Water plays a critical role in many steps of the pharmaceutical development as this small molecule has the ability to interact with compounds in numerous ways and may therefore significantly affect manufacturing processes and finally the quality of (pharmaceutical) products. The formation of a molecular compound (hydrate), where water becomes a part of the crystal lattice, is mostly accompanied with a significant change in the solid-state properties, and therefore this type of interaction must be seen as critical [1]. Hydrate formation itself is a widespread phenomenon and is known to occur for at least one third of drug molecules [2,3], and this trend is increasing significantly for new drug substances. Nevertheless, we are still not able to predict hydrates, their stability and dehydration mechanisms based on the molecular diagram only.

This talk will emphasise on the efforts that are sometimes required to identify solid forms of complex hydrate forming systems. Examples from our research will be used to illustrate how the combination of a variety of experimental techniques, covering temperature- and moisture-dependent stability, and computational modelling allows to generate sufficient kinetic, thermodynamic and structural information to understand hydrate formation and its impacts on relevant physicochemical properties.

The solid form landscape of brucine sulphate was elucidated, resulting in three hydrate forms and amorphous brucine sulphate. HyA was produced from water and the other two by dehydration starting from HyA. Removal of the essential water molecules stabilising the hydrate structures causes a collapse to the amorphous state [4]. Eight hydrate forms were verified for the related compound, strychnine sulphate. Three of the hydrates were found to be stable at ambient conditions. The other five hydrates are only observable at low(est) relative humidity (RH) levels at room temperature. Some of the hydrates can only exist within a very narrow RH range and are therefore regarded as intermediate phases. The specific moisture and temperature conditions of none of the applied drying conditions yielded a crystalline water-free form, highlighting the essential role of water molecules for the formation and stability of crystalline strychnine sulphate [5].

Despite their structural similarity, marked differences in the formation of solid forms are seen for brucine and strychnine. One anhydrous form and 1,4-dioxane solvates were crystallized for strychnine, whereas two non-solvated polymorphs, four hydrates, an isostructural dehydrate, twelve solvates and two hetero-solvates are known to exist for brucine [6-8]. One of the brucine hydrates shows a non-stoichiometric (de)hydration behaviour, one collapses to an amorphous phase, and the third one to the polymorph which is stable at room temperature. Interestingly, each of the three hydrates may become the most stable form depending on temperature and water activity.

To conclude, this study demonstrates the importance of applying complimentary analytical techniques and appropriate approaches for understanding the stability ranges and transition behaviour between the solid forms of compounds with multiple hydrates.

[1] Khankari, R. K. & Grant, D. J. W. (1995). Thermochim. Acta, 248, 61.

[2] Stahly, G. P. (2007). Cryst. Growth Des, 7, 1007.

[3] Reutzel-Edens, S. M., Braun D. E. & Newman A. W. (2019). Polymorphism in the Pharmaceutical Industry: Solid Form and Drug Development, edited by R. Hilfiker & M. Von Raumer: Wiley-VCH, pp. 159-188.

[4] Braun, D. E. (2020). CrystEngComm, 22, 7204.

[5] Braun, D. E., Gelbrich, T., Kahlenberg, V. & Griesser, U. J. (2020). Cryst. Growth Des., 20, 6069.

[6] Braun, D. E. and Griesser, U. J. (2016). Cryst. Growth Des., 16, 6405.

[7] Braun, D. E. and Griesser, U. J. (2016). Cryst. Growth Des., 16, 6111.

[8] Watabe, T., Kobayashi, K., Hisaki, I., Tohnai, N. & Miyata, M.Bull. (2007). Chem. Soc. Jpn., 80, 464.

We have recently studied a diarylamine compound, tolfenamic acid (TFA), and examined its solution chemistry, crystallization kinetics, and molecular interactions. The polymorphic system typically crystalizes as From I or Form II, or both concurrently, with Form I being the most stable at room temperature. Both polymorphs are composed of hydrogen-bonded, carboxyl homodimers as the supramolecular synthon in their respective crystal structures. One interesting kinetic phenomenon that we experimentally discovered was an intermediate or transitional retraction of the mass composition of Form I in crystallized samples over the course of concomitant crystallization. The composition retraction bears two characteristic attributes, the retraction depth and the onset fraction. The former quantifies the maximal extent to which the Form I composition retracts prior to elevation, whereas the later attribute characterizes the initially measured Form I composition. Conversely, during solvent-mediated phase transformation, the mass composition of Form I monotonically increases and only Form II nucleates initially. We further learned through population balance simulations that this characteristically kinetic phenomenon is a sufficient condition indictor of concomitant crystallization of polymorphic systems. Interestingly, when experimental observation is made at a later time after the kinetic retraction, it seems unlikely to kinetically differentiate the two crystallization pathways.

In recent years, considerable investment has been made towards advancing pharmaceutical development and manufacturing through Digital Design approaches.¹ Industrial scientists are moving away from time and resource intensive screening techniques to more rapid in silico methods to inform key decisions throughout the drug manufacturing process.

The links between solid form and structural properties are well developed,² but our understanding of the relationship between particle and surface properties and downstream manufacturability of an Active Pharmaceutical Ingredient (API) are considerably less established. By providing new methods for visualising and describing these key attributes, we can gain a deeper insight into properties that contribute to the way particles flow or how they form tablets under compression.

Since describing these approaches and their application to the drug lamotrigine,³ we have continued to develop and refine the way that we can describe a particle and its properties. This presentation will discuss those advances and the challenges that lie ahead.

References

1. www.addopt.org

2. P.T.A Galek et al., CrystEngComm 2012, 14, 2391–2403

3. M.J. Bryant et al., Cryst. Growth Des. 2019, 19, 9, 5258–5266

Pharmaceutical industries have witnessed an increasing trend towards poor aqueous solubility and according to a report 75% of the marketed drugs belong to BCS class II or IV. Efforts to improve aqueous solubility by modifying the chemical structures are carried out during lead optimization in early drug discovery stage while trying to maintain desired potency and ADME properties. However, experimental aqueous solubility assays available during lead optimization are prone to overestimate solubility to a variable extent. This overprediction of aqueous solubility can result in overly optimistic view of developability with negative implications for compounds differentiation and candidate selection for development. On the other hand, failure to improve aqueous solubility could lead to inadequate evaluation of safety and efficacy profile of candidates and resource intensive formulation approaches. With the advancement of computations as well as due to immense pressure to shorten development timelines, in-silico approaches to predict aqueous thermodynamic solubility are of greater importance. In this presentation a physics-based ensemble modeling approach consisting of high-fidelity cloud-based crystal structure prediction (CSP) methodology optimized for computational cost and a novel free energy perturbation (FEP+) workflow is discussed to predict aqueous thermodynamic crystalline solubility of chemically structurally related compounds during lead optimization stage using just the 2-D structure as an input.

The various aspects of drug design or catalysis is compartmentalized within defined research fields, i.e. bioactivity testing versus pure synthetic chemistry; homogeneous versus heterogeneous catalysis. These are independent and often non-interactive specialities which have developed along parallel pathways with a common objective. The world economic drive towards the 4^th industrial revolution captures the idea of the confluence of new technologies and their cumulative impact on our world. Hence the ability to merge, bridge and remove boundaries will result in the establishment of interoperable research.

Drug design, particularly the development of target specific radiopharmaceuticals which involves the selective receptor binding of a radioactive organometallic complex to a possible disease site involves multiple facets. Simple manipulation of the ligand system bound to the metal centre can significantly alter parameters such as steric and electronic character, chirality, stability, biological and hydro/lipophilicity properties. Our organometallic research utilising the group 7 transition metal triad of manganese, technetium and rhenium for nuclear medical imaging and therapeutic agents, includes the interactions with proteins using protein crystallography. This provides valuable structural information in a similar vein to fragment based drug discovery (FBDD). The domain of chemical versus macromolecular crystallography has resulted in multiple discipline variations, such as incompatible software, data formatting and terminology. A key challenge which hinders research advancement is the lack of interoperability between chemical and biological crystallographic data.

This perspective will highlight the opportunities of harvesting both small molecule and macromolecular structural data, the joint usage of the CSD and PDB databases, as well as the advantages of software which can convert organometallic small molecule structural data for use in protein refinement software. This multidiscipline approach to radiopharmaceutical development will include kinetic reactivity studies highlighting how subtle structural changes can significantly affect chemical reactivity and hence the protein coordination in macromolecular structures. Trends similarly witnessed in catalysis research.

Multiple scattering in 3D electron diffraction (3D ED) experiments is responsible for deviations of diffracted intensities from intensities expected from kinematical diffraction theory [1]. Though this is usually considered a disturbing factor in routine structure determinations, these deviations also contain valuable information on the absolute structure [2]. Analysing 3D ED measurements from different laboratories around the world, we demonstrate that the absolute structure of single submicrometric crystals can be reliably and easily determined in a routine way if dynamical diffraction effects are incorporated in the refinement of the structure model.

We investigated and reinvestigated data sets of non-centrosymmetric samples recorded with beam-precession (precession-assisted 3D ED) and with continuous-rotation 3D ED (IEDT, MicroED, cRED) to determine the absolute structure, which directly determines the absolute configuration of chiral molecules in the unit cell. Dynamical effects are very sensitive to the absolute structure due to the interference of multiple beams contributing to each reflection [3]. In comparison to X-ray diffraction-based methods, the requirements for a successful determination of the absolute structure are strongly reduced. We demonstrate that with a completeness as low as 25% (Figure 1), a limited resolution d_min > 1 Šand only a preliminary structure model the correct chirality can still be identified. The low requirements also allow significantly reducing the number of refinement parameters so that the computationally demanding calculations applying dynamical diffraction theory are only a matter of minutes even for unit cells with a volume of several thousand ų. The determination is based on a simple comparison of residual factors (R_obs and wR_all) of the refined, inversion-related models (Figure 1). With this approach, the routine determination of the chirality of molecules in submicrometric crystals is ready to be implemented in any laboratory with access to 3D electron diffraction measurements. Considering ongoing developments, improvements and increasing level of automatization of data acquisition and analysis [1], we believe that especially the pharmaceutical industry will strongly benefit from the presented approach.

[1] Gemmi, M., Mugnaioli, E., Gorelik, T., Kolb, U., et al. (2019). ACS Cent. Sci. 5, 1315−1329.

[2] Brázda, P., Palatinus, L., Babor, M. (2019). Science. 364, 667−669

[3] Spence, J.C.H., Zuo, J.M., O'Keeffe, M., Marthinsen, K., Hoier, R. (1994). Acta. Cryst. A50, 647−650

Support by the Czech Science Foundation (project number 21-05926X), and by Operational Programme Research, Development and Education financed by European Structural and Investment Funds and the Czech Ministry of Education, Youth and Sports (Project No. SOLID21 CZ.02.1.01/0.0/0.0/16_019/0000760) is highly appreciated.