Minimal effects of macromolecular crowding on an intrinsically disordered protein: a small-angle neutron scattering study

Abstract

Small-angle neutron scattering was used to study the effects of macromolecular crowding by two globular proteins, i.e., bovine pancreatic trypsin inhibitor and equine metmyoglobin, on the conformational ensemble of an intrinsically disordered protein, the N protein of bacteriophage λ. The λ N protein was uniformly labeled with (2)H, and the concentrations of D2O in the samples were adjusted to match the neutron scattering contrast of the unlabeled crowding proteins, thereby masking their contribution to the scattering profiles. Scattering from the deuterated λ N was recorded for samples containing up to 0.12 g/mL bovine pancreatic trypsin inhibitor or 0.2 g/mL metmyoglobin. The radius of gyration of the uncrowded protein was estimated to be 30 Å and was found to be remarkably insensitive to the presence of crowders, varying by <2 Å for the highest crowder concentrations. The scattering profiles were also used to estimate the fractal dimension of λ N, which was found to be ∼1.8 in the absence or presence of crowders, indicative of a well-solvated and expanded random coil under all of the conditions examined. These results are contrary to the predictions of theoretical treatments and previous experimental studies demonstrating compaction of unfolded proteins by crowding with polymers such as dextran and Ficoll. A computational simulation suggests that some previous treatments may have overestimated the effective volumes of disordered proteins and the variation of these volumes within an ensemble. The apparent insensitivity of λ N to crowding may also be due in part to weak attractive interactions with the crowding proteins, which may compensate for the effects of steric exclusion.

Figure 1

Small-angle neutron scattering from BPTI (A) and myoglobin (B) with and without solvent contrast matching. Data for the different samples are displaced vertically for clarity, in the order indicated by the key. Unless indicated otherwise, the D₂O concentration was 56% for BPTI samples and 43% for myoglobin samples. The error bars in this and other SANS profiles represent standard deviations calculated from counting statistics. To see this figure in color, go online.

Figure 2

Small-angle neutron scattering from 0.01 g/mL λ N in the presence and absence of globular crowding proteins. D₂O concentrations were 46, 56, and 43% for the samples containing no crowding protein, BPTI, or myoglobin, respectively. Data for different samples are displaced vertically for clarity. (Curves) Fits of the experimental data to the SANS profile predicted for a calculated ensemble of random λ N conformations, with the conformations weighted by a solvation free energy parameter of 1 cal/mol. Each experimental profile was fit to the same predicted curve using the method of least squares, with the experimental data points inversely weighted by their square roots to reflect counting uncertainties. The fits incorporated two adjustable parameters—a multiplicative scaling factor to account for differences in total scattering intensity and an additive constant to account for errors in background subtraction.

Figure 3

Small-angle neutron scattering from λ N in the presence and absence of globular crowding proteins, presented as Guinier (A) and log-log (B) plots. In panel A, the radius of gyration was estimated by fitting data for 0.022 Å⁻¹ ≤ q ≤ 0.049 Å⁻¹ to the Guinier relationship (Eq. 3). In panel B, the fractal dimension, D_m, was estimated by fitting the data for 0.078 Å⁻¹ ≤ q ≤ 0.21 Å⁻¹ to the power-law relationship of Eq. 4. The uncertainties in R_g and D_m are the standard errors derived from the least-squares fits.

Figure 4

Simulation of solvation and crowding effects on λ N. (A and B) Distributions of λ N conformations in the presence of increasing concentrations of spherical particles with radii of 15 Å (A) or 20 Å (B). An unweighted distribution composed of ≈225,000 conformations was weighted to represent the effects of an unfavorable solvation free energy of 1 cal/mol and the indicated crowding density. (Dashed lines) Distributions for the different crowding densities, which were calculated from the excluded covolumes of the individual conformations and a sphere of the indicated radius, as described in the text. The distributions are represented as the number of chains with radii lying within intervals of 2 Å. (C and D) RMS (R_g) of weighted ensembles as a function of crowding density, for different values of the solvation free energy. To see this figure in color, go online.

Figure 5

Two-dimensional histogram representing the unweighted distribution of ∼225,000 simulated λ N conformations with different radii of gyration (R_g) and equivalent sphere radii (r_eff), calculated from the excluded covolumes with a crowding sphere of radius 15 Å (Eq. 1). The histogram is divided into 20,000 bins representing intervals of 0.35 Å in each dimension. The occupancies of the bins are represented by the color key, as the fraction of the total number of conformations. (Dashed line) Bins for which r_eff = R_g. Structures of representative conformations in the ensemble are shown as space-filling models, each oriented so that the larger two dimensions lie approximately in the plane of the page. The sphere represents the crowder used to calculate the covolume, with its position on the graph indicating its own surface radius (15 Å) and radius of gyration (11.6 Å). To see this figure in color, go online.