High cooperativity of the SV40 major capsid protein VP1 in virus assembly

Abstract

SV40 is a small, non enveloped DNA virus with an icosahedral capsid of 45 nm. The outer shell is composed of pentamers of the major capsid protein, VP1, linked via their flexible carboxy-terminal arms. Its morphogenesis occurs by assembly of capsomers around the viral minichromosome. However the steps leading to the formation of mature virus are poorly understood. Intermediates of the assembly reaction could not be isolated from cells infected with wt SV40. Here we have used recombinant VP1 produced in insect cells for in vitro assembly studies around supercoiled heterologous plasmid DNA carrying a reporter gene. This strategy yields infective nanoparticles, affording a simple quantitative transduction assay. We show that VP1 assembles under physiological conditions into uniform nanoparticles of the same shape, size and CsCl density as the wild type virus. The stoichiometry is one DNA molecule per capsid. VP1 deleted in the C-arm, which is unable to assemble but can bind DNA, was inactive indicating genuine assembly rather than non-specific DNA-binding. The reaction requires host enzymatic activities, consistent with the participation of chaperones, as recently shown. Our results demonstrate dramatic cooperativity of VP1, with a Hill coefficient of approximately 6. These findings suggest that assembly may be a concerted reaction. We propose that concerted assembly is facilitated by simultaneous binding of multiple capsomers to a single DNA molecule, as we have recently reported, thus increasing their local concentration. Emerging principles of SV40 assembly may help understanding assembly of other complex systems. In addition, the SV40-based nanoparticles described here are potential gene therapy vectors that combine efficient gene delivery with safety and flexibility.

Figure 1. Parameters affecting the in vitro packaging reaction.

Effect of temperature (A) and concentration of monovalent salts (B) present during the re-association reaction (step B) on the yield, measured as titer of luc transducing units.

Figure 2. The reaction requires cellular factors.

A–VLPs were purified on sucrose gradient from nuclear extracts harvested on day 3. M-molecular weight marker; VLPs–purified VLPs used in the packaging experiment (Table 1). B-Nuclear extracts were harvested from infected Sf9 cells at different time points after infection, as designated on top, and analyzed by SDS-PAGE and Coomassie-blue staining. M-molecular weight marker; Mock-nuclear extract of Sf9 cells infected with wild type baculovirus; SV40-nuclear extract of CV1 cells infected with wild type SV40; BSA-1 µg BSA, a standard for band intensity. C-Activity of the nuclear extracts shown in part B. ○-Packaging activity measured as titer (left ordinate); □-VP1 level estimated from A (right ordinate).

Figure 3. Cooperativity of VP1 in the reaction.

Effect of substrate concentration on the reaction yield. A–DNA; B–VP1. The amounts used per reaction, as described in Materials and Methods, is indicated. C-Fourteen different batches of nuclear extracts were assayed for packaging activity. Only 11 distinct data points are seen because of overlap of some of the data points. D-Calculation of Hill coefficient from the same set of data as in C. Y is the fraction of fully assembled VP1, measured as TU.

Figure 4. Equilibrium sedimentation in CsCl gradient.

The reaction products, 2 ml, were fractionated on a CsCl density gradient in stabilization buffer at pH 5.2 (Materials and Methods). VP1 was analyzed by Coomassie and Bradford, DNA by real-time quantitative PCR and luc TU as described in Materials and Methods.

Figure 5. Analysis of the particles.

A,B-Proteins were separated by elecrtrophoresis in Tricine buffer on 16% polyacrylamide gel. Western blotting was performed with polyclonal antibody against histone H3 (Upstate) (A) and against VP1 (B). M–size marker; 1–Nuclear extracts used in the packaging reaction; 2–In vitro assembled particles, fraction 9 of Fig. 4. 3–wild type SV40. C-Analysis of the packaged DNA. Particles purified by ultrafiltration were treated with Tris base (200 mM) in presence of 25 mM EGTA and 25 mM DTT for 1 hr at 37°C. DNA was extracted by phenol-chloroform treatment in presence of 1% SDS, and analyzed by Southern blotting with pGL3-control as a probe.

Figure 6. Structure of the nanoparticles.

Transmission electron microscopy pictures of (A) VLPs, (B) in vitro packaged nanoparticles, and (C) wild type SV40. Samples were adsorbed onto formvar-carbon-coated copper grids and stained with 1% sodium phosphotungstate, pH 7.0. The samples were viewed in a Philips CM-12 electron microscope, using a voltage of 100 kV, and photographed at a magnification of 53,000×.

Figure 7. A model for the in vitro assembly reaction.

The addition of DTT to nuclear extracts (A) leads to disassembly of the VLPs. Following the addition of supercoiled DNA (B) pentamers bind along each DNA molecule. This increases the local concentration of VP1, facilitating concerted assembly. Reassembly may be facilitated by presence of chaperones in the nuclear extracts. Assembly is accompanied by DNA condensation, presumably via the action of topo II. In Step C the capsids are stabilized at pH 5.2.