Data for ion and seed dependent fibril assembly of a spidroin core domain

This data article includes size exclusion chromatography data of soluble eADF4(C16), an engineered spider silk variant based on the core domain sequence of the natural dragline silk protein ADF4 of Araneus diadematus, in combination with light scattering; the protein is monomeric before assembly. The assembled mature fibrils were visualized by transmission electron microscopy (TEM) and atomic force microscopy (AFM). Sonicated fibrils were used as seeds to by-pass the nucleation lag phase in eADF4(C16) assembly. We also provide data on the sedimentation kinetics of spider silk in the presence of different NaCl concentrations revealing very slow protein aggregation in comparison to the fast assembly triggered by phosphate ions published previously [1]. Experiments in the Data article represent supporting material for our work published recently [1], which described the assembly mechanism of recombinant eADF4(C16) fibrils.


Value of the data
Published work [1] and this additional material provide insights into how to analyze the nucleated assembly mechanism of recombinant spider silk protein eADF4(C16). The assembly mechanism of the silk compares to that of many other fibril forming proteins such as human Aβ peptides [2], huntingtin peptides [3] or yeast prion Sup35-NM [4] which all of them possess cross-β fibril structures. Data in the article and in the related publication [1] provide evidences that potassium and phosphate ions specifically trigger both nucleus formation as well as fibril growth. Both potassium and phosphate are strong kosmotropic ions and they also play a crucial role in the assembly of natural spider silks [5] in the spinning duct, whereas less kosmotropic NaCl is present in the ampulla stabilizing the protein during the storage. Therefore, the provided data could trigger further studies with amyloidogenic proteins concerning the influence of kosmotropic salts on fibril formation. In contrast, NaCl shows only marginal effect on the assembly of recombinant eADF4 (C16). Nevertheless, kinetic data suggest that the self-assembly of the cross-β fibrils as described here and in [1] is not related to the assembly of natural spider silk fibers. Nucleation is accompanied by a long lag phase (hours) followed by fibril elongation possessing perpendicularly oriented β-sheets [6,7]. Natural silk assembly is very fast (in the millisecond regime) and leads to β-sheet alignment along the fiber axis [5,8].

Data
In the data we included size exclusion chromatography data of soluble eADF4(C16) in a combination with light scattering revealing that the protein is monomeric before assembly (Fig. 1). A defined structural state is important before starting any kinetic study, since assemblies may significantly influence the kinetics by accelerating the nucleation (Fig. 2). The assembled mature fibrils were visualized by transmission electron microscopy (TEM) (Fig. 3) and atomic force microscopy (AFM) (Fig. 4), revealing fibrils typically 10 nm in diameter and 1 mm in lengths. Sonication led to significantly shorter fibrils, as shown by AFM, increasing the number of the active fibril ends. Sonicated fibrils can be used as seeds to by-pass nucleation in eADF4(C16) assembly (Fig. 5). We further show sedimentation kinetics of spider silk in the presence of different NaCl concentrations revealing very slow protein aggregation (Fig. 6) in comparison to the fast assembly triggered by phosphate ions [1], which indicates that ion masking events are less important for the protein-protein interation during spider silk assembly.

Multi angle light scattering (MALS) measurements
All buffers and the eADF4(C16) solution were filtered using 0.02 mm filters. The MALS measurements were carried out using an Agilent 1100 Series HPLC system connected to a multiangle light scattering detector DAWN EOS (Wyatt, Germany). The system was additionally equipped with a Rheodyne valve and a 1 mL loop to apply the protein solution at a flow rate of 0.2 mL/min to the K5 flow cell of the MALS detector. The flow was stopped after two minutes, the inlet and the outlet tubing were sealed, and light scattering was monitored over time. Molecular masses from light scattering signals were calculated using the ASTRA 6 software (Wyatt, Germany) (Data in Fig. 1).

Transmission electron microscopy
A 5 mL droplet was placed on a Formvar grid and allowed to settle for 30 s before the excess solution was blotted off. Then, 2 Â 5 mL water were added and blotted off. 5 mL uranyl acetate solution were added to the grid and left for 30 s before being blotted off. Finally, a 5 mL solution of water was added and blotted off. Samples were air dried for 16 h prior to observation on a JEOL JEM-2100 transmission electron microscope (Data in Fig. 3).

Atomic force microscopy
AFM scanning of dry samples was performed on a Dimension™ 3100 equipped with a NanoScope s V controller (Veeco Instruments Inc., USA) operating in Tapping Mode ™ using Si 3 N 4 cantilevers (OMCL-AC160TS, Olympus, spring constant of 42 N/m, resonance frequency of 300 kHz, tip radius less than 7 nm). For imaging, tapping mode (ratio of setpoint amplitude to free amplitude $ 0.7-0.9) was applied. AFM scans were processed using NanoScope Analysis software Version 1.40r3 (Brucker, Santa Barbara, CA) (Data in Fig. 4).