Fundamentals of Protein Biotechnology

Stein

• Verlag: CRC Press
• Jahr: 2017
• ISBN:
  9781351446358
• Auflage: 1
• Thema: Medicine
• Thema: Biochemistry in Medicine

Fills a gap between the existing studies of proteins, which tend to be highly technical and geared toward the practicing protein chemist, and biochemistry textbooks, which focus on general principles. Scientists cover a dozen topics by presenting fundamental principles, an overview, and the practica

1

An Overview and Historical Perspective of Protein Biotechnology

Sidney Udenfriend

Roche Institute of Molecular Biology
Nutley, New Jersey

THE RESURGENCE OF PROTEIN CHEMISTRY

The techniques of molecular biology were introduced in the early 1970s. It appeared to many then that this spelled the end of protein chemistry. Why isolate a protein or peptide if all one had to do was clone the corresponding cDNA and sequence it? The rapidity with which proteins were cloned appeared to corroborate this sentiment. However, most of the proteins that were cloned initially, such as insulin and growth hormone, had already been totally characterized by chemical methods prior to the advent of molecular biology.

Molecular biologists, at the time (mid-1970s), were somewhat justified in bypassing protein chemistry because it was not useful in investigations of proteins of importance but of low abundance. Why was this so? Because at that time biological activities were already measurable in the picomole range and below. While molecular biology also operated in this range, protein chemistry was still operating in the nanomole range. This disparity in sensitivities led molecular biologists to develop ingenious methods to bypass protein isolation and characterization. Today with expression cloning methods it is possible to clone and sequence the cDNA of a protein merely through its biological or immunological activity.

However, over time, there was a resurgence of interest in protein chemistry. Microanalytical methods were introduced at every step, including chemical assay, isolation, amino acid assay, and sequencing. Today protein chemistry also operates in the picomole range, making it an equal partner with molecular biology for elucidating structure and function. Furthermore, protein chemistry contributes information that is not obtainable by molecular biology alone. First of all, obtaining even limited sequence information on an isolated protein permits the preparation of synthetic deoxynucleotides that can be of considerable help as probes in cloning the corresponding cDNA. This is particularly important in attempts to clone rare species of proteins such as hormones, growth factors, and receptors. Knowledge derived from protein sequencing is generally also more precise in determining the amino terminus of a protein than is cDNA sequencing alone. Finally, the active protein or peptide frequently undergoes considerable posttranslational processing that cannot be predicted from the cDNA; namely cleavage by peptidases, phosphorylation, glycosylation, subunit aggregation, and so on. DNA has been referred to as the blueprint of life. If this is so, then each protein coded for by a specific cDNA may be considered to be an individual structure. In real life we utilize the structure not its blueprint.

The advent of molecular biology and genetics brought with it a revolution in technology. Because of the high profile of molecular biology most individuals are aware of the advances in methods involving DNA and RNA. However, there have been comparable advances in the methodology of protein and peptide chemistry, although most of these are still not routinely available to biologists. Unfortunately, while our graduate schools are training molecular biologists in great numbers today, few are being trained in modern protein and peptide chemistry, particularly at the level of purification and analysis. It is hoped that this volume will help fill this educational gap.

AN HISTORICAL VIEW

Before detailing the different procedures of modern protein and peptide chemistry, it would be interesting to look at their history and development. It is only within the last 35 years that we have come to realize that each protein is composed of stoichiometric amounts of 20 different amino acids, and that the latter are further arranged in a fixed sequence that is unique for that protein. Until the late 1940s and early 1950s there were no analytical methods that were suitable for dealing with even large amounts of proteins or peptides. While resourceful biochemists at that time were able to purify some proteins that were major tissue constituents (i.e., hemoglobin, casein, insulin, etc.) they were not able to determine even their amino acid compositions.

The earliest methods for amino acid assay utilized “specific” precipitants for each of the amino acids in a protein hydrolysate. Such assays required gram quantities of protein and were not readily reproducible from laboratory to laboratory. Furthermore, they were not sufficiently specific, and reagents were not available for all the amino acids. In 1950, microbiological assays for amino acids were introduced. Certain bacteria could be grown in media with one of the amino acids missing and all the others in excess, so that, on addition of a protein hydrolysate, the organism grew in proportion to the concentration of that amino acid. For example, to measure leucine, a culture medium was used containing all the amino acids except leucine. On addition of a protein hydrolysate, growth was proportional to the leucine content. Assay of each of the amino acids for one protein hydrolysate required incubation of the organism with a specific medium along with controls, blanks, etc. Microbiological assay was more sensitive (100–200 μg) than precipitation methods, but it was time consuming and did not have the precision to yield convincing stoichiometric data. At about the same time the isotope derivative method for amino acid assay was introduced (1). Aliquots of a labeled hydrolysate were used to assay each amino acid individually by isotope dilution methods. This method was sufficiently precise and specific to show for the first time that all proteins then available, except collagen, contained less than 0.05 residue per mole of hydroxyproline. The method was sensitive (μg quantities of protein), but again laborious. It was not until Stein and Moore (2) introduced the ninhydrin amino acid analyzer and helped introduce commercial automated analysis that amino acid assay attained the sensitivity, precision, and relative simplicity required for chemical analysis. Of course, it was the introduction of column chromatography that made the amino acid analyzer possible. The introduction of commercial amino acid analyzers by Beckman Instrument Company and others finally gave protein chemists the precision necessary to prove convincingly that each protein possesses a unique amino acid composition.

The ninhydrin amino acid analyzer, little modified from the original Stein and Moore instrument, represented the dominant technology until about 10 years ago. However, the limits of sensitivity of most commercial amino acid analyzers until fairly recently was about 1 nanomole of each amino acid. The introduction of high-performance liquid chromatography (HPLC), newer fluorescent and colorimetric reagents, and modern computer technology has now pushed the limits of sensitivity to a few picomoles of each amino acid in a hydrolysate. There are now several procedures and instruments that routinely operate in the picomole range.

For analysis to be meaningful, it is necessary to purify a protein. In the 1960s Burgess et al. (3) and Schally et al. (4) purified and characterized thyroid-releasing factor (TRF) the first of many hypothalamic releasing factors. As it turned out, TRF represented less than 0.0004% of beef hypothalamic protein. To isolate sufficient amounts for analysis by the procedures available at the time (μmol quantities) these investigators had to start with several hundred thousand hypothalami (25–50 kg). Similar quantities of tissue were used in several trial runs prior to the final isolation. The large amounts of tissue were required because the overall yields were low by the procedures that were used at the time. In 1983, when Guillemin and colleagues (5) isolated growth hormone-releasing factor (GRF) they used only 7.2 g of tissue and isolated 1–5 nmol of peptide which provided more than enough material for chemical and biological characterization. More recently Esch et al. (6) in Guillemin’s laboratory isolated and sequenced the follicle-stimulating hormone- (FSH) releasing peptide from a few liters of porcine follicular fluid. About 10 pmol was used to provide sufficient amino terminal sequencing to prepare a probe for cloning.

The practical advantages of this phenomenal decrease in scale of operation (micromoles to picomoles or 1,000,000-fold) should be pointed out. The use of less tissue and less reagent is important for financial reasons. Hundreds of thousands of beef hypothalami, the amount used for TRF isolation and the reagents, equipment, and manpower necessary to process such large amounts of tissue approached one million dollars. Modern micromethods are therefore highly cost effective. Another advantage is speed. A single run on an HPLC column takes an hour or two compared with days for older column methods. Furthermore, the efficiency of HPLC columns provides a higher degree of purification per run. Current micromethods for amino acid assay and sequencing are not only more sensitive, but also far more rapid compared with methods of only 10 years ago. In fact, the rate-limiting factor in most isolations is no longer the chemical methodology but the procedure required to monitor the biological properties of a newly discovered protein or peptide. Referring to advances in protein and peptide isolations over the years Roger Guillemin said, “Things have changed (and improved) over the years. The only thing that remains constant and an absolute requisite for all these purification procedures is a truly specific and reliable bioassay and people to do it right.”

Isolation of proteins, until the 1950s, required precipitation procedures. Salts, solvents, and pH were used to produce differential precipitation of the desired protein in the presence of other proteins extracted from a tissue. Such procedures were obviously limited to major proteins and also relatively stable ones. Differential adsorption and elution was followed by column chromatography on various types of gels in the late 1950s and early 1960s. With detection by absorption at 280 nm, sensitivity was pushed down to the nanomole level. However, the limited resolving powers of gels required large columns and long running times (days). Volumes were accordingly large and elution was limited to aqueous solutions. HPLC radically changed the nature of protein and peptide purification.

The development of rapid methods for isolation of proteins and peptides was made possible by the smaller and smaller amounts required for sequencing. Before Sanger elucidated the primary structure of insulin in the early 1950s (7), no one had demonstrated that each protein had a unique sequence. Although most scientists know that Sanger used overlapping peptides to attain the overall sequence of insulin, few realize that he never sequenced a peptide. He converted each of the two insulin chains to tri- and tetrapeptides by random partial hydrolysis in acid, separated peptides by paper chromatography, and determined the amino terminus and amino acid composition of each small peptide. From the amino acid compositions of these small randomly produced peptides Sanger was able to deduce the sequences of the A and B chains of insulin. This represented quite a feat! The first true sequencing from the amino terminus, utilizing the reagent phenylisothiocyanate was introduced by Edman and Begg (8). Largely as a result of advances in instrumentation, the Edman procedure has now achieved sensitivity in the low picomole range. Newer instrumentation has also made it possible to obtain long sequences in relatively short periods of time. With some commercial instruments, and with some proteins, it is now possible to obtain 40 to 50 sequences on as little as 100 pmol in about two days. Because of these advances in peptide chemistry, the Food and Drug Administration now requests partial sequence information on every batch of a recombinant protein product that is used clinically.

APPLICATION TO BIOTECHNOLOGY

Sequence data, whether obtained by protein or cDNA sequencing provide information for peptide synthesis. Recent developments in methods and instrumentation for peptide synthesis have also been formidable. The solid-state method, which is the heart of today’s methodology in peptide synthesis, was slow to gain acceptance. When Merrifield first introduced this procedure in 1963 (9), synthetic peptide chemists of the day were not receptive. They acknowledged it as an ingenious concept but felt that its use would be limited because the partial reaction products that accumulated at each step, even though 99% complete, could not be separated from the desired peptide by the methods then available. It was generally believed that synthesis in solution would not be supplanted because it permitted isolation and purification of intermediates at each step. Merrifield’s answer to this criticism by most of the peptide chemists of the day clearly shows his vision. “Recent improvements in purification techniques suggest that future developments will allow the ready separation of product mixtures that presently appear intractable. In addition, new and refined analytical methods should permit much greater precision in assessing the homogeneity, composition, and conformation of synthetic product” (10). HPLC was introduced not long after his remarks. Purifying even the largest peptides produced by solid-phase methods then became almost a trivial matter. Based on these synthesis and purification procedures, the production of large peptides (30–40 residues) for use as drugs is now economically feasible. Chemical synthesis of even larger peptide residues can now be carried out, but as yet, only recombinant DNA methodology can produce them in sufficient quantity and at a cost suitable for commercial applications.

Large peptides and proteins produced by recombinant DNA (rDNA) technology require purification from the normal products of the cells in which they are produced. Here, again, recent advances in purification procedures make this a relatively simple matter. The high specificity of monoclonal antibodies has been utilized to develop affinity chromatography procedures that, in one step, can isolate a cloned protein in 80% to 90% purity from a cell culture. For example, in the case of α interferon, one step of antibody affinity chromatography and a second of HPLC can yield a product that is over 99% pure and free of all detectable cell products (11). Such procedures have already been scaled up to produce and purify peptides and proteins for commercial use in 25–100g batches. Application of newer bioengineering principles to biological materials should permit scale-up to kilogram quantities.

Purification, characterization, and synthesis of a peptide or protein is quite an achievement. Besides having the polypeptide available for research and clinical study, it permits chemists and molecular biologists to investigate structure-function relationships. The purpose of structural modifications might be to increase the inherent biological activity, stabilize the molecule to the actions of tissue enzymes, or map the active site. A plethora of modifications to chemically synthesized peptides, including substitution with D-amino acids, is found in the scientific literature. Site-directed mutagenesis and other genesplicing techniques can be used to generate a myriad of polypeptides with defined structural differences.

Sequencing information gives the chemist only a two-dimensional picture of a protein or peptide when the latter actually exist and function as three-dimensional entities. Anfinsen (12) showed that the information required to fold a protein into its unique, biologically active, three-dimensional conformation already exists in its primary structure. Attempts are being made to determine three-dimensional structure from known b...