Making New Out of (G)old
Osmium and Epon as Standard for Immunogold Localization
- Fig. 1: Immunogold localization of the S-layer protein Msed_1806 of M. sedula, grown on pyrite (FeS2). The protein is located at the cell surface, on top of the cytoplasmic membrane. (HPF; FS: E.G.F.U.: epon). Scale bar: 200 nm.
- Fig. 2: Transmission electron micrograph of a thin sectioned P. tricornutum cell (A; HPF, FS: A.O.U.H.: epon). (B) At higher magnification of the indicated area in (A), the four plastid membranes can be easily distinguished (also see ). (C) Localization of the β-glucan chrysolaminarin within the marbled vacuole. Some of the 5 nm gold particles are indicated (arrows). Mt, mitochondrium; Nu, nucleus; PI, plastid; V, vacuole. Scale bars: 1 µm (A) and 200 nm (B).
- Fig. 3: Immunogold localization of the S-layer protein Msed_1806 of M. sedula, grown on pyrite (FeS2) (HPF; FS: E.G.F.U; epon). The majority of protein is located at the cell surface and to a minor portion within the cytoplasm where the protein is produced (D). Py, pyrite. Scale bars: 500 nm (A, D) and 200 nm (B, C).
The results of several independent electron microscopic studies could show that the use of epon and osmium tetroxide do not exclude subsequent immunogold labeling of antigens. The key to success is high-pressure freezing and freeze substitution in acetone or ethanol containing osmium and/or uranyl acetate. A subsequent embedding in epon resins not only leads to excellent structural preservation of biological material, it still enables immunogold localization studies of proteins or sugars.
High-pressure Freezing as Prerequisite
As a prerequisite for any kind of ultrastructure research on biological samples in the transmission electron microscope (TEM), the appropriate and careful preparation of biological material like prokaryotic and eukaryotic cells or tissues is essential. The final resolution in the resulting images or models is not limited by the properties of the TEM, which would theoretically set this limit in a range of about 4 pm, but is highly influenced by the general quality of the preparation method and in this regard also with negative effects and artifacts caused by improper handling.
The combination of high-pressure freezing (HPF) and automatic freeze substitution (FS) was shown to be essential for a preservation of biological specimens close to their natural state [1,2,3]. It is therefore used as standard preparation method for embedding and sectioning at room temperature as well as in cryo-sectioning and subsequent immunogold labeling. In a standardized application of two different freeze substitution media with or without osmium tetroxide and the following embedding in epon resins it could be shown that it is possible to have good structural preservation without completely damaging epitopes for immunogold localization studies. These protocols found application on several organisms ranging from multicellular eukaryotic systems down to unicellular eukaryotes (e.g. diatomes) and prokaryotes (Bacteria and Archaea). Three selected examples from the eukaryotic photosynthetic diatome Phaeodactylum tricornutum (fig. 2A) and the thermoacidpohilic coccoid crenarchaeum Metallosphaera sedula (fig.
1) were involved in this short survey. In P. tricornutum, photosynthesis is carried out by a plastid that has evolved out of a so called secondary endosymbiosis [4,5]. This plastid is therefore surrounded by four consecutive membranes (fig. 2B) and proteins which are encoded in the nucleus and should be directed in one certain compartment inside the plastid have to cross the membranes via distinct transporters (e.g. Sec61, SELMA, Omp85, TIC). For a definite localization of proteins, the visibility of membranes which more or less implies optimal structural preservation by the use of osmium tetroxide (OsO4) is indispensable. With this study it should be shown that the combination of OsO4, epon resins and immunogold localization is not an absolute no-go. In most cases, acetone containing 0.2 or 2% OsO4, 0.1% uranyl acetate and 9.3% water (A.O.U.H.) was used as freeze substitution medium [6,7,8]. As an alternative, ethanol containing 0.5% uranyl acetate, 0.5% glutaraldehyde, 1.0% formaldehyde and 5.4% water (E.G.F.U.) was used. In all described experiments, epoxid resin (epon) was used as embedding substrate [7,8,9].
1. GFP-fusion Proteins in P. tricornutum
The majority of protein localization studies in the model organism P. tricornutum were performed by the utilization of GFP-fusion proteins. By transfection of the cells with the respective DNA constructs, the resulting GFP fluorescence can be detected in the light microscope and first estimations about the localization can be drawn. After high-pressure freezing and freeze substitution with A.O.U.H. (0.2 % OsO4), the cells are embedded in epon resin. Finally, thin sectioning of the samples is followed by immunogold labeling, using primary antibodies against GFP (goat-α-GFP, dilution 1:1000, polyclonal) and secondary antibodies rabbit-α-goat IgG (dilution 1:20) coupled to gold particles (10 nm) for detection in the electron microscope. As a result, GFP-fusions of the three proteins Glx, Syn and MutS could all be localized within the photosynthetically active plastid of P. tricornutum .
2. Detection of β-Glucan in P. tricornutum
For the retention of sugars and sugar moieties within the cells, high-pressure freezing and especially the subsequent freeze substitution is the method of choice for further analyzes. One reason is that carbohydrates are usually washed out and get diluted in the graded ethanol series that is used for dehydration of biological material after conventional chemical fixation with aldehydes. The application of cryo-fixation methods and freeze substitution with acetone containing 0.2 % OsO4 and embedding in epon leads to excellent structural preservation, especially of the membranous structures within the plastid. Herein, the positive effects of this advanced preparation not only keep the carbohydrate portions within the cells, it also preserves the epitopes in a way that they can still be detected via immunogold labeling: i) With the lectin concanavalinA (ConA), glycoproteins like those involved in the secretory system, could be detected in epon embedded P. tricornutum cells, even when chemical fixation was performed. The ConA biotin conjugates were detected with streptavidin coupled to 20 nm gold . ii) The β-glucan chrysolaminarin could be located within the marbled vacuole of P. tricornutum (fig. 2C): primary antibody, mouse-α-betaglucan (monoclonal, dilution 1:10.000); secondary antibody goat-α-mouse IgG (5 nm gold, dilution 1:20). Not only that the latter experiment even worked with a monoclonal antibody at dilutions of 1:10.000, the localization was impossible with chemical fixation and dehydration with ethanol which pronounces the importance of high-pressure freezing and freeze substitution.
3. The S-Layer of M. sedula
The purification of the S-layer protein Msed_1806 of the thermoacidophilic crenarchaeon M. sedula was carried out according to the protocols described previously [8,10], polyclonal antibodies against the purified protein were generated in rabbits. After high-pressure freezing and freeze substitution of cellular water with E.G.F.U.-medium, the cells were embedded in conventional epon resin. On section immunogold labeling with the primary antibody rabbit-α-Msed_1806 (dilution: 1:500) and a secondary antibody goat-α-rabbit IgG (6 nm gold; dilution: 1:20) led to the expected localization of the S-layer protein at the cell surface (fig. 3) .
There are several advantages of epon and osmium tetroxide like the structural preservation and also the visibility of cellular components (e.g. membranes) by the application of the described protocols. One major goal of this study was to proof that immunogold labeling on epon thin sections is still possible after high-pressure freezing combined with freeze substitution, even when osmium tetroxide is involved. An important fact about epon resins is that there have been some significant changes of the chemical nature of the resin in the late 1970s and early 1980s and also some substitutes were introduced by companies under the same name. Although it has to be noted that the quality of antibodies (polyclonal should be favored), quality and number of epitopes and their sensitivity against the preparation plays an important role, this overview should encourage people to give this method at least a try.
For their support in this work, the following people should be acknowledged: Uwe-G. Maier for allocation of the electron microscopic facilities in Marburg, K. Bolte for assistance in microscopy, D. Märte, M. Debus and M. Johannsen for technical assistance and F. Hempel for providing biological material. A.K. was supported by the LOEWE program of the state of Hessen.
 Möbius W.: Ann. Anat. 191, 231-47 (2009)
 Schwarz H. & Humbel B.M.: Methods Mol. Biol. 369, 229-56 (2007)
 McDonald K.: Protoplasma, DOI 10.1007/s00709-013-0575-y (2013)
 Stork S. et al.: Protoplasma 250, 1013-23 (2013)
 Hempel F. et al.: Biol. Chem. 388, 899-906 (2007)
 Peschke M. et al.: PNAS 110, 10860-65 (2013)
 Klingl A. et al.: Proceedings MC 2013, urn:nbn:de:bvb:355-epub-287343 (2013)
 Klingl A.: PhD thesis, urn:nbn:de:bvb:355-epub-196832 (2011)
 Rachel R. et al.: Methods Cell Biol. 96, 47-69 (2010)
 Klingl A. et al.: Arch. Microbiol. 193, 867-82 (2011)
Dr. Andreas Klingl (corresponding author via e-mail request)
LOEWE Centre for Synthetic Microbiology (Synmikro)
Department of Cell Biology