The goal of the NYSBC electron microscopy facility is to help researchers elucidate the intermolecular interactions and domain architectures of macromolecules within their native cellular assemblies. Towards this goal, the facility has brought together a combination of instrumentation and staff expertise that supports the determination three-dimensional structures using the major techniques available to the field.
Beginning an electron microscopy project
Usually, a project aimed to determine structural information of a given material is ripe for study when a high degree of purity has been achieved. High purity means PAGE gels show the band(s) corresponding to the material that will be studied, with little or no traces of contaminants.
Biochemically homogeneous sample
If your sample is a macromolecular complex, such as a protein or a nucleic acid/protein complex, that i) has a defined composition, ii) can be isolated in multiple, identical copies, and iii) can be purified to homogeneity, electron microscopy may be used for structure elucidation. The size and conformational properties of your complex will determine which technique(s) could be used.
Flowchart 1: Determining the EM technique(s) to use
How big is your protein?
At this stage, if the molecule is large enough (more than ~300kDa), one can attempt to visualize it in the electron microscope, and thus make an initial assessment of the samples. This is done usually preparing samples with heavy metal salts, generally known as negative stains. In these samples, one can get usually an adequate idea of the steps to follow, and if the molecules are deemed to be structurally homogeneous, several steps can be taken to proceed further. If the macromolecular complex of interest is smaller than 200kDa the techniques of electron crystallography or helical reconstruction are ideally suited.
If the complex of interest is larger than 300kDa single particle analysis or electron tomography (see below) can be used for structure determination without requiring crystallization.
If the purified sample consists of a single conformation of the macromolecular complex of interest then crystallization attempts should be made. 2D crystallography is particularly effective for membrane proteins, and helical symmetry is adopted by a variety of proteins in vivo. Examples of macromolecules studied by 2D crystallography are bacteriorhodopsin, aquaporin, and light harvesting complex. Macromolecules studied using helical reconstruction are actin/myosin, tubulin/kinesin, bacterial flagella, and acetyl choline receptor. Using 2D crystals or ordered helical arrays electron crystallographyor helical reconstruction can yield potentially high resolution structures.
If the complex does not crystallize and is smaller than 150nm single particle analysis is the method of choice for structure determination. Otherwise, electron tomography may be used.
Single Particles, Ordered Arrays or Tissue?
If the sample to be studied is of a relatively low molecular weight, a way to improve the chances of success is to produce ordered arrays, such as two dimensional crystals or helical tubes. Then image processing can be used to obtain the average structure of the molecules, in some cases to near atomic resolution. If the sample is homogeneous but cannot be crystallized, then it can be convenient to concentrate on one of its large, oligomeric forms and proceed with single particle analysis. In this case, if the sample presents one view more prominently than any other, a three-dimensional reconstruction can be determined using random conical tilt reconstructions, which is a procedure included in single particle analysis, and which provides a relatively fast and reliable method to compute an initial three-dimensional model of the sample. For molecules smaller than ~200kDa, it is difficult to extract data from single particle analysis, but this should be ruled out on an individual basis.
In the case of completely inhomogeneous, large samples, such as retroviruses or tissue, structural information can be collected using electron tomography. In this case, the sample is tilted in small angular increments and the individual views of the same area are used to form a three-dimensional volume. Due to radiation damage, as well as to the harsh preparative techniques required for tissue samples, the structural information from tomography is rather coarse.
Single particle analysis is ideally suited for structure determination of biochemically purified samples displaying a small degree of structural and conformational heterogeneity (2-5 discrete species). However, sorting out this heterogeneity represents the cutting edge in the field. Complexes whose structures have been determined using single particle analysis include GroEL, the proteosome, and the ribosome in various functional states.
Electron tomography can be useful in characterizing structural heterogeneity, such as in the case of the nuclear pore complex, albeit yielding lower resolution information.
Biochemically unique/heterogeneous (pleiomorphic) sample
If you are studying an organelle, a macromolecular assembly in situ , or if your sample is biochemically or structurally heterogeneous, then electron tomography is the preferred technique for structure determination. The golgi apparatus, mitochondria, desmosomes, and the distribution of glyocoproteins on the surface of HIV have been studied using electron tomography .
How to prepare sample?
In order to visualize a sample in a transmission electron microscope the sample must (a) be thin enough such that a beam of electrons can penetrate it (<250nm, ideally <100nm), (b) be deposited onto an EM grid, which is a thin circular copper grid that is 3mm in diameter, and (c)withstand high vacuum and electron radiation within the microscope column. For tissue, samples are prepared by cutting thin sections (Sectioning). For aqueous suspensions of macromolecules, including 2D crystals and ordered helical arrays, 1-5 microliters of the solution is pipetted onto the EM grid, which is then subjected to either negative staining, plunge freezing, or a combination of these sample preservation techniques, cryo-negative staining.
Negative Stain or Frozen-hydrated?
As informative as stained samples might be, it is desirable eventually to collect data from specimens frozen in their native buffers. To this end, once the ideal conditions for negative staining have been determined, we can freeze solutions of sample for imaging. These samples are visualized at low temperatures, to preserve the samples in a vitrified state, and to provide them some protection from the noxious effects of the electron bombardment.
Negative staining involves the addition of a heavy metal salt solution that forms an electron-dense mould around individual macromolecular complexes. Normally, this mould is formed by simply air drying the EM grid. The resulting samples are easy to manipulate and can be stored for long periods. In the electron microscope, this mould produces a high contrast image and is resistant to radiation damage. Thus, negative staining is the preferred method for screening samples and can also be used for low-resolution structure determination of the molecular envelope.
Plunge freezing results in a sample that is preserved in a physiological buffer, thus preserving not only its native conformation but also high resolution structural information. The trade offs involve severe radiation sensitivity and substantially lower contrast of the native biological material as well as the technical demands of handling frozen samples and ensuring their mechanical stability while in the microscope. For this technique, samples are pipetted onto an EM grid and, after blotting away excess solution, the grid is then plunged into liquid ethane. This procedure vitrifies the aqueous solvent, thus preserving the hydrogen-bonding networks that normally surround a macromolecule in liquid water. However, freezing must be rapid enough to prevent ice crystal formation, which will displace these hydrogen bonds and produce severe physical damage as the ice crystals push on neighboring biological material.
Negative staining vs. Plunge freezing
While preserving a sample in an unstained frozen-hydrated state is desirable, the use of negative staining is more practical at the beginning of new projects when dealing with small macromolecular complexes. Screening of samples prepared under different conditions is considerably faster by negative stain and the resulting high-contrast images are easier to evaluate. Determination of a refinement reference for single particle analysis is much more reliable due to higher signal-to-noise ration from negatively stained samples. Cryo-negative staining represents a compromise that may be useful in some situations.
A high concentration of negative stain is added to the sample, which is then plunge frozen in liquid ethane in the fully hydrated state. This technique prevents the dehydration and specimen collapse associated with conventional negative staining and increases the contrast of macromolecular complexes relative to conventional plunge freezing. Some complexes for which this method has been used include TFIIE, RNA polymerase II, and GroEL.
Biological specimens, such as tissue, cultured cells, or organelles, which are thicker than 250 nm must be sectioned prior to electron microscopy. The challenge is to preserve the structural integrity of macromolecular complexes within these sections. Conventional techniques employ chemical fixation, staining, dehydration and embedding in polymer resins prior to cutting sections with an ultramicrotome. The conventional protocols involve harsh treatments that exract substantial biological materials and fail to preserve details finer than ~20 nm. Cryogenic methods offer substantially better preservation. Although plunge-freezing is limited to samples <5-10 micrometers in thickness, high-pressure freezing is suitable for samples up to 100 micrometers in thickness. After freezing, the sample can either be directly sectioned (cryo-ultramicrotomy) for visualization in the frozen, unstained state, or subjected to freeze substitution and resin infiltration followed by conventional ultramicrotomy.