Ion-specific control of the self-assembly dynamics of a nanostructured protein lattice

Abstract

Self-assembling proteins offer a potential means of creating nanostructures with complex structure and function. However, using self-assembly to create nanostructures with long-range order whose size is tunable is challenging, because the kinetics and thermodynamics of protein interactions depend sensitively on solution conditions. Here we systematically investigate the impact of varying solution conditions on the self-assembly of SbpA, a surface-layer protein from Lysinibacillus sphaericus that forms two-dimensional nanosheets. Using high-throughput light scattering measurements, we mapped out diagrams that reveal the relative yield of self-assembly of nanosheets over a wide range of concentrations of SbpA and Ca(2+). These diagrams revealed a localized region of optimum yield of nanosheets at intermediate Ca(2+) concentration. Replacement of Mg(2+) or Ba(2+) for Ca(2+) indicates that Ca(2+) acts both as a specific ion that is required to induce self-assembly and as a general divalent cation. In addition, we use competitive titration experiments to find that 5 Ca(2+) bind to SbpA with an affinity of 67.1 ± 0.3 μM. Finally, we show via modeling that nanosheet assembly occurs by growth from a negligibly small critical nucleus. We also chart the dynamics of nanosheet size over a variety of conditions. Our results demonstrate control of the dynamics and size of the self-assembly of a nanostructured lattice, the constituents of which are one of a class of building blocks able to form novel hybrid nanomaterials.

Keywords: Ca2+ binding; biomaterials; nanostructures; protein interactions; self-assembly dynamics.

Figure 1

Light scattering is a measure of SbpA nanosheet growth in solution. (A) Representative traces of A_340nmvs time from light-scattering assays containing SbpA (6.8 μM) in solution in the absence (dashed line) or presence of 10 mM Ca²⁺ (solid line), indicating that increases in light scattering are associated with nanosheet formation. The colored, solid circles represent time points at which samples were imaged by STEM. (B) STEM images of negatively stained nanosheets of SbpA (6.8 μM) formed in the presence of 10 mM Ca²⁺after 800 min. The white arrows point to stacks of square protein sheets. The scale bar represents 1 μm. (C) Histogram of the side lengths of sheets visualized by STEM at increasing reaction times. For each measurement, the total number of nanosheets measured (N) is indicated. The dashed curves represent Gaussian fits to each data set. (D) The side lengths of nanosheets as a function of time, indicating the growth of sheets with time. The average and standard deviation of the side length are determined from the Gaussian fits shown in C. The solid line is a fit of eq 6 to the data.

Figure 2

Scattering by growing nanosheets is sufficient to describe the sigmoidal behavior in light scattering. Plot of the attenuation coefficient measuring self-assembly of SbpA into nanosheets at a concentration of 6.8 μM with a Ca²⁺ concentration of 10 mM is shown. The calculated scattering values from the mean size and error of nanosheets of SbpA from Figure 1D are overlaid as closed circles with the standard error of the mean shown in black and the standard deviation shown in gray.

Figure 3

The dynamics and yield of nanosheets of SbpA are dependent on both the protein and Ca²⁺ concentration. (A) Normalized A_340nm (A_norm) as a function of Ca²⁺ and SbpA concentration at 100 (left panel), 400 (middle panel), and 780 min (right panel) after starting the reaction. At 780 min, the data show a localized region of nanosheet assembly. The color scale indicates the value of A_norm (calculated as detailed in Methods) and is identical for each panel. (B) Representative STEM images of SbpA reactions at three different Ca²⁺ concentrations: 0.25 mM (circle), 5 mM (square), and 150 mM (triangle). The scale bar represents 2 μm for the left and middle panel and 4 μm for the right panel.

Figure 4

Formation of SbpA nanosheets in solution requires a minimum amount of free protein monomer. The end point A_norm plotted as a function of the SbpA concentration at 50 mM Ca²⁺ (black circles). The error bars represent the standard deviation of at least 3 measurements. The critical concentration of SbpA was determined by fitting a line through the data, and determining the intercept with the x-axis.

Figure 5

Formation of SbpA nanosheets in solution is not linearly dependent on the divalent ion concentration. (A) End point absorbance measurement from light scattering assays as a function of the divalent cation concentration for Ca²⁺ only (blue circles), 5 mM Ca²⁺ + Mg²⁺ (black circles), or 5 mM Ca²⁺ + Ba²⁺ (red circles). (B) End point absorbance measurements from light scattering experiments in which the total divalent ion concentration was kept at 50 mM total, while varying the relative concentrations of Ca²⁺ and Mg²⁺. A linear fit of the data gives a slope of −10^–3 (black line). The x-axis is linearly scaled before the break, and is on a logarithmic scale after the axis break. Error bars represent standard deviation from 2 to 3 measurements.

Figure 6

Tb³⁺ luminescence can measure binding of SbpA to Ca²⁺. (A) A solution of SbpA (6.8 μM) with 10 mM Tb³⁺ was excited at 280 nm and the emission of the sample was observed from 450 to 700 nm. Four peaks are observed at 490, 546, 585, and 620 nm, indicating that Tb³⁺ binds to SbpA. (B) Titration binding curve of Tb³⁺ to 1.51 μM SbpA as measured by Tb³⁺ luminescence. The complex was excited at 280 nm and the emission observed at 544 nm. The solid circles represent data points and the solid line is a fit of a one-site binding curve to the data. The dashed lines are fits to linear portions of the data. The intersection of the dashed lines was used to determine the stoichiometric break-point for this binding curve. (C) Competition of Tb³⁺-SbpA complexes with Ca²⁺ as measured by a decrease in the Tb³⁺ luminescence. The open circles are titration data, which are fit by a competition binding curve shown by a solid line. The error bars represent the standard deviation determined from two experiments.

Figure 7

Self-assembly by SbpA corresponds to irreversible growth. (A) Estimated sheet lengths from the end point light scattering data are plotted as a function of the [Ca²⁺] and [SbpA] at 780 min (Figure 2A). The lengths were calculated using the equation for scattering by dielectric discs and using a sheet thickness of 10 nm. The color map is capped at 3000 nm due to the variability of the lengths calculated with high light scattering. (B) A_norm as a function of time for SbpA self-assembly at 50 mM Ca²⁺ and increasing protein concentration (colored curves). Fits to eq 4 are shown as dashed curves in the same color.