Research data supporting: Applying 3D correlative structured illumination microscopy and X-ray tomography to characterise herpes simplex virus-1 morphogenesis

This dataset consists of 3D tomographic data collected using cryo-soft-X-ray tomography (cryoSXT) and cryo-structured illumination microscopy (cryoSIM) data. These data were collected for a study designed to assess (a) the morphogenetic pathway of herpes simplex virus (HSV)-1 assembly and (b) how the loss of different HSV-1 proteins known or suspected to be involved in virus assembly and egress differentially affected virus morphogenesis Cryo-soft-X-ray tomography (cryoSXT) collection parameters An UltraXRM-S/L220c X-ray microscope (Carl Zeiss Xray microscopy) was used to collect the datasets at Beamline B24 at Diamond Light Source. Incident soft X-rays generated at the synchrotron (500 eV, λ = 2.48 nm) were used to illuminate the samples and transmitted X-rays were detected using a 1024B Pixis CCD camera (Princeton Instruments). Transmitted light was focused using a 25 nm zone plate objective with a nominal resolution limit of 25 nm. Samples were focused by Z translations of the zone plate and samples were centred along a rotational axis by Z translations of the sample grid. Tomographic data were collected at fields of view measuring 9.46×9.46 μm. In each case, a collection of X-ray projections known as a tilt series were acquired by rotating the sample with maximum tilt angles of –60/+60 and acquiring images at increments of 0.2 or 0.5. A 0.5 or 1 second exposure time was used depending on the intensity of the transmitted X-rays. Tilt series were reconstructed using IMOD (version 4.9.2). For each field of view, a tilt series and a reconstructed tomogram are provided. Cryo-structured illumination microscopy (cryoSIM) collection parameters A cryoSIM developed in-house was used to collect datasets at Beamline B24 at Diamond Light Source. Hoescht stain fluorescence was excited using a 405 nm laser and was detected using an EM-452-45 filter (452 ± 22.5 nm). eYFP-VP26 fluorescence was excited using a 488 nm laser and was detected using an EM-525-50 filter (525 ± 25 nm). gM-mCherry fluorescence was excited using a 561 nm laser and was detected using an EM-605-70 filter (605 ± 35 nm). MitoTracker Deep Red fluorescence was excited using a 647 nm laser and was detected using an EM-655-lp filter (≥ 655 nm). CryoSIM data were reconstructed using SoftWoRx (AppliedPrecision Inc., Issaquah, WA) and the fluorescent channels were aligned using Chromagnon. Correlation of cryoSIM and cryoSXT CryoSIM data was correlated onto CryoSXT data using easyCLEMv0. CryoSXT tomograms were used as the target for the correlation and a frame was added around the tomogram, increasing the XY dimensions to 1200×1200 voxels. This increased the canvas size of the transformed cryoSIM data and reduced the probability that edges of the cryoSIM data would overlap with the tomogram at the centre of the frame during XY translations and rotations of the cryoSIM data. To avoid the need for an affine transformation model, maximum Z projections of cryoSIM data were used to generate transformation files using the rigid transformation model. The X-ray mosaic and a minimum Z projection of the tomogram were used as targets for these transformations. Rigid-transformation files generated using the 2D maximum Z projections of the cryoSIM data were applied to the 3D Z stacks to prevent the anisotropic stretching of signal generated by affine transformations in lieu of rotation around the X or Y axis. Before conducting a 3D rigid transformation of cryoSIM Z stacks onto the target tomogram, the number of slices in the cryoSIM image was increased 5-fold (reducing the voxel depth from 125 nm to 25 nm). This made it easier to correlate the front and back edges of mitochondrial fluorescence and tomographic mitochondria together. The TransformJ plugin in Fiji was used to further fine-tune the correlation where needed.