Human transthyretin (TTR), a thyroxine and retinol-binding protein transporter in the blood and cerebrospinal fluid, can become amyloidogenic with age or by single point mutations causing hereditary and early-onset forms of the disease. In vitro studies have identified tetramer dissociation followed by monomer misfolding as the rate-limiting step in TTR amyloidogenesis. X-ray and neutron diffraction studies have been fundamental in understanding the architecture of TTR and its propensity to become amyloidogenic. In particular, neutron studies have provided unique details on how point mutations can alter hydrogen bonding underlying loss of tetramer stability in pathogenic mutants [1].
Due to the weaker flux of neutron beams, as compared to X-rays, much larger crystals are required for neutron studies, limiting its application for protein targets that are difficult to crystallize. Thus, the growth of large crystals remains a constant technical challenge for ‘new’ targets. Counter diffusion (CD) crystallization is an alternative method to obtain protein crystals. In this approach, the protein is filled in a capillary and immersed by its open end in a highly saturated precipitant solution. Diffusion of the precipitant in the capillary creates a ‘supersaturation wave’ that screens a wide range of crystallization conditions within a single capillary. This culminates in the formation of a metastable region (typically at the furthest end of the capillary from the precipitant) where large and high quality crystals grow [2].
As part of a collaborative study involving ILL, EMBL, ESRF and the Instituto Andaluz de Ciencias de la Tierra (IACT) in Granada, De’Ath, Oliva and colleagues [3] used CD crystallization to study TTR, defining a pipeline to optimize crystal growth and obtain large crystals suitable for neutron studies. Feasibility experiments were first carried out in small diameter capillaries (0.2 mm) whereby the typical CD growth pattern was observed alongside with an unexpected growth behavior. This produced rod-likecrystals that extended alongside the capillary wall. Samples were characterized by X-ray diffraction on the MASSIF-1 beamline (ESRF), using a combination of mesh scans and full dataset collection. The data obtained confirmed previous observations that crystal size and quality increased at the top end of the capillary when the typical CD crystal growth was obtained (Figure 1). Thanks to the small size of the beam, the rod-like crystals were analyzed by acquiring several datasets at regular intervals along their entire length, showing their high overall homogeneity. When larger diameter capillaries (1 mm) were used, crystals grew to a sufficient volume for preliminary neutron diffraction tests. Data were acquired on the LADI-III beamline (ILL). The same crystal was recorded first in hydrogenated buffer and again after exchangeinto fully deuterated buffer, revealing a clear increase in the diffraction quality and thus proving that such crystals are suitable for neutron crystallography studies (Figure 2).
This work was possible thanks to the collaboration with Dr. José Gavira (IACT) and the use of different PSB facilities: the D-LAB (ILL) for protein production and labelling, the ESRF and EMBL for the data acquisition on MASSIF-1, the ILL for the data collection on LADI-III.
C. De’Ath (ESRF / ILL), M. F. Oliva (ILL), M. P. Blakeley (ILL), M. Moulin (ILL), E. P. Mitchell (ESRF), M. W. Bowler (EMBL), T. Forsyth (ILL)
[1] Yee AW, Aldeghi M, Blakeley MP, Ostermann A et al. (2019) Nat Commun., 10, 1–10
[2] Otálora F, Gavira JA, Ng JD, García-Ruiz JM (2009) Prog Biophys Mol Biol., 101, 26–37
[3] De’Ath C, Oliva MF, Moulin M, Blakeley MP et al. (2025) J Appl Crystallogr., 58, 107–118
Figure 1: Variation of crystal quality ov er the supersaturation gradient within a capillary. (a) Photograph showing the 0.2 mm capillary (TTR at 60 mg/mL) under cross polarized light. The dashed lines indicate the regions used to acquire the mesh scan (bottom end of the capillary show n on the left at the site of precipitant entry). (b) The mesh scans are plotted as total integrated signal over the defined area (X, Y) with the relative intensity signal scale bar on the right and the N scores calculated for each area indicated above. (c) Diffraction pattern corresponding to the mesh scans above. The red and green box regions show enlarged portions to emphasize
differences in the quality of diffraction spots (d spacing at t he corner of the diffraction images is 1. 25 Å).
Figure 2: Neutron diffraction patterns recorded at the L ADI-III diffractometer (ILL), using a crystal of ~ 0.7 mm3 obtained in the 1 mm capillary set-up using 30 mg/mL protein with 0. 15% (w/v) agarose gel. (a) The position of the crystal in the capillary is marked as a bo xed area in the schematic (B = bottom; T = top). (b) Test neutron diffraction pattern recorded from TTR crystal hydrated with natural-abundance H2O. (c) Neutron diffraction pattern recorded from the same sample foll owing stepwise exchange of H2O by D2O.