View metadata, citation and similar papers at core.ac.uk brought to you by CORE provided by Springer - Publisher Connector Histochem Cell Biol (2008) 130:185–196 DOI 10.1007/s00418-008-0459-y REVIEW Cryo-electron tomography of cells: connecting structure and function Vladan LubiT · Andrew Leis · Wolfgang Baumeister Accepted: 3 June 2008 / Published online: 20 June 2008 © The Author(s) 2008 Abstract Cryo-electron tomography (cryo-ET) allows molecular resolution (rev. in Lucic et al. 2005; Nicastro the visualization of cellular structures under close-to-life et al. 2005). Rapid freezing followed by the investigation of conditions and at molecular resolution. While it is inher- the frozen-hydrated samples avoids artifacts notorious to ently a static approach, yielding structural information chemical Wxation and dehydration procedures (Dubochet about supramolecular organization at a certain time point, it et al. 1988). Furthermore, the biological material is can nevertheless provide insights into function of the struc- observed directly, without heavy metal staining, avoiding tures imaged, in particular, when supplemented by other problems in interpretation caused by an unpredictable accu- approaches. Here, we review the use of experimental meth- mulation of staining material. Consequently, cryo-ET of ods that supplement cryo-ET imaging of whole cells. These whole cells has the advantage that the supramolecular include genetic and pharmacological manipulations, as well architecture can be studied in unperturbed cellular environ- as correlative light microscopy and cryo-ET. While these ments (Baumeister 2002). methods have mostly been used to detect and identify struc- We can distinguish two types of cryo-preparation. Thin tures visualized in cryo-ET or to assist the search for a fea- samples (typically below 0.5 m in thickness) can be rap- ture of interest, we expect that in the future they will play a idly frozen by plunging into liquid ethane (Dubochet et al. more important role in the functional interpretation of cryo- 1988) and imaged in an electron microscope without fur- tomograms. ther processing. While plunge-freezing is limited to small cells and thin regions of cells, thicker samples are frozen by Keywords Cryo-electron tomography · high-pressure freezing (Moor 1987) and then cryo-sec- Correlative light microscopy · Electron microscopy tioned, that is sectioned to typically 50–200 nm thick sec- tions at cryogenic temperatures (Al-Amoudi et al. 2004). In both cases the high freezing rates ensure that a sample is Introduction vitriWed, i.e. formation of ice crystals is avoided. Also, sam- ples are kept and imaged in EM at cryogenic temperatures Among techniques used in cell biology, cryo-electron to avoid devitriWcation. tomography (cryo-ET) is a relatively recent one. It com- Electron microscopy (EM) in general, and cryo-ET bines the advantages of 3D imaging with a close-to-life being no exception, delivers static images of a biological preservation and it allows studying biological material at system, as does X-ray crystallography. In spite of this apparent limitation, both methods can provide a structural Histochemistry and Cell Biology Lecture presented at the 50th framework for a mechanistic understanding of molecular Symposium of the Society for Histochemistry in Interlaken, and cellular functions. Cryo-ET has already pushed the reso- Switzerland, 1–4 October 2008. lution limits and increased the Wdelity of the images, and we can expect this trend to continue. In this review, we V. LubiT · A. Leis · W. Baumeister (&) focus on the experimental methods that supplement cryo- Max Planck Institute for Biochemistry, Am Klopferspitz 18, 82152 Martinsried, Germany ET imaging of whole cells. Various approaches were used e-mail: [email protected] to assist the detection or identiWcation of imaged structures 123 186 Histochem Cell Biol (2008) 130:185–196 and also to allow cryo-ET imaging of intracellular struc- ET is to obtain enough cryo-structures so that they can be tures arrested in a particular functional state. This is a very statistically analyzed at a reasonable conWdence level. DiY- young Weld, and consequently the current literature is not culties in sample preparation, as well as the time needed to extensive. We paid particular attention to the methods used record, reconstruct and analyze a cryo-tomogram limit and to the development of methods that might be relevant throughput. For example, the investigation of actin net- for the future. The methods were separated in two main cate- works in Wlopodia of slime mold Dictyostelium cells was gories. On one side are genetic and pharmacological facilitated by using genetically modiWed cells that over- manipulations that are used along the lines of the well- express small GTPase Rac1A and show an abundance of proven basic paradigm that involves investigations of Wlopodia (Medalia et al. 2007). Tomograms of wild-type untreated, treated and (when applicable) control samples. Wlopodia were used as a control, to make sure that the over- When applied to cryo-ET, a number of tomograms are expression of Rac1A did not alter actin networks. recorded for samples in each group and a value is associ- Cytoskeletal Wlament bundles observed in cryo-tomo- ated with each tomogram in order to assess the eVect of a grams of wild-type Caulobacter crescentus were classiWed treatment. Correlative light microscopy (LM) and EM in four types based on their dimensions and cellular locali- methods fall in the second category. The advantage of this zation (Briegel et al. 2006). In an attempt to identify at least approach is that LM provides information about the larger some of the Wlaments, the authors also imaged cells lacking environment and the functional states of structures visual- crescentin, one of the previously identiWed cytoskeletal pro- ized in cryo-ET, or alternatively, cryo-ET eVectively teins. Additionally, they imaged both wild-type and cres- extends the resolution achievable in LM. centin knockout cells that were treated with a drug that is known to depolymerize Wlaments composed of another cytoskeletal protein, MreB. While crescentin and MreB Genetic, pharmacological and related treatments could be ruled out as constituents of some Wlament types, integrated with cryo-ET the interpretation of the results was complicated by ambig- uous results. For example, the lack of one type of Wlament Finding and identifying a structure of interest in tomograms in the MreB depolymerizing-drug-treated crescentin knock- recorded under low-dose conditions having a poor signal- out cells, but not in the non-treated knockout and treated to-noise ratio is one of the basic problems in cryo-ET. It is wild-type cells, points to a particular interaction between also a prerequisite for providing insights to its function. crescentin, MreB and an as yet unidentiWed molecule form- Although a large number of observations obtained from ing these Wlaments. Also, although the authors analyzed a conventional EM are certainly very useful in this respect, fair number of tomograms, another type of Wlament could more sophisticated detection methods are needed if cryo- not be identiWed because its abundance in wild-type cells ET is expected to reach beyond ultrastructure to the molec- was too low. Consequently, this example shows how even a ular architecture level. In straightforward cases, when moderately complicated experimental system (four condi- features of interest possess strong intrinsic contrast, they can tions, four Wlament types) can go beyond current limits of be easily identiWed. For example, the location and spatial the interpretability of cellular cryo-ET. An even more elab- arrangement of magnetosomes, small organelles containing orate version of the same wild-type—mutant—rescue magnetite crystals present in magnetotactic bacteria, were scheme was applied in the investigation of Escherichia coli compared in wild-type and mutant cells lacking proteins chemotaxis receptor arrays (Zhang et al. 2004, 2007). It that were expected to be relevant for magnetosome organi- involved genetic deletion of chemoreceptors and/or another zation and function (Komeili et al. 2006; ScheVel et al. two proteins of the chemosensory system, CheA and 2006). In both studies, the wild-type organization of mag- CheW, and their subsequent expression at diVerent relative netosomes was restored after overexpression of the deleted levels. As a result, it was conWrmed that the periodic struc- proteins. Cryo-preparation was necessary for the preserva- tures located at cell poles represent chemoreceptors, illus- tion of cytoskeletal Wlaments of Spiroplasma melliferum trating how a genetic approach can assist the identiWcation (Kurner et al. 2005) and of cell surface Wlaments of Flavo- of structures observed in cryo-tomograms. bacterium johnsoniae (Liu et al. 2007) and allowed a rela- In all examples presented so far, various experimental tively straightforward detection. This in turn made an assays were used to assist detection and identiWcation of integrative approach possible, whereby the genetic deletion structures of interest in cryo-tomograms, but not to investi- of GldF, one of the envelope-associated gliding proteins, gate cells in a well-deWned functional or developmental led to a complete loss of Wlaments (Liu et al. 2007). The stage. The latter approach was implemented by partially reintroduction of the deleted protein rescued the Wlaments, synchronizing the growth phase of the unicellular eukary- thus, providing new evidence for the involvement of GldF ote Ostreococcus tauri by light–dark cycles and revealed in Wlament formation. Another challenge inherent to cryo- the detailed ultrastructure of whole cells, tightly packed 123 Histochem Cell Biol (2008) 130:185–196 187 with organelles, even though the resolution was eVectively good structural preservation and good labeling. In general, limited by the thickness of this sample (Henderson et al. pre-embedding labeling methods (typically a sample is 2007). mildly Wxed, permeabilized and labeled by primary Rapid freezing techniques currently allow EM imaging reagents before it is embedded and sectioned) can achieve of the same process at well-deWned time points.