High-resolution mass spectrometry of viral assemblies: molecular composition and stability of dimorphic hepatitis B virus capsids

Abstract

Hepatitis B virus (HBV) is a major human pathogen. In addition to its importance in human health, there is growing interest in adapting HBV and other viruses for drug delivery and other nanotechnological applications. In both contexts, precise biophysical characterization of these large macromolecular particles is fundamental. HBV capsids are unusual in that they exhibit two distinct icosahedral geometries, nominally composed of 90 and 120 dimers with masses of approximately 3 and approximately 4 MDa, respectively. Here, a mass spectrometric approach was used to determine the masses of both capsids to within 0.1%. It follows that both lattices are complete, consisting of exactly 180 and 240 subunits. Nanoindentation experiments by atomic-force microscopy indicate that both capsids have similar stabilities. The data yielded a Young's modulus of approximately 0.4 GPa. This experimental approach, anchored on very precise and accurate mass measurements, appears to hold considerable potential for elucidating the assembly of viruses and other macromolecular particles.

Fig. 1.

EM and mass spectrometry of HBV capsids. (A) Capsids visualized by cryo-EM at magnification ×38,000. Two sizes of capsids are observed; large ones composed of nominally 120 dimers with T = 4 geometry (blue arrowhead) and small ones composed of nominally 90 dimers with T = 3 geometry (pink arrowhead). (Scale bar: 50 nm.) (B and C) Mass spectra of cp149 3^C→A capsids (composed of dimers without an intramolecular disulfide bond) (B) and cp149 61^C capsids (with an intramolecular disulfide bond) at 150 V accelerating voltage (C). The samples were at ≈0.04 μM capsids (≈8 μM monomer of capsid protein) in 200 mM ammonium acetate (pH 6.8). The distributions of peaks around m/z 22,000 and 25,000 represent the T = 3 and T = 4 capsids, respectively, as marked by the corresponding cryo-EM reconstructions (12) in pink and blue. For each distribution, the main charge state is indicated. The Insets on the right show spectra convoluted to uncharged species. For a comparison of detected peaks with ones calculated for capsids composed of either 89, 90, 119, or 120 dimers, see Table S1.

Fig. 2.

Mass spectrum of HBV cp183 capsids recorded at 200 V accelerating voltage. The sample was 6 μM (cp monomer) in 200 mM ammonium acetate (pH 6.8). Cp183 incorporates small RNAs upon assembly in E. coli (10). The amount of RNA per particle is variable, giving rise to heterogeneity in masses so that the resulting mass spectra are not well resolved. They are, nevertheless, highly reproducible, exhibiting a broad maximum at ≈1m/z 32,000 and a small shoulder on the high-m/z side. The absence of resolved charge states precludes a mass assignment. Furthermore, no fragmentation products (that could hint at the capsid masses) were observed, even at the highest applicable accelerating voltage (200 V). However, we were able to estimate that the masses are between ≈5,000 and 6,500 kDa (see Results). The mass spectrum of cp149 61^C is given in the back for comparison.

Fig. 3.

Dissociation behavior of HBV cp149 capsids in vacuo. Tandem mass spectra of T = 3 3^C→A (A) and 61^C (B) capsids at m/z 21,700 show dissociated monomers (3^C→A) and dimers (61^C) in the low-mass range. As marker symbols, one or two subunits of the capsid crystal structure (15) are used, although we note that there is no evidence of a monomeric species that has this fold. Undissociated capsids [marked by EM reconstruction (12)] and dissociated countercomplexes (number of lost monomers or dimers, as indicated) are observed in the high-mass range. Collision voltages were 150, 180, and 200 V from front to back for both constructs. Charge states of the base peaks for each distribution are given. At the highest applicable collision voltage (200 V), T = 3 3^C→A loses up to seven monomers and 61^C up to three dimers. The spectra are normalized on their base peaks. Magnification levels in the high mass range are indicated when used.

Fig. 4.

AFM data on HBV capsids. (A) AFM image of a T = 4 capsid, with a T = 3 particle at lower magnification (Upper Left Inset). The corresponding height profiles (Lower Left Inset) distinguish between the T = 3 and T = 4 capsids, detecting the 15% difference in diameter. The lateral apparent diameters are smeared by AFM tip convolution effects. (B) Example of force–distance (FZ) curves for a T = 4 and a T = 3 (Inset) particle. The reference curve “FZ glass” is shown together with the 1st, 2nd, 10th, and 35th FZ curves. Particle indentation is denoted by Δh. We measured average spring constants of 0.143 ± 0.007 N/m (±SEM, n = 55) and 0.093 ± 0.007 N/m (±SEM, n = 46) for the T = 3 and T = 4 capsids. (C) Spring constant k for one of the particles as a function of the indentation cycle. To test whether subunits can be removed, the capsids were subjected to a sequence of small indentations. Even after 35 indentations, their stiffness was unchanged; hence, it can be concluded that no subunits were removed. Furthermore, no significant decrease in height was observed after 35 indentations. Error bars show the standard deviation.