Introduction

Protein crystallization serves as a crucial step in determining 3D protein structures, which are necessary for a better understanding of their structure and function. Protein crystallization primarily aims to produce crystals with sufficient diffraction properties to permit protein structure determination.

Microgravity provides a unique environment for the growth of high-quality protein crystals compared to classical Earth-based approaches. In some cases protein crystals grown under microgravity conditions in space exhibit improved order, fewer defects, and diffract better than Earth-grown crystals1,2,3. In a space microgravity environment, factors like buoyancy-driven convection, sedimentation, and hydrostatic pressure are greatly reduced during crystallization, thereby minimizing protein mass transport constraints and fostering optimal crystal packing2,4,5,6,7.

Although protein crystallization in microgravity was first performed on spacecraft decades ago, technological advancements and an improved understanding of the underlying mechanisms have sparked renewed interest in protein crystallization in microgravity. Specialized equipment and methods have been developed to optimize the crystallization process in space1,2,8,9. However, despite substantial advances in protein crystallization techniques in space6,9, the preparation, packaging, and transportation of protein samples to the space environment introduce challenges, as there are potential sample degradation, contamination, or mishandling during these processes which impact the reliability and quality of the crystallization experiments. An additional challenge is the limited ability to handle samples on the space station and inspect crystal growth during crystallization.

Currently, 3D bioprinting technology in space microgravity experiments has drawn attention. Bioassembler is a novel bioprinting system that has gained interest in tissue engineering and regenerative medicine in space by producing exact and accurate 3D tissue structures10. This system is designed to create complex 3D structures by using magnetic forces to assemble cells and biomaterials into functional tissues, making it a promising technique for controlling the bioassembly process in space.

In this study, we introduced the novel approach of utilizing the bioassembler technique, specifically the “Organ.Aut”11, for successful crystallization of proteins in space microgravity environment, taking hen egg white lysozyme (HEWL) as a model object. We analyzed the quality of the acquired crystals, solved the protein structures, and conducted a comparative analysis with crystals grown on Earth. We report the ultrahigh-resolution structures of HEWL grown on Earth and in space microgravity with a resolution of 0.80 Å and 1.09 Å, respectively. We further subject the structures to a comparative analysis with previously published models and finally conclude on the utility of the bioassembler “Organ.Aut” for growing high-quality crystals in space. The proposed device has the advantage of allowing easy mixing of protein and precipitant solutions directly on the space station, as well as the real-time observation of crystal growth with specially dedicated cameras.

Results and Discussion

Experimental design

The magnetic bioassembler “Organ.Aut” has been specifically engineered and certified for life science research in space. Detailed technical specifications and operational procedures are outlined in recent successful use examples10,12. In short, the system comprises two primary components: rechargeable cuvettes and the magnetic bioassembler chamber. The cuvette (Fig. 1A) features one main volume and an additional two supporting volumes. The supporting volumes can be injected into the main one by specially designed pistons when activated by pressing the buttons. Typically, these volumes are filled with solutions on Earth and mixed in space according to the requirements of a specific experiment. The solution in the main volume requires the presence of Gd3 + -HPDO3A as a paramagnetic component to facilitate self-assembly of a diamagnetic object at the center of the cuvette. The magnetic bioassembler chamber produces an inhomogeneous magnetic field around each cuvette, which is characterized by a high gradient and a central magnetic pit. The bioassembler can accommodate up to six independent cuvettes delivered to the International Space Station (ISS) (Fig. 1C).

Fig. 1: Schematic illustration of the space microgravity experiment.
figure 1

A Cuvette preparation procedure. The cuvette is consequently filled with agarose plug, protein solution, and precipitant. B Six cuvettes are loaded in the transportation box and delivered to the ISS with the Soyuz MS-14 spacecraft. The box also contains a thermologger that continuously records the temperature. C Key stages of the experiment performed aboard the ISS. The cuvettes are loaded into the bioassembler “Organ.Aut” to sustain the crystallization process for 10 days. Three video cameras are installed on the bioassembler to capture the crystallization process. D Transportation of the cuvettes with the crystals in the transportation box back to Earth with the Soyuz MS-14 spacecraft. The data from the thermologger are shown in blue. Figure adapted from previous work10.

Full size image

A transportation box is designed to accommodate the six cuvettes (Fig. 1B, D) to and from the ISS. A thermologger is installed at the bottom of the transportation box to record the temperature around the cuvettes from the moment they are loaded into the box.

The bioassembler features autonomic video cameras installed to capture the processes that take place within the cuvettes through a specially designed transparent observation window (Fig. 1C).

In our study, we repurposed the bioassembler “Organ.Aut” system for protein counterdiffusion crystallization in space13. To achieve this, we deactivated one of the supporting volumes along with both buttons while utilizing the main and second volumes to establish a counterdiffusion crystallization setup (see left-hand part of Fig. 1A). As we will discuss in further detail in the section entitled “Gd3+-HPDO3A interaction with the protein,” the Gd3+-HPDO3A and the magnetic field did not affect this setup, although they form an intrinsic part of the original “Organ.Aut” system.

We used HEWL as a protein for crystallization as it is an ideal model for structural studies, and its crystallization process has been studied in great detail14,15,16,17,18,19,20,21.

The experimental design of our study encompassed a series of sequential steps. Initially, the cuvettes, comprising the precipitants and the protein solution were prepared on Earth (Fig. 1A). To prevent undesired interaction during transportation to the ISS, we implemented an agarose plug (in orange color) to separate the precipitant (in green color) from the protein-containing solution in the main volume (blue color) within sealed cuvettes (Fig. 1A). The calibration of the agarose plug, focusing on its size and composition, was meticulously conducted to ensure that crystallization would exclusively commence upon the cuvettes reaching the ISS. These crucial parameters were determined through a preliminary experiment performed on Earth and showed the appearance of initial protein crystals approximately three days after the cuvette preparation. In addition, to stabilize the position of protein crystals during their return journey to Earth, the protein solution contained 0.5% agarose.

On the 27th of August 2019, the cuvettes were transported to the Russian Orbital Segment (ROS) of the ISS utilizing the Soyuz MS-14 spacecraft. Subsequently, they were integrated into the “Organ.Aut”, which had been delivered during the preceding expedition10. Thus, the bioassembler had been present in space for eight months before the cuvettes integration for our experiment.

Finally, the cuvettes containing the protein crystals were returned to Earth by the Soyuz MS-14 spacecraft on September 6th 2019, after a 10-day stay at the ISS. Examination of the camera records from the ISS revealed the presence of protein crystals (Fig. 2A). The thermologger indicated a stable temperature of 18-19 °C throughout the experiment, with abrupt spikes reaching around 25 °C during pre-flight preparations and transit to and from the ISS.

Fig. 2: Typical crystal images obtained with the bioassembler “Organ. Aut”.
figure 2

A Crystals in a cuvette with a 60 mg/ml protein concentration. The image was obtained by the camera installed in the bioassembler “Organ.Aut” on the ISS. B, C Typical crystals selected respectively from Earth and space for crystallographic data collection. D–F Crystals obtained in space in the cuvettes with 10, 30, and 60 mg/ml protein concentrations. B–F were obtained on Earth with the Nikon microscope.

Full size image

A parallel control experiment was conducted on Earth utilizing identical bioassembler “Organ.Aut”, cuvettes, and protein batch. Records from Earth control cameras showed the crystal growth process (Supplementary Movie 1). Crystals from cuvettes were subjected to detailed morphological and diffraction analyses.

Characterization of crystals

The obtained crystals were visually inspected with a microscope and subjected to crystallographic analysis at the European Synchrotron (ESRF). Figure 2 shows typical crystals obtained with the bioassembler. The six cuvettes were arranged into three pairs for experimental replication, with each pair assigned a specific protein concentration: 10, 30, and 60 mg/ml. All cuvettes contained protein crystals, and the quantity of the crystals showed no visible dependence on protein concentration. However, crystal size significantly increases with concentration: the size of most crystals in the cuvettes ranges from 300–400 μm, 700–1000 μm, and 800–1500 μm (Fig. 2A–C). The crystals exhibited a tetragonal shape characteristic of HEWL. These observations were consistently reproduced within each pair for space and Earth crystals. Despite their similar size, a notable distinction was observed. Larger crystals from space displayed visible cracks, while Earth crystals maintained well-defined shapes. During crystallization at ISS, crystals could undergo microgravity with quasi-static accelerations of up to 10−6 g and harmonic vibrations ranging from 5 to 20 Hz10. Upon landing, however, they may be subjected to abrupt shocks with pulses reaching 67 g, lasting from 1 ms to 2 ms10. This sudden impact, as well as the temperature fluctuations recorded, have the potential to induce stress, resulting in crystal damage and the formation of cracks, as seen in Fig. 2A. It is essential to clarify that the high g-jump referred to in the earlier study10 represents extreme cases. It does not reflect the actual acceleration experienced by our crystals during this expedition, as it was not directly measured during the experiment and is expected to fall within a milder range.

Well-shaped crystals 100–500 µm in size (Fig. 2B, C) from samples with 30 mg/ml protein concentration were selected for the data collection at cryogenic temperature. Ideally, crystals obtained at microgravity should be examined at room temperature to accurately assess the impact of microgravity on crystal quality1. However, in our case the space-grown crystals exhibited cracking due to temperature fluctuations. These defects were clearly observed under the microscope, indicating a deterioration in crystal quality. A detailed comparison of the quality between space-grown and Earth-grown crystals was deemed unfeasible. Meanwhile, the crystallization conditions employed were notably distinctive: crystals were obtained in an agarose environment, supplemented with Gd³⁺-HPDO3A, subjected to a magnetic field, and grown under microgravity. Given these unique parameters, we opted to collect the highest quality X-ray diffraction data to conduct a precise investigation of the protein structure. In addition, comparing protein structures from microgravity crystals with ground-grown controls holds significance, as structural differences within these environmental conditions have been detected previously22 and may be expected in our case as well. To achieve this, data collection was performed at cryogenic temperatures, which is a standard practice as it enhances data quality. Table 1 provides a summary of the crystallographic data collection statistics for Earth and space crystals.

Table 1 Summary of the data processing statistics
Full size table

Typically for HEWL, all the crystals exhibit the P43212 space group with similar unit cell constants regardless of their growth condition. However, the unit cell volumes of the space-grown crystals consistently exceed those of the earth-grown crystals. All the crystals show a similar range of mosaicity, with an average of 0.07° for the space crystals and 0.1° for the Earth crystals, suggesting comparable crystal quality. It is important to note that the crystals were examined under cryocooled conditions, which are known to significantly increase mosaicity, potentially obscuring any improvements achieved through microgravity1. The diffraction quality of Earth-grown crystals, however, appeared slightly better, with best resolutions of approximately 0.88 Å compared to 1.09 Å for the space crystals. This observation aligns with the slightly higher Wilson B-factors of the space crystals (~18 Å2) compared to the control crystals ( ~ 13Å2), suggesting a potential influence from the landing stress mentioned earlier, which may have affected their quality. Meanwhile, the data collection wavelength for crystals from both groups was arbitrarily selected differently, which can partially explain the difference in the diffraction statistics.

Comparative analysis of crystal structures

Further structure solutions were carried out using the best space and Earth datasets, with resolutions of 1.09 Å and 0.80 Å, respectively (Table 2). In both cases, the HEWL structure comprises all 129 amino acid residues expected from the primary sequence. As expected for HEWL and consistently observed for both space and Earth data, the secondary structure consists of seven α-helices (residues 5–14, 25–36, 80–84, 89–100, 104–107, 109–114, and 121–123) and three antiparallel β-sheets (residues 43–45, 51-53 and 58–59) (Fig. 3A). Additionally, the protein folding is stabilized by four intermolecular disulfide bonds (residues Cys6–Cys127, Cys30–Cys115, Cys64–Cys80, and Cys76–Cys94).

Fig. 3: HEWL crystal structures from Earth and space.
figure 3

A Overall structural alignment of the HEWL with Cα-Cα RMSD of 0.071 Å. The structures are comparable with a minute difference from residues 100 to 103. B, C Crystal contact interaction of Gd3+-HPDO3A with Earth and space HEWL, respectively. In both models, the Gd3+-HPDO3A molecules occupy the interstitial space between three protein molecules, stabilizing the crystal contacts. The Gd3+-HPDO3A complex exhibits hydrophobic interaction with proximity aromatic side chains of Trp 62, Trp 63, and Trp123. Leu 75 within 4 Å is also involved in the nonpolar interaction stabilizing the protein contacts.

Full size image
Table 2 Crystallographic data and refinement statistics for the final models
Full size table

The overall RMSD for Cα atoms of the space and Earth structures is 0.069 Å, indicating that the overall models were very similar. The number of modeled atoms is also very similar for the two models, with more of them built in the Earth model (see Table 2). These additional atoms mostly represent alternative conformations of specific amino acids. Thirty-one amino acids take on multiple conformations in the Earth model, whereas in the space model, this number is 24. The most pronounced backbone displacement is ~1 Å in the 100–103 amino acid region where HEWL interacts with HPDO3A (Fig. 3). This interaction is described in more detail below.

Gd3+-HPDO3A interaction with the protein

As it was previously designed10, the standard bioassembler “Organ.Aut” setup implies a nonhomogeneous static magnetic field (1.21 T), overlaid on the sample chamber and a Gd3+-HPDO3A containing solution sample. This measure was originally designed to facilitate magnetic levitational bioassembly of 3D tissue in space. In our experiment, the magnetic field did not create a significant force that could influence the crystallization process—the magnetic field energy per molecule is estimated to be 1.1×10-24J, which is > 5000 times smaller than its kinetic energy (6 × 10−21J) at room temperature. However, Gd3+-HPDO3A can be readily found in the HEWL crystal structure for both Earth and space crystals (Fig. 3). Two Gd3+-HPDO3A molecules are found in the asymmetric unit. The HPDO3A macrocyclic molecule forms four nitrogen and four oxygen coordination contacts with Gd3+. The unpaired electrons in the gadolinium ion interacted with the nearby water molecules to satisfy the 9th coordination valency of the ion (Figs. 3B and 3C). The resultant geometry is in close agreement with the crystal structure of the pure Gd3+-HPDO3A23. For both Gd3+-HPDO3A molecules, the anomalous density clearly shows the presence of two alternative Gd3+ positions with partial occupancies of 0.73/0.27 and 0.68/0.32 for the Earth and space structures, respectively. The Fo-Fc-maps in this region also show the presence of alternative minor conformation of the chelate molecules. However, the minor conformation’s occupancy of HPDO3A is insufficient for unambiguous modeling; consequently, they were not incorporated into the model.

The Gd3+-HPDO3A molecules occupy space between three protein molecules, evidently stabilizing crystal contacts. In both Earth and space HEWL structures (Fig. 3), the Gd3+-HPDO3A complex establishes water-mediated contacts with positively charged side-chains of Arg73 and Lys33’, as well as the carboxyl side-chain of Asp119” and the amine backbone of Val120” and Gln121”(‘ and “ denote symmetry-related HEWL molecules). These water-mediated interactions play a vital role in indirectly connecting the residues to the Gd3+-HPDO3A complex. Interestingly, while the Asp101 carbonyl backbone of the Earth HEWL interacts with the carboxyl groups of the Gd3+-HPDO3A complex via a water molecule, the space HEWL exhibits a hydrophilic interaction using the alternative conformations of Asp101 (Fig. 3). This residue (Asp101) in the space structure, directly and indirectly (via water), interacts with the chelate using its backbone and side chain together. Furthermore, the Gd3+-HPDO3A complex in both Earth and space HEWL structures forms hydrophobic contacts with Trp62, Trp63, Leu75, and Trp123’ to adjacent protein molecules, using its CH–π. These CH–π interactions reinforce the protein-protein interaction. Notably, the Earth HEWL structure exhibits an additional direct polar interaction between the guanidine group of an alternative conformation of Arg5’ and the Gd3+-HPDO3A complex, further strengthening the interaction.

Comparison of Gd³⁺-HPDO3A complex in HEWL crystallization

Multiple structures of HEWL complexed with Gd³⁺-HPDO3A have been previously solved, highlighting the gadolinium complex’s utility in crystallography for molecular phasing enhancement24,25,26,27,28,29,30. The Gd³⁺-HPDO3A derivative has also been employed as a strategy to investigate radiation damage28. In our study, we compared our crystallographic data with published Gd³⁺-HPDO3A-HEWL structures, revealing a maximum Cα–Cα RMSD of 0.283 Å between our Earth structure and PDB ID: 5C6L. Notably, the Gd³⁺-HPDO3A complexes occupy the same region of the binding site and exhibit nearly identical conformations across all analyzed structures. The primary distinction lies in the conformation of the Arg73 side chain; in our space and Earth models, Arg73 displays two conformations, whereas only one conformation is observed in other models. It is important to highlight that the previous models exhibiting a single conformation were derived from data collected at room temperature.

Gd³⁺-HPDO3A, together with similar Gd³⁺-based compounds, has previously garnered interest for its role in obtaining high-phasing-power heavy-atom derivatives for SAD data collection31,32,33. Gd³⁺-HPDO3A can be readily incorporated into crystallization protocols through co-crystallization or soaking methods. Its presence in our experimental system allowed us to obtain high-quality crystal diffraction data with strong anomalous signals, facilitating the generation of high-quality electron density maps and enabling de novo phasing via the SAD method. While our findings and previous studies indicate that Gd³⁺-HPDO3A does not adversely impact crystal formation, it is crucial to recognize that the effects of Gd³⁺-HPDO3A may be protein-specific and can vary significantly among different proteins.

The crystallophore, a Tb-Xo4 complex developed from Gd³⁺-HPDO3A, has recently been recognized as a valuable technique in protein crystallography. It integrates protein nucleating effects, phasing properties, and luminescence, facilitating both protein crystallization and structure determination, while also simplifying crystal detection34,35,36. Incorporating Tb-Xo4 into pre-flight protein crystallization screens may enhance the efficiency of crystal formation and structure determination. However, further investigation is warranted to fully assess its potential benefits across a diverse range of proteins.

Bioassembler “Organ.Aut” for space protein crystallization

Our study describes, for the first time, the successful application of the innovative bioassembler “Organ.Aut” technique for crystallizing proteins both on Earth and in space. The results demonstrate that the crystals grown using this device exhibit identical morphology and ultrahigh-diffraction power, indicating the potential of the bioassembler “Organ.Aut” as an alternative method for producing high-quality protein crystals with minimal defects.

To obtain high-quality crystals during crystallization experiments, precise control over crystal nucleation and growth is crucial. The bioassembler “Organ.Aut,” showcased herein for protein crystallization, enables the simultaneous execution of six experiments, offering a valuable opportunity to obtain reproducible crystals exhibiting consistent properties. A notable feature of the bioassembler is its capability to mix protein and precipitant solutions directly on the ISS. This functionality facilitates precise control over the initiation of crystallization. Additionally, the integration of three GoPro Hero4 action cameras focused on the crystallization samples enables real-time recording of the crystallization process, allowing for meticulous inspection and analysis of the crystallization process.

In space exploration, sample systems are subject to temperature fluctuations and acceleration-induced g-jumps. To mitigate the potential impact of g-jumps on crystal integrity, we incorporated agarose gel into the crystallization chamber, following the approach outlined in previous study37 in the exceeding concentration. This method has provided sufficient protective environment for the crystals during landing, shielding them from typical g-jumps. Despite this precaution, our sample system still encountered significant temperature fluctuations, which likely contributed to the visible cracks observed in the protein crystals. It is plausible that the system reached a critical temperature for the tetragonal form of the protein, beyond which only the orthorhombic lysozyme can crystallize38. Although all tested crystals were tetragonal, we suspect that this temperature associated with solubility differences, may have facilitated the formation of cracks. To fully leverage the advantages of microgravity for crystallization experiments, the bioassembler “Organ.Aut” system should be equipped with temperature-controlled features that utilize thermostatic materials to maintain stable temperature, thereby protecting samples from external variations8,39,40.

In microgravity environments, such as space, crystal formation can be significantly enhanced due to the absence of sedimentation and convection4,5,7. This environment presents a promising opportunity for achieving high-quality crystals. However, challenges arise when transporting crystals from space to Earth (Fig. 1D). During re-entry, the crystals may be exposed to extreme overloads and temperatures, potentially causing them to fracture or shatter, leading to data with lower resolution. Therefore, careful post-crystallization handling and manipulation of crystals are necessary, particularly for those transported from space, as the extreme conditions experienced during their time in microgravity may have rendered them more fragile.

Methods

Protein crystallization

The bioassembler “Organ.Aut”11 as designed, consists of six special cuvettes embedded in magnetic chambers and three video cameras GoPro Hero4 action cameras (Woodman Labs Inc., USA) (Fig. 1C). The cuvettes house the samples while the magnet chambers generate a magnetic field around the cuvettes. This innovation is specifically designed for levitational bioassembly, facilitating the creation of 3D tissues in the unique microgravity environment of space. In this study, we adopted the “Organ.Aut” for protein crystallization. The procedure starts with the cuvette preparation, described as follows. A cuvette (Fig. 1A) was filled with a 2% agarose gel (low-melting agarose, Agarose LM, cat.# 1925.0025, Dia-m) in 0.1Μ acetate buffer (pH = 4.8). The agarose concentration and the plug size were predetermined on Earth. To examine crystal formation in microgravity, it was crucial to establish conditions where the precipitant would interact with the protein solution and initiate the crystallization process upon the cuvettes’ delivery to the ISS, rather than initiating it beforehand. Pre-flight preparations and the journey to the ISS take approximately 72 h. To ensure that crystallization commences after this 72 h duration, an agarose plug was strategically positioned between the precipitant and the protein solution. Multiple concentrations (0.5%, 1%, and 2%) and volumes (200 μL, 400 μL, and 600 μL) of agarose were evaluated. Within the 0.5% and 1% agarose solutions, the precipitant diffused so rapidly that even with the maximum plug size (600 µL) allowed by the cuvette design, the first crystals became apparent within the 72 h timeframe. However, when the 2% agarose solution was utilized, the 600 µL plug effectively delayed the crystal formation by 72–80 h, ensuring that crystals grow directly aboard the ISS.

Subsequently, 1 ml solution of paramagnetic protein solution containing HEWL, 250 mM Gadovist (Gd3+-2,2’,2”-[10-(2-hydroxypropyl)-1,4,7,10-tetraazacyclododecane-1,4,7-triyl]triacetate, Gd3+-HPDO3A, Bayer AG), 0.5% agarose in 0.1Μ acetate buffer was placed in the lower section (biofabrication chamber) of the cuvette (Fig. 1A). The presence of agarose in the biofabrication chamber is to protect the crystals during landing. The cuvettes were arranged into three pairs for experimental replication, with each pair assigned a specific protein concentration: 10, 30, and 60 mg/ml. Finally, an upper chamber of the cuvette (Fig. 1A) was filled with 1 ml of 0.1 М acetate buffer containing 13% w/v NaCl (Helicon) as the precipitation solution. In total, six cuvettes were prepared, and once loaded, they were transported to the ROS of the ISS and installed in the bioassembler “Organ.Aut” (already in space) for the crystallization process. The temperature during the crystallization process was monitored by the thermologger (DS1922L-F5, iButton). The thermologger was installed on the bottom side of the transportation box (Fig. 1B). Upon arrival at the ISS, the transportation box was positioned near the bioassembler “Organ.Aut”. Consequently, the thermologger tracked the room temperature where cuvettes were housed, starting from their insertion into the transportation box until their removal upon return to Earth. The temperature was 18–19 °C, with abrupt spikes reaching around 25 °C during pre-flight preparations and transit to and from the ISS (Fig. 1D). A parallel control experimental setup was run on Earth at room temperature (21 °C), where crystals appeared in 3 days and reached their final size within 10 days. The crystals were viewed with an inverted microscope Nikon Eclipse Ti-S, and the resulting images are presented in Fig. 2B–F.

Data collection and processing

The crystal-containing cuvette was carefully disassembled, as demonstrated in Supplementary Movie 2, and the agarose gel with embedded HEWL crystals was delicately transferred onto a glass slide. To prevent the agarose gel from drying out and to ensure crystal protection, the precipitant used for the crystallization was added. Under microscopic observation, the agarose gel surrounding the crystals was manually and gently removed using manipulators from the Micro-Tools Set (Hampton Research).

Crystals from one cuvette originating from Earth and one cuvette from space, both corresponding to the protein concentration of 30 mg/ml, were transferred into nylon CryoLoops (Hampton Research) and flash-frozen in liquid nitrogen without additional use of cryoprotectants.

The crystals’ diffraction data were collected at the ID23-1 beamlines of ESRF at 100 K. The data were obtained using a Dectris Eiger2 CdTe 16 M with a single wavelength of 0.6199 Å for the Earth crystals and 0.9677 Å for the space crystals. The datasets were integrated and scaled using XDS Suite41 (Table 1).

Structural solution

The highest-resolution data from Earth was utilized to solve the structure with the Phenix package tools of experimental phasing42 due to higher anomalous strength. The Gd3+ positions were determined using Autosol43 with a resolution cut recommendation by phenix.xtriage44. Following phase calculations, the initial generated map was improved by a density modification prior to phenix.autobuild45. In COOT46 and PHENIX42,44,47 suites, different rounds of manual model building and refinements were done to add missing residues, Gd3+-HPDO3A and make changes to fit the residues into the density maps. The Fo-Fo Isomorphous Difference Map utility of Phenix was used to compare the datasets, which showed similarities across all crystal data. Owing to the structural isomorphism, the phase from the Earth model was used to solve the structure for the space datasets utilizing the molecular replacement technique. During the refinement process, Friedel’s pairs were treated independently due to a large anomalous signal (Table 1). This resulted in R-free improvements over merged data of 4% and 5% for the Earth and space models, respectively.

The initial refinement of the Earth model reached a resolution of 0.88 Å. However, during a pairwise refinement process, valuable data were obtained beyond this resolution. As a result, the final structures for both Earth and space models were refined to resolutions of 0.80 Å and 1.09 Å, respectively. The resulting models were validated with phenix.molprobity48. The overall statistical output of the structures is presented in Table 2. The visualization figures were created with PyMol (version 2.5.0).