An adage of biochemistry is “Never waste pure thoughts on an impure protein.” Starting from pure proteins, we can determine amino acid sequences and investigate biochemical functions. From the amino acid sequences, we can map evolutionary relationships between proteins in diverse organisms (Chapter 6). By using crystals grown from pure protein, we can obtain x-ray data that will provide us with a picture of the protein’s tertiary structure—the shape that determines function.

67

Purification should yield a sample containing only one type of molecule—the protein in which the biochemist is interested. This protein sample may be only a fraction of 1% of the starting material, whether that starting material consists of one type of cell in culture or a particular organ from a plant or animal. How is the biochemist able to isolate a particular protein from a complex mixture of proteins?

A protein can be purified by subjecting the impure mixture of the starting material to a series of separations based on physical properties such as size and charge. To monitor the success of this purification, the biochemist needs a test, called an assay, for some unique identifying property of the protein. A positive result on the assay indicates that the protein is present. Although assay development can be a challenging task, the more specific the assay, the more effective the purification. For enzymes, which are protein catalysts (Chapter 8), the assay usually measures enzyme activity—that is, the ability of the enzyme to promote a particular chemical reaction. This activity is often measured indirectly. Consider the enzyme lactate dehydrogenase, which catalyzes the following reaction in the synthesis of glucose:

Reduced nicotinamide adenine dinucleotide (NADH, Figure 15.13) absorbs light at 340 nm, whereas oxidized nicotinamide adenine dinucleotide (NAD⁺) does not. Consequently, we can follow the progress of the reaction by examining how much light-absorbing ability is developed by a sample in a given period of time—for instance, within 1 minute after the addition of the enzyme. Our assay for enzyme activity during the purification of lactate dehydrogenase is thus the increase in the absorbance of light at 340 nm observed in 1 minute.

To analyze how our purification scheme is working, we need one additional piece of information—the amount of protein present in the mixture being assayed. There are various rapid and reasonably accurate means of determining protein concentration. With these two experimentally determined numbers—enzyme activity and protein concentration—we then calculate the specific activity, the ratio of enzyme activity to the amount of protein in the mixture. Ideally, the specific activity will rise as the purification proceeds and the protein mixture will contain the protein of interest to a greater extent. In essence, the overall goal of the purification is to maximize the specific activity. For a pure enzyme, the specific activity will have a constant value.

Having found an assay and chosen a source of protein, we now fractionate the cell into components and determine which component is enriched in the protein of interest. In the first step, a homogenate is formed by disrupting the cell membrane, and the mixture is fractionated by centrifugation, yielding a dense pellet of heavy material at the bottom of the centrifuge tube and a lighter supernatant above (Figure 3.1). The supernatant is again centrifuged at a greater force to yield yet another pellet and supernatant. This procedure, called differential centrifugation, yields several fractions of decreasing density, each still containing hundreds of different proteins. The fractions are each separately assayed for the desired activity. Usually, one fraction will be enriched for such activity, and it then serves as the source of material to which more-discriminating purification techniques are applied.

68

Several thousand proteins have been purified in active form on the basis of such characteristics as solubility, size, charge, and specific binding affinity. Usually, protein mixtures are subjected to a series of separations, each based on a different property. At each step in the purification, the preparation is assayed and its specific activity is determined. A variety of purification techniques are available.

Most proteins are less soluble at high salt concentrations, an effect called salting out. The salt concentration at which a protein precipitates differs from one protein to another. Hence, salting out can be used to fractionate proteins. For example, 0.8 M ammonium sulfate precipitates fibrinogen, a blood-clotting protein, whereas a concentration of 2.4 M is needed to precipitate serum albumin. Salting out is also useful for concentrating dilute solutions of proteins, including active fractions obtained from other purification steps. Dialysis can be used to remove the salt if necessary.

69

Proteins can be separated from small molecules such as salt by dialysis through a semipermeable membrane, such as a cellulose membrane with pores (Figure 3.2). The protein mixture is placed inside the dialysis bag, which is then submerged in a buffer solution that is devoid of the small molecules to be separated away. Molecules having dimensions significantly greater than the pore diameter are retained inside the dialysis bag. Smaller molecules and ions capable of passing through the pores of the membrane diffuse down their concentration gradients and emerge in the solution outside the bag. This technique is useful for removing a salt or other small molecule from a cell fractionate, but it will not distinguish between proteins effectively.

More-discriminating separations on the basis of size can be achieved by the technique of gel-filtration chromatography, also known as molecular exclusion chromatography (Figure 3.3). The sample is applied to the top of a column consisting of porous beads made of an insoluble but highly hydrated polymer such as dextran or agarose (which are carbohydrates) or polyacrylamide. Sephadex, Sepharose, and Biogel are commonly used commercial preparations of these beads, which are typically 100 μm (0.1 mm) in diameter. Small molecules can enter these beads, but large ones cannot. The result is that small molecules are distributed in the aqueous solution both inside the beads and between them, whereas large molecules are located only in the solution between the beads. Large molecules flow more rapidly through this column and emerge first because a smaller volume is accessible to them. Molecules of medium size occasionally enter the beads and will flow from the column at an intermediate position, while small molecules, which take a longer, tortuous path, will exit last.

70

To obtain a protein of high purity, one chromatography step is usually not sufficient, because other proteins in the crude mixture will likely co-elute with the desired material. Additional purity can be achieved by performing sequential separations that are based on distinct molecular properties. For example, in addition to size, proteins can be separated on the basis of their net charge by ion-exchange chromatography. If a protein has a net positive charge at pH 7, it will usually bind to a column of beads containing carboxylate groups, whereas a negatively charged protein will not (Figure 3.4). The bound protein can then be eluted (released) by increasing the concentration of sodium chloride or another salt in the eluting buffer; sodium ions compete with positively charged groups on the protein for binding to the column. Proteins that have a low density of net positive charge will tend to emerge first, followed by those having a higher charge density. This procedure is also referred to as cation exchange to indicate that positively charged groups will bind to the anionic beads. Positively charged proteins (cationic proteins) can be separated by chromatography on negatively charged carboxymethylcellulose (CM-cellulose) columns. Conversely, negatively charged proteins (anionic proteins) can be separated by anion exchange on positively charged diethylaminoethylcellulose (DEAE-cellulose) columns.

Affinity chromatography is another powerful means of purifying proteins that is highly selective for the protein of interest. This technique takes advantage of the high affinity of many proteins for specific chemical groups. For example, the plant protein concanavalin A is a carbohydrate-binding protein, or lectin (Section 11.4), that has affinity for glucose. When a crude extract is passed through a column of beads containing covalently attached glucose residues, concanavalin A binds to the beads, whereas most other proteins do not (Figure 3.5). The bound concanavalin A can then be released from the column by adding a concentrated solution of glucose. The glucose in solution displaces the column-attached glucose residues from binding sites on concanavalin A. Affinity chromatography is a powerful means of isolating transcription factors—proteins that regulate gene expression by binding to specific DNA sequences. A protein mixture is passed through a column containing specific DNA sequences attached to a matrix; proteins with a high affinity for the sequence will bind and be retained. In this instance, the transcription factor is released by washing with a solution containing a high concentration of salt.

In general, affinity chromatography can be effectively used to isolate a protein that recognizes group X by (1) covalently attaching X or a derivative of it to a column; (2) adding a mixture of proteins to this column, which is then washed with buffer to remove unbound proteins; and (3) eluting the desired protein by adding a high concentration of a soluble form of X or altering the conditions to decrease binding affinity. Affinity chromatography is most effective when the interaction of the protein and the molecule that is used as the bait is highly specific.

Affinity chromatography can be used to isolate proteins expressed from cloned genes (Section 5.2). Extra amino acids are encoded in the cloned gene that, when expressed, serve as an affinity tag that can be readily trapped. For example, repeats of the codon for histidine may be added such that the expressed protein has a string of histidine residues (called a His tag) on one end. The tagged proteins are then passed through a column of beads containing covalently attached, immobilized nickel(II) or other metal ions. The His tags bind tightly to the immobilized metal ions, binding the desired protein, while other proteins flow through the column. The protein can then be eluted from the column by the addition of imidazole or some other chemical that binds to the metal ions and displaces the protein.

71

A technique called high-performance liquid chromatography (HPLC) is an enhanced version of the column techniques already discussed. The column materials are much more finely divided and, as a consequence, possess more interaction sites and thus greater resolving power. Because the column is made of finer material, pressure must be applied to the column to obtain adequate flow rates. The net result is both high resolution and rapid separation. In a typical HPLC setup, a detector that monitors the absorbance of the eluate at a particular wavelength is placed immediately after the column. In the sample HPLC elution profile shown in Figure 3.6, proteins are detected by setting the detector to 220 nm (the characteristic absorbance wavelength of the peptide bond). In a short span of 10 minutes, a number of sharp peaks representing individual proteins can be readily identified.

How can we tell that a purification scheme is effective? One way is to ascertain that the specific activity rises with each purification step. Another is to determine that the number of different proteins in each sample declines at each step. The technique of electrophoresis makes the latter method possible.

A molecule with a net charge will move in an electric field. This phenomenon, termed electrophoresis, offers a powerful means of separating proteins and other macromolecules, such as DNA and RNA. The velocity of migration (v) of a protein (or any molecule) in an electric field depends on the electric field strength (E), the net charge on the protein (z), and the frictional coefficient (f).

The electric force Ez driving the charged molecule toward the oppositely charged electrode is opposed by the viscous drag fv arising from friction between the moving molecule and the medium. The frictional coefficient f depends on both the mass and shape of the migrating molecule and the viscosity (η) of the medium. For a sphere of radius r,

Electrophoretic separations are nearly always carried out in porous gels (or on solid supports such as paper) because the gel serves as a molecular sieve that enhances separation (Figure 3.7). Molecules that are small compared with the pores in the gel readily move through the gel, whereas molecules much larger than the pores are almost immobile. Intermediate-size molecules move through the gel with various degrees of facility. The electric field is applied such that proteins migrate from the negative to the positive electrodes, typically from top to bottom. Electrophoresis is performed in a thin, vertical slab of polyacrylamide gel. Polyacrylamide gels are choice supporting media for electrophoresis because they are chemically inert and readily formed by the polymerization of acrylamide with a small amount of the cross-linking agent methylenebisacrylamide to make a three-dimensional mesh (Figure 3.8). Electrophoresis is distinct from gel filtration in that, because of the electric field, all of the molecules, regardless of size, are forced to move through the same matrix.

72

Proteins can be separated largely on the basis of mass by electrophoresis in a polyacrylamide gel under denaturing conditions. The mixture of proteins is first dissolved in a solution of sodium dodecyl sulfate (SDS), an anionic detergent that disrupts nearly all noncovalent interactions in native proteins. β-Mercaptoethanol (2-thioethanol) or dithiothreitol is added to reduce disulfide bonds. Anions of SDS bind to main chains at a ratio of about one SDS anion for every two amino acid residues. The negative charge acquired on binding SDS is usually much greater than the charge on the native protein; the contribution of the protein to the total charge of the SDS–protein complex is thus rendered insignificant. As a result, this complex of SDS with a denatured protein has a large net negative charge that is roughly proportional to the mass of the protein. The SDS–protein complexes are then subjected to electrophoresis. When the electrophoresis is complete, the proteins in the gel can be visualized by staining them with silver nitrate or a dye such as Coomassie blue, which reveals a series of bands (Figure 3.9). Radioactive labels, if they have been incorporated into proteins, can be detected by placing a sheet of x-ray film over the gel, a procedure called autoradiography.

73

Small proteins move rapidly through the gel, whereas large proteins stay at the top, near the point of application of the mixture. The mobility of most polypeptide chains under these conditions is linearly proportional to the logarithm of their mass (Figure 3.10). Some carbohydrate-rich proteins and membrane proteins do not obey this empirical relation, however. SDS– polyacrylamide gel electrophoresis (often referred to as SDS-PAGE) is rapid, sensitive, and capable of a high degree of resolution. As little as 0.1 μg (~2 pmol) of a protein gives a distinct band when stained with Coomassie blue, and even less (~0.02 μg) can be detected with a silver stain. Proteins that differ in mass by about 2% (e.g., 50 and 51 kDa, arising from a difference of about 10 amino acids) can usually be distinguished with SDS-PAGE.

We can examine the efficacy of our purification scheme by analyzing a part of each fraction by electrophoresis. The initial fractions will display dozens to hundreds of proteins. As the purification progresses, the number of bands will diminish, and the prominence of one of the bands should increase. This band should correspond to the protein of interest.

Proteins can also be separated electrophoretically on the basis of their relative contents of acidic and basic residues. The isoelectric point (pI) of a protein is the pH at which its net charge is zero. At this pH, its electrophoretic mobility is zero because z in equation 1 is equal to zero. For example, the pI of cytochrome c, a highly basic electron-transport protein, is 10.6, whereas that of serum albumin, an acidic protein in blood, is 4.8. Suppose that a mixture of proteins undergoes electrophoresis in a pH gradient in a gel in the absence of SDS. Each protein will move until it reaches a position in the gel at which the pH is equal to the pI of the protein. This method of separating proteins according to their isoelectric point is called isoelectric focusing. The pH gradient in the gel is formed first by subjecting a mixture of polyampholytes (small multi-charged polymers) having many different pI values to electrophoresis. Isoelectric focusing can readily resolve proteins that differ in pI by as little as 0.01, which means that proteins differing by one net charge can be separated (Figure 3.11).

74

Isoelectric focusing can be combined with SDS-PAGE to obtain very high resolution separations. A single sample is first subjected to isoelectric focusing. This single-lane gel is then placed horizontally on top of an SDS–polyacrylamide slab. The proteins are thus spread across the top of the polyacrylamide gel according to how far they migrated during isoelectric focusing. They then undergo electrophoresis again in a perpendicular direction (vertically) to yield a two-dimensional pattern of spots. In such a gel, proteins have been separated in the horizontal direction on the basis of isoelectric point and in the vertical direction on the basis of mass. Remarkably, more than a thousand different proteins in the bacterium Escherichia coli can be resolved in a single experiment by two-dimensional electrophoresis (Figure 3.12).

Proteins isolated from cells under different physiological conditions can be subjected to two-dimensional electrophoresis. The intensities of individual spots on the gels can then be compared, which indicates that the concentrations of specific proteins have changed in response to the physiological state (Figure 3.13). How can we discover the identity of a protein that is showing such responses? Although many proteins are displayed on a two-dimensional gel, they are not identified. It is now possible to identify proteins by coupling two-dimensional gel electrophoresis with mass spectrometric techniques. We will examine these powerful techniques shortly (Section 3.3).

75

To determine the success of a protein purification scheme, we monitor each step of the procedure by determining the specific activity of the protein mixture and by subjecting it to SDS-PAGE analysis. Consider the results for the purification of a fictitious protein, summarized in Table 3.1 and Figure 3.14. At each step, the following parameters are measured:

Total Protein. The quantity of protein present in a fraction is obtained by determining the protein concentration of a part of each fraction and multiplying by the fraction’s total volume.

Total Activity. The enzyme activity for the fraction is obtained by measuring the enzyme activity in the volume of fraction used in the assay and multiplying by the fraction’s total volume.

Specific Activity. This parameter is obtained by dividing total activity by total protein.

Yield. This parameter is a measure of the activity retained after each purification step as a percentage of the activity in the crude extract. The amount of activity in the initial extract is taken to be 100%.

Purification Level. This parameter is a measure of the increase in purity and is obtained by dividing the specific activity, calculated after each purification step, by the specific activity of the initial extract.

As we see in Table 3.1, the first purification step, salt fractionation, leads to an increase in purity of only 3-fold, but we recover nearly all the target protein in the original extract, given that the yield is 92%. After dialysis to lower the high concentration of salt remaining from the salt fractionation, the fraction is passed through an ion-exchange column. The purification now increases to 9-fold compared with the original extract, whereas the yield falls to 77%. Gel-filtration chromatography brings the level of purification to 110-fold, but the yield is now at 50%. The final step is affinity chromatography with the use of a ligand specific for the target enzyme. This step, the most powerful of these purification procedures, results in a purification level of 3000-fold but lowers the yield to 35%. The SDS-PAGE analysis in Figure 3.14 shows that, if we load a constant amount of protein onto each lane after each step, the number of bands decreases in proportion to the level of purification, and the amount of protein of interest increases as a proportion of the total protein present.

76

A good purification scheme takes into account both purification levels and yield. A high degree of purification and a poor yield leave little protein with which to experiment. A high yield with low purification leaves many contaminants (proteins other than the one of interest) in the fraction and complicates the interpretation of subsequent experiments.

We have already seen that centrifugation is a powerful and generally applicable method for separating a crude mixture of cell components. This technique is also valuable for the analysis of the physical properties of biomolecules. Using centrifugation, we can determine such parameters as mass and density, learn about the shape of a molecule, and investigate the interactions between molecules. To deduce these properties from the centrifugation data, we require a mathematical description of how a particle behaves when a centrifugal force is applied.

A particle will move through a liquid medium when subjected to a centrifugal force. A convenient means of quantifying the rate of movement is to calculate the sedimentation coefficient, s, of a particle by using the following equation:

where m is the mass of the particle, is the partial specific volume (the reciprocal of the particle density), ρ is the density of the medium, and f is the frictional coefficient (a measure of the shape of the particle). The term is the buoyant force exerted by liquid medium.

Sedimentation coefficients are usually expressed in Svedberg units (S), equal to 10⁻¹³ s. The smaller the S value, the more slowly a molecule moves in a centrifugal field. The S values for a number of biomolecules and cellular components are listed in Table 3.2 and Figure 3.15.

Several important conclusions can be drawn from the preceding equation:

1. The sedimentation velocity of a particle depends in part on its mass. A more massive particle sediments more rapidly than does a less massive particle of the same shape and density.

2. Shape, too, influences the sedimentation velocity because it affects the viscous drag. The frictional coefficient f of a compact particle is smaller than that of an extended particle of the same mass. Hence, elongated particles sediment more slowly than do spherical ones of the same mass.

77

3. A dense particle moves more rapidly than does a less dense one because the opposing buoyant force is smaller for the denser particle.

4. The sedimentation velocity also depends on the density of the solution (ρ). Particles sink when , float when , and do not move when .

A technique called zonal, band, or most commonly gradient centrifugation can be used to separate proteins with different sedimentation coefficients. The first step is to form a density gradient in a centrifuge tube. Differing proportions of a low-density solution (such as 5% sucrose) and a high-density solution (such as 20% sucrose) are mixed to create a linear gradient of sucrose concentration ranging from 20% at the bottom of the tube to 5% at the top (Figure 3.16). The role of the gradient is to prevent convective flow. A small volume of a solution containing the mixture of proteins to be separated is placed on top of the density gradient. When the rotor is spun, proteins move through the gradient and separate according to their sedimentation coefficients. The time and speed of the centrifugation is determined empirically. The separated bands, or zones, of protein can be harvested by making a hole in the bottom of the tube and collecting drops. The drops can be measured for protein content and catalytic activity or another functional property. This sedimentation-velocity technique readily separates proteins differing in sedimentation coefficient by a factor of two or more.

78

The mass of a protein can be directly determined by sedimentation equilibrium, in which a sample is centrifuged at low speed such that a concentration gradient of the sample is formed. However, this sedimentation is counterbalanced by the diffusion of the sample from regions of high to low concentration. When equilibrium has been achieved, the shape of the final gradient depends solely on the mass of the sample. The sedimentation-equilibrium technique for determining mass is very accurate and can be applied without denaturing the protein. Thus the native quaternary structure of multimeric proteins is preserved. In contrast, SDS–polyacrylamide gel electrophoresis provides an estimate of the mass of dissociated polypeptide chains under denaturing conditions. Note that, if we know the mass of the dissociated components of a multimeric protein as determined by SDS–polyacrylamide analysis and the mass of the intact multimer as determined by sedimentation-equilibrium analysis, we can determine the number of copies of each polypeptide chain present in the protein complex.

In Chapter 5, we shall consider the widespread effect of recombinant DNA technology on all areas of biochemistry and molecular biology. The application of recombinant methods to the overproduction of proteins has enabled dramatic advances in our understanding of their structure and function. Before the advent of this technology, proteins were isolated solely from their native sources, often requiring a large amount of tissue to obtain a sufficient amount of protein for analytical study. For example, the purification of bovine deoxyribonuclease in 1946 required nearly ten pounds of beef pancreas to yield one gram of protein. As a result, biochemical studies on purified material were often limited to abundant proteins.

Armed with the tools of recombinant technology, the biochemist is now able to enjoy a number of significant advantages:

1. Proteins can be expressed in large quantities. The homogenate serves as the starting point in a protein purification scheme. For recombinant systems, a host organism that is amenable to genetic manipulation, such as the bacterium Escherichia coli or the yeast Pichia pastoris, is utilized to express a protein of interest. The biochemist can exploit the short doubling times and ease of genetic manipulation of such organisms to produce large amounts of protein from manageable amounts of culture. As a result, purification can begin with a homogenate that is often highly enriched with the desired molecule. Moreover, a protein can be easily obtained regardless of its natural abundance or its species of origin.

2. Affinity tags can be fused to proteins. As described above, affinity chromatography can be a highly selective step within a protein purification scheme. Recombinant DNA technology enables the attachment of any one of a number of possible affinity tags to a protein (such as the “His tag” mentioned earlier). Hence, the benefits of affinity chromatography can be realized even for those proteins for which a binding partner is unknown or not easily determined.

3. Proteins with modified primary structures can be readily generated. A powerful aspect of recombinant DNA technology as applied to protein purification is the ability to manipulate genes to generate variants of a native protein sequence (Section 5.2). We learned in Section 2.4 that many proteins consist of compact domains connected by flexible linker regions. With the use of genetic-manipulation strategies, fragments of a protein that encompass single domains can be generated, an advantageous approach when expression of the entire protein is limited by its size or solubility. Additionally, as we will see in Section 9.1, amino acid substitutions can be introduced into the active site of an enzyme to precisely probe the roles of specific residues within its catalytic cycle.

79