Drug-drug interactions (e.g., by induction or inhibition of certain enzymes) may contribute to disruption of the normal kinetics of drugs which may result in increased risk of occurrence of adverse effects and as a result affect the clinical development of novel therapeutics. Thus, HTS approaches have been developed in order to determine the activity of chemicals towards modulating the drug metabolizing enzymes. HTS assays which utilize the activity of cytochrome CYP enable the establishment of pharmacokinetic drug interaction and are of vital importance in the drug development industry.

Low-throughput methods combined with liquid chromatography/mass spectrometric methods (LC/MS/MS) allow the determination of the effect of CYP enzymes on examined molecules [51,52]. In these assays hepatocytes are frequently used which enable the acquirement of information concerning II phase of a drug’s metabolism. It is possible to compare in one assay the cytotoxicity of chemicals to the cell line with and without CYP enzymes [53]. Apart from the LC/MS/MS method, there is a possibility to measure the CYP-mediated metabolism by means of fluorescent method. Such assays can be miniaturized using low reaction volumes which enable screening in 1536-well plates [54,55].

There are also available tests which couple measurement of cytotoxicity endpoints with CYP catalysis (e.g., the MetaChip system) [56]. The IdMOC system, a co-culture of five cell types and hepatocytes, enables the evaluation of the effect of metabolites of studied chemicals on a variety of cells which represent different tissues [57].

HTS luciferase reporter gene assays performed in liver cell lines, are used to identify activators of PXR (pregnane X receptor), a nuclear receptor, which contributes to induction of mRNA for a variety of CYPs. There have been reported strong correlations between inducers of the PXR reporter gene activity and the ability to induce CYP3A. For confirmation lower throughput enzymatic activity assays are used [58,59].

Nowadays we are observing an increasing demand for the understanding of individual genetic variability in the development of various diseases and in drug response and toxicities. Thus, screening of different genetic polymorphisms in large populations is a major goal that would facilitate this process. The increasing interest in these pathogenetic studies has resulted in greater demand for broad genome association studies and as a consequence led to the development of new and robust high throughput screening methods for genotype analysis. One of these methods is matrix-assisted laser-desorption ionization time-of-flight mass spectrometry (MALDI-TOF MS), which is a powerful technique for DNA analysis. MALDI-TOF MS approaches have been developed for rapid screening of single nucleotide polymorphisms (SNPs), epigenotype analysis, quantitative allele studies, and for the discovery of new genetic polymorphisms. These methods are based on single base primer extension and minisequencing implemented with new chemicals using MALDI-TOF MS and include photochemical and other chemical and enzyme cleavage strategies that facilitate sample automation and MS analysis for real-time genotyping and resequencing screening [60].

Cytochrome P450 (CYP) enzyme system can be encoded by the P450 gene family. It is one of the widely studied topics in drug development. Several of the drugs metabolizing enzymes, which belong to the CYP family, are polymorphic which means that they have more than one variant of the gene. Despite similar functional properties each of the CYP isozymes is different and has a distinct role. Therefore, individual differences in the efficacy of drug treatment, adverse effects and toxic effects may occur. It is also worth emphasizing that some drugs are metabolized by more than one isozyme, which may contribute to metabolism-based drug-drug interactions. Furthermore, some drugs might be inducers or inhibitors of specific isozymes which lead to alteration in metabolism of other drugs—substrates of these isozymes. There is a wide variety in the expression, activity and concentration of specific isozymes among individuals, species and ethnic groups. The expression of these enzymes is altered by species specificity, genetic polymorphism, age, diseases and environmental inducers. A possible example is CYP450 enzymes, variability associated with this enzyme group results in marked differences in response when the same drug is administered to different people. In the case of CYP450 enzymes there are three major groups of metabolizers: extensive metabolizers, poor metabolizers and ultra-sensitive metabolizers. Genetic differences among individuals can lead to an excessive or prolonged therapeutic effect or toxic overdose [61]. In the drug development process it is crucial to have information on the enzymes responsible for the metabolism of a drug candidate at the earliest stage. Generally, genetic polymorphism contributes to high inter-individual variability and potential for drug-drug interactions. Genetic information is used to establish the response of individuals and populations to drugs. Furthermore, metabolite profiles are important for the design of pro-drugs and pharmacologically active metabolites. For in vitro studies before pre-clinical screening low-throughput assays are performed. Information obtained by incubating a tested drug with an appropriate system can be used to design safer and more metabolically stable drugs. Currently there is a wide variety of hepatic in vitro systems which differ in biological intricacy. To study multiple aspects of drug metabolism cell cultures or cell suspensions are used. Hepatocytes are utilized for studying Phase I and Phase II reactions. For drug metabolism studies primary cell lines are used which are isolated from fresh liver tissue. These systems can be used immediately after isolation or culture for long-term studies. However, cultured cells lose the enzymatic activity rapidly with time. Thus, there is a great need to improve stabilization of P450 activity [61]. Marks et al. developed and characterized a fluorescence-based HTS assay employing recombinant human CYP2B6 and 2 novel fluorogenic substrates (the Vivid CYP2B6 Blue and Cyan Sub-strates). Developed assays have been proven to be robust and sensitive, and allow screening in HTS mode of a large panel of compounds for CYP2B6 metabolism and inhibition [62].

Genetic toxicology is the scientific discipline the aim of which is to establish the effects of chemical, physical and biological agents on the heredity of living organisms. For measurement of genotoxicity of chemicals the use of the Ames bacterial reverse mutation test, the mouse lymphoma tk gene mutation assay (a negative selection for loss of the functional thymidine kinase gene), and the micronucleus clastogenicity assay are employed. The Ames test, the simplest and quickest of the existing genotoxicity assays, is capable of detecting point mutations and frame shift mutations. However, it does not detect chromosomal rearrangements or double strand breaks. In the micronucleus assay double strand breaks contribute to formation of chromosomal fragments that are not attached to microtubules during metaphase, and are not pulled to opposite poles before cell division. These chromosome fragments migrate outside the normal nucleus and can be observed microscopically as micronuclei. This assay is prone to false positive results which occur when an undamaged but lagging chromosome forms a micronucleus and false negative results which are caused because the micronucleus assay detects only double strand breaks [63]. Methods listed above possess several drawbacks, such as high costs, low specificity and sensitivity. Furthermore, these tests do not allow the screening of a large number of compounds [64].

Scientists’ efforts have led to the development of high-throughput genotoxicity assays which allow screening of a greater number of chemicals. One of such tests is the Ames II assay, adaptation of the previous test which has a very good conformity to the standard Ames testing procedure, decreases the amount of test compound required for a study and is compatible with limited automation [65,66].

Replacement of traditional microscopy by automated cellular imaging enabled higher throughput and contributed to a lower amount of compound (approximately 3 mg) in the micronucleus clastogenicity assay [67].

Ritter et al. developed an integrated higher throughput method for the comet assay which is a method for determination of DNA damage in vivo and in vitro. The scientists’ efforts have contributed to a faster and easier slide-production, smaller amount of cells needed, higher amount of comets quantified, and a fully automated analysis of comets. According to the research results the introduced method was characterized by high reproducibility, flexibility, and efficiency. Thus this procedure can be used as an automated analysis method in HT genotoxicity studies in vitro [68].

Evan’s team introduced a new assay system in which the chicken DT40 B cell line was used. This cell line has several significant advantages that make it suitable for genotoxicity studies. It was reported that the implemented assay provides enhanced sensitivity by means of genetically defined and phenotypically characterized mutants, which are defective in DNA repair pathways. Moreover, utilization of DNA repair proficient wild-type cells as a negative control, contributes to minimization of false negative outcomes. Evan’s method enables the establishment of mechanisms of genotoxicity and the extrapolation of the results to the human context [63].

The effects of chemical exposure on gene regulation might be measured also by means of reporter gene-based systems. These systems monitor expression of toxicity markers such as those associated with tumorigenesis, cytokine release and transcriptional activation, which relate to carcinogenicity, mutagenicity, inflammation and endocrine disruption. The use of in vivo reporter based assays and the development of transgenic animals constitute a reduction and refinement of traditional rodent bioassays. These in vitro assays do not only provide relevant toxicological information, but also facilitate the establishment of mechanisms of toxicity. Despite the fact that reporter gene-based systems have been developed, it is clear that there is no ideal in vitro or in vivo system to reliably assess genotoxicity. Thus, there is an urgent need to focus on the discovery of tissue-specific toxicity biomarkers so that sophisticated human cell line-based assays can be developed, and validated for regulatory toxicity testing [69].

HTS is a testing strategy that enables a broad probe of biological targets, pathways, and mechanisms in relation to toxicity endpoints, including genetic toxicity, for a large number of compounds. Genotoxicity screening, possessing a unique placement in toxicology, might be regarded as a front-line safety assessment tool. Of vital importance is also the relation of genotoxicity mechanisms with carcinogenicity.

Ion channels are a class of membrane-spanning protein pores that mediate the flux of ions in different cell types. There are different types of ion channels throughout the body, and apart from setting the resting membrane potential and controlling cellular excitability, ion channels also mediate critical physiological functions such as heartbeat, signal transduction, cell secretion, and gene expression. Ion channels use a variety of “gating” mechanisms to open and close their pores in response to biological stimuli such as ligand binding or membrane potential changes. Some of these channels have emerged as attractive drug targets and to realize their full potential as a target class, scientists are actively searching for drugs which could target the specific states of ion channels. It would allow the fine-tuning of drug effects as a function of the degree and frequency of channel activity [68,70,71].

The patch-clamp method is the current gold standard technique for probing ion channel activity. This method allows the measurement of small ion currents with millisecond temporal resolution with simultaneous control of the membrane potential. Automated patch clamp (APC) instruments, in comparison with manual patch clamp, have a 10–100-fold improved throughput [72]. On the other hand, the highest throughput using commercially available APC instruments is only 2000 data points per day. Furthermore, these instruments are rather costly. Thus, APC instruments are primarily used for lead optimization, hit confirmation, and safety screening. As a result, developments in APC technologies are focused on higher throughput, lower cost, and miniaturization [73].

In comparison with the patch-clamp method, optical ion channel assays can achieve a throughput of 100,000 analyses per day. Moreover, they are cheap, simple to perform, non-invasive, amendable to miniaturization and can test multiple cells at a time. A major problem for optical assays is the inability to control the membrane potential and interrogate ion channels in different conformational states [74]. Calcium flux assays are important techniques for Ca²⁺-permeable ion channels because of the availability of excellent Ca²⁺-specific indicators and a large dynamic range of intracellular Ca²⁺ concentrations. There are two main approaches to measure intracellular concentration of Ca²⁺: using organic dyes or genetically encoded proteins with either natural or engineered Ca²⁺ sensitivity [74,75].

Potassium ion channels have attracted huge attention of scientists, as targets for therapeutic indications as well as for safety profiling. In contrast to Ca²⁺-permeable ion channels, no equivalent resources are available for potassium channels. In order to facilitate the pharmaceutical development of potassium channel modulators, high throughput potassium-specific optical assays are critical. One of the major obstacles in designing such assays is the shortage of K⁺-specific fluorescent indicators which are capable of detecting narrow physiological variations of extracellular K⁺ concentration [73].

Assays have also been developed which are aimed at ion channels in order to establish whether a tested compound causes cardiac arrhythmias. In use are binding assays, ion flux assays, fluorescence-based assays, and automated patch-clamp instrumentation. For example the receptor-binding assay detects chemicals that bind to the receptor and is sensitive, whereas it may not induce prolongation of a QT section. As a consequence, compounds active in this test must be examined also by voltage-clamping electrophysiology measurements [76].

Another test is the rubidium flux assay. It guarantees limited throughput and a functional endpoint. However, the main problem is its automation because a flame atomic absorption spectrometer as a detector is required [77].

There are also used fluorescent, voltage-sensitive dyes combined with a fluorometric measurement for potassium ion channel hERG which are sensitive, but simultaneously unreliable with regard to patch-clamp values [78].

By means of high-throughput, gene reporter assays utilizing luciferase or β-lactamase can easily screen compounds with both agonist and antagonist activity for these enzymes [79,80].

Tests have been developed based on protein:protein interactions (domains of the nuclear receptor react with the coactivator protein). These interactions can be measured by the following methods: fluorescence polarization with a fluorescently labeled small peptide [81], fluorescence resonance energy transfer using fluorescently labeled receptor and coactivator [82], or Amplified Luminescence Proximity Homogenous Assay [83].

Traditional receptor binding assays utilizing radio-labeled ligands can be configured for high-throughput tests by using scintillation proximity assay beads [84].

To improve and accelerate the development of new ion channel modulators, there is a need for screening technologies that would offer both high throughput and high information content in a cost effective way.

Apart from well-defined and characterized molecular targets with established associations towards toxicity, there is a need for broad screening of chemicals against a panel of molecular targets. Extensive in vitro pharmacology profiling of new chemical compounds during early phases of drug discovery has become an essential tool to predict a broad spectrum of clinical adverse effects. State-of-the-art assays and rapidly expanding knowledge about different types of receptors, ion channels and enzymes have made it possible to implement a large number of assays addressing possible clinical activities. These assays can potentially avoid late, high attrition rates, making drug development more cost effective. Furthermore, safety pharmacology profiling of compounds enables the acquirement of receptor-, enzyme-, transporter- and channel-related liabilities of compounds. It has been suggested that this approach would be useful for the estimation of the toxicological potential of environmental chemicals [85]. There is available a large number of these tests, e.g., kinases, proteases, nuclear receptors, phosphatases, phosphodiesterases. A panel of 92 ligand-binding assays was used for identification of relationships between molecular structure and biological activity profiles [86].

An example of the importance of broad pharmacological profiling is the case of CGP 71683A, which is a potent and highly selective neuropeptide Y (NPY) Y5 receptor antagonist. In animal experiments this compound was shown to reduce food intake and lower body weight [87,88]. CGP 71683A exhibited very little activity towards Y1,2,4 receptors and reduced NPY-induced feeding. Thus, this tested compound was regarded as a gold-standard reference compound to study the effects of a selective Y5 receptor inhibition on food intake. Chronic treatment with CGP 71683A caused loss of activity and food intake returned to normal which was explained by the effect of the activation of counter regulatory systems. Although this seemed to be reliable, it remained unclear why the body weight remained low. Although Zuana et al. confirmed the selectivity of CGP 71683A within the NPY receptor family, in a broad binding assay panel they estimated high affinity to the muscarinic receptor, the serotonin uptake recognition site and to a lesser extent to α2-adrenergic receptors. Furthermore, morphological observations suggested that this compound caused brain inflammation, which could have contributed to the prolonged weight loss in animals [89]. The study performed by Zuana and co-workers highlights the importance of determining the full pharmacological profile of compounds, especially when complex mechanisms contribute to the development of certain biological activity and when experimental data are only interpreted with some difficulties.

Together with other in vitro assays focusing on toxicology and bioavailability, broad pharmacological assays provide a powerful tool to aid drug development. Several commercial databases have been created containing HTS data which are extremely helpful to the development of computational models to establish toxicological potential [47].

Cell-based assays have gained an established position in drug screening in the pharmaceutical industry. These assays enable the evaluation of potential drug targets by functionally characterizing their effect in cells and by assessing specificity and efficacy of drug leads. Cell-based assays provide information on the nature of the pharmacological activity of a compound at a specific receptor, ion channel or intracellular target. Cell-based screening approaches give information on the ability of the chemical compounds to penetrate the cell membrane, as well as an acute cytotoxicity profile. Furthermore, these assays allow the identification of the targets for drugs of unknown mechanism of action and performance of ADME-Tox of potential drugs. It is worth emphasizing that the membrane permeability and cytotoxicity data obtained from cell-based assays cannot be considered as definitive indicators of absorption or toxicity properties of lead compounds. However, such information provides an alert for an encumbered chemical series early in the lead generation process.

Cell-based assays contribute to a high level of complexity and possess several advantages over simple biochemical tests. Cellular models may be regarded as more suitable for detecting toxicity because in opposite to in vivo environment, there are no homeostatic mechanisms which could provide a buffering capacity against damages caused by xenobiotics. A variety of different cell lines have been used to predict organ specific toxicity. Furthermore, different species of cell lines are relatively easily accessible and this diversity enables comparison of compounds’ activities among various species. Such research might constitute valuable help in eliciting the reliability of the extrapolation of results from animal tests to humans [47].

In the early stages of the use of cellular models utilization of ATP content, membrane leakage, and cell number as cytotoxicity endpoints was made. These studies were appropriate for evaluating acute toxicity, but did not deliver information regarding the mechanisms of toxicity. Recent efforts and new procedures such as a wide array of cell signaling, stress-response pathways, contributed to greater understanding of toxicity mechanisms.

In Table 3 classification of cell-based assays is presented [1].

There are two types of HCS (High Content Screening). The first one uses fixed cells with fluorescent antibodies, ligands, and nucleic acid probes; the second utilizes live cells with multicolor fluorescent indicators and biosensors. Recent examples of HCS include a screen for cell migration inhibitors using automated microscopy, a screen to detect chemical suppression of a genetic mutation in zebra fish and the use of gene-expression signatures to identify compounds that induce differentiation of acute myeloid leukemia cells [1,90].

Nowadays, mainly due to high-content imaging (HCI) it is possible to conduct high-throughput, quantitative analysis of cellular phenotypic assays. HCI is a combination of epifluorescence imaging platforms and robust image analysis algorithms. Giuliano et al. stated that all gathered information at the cell level might be regarded as high-content screening. Such information include the following parameters: cell signaling pathways, protein expression levels, cell cycle status, receptor internalization, cytoskeletal integrity, energy metabolism status, nuclear morphology, post-translational protein modifications, cell movement, and cell differentiation [91]. HCI is usually used to evaluate the effect of chemicals on a set of parameters measuring cell homeostasis. Such procedures enable efficient screening of various compounds. Furthermore, simultaneous measurement of different parameters results in increased sensitivity [92].

The National Cancer Institute created a database of tens of thousands of chemicals which showed activity of growth inhibition on 80 tumor cell lines. This database enabled the understanding of the mechanism of toxic activity of studied chemicals which were grouped on the basis of structural information and biological properties by means of self organizing maps (SOM’s) [93]. This technique was used by Glover’s team. They studied a group containing a number of known inhibitors of mitochondrial complex I of the electron transport chain. They selected ten chemicals with unknown mechanism of action and subsequently five of these appeared to be potent inhibitors of complex I activity [94].

This procedure was successfully used also by Berg et al. in order to group compounds by mechanism of action. Exploited HTS panel comprised four human cellular assays and enabled the measurement of multiple inflammation related endpoints. Forty four compounds were classified successfully according to mechanisms unrelated to inflammatory system modulation and thus it could be concluded that this approach may contribute to understanding of mechanisms of toxicity [95]. In 2006 MacDonald’s group established the activity of over 100 drugs in 49 different human cell-based assays at several time points. These tests were based on protein (involved in cell-cycle, apoptosis, mitogenesis, proteolysis, GPCR signaling, cytoskeletal function, DNA damage, nuclear receptor signaling, stress and inflammation) interactions. After analysis it was reported that for some of the drugs new activities were discovered [96].

Dynamic monitoring of cytotoxicity by microelectronic sensors might also be used for characterization of the toxicological mechanism. For this purpose 96- or 384-well plates can be used, where cell viability, morphology and adherence are measured. One of the most important features of this procedure is the possibility to use any attached cell type [97].

Due to the fact that even complex cell culture assays cannot reliably model the higher level interactions which are presented in living organisms, HTS approaches have been developed using whole animals in the form of model organisms.

As model organisms invertebrates can be used, e.g., the yeast, Saccharomyces cerevisiae [98], and nematode, Caenorhabditis elegans [99]; and vertebrates, such as the zebrafish, Danio rerio [100]. In Table 4 there is presented the characterization of model organisms used in HT screening.

These model organisms, mainly during developmental stages, may interrogate vertebrate animals and thus contribute to easier access to a wide array of chemicals. By explicating the mechanisms involved in the response of the organisms to chemicals, then evaluating and testing that mechanism across species against any molecular targets, it is possible to establish a risk assessment.

In comparison with studies on larger vertebrates, these studies are inexpensive and capable of testing moderate numbers of chemicals, although not in the traditional HTS range, but are still laborious.

Whole organisms, such as zebrafish embryos, have become attractive model systems for developing disease-related assays for toxicology studies. Limitations in performing these assays are the availability of fast image acquisition systems with sufficient resolution and depth of field and the analysis of the complicated images to provide fast and accurate quantification of the biology of interest in the assay [109].

Nowadays the trend towards miniaturization of assays may be observed. At present most HTS is carried out in 96-well plate format, but utilization of 384-well plate formats is increasing. This plate format has been established as the format of choice for compound storage and screening assays and is used in various types of biochemical and cell-based assays. An 864-well and 1536-well plate format has also been implemented. However, the utilization of such high density formats is limited by numerous obstacles [1]. Benefits of miniaturization include lower volume of reagents required and faster experimental processing, and as a consequence reduction of cost and time [110].

In summary, computational toxicology, SAR prediction models, toxicogenomics, and HTS testing programs significantly alter the current paradigm of toxicity screening. A major challenge for the field of contemporary toxicology is to embrace computational toxicology, structure-based and in silico prediction methods, and new assay technologies that are able to efficiently screen thousands of chemicals [111].