SizeandShapeofProteinMolecules4

Determining the Molecular Weight of a Protein Molecule—Combining S and R _s à la Siegel and Monte

With the completion of multiple genomes and increasingly good annotation, the primary sequence of almost any protein can be found in the databases. The molecular weight of every protein subunit is therefore known from its sequence. But an experimental measure is still needed to determine if the native protein in solution is a monomer, dimer, or oligomer, or if it forms a complex with other proteins. If one has a purified protein, the molecular weight can be determined quite accurately by sedimentation equilibrium in the analytical ultracentrifuge. This technique has made a strong comeback with the introduction of the Beckman XL-A analytical ultracentrifuge. There are a number of good reviews (14, 15), and the documentation and programs that come with the centrifuge are very instructive.

What if one does not have an XL-A centrifuge or the protein of interest is not purified? In 1966, Siegel and Monte (4) proposed a method that achieves the results of sedimentation equilibrium, with two enormous advantages. First, it requires only a preparative ultracentrifuge for sucrose or glycerol gradient sedimentation and a gel filtration column. This equipment is available in most biochemistry laboratories. Second, the protein of interest need not be purified; one needs only an activity or an antibody to locate it in the fractions. This is a very powerful technique and should be in the repertoire of every protein biochemist.

The methodology is very simple. The protein is run over a calibrated gel filtration column to determine R _sand hence f. Separately, the protein is centrifuged through a glycerol or sucrose gradient to determine S. One then uses the Svedberg equation (Eq. 4.1 ) to obtain M as a function of R _s and S.

setting η = 0.01, v ₂ ρ = 0.73, converting S to Svedberg units and R _s to nanometer, we can simplify further:

where S is in Svedberg units, R _s is in nanometer, and M is in Daltons.

This is pretty simple! Importantly, in typical applications, this method gives the protein mass within about ±10%. This is more than enough precision to distinguish between monomer, dimer, or trimer.

Electron Microscopy of Protein Molecules

Since the early 1980s, electron microscopy has become a powerful technique for determining the size and shape of single protein molecules, especially ones larger than 100 kDa. Two techniques available in most EM laboratories, rotary shadowing and negative stain, can be used for imaging single molecules. Cryo-EM is becoming a powerful tool for protein structural analysis, but it requires special equipment and expertise. For a large number of applications, rotary shadowing and negative stain provide the essential structural information.

For rotary shadowing, a dilute solution of protein is sprayed on mica, the liquid is evaporated in a high vacuum, and platinum metal is evaporated onto the mica at a shallow angle. The mica is rotated during this process, so the platinum builds up on all sides of the protein molecules. The first EM images of single protein molecules were obtained by Hall and Slayter using rotary shadowing (16). Their images of fibrinogen showed a distinctive trinodular rod. However, rotary shadowing fell into disfavor because the images were difficult to reproduce. Protein tended to aggregate and collect salt, rather than spread as single molecules. In 1976, James Pullman, a graduate student at the University of Chicago, then devised a protocol with one simple but crucial modification—he added 30% glycerol to the protein solution. For reasons that are still not understood, the glycerol greatly helps the spreading of the protein as single molecules.

Pullman never published his protocol, but two labs saw his mimeographed notes and tested out the effect of glycerol, as a part of their own attempts to improve rotary shadowing (17, 18). They obtained reproducible and compelling images of fibrinogen (the first since the original Hall and Slayter study and confirming the trinodular rod structure) and spectrin (the first ever images of this large protein). The technique has since been used in characterizing hundreds of protein molecules.

Figure 4 shows rotary shadowed SMC protein from B. subtilis, fibrinogen, and hexabrachion (tenascin). SMC protein from B. subtilis is highly elongated, consistent with its high S _max/S discussed above (19). The fibrinogen molecules show the trinodular rod, but these images also resolved a small fourth nodule next to the central nodule (20), not seen in earlier studies. The central nodule is about 50 kDa, and the smaller fourth nodule is about 20 kDa. The “hexabrachion” tenascin molecule (21) illustrates the power of rotary shadowing at two extremes. First, the molecule is huge. Each of its six arms is made up of ∼30 repeating small domains, totaling ∼200,000 Da. At the larger scale, the EM shows that each arm is an extended structure, matching the length expected if the repeating domains are an extended string of beads. At the finer scale, the EM can distinguish the different sized domains. The inner segment of each arm is a string of 3.5-kDa epidermal growth factor domains, seen here as a thinner segment. A string of 10-kDa FN-III domains is clearly distinguished as a thicker outer segment. The terminal knob is a single 22-kDa fibrinogen domain. The R _min of these domains are 0.8, 1.7, and 2.8 nm, and these can be distinguished by rotary shadowing. Rotary shadowing EM can visualize single globular domains as small as 10 kDa (3.5 nm diameter) and elongated molecules as thin as 1.5 nm (collagen).

Fig. 4 Rotary shadowing EM of three highly elongated protein molecules: the SMC protein from B. subtilis (19), fibrinogen (20), and the hexabrachion protein, tenascin (21).

Negative stain is another EM technique capable of imaging single protein molecules. It is especially useful for imaging larger molecules with a complex internal structure, which appear only as a large blob in rotary shadowing. Importantly, noncovalent protein–protein bonds are sometimes disrupted in the rotary shadowing technique (8), but uranyl acetate, in addition to providing high resolution contrast, fixes oligomeric protein structures in a few milliseconds (22). An excellent review of modern techniques of negative staining, with comparison to cryo-EM, is given in (23).

The simple picture of the molecule produced by EM is frequently the most straightforward and satisfying structural analysis at the 1–2-nm resolution. When the structure is confirmed by hydrodynamic analysis, the interpretation is even more compelling.