 28.80 3.92 11.52 34.56 16 3.40 € 0.09 € 384 16 × 24 9 18 30 100 3.46 11.52 38.40 4.15 13.82 46.08 64 6.90 € 0.11 € 1536 32 × 48 2 4 4 10 3.07 6.14 15.36 3.69 7.37 18.43 256 30.40 € --/-- Going for high-density formats (i.e. ≥ 96 well) should result in a reduction of media consumption by about 20-30%. In addition, cell culture media requirements do not increase even when going for still higher-density formats, i.e. >384-well plates. On the other hand, the numbers of tests being performed (productivity) increases at least 16-fold. Optimal* is adding cell culture media according to generally accepted standards for optimum gas exchange (fill level = 2 mm). Plates** are price per plate, when ordering 1,000 plates of a standard, high-quality, sterile cell culture plate with lid (for use in assay development). Please note, however, that once established, most cell-based assays do not need to be performed in sterile cell culture plates! Tips*** are list price per tip when ordering standard, high-quality, nonsterile 50/125 μL tips predominantly used in the 96-well/384-well experiments. (a) -/- not given: only the tip prices for use in high-density plate formats are listed. (b) --/-- not given: 1536-well format pipetting usually requires the use of multichannel dispensers, although tips used in 384-well formats can be used Fig. 1 (a) Variation of key parameters and conditions in order to establish the linear range of a cell-based assay. Some definitions: the linear range comprises the optimal region in which the actual assay read-out is proportional to a particular cell number (i.e. targets/cell) and the parameters/conditions varied, respectively. In contrast to the linear range, the dynamic range of an assay is defined as the area between the mean maximal and mean minimal signals (background). The crucial working range is obtained by subtracting 3 standard deviations (SDs; σ) from the maximal signal and adding 3 SDs to the minimal signal/background. Any value within the working range can be considered statistically significantly different with 99.7% confidence. (b) In cases in which high and low controls are available only, the calculation of the actual size of the assay window follows exactly the same procedure used to determine the working range of a given assay. In both cases it is obvious immediately that the degree of variation in the highest and lowest signals determines the actual assay quality: even assays with quite low S/B values can be evaluated with high precision once their maximal and minimal values do not oscillate very much - something generally achieved in a straightforward manner when switching from manual well-to-well pipetting to highly accurate electronic multichannel pipettes and dispensers.