Strategy

Initial considerations

Any cell, whether prokaryotic or eukaryotic, contains tens of thousands of different proteins with a wide range of biological activities. Purification of individual proteins is therefore central to any detailed biological analysis of their structure and function. There is no single procedure whereby any and every protein can be isolated in pure form, but rather a set of procedures each of which attempts to distinguish between proteins on the basis of specific structural or functional characteristics. By selecting appropriate procedures and applying these in a carefully considered sequence of steps, it is usually possible to obtain a high degree of purification of the desired protein and in an acceptable yield. There is no standard order in which these procedures should be used to obtain optimal purifications. Because of structural and functional differences between proteins, an ideal sequence of steps for one protein will, quite possibly, be unsuccessful for another. A knowledge of the theoretical basis of each procedure will allow the researcher to choose an initial sequence of techniques with which to attempt any given purification. However, the development of an optimised protocol involves considerable trial-and-error experimentation to assess the potential of each step.

Irrespective of the purification scheme adopted, it is essential to realize that most proteins are quite fragile molecules. Exposure to even moderate temperatures (e.g. 37ºC) causes slow denaturation so that all purification steps should be carried out at low temperature (0 - 4ºC) unless specified otherwise. Proteins are also sensitive to extremes of pH so that purification procedures are always carried out in buffered solutions and pH extremes are avoided. Finally, since many proteins are more unstable in dilute solution, the protein concentration should not be allowed to fall to unreasonably low levels.

Although this program simulates the purification of proteins from solution, the starting point for protein purification is usually a culture of bacterial, plant or animal cells or a particular animal tissue or organ. The initial requirement is to release the protein from the cells without causing loss of its biological activity. A widely used gentle procedure for liberating proteins from animal cells is homogenization. Bacteria, yeasts and plant cells have thick cell walls which require the use of more disruptive procedures such as sonication (sonic radiation) or the use of a high-pressure press. A major advantage of homogenization, if it can be used, is that the integrity of most intracellular organelles is preserved. Thus, if the desired protein is restricted to a particular organelle such as the nucleus, mitochondrion or lysosome, or is perhaps bound to the plasma membrane, substantial purification can be achieved by isolating that cellular fraction before releasing the protein. Common procedures to achieve this are differential centrifugation and density-gradient centrifugation. If the desired protein is located in soluble form inside the purified organelle, it can now be released by disrupting the organelle either using sonication or the addition of non-ionic detergent such as Triton X-100. If the protein is an integral part of the organelle membrane or plasma membrane, it can be released by the use of detergent. Once the protein is available in soluble form it can be purified by using any of the standard techniques described in this Help file and simulated in this computer package.

There are numerous protein purification procedures available, some of which are much more commonly used than others. The optimal purification scheme for a particular protein depends mainly upon the structural and functional properties of the desired protein and of the protein contaminants from which it is to be purified. Indicated below are some general guidelines in the selection of techniques but it must be remembered that the precise conditions for the application of those methods are largely defined by trial-and-error experimentation.

(i) The early steps in a purification scheme are usually those which have the lowest resolving power such as ammonium sulfate precipitation or, for some proteins, heat denaturation. These procedures have the advantage of greatly reducing the total amount of protein in the sample so that chromatographic procedures, which are theoretically capable of high resolution, can then be applied to maximum advantage. Very high-resolution but low-capacity procedures such as isoelectric focusing are often applied only late in a purification scheme when the total amount of protein is relatively small and contaminating proteins are present in quite low concentrations. This type of high-resolution procedure, usually conducted on a fairly small scale, is therefore intended to remove the remaining few impurities.

(ii) It is typically more efficient to choose a series of methods which exploit different physical or functional properties of the proteins rather than a series of steps exploiting the same property. Thus, purification schemes employing column chromatography, usually include at least one size-fractionation procedure (gel filtration) and at least one charge-fractionation step (ion-exchange chromatography).

(iii) During chromatography, it is common practice to sacrifice some yield for increased purity. In other words, do not try to recover every last microgram of desired protein if this means including fractions containing relatively large amounts of protein contaminants. The fractions to be pooled ready for the next purification step should be chosen on the basis of highest possibly purity at reasonable yield.

(iv) Some protein fractionation techniques, in particular SDS-polyacrylamide gel electrophoresis (SDS-PAGE), isoelectric focusing and two-dimensional gel electrophoresis, are not usually employed as preparative procedures for protein purification but rather are used to check the progress of purification. They are high-resolution analytical techniques. However, increasingly in modern biochemistry it may be sufficient to purify only a few micrograms of the protein of interest for analysis. In such situations, these 'analytical' techniques may yield enough protein for research purposes and therefore be considered preparative.

(v) Another important consideration when choosing a fractionation technique is its cost. Both the materials used and the staff time consumed have to be considered. Some techniques are relatively cheap and quick, such as heat treatment or ammonium sulfate fractionation, while others tend to be expensive and time-consuming. Particularly expensive are those techniques that consume ampholytes, such as preparative isoelectric focusing. Therefore, while several techniques seem at first sight to be very attractive in terms of the high resolution that they offer, generally these advantages tend to be offset by their cost when applied to large samples. Indeed, equally good results can often be obtained more economically by the combination of several less expensive techniques.

Each purification procedure subdivides the protein mixture into a number of fractions one or more of which will contain the protein of interest whilst other fractions will contain proteins which are not required and so are discarded. Clearly, to determine which fractions are to be kept one needs to be able to detect and quantify the protein of interest. If this protein is an enzyme, the requirement is for a specific and sensitive assay for its catalytic activity. If the protein is a hormone, one may be able to use a hormone-binding assay, radioimmunoassay or even a bioassay. Alternatively, the protein may contain a characteristic moiety, such as a porphyrin ring, or a trace metal, such as copper or molybdenum, which can be measured accurately and so quantify the amount of that protein present. This simulation software assumes, for simplicity, that the protein to be purified is an enzyme. Therefore more background information on enzyme assays is given below.

In addition to measuring the amount of the desired protein in all fractions at each stage of purification, it is also necessary to determine the total amount of protein present in each fraction. This information is needed not only to assist in the decision as to which fractions should be pooled for subsequent purification steps, but also allows the investigator to judge the degree of purification at each stage so that the effectiveness of each procedure can be properly assessed.

Two approaches can be used. It may be sufficient to measure the light absorbance at 280 nm to estimate the protein content of individual fractions obtained by, say, column chromatography. Provided the necessary equipment is available, it is good practice to monitor the fractionation of total proteins by column chromatography by using a continuous-flow cell in a spectrophotometer set at 280 nm and linked to a chart recorder. This gives a continuous trace of absorbance at this wavelength as a measure of protein fractionation and so has far higher resolution than collecting fractions of known volume and measuring the protein concentration of each individually. Despite this continuous absorbance analysis, one still needs to collect fractions, of course, to enable enzyme assays to be carried out and for those parts of the column eluate containing the enzyme to be pooled for subsequent purifications.

The alternative approach to a direct spectrophotometric estimation of protein content is to assay the total protein content of each fraction using a chemical assay such as the Lowry procedure or the dye-binding assays such as the Bradford procedure.

The amount of enzyme in a solution can be determined accurately by measuring its catalytic activity, i.e. the amount of substrate consumed or product produced in unit time. Ideally, an enzyme should be assayed in a reaction mixture where the pH is optimal for that enzyme and any cofactor or coenzymes required, as well as the substrate, are present at levels above those needed to saturate the enzyme. Under these conditions, the initial reaction rate is zero order for substrate and so the initial rate of reaction is proportional only to the concentration of enzyme present.

In order to monitor the rate of reaction one needs a simple, specific analytical method for measuring the consumption of the substrate or the appearance of the products. Since the substrate usually needs to be present at quite high concentration to ensure saturation of the enzyme, it is often more accurate to monitor the progress of the enzyme reaction by measuring changes in the product concentration rather than substrate concentration. The precise method for achieving this depends upon the particular enzyme under consideration. The simplest situations are those where the substrate or product absorbs light of a particular wavelength and so changes in its concentration can be followed by measuring light absorbance at that wavelength using a spectrophotometer. An example is lactate dehydrogenase which uses lactate and NAD⁺ as substrates to produce pyruvate and NADH. Since the product, NADH, absorbs light at 340 nm whereas lactate, pyruvate and NAD⁺ do not, the rate of reaction can be followed by monitoring the increase in light absorbance at this wavelength. If the enzyme being purified catalyses a reaction in which no substrate or product can be assayed spectrophotometrically, it may be possible to use a second reaction to convert one of the products to a compound which does absorb light and so can be measured. This is called a coupled assay. Many other methods of enzyme assay exist other than those dependent on the absorbance of light by a substrate or a product - too many to consider in detail here. However the principles of enzyme assay are the same irrespective of the assay procedure adopted.

Traditionally, one unit of enzyme activity is defined as that which causes the conversion of 1 micromol of substrate per minute under the (optimal) conditions of measurement. In recent years, a new international unit of enzyme activity has been recommended. This is the Katal (abbreviated kat) and is defined as the amount of enzyme which causes the consumption of 1 mol of substrate per second. Thus an 'old' enzyme unit is equal to 16.67 nanokatals (i.e. 16.67 nkat).

Since enzyme activity itself is a measure only of the amount of the enzyme present and says nothing of the presence of other contaminating proteins, an additional term, specific activity, is also required. Using the traditional unit of enzyme activity, the specific activity is the number of enzyme units per milligram of protein. Alternatively, using the new international unit of enzyme activity, specific activity is quoted as katals per kilogram of protein. The specific activity of the desired enzyme will be low in the original tissue or cell extract and will increase during purification to a maximum when the enzyme is pure. Specific activity is thus a measure of enzyme purity.

In monitoring the effectiveness of an enzyme purification, it is not the absolute value of specific activity which is important since this will depend upon the intrinsic catalytic capabilities of the particular enzyme under study. Rather, the important aspect is how the specific activity changes during the purification. The degree of enrichment of the enzyme achieved in any given step can be calculated as follows:

A successful purification step should not only increase the specific activity of the enzyme substantially but the enzyme should also be recovered in good yield. This can be calculated as follows:

Note that the yield of enzyme after a particular purification procedure may be low not because the procedure is failing to purify that protein, but because it is causing some inactivation of the enzyme. An additional problem sometimes experienced is that the enzyme may have been active immediately subsequent to the purification step but may lose substantial activity if stored for long periods prior to assay. A lower yield than expected indicates reason for further investigation.

At a late stage of protein purification when the protein and enzyme profiles of fractions coincide, one may believe that the enzyme has been purified. Proof that the protein is pure can only really be obtained by protein sequencing. However, considerably simpler analytical procedures can be used to provide a good indication of the degree of purity. The commonest procedure is SDS-PAGE which is both rapid and sensitive and separates polypeptides solely on the basis of size. Alternatively, one can attempt analytical isoelectric focusing which separates proteins by charge differences. Ideally, however, one should test the apparently pure protein by two-dimensional gel electrophoresis which separates proteins by charge in the first dimension (isoelectric focusing) and size in the second dimension (SDS-PAGE). The component polypeptides of a protein mixture are displayed as spots on a rectangular gel slab. Given the very high resolution of this technique, a sample which displays a single polypeptide spot is probably (but not certainly) pure.