Structures 4-n-propyl Piperazines as Non-Imidazole Histamine H3 Antagonists

Seven new low-temperature structures of 4-n-propylpiperazine derivatives, potential H3 receptor antagonists, have been determined by X-ray crystallography, with the following symmetry and unit cell parameters: 2-(4-propyl-piperazin-1-yl)oxazolo[4,5-c]pyridine (compound 1), P-1, 5.9496 Å, 12.4570 Å, 12.8656 Å, 112.445°, 95.687°, 103.040°; 2-(4-propyl-piperazin-1-yl)thia-zolo[4,5-c]pyridine (compound 2), I2/a, 22.2087 Å, 7.5519 Å, 19.9225 Å, β = 92.368°; 2-(4-propyl-piperazin-1-yl)oxazolo[5,4-c]pyridine (compound 3), C2/c, 51.1351 Å, 9.36026 Å, 7.19352 Å, β = 93.882°; 2-(4-propyl-piperazin-1-yl)thiazolo[5,4-c]pyridine (compound 4), Pbcn, 19.2189 Å, 20.6172 Å, 7.4439 Å; 2-(4-propylpiperazin-1-yl)[1,3]oxazolo[4,5-b]pyridine, hydrate (structure 5), Pbca, 7.4967 Å, 12.2531 Å, 36.9527 Å; 2-(4-propylpiperazin-1-yl)[1,3]oxazolo[4,5-b]pyridine, first polymorph (structure 6), P-1, 7.2634 Å, 11.1261 Å, 18.5460 Å, 80.561°, 80.848°, 76.840°; 2-(4-propylpiperazin-1-yl)[1,3]oxazolo[4,5-b]pyridine, second polymorph (structure 7), P21, 8.10852 Å, 7.06025 Å, 12.41650 Å, β = 92.2991°. All the compounds crystallized out as hydrobromides. Oxazole structures show a much greater tendency to form twin crystals than thiazole structures. All the investigated structures display N—H···Br hydrogen bonding. (ADME) analysis, including the assessment of absorption, distribution, metabolism, and excretion, determined the physicochemical properties, pharmacokinetics, drug similarity, and bioavailability radar, and confirmed the usefulness of the compounds in question for pharmaceutical utility. This work is a continuation of the research searching for a new lead of non-imidazole histamine H3 receptor antagonists.


Introduction
The histamine H3 receptor has been the subject of much recent interest, due to its central role in regulating neurotransmitter levels. This G-protein-coupled receptor acts as presynaptic auto-and heteroreceptor, mainly in the central nervous system (CNS) [1], controlling the synthesis and release of histamine, but also modulating several other neurotransmitter systems, e.g., acetylcholine [2,3], dopamine [4], noradrenalin [5], and serotonin [6]. A variety of potential therapeutic applications for H3 receptor antagonists/inverse agonists have been proposed as potential drugs for the treatment of several CNS disorders, such as attention-deficit hyperactivity disorder (ADHD) [7], Alzheimer's disease [8], and schizophrenia [9]. The physiological and pathophysiological implications of histamine H3 receptors increase the need for potent and selective ligands as pharmacological tools and for potential drug development. The first generation of H3 antagonists was characterized by the presence of an imidazole ring as in histamine, many of which have found utility as pharmacological tools [10]. In contrast to the early work in the field, most chemical series of current interest appear to be non-imidazole compounds because of the major disadvantages of the 4-substituted imidazole moiety, including poor brain penetration and issues related to hepatic cytochrome P450 enzyme inhibition, such as drug-drug interactions, liver toxicity, and inhibition of adrenal synthesis [11].
Some years ago, we reported the synthesis and pharmacological characterization of the series of 2-(4-propylpiperazin-1-yl)thiazolopyridines and their analogue 2-(4-propylpiperazin-1-yl)-oxazolopyridines in vitro. Thiazoles displayed a higher activity than their oxazole analogs [12]. The most active compounds of both series are presented in Figure 1.
most chemical series of current interest appear to be non-imidazole compounds because of the major disadvantages of the 4-substituted imidazole moiety, including poor brain penetration and issues related to hepatic cytochrome P450 enzyme inhibition, such as drug-drug interactions, liver toxicity, and inhibition of adrenal synthesis [11].
Some years ago, we reported the synthesis and pharmacological characterization of the series of 2-(4-propylpiperazin-1-yl)thiazolopyridines and their analogue 2-(4propylpiperazin-1-yl)-oxazolopyridines in vitro. Thiazoles displayed a higher activity than their oxazole analogs [12]. The most active compounds of both series are presented in Figure 1.

Materials and Methods
The synthesis of all compounds was described by Walczyński [12]. Single crystals of compounds 1, 2, 3, 4, 5, 6 and 7, suitable for X-ray diffraction, were obtained from methanol-water-DMF (1:1:1 v/v) solutions by slow evaporation of the solvents at room temperature. The crystal measurements were performed on a XtaLAB Synergy diffractometer, Dualflex, Pilatus (Bioz Stars, Los Altos, CA, USA) 300 K [13]. All diffraction experiments were carried out with CuKα radiation. Diffraction data were processed with CrysAlis The rational design of new highly active compounds that selectively bind to specific receptors requires both knowledge of the active site structure of the receptor and the ligand itself. The compound represented by structures 5, 6, and 7, having no significant activity, was added for comparative purposes.

Materials and Methods
The synthesis of all compounds was described by Walczyński [12]. Single crystals of compounds 1, 2, 3, 4, 5, 6 and 7, suitable for X-ray diffraction, were obtained from methanolwater-DMF (1:1:1 v/v) solutions by slow evaporation of the solvents at room temperature. The crystal measurements were performed on a XtaLAB Synergy diffractometer, Dualflex, Pilatus (Bioz Stars, Los Altos, CA, USA) 300 K [13]. All diffraction experiments were carried out with CuKα radiation. Diffraction data were processed with CrysAlis PRO (Rigaku Oxford Diffraction. CrysAlis PRO; Rigaku Oxford Diffraction Ltd: Yarnton, Oxfordshire, England, 2020) [14]. Crystal structure solution and refinement were carried out with SHELX [15,16]. All H atoms (except majority of those engaged in hydrogen bonds) were geometrically optimized and allowed as riding atoms, with C-H = 0.95 Å for aromatic CH groups, 0.97 Å for secondary CH 2 groups and 0.96 Å for methyl groups, and N-H = 0.86 Å, with U iso (H) = 1.2 U eq (C, N). In all studied structures, the methyl H atoms were refined with U iso (H) = 1.5 U eq (C).
Quantum calculations were performed with GAMESS-US quantum computing package [17] using DFT/B3LYP [18][19][20][21] functional with the base functions 6-311 G(d,p) to optimize the geometry of the studied compounds, taking into account the solvent effect of water using polarizable continuum model. The MOLDEN package [22] was used for the preparation of input files and for visualization purposes. MULTIWFN was used for calculation of CHELPG charges [23].
Additionally, ADME analysis was performed to test the use of the studied compounds as drugs. ADME analysis was performed using SwissADME service (Swiss Institute of Bioinformatics 2021) [24][25][26] and ProTOX II service to predict toxicities of tested compounds [27].

Crystallography
The crystal data, data collection, and structure refinement details are summarized in Table 1 (full details are deposited in supplementary materials). Except structure 7, all the oxazole structures revealed a propensity for twinning. For structure 1, two components of the crystal were rotated approximately 2 • , and were difficult to separate, hence the data were processed as a single crystal, giving rise to high mosaicity. For structure 3, many components were identified in all the studied crystals. Finally, a crystal with four components was chosen for the diffraction experiment. The contributions of the components to the total intensity were the following: 0.51, 0.23, 0.02, and 0.24. Moreover, in this structure, very high disorder for Br anions, and probably DMF and water molecules, was observed, which could not be resolved, and, consequently, the SQUEEZE procedure [28] has been applied. For structure 5, four components were also identified, with the following contributions to the total intensity: 0.37, 0.09, 0.25, and 0.29. For structure 6, only two components were identified, with contributions of 0.58 and 0.42.
In all the oxazole structures, high values of residual electron density were located in the difference electron density maps in the vicinity of Br anions (Table 1, Figure 2a). This is probably (at least partly) due to the anharmonicity of the thermal motion of the ions, because taking into account the third and fourth orders of displacement parameters (25 additional parameters) (Olex2 v.1.3.0) [29] leads to a significant reduction in the residual maxima for structure 1 (from 1.6 to 1.1 e/A 3 , and from −1.0 to −0.4 e/A 3 ) (Figure 2b), and to a reduction in the R factor from 4.2% to 3.5%, with a still relatively high N ref /N par ratio of 14. Similar behavior is observed for, e.g., structure 7, where the reduction in the residual maxima amounts to 1.4 e/A 3 (from 2.2 to 0.8 e/A 3 ) and the reduction in the R factor from 3.3% to 2.1%, with an N ref /N par ratio of 15. On the contrary, for thiazole derivatives (2, 4), such anomalies in the difference Fourier maps were not observed.  The molecular structures of the studied compounds in their crystals are shown i Figure 3. The molecular structures of the studied compounds in their crystals are shown in Figure 3.  crystallizes as a solvate, with a disordered DMF molecule built into the structure. In contrast to the 2, 4, 6, and 7 compounds, 1 and 5 crystallized out as a diprotonated hydrate.
For structure 3, very high structural disorder is observed. Large maxima in the difference Fourier map with electron densities of~19.2 and 11.2 e/A 3 ( Figure 4) can be attributed to sites partially occupied by bromide anions, with displacement factors~3 times higher than that for the Br1 anion. The remaining maxima seem to be uninterpretable. The structures 2, 4, 6, and 7 crystallized out in the monoprotonated form. In all the investigated compounds, the piperazine nitrogen atom N11 is protonated. Compound 2 crystallizes as a solvate, with a disordered DMF molecule built into the structure. In contrast to the 2, 4, 6, and 7 compounds, 1 and 5 crystallized out as a diprotonated hydrate.
For structure 3, very high structural disorder is observed. Large maxima in the difference Fourier map with electron densities of ~19.2 and 11.2 e/A 3 ( Figure 4) can be attributed to sites partially occupied by bromide anions, with displacement factors ~3 times higher than that for the Br1 anion. The remaining maxima seem to be uninterpretable. showing the area of electron density representing delocalized solvent molecules (probably DMF and water) and probable sites of disordered bromide anions (two big yellow ellipsoids), which were excluded from the refinement and instead SQEEZE procedure was applied. Drawing prepared with ShelXle.
Omitting this whole part of the structure, one can observe large voids forming channels throughout the crystal structure along the [001] direction ( Figure 5). Such channel-like structures may lead to high disorder that is difficult to resolve, which is why we decided to use the SQUEEZE procedure in PLATON. The resulting voids constitute approximately 18% of the volume of the unit cell, and there were 149 recovered electrons in a single void. As we can observe, e.g., for structures 1, 2, or 5, both water and DMF molecules can be built into the structure of the studied compounds. Hence, we assume that there are two bromide anions in each channel, one DMF, and four water molecules, which gives 150 electrons per void.  showing the area of electron density representing delocalized solvent molecules (probably DMF and water) and probable sites of disordered bromide anions (two big yellow ellipsoids), which were excluded from the refinement and instead SQEEZE procedure was applied. Drawing prepared with ShelXle.
Omitting this whole part of the structure, one can observe large voids forming channels throughout the crystal structure along the [001] direction ( Figure 5). Such channel-like structures may lead to high disorder that is difficult to resolve, which is why we decided to use the SQUEEZE procedure in PLATON. The resulting voids constitute approximately 18% of the volume of the unit cell, and there were 149 recovered electrons in a single void. As we can observe, e.g., for structures 1, 2, or 5, both water and DMF molecules can be built into the structure of the studied compounds. Hence, we assume that there are two bromide anions in each channel, one DMF, and four water molecules, which gives 150 electrons per void.
The structures 2, 4, 6, and 7 crystallized out in the monoprotonated form. In all the investigated compounds, the piperazine nitrogen atom N11 is protonated. Compound 2 crystallizes as a solvate, with a disordered DMF molecule built into the structure. In contrast to the 2, 4, 6, and 7 compounds, 1 and 5 crystallized out as a diprotonated hydrate.
For structure 3, very high structural disorder is observed. Large maxima in the difference Fourier map with electron densities of ~19.2 and 11.2 e/A 3 ( Figure 4) can be attributed to sites partially occupied by bromide anions, with displacement factors ~3 times higher than that for the Br1 anion. The remaining maxima seem to be uninterpretable. Figure 4. Difference electron density map for 3 showing the area of electron density representing delocalized solvent molecules (probably DMF and water) and probable sites of disordered bromide anions (two big yellow ellipsoids), which were excluded from the refinement and instead SQEEZE procedure was applied. Drawing prepared with ShelXle.
Omitting this whole part of the structure, one can observe large voids forming channels throughout the crystal structure along the [001] direction ( Figure 5). Such channel-like structures may lead to high disorder that is difficult to resolve, which is why we decided to use the SQUEEZE procedure in PLATON. The resulting voids constitute approximately 18% of the volume of the unit cell, and there were 149 recovered electrons in a single void. As we can observe, e.g., for structures 1, 2, or 5, both water and DMF molecules can be built into the structure of the studied compounds. Hence, we assume that there are two bromide anions in each channel, one DMF, and four water molecules, which gives 150 electrons per void.  Thus, it seems that this structure contains 1.5 bromide anions per molecule with the protonated piperazine nitrogen atom N11 and the half-protonated nitrogen atom N25, which ensures that charge balance is preserved. Hence, the crystal structure is a 1:1 mixture of mono-cations and di-cations of compound 3. The disordered Br − anion (1/2 per molecule of 3) and solvent molecules (0.25 DMF per molecule and 1 H 2 O per molecule) are located in the channel of the blur of electron density (Figure 4).
Thus, it seems that this structure contains 1.5 bromide anions per molecule with protonated piperazine nitrogen atom N11 and the half-protonated nitrogen atom N which ensures that charge balance is preserved. Hence, the crystal structure is a mixture of mono-cations and di-cations of compound 3. The disordered Br − anion (1/2 p molecule of 3) and solvent molecules (0.25 DMF per molecule and 1 H2O per molecu are located in the channel of the blur of electron density (Figure 4).
Materials 2021, 14, x FOR PEER REVIEW 12 of Structures 5, 6, and 7 were obtained from the same sample. Structure 5 is in the for of a dibromohydrate. The packing of this structure differs from the analogo dibromohydrate (compound 1) (Figure 14). In this case, all the strong hydrogen bon (Table 6 and Figure 15) form a chain. Additionally, the weak hydrogen bonds (C-H… Br, or N type) stabilize the packing of the molecules.   Structures 6 and 7, on the other hand, are polymorphs of this compound. Both these structures are in the form of bromide, and have the same strong hydrogen bon between bromine and the piperazine system (Br...H-N) (Figures 16 and 17, and Tables and 8). Analysis of molecular packing (Figures 18 and 19) reveals different polymorph In structure 7, a layered arrangement of molecules is observed (Figure 19). Analysis of molecular packing (Figures 18 and 19) reveals different polymorphs. In structure 7, a layered arrangement of molecules is observed ( Figure 19). (Table 6 and Figure 15) form a chain. Additionally, the weak hydrogen bonds (C-H… Br, or N type) stabilize the packing of the molecules.   Structures 6 and 7, on the other hand, are polymorphs of this compound. Both these structures are in the form of bromide, and have the same strong hydrogen bo between bromine and the piperazine system (Br...H-N) (Figures 16 and 17, and Table  and 8). Analysis of molecular packing (Figures 18 and 19) reveals different polymorp In structure 7, a layered arrangement of molecules is observed (Figure 19).      The greatest variation in the geometry of the studied molecules can be observed on the nitrogen atom N14, the conformation of which determines the direction of the tworing aromatic system (Figure 20). This conformation can be defined as the angle, let us call it δ, between the N14-C21 bond and the plane formed by the following atoms: C13, N14,        The greatest variation in the geometry of the studied molecules can be observed on the nitrogen atom N14, the conformation of which determines the direction of the tworing aromatic system (Figure 20). This conformation can be defined as the angle, let us call     The greatest variation in the geometry of the studied molecules can be observed on the nitrogen atom N14, the conformation of which determines the direction of the two ring aromatic system (Figure 20). This conformation can be defined as the angle, let us cal The greatest variation in the geometry of the studied molecules can be observed on the nitrogen atom N14, the conformation of which determines the direction of the two-ring aromatic system (Figure 20). This conformation can be defined as the angle, let us call it δ, between the N14-C21 bond and the plane formed by the following atoms: C13, N14, and C15. For the most planar case (structure 5), the angle is less than 1 • , and the highest value of this angle, almost 30 • , is for structure 2. The structural feature that clearly distinguishes thiazole derivatives from oxazole derivatives is the geometry of the five-membered ring. This geometry, in turn, determines the position of a piperazine ring (Figure 21), which can influence the biological activity of the compounds. It is known that, among the studied structures, the thiazoles are more active than the oxazoles [12]. Overlay of molecules for all determined structures; 1 is blue, 2 is red, 3 is cyan, 4 is orange, 5 is green, 6 is light green and 7 is grey.

Density Functional Theory (DFT) Calculations
We were interested in how many conformations of the studied molecules in their crystal environment were different from those optimized in solution, where molecules experience more conformational freedom, to observe how the interactions in the crystal modify these conformations. For calculations, we used the GAMESS-US quantum computing package with the parametrization described in the Materials and Methods section.
It turned out that the optimized conformations were very close to the crystallographic conformations for all the examined structures. Thus, it seems that the degree of freedom associated with the above-mentioned angle δ is very "soft", and, therefore, its variability among the examined structures is most likely due to the intermolecular interactions in the crystalline state. The other parameters determining the geometry of the optimized molecules are also very similar to those observed in crystals; hence, in the case of the tested compounds, crystallographic studies provide structural The structural feature that clearly distinguishes thiazole derivatives from oxazole derivatives is the geometry of the five-membered ring. This geometry, in turn, determines the position of a piperazine ring (Figure 21), which can influence the biological activity of the compounds. It is known that, among the studied structures, the thiazoles are more active than the oxazoles [12]. The structural feature that clearly distinguishes thiazole derivatives from oxazole derivatives is the geometry of the five-membered ring. This geometry, in turn, determines the position of a piperazine ring (Figure 21), which can influence the biological activity of the compounds. It is known that, among the studied structures, the thiazoles are more active than the oxazoles [12]. Overlay of molecules for all determined structures; 1 is blue, 2 is red, 3 is cyan, 4 is orange, 5 is green, 6 is light green and 7 is grey.

Density Functional Theory (DFT) Calculations
We were interested in how many conformations of the studied molecules in their crystal environment were different from those optimized in solution, where molecules experience more conformational freedom, to observe how the interactions in the crystal modify these conformations. For calculations, we used the GAMESS-US quantum computing package with the parametrization described in the Materials and Methods section.
It turned out that the optimized conformations were very close to the crystallographic conformations for all the examined structures. Thus, it seems that the degree of freedom associated with the above-mentioned angle δ is very "soft", and, therefore, its variability among the examined structures is most likely due to the intermolecular interactions in the crystalline state. The other parameters determining the geometry of the optimized molecules are also very similar to those observed in crystals; hence, in the case of the tested compounds, crystallographic studies provide structural information that can be transferred directly to the non-crystalline state.
Using the geometry of molecule 3 as a starting point, we optimized its geometry in an unprotonated form, and calculated the electrostatic potential (ESP) charges with the Figure 21. Overlay of molecules for all determined structures; 1 is blue, 2 is red, 3 is cyan, 4 is orange, 5 is green, 6 is light green and 7 is grey.

Density Functional Theory (DFT) Calculations
We were interested in how many conformations of the studied molecules in their crystal environment were different from those optimized in solution, where molecules experience more conformational freedom, to observe how the interactions in the crystal modify these conformations. For calculations, we used the GAMESS-US quantum computing package with the parametrization described in the Materials and Methods section.
It turned out that the optimized conformations were very close to the crystallographic conformations for all the examined structures. Thus, it seems that the degree of freedom associated with the above-mentioned angle δ is very "soft", and, therefore, its variability among the examined structures is most likely due to the intermolecular interactions in the crystalline state. The other parameters determining the geometry of the optimized molecules are also very similar to those observed in crystals; hence, in the case of the tested compounds, crystallographic studies provide structural information that can be transferred directly to the non-crystalline state.
Using the geometry of molecule 3 as a starting point, we optimized its geometry in an unprotonated form, and calculated the electrostatic potential (ESP) charges with the MULTIWFN [23] package for this unprotonated molecule, to see which of the four nitrogen atoms is most willing to attach protons (Table 9). In Figure 22, a graphical representation of the electrostatic potential is presented. Table 9. Energy of the optimized geometry of molecule 3 in the following different mono-protonated states: at N25, N29 and N11 nitrogen atoms, and CHELPG ESP fitted atomic charges for the optimized molecule 3 in unprotonated state.

Atoms
Energy ( nitrogen atoms is most willing to attach protons (Table 9). In Figure 22, a graphic representation of the electrostatic potential is presented.  Unfortunately, these results do not correlate well with the structures determine experimentally, where the first protonation candidate is the N11 atom, followed by th N25 atom. Moreover, protonation of the N29 atom was not observed in any of th experimentally determined structures, which, according to ESP calculations, is best suite for the attachment of a proton.
We also calculated the free energies of the molecules in the mono-protonated stat for all the nitrogen atoms, by the PCM method, with water as a solvent. The results a presented in Table 9. The lowest energy occurs for N25-H and N11-H (with a differenc of only 1.5 kcal/mol between the two). The energy for N29-H is higher (10 kcal/mol), an it is much higher for N14-H (~30 kcal/mol). It seems that for the investigated structure the calculated energies are much better suited for predicting protonation sequences tha the ESP potentials.

ADMET Analysis
The bioavailability radars for all the studied compounds are very similar (Figure 2  for compound 1). The pink-colored zone on the bioavailability radar (SwissADME presented the optimal range for each property, indicating the drug-likeness of a molecul All the compounds meet the rules of Lipinski [32], Ghose [33], Egan [34], Veber [35], an Muegge [36]. All the compounds were found to be highly absorbed in the gastrointestin tract, making them effective drugs ( Figure 24). An important element is that, having hig Unfortunately, these results do not correlate well with the structures determined experimentally, where the first protonation candidate is the N11 atom, followed by the N25 atom. Moreover, protonation of the N29 atom was not observed in any of the experimentally determined structures, which, according to ESP calculations, is best suited for the attachment of a proton.
We also calculated the free energies of the molecules in the mono-protonated state, for all the nitrogen atoms, by the PCM method, with water as a solvent. The results are presented in Table 9. The lowest energy occurs for N25-H and N11-H (with a difference of only 1.5 kcal/mol between the two). The energy for N29-H is higher (10 kcal/mol), and it is much higher for N14-H (~30 kcal/mol). It seems that for the investigated structures, the calculated energies are much better suited for predicting protonation sequences than the ESP potentials.

ADMET Analysis
The bioavailability radars for all the studied compounds are very similar (Figure 23 for  compound 1). The pink-colored zone on the bioavailability radar (SwissADME) presented the optimal range for each property, indicating the drug-likeness of a molecule. All the compounds meet the rules of Lipinski [32], Ghose [33], Egan [34], Veber [35], and Muegge [36]. All the compounds were found to be highly absorbed in the gastrointestinal tract, making them effective drugs ( Figure 24). An important element is that, having high activity towards the H3 receptor, it crosses the blood-brain barrier. All of the studied compounds are not inhibitors of the CYP3A4 isoenzyme, which is largely responsible for the metabolism and elimination of most clinically used drugs [37]. Negative logKp values indicate that the compounds are not available through the skin.

Conclusions
Seven new crystal structures (histamine H3 antagonists) were determined, including two polymorphs and one hydrate of the same compound. Interestingly, polymorphs 6 and 7 were both obtained from the same batch of crystallization. The two main factors differentiating the conformation of the studied molecules are as follows: (i) chain conformation, defined by the torsion angle N11-C3, and (ii) conformation at N14, defined by the angle between the N14-C21 bond and the C13-N14-C15 plane. The variability in the latter parameter is probably due to the intermolecular interactions occurring in the crystal structures, which is confirmed by QM calculations. The ADME analysis confirmed that the tested compounds are good drug candidates. For thiazole derivatives (compounds 2 and 4), which show higher activity (as non-imidazole antagonists of histamine H3) than their oxazole analogues, the relative position of the aromatic bicyclic system and the piperazine ring is slightly different to that of oxazoles, which may affect their biological activity.
Supplementary Materials: The following are available online at www.mdpi.com/xxx/s1, Cif files containing all crystallographic data.

Conclusions
Seven new crystal structures (histamine H3 antagonists) were determined, including two polymorphs and one hydrate of the same compound. Interestingly, polymorphs 6 and 7 were both obtained from the same batch of crystallization. The two main factors differentiating the conformation of the studied molecules are as follows: (i) chain conformation, defined by the torsion angle N11-C3, and (ii) conformation at N14, defined by the angle between the N14-C21 bond and the C13-N14-C15 plane. The variability in the latter parameter is probably due to the intermolecular interactions occurring in the crystal structures, which is confirmed by QM calculations. The ADME analysis confirmed that the tested compounds are good drug candidates. For thiazole derivatives (compounds 2 and 4), which show higher activity (as non-imidazole antagonists of histamine H3) than their oxazole analogues, the relative position of the aromatic bicyclic system and the piperazine ring is slightly different to that of oxazoles, which may affect their biological activity.
Supplementary Materials: The following are available online at www.mdpi.com/xxx/s1, Cif files containing all crystallographic data. Servis ProTox II classified the compounds 1, 3, and 5 into toxicity class 4 (harmful if swallowed), with a predicted LD50 of 1000 mg/kg. Compounds containing a sulfur atom (2 and 4) are in toxicity class 3 (toxic if swallowed), with a predicted LD50 of 300 mg/kg.

Conclusions
Seven new crystal structures (histamine H3 antagonists) were determined, including two polymorphs and one hydrate of the same compound. Interestingly, polymorphs 6 and 7 were both obtained from the same batch of crystallization. The two main factors differentiating the conformation of the studied molecules are as follows: (i) chain conformation, defined by the torsion angle N11-C3, and (ii) conformation at N14, defined by the angle between the N14-C21 bond and the C13-N14-C15 plane. The variability in the latter parameter is probably due to the intermolecular interactions occurring in the crystal structures, which is confirmed by QM calculations. The ADME analysis confirmed that the tested compounds are good drug candidates. For thiazole derivatives (compounds 2 and 4), which show higher activity (as non-imidazole antagonists of histamine H3) than their oxazole analogues, the relative position of the aromatic bicyclic system and the piperazine ring is slightly different to that of oxazoles, which may affect their biological activity.