Figure 1.
Normalized autocorrelation curves for HRV2, viral RNA, and oligonucleotides obtained by fluorescence correlation spectroscopy.
(A) DyLight 488-labeled HRV2 (cyan), YOYO-1 iodide509-labeled in vitro-transcribed viral RNA (red), and the labeled oligonucleotides (DyLight 488-labeled 3′-oligo-nucleotide and FAM-labeled 5′-oligo-nucleotide had the same value; both in black) exhibit different diffusion times. The respective oligonucleotides (B and C, black) were separately hybridized to in vitro-transcribed viral RNA (3′ red and 5′ green) and measured before and after digestion with RNase A/RNase H (blue), respectively. Note that essentially the same results were obtained with RNA released from virions by heating to 56°C for 20 min. Dots, measured data; continuous lines, one-component (A, and black lines in B, C) and two-component fit (colored lines in B, C; contribution of the 2nd component calculated from the fit is given in %). Note the complete overlap of the curves corresponding to the free oligonucleotides and to the RNase-digested samples.
Table 1.
Diffusion coefficients (D) of labeled free oligonucleotides (probe), YOYO-1 iodide509-labeled in vitro-transcribed RNA, and DyLight 488-labeled virus.
Table 2.
Diffusion coefficients of free probes, probes hybridized to RNA, and probes hybridized to RNA followed RNaseA/RNaseH digestion.
Figure 2.
RNA sequences at the 3′ and the 5′-ends become accessible for hybridization with different kinetics.
HRV2 was incubated at 56°C for 0, 7, and 20 min. Dylight 488-labeled oligonucleotide specific for 3′ sequences (A) and 5′ sequences (B) was added and the autocorrelation function was measured (C, D). Identical experiments were carried out again with additional time points. The percentage of free oligonucleotide (red), oligonucleotide hybridized to free RNA (ochre) and oligonucleotide hybridized to RNA connected to the virus (blue) was calculated for each time point by employing a three-component fitting procedure (symbols). For better appreciation, the respective values for each constituent (free oligo, oligo bound to RNA, and oligo bound to RNA partially extruded from the virion) were fit by a sum of Boltzmann functions (lines). Data points are from two independent experiments. Note that repeated experiments showed the same trend over time, but there were differences in background and in the absolute values of the measurements.
Figure 3.
Subviral particles containing rod-like density accumulate on incubation of psoralen UV-crosslinked virions at 56°C.
A) Electron micrographs of phosphotungstate-stained HRV2 without a) and with b) psoralen UV crosslinking. Panels c) and d) samples were treated as a) and b) but heated to 56°C for 10 min. B) Incubation of psoralen UV crosslinked HRV2 as in Ad) resulted in the generation of various subviral particles. Examples selected visually are depicted in the insets. C) Particles representative for the examples in B) found at various incubation times were visually identified and counted on micrographs (number of particles counted at 2 min, 6,672; 4 min, 6,026; 6 min, 10,847; 8 min, 5,000; 10 min, 7,470; total count at time 0 was taken as 100% native). The mean of three separate experiments is shown. Since it is not possible to reliably distinguish native virus from full 135S A-particles by this method, we conjecture that particles classified as ‘full’ might also include virions that are still in the native conformation. Note that some broken particles are also visible in the micrographs. Bar, 100 nm.
Figure 4.
Cryo-electron microscopy image analysis of subviral HRV2 particles obtained on heating psoralen UV crosslinked virions.
A) Examples of micrographs of particles visually classified as full (left column), ‘rod-containing’ (middle column), and empty (right column). B) Distribution of the relative RNA density in the individual particle images. Core density (r = 0–113 Å) was related to part of the density of the protein shell (r = 113–143 Å; see Fig. S1) for all particle images (all) and the classes obtained by ML3D-classification. Class1 (full), class2 (rods), class3 (empty) relate to the 3DR shown in Fig. 5.
Figure 5.
Cryo-electron microscopy image reconstruction of class1, 2, and 3 as obtained by maximum likelihood 3D-classification of psoralen UV crosslinked and heated virions viewed down a 2-fold axis.
(a)–(c), radially color-coded volumes obtained on 3DR imposing icosahedral symmetry (upper half) or without imposing symmetry (lower half); (d) and (f), transverse central sections; (e) for better appreciation of the internal rod-like density, the particle was cut open at about 6 nm above the center of the virion; (g)–(i), central slabs (∼3.8 Å thick); in (e) and (j)–(l), class2 particles are shown with an arrow along the longitudinal axis of the rod indicating the orientation of the internal density with respect to the icosahedral symmetry. Note that the views are on different 2-fold axes. Rotation of the model (k) by 180° around the y-axis indicates that the ‘rod’ contacts the inner face of the virion shell at positions close to a 2-fold and a 3-fold axis at roughly opposite sides. All volumes are depicted with sigma = 1 above the mean density. In (h) the positions of 2-fold (red arrow), 3-fold, and 5-fold axes are indicated. In (a), an obvious deviation from symmetry is indicated with a small black arrow.
Figure 6.
Capillary electrophoretic analysis and quantification of (sub)viral particles obtained on incubation at 56°C.
HRV2, non-crosslinked (A) and crosslinked (B), was incubated at 56°C for the times indicated. Note that about 50% and 100% of all non-crosslinked native virus (N) was converted into empty particles and free RNA within 5 min and 20 min, respectively. For crosslinked virus (Nx) the conversion did not proceed beyond RNA-containing intermediates (Rx). Nevertheless, some free RNA, presumably with different degrees of degradation as reflected in different migration behaviour, was also seen. Only 3 time points are shown for clarity. In (C) and (D), the three components present in these samples were quantified including more time points. Native virus (crosslinked or non-crosslinked), red; intermediate particles including ‘rod-particles’, blue; empty particles, green. The presence of RNA was ascertained by extinction at 260 nm (not shown but compare to Fig. 7). IS, internal standard; DMSO (3.5 µM). Weiss and colleagues [36] found mobilities [in 10−9 m2/Vs] of −6.6 for native HRV2, of −8.2 for empty particles, and of −17.1 for intermediate particles. Whereas the present values for native and empty particles were almost identical with the ones determined in this previous work, the intermediate particles migrated as a very broad peak with mobilities between −19.2 and −30.2×10−9 m2/Vs suggesting that they carry RNA segments of different lengths externally. Continuous lines are derived from modelling the time-dependent equilibrium concentrations of the components with GEPASI [48].
Figure 7.
Capillary electrophoresis of crosslinked HRV2 heated to 56°C for 10 min reveals partial exposure of the RNA.
A) Control; without antibody. B) As in A, but incubated with mAB J2 directed against dsRNA. C) As in A, but incubated with micrococcal nuclease. D) As in B, but without heat treatment. Note that the antibodies did not bind to the control sample kept at room temperature (D), as the crosslinked virus had the same electrophoretic mobility as native virus [36]. All subviral particles contained RNA as seen from extinction at 260 nm (values multiplied by 3 for better appreciation). Internal standards, DMSO (I.S.) and benzoic acid (B.A.). Blue arrow, uncoating intermediate; red arrow, intact virion.
Figure 8.
Viral RNA sequences close to the 5′-end are protected against RNases by the protein shell of subviral particles.
A) HRV2 was subjected to psoralen UV crosslinking and electrophoresed on a native 0.7% agarose gel without heating (Nx) and after heating to 56°C for 10 min and incubated with micrococcal nuclease at 37°C for 20 min (Rx). RNA was identified by ethidium bromide staining and the bands were cut out. B) Aliquots of the excised bands were boiled in reducing sample buffer and the proteins were separated on a 12–20% gradient SDS-PAGE gel followed by silver staining. V, untreated HRV2 used as a control; M, markers. C) RNA was extracted from aliquots of the gel pieces and subjected to RT-PCR using primer pairs complementary to sequences at the positions indicated in the scheme at the bottom. RNA bracketed by the sequences of the respective oligonucleotide pairs was revealed by amplification followed by agarose-gel electrophoresis and staining with ethidium bromide and correlated with the bands of marker DNA run on the same gel.
Figure 9.
Schematic representation of the time-dependent conversion (with rate constant k1) of native virus to intermediate particles (that are in the process of RNA release and include ‘rod-particles’) and of such intermediates particles further to empty capsids (with rate constant k2).
Intermediates with partially-released RNA (with diffusion properties similar as native virus) and free RNA are distinguished by FCS (Fig. 1) and their relative amounts as a function of the incubation time at 56°C of native virus were determined by FCS using oligonucleotides hybridizing to sequences close to the 3′- and the 5′-end, respectively (Fig. 2). The percentage of ‘rod-particles’ (shown in Figs. 3 to 5) contributing to the RNA-release intermediates was strongly enhanced by RNA crosslinking (Fig. 3); almost no (completely) empty particles were detected by CE (Fig. 6). Like FCS (Fig. 2), CE demonstrated the presence of externalized RNA in intermediate particles (Fig. 7). By using RT-PCR, only sequences close to the 5′-end of the RNA were detected in RNase-digested uncoating intermediates (Fig. 8), indicating that the 3′-end had become accessible.