Abstract text (incl. figure legends and references)

Morphogenesis of human cytomegalovirus (HCMV) is still only partially understood. Previous studies have shown that the HCMV tegument protein pUL71 plays an important role in viral replication and morphogenesis (1). To better understand the function of the protein, pUL71 should be localized in infected cells by pre-embedding immunogold labelling and transmission electron microscopy (TEM). In recent years, EM preparation methods have been published that combine good structural preservation by high-pressure freezing and freeze substitution with subsequent rehydration allowing immunolabeling before resin embedding (e.g., 2,3).

We describe the establishment of this method in our laboratory and its adaptation for HCMV-infected fibroblasts. The major aim is to achieve the best possible ultrastructural preservation while providing sufficient permeabilization for efficient pre-embedding immunogold labeling.

After high-pressure freezing of firoblasts infected with HCMV expressing an HA-tagged variant of pUL71, we tested freeze-substitution (FS) media with different concentrations of glutaraldehyde, osmium tetroxide and uranyl acetate, as well as of the detergents TritonX100 and Digitonin. We then evaluated the ultrastructural preservation of these samples by TEM. Labelling efficiency was then first tested by fluorescence microscopy experiments. Rehydration and permeabilization approaches that provide good labelling efficiency will then be adapted for pre-embedding immunogold labeling of pUL71-HA with anti-HA IgG as primary and Nanogold Fab fragments as secondary antibody.

The most promising tradeoff between ultrastructural preservation, sufficient heavy metal contrast, and antibody penetration was observed with a FS solution containing 0.2% osmium tetroxide, 0.1% uranyl acetate and 5% water in acetone. Fluorescence microscopy showed specific antibody labelling of pUL71-HA after exchanging the FS medium at -20 °C for a medium containing 0.2% glutaraldehyde instead of osmium tetroxide, while it made no difference whether permeabilization was performed with TritonX or Digitonin.

Immunolabeling after FS but before embedding is a promising but complex method, as the fixatives as well as the acetone can interact with the antigens and mask the epitopes. Specifically, the proteolytic effect of osmium tetroxide has a strong effect on antigenicity. However, since this effect is strongly temperature dependent, removal of osmium tetroxide at negative temperatures should have less effect on antigenicity (4,5). The identification of a suitable anti-HA antibody in fluorescence microscopy experiments by using optimized fixation methods enables us to perform immunogold labelling for detection of the antibody by TEM.

1. Schauflinger, M. et al. 2013. Analysis of human cytomegalovirus secondary envelopment by advanced electron microscopy. Cellular Microbiology.

2. Hess, M.W. et al., 2018. Combining high-pressure freezing with pre-embedding immunogold electron microscopy and tomography. Traffic.

3. Tsang, T.K. et al., 2018. High-quality ultrastructural preservation using cryofixation for 3D electron microscopy of genetically labeled tissues. eLife.

4. Flechsler, J. et al., 2020. 2D and 3D immunogold localization on (epoxy) ultrathin sections with and without osmium tetroxide. Microscopy Research and Technique.

5. Koga, D. et al., 2020. Optimizing the reaction temperature to facilitate an efficient osmium maceration procedure. Biomed. Res. 41, 161–168.