Figure 1: Clonable, gold-clustering labels. A: Atomic structure of rat liver MTH (isoform II) at 2.0 Angstroms resolution (Braun et al., 1992). B: Metallothionein-Eg5 microtubule complexes with a helically averaged 3-D reconstruction highlighting the MTH-gold densities (yellow) and their position relative to Eg5 (blue). A helical average of an undecorated microtubule is shown in red. C: Desmin-MTH chimera polymerized into intermediate filaments present the gold clusters on their surface. D: 3-D model of the center filament in C.
Cryo-electron microscopy is expanding its scope from macromolecules towards much larger and more complex cellular specimens such as organelles, cells and entire tissues. While isolated macromolecular specimens are typically composed of only very few different components that may be recognized by their shape, size or state of polymerization, cellular specimens combine large numbers of proteinaceous structures as well as nucleic acids and lipid arrays. Consequently, an unambiguous identification of these structures within the context of a whole cell may create a very difficult challenge. On plastic-embedded specimens, or Tokuyasu sections epitopes that are exposed at the surface can be tagged by antibodies. However, vitrified sections have to be kept at strict cryo-conditions (below -140 degrees C) and therefore do not allow any post-sectioning treatment of the specimens other than data acquisition in the microscope. Hence, the labels have to be placed into the specimen before freezing. Here we report on the application of a small metal-clustering protein, metallothionein (MTH), as a clonable label capable of clustering metal atoms into a high-density particle with high spatial resolution. We tested MTH as a label for kinesin-decorated microtubules (MTs) as well as the building blocks of desmin intermediate filaments (IFs).
Braun, W., Vasak, M., Robbins, A.H., Stout, C.D., Wagner, G., Kagi, J.H., Wuthrich, K. (1992) Comparison of the NMR solution structure and the x-ray crystal structure of rat metallothionein-2. Proc. Natl. Acad. Sci. USA v89:10124-10128. PMID: 1438200 PMCID: PMC50290
Bouchet-Marquis C, Pagratis M, Kirmse R, Hoenger A. (2012) Metallothionein as a clonable high-density marker for cryo-electron microscopy J Struct Biol. 177:119-27.
2: Novel microtubule-associated structures in the ventral disc, the attachment organelle of Giardia intestinalis
In collaboration with: Scott C. Dawson, Dept. Microbiology, University of California, Davis, CA 95616, USA.
Figure 2: Regular arrays within the unique Giardia cytoskeleton. These arrays constitute excellent test grounds for our labeling experiments as they allow averaging of subvolume elements and calculating difference maps for a precise detection of clonable labels. A: Tomographic 20 nm thick X-Y slice of the Giardia ventral disc (green organelle in the Giardia cell shown in (Fig. 1C and D) at the level of the microtubule arrays (red line in C). B: 20 nm tomographic slice about 50 nm towards the dorsal side showing the arrangement of the Giardia microribbons (yellow line in C). The ventral disc is a large regular array of microribbons and microtubules with numerous associated proteins that await identification. C: Tomographic 20 nm thick X-Z slice of the Giardia ventral disc showing microtubules and associated microribbons end-on. D: Tomographic slice of the regular crystalline array of the marginal plate, which is found in the area where the anterior flagella cross each other. Elongated features on top are two anterior flagella. E: 3-D volume averaging of the repetitive units of the marginal plate (red square). F: Fourier-filtered 2-D projection of the area marked by the red square in D. G: diffraction spots demonstrate the highly regular arrangement of the marginal plate.
Figure 3: 3-D reconstruction of the microtubule-microribbon complex of the Giardia ventral disc. End-on (towards the microtubule plus-end) and side view of a grand average over 4,700 individual tomographic subvolumes. MT protofilaments are numbered, starting at the position of the seam. The largest associated densities are called side-arms (green). We do not know how many individual protein domains are within this structure. Insets to the right show cross-sections at their corresponding position in the 3-D map. The cross-section through the microribbons reveals a distinct 16 nm repeat, corresponding to two consecutive tubulin dimers along a protofilament. Side-arms repeat in register with the tubulin dimer repeat. The 3-D map still suffers from a missing cone of data, visible on the clear separation of protofilaments horizontally, but not vertically. Figure adapted from Schwartz et al. (2012).
Beyond the construction of canonical axonemes and spindles, many eukaryotes have evolved elaborate microtubule structures than enable specialized functions throughout their complex life cycles. Each of these complex microtubule-based structures, whether the subpellicular microtubules of apicomplexans (Cyrklaff et al., 2007) and trypanosomes (Vickerman and Preston, 1976) or the conoid of Toxoplasma (Hu et al., 2002) have unique supramolecular architectures. Very few structures of large microtubule-based organelles have been imaged in vivo, although this is obviously a critical step in ascertaining their function.
The complex ventral disc of the widespread protist parasite, Giardia intestinalis, is one such microtubule-based structure whose function in attaching to the host is critical to giardial pathogenesis (Cotton et al., 2011). Acute giardiasis resulting from ingestion of the giardial cyst form and subsequent colonization of the small intestine by the flagellated trophozoite form is the most common cause of protozoan intestinal infection in the U.S. and worldwide (Savioli et al., 2006). An estimated one billion people are currently infected with Giardia, mainly in the developing world. The current lack of concerted research effort has resulted in giardiasis being recently classified as a World Health Organization (WHO) neglected disease. Recent evidence of resistance to the widely used anti-giardial drugs (Land and Johnson, 1999), underscores the necessity of understanding giardial virulence and identifying alternatives to the limited number of known anti-giardial compounds.å
Cotton JA, Beatty JK, Buret AG (2011). International journal for parasitology 41: 925 to 933.
Cyrklaff M, Kudryashev M, Leis A, et al. (2007) The Journal of experimental medicine 204: 1281 to 1287.
Hu K, Roos DS, Murray JM (2002). The Journal of cell biology 156: 1039 to 1050.
Land KM, Johnson PJ (1999) reviews and commentaries in antimicrobial and anticancer chemotherapy 2: 289 to 294.
Savioli L, Smith H, Thompson A (2006) Trends in parasitology 22: 203 to 208.
Vickerman K, Preston TM (1976) Comparative cell biology of the kinetoplastid flagellates. 1st ed. Lumsden WCA, Evans DA, editors New York: Academic Press.
Schwartz CL, Heumann JM, Dawson SC, Hoenger A. (2012) A detailed, hierarchical study of Giardia lamblia's ventral disc reveals novel microtubule-associated protein complexes. PLoS One. 7:e43783.
Hagen K, Hirakawa M, House S, Schwartz CL, Pham JK, et al. (2011) Novel structural components of the ventral disc and lateral crest in Giardia intestinalis. PLoS Neglected Tropical Diseases 5: e1442.
3: Unsupervised Classification of Subvolume Averages in PEET Using Wedge-Masked Differences
After alignment and averaging of 3D subvolumes, unsupervised classification is prudent to check for heterogeneity within the population and, in some cases, to identify important structural or conformational variants. Principal components analysis (PCA), or related methods, followed by k-means, hierarchical, associative mapping or similar techniques are typically used for this purpose in other fields. In cryo-ET, however, missing tomographic data lead to artifactual clusters based primarily on missing wedge orientation and obscuring any real population heterogeneity that may be present. We have developed a new approach to dealing with this (Heumann et al., 2006) which, along with 2 prior techniques (Forster et al., 2008, Schmid and Booth 2008), now ships with PEET. This new method is considerably faster and, in some cases, appears to be at least as accurate and possibly more so than competing methods.
Forster, F. Pruggnaller, S., Seybert, A., et al. (2008). J. Struct. Biol. 161:276-286.
Schmid ,M. and Booth, C. (2008). J. Struct. Biol. 161:243-248.
Heumann, J., Mastronarde, D., and Hoenger, A. (2011). Clustering and variance maps for cryo-electron tomography using wedge-masked differences, J. Struct. Biol.175:288-299.
4: 3-D Cryo-EM on Microtubules complexed with Kar3Vik1 Heterodimer: Docking atomic X-ray Data into EM 3-D Volumes at different Nucleotide States
In collaboration with: Ivan Rayment, Dept. of Biochemistry, University of Wisconsin, Madison, WI 53706, USA, Susan Gilbert, Dept. of Biology, Rensselaer Polytechnic Institute, Troy, NY 12180, USA
Figure 4: A-B: Helical 3-D maps of Kar3Vik1 complexed to microtubules in the absence of nucleotide (A), and in the presence of AMP-PMP (B: see Cope et al., 2013). The Vik1 domain (yellow) is labeled with a maleimide-linked nanogold particle that unambiguously identified its position with regard to Kar3 (grey, underneath Vik1). The most striking difference between the two nucleotide-states is the stalk region (light blue) that swings towards the minus-end (to the top in B; towards the viewer in D) upon AMP-PNP uptake. C-D: Cross-sectional view of the 3-D electron density map of Kar3Cik1 docked into the isosurface mesh representation of the 3-D electron density map of Kar3Vik1 in the absence of nucleotide (C), and in the presence of AMP-PMP (D: see Gonzalez et al., under revision). Both maps are quite similar, but show some significant density differences at the location of the yellow volumes. Overall, Kar3Cik1 and Kar3Vik1 appear structurally identical at this resolution, and it seems that they utilize the same mechanism of movement to perform different functions in the cell. Wire mesh: Kar3Vik1; red/green/blue diffuse density: Kar3Cik1; solid yellow density: difference map. Figure adapted from Cope et al. (2013; A/B) and Gonzalez et al. (under revision; C/D).
Kinesin-14 motors generate minus-ended force that is utilized in mitosis, meiosis, and karyogamy. These motors are dimeric and operate with a non-processive powerstroke mechanism, but the role of the second motor head in motility has been unclear. In Saccharomyces cerevisiae, the Kinesin-14, Kar3, forms a heterodimer with either Vik1 or Cik1. Vik1 contains a motor homology domain that retains microtubule-binding properties but lacks a nucleotide-binding site. In contrast to the microtubule plus-end-directed processive motors, the Kinesin-14 motors are non-processive, minus-end-directed, and contain C-terminal motor domains connected by a N-terminal coiled coil. They utilize a powerstroke mechanism involving rotation of the coiled-coil stalk toward the minus end of microtubules instead of a stepping mechanism (Wendt et al., 2002; Endres et al., 2005). Sequence analysis suggests that all motors in this class are dimeric. This raises the question of whether the second motor head of the Kinesin-14 dimer plays a direct role in motility.
Our main contribution to this project was the generation of 3D maps from Kar3Vik1-microtubule complexes by cryo-electron microscopy and helical 3-D reconstruction. We produced three different maps at a nucleotide-free state, an ATP-mimicking state (with AMP-PNP) and an ADP-Pi mimicking state (with ADP-AlF2) respectively. The X-ray structure was solved from Kar3Vik1 crystals soaked with ADP, which is generally accepted as a solution state for kinesin motors.
Endres, N.F., Yoshioka, C., Milligan, R.A., et al. (2005). Nature. 439:875 to 878
Wendt, T.G., Volkmann, N., Skiniotis, G., et al. (2002). EMBO J 21:5969 to 5978.
Vale, R.D., and Fletterick, R.J. (1997). Annu Rev Cell Dev Biol 13:745 to 777.
Cope J, Rank KC, Gilbert SP, Rayment I, Hoenger A. (2013) Kar3Vik1 Uses a Minus-End Directed Powerstroke for Movement Along Microtubules. PLOS-one 8:e53792.
Rank KC, Chen CJ, Cope J, Porche K, Hoenger A, Gilbert SP, Rayment I. (2012) Kar3Vik1, a member of the Kinesin-14 superfamily, shows a novel kinesin microtubule-binding pattern. J Cell Biol. 197:957 to 70
Cope BJ, Heumann J, Hoenger A. (2011) Cryo-electron tomography for structural characterization of macromolecular complexes. Curr Protoc Protein Sci. Aug; Chapter 17:Unit17.13.
Cope J, Gilbert S, Rayment I, Mastronarde D, Hoenger A. (2010) Cryo-electron tomography of microtubule-kinesin motor complexes. J Struct Biol. 170:257 to 65.
5: Structural insight on mPyV replication and assembly factories by electron microscopy
In collaboration with: Robert L. Garcea, Univ. of Colorado at Boulder, MCD-Biology
Murine Polyomavirus (mPyV) is part of the polyomaviridae family. This family also includes, simian virus 40 (SV40), human BK and JC viruses (BKV, JCK). More recently, Merkel cell virus (MCV) was characterized (1) and shown to be closely related to murine PyV. Murine PyV was the first one discovered by Ludwik Gross in 1953. It is a double stranded, none enveloped virus. Its capsid (50 nm in diameter) is composed of 72 pentamers of the major coat protein VP1 and of VP2 and VP3 proteins buried within the capsid (2). After entry and migration to the cell host nucleus, the mPyV gene expression can be divided into two phases (3, 4). The early phase sees non-structural viral proteins such as small, middle and large T antigen to be synthesized. Those proteins highly interact with the host cell key proteins. As a result, the virus takes control of cellular mechanisms like; cell cycle progression towards S-phase as well as preventing the cell to engage apoptosis process. The early phase leads to viral DNA replication exploiting the host nuclear machinery. During the later phase, the structural proteins (VP1, VP2, VP3) are expressed and de novo viral particles are assembled in the nucleus.
In order to understand better the assembly process of the newly synthesized viral particles, we applied various electron microscopy approaches on A31-3T3 cells 36h after infection by the murine polyomavirus. Our study shows for the first time the presence of viral assembly factories and reveals their three-dimensional ultrastructural organization using electron tomography.
Erickson, K.D., Bouchet-Marquis, C., Heiser, K., Szomolanyi-Tsuda, E., Mishra, R., Lamothe, B., Hoenger, A., and Garcea, R.L. (2012) Nuclear Virus Factories in Polyomavirus-Infected Cells. PLOS Pathogens, Apr;8(4):e1002630.
6: Solving the Vimentin and Desmin Intermediate Filament Structure
Figure 6: Cryo-electron tomography on unfixed and unstained vimentin IFs. A: Cryo-electron micrograph of frozen-hydrated vimentin IFs that shows a remarkable stiffness in compacted filaments. At zones of partial unraveling four protofibril strands can be identified (box). B: Vimentin IFs in frozen-hydrated CHO cells prepared by vitreous sectioning. Note that, unlike the vimentin filaments in A and E, these native filaments show a central density. C: In unidirectional metal-shadowing preparations this twist could be identified as right-handed. D: Cryo-electron tomographic slice of frozen-hydrated filaments identical to the ones shown in A. Filaments are either compacted and appear rather stiff, or partially unraveled, exposing individual protofibrils that show a twisted arrangement (arrows). E: Slicing the filaments in x-z direction exposes their packing pattern with four protofibrils within each filament (black boxes).
In collaboration with: Harald Herrmann Lerdon, German Cancer Research Center, Heidelberg, Germany.
The cell's cytoskeleton consists of three individual filamentous systems: Microtubules, actin filaments and intermediate filaments (IFs). Together they coordinate a wide array of cellular functions such as cell division and material transport. IFs are still the least understood of the three regarding filament structure and assembly. With a diameter of about 10 nm they line up in between actin filaments ( about 8 nm) and microtubules ( about 20 nm). The classical view of IF function as a stress absorber and mechanical stabilizer of the cell has been gradually overcome by the growing knowledge of their interaction with various cellular components and how they integrate in various regulatory functions. The term IF comprises a large protein family with about 70 members in humans and IFs are found throughout the body in various tissues. IFs give rise to a large variety of genetic diseases ranging from premature aging (progeria) to muscular dystrophies with most of them based on point mutations in the individual proteins.
The assembly mechanism of IFs differs considerably from that of actin and microtubules. The rod shaped IF proteins form a coiled-coil dimer that assembles laterally into tetramers. Further lateral association generates a short barrel like structure of about 60 nm long and 16 nm in diameter termed Unit-Length-Filament (ULF). After ULF formation longitudinal assembly into full length IFs proceeds. Structural investigation of IFs is hindered by this elongated rod-like shape, which makes them nearly impossible to crystallize. In fact only fragments of some members of the IF family are crystallized and these crystal structures are only describing the central coiled-coil domain. The flanking regions are randomly coiled and can't be crystalized.
Kirmse R, Bouchet-Marquis C, Page C, Hoenger A. (2010) Three-dimensional cryo-electron microscopy on intermediate filaments. Methods Cell Biol. 96:565-89.
Kirmse R, Qin Z, Weinert CM, Hoenger A, Buehler MJ, Kreplak L. (2010) Plasticity of intermediate filament subunits. PLoS One. 5:e12115.
7: Technical advancements in cutting and handling frozen-hydrated cryo-sections:
Figure 7: Intracellular molecular details preserved in vitrified sections. A and B: 70 nm, frozen-hydrated cryo-section observed in a high-pressure frozen A31-3T3 cell. A: Approximately 10 nm thick tomographic slice that shows mitochondria, ribosomes, endosomes and an actin-coated microtubule. Inset: end-on view of such a coated microtubule, and side view of an actin bundle. B: An important advantage of CEMOVIS is illustrated in these triplet microtubules of the centriole where the technique allows for a direct determination of the microtubule polarity due to the characteristic slew of the protofilaments in end-on views (compare to helical averaging of a set of microtubules in the inset). C: Cross-sections of a microtubule next to vimentin intermediate filaments (IFs) in a vitrified section of CHO cells. All IFs show a clear central density, which is different from data obtained on in vitro polymerized vimentin IFs (see: Goldie et al., 2007). D: Frozen cell pellet in a dome-shaped carrier approaching the cryo-diamond knife for trimming. E: 100 micrometer wide, 50 nm thick vitrified sections are shown on the surface of a 45 degree cutting cryo-diamond knife ready to be collected on a carbon-coated, copper grid.
Our work on the improvement of vitrified sectioning (e.g. see Bouchet-Marquis 2011) has been a continuous effort in the lab, pursued initially by Mark Ladinsky (see Fig. 3B, C and E; see also Ladinsky et al., 2006) and more recently by Cedric Bouchet-Marquis who joined us from the Dubochet lab in Lausanne, Switzerland. While much of our progress has been incremental, we have innovated the use of a micromanipulator (Fig. 3C) for handling the ribbons formed by cryo-sections as they come off the knife. This has facilitated the accumulation of many serial cryo-sections, a useful advance for finding specific intracellular structures, such as a centrosome or a Golgi apparatus.
Vitrified sections are now produced on a regular basis. The microtome setup designed by Mark Ladinsky and Cedric Bouchet-Marquis consists of a commercial Leica FCS cryo-ultramicrotome supplemented with an ionizer device that charges the vitrified sections so they adsorb better to the grid surface, and a micromanipulator to handle the sections when they come of the knife. We have experimented with knives that exhibit different angles (25 degrees with cryo-platform, 35 with cryo-platform to 45 degrees included angle). While an angle of 45 degrees constitutes the standard for regular diamond knives, the smaller angles were examined with the anticipation that they might reduce the formation of crevasses. We are constantly experimenting with new tools to improve the quality of vitrified sections.
Hoog, J.L., Bouchet-Marquis, C., McIntosh, J.R., Hoenger, A., and Gull, K. (2012). Cryo-Electron Tomography and 3-D Analysis of the Intact Flagellum in Trypanosoma brucei. J. Struct. Biol. 178:189-98.
Bouchet-Marquis C, & Hoenger A. (2011) Cryo-electron tomography on vitrified sections: a critical analysis of benefits and limitations for structural cell biology. Micron. 42:152-62.
Hoenger A, & Bouchet-Marquis C. (2011) Cellular tomography. Adv Protein Chem Struct Biol. 82:67-90.
8: The Role of Septins in Glucose starved S. pombe cells.
Figure 8: Relationship between septins and cell-freezing A: Diffusion of cytoplasmic vesicle in wildtype and septin-2 deletion mutant. Substantial variations in diffusion are seen in wildtype, while the septin-2 deletion mutant remains less affected by starvation (unpublished data produced by E.L. Florin, Univ. of Texas at Austin). B: After five days of starvation septin-2 deletion mutants, expressing Septin-3-GFP show bundles of septin-3. The driving force behind these bundles is unclear, but they can easily be seen by cryo-electron microscopy and tomography (B) as well as fluorescence microscopy (see Fig. 3). The left panel shows a tomographic slice at 1.4 nm thickness, and at 35 nm on the right. While the left panel focuses on a very limited depth molecular detail are well visible. The right panel presents the superposition of almost half of a plastic-thin section and reveals more in-depth data, but with less molecular resolution.
Figure 9: Vitrified sections of yeast cells showing putative bundles of septin fibers (A, B) Septin bundles in 8 days starved cells lacking septin-2 in a septin-3-GFP background. C) Shows a wild type cell after eight days of glucose starvation as in figure 3 below. Panel D) is a RF-FSF preparation of a 3-day starved septin-2 deletion mutant (also expressing septin-3-GFP). E) Vitrified section of a 3T3 fibroblast cell showing large bundles of actin stress fibers, either in cross-section (below the label "actin stress fibers"), or at an oblique angle (the two bundles above and left of the label). In cross-section the filament bundles reveals a relatively regular, almost orthogonal packing pattern. Oblique cuts reveal straight filament bundles. Two microtubules in cross-section can be seen on upper left corner. At least when bundled in stress fibers actin filaments appear better ordered than the putative septin fiber bundles in A - D.
Septins are a group of highly conserved GTP binding proteins found in eukaryotes and are involved in many cellular processes, often regulated through their capability of forming filamentous and ring-like structures. Certain polymeric septin assemblies can be imaged by electron microscopy (see: Bertin et al., 2008; Garcia et al., 2011) within intact cells, which is the core technology of our lab. A previous collaboration between the labs of Prof. E.L. Florin, Prof. Damian Brunner and myself led to the discovery of an unexpected, striking behavior of glucose starved S. pombe cells: A "viscosity" increase of their cytoplasm that is reversible by adding glucose back to the cells. During cell stasis the observed motion of endogenous lipid granules and other structures became strongly restricted. Such a strong immobilization requires a dense network, and as all observable structures were affected in the same way, we hypothesize that this filamentous network is generated throughout the cell by septin polymerization. In yeast cells, septins build scaffolding to provide structural support during cell division and compartmentalize parts of the cell. S. pombe cells react to glucose starvation with a "freezing" process of the cytosol, which abolishes all intracellular movements and also resists actively moving vesicles with a laser trap. We produced preliminary evidence that this process is caused by septin polymerization, and the creation of a dense filamentous septin meshwork throughout the entire cytosol. This process is unrelated to mitotic polymer formation as seen during cytokinesis (Bertin et al., 2012). By now we have created several septin deletion mutants that show much reduced cell-freezing during starvation, while the viscosity of the cytosol remains low.
Bertin A, McMurray MA, Grob P, Park SS, Garcia G 3rd, Patanwala I, Ng HL, Alber T, Thorner J, Nogales E. (2008) Saccharomyces cerevisiae septins: supramolecular organization of hetero-oligomers and the mechanism of filament assembly. Proc Natl Acad Sci U S A. 105:8274-9.
Bertin A, McMurray MA, Pierson J, Thai L, McDonald KL, Zehr EA, Garcia G 3rd, Peters P, Thorner J, Nogales E. (2012) Three-dimensional ultrastructure of the septin filament network in Saccharomyces cerevisiae. Mol Biol Cell. 23:423-32.
Garcia G 3rd, Bertin A, Li Z, Song Y, McMurray MA, Thorner J, Nogales E. (2011) Subunit-dependent modulation of septin assembly: budding yeast septin Shs1 promotes ring and gauze formation. J Cell Biol. 195:993-1004.