Virtual screening to identify calcium channel blockers


The first marketed selective T-type calcium channel blocker mibefradil was withdrawn due to adverse drug interactions, prompting the search for new scaffolds.

Seeking improved safety from Medicinal Chemistry scaffolds

This paper describes a workflow using Reaxys® Medicinal Chemistry to assess chemical and biological information and the Reaxys Flat File as a source of chemical diversity for ligand-based virtual screening.

New TCCBs with new scaffolds are required to understand the exact role of T-type Ca2+ channels in cellular functions

Neuronal voltage-gated calcium channels play a central role in the control of cellular excitability and a number of calcium-dependent cellular functions, including gene transcription and transmitter release. Subtypes of high voltage-activated Ca2+ channels play important roles in synaptic transmission in the nervous system. However, the main proposed functions of low voltage-activated Ca2+ channels—also called transient or T-type calcium channels—include promotion of Ca2+-dependent burst firing, generation of low-amplitude intrinsic neuronal oscillations, elevation of Ca2+entry and boosting of dendritic signals, which contribute to neuronal pacemaker activity, wakefulness, seizure susceptibility or integration of sensory information, including pain (1–3)
Huguenard, J.R. and Prince, D.A. (1992)
A novel T-type current underlies prolonged Ca2+- dependent burst firing in GABAergic neurons of rat thalamic reticular nucleus. J Neurosci 12: 3804–3817.
Perez-Reyes E. (2003)
Molecular physiology of low-voltage-activated t-type calcium channels. Physiol Rev 83: 117–161.
Bourinet, E. and Zamponi, G.W. (2005)
Voltage-gated calcium channels as targets for analgesics. Curr Top Med Chem 5: 539-546.
.
Three structurally and functionally distinctive T-type calcium channels have been cloned. These are denoted as voltage-dependent Ca2+ channel Cav3.1, Cav3.2 and Cav3.3. Of these isoforms, Cav3.2 has been identified as a good target to identify new classes of analgesic drugs for pathological pain syndromes (3)
Bourinet, E. and Zamponi, G.W. (2005)
Voltage-gated calcium channels as targets for analgesics. Curr Top Med Chem 5: 539-546.
. However, only limited progress has been made in the quest to identify both potent and selective compounds for T-type channel blockade. The Ca2+ channel blockers such as flunarizine, U-92032, nicardipine and mibefradil (Figure 1) have been reported as active T-type Ca2+channel blockers (TCCBs).

Structures of some known active TCCBs - Elsevier Whitepaper | R&D Solutions
Figure 1. Structures of some known active T-type calcium channel blockers (TCCBs)

Mibefradil, the first marketed selective TCCB inhibits the T-type Ca2+ channels 10−30 times more potently that L-type Ca2+ channels. It was finally withdrawn due to its pharmacokinetic interactions with other drugs metabolized by cytochromes P450 3A4 and 2D6. Therefore, new TCCBs having new scaffolds are required to understand the exact role of T-type Ca2+ channel in cellular functions.
Herein, we report the application of Reaxys Medicinal Chemistry as an essential starting point for gathering critical biological and chemical information required to feed into ligand-based virtual screening (Figure 2). The data generated was used with bi-dimensional 2D pharmacophoric fingerprint software to identify a new series of TCCBs. The virtual hits were put through a functional assay on the Cav3.2 isoform to assess their biological activity.

Workflow to identify new TCCBs - Elsevier Whitepaper | R&D Solutions
Figure 2. Workflow to identify new T-type calcium channel blockers (TCCBs)

Dataset selection

The first step in the search for a new TCCBs is to collect all pertinent chemical and biological information related to TCCBs, in particular the Cav3.2 isoform (Figure 3). Searching by target name, initially under ion channels and then by narrowing the search down to voltage-dependent calcium channels, followed by selection of the specific isoform enables easy retrieval of all the relevant chemical and biological data from Reaxys Medicinal Chemistry.

Selection of relevant data to retrieve from Reaxys Medicinal Chemistry | Elsevier Whitepaper
Figure 3. Selection of relevant data to retrieve from Reaxys Medicinal Chemistry

To facilitate comparisons of bioactivity data from different publications and assay types, all the data points in Reaxys Medicinal Chemistry have pX values. pX values are calculated by transforming parameters such as EC50, IC50 and Ki into the –Log equivalent (pEC50, pIC50, pKi). These are normalized values assigned to the data that enable easily quantification of compound–target affinity and compare information from all around the world.
The target search for the Cav3.2 isoform of TCCBs in Reaxys Medicinal Chemistry retrieved the chemical structure of 1,854 ligands active against that specific target. These 1,854 ligands have 2,285 bioactivities associated with them and were extracted from 77 citations (Figure 4A).

A

The Heatmap for all the ligands with activity against the Cav3.2 isoform of TCCBs | R&D Solutions

B

The Heatmap for ligands with a pX activity above 6.0 - Industry Insights | Elsevier R&D Solutions
Figure 4. A The Heatmap for all the ligands with activity against the Cav3.2 isoform of TCCBs. B The Heatmap for ligands with a pX activity above 6.0 (affinity < 1 μM).

An affinity profile for the most potent ligands active against the Cav3.2 isoform of TCCBs, with pX values greater than 6.0 (affinity < 1 μM), can be generated and viewed as a Heatmap (Figure 4B). The Heatmap visualizes the relationships between ligands and their targets in terms of key parameters, allowing rapid identification of relevant ligand–target interactions. The map displays all ligands with a pX above 6.0 and the associated target TCCB protein for which in vitro biological data has been mined from the literature. In the Heatmap, biological affinities or activities are quantified as a pX value and displayed from 1 (low activity) in blue to 15 (high activity) in red.
At the time of this analysis, this query resulted in 471 ligands with a pX value of above 6.0 (affinity < 1 μM) against the TCCB target. Some of these molecules are depicted in Figure 5. These molecules were then clustered using variable-length Jarvis-Patrick clustering. The query set was narrowed down and the central molecules of resulting clusters and singletons was used for building the 2D pharmacophore in the virtual screening analysis.

Molecules - Industry Insights | Elsevier R&D Solutions
Figure 5. Some representative molecules from the query

A multitude of key parameters can be explored to aid understanding of the ligand−target interactions, such as drug-like grading, in vitro efficacy, in vivo animal models, metabolism, pharmacokinetic and pharmacodynamic data, and clinical use/application, as well as extensive information on the chemical structure itself. For example, Reaxys Medicinal Chemistry can be used to get detailed metabolic data for each compound/ligand (Figure 6). For each data point, the parameter measured, the value, target, target species, tissue/organ, dose and reference are all shown and both quantitative and qualitative results available in the database. This aspect would be particularly beneficial considering that Mibefradil was withdrawn from the market due to the inhibition of cytochrome P450 enzymes 3A4 and 2D6, which had the potential to lead to serious drug−drug interactions.

Number of 1,4 benzodiazepines with a pX value greater than 7 per GPCR target | R&D Solutions
Figure 6. All of the detailed information on the metabolism of mibefradil can easily be retrieved: the IC50 value for activity against CYP3A4 and CYP2D6 along with the corresponding reference is highlighted.

Virtual screening

Over the past few years, virtual screening has emerged as a complementary approach to high-throughput screening and has become an important in silico technique in the pharmaceutical industry. It includes ligand- and structure-based methods. Pharmacophore modeling is one such approach where known active molecules are analyzed for common steric and electronic features responsible for drug−receptor interactions. Alternatively, protein−ligand crystal structure complexes can be used to construct specific receptor-based pharmacophore models. In this workflow, a ligand-based approach was selected to quickly identify novel TCCBs using bi-dimensional (2D) pharmacophoric fingerprints (FP) based on the clusters obtained from the Reaxys Medicinal Chemistry data retrieval.

Prior to this, an external molecular database was constructed and maintained. Millions of compounds coming from Reaxys Flat File were loaded into a database. After applying Lipinski drug-like filters, subsequent processing done on the screening database, included salt removal, duplicates suppression and standardization.
Next, by calculating the pharmacophoric fingerprints, the chemical space representation of these molecules was assessed. Ideally, two sets of fingerprints will be computed: ChemAxon’s Pharmacophoric Fingerprints or PF and CCG’s GpiDAPH3 fingerprints implemented in the MOE suite. Particularly for PF, several parameters have to be optimized for FP generation.

The subsequent analysis was limited to the software where the hits had higher chemical diversity. The virtual screening campaign provides 314 hits that are predicted to be active on the Cav3.2 channel. Some parameters were taken into account for compound selection:

  1. Molecular diversity and chemical originality.
  2. Compound availability and pricing.

Finally, based on a visual inspection, a subset of 39 unique molecules was selected for further biological evaluation.

Biological assay

The subset of 39 unique molecules was tested on HEK293 cells transfected with the human Cav3.2 isoform. The compounds were purchased and tested at 10 µM for their ability to affect the functional activity of recombinant human Cav3.2 and the results are displayed in Figure 7. In total, 15 compounds were found to inhibit the Cav3.2 channel at >50% inhibition and of those 9 displayed very promising activity with >75% inhibition (Figure 8).

All of the detailed information on the metabolism of mibefradil can easily be retrieved | Elsevier
Figure 7. The biological screening results for the 39 compounds on Cav3.2 channels, obtained using an electrophysiology experiment with a single concentration of 10 µM
The chemical structure of the 9 more active compounds in the Cav3.2 channel | Elsevier Whitepaper
Figure 8. The chemical structure of the 9 more active compounds (% inhibition >75%) in the Cav3.2 channel

Conclusion

In this study, a proposed workflow was described for identifying new TCCBs from a structurally diverse dataset of known active compounds using virtual screening procedures incorporating various bi-dimensional chemical and pharmacophoric fingerprints. Reaxys Medicinal Chemistry was used to easily and efficiently retrieve all relevant chemical and biological information for existing TCCBs and Reaxys Flat File as a source of chemical diversity. During hit selection from the virtual screening, further analysis of the potential ligands can be done, for example, to identify potential drug−drug interactions earlier in the discovery process.

Essential drug discovery solution

Reaxys Medicinal Chemistry is an extensive databasecontaining chemical information linked to in vitro and in vivo biologicalactivities extracted from over 300,000 articles, 90,000 patents and 5,000journals. More than 6 million chemical compounds are associated with theirbiological data (> 29 million bioactivity data points) and linked toinformation on 12,700 pharmacological targets, allowing the scientists toreveal connections between compounds, effects and targets. The data is indexedand normalized for maximum searchability and consistency.

References

  1. Huguenard, J.R. and Prince, D.A. (1992) A novel T-type current underlies prolonged Ca2+-dependent burst firing in GABAergic neurons of rat thalamic reticular nucleus. J Neurosci 12: 3804–3817.
  2. Perez-Reyes E. (2003) Molecular physiology of low-voltage-activated t-type calcium channels. Physiol Rev 83: 117–161.
  3. Bourinet, E. and Zamponi, G.W. (2005) Voltage-gated calcium channels as targets for analgesics. Curr Top Med Chem 5: 539-546.
Share
Tweet
Share
Share
Share