Abstract
Background
The serotonin 5-HT5A receptor has attracted much more research attention, due to the therapeutic potential of its ligands being increasingly recognized, and the possibilities that lie ahead of these findings. There is a growing body of evidence indicating that these ligands have procognitive, pro-social, and anti-depressant properties, which offers new avenues for the development of treatments that could address socially important conditions related to the malfunctioning of the central nervous system. The aim of our study was to unravel the molecular determinants for 5-HT5AR ligands that govern their activity towards the receptor.
Methods
In response to the need for identification of molecular determinants for 5-HT5AR activity, we prepared a comprehensive collection of 5-HT5AR ligands, carefully gathering literature and patent data. Leveraging molecular modeling techniques, such as pharmacophore hypothesis development, docking, and molecular dynamics simulations enables to gain valuable insights into the specific interactions of 5-HT5AR ligand groups with the receptor.
Results
The obtained comprehensive set of 2160 compounds was divided into dozens of subsets, and a pharmacophore model was developed for each group. The results from the docking and molecular dynamics simulations have enabled the identification of crucial ligand–protein interactions that are essential for the compound's activity towards 5-HT5AR.
Conclusions
The findings from the molecular modeling study provide valuable insights that can guide medicinal chemists in the development of new 5-HT5AR ligands. Considering the pharmacological significance of these compounds, they have the potential to become impactful treatments for individuals and communities in the future. Understanding how different crystal/cryo-EM structures of 5-HT5AR affect molecular modeling experiments could have major implications for future computational studies on this receptor.
Graphical abstract
![](http://media.springernature.com/lw685/springer-static/image/art%3A10.1007%2Fs43440-024-00622-4/MediaObjects/43440_2024_622_Figa_HTML.png)
Similar content being viewed by others
Avoid common mistakes on your manuscript.
Introduction
The 5-hydroxytryptamine receptor 5A (5-HT5AR) is a G protein-coupled receptor (GPCR) belonging to the 5-HT receptor family [1, 2]. These receptors, categorized as aminergic receptors, bind to amines such as serotonin, dopamine, histamine, and norepinephrine [3, 4]. Their structural characteristic is a heptahelical domain, which consists of seven transmembrane (7TM) helices embedded in the cell membrane, allowing for communication between the extracellular environment and cytoplasmic signaling cascades and the nuclear cell-cycle machinery [5]. The 7TM domain is responsible for interacting with ligands, which trigger signal transduction and activate intracellular signaling pathways [6].
The 5-HT5AR was first cloned in 1992, and it is encoded by the 5-HT5A gene [7, 8]. It belongs to the 5-HT5R subfamily, together with the 5-HT5BR. Although two 5-HT5R genes (5-HT5A and 5-HT5B) have been described in the rodents, only 5-HT5AR is expressed in humans, whereas the coding sequence of 5-HT5BR is interrupted by stop codons, leading to a non-functional 5-HT5B gene [9].
A wide distribution of 5-HT5AR has already been proven in the human brain, with the highest expression rate in the cortex and limbic regions [10]. Its activity is regulated by serotonin, a neurotransmitter that governs various physiological functions, such as mood, cognition, sleep, circadian rhythms, thermoregulation, pain, etc. It also has a profound impact on hemostasis, vascular tone, heart rate, respiratory drive, cell growth, and immunity [11, 12]. Recent findings on its expression patterns in regions associated with important physiological functions have opened up new avenues for exploring 5-HT5AR as a promising therapeutic target for psychiatric disorders [13,14,15]. These studies have confirmed its dysregulation in depression, anxiety, and schizophrenia, as well as its role in circadian rhythm regulation [7, 16, 17].
It has been demonstrated that 5-HT5AR activation initiates multiple signal transduction cascades. It is negatively coupled to adenylyl cyclase, protein kinase A, and cAMP signaling via Gαi/o protein [18,19,20]. Moreover, the activation of 5-HT5AR transiently opens potassium channels, causing hyperpolarization of the cell membrane and a decrease in neuronal excitability [21].
The 5-HT5AR displays an intriguing pharmacological profile: it binds with the highest affinity to ergots, ergolines, and the synthetic agonist 5-carboxamidotryptamine (5-CT); however, most other agents acting on the 5-HT receptors lack the 5-HT5AR affinity [8, 22, 23]. Furthermore, the development of selective 5-HT5AR agents has been challenging and only a limited number of such compounds have been described so far [23, 24] (the detailed history of 5-HT5AR ligand discoveries can be found in the Supplementary Information, File S1). Nevertheless, the number of known 5-HT5AR ligands is continuously growing and the compound dataset associated with this receptor in the ChEMBL database [25] expands with each subsequent version of this repository. The growing interest in 5-HT5AR-oriented research is also reflected in the wealth of data available in the recently released patents, significantly expanding the collection of known 5-HT5AR ligands.
A growing number of data is driving the use of various statistical and machine learning approaches to uncover hidden patterns and connections within them. The swift and efficient processing of large datasets has established computational approaches as essential tools in drug design campaigns. They support drug design pipelines at every stage, starting from target identification and validation, via the detection of new potential ligands, through the optimization of their activity, physicochemical, and pharmacokinetic profile. They also play a crucial role in analyzing the outcome of in vitro and in vivo tests, results of clinical trials, and monitoring the drug effectiveness after its introduction to the market. The integration of in silico tools in the drug discovery process reduces costs and time devoted to the identification of successful drug candidates, sparking a revolution in the field of drug design [26,27,28,29].
Our study thoroughly identifies molecular determinants for 5-HT5AR activity from ligand- and structure-based perspectives. Our work also involves the compilation of the latest information on 5-HT5AR ligands present in both literature and patent data. The manually extracted information from patents (of several hundreds of records) greatly enhanced the ChEMBL-based dataset, resulting in a valuable collection of 5-HT5AR ligands. This resource is of great importance to the whole medicinal chemistry community, significantly supporting further research within this target (we share all data via the Supplementary Information, File S2). We analyzed the dataset of 5-HT5AR agents globally, but we also focused on particular chemical groups to gain a comprehensive understanding of the activity profile of particular compounds. By employing pharmacophore modeling, docking, and molecular dynamics (MD) simulations, we were able to gain a broad understanding of the activity profile of specific compounds and their ligand-receptor contacts using the recently released crystal/cryo-EM structures of this receptor [23, 30]. The comprehensive compilation, analysis, and comparison of existing ligands, along with a detailed examination of ligand-receptor contacts, will undoubtedly aid in developing new drug candidates displaying activity towards 5-HT5AR. The scheme of the whole study is presented in Fig. 1.
Materials and methods
At first, all records referring to the human 5-HT5AR were retrieved from the ChEMBL database version 33 (Target ChEMBLID: CHEMBL3426) [25]. Such a compound database was then enriched with the manually extracted patent data. The resulting dataset was subjected to manual clustering aided by the results of the automated grouping carried out in the Canvas module from the Schrödinger Suite 2023 [31]. For each compound group with at least ten ligands, the pharmacophore model was developed using Phase [32]. The compounds were docked to the available crystal structures of 5-HT5AR (7UM4 [23], 7UM5 [23], 7UM6 [23], 7UM7 [23]; 7X5H [30] was not considered due to crystallization with the same ligand as in 7UM5, and 7UM5 had better resolution). Docking was performed in two modes: separately for particular clusters (those with the highest number of examples) and collectively using the ChEMBL-based dataset. Due to a lack of information about the inactive compounds in patents, the collective analysis was limited to data included in the ChEMBL database. Glide [33] in extra precision mode was used for docking, with compounds prepared using LigPrep [34] from the Schrödinger Suite (protonation states generated at pH 7.4 ± 0.0, and all possible stereoisomers were enumerated; other settings remained at default). The reliability of docking was verified by re-docking the co-crystallized ligands and examination of the RMSD of the obtained conformations with reference to the co-crystallized pose. The resulting ligand-receptor complexes were encoded in the form of the structural interaction fingerprint (SIFt) [35], and the contact frequencies for a particular group of compounds were measured. Results obtained via the statistical analysis of docking were confronted with the mutagenetic data reported for the 5-HT5AR (GPCRdb [36] data were used). To gain a detailed understanding of the ligand–protein contacts in the compound complexes with 5-HT5AR for different crystal structures, a series of MD simulations was carried out for the selected set of quinoline derivatives (CHEMBL5077128, CHEMBL5082447, CHEMBL2005743), starting from the ligand-receptor complexes obtained in docking. To capture differences among the available crystal structures, only EC50 data were taken into account, as there is only one crystal structure available for 5-HT5AR in its inactive conformation. The MD simulations were carried out in Desmond [37], using TIP3P solvent model [38] and POPC (palmitoyl-oleil-phosphatidylcholine) as a membrane model and OPLS3e force-field under the pressure of 1.01325 bar and in the temperature of 300 K. The box shape was orthorhombic with dimensions of 10 Å × 10 Å × 10 Å. In each case, the system was neutralized by the addition of the respective number of Cl- ions and relaxed before the simulation; the duration of each simulation was equal to 1000 ns. At first, the results were analyzed in terms of the stability of the compound pose in the binding pocket. The assessment was carried out qualitatively through visual inspection of the compound orientations at specific time steps of the simulation (starting pose, after 250 ns, after 500 ns, after 750 ns, and after 1000 ns) and quantitatively by examining the compound's RMSF during the simulation. Finally, correlational studies were conducted to identify the positions with the highest correlation (expressed by Spearman’s correlation coefficient) between the frequency of compound interaction with particular amino acids and the outcome of the experimental verification of compound activity.
Results
Analysis of ChEMBL data
The distribution of activity parameter values for ChEMBL records referring to human 5-HT5AR is presented in Fig. 2.
Distribution of activity parameters values of 5-HT5AR ligands deposited in the ChEMBL database (orange lines refer to step changes in the scale), a Ki, b IC50, c EC50, d Venn diagram for particular parameters values determined for 5-HT5AR ligands, created on-line at https://bioinformatics.psb.ugent.be/webtools/Venn/; e the number of records present in the ChEMBL database referring to different serotonin receptor subtypes
Overall, there are 491 records with Ki values reported (Fig. 2a), 231 of them related to EC50 (Fig. 2c), and only 52 providing information about the IC50 values of 5-HT5AR ligands (Fig. 2b). These refer to 447 unique compounds with determined Ki values, 230 unique ligands with known EC50 values, and 45 with known IC50 values (Fig. 2d). Among these compounds, 135 have Ki values below 100 nM, and 3 compounds have IC50 values below 100 nM. Interestingly, for all ligands with EC50 values determined, they exceeded 100 nM, with the most active compound having an EC50 value of 354.8 nM).
When comparing the number of identified 5-HT5AR ligands to other related targets (Fig. 2), we find that it is significantly lower than the highest populated in ligands 5-HTR subtypes. The ChEMBL database currently contains only 1163 records for this receptor subtype, which indicates the need for further research and development in this area. In contrast, the number of records for 5-HT2AR and 5-HT1AR are almost 8000 (7904) and 7073, respectively. The other receptor subtypes with a high number of ligands include 5-HT2CR, 5-HT2BR, 5-HT6R, and 5-HT7R. However, some subtypes have relatively lower amounts of data, such as 5-HT1ER (607 records), 5-HT1FR (175 records), and 5-HT4R (942 records).
The ChEMBL-based compound database prepared within the study was also enriched with the manually extracted patent data, providing several hundreds of additional data points thoroughly described in further sections of the article and shared via the Supplementary Information. Such a huge amount of information on 5-HT5AR in patents is a promising sign of a growing interest in this target by pharmaceutical companies and suggests potential therapeutic applications of its ligands.
Clustering and characterization of different groups of 5-HT5AR ligands
Groups of 5-HT5AR ligands organized on the basis of their chemical structure are summarized in Figs. 3 and 4. The latter presentation also encompasses the distribution of Ki values associated with the specific cluster and pharmacophore model developed for each group.
The largest chemical class of known 5-HT5AR binders (as presented in Fig. 3) consists of acylguanidines with over 900 compounds reported [39,40,41,42,43]. It is worth pointing out that these compounds display high 5-HT5AR affinity, with the great majority of them having Ki values below 50 nM. Another significant chemical class of 5-HT5AR ligands includes 2-aminoquinolines (448) [44,45,46,47,48,49,50,51] and benzoxazines (172) [52, 53], many of which also have a significant number of compounds with Ki values towards 5-HT5AR below 50 nM (391 and 146, respectively). It is worth mentioning that these two classes of 5-HT5AR ligands are not present in the ChEMBL database; they are covered in patents. Among 5-HT5AR ligands, there are also 153 2-amino-dihydroquinazolines [54,55,56,57,58] with high-to-moderate affinity to the receptor, 97 N’-benzyl-N-(pyridin-2-yl)guanidines [59], 90 benzimidazoles [60], 59 biarylmethylamines [61], and 54 aminoimidazoles. The exact values of activity parameters are not provided for the aminoimidazoles, but the information included in the patent indicates that Ki values towards 5-HT5AR for the whole group of compounds are below 600 nM. There is also a group of 30 carbolines with moderate 5-HT5AR affinity [62, 63]. Additionally, there are three sets of 20-membered compound groups of ligands, namely arylpiperazines (ChEMBL data), lysergic acid derivatives (PDSP data), and isoquinoline amides (patented) [64]. Moreover, the ChEMBL database contains information about 13 1‐(quinolin‐8‐yl)methanamines [24], 12 quinazolines, 10 tricyclic compounds, and 10 2‐chloro‐2'‐methoxy‐1,1'‐biphenyls.
In addition, there are over a dozen groups of 5-HT5AR ligands with only a few example compounds reported so far (Fig. 4). The PDSP database contains arylpiperidines, 1,2,3,4-tetrahydronaphthalen-1-one- and mellpalladine B-based 5-HT5AR ligands, whereas in the ChEMBL database, we can find additionally tryptamines, 1,3,5-triazine-2,4-diamines, aplysinopsin analogs, diarylalkylamines, 1‐{2‐[1‐(arylsulfonyl)pyrrolidin‐2‐yl]ethyl}piperidines, 5-OH-DPAT analogs, 1,2,3,4-tetrahydroisoquinolines, and one N‑methyllaurotetanine-based ligand.
Docking
In 2022, significant progress was achieved in comprehending the crystal structures of the 5-HT5AR. As a result of extensive research, two papers revealed a total of 5 crystal/cryo-EM structures of this receptor [23, 30]. These findings have been instrumental in advancing our understanding of the receptor and its role in biological processes. When developing new ligands for a specific biological target through structure-based drug design, understanding the spatial structure of a protein is critical. This knowledge enhances the reliability of docking studies and paves the way for the development of new effective ligands.
The available crystal structures of the 5-HT5A receptor are summarized in Table 1 and the example structures with the co-crystallized ligands are presented in Fig. 5.
The reliability of docking studies was confirmed through the re-docking of the co-crystallized ligands. The RMSD values between the compound pose obtained in docking and the co-crystallized conformation were as follows: 0.205 Å, 0.041Å, 0.224 Å, and 0.283Å for 7UM4, 7UM5, 7UM6, and 7UM7 crystal structures, respectively.
The docking study outcomes were initially analyzed in a cluster-based manner, focusing on the most populated clusters of 5-HT5AR ligands. Such an approach enabled the indication of the amino acids that most frequently interact with particular groups of compounds (Fig. 6). For example, it revealed the consistent contact of all the analyzed groups of compounds with amino acids such as D3x32, V3x33, C3x36, V45x52, S5x43, A5x461, F6x51, F6x52, and E6x55. On the other hand, there is a noticeable consequent contact pattern for acylguanidines, which is not so typical for the other considered groups of 5-HT5AR ligands. For example, almost all compounds from this group make interact with E2x64, W3x28, S45x53, T5x44, L7x38, and Y7x42, which is not so frequent for 2-aminoquinolines, benzoxazines, 2-amino-dihydroquinazolines, and N’-benzyl-N-(pyridin-2-yl)guanidines. Docking results to 7UM4 of representatives of 2-aminoquinolines and acylguanidines (compounds presented in Fig. 3) are visualized in Fig. 7. Although both compounds display very high 5-HT5AR affinity (Ki = 3.30 nM and Ki = 0.68 nM for 2-aminoquinolines acylguanidine example, respectively) and occupy a similar area of the 5-HT5AR binding pocket, their orientation in the binding site, as well as the ligand–protein contacts formed are slightly different. In particular, both compounds form π-π contacts with the phenylalanine cluster (F6x51 and F6x52) and hydrogen bonds with E2x64; however, CHEMBL3654198 with the acylguanidine core makes also several hydrogen bonds with D3x32 whereas the 2-aminoquinoline-based ligand forms a hydrogen bond with Q45x51.
Docking poses of representatives of 2-aminoquinolines (orange) and acylguanidines (magenta) of structures presented in Fig. 3
When analyzing the docking studies outcome globally, the 5-HT5AR ligands extracted from the ChEMBL database (only ChEMBL data were used due to the biased activity distribution in patents) were categorized into active/inactive groups based on their Ki values: compounds with Ki values below 1000 nM were classified as active, while those with Ki values above 1000 nM were considered inactive.. The absolute difference between the fraction of interacting active compounds and the fraction of interacting inactive compounds was determined. Positions with the highest difference in interaction contact frequency between the two groups of compounds were indicated in Fig. 8.
In general, the results are slightly different when various 5-HT5AR crystal structures are considered, although in general, the indicated positions align with the crucial amino acids reported for other 5-HT receptor subtypes [65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80]. For example, the consistent variance in interaction frequency between active and inactive compounds across different crystal structures highlights the importance of the aspartic acid located in the third transmembrane helix (D3 × 32 according to the GPCRdb numbering). Still, only for 7UM4 (referring to the inactive conformation of 5-HT5AR) and 7UM5 (reflecting the active state of 5-HT5AR) the difference exceeded 10%, being equal to 18% and 14%, respectively. In contrast, for 7UM6 and 7UM7, the interaction rates of active and inactive compounds with D3x32 exhibited a more similar pattern, with differences of 7% for both crystal structures. Another position that was consistently identified for its discriminative power between active and inactive compounds was F6x52, although, for none of the studied crystal structures, the difference exceeded 5%.
Overall, the highest discriminative power was observed for the 7UM4 crystal structure with 3 positions (E2x64, D3x32, V3x33) showing active/inactive contact frequency difference exceeding 15% (21%, 18%, and 20%, respectively). Also for 7UM4 there was the highest total number of indicated positions (8). The crystal structures 7UM5 and 7UM6 displayed the highest discriminative power within the 7th transmembrane helix with differences of 18% for L7x38 in 7UM5 and 15% for Y7x42 in 7UM6. On the other hand, S5x43 was identified as the most discriminative position in the 7UM7-based studies.
The docking data were also analyzed for the functional experiments. At first, the potency of compounds to inhibit 5-HT5AR (IC50 data) was taken into account. As this parameter expresses the compound’s potency to inhibit the receptor, only its inactive conformation was considered (the 7UM4 crystal structure (Fig. 9)). The most evident observation regarding this data is that the indicated differences between the contact frequency of the confirmed 5-HT5AR antagonists and non-antagonists were much higher than in the Ki-based analysis. There are four positions (W3x28, I3x29, D3x32, and F6x51) for which the difference between the contact frequency of the two groups of compounds exceeded 30%. Qualitatively, the positions identified in the IC50-based studies align with those from the Ki-based experiments.
The analysis of the EC50 data was carried out in an analogous manner with some unique considerations due to the presence of only one active compound (Fig. 10). Figure 10 presents the positions for which the highest interaction frequency difference was observed, but with the assumption that particular contact did not occur for the examined agonist. Interestingly, for 7UM5- and 7UM6-based experiments, indeed the D3x32 position was not indicated, but for 7UM7-based studies, it is present on the diagram, which means that the one known 5-HT5AR agonist does not make contact with this residue when docked to 7UM7. Also, the phenylalanines from the 6th TM helix (F6x51 and F6x52) are not present in this comparison (as they make contact also with the active compound), but there is also a set of residues which were indicated in the previous analysis (Fig. 8), but they are not forming interaction with the examined agonist, such as W3x28, W6x48, V3x33, etc.
Molecular dynamics simulations
The structures of ligands that underwent MD simulations are presented in Fig. 11. The figure indicates that the inactive CHEMBL2005743 did not preserve the obtained docking pose and in the initial time of the simulation its conformation changed. On the other hand, the outcome of the 7UM7-based studies contrasts with these results. In this case, CHEMBL2005743 was found to be most stably fit in the 5-HT5AR binding pocket, whereas the most active CHEMBL5077128 changed its initial pose and remained in the new conformation until the end of the simulation.
The MD data were also analyzed more formally by the analysis of compound RMSF (Table 2).
The data provided in Table 2 revealed that although in the case of the 7UM5-based studies the inactive CHEMBL2005743 initially underwent a significant conformational change, but then stabilized in the subsequent frames of the simulation, with an average RMSF similar to that of the most active CHEMBL5077128. On the other hand, the conformation of CHEMBL5082447 varied significantly in the 5-HT5AR binding site when the 7UM5 crystal structure was taken for studies. The only crystal structure for which the activity-RMSF relationship for the examined compounds was preserved was 7UM6. However, before making the general recommendation for its usage for compound EC50 prediction, more extensive studies should be carried out, which is not possible at this moment due to the limited availability of agonistic structures reported so far.
Finally, the correlational studies between the frequency of compound interaction with particular amino acids were carried out to pinpoint positions with the highest Spearman’s correlation coefficient with the experimental activity verification results. This analysis led to the identification of the set of positions for which the correlation was equal to ± 1 (Table 3).
Confrontation with mutagenetic data
Results obtained via the computational studies were confronted with the mutagenetic data reported for the 5-HT5AR (visualization prepared in the GPCRdb service, Fig. 12). The list of amino acids indicated both in docking and MD simulations were consistent with the experimental outcome of the mutagenesis experiments, although there are variations in the observed effect of amino acid substitution (increase/decrease of a compound potency).
Mutagenetic data available for the 5-HT5AR; figure created with the use of the GPCRdb service [36]; yellow color indicates change of the ligand binding affinity upon particular residue mutation
There is clear evidence supporting the importance of the D3x32 for ligand binding, as all available data report the activity loss of a ligand upon the substitution of this amino acid by alanine [18, 25]. Both papers also report the activity change upon the F6x52A mutation; however, this effect is much less pronounced compared to the D3x32 substitution, as it is approx. 1.5-fold decrease in activity related to this mutation. E2x64 was the position for which high discriminative power was indicated for the 7UM4-based studies. The effect of the mutation at this position strongly depended on the substituted amino acid and the specific ligand being examined [23]. For example, the E2x64D mutation led to a slight increase in the activity of CHEMBL3654198, whereas the E2x64 mutation to glutamine or alanine led to a significant decrease of this ligand affinity to 5-HT5AR. On the other hand, when lysergide was used as the examined ligand, all E2x64D, E2x64Q, and E2x64A substitutions resulted in the worsening of the ligand affinity. Another position with high discriminative power for 7UM4-based docking experiments was V3x33 (20% of the difference between the contact frequency of active and inactive compounds) and for this amino acid, the mutational data consistently indicate a significant decrease in activity for all examined ligands. The mutation of L7x38 was extensively examined for different ligands and varying substitution schemes, leading to a slight decrease of affinity in the majority of cases and a lack of change in one case. The importance of the Y7x42 position is undeniable, as its mutation led to either a significant decrease in ligand affinity or a complete loss of activity. Position S5x43, which displayed high discriminative power in 7UM7-based docking studies, is related to inconsistent mutagenetic data – the influence on ligand affinity (increase/decrease) depended on the specific ligand and the substituted amino acid.
Discussion
The conducted study on 5-HT5AR ligands has produced a wealth of constructive insights into the nature of this target. The results obtained have confirmed the significance of 5-HT5AR in drug design applications and provided valuable structural and interaction-based information for designing agents with the desired activity profile. The extensive range of reported compounds and their diverse structural characteristics and biological activity towards the target showcase the potential of 5-HT5AR in pharmacological research. The distribution of ligands between particular data sources highlights the importance of the conducted search and manual extraction of compound structures from patents. The most populated classes of 5-HT5AR ligands, that is acylguanidines, 2-aminoquinolines, and benzoxazines, which contained in total over 1500 records, were fetched exclusively from patents, which emphasizes the necessity of manual extraction for incorporating such compounds in the dataset. Despite quite significant structural diversity of the presented ligands, there are common features which were revealed by pharmacophore modeling, such as aromatic rings together with hydrophobic and donor moieties. The geometric relationships between these features vary for different chemical classes considered; however, there is a set of ligand–protein contacts, which are supposed to be formed in order to maximize the probability of 5-HT5AR activity. These interactions include contacts with D3x32, V3x33, C3x36, V45x52, S5x43, F6x51, and F6x52 amino acids, which were confirmed both in in silico studies (docking and molecular dynamics simulations), as well as in the experimental research (mutagenetic studies). The alignment of the docking outcome with the previously reported mutagenetic data confirms the reliability of the computational approaches applied. Further validation of the applied computational protocols was obtained through the correlational studies of the MD output with the in vitro reports. All the obtained results can be utilized in the design of agents with selective or non-selective activity profiles towards 5-HT5AR.
Conclusions
In the study, we prepared a comprehensive collection of 2160 5-HT5AR ligands supplied with a detailed examination of their potential interactions with the target protein. The systematic description of the chemical groups of already known ligands, supplied with the pharmacophore modeling and in-depth analysis of docking results and MD simulation studies for selected ligands, enabled the determination of key factors to consider when designing new 5-HT5AR agents. By focusing on different chemical classes of 5-HT5AR ligands during modeling, we have enhanced the utility gained knowledge by capturing in detail the ligand–protein contacts characteristic for a given scaffold. The incorporation of different crystal structures and bioactivity data from various sources makes the study comprehensive and highly valuable for the medicinal chemistry community, providing guidance for successful development of compounds displaying activity towards 5-HT5AR. Furthermore, the MD-based analysis is supported by the correlation of the computational outcome with the biological results measured in vitro, which further validates our findings.
Data availability
The 5-HT5AR dataset prepared in the study is available in the Supplementary Information. Detailed molecular modeling outcomes are available from authors upon request.
Abbreviations
- 5-CT:
-
5-Carboxamidotryptamine
- 5-HT5AR:
-
5-Hydroxytryptamine receptor
- 7TM:
-
Seven transmembrane
- GPCR:
-
G protein-coupled receptor
- MD:
-
Molecular dynamics
- SAR:
-
Structure–activity relationship
- SIFt:
-
Structural interaction fingerprint
References
Pytliak M, Vargová V, Mechírová V, Felšöci M. Serotonin receptors—From molecular biology to clinical applications. Physiol Res. 2011;60(1):15–25.
Volk B, Nagy BJ, Vas S, Kostyalik D, Simig G, Bagdy G. Medicinal chemistry of 5-HT5A receptor ligands: a receptor subtype with unique therapeutical potential. Curr Top Med Chem. 2010;10(5):554–78.
Michino M, Beuming T, Donthamsetti P, Newman AH, Javitch JA, Shi L. What can crystal structures of aminergic receptors tell us about designing subtype-selective ligands? Pharmacol Rev. 2015;67(1):198–213.
Shi L, Javitch JA. The binding site of aminergic G protein-coupled receptors: the transmembrane segments and second extracellular loop. Annu Rev Pharmacol Toxicol. 2002;42:437–67.
Peeper DS, Bernards R. Communication between the extracellular environment, cytoplasmic signalling cascades and the nuclear cell-cycle machinery. FEBS Lett. 1997;410(1):11–6.
Lee KH, Manning JJ, Javitch J, Shi L. A novel, “activation switch” motif common to all aminergic receptors. J Chem Inf Model. 2023;63(16):5001–17.
Plassat JL, Boschert U, Amlaiky N, Hen R. The mouse 5HT5 receptor reveals a remarkable heterogeneity within the 5HT1D receptor family. EMBO J. 1992;11(13):4779–86.
Rees S, den Daas I, Foord S, Goodson S, Bull D, Kilpatrick G, Lee M. Cloning and characterisation of the human 5-HT5A serotonin receptor. FEBS Lett. 1994;355(3):242–6.
Grailhe R, Grabtree GW. Hen R Human 5-HT(5) receptors: the 5-HT(5A) receptor is functional but the 5-HT(5B) receptor was lost during mammalian evolution. Eur J Pharmacol. 2001;418(3):157–67.
Pasqualetti M, Ori M, Nardi I, Castagna M, Cassano GB, Marazziti D. Distribution of the 5-HT5A serotonin receptor mRNA in the human brain. Brain Res Mol Brain Res. 1998;56(1–2):1–8.
Barnes NM, Ahern GP, Becamel C, Bockaert J, Camilleri M, Chaumont-Dubel S, et al. International union of basic and clinical pharmacology. CX. Classification of receptors for 5-hydroxytryptamine; pharmacology and function. Pharmacol Rev. 2021;73(1):310–520.
Berger M, Gray JA. Roth BL The expanded biology of serotonin. Annu Rev Med. 2009;60:355–66.
Vidal-Cantú GC, Jiménez-Hernández M, Rocha-González HI, Villalón CM, Granados-Soto V, Muñoz-Islas E. Role of 5-HT5A and 5-HT1B/1D receptors in the antinociception produced by ergotamine and valerenic acid in the rat formalin test. Eur J Pharmacol. 2016;781:109–16.
Sagi Y, Medrihan L, George K, Barney M, McCabe KA, Greengard P. Emergence of 5-HT5A signaling in parvalbumin neurons mediates delayed antidepressant action. Mol Psychiatry. 2020;25(6):1191–201.
Yamazaki M, Harada K, Yamamoto N, Yarimizu J, Okabe M, Shimada T, Ni K, Matsuoka N. ASP5736, a novel 5-HT5A receptor antagonist, ameliorates positive symptoms and cognitive impairment in animal models of schizophrenia. Eur Neuropsychopharmacol. 2014;24(10):1698–708.
Geurts FJ, De Schutter E, Timmermans JP. Localization of 5-HT2A, 5-HT3, 5-HT5A and 5-HT7 receptor-like immunoreactivity in the rat cerebellum. J Chem Neuroanat. 2002;24(1):65–74.
Jouvet M. Sleep and serotonin: an unfinished story. Neuropsychopharmacology. 1999;21(2 Suppl):24S-27S.
Masson J, Emerit MB, Hamon M. Darmon M serotonergic signaling: multiple effectors and pleiotropic effects. Wiley Interdiscipl Rev Membr Transp Signal. 2012;1(6):685–713.
Francken BJ, Josson K, Lijnen P, Jurzak M, Luyten WH. Leysen JE human 5-hydroxytryptamine(5A) receptors activate coexpressed G(i) and G(o) proteins in Spodoptera frugiperda 9 cells. Mol Pharmacol. 2000;57(5):1034–44.
Noda M, Yasuda S, Okada M, Higashida H, Shimada A, Iwata N, et al. Recombinant human serotonin 5A receptors stably expressed in C6 glioma cells couple to multiple signal transduction pathways. J Neurochem. 2003;84(22):222–32.
Noda M, Higashida H, Aoki S, Wada K. Multiple signal transduction pathways mediated by 5-HT receptors. Mol Neurobiol. 2004;29(1):31–9.
Erlander MG, Lovenberg TW, Baron BM, de Lecea L, Danielson PE, Racke M, et al. Two members of a distinct subfamily of 5-hydroxytryptamine receptors differentially expressed in rat brain. Proc Natl Acad Sci U S A. 1993;90(8):3452–6.
Zhang S, Chen H, Zhang C, Yang Y, Popov P, Liu J, et al. Inactive and active state structures template selective tools for the human 5-HT5A receptor. Nat Struct Mol Biol. 2022;29(7):677–87.
Levit Kaplan A, Strachan RT, Braz JM, Craik V, Slocum S, Mangano T, et al. Structure-based design of a chemical probe set for the 5-HT5A serotonin receptor. J Med Chem. 2022;65(5):4201–17.
Zdrazil B, Felix E, Hunter F, Manners EJ, Blackshaw J, Corbett S, et al. The ChEMBL Database in 2023: a drug discovery platform spanning multiple bioactivity data types and time periods. Nucleic Acids Res. 2024;52(D1):D1180–92.
Sliwoski G, Kothiwale S, Meiler J, Lowe EW Jr. Computational methods in drug discovery. Pharmacol Rev. 2013;66(1):334–95.
Shaker B, Ahmad S, Lee J, Jung C, Na D. In silico methods and tools for drug discovery. Comput Biol Med. 2021;137:104851.
Velmurugan D, Pachaiappan R, Ramakrishnan C. Recent trends in drug design and discovery. Curr Top Med Chem. 2020;20(19):1761–70.
Sadybekov AV. Katritch V computational approaches streamlining drug discovery. Nature. 2023;616(7958):673–85.
Tan Y, Xu P, Huang S, Yang G, Zhou F, He X, Ma H, Xu HE, Jiang Y. Structural insights into the ligand binding and Gi coupling of serotonin receptor 5-HT5A. Cell Discov. 2022;8(1):50.
Schrödinger Release 2023–3: Canvas, Schrödinger, LLC, New York, NY, 2023.
Schrödinger Release 2023–3: Phase, Schrödinger, LLC, New York, NY; 2023.
Schrödinger Release 2023–3: Glide, Schrödinger, LLC, New York, NY; 2023.
Schrödinger Release 2023–3: LigPrep, Schrödinger, LLC, New York, NY; 2023.
Deng Z, Chuaqui C, Singh J. Structural interaction fingerprint (SIFt): a novel method for analyzing three-dimensional protein-ligand binding interactions. J Med Chem. 2004;47(2):337–44.
Pándy-Szekeres G, Caroli J, Mamyrbekov A, Kermani AA, Keserű GM, Kooistra AJ, Gloriam DE. GPCRdb in 2023: state-specific structure models using AlphaFold2 and new ligand resources. Nucleic Acids Res. 2023;51(D1):D395–402.
Schrödinger Release 2023–3: Desmond, Schrödinger, LLC, New York, NY; 2023.
Jorgensen WL, Chandrasekhar J, Madura JD, Impey RW. Klein ML comparison of simple potential functions for simulating liquid water. J Chem Phys. 1983;79(2):26.
Madapa S, Harding WW. Semisynthetic studies on and biological evaluation of N-methyllaurotetanine analogues as ligands for 5-HT receptors. J Nat Prod. 2015;78(4):722–9.
Hamaguchi W, Kinoyama I, Koganemaru Y, Miyazaki T, Kaneko O, Sekioka R, Wasio T. Tetrahydroisoquinoline derivative. US-8962612B2. 2015.
Kinoyama I, Miyazaki T, Koganemaru Y, Shiraishi N, Kawamoto Y, Washio T. Acylguanidine derivative. WO2010090304A1. 2010.
Kinoyama I, Koganemaru Y, Miyazaki T, Washio T. Substituted acylguanidine derivative. WO2010090305A1. 2010.
Kinoyama I, Miyazaki T, Koganemaru Y, Washio T., Hamaguchi W. Nitrogenous-ring acylguanidine derivative. WO2011016504A1. 2011.
Kolczewski S, Riemer C, Roche O, Steward L, Wichmann J, Woltering T. 2-aminoquinolines. WO2009109493A2. 2009.
Kolczewski S, Riemer C, Steward L, Wichmann J, Woltering T. 2-aminoquinoline derivatives. US7825253B2. 2010.
Kolczewski S, Riemer C, Steward L, Wichmann J, Woltering T. 2-aminoquinolines as 5-ht(5a) receptor antagonists. WO2008068157A1. 2008.
Kolczewski S, Riemer C, Steward L, Wichmann J, Woltering T. Quinoline derivatives as 5ht5a receptor antagonists. WO2009040290A1. 2009.
Kolczewski S, Riemer C, Roche O, Steward L, Wichmann J, Woltering T. 2-aminoquinolines.WO2009109477A1. 2009.
Kolczewski S, Riemer C, Roche O, Steward L, Wichmann J, Woltering T. 2-aminoquinoline derivatives. WO2009109491A1. 2009.
Kolczewski S, Riemer C, Roche O, Steward L, Wichmann J, Woltering T. 2-aminoquinolines. WO2009109502A1. 2009.
Kolczewski S, Riemer C, Roche O, Steward L, Wichmann J, Woltering T. 2-aminoquinolines as 5-ht5a receptor antagonists. WO2009112395A1. 2009.
Kolczewski S, Roche O, Steward L, Wichmann J, Woltering T. 6-substituted benzoxazines. WO2010026110A2. 2010.
Kolczewski S, Roche O, Steward L, Wichmann J, Woltering T. 5-substituted benzoxazines. WO2010026112A1. 2010.
Alanine A, Gobbi LC, Kolczewski S, Leubbers T, Peters J-U, Steward L. Use of 2 -anilino - 3 , 4 -dihydro - quinazolines as 5ht5a receptor antagonists. WO2006097391A1. 2006.
Peters JU, Lübbers T, Alanine A, Kolczewski S, Blasco F, Steward L. Cyclic guanidines as dual 5-HT5A/5-HT7 receptor ligands: optimising brain penetration. Bioorg Med Chem Lett. 2008;18(1):262–6.
Alanine A, Gobbi LC, Kolczewski S, Leubbers T, Peters J-U, Steward L. (3,4-dihydro-quinazolin-2-yl)-indan-1-yl-amines. US7348332B2. 2008.
Alanine A, Gobbi LC, Kolczewski S, Leubbers T, Peters J-U, Steward L. 8-alkoxy or cycloalkoxy-4-methyl-3,4-dihydro-quinazolin-2-ylamines. US7790733B2. 2010.
Alanine A, Gobbi LC, Kolczewski S, Leubbers T, Peters J-U, Steward L. 5-Chloro-4-alkyl-3,4-dihydro-quinazolin-2-ylamine derivatives. US7790732B2. 2010.
Amberg W, Netz A, Kling A, Ochse M, Lange U, Hutchins CW, et al. Hetaryl-substituted guanidine compounds and use thereof as binding partners for 5-HT5-receptors. US9296697, 2016.
Suzuki R, Katayama K, Ueno S, Sugimoto Y, Watanabe H. Benzimidazole derivative. WO/2020/095912, 2020.
Bromidge SM, Corbett DF, Heightman TD, Moss SF. Biaryl Compounds Having activity at the 5HT5A receptor. WO/2004/096771, 2004.
Khorana N, Smith C, Herrick-Davis K, Purohit A, Teitler M, Grella B, et al. Binding of tetrahydrocarboline derivatives at human 5-HT5A receptors. J Med Chem. 2003;46(18):3930–7.
Rajagopalan R, Bandyopadhyaya A, Rajagopalan DR, Rajagopalan P. The synthesis and comparative receptor binding affinities of novel, isomeric pyridoindolobenzazepine scaffolds. Bioorg Med Chem Lett. 2014;24(2):576–9.
Hamaguchi W, Koganemaru Y, Sekioka R, Osamu K, Kato K. Isoquinoline derivative. JP2014076948. 2014.
Pei Y, Wen X, Guo SC, Yang ZS, Zhang R, Xiao P, Sun JP. Structural insight into the selective agonist ST1936 binding of serotonin receptor 5-HT6. Biochem Biophys Res Commun. 2023;671:327–34.
Cao C, Barros-Álvarez X, Zhang S, Kim K, Dämgen MA, Panova O, et al. Signaling snapshots of a serotonin receptor activated by the prototypical psychedelic LSD. Neuron. 2022;110(19):3154–67.
Gumpper RH, Fay JF, Roth BL. Molecular insights into the regulation of constitutive activity by RNA editing of 5HT2C serotonin receptors. Cell Rep. 2022;40(7):111211.
Huang S, Xu P, Shen DD, Simon IA, Mao C, Tan Y, et al. GPCRs steer Gi and Gs selectivity via TM5-TM6 switches as revealed by structures of serotonin receptors. Mol Cell. 2022;82(14):2681-2695.e6.
Cao D, Yu J, Wang H, Luo Z, Liu X, He L, et al. Structure-based discovery of nonhallucinogenic psychedelic analogs. Science. 2022;375(6579):403–11.
Chen Z, Fan L, Wang H, Yu J, Lu D, Qi J, et al. Structure-based design of a novel third-generation antipsychotic drug lead with potential antidepressant properties. Nat Neurosci. 2022;25(1):39–49.
Wang C, Jiang Y, Ma J, Wu H, Wacker D, Katritch V, et al. Structural basis for molecular recognition at serotonin receptors. Science. 2013;340(6132):610–4.
Xu P, Huang S, Zhang H, Mao C, Zhou XE, Cheng X, et al. Structural insights into the lipid and ligand regulation of serotonin receptors. Nature. 2021;592(7854):469–73.
Kim K, Che T, Panova O, DiBerto JF, Lyu J, Krumm BE, et al. Structure of a hallucinogen-activated Gq-coupled 5-HT2A serotonin receptor. Cell. 2020;182(6):1574-1588.e19.
McCorvy JD, Wacker D, Wang S, Agegnehu B, Liu J, Lansu K, et al. Structural determinants of 5-HT2B receptor activation and biased agonism. Nat Struct Mol Biol. 2018;25(9):787–96.
García-Nafría J, Nehmé R, Edwards PC, Tate CG. Cryo-EM structure of the serotonin 5-HT1B receptor coupled to heterotrimeric Go. Nature. 2018;558(7711):620–3.
Peng Y, McCorvy JD, Harpsøe K, Lansu K, Yuan S, Popov P, et al. 5-HT2C receptor structures reveal the structural basis of GPCR polypharmacology. Cell. 2018;172(4):719-730.e14.
Ishchenko A, Wacker D, Kapoor M, Zhang A, Han GW, Basu S, et al. Structural insights into the extracellular recognition of the human serotonin 2B receptor by an antibody. Proc Natl Acad Sci U S A. 2017;114(31):8223–8.
Wacker D, Wang S, McCorvy JD, Betz RM, Venkatakrishnan AJ, Levit A, et al. Crystal structure of an LSD-bound human serotonin receptor. Cell. 2017;168(3):377-389.e12.
Stauch B, Cherezov V. Serial femtosecond crystallography of G protein-coupled receptors. Annu Rev Biophys. 2018;47:377–97.
Wacker D, Wang C, Katritch V, Han GW, Huang XP, Vardy E, et al. Structural features for functional selectivity at serotonin receptors. Science. 2013;340(6132):615–9.
Funding
The study was supported by the statutory funds of the Maj Institute of Pharmacology Polish Academy of Sciences, Department of Medicinal Chemistry.
Author information
Authors and Affiliations
Contributions
The manuscript was written through the contributions of all authors. All authors have given approval to the final version of the manuscript.
Corresponding author
Ethics declarations
Conflict of interest
The authors declare no conflict of interest.
Additional information
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Information
Below is the link to the electronic supplementary material.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Kordylewski, S.K., Bugno, R., Bojarski, A.J. et al. Uncovering the unique characteristics of different groups of 5-HT5AR ligands with reference to their interaction with the target protein. Pharmacol. Rep (2024). https://doi.org/10.1007/s43440-024-00622-4
Received:
Revised:
Accepted:
Published:
DOI: https://doi.org/10.1007/s43440-024-00622-4