ISSN:0144-557X    

DOI:10.1039/AP99128FX017

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

ANPRDI 28(5) 137-1 68 (1 991 ) Proceedings of the Analytical Division of The Royal Society of Chemistry CONTENTS 137 Report of Meeting 137 New Government Chemist 137 Open Competition 138 Analytical Viewpoint 138 'Ad HOC Analysis: How Do You Know You Have Got it Right?' by N. T. Crosby 140 SUMMARIES OF PAPERS 140 Affinity Chromatography 140 141 143 'Affinity Chromatography-An Overview' by Ken Jones 'Immunoaffinity Purification of Industrially Relevant Enzymes' by Mark J. Berry 'Dye Affinity Chromatography' by Michael D. Scawen 145 Chemometrics Papers 145 147 150 'Chemometrics as an Aid to Liquid Chromatography Optimization' by Richard G. Brereton 'Cluster Analysis With Infrared Spectra' by Michael J. Adams 'Principal Component Analysis: An Introduction' by Gordon L. Smith 152 Polymerase Chain Reaction-Technology and Applications 152 153 154 'Molecular Archaeology' by Erika Hagel berg 'Strategies for the Use of the Polymerase Chain Reaction in Forensic Analysis' by K. M. Sullivan, G. Tully, R. Hopgood, D. Lingard and P. Gill 'Strategies of the Polymerase Chain Reaction' by Alan D. B. Malcolm 1 56 Equipment News 161 Conferences and Meetings 163 Courses 164 Publications Received 167 Analytical Division Diary Typeset and printed by Black Bear Press Limited, Cambridge, England May 1991 Analytical Proceedings

ISSN:0144-557X    

DOI:10.1039/AP99128BX019

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

ANALYTICAL PROCEEDINGS, MAY 1991, VOL 28 137 Open Competition The theme of this competition is ‘Ana- lytical Chemistry-Today’s Definition and Interpretation’. The aim of the com- petition is to generate fresh ideas about analytical chemistry and to achieve an accepted and comprehensive definition of the field. Of particular importance are the interpretations of the definition, which should demonstrate the intellectual con- cept of analytical chemistry. The four best papers, in the opinion of an international jury, will be recognized by an award consisting of a scroll and a prize of AS20000, 8000, 4000 and 4000, respectively. These papers will be publi- shed in Fresenius Journal of Analytical Chemistry. The papers should be divided into three sections: (1) Definition (10-15 lines; approximately 150 words); (2) Interpreta- tion (2, maximum 3, pages); and (3) Conclusion (10 lines). The deadline for submission is May 31 , 1991, and the papers will be evaluated by October 1, 1991. Papers should be sent to: Professor Dr. Manfred Grasserbauer, Institute of Analytical Chemistry, Technical Univer- sity Vienna, Getreidemarkt 9, A- 1060 Vienna, Austria.

ISSN:0144-557X    

DOI:10.1039/AP991280137b

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

138 ANALYTICAL PROCEEDINGS, MAY 1991, VOL 28 Analytical Viewpoint The following is a member of a continuing series of articles providing either a personal view of part of one discipline in analytical chemistry (its present state, where it may be leading, etc.), or a philosophical look at a topic of reievance t o chemists in general or analytical chemists in particular. These contributions need not have been the subject of papers at Analytical Division Meetings. Persons wishing to provide an article for publication in this series are invited t o contact the editor of Analytical Proceedings, who will be pleased to receive manuscripts or to discuss outline ideas with prospective authors. Ad Hoc Analysis: How Do You Know You Have Got it Right?* N. T. Crosby DeDartment of Trade and lndustrv, Laboratorv of the Government Chemist, Queens Road, Teddington, Middlesex TWI 1 01 Y ,- The importance of chemical measurements in today’s world is increasingly being recognized not just by chemists themselves but also by politicians, administrators and the general public.Many aspects of trade depend on analytical chemistry but in recent years it has been an increased awareness and interest in the safety of food we eat and the environment we live in that has stimulated developments in analysis. There is a growing realization that errors can be just as damaging, if not more so, than a complete lack of information. While analytical mistakes can obviously have serious consequences in the clinical and forensic fields, accuracy is equally important in other areas where only finance or administrative action is involved.Nevertheless, not all analyses are required to be performed to the highest standards of accuracy and precision. Where it is necessary only to prove that a sample satisfies a particular specification or falls within a given compositional range, the exact value of a given parameter may be of little significance unless the value approaches the permitted tolerance. Most analytical chemists these days are constrained primarily by costs. Hence, the practice of analytical chemistry must take into account not only the professional and academic aspects of the science, but also the purpose for which the analysis is being performed in order to obtain best value for money. Analytical chemistry covers such a wide range of analytes and materials.Analytes comprise around 100 inorganic ele- ments and further problems arise from speciation and anions, whilst in the field of organic chemistry several million compounds exist. Materials presented for analysis span an equally wide range of compositions and properties. The tools available to the analytical chemist are many and varied. Each technique has its own advantages and disadvantages and it is not always clear which is the best approach for any one particular circumstance. Furthermore, the analyst is under pressure to produce his results more quickly and more cheaply, and to achieve lower detection limits and greater accuracy and precision with fewer, less well trained and experienced staff. In recent years, therefore, much thought has been given to the standardization and validation of analytical methods , on the one hand, and to laboratory quality assurance on the other.Both of these aspects are designed to improve the quality and reliability of analytical results and reduce the time, money and effort wasted in the production of misleading or incorrect data. A number of official bodies have been established to produce standardized, recommended or official methods of * Based on a lecture given by the author to Analyticon 90 on September 26, 1990, at the Olympia Conference Centre, London. analysis. The Analytical Methods Committee of the Royal Society of Chemistry was constituted in 1935, although similar activities in restricted fields of analysis had been reported in The Analyst for many years without committing the Society to any corporate responsibility.The Association of Official Analytical Chemists in the USA and The International Organization for Standardization (KO) based in Geneva are examples of other organizations with a long history of work in this area. Many other bodies have developed methods applic- able to a particular commodity. The approach is invariably to take a published method, check its applicability and scope, improve the ruggedness if possible and then submit the procedure to a critical study in as many laboratories as possible using a wide range of materials. The repeatability and reproducibility of the method are established by collaborative studies carried out according to recognized protocols. The method is then written up in accordance with an approved format and any competent laboratory should then be able to use the method and obtain similar results.However, the use of a reliable method is only one requirement for obtaining satisfactory results. No laboratory can produce reliable results even with approved methods if the equipment used is not calibrated, the staff are unqualified or the environment is contaminated. Hence, increasing attention is being given to quality assurance schemes by which a particular laboratory has to prove its competence to indepen- dent third parties. In the UK, the Government has established the National Measurement Accreditation Service (NAMAS) primarily for testing and calibrating laboratories involved in physical measurement, but increasingly becoming concerned with chemical measurements.The regulations adopted by NAMAS are generally compatible with IS0 Guide 25, EN45001 and Good Laboratory Practice. Much emphasis is placed on documentation, calibration and traceability of standard materials. Participation in proficiency testing schemes relevant to the area of accreditation is also required where possible. The use of reference materials is another important safeguard for the quality of analytical work but, unfortunately, in the field of food and agriculture, suitable materials may not be available. The Laboratory of the Government Chemist has established a programme entitled Valid Analytical Measurement (VAM) to promote a National and European infrastructure designed to demonstrate and improve the validity of analytical measure- ments.This programme will support the development of better methods, the production of new reference materials, pro- ficiency testing schemes and laboratory accreditationANALYTICAL PROCEEDINGS, MAY 1991, VOL 28 139 (NAMAS), as well as research and co-operation amongst the wider analytical community. Ad Hoe Analysis Quality assurance schemes are designed to monitor the performance of laboratories over a period of time. In this way one can check for calibration shifts, reagent changes, instrumental drift, detector poisoning, etc., but such schemes are of little use for one-off analyses where no quality control data exist. For example, I am head of the Fertilisers and Feeding Stuffs Section of LGC. Under the Agriculture Act (1970) the Government Chemist is designated the referee analyst.The Fertiliser and Feeding Stuffs Sampling and Analysis Regulations contain nearly 120 separate methods of analysis. Other analytes for which no approved method exists may also have to be determined. There is no realistic way in which any laboratory can operate a continuous quality control system for such a large number of methods, particularly when the sample throughput is very low. How then can we be assured that the result obtained is meaningful? Table 1 lists some of the factors which could give rise to incorrect results. Perhaps the analytical method, contamination and interferences deserve the most attention in this type of work. The method used may well have had to be adapted for the particular purpose and the analyst will not be familiar with its use in routine analysis.The critical operations in the analysis may not have been identified, as with those procedures in use over many years, although wider analytical experience will highlight the likely problem areas. Contamination is of prime importance for analysis at very low levels and can arise from a number of different sources. Impurities present in the supply of demineralized water or contaminants in the laboratory atmosphere or on glassware can give rise to problems in such situations. For example, for the determination of very low levels of lead it is best to reserve a complete set of glassware kept covered by dilute nitric acid when not in use. For trace organic work, it is vital to ensure that solvents used by other chemists in the same building are not entering the atmosphere, thus producing spurious peaks on a gas chromatogram. Equally, plasticizers can leach out from containers and contribute to contamination.Interfering substances are difficult to eliminate however good the ‘clean-up’ stage built into the analytical method. The problem is compounded in ad hoc analysis as frequently ‘blank’ materials are not available, nor are suitable reference materials. Hence, it is often not possible to determine whether interference has occurred in a single analysis. Sampling is very important although frequently not under the control of the chemist making the analysis. Nevertheless, it is fairly obvious that unless the portion taken for analysis is truly representative of the bulk product, the result obtained could be erroneous and misleading, however well the determi- nation has been carried out. Whilst some sampling schemes and protocols have been worked out in great detail, much work remains to be done.Products must always be assumed to be heterogeneous without evidence to the contrary. Of equal importance is sample pre-treatment prior to analysis. This can include grinding, washing and cooking preparations, filtration of river waters containing sediment, drying or de-fatting to facilitate solvent extraction. All of these operations can influence the final analytical result. Whether they are necess- ary or desirable, and to what extent, can only be decided by reference to the problem put to the analyst. Validation of Analytical Data Having described some o f the pitfalls encountered in nd hoc.analysis, I would now like to describe some precautions that can be taken to ensure that reliable and meaningful results are obtained even in non-routine analytical operations. In the first place, no reliance should be placed on a single determination. If the analysis is repeated a number of times and consistent results are obtained, it is to some extent reassuring, although the values could still be consistently wrong! However, replicate analyses also serve to check on the homogeneity of the sample and random variations arising from contamination or misuse of the equipment. They provide some evidence as to the competence of the analyst. Naturally, some assays are more difficult than others and so relative standard deviations will vary from one determination to another, depending on the concentration of the analyte and interferences present.Never- theless, failure to obtain consistent replicates must always serve as a warning that something is going wrong. It is vital to make replicate measurements starting with a separate portion of sample, rather than repeated determination on the final solution. Even if identical material containing no analyte is unavail- able, it is essential to add a known amount of the analyte (at the expected concentration) to a separate portion of the sample, carry out the determination and then compute the recovery achieved. One must always remember that a simple addition of an analyte in solution to a solid matrix may not reproduce the situation encountered with the real sample where the analyte may exist in a combined state with a constituent of the matrix.It is good practice to make additions using a solvent which differs from that used in the method at the extraction stage and to allow contact between analyte added and the matrix for 24 h prior to analysis so that such interactions have a chance to occur. Where recoveries are low, it is necessary to use more rigorous extraction-digestion conditions and/or to check for losses at other stages of the procedure. The use of suitable radiolabelled compounds is advocated, but seldom possible in this type of work. Reference materials provide an alternative option if available. Blank tests, in which the determination is carried out without the sample being present, are essential as a check on possible contamination. Some analysts, using a spectrophotometric procedure, make measurements on sample solutions against a blank solution in the reference beams.However, in my view it is preferable always to measure against the solvent and obtain a separate figure for the blank, which can then be subtracted from the value obtained for the sample so that a separate measure of contamination is obtained. Calibration graphs are to some extent self regulating as any deviation from the ‘straight and narrow’ must be investigated. It is worth checking the purity of the primary reference material whenever possible. When an analyst has completed all useful work with the given method, consideration of other safeguards must then be given.First of all, it is important to check all the calculations and this must be done independently. In addition, the experimental work can be repeated by a second analyst, even if the same equipment and reagents are used. An independent result reported by a second laboratory is ideal, although seldom possible to achieve. Participation in collaborative studies is the best way of evaluating the competence of a laboratory and of individual analysts, although for ad hoc analysis the opportunity for the same determination to be tested will seldom arise. However, an analyst who performs well in a collaborative study of a method based on, say, high-performance liquid chromatography (HPLC) has at least demonstrated his competence with the technique, and even if the ad hoc determination is for a completely different analyte also using HPLC, he would give one more confidence in his results than if the analyst had never used HPLC before. However many safeguards are built into analytical opera- tions, in the end the quality o f analytical work depends primarily o n the quality of the analytical scientists at the bench and in charge of the work. This is especially true of ud hoc analysis, which relies even more heavily on the ability, experience, motivation and quality o f the individual. An analytical chemist can also improve a poor analytical method and so obtain satisfactory results. Even the best methods fdl prey to incompetent staff. Therefore, the most important safeguard must always remain the experience, reliability and reputation of the analytical chemist performing the analysis.

ISSN:0144-557X    

DOI:10.1039/AP9912800138

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

140 ANALYTICAL PROCEEDINGS, MAY 1991, VOL 28 Affinity Chromatography The following are summaries of three of the papers presented at a Joint Meeting ofthe South East Region and the Chromatography and Electrophoresis Group held on November 20th, 1990, in Glaxo Group Research, Greenford, Middlesex. Affinity Chromatography-An Overview Ken Jones Affinity Chromatography Ltd., Freeport, Ballasalla, Isle of Man Affinity chromatography is only one of many disciplines involved in the development of biotechnology processes. Despite its unobtrusive role, it is regarded as a most important interface between biochemistry and the separation sciences, and is set to play a most important part in the over-all advancement of biotechnology. It was proposed that the affinity method was the key that would overcome the major current restriction in the growth of biotechnology, i.e., the selective separation of a specific target protein from the myriads of similarly structured proteins from the crude mixtures in which they were formed.Affinity chromatography is one of the simplest of all chromatographic methods, but has generally been restricted to low-pressure systems. Although not a mainstream analytical chromatographic method, its role in large-scale macromol- ecular separations is highly valued. Nevertheless, biochemists tend to use the well tried strategies, sequencing non-selective, low-performance techniques such as ion-exchange, size exclu- sion and hydrophobic interaction methods. Basic Mechanisms Ion-exchange, hydrophobic interaction and affinity methods are effected by utilizing the basic intermolecular binding forces (selected from electrostatic, hydrophobic, hydrogen bonding and van der Waals).If only one is in play then the reagent can only exert a non-selective effect. These, in conjunction with size exclusion, constitute the main methods of separation of proteins. In contrast, affinity processes utilize the common feature of all biological macromolecules-their ability to recognize and bind to other molecules, often in a highly specific manner. It is this binding ability which allows all proteins to be separated and purified by the affinity process. The more interactions are brought into play, the better is the selectivity. Protein structures are immensely complex, and in most instances very little is known about their exact structure.Often only the isoelectric point (PI) and relative molecular mass are known. Although HPLC is extensively used for the analysis of proteins, in most instances the method cannot be scaled up, nor is there a parallel method for selective protein separations which exhibits the same power and convenience as HPLC. Consequently, large-scale protein separation strategies are usually developed empirically. Unless an effective affinity ligand is immediately to hand, the well tried non-selective technologies are applied. Once established, the regulatory process reduces the options in seeking alternative strategies, and usually results in very high separation costs. As about 5040% of the production cost of a therapeutic protein is incurred at the separation stage, it clearly is in the interests of all users to examine the affinity mode of separation at the earlier stages.In addition, affinity chromatography is uniquely capable of concentrating very dilute solutions whilst simul- taneously stabilizing the protein on the column. Despite these obvious advantages, the development of affinity chromato- graphy has been hampered by two problems: the high cost of current affinity media and the difficulties of making such media operationally stable in a multi-cycle pyrogen-free environ- ment. These complications have been compounded by the extensive use of ‘natural’ ligands, i.e., other biomolecules such as cofactors and antibodies. Such ligands are difficult to identify, tend to be rare, difficult to purify, are catalytically and enzymically unstable and are generally very expensive.As an example, protein A adsorbents cost upwards of &20 ml-1. Fortunately, it has proved possible to overcome these difficul- ties by the use of novel synthetic structures that had their origins in textile dye chemistry. Natural Versus Synthetic Affinity Ligands One of the major advantages of working in the affinity mode is the vast range of potential ligands available, theoretically offering an infinite range which can separate all other proteins. However, of the many affinity ligands used in recent decades, relatively few are commercially available. Militating against the ‘natural’ proteinaceous ligands is their general instability and high costs. Through the design of specific structures, it is now possible to design de nuvu geometrically designed backbones which can be flexed into the folds which are intrinsic in all protein structures.These frameworks can then be manipulated to carry the selected chemical groups which will maximize binding effects. The result is high selectivity. The accidental discovery that many different proteins could be separated by Cibacron Blue F3G-A led to the publication of over 1000 papers, incorporating over 90 different trade- marked dyes. Today the CI Reactive Blue 2 class of dyes, to which Cibacron F3G-A belongs, are the most widely used synthetic affinity ligands. Despite this success. it has not proved to be the breakthrough so eagerly awaited. An essential feature of all chromatographic processes is the necessity for exact repeatability from column to column, year after year.Textile dyes are bulk chemicals, most of which contain many by- products which are co-produced at every stage of the dye manufacturing process. This fact alone makes reproducibility problematical. Of even greater significance is that the bonding process between dye and matrix has been insufficiently researched. For these reasons, all commercially available textile dye-based ligands leak to various extents, the leakage rate being dependent on the degree of impurities present and the bonding technology used. However, dyes do have an underlying power to separate a very diverse range of proteins. They also have sufficient structural complexity to allow extensive manipulation of their structures, providing an opportunity to enhance these basic properties.By simul- taneously eliminating leakage, the way was clear for theANALYTICAL PROCEEDINGS, MAY 1991, VOL 28 I41 creation of a new generation of stable, inexpensive, leak-tight ligands. New Biomimetic Ligands Over several years, and by initially concentrating on well known dye structures, Cambridge University’s Institute of Biotechnology demonstrated that novel variants of such dyes would not only be universal in their application, but also that the undesirable features exhibited by textile dyes could .be eliminated. By extensive use of computer-aided molecular modelling methods, in combination with much practical experimentation , this group developed considerable insights into the synthesis of specific ligands as a function of the proteins of interest.This work led to a totally new series of MIMETIC ligands. These show a remarkable ability to separate a multifarious range of proteins, and are very stable over a wide pH range. The severe leakage experienced with a textile dye under depyrogenating conditions is clearly seen in Fig. 1, as is the over-all stability of a MIMETIC product. By packaging ten specifically designed MIMETIC ligands into a rapid screening PIKSI module (Protein Identification Kit for Sorbent Isolation), a rapid evaluation can be made as to the suitability of this system. De Novo Design There are six factors in the creation of a new ligand: design, synthesis, immobilization, characterization, in-depth testing and product evaluation.In designing the ligand, it must feature ionic, polar and non-polar groups in positions similar to the equivalent groups on the molecule of interest. Unless the X-ray crystallographic structure is known, the initial targeting of these specific groups is probably the most difficult part of the over-all problem. When X-ray data are not available, fragmentation data, produced for example by tryptic mapping, are useful, or simulation techniques can be applied. About 10 000 protein sequences are known, and about 400 three-dimensional protein structures are available on databases. By interchange of molecular modelling facilities such as the Evans and Sutherland, the Brookhaven protein database and the inter- active Macromodel (and Sybyl) programs, protein models can be projected on to a screen, then overlaid by proposed ligands.Manipulation of these models can then expose those elements which may be binding centres on the protein, to be juxtaposed with appropriate groups designed on the ligand backbone. Several examples were given, some with fully designed ligands, others which utilized a graft technique. The latter involved grafting an active part of a simpler affinity ligand, 0.9 0 2 0.8 8 0.7 .- .I- c) 8 0.6 C m g 0 . 5 - m .- .z 0.4 0.3 s 6 0.2 .- w- u) Y -I g 0.1 0 2 4 6 8 10 12 14 PH Fig. 1 Comparison of ligand leakage from MIMETIC ligand A6XL and conventional textile dye agarose. A, Cibacron Blue F3G-A- agarose (4% cross-linked); B, MIMETIC Orange 2 A6XL. Data determined from 1.0 g adsorbent samples shaken in 5 ml of test solution for 48 h at 4°C.Assay detection limit: 0.02% of initial immobilized ligand concentration, equivalent to 5 x 10-10 mol g-1 moist gel ‘ e . g . , benzamidine, on to a backbone of a known binding dye. The difficult-to-separate alkaline phosphatase was shown to be extracted from a crude calf intestinal preparation as a one-step ‘330-fold purification by the addition of a phosphonic acid group into the terminal aminobenzene ring of CI Reactive Blue 2. In conclusion, it was emphasized that the widespread adoption of non-selective media contributes significantly to the high cost of manufacture of most proteins. Replacement of conventional media by highly selective affinity media using ‘natural’ ligands is technically successful, but in general such media are uneconomic and unstable under the harsh regenera- tion procedures commonly used.Reagents based on textile dyes have had some success, but they in turn suffered from significant disadvantages, not least of which was leakage. Until recently, computer-assisted modelling techniques capable of creating de rzovo synthetic highly selective ligands were unavailable. Synthetic stable state of the art ligands can now be selected by use of a simple but rapid screening module. Whenever a greater performance is demanded, powerful and sophisticated design technologies can re-design any of the base structures, and custom-specific selective affinity media can be synthesized. lmmunoaff inity Purification of Industrially Relevant Enzymes Mark J. Berry Immunology Department, Unilever Research, Colworth House, Sharnbrook, Bedfordshire MK44 I LO Background The exquisite specificity of a monoclonal antibody may be used as the basis for very high-resolution separation systems.This is the principle of immunoaffinity purification. The diversity of the immune system is enormous: it has been estimated to have the capacity for making 108 different antibody specificities.1 In practical terms, this means that it is possible to raise monoclo- nal antibodies (usually from the mouse) that are uniquely specific for any enzyme, serum protein, carbohydrate, cell, virus or small organic compound that is likely to be encoun- tered. The technique is widely used for laboratory-scale prepara- tive work but is rarely used either for analytical applications or for industrial-scale processes.One reason for this is the lack of chromatographic media with the dual properties of large pores (to accommodate the antibody molecule) and rigidity (to permit high flow-rates). Agarose is normally used but is too soft for rapid analytical separations or industrial-scale processes. However, new media are being developed for immunoaffinity purification2 which will alleviate this problem; therefore, the role of immunoaffinity chromatography in analytical biochemistry should be reconsidered. This paper describes two project areas: the purification of a-galactosidase from guar seeds and the purification of alkaline phosphatase from calf intestine. Both of these projects were laboratory-scale preparative studies. However, the main find-142 ANALYTICAL PROCEEDINGS, MAY 1991, VOL 28 ings and conclusions are equally relevant to analytical chroma- tography.Experimental Purification of a-Galactosidase from Guar Seeds a-Galactosidase can be used to improve the properties of guar gum, in particular its readiness to form gels. Guar gum that has been modified by this enzyme may be used as a food stabilizer in low-fat spreads, ice cream and pet foods. a-Galactosidase may be purified from guar seeds but the previously published methods required two chromatographic steps3 and in our hands were very time consuming. The aim of this study was to develop a rapid single-step process of purifying a-galactosidase from guar seeds using immunoadsorbents. Three monoclonals to a-galactosidase were raised and immobilized on Sepharose. All three immunoadsorbents exhibited retardation of a-galactosidase in phosphate-buffered saline (PBS) rather than tight binding.This is in marked contrast to other immunoadsorbent processes designed in our laboratory and by others.4.5 As all three antibodies isolated in the study bind a-galactosidase with lower affinity than would be typical, it seems that there is some generic effect brought about by a peculiar immunogenicity of this enzyme. The antibody with the highest affinity was determined by measuring the selectivity, (V, - Vo)/(V, - Vo), for the three immunoad- sorbents. For this immunoadsorbent , the resolution was improved further by optimizing the loading buffer. It was found that the antibody-enzyme interaction was stronger in high ionic strength buffer (PBS + 1 mol dm-3 NaCI).This increased binding was sufficient to adsorb, i. e . , immobilize, a-galactosidase and therefore complete separation from other enzyme activities was now possible. The fact that binding was increased in high ionic strength buffer suggests that hydro- phobic interactions make an important contribution to this particular antibody-antigen interaction. Binding of wgalacto- sidase was confirmed to be immunospecific by passing the same feedstock down an identical column with no antibody bound and down a column with a non-specific antibody bound (Fig. 1). - LD - 2 1.0 f 0.8 2 0.6 2 0.4 0 ? 0.2 2 0 9 2 0 20 40 60 0 20 40 60 E I ut io n volume/m I Elution volurne/m I Fig. 1 Interaction of a-galactosidase with columns in high-salt buffer.( a ) Column functionalized with monoclonal antibody specific for a-galactosidase; (6) column not functionalized Purification of Alkaline Phosphatase from Calf Intestine The aim of this study was to purify alkaline phosphatase for use as a marker enzyme for diagnostic tests. A particular objective was to isolate a thermostable isoenzyme to improve the product shelf-life . Five monoclonals to alkaline phosphatase were raised and immobilized on Sepharose. All five immunoadsorben ts bound enzyme tightly in PBS, which is more typical than in the a-galactosidase project. Enzyme was eluted in 50% ethylene glycol, which was found to desorb enzyme without inactivating it. Lower affinity antibodies released enzyme on elution with 50% ethylene glycol only; others required 50% ethylene glycol and elevated pH (pH 10.5).The advantage of using lower affinity antibodies which release enzyme with 50% ethylene glycol (pH 7) is that these less harsh elution conditions were found to prolong the lifetime of the immunoadsorbent. Alkaline phosphatase is expressed as many different isoenzymes in calf intestine. These are formed by different carbohydrate modifications giving rise to a number of different ‘glycoforms’.~ These may be identified by isoelectrofocusing in acrylamide gels and staining with a coloured enzyme substrate 5-bromo-4-chloroindo1-3-yl phosphate (BCIP) . Commercial preparations of calf intestine phosphatase were found to contain about six isoenzymes and there was a high degree of inter-batch variation.It was found that of three antibodies studied, each was specific for a different sub-group of isoenzymes. Isoenzymes that did not bind the immunoadsor- bent passed straight through, others eluted as a specific peak (Fig. 2). The isoenzyme specificity of each immunoadsorbent was determined by running enzyme fall-through and peaks on isoelectrofocusing gels. We failed to isolate a thermostable isoenzyme routinely; however, this model system clearly demonstrates the potential for separating different glycoforms by immunoaffinity chro- matography. 1600 P . I - E 800 0) Q s (I E L + 400 g m - 0 0 20 40 60 80 Y O 0 Elution volume/ml Fig. 2 Recovery of alkaline phosphatase from calf intestine with immobilized monoclonal antibody (2405 4. Only 70% of isoenzymes bind to the column Conclusions There are two important optimization stages in the design of an immunoaffinity separation system.First, the most appropriate antibody should be selected from a library rather than seizing upon the first antibody which works adequately. Important criteria for selecting antibodies include affinity and specificity. Examples of how these may be determined in practical terms have been described. Second, the loading and elution buffers should be optimized to maximize the resolution and column lifetime. Immunoaffinity separation is a very high-resolution tech- nique and one area where it may play an important role in the future is in the detection and/or purification of protein glycoforms. The separation of glycoforms (especially of recombinant therapeutic proteins) is becoming an important issue in biotechnology.7 Immunoaffinity separation is an appropriate technique to use as antibodies may be raised against different carbohydrate patterns on the surface of glycoproteins, making the complete separation of different glycoforms possible. It is unlikely that such high resolution could be achieved by any other technique that is currently available.References Schultz, G . , paper presented at RSC Meeting on Catalytic Antibodies, 1990. Nakamura, K., Hashimoto, T., Kato, Y., Shimura, K., and Kasai, K., J. Chrornatogr.. 1990, 510. 101. McCleary, B. V., Phytochemistry. 1983, 22, 649. May, K., Gani, M. M.. and Senior, S. J., Eur. Pat., 1985, PN 0151 320. Bureau, D.. and Daussat, J., J. Immunol. Methods, 1981.41, 387. Coleman, J. E., and Gettins, P., Met. Ions Biol., 1983, 5 , 156. Michaels, A., in Separations for Biotechnology, ed. Pyle. D. L., Elsevier, London, 190, pp. 3-8.ANALYTICAL PROCEEDINGS, MAY 1991, VOL 28 143 Dye Affinity Chromatography Michael D. Scawen Division of Biotechnolog y, Centre for Applied Microbiology and Research, Porton, Salisbury, Wiltshire SP4 OJG For the purification of proteins, affinity chromatography offers perhaps the ultimate in high resolution. As a result, proteins can be purified in fewer steps, so giving higher recoveries than when conventional techniques such as ion exchange or gel filtration are used. A wide variety of compounds can be used as affinity ligands for the purification of proteins; examples include substrate, substrate analogue or inhibitor, cofactor, receptor, antibody or textile dye.Affinity chromatography, using biological ligands , suffers from the disadvantages that such ligands can be unstable, expensive and difficult to immobilize. The textile dyes have the advantage that they are inexpensive bulk chemicals, whereas the other ligands are often expensive biochemicals. The textile dyes can be viewed as an example of a group-specific ligand that can bind more than one protein. They were originally developed by ICI during the 1950s to provide colourfast dyes for cotton fabrics, by forming a covalent bond between the dye and the fabric.' The dye molecules consist of three main elements: a chromo- phore, which may be an anthroquinone, an azo compound, a phthalocyanine or a triphendioxazine compound, a spacer ring, usually diaminobenzenesulphonic acid, and a reactive group, based on trichloro-s-triazine; this is often also linked to an aminobenzenesulphonic acid group.The reactive triazinyl chloride group is able to react with hydroxyl groups on cellulose or agarose under mildly alkaline conditions. The structures of many of the dyes are proprietary, but Procion Blue H-B, shown in Fig. 1, is a typical example of a blue dye. The structures of several other dyes are given in reference 1. 0 NH, ' S O f so3- (m/p mixture) Fig. 1 Structure of Procion Blue H-B Use of Dyes in Affinity Chromatography The interaction between proteins and immobilized textile dyes was first observed by accident; yeast pyruvate kinase was found to elute from a Sephadex G-200 gel filtration column in an anomalous position when it was co-chromatographed with Blue Dextran.2 Blue Dextran is a high relative molecular mass dextran (M, > 2 x 106) linked to the blue dye Cibacron Blue, which is widely used as a void volume marker in gel filtration.It was quickly shown that it was the dye, not the dextran, that was responsible for the unusual behaviour, and that the interaction could be abolished at high salt concentrations. This observa- tion quickly led to the use of Blue Dextran, either alone or immobilized on to agarose, for the affinity purification of several enzymes, including phosphofructokinase ,3 pyruvate kinase4 and lactate dehydrogenase.5 Later it was found that the dye alone could be immobilized on to agarose and other supports ,6 and dye affinity chromatography rapidly became established as a purification method.There are a large number of dye compounds available, and soon protein chemists armed themselves with arrays of dyed agarose columns. As there is very little guidance as to which dye to use for a particular protein, the classical approach has been to screen large numbers of different dye columns for binding and elution of a desired protein. The methods of elution are similar to those used for classical affinity chromato- graphy, and are summarized in Table 1. In general, the most specific method of elution should be used, but many proteins have been successfully purified using salt elution. Table 1 Methods of elution from affinity matrices Biospecific elution (cofactor, substrate, inhibitor, free ligand) Change in ionic strength Change in pH Chaotropic agent (KSCN) Denaturing agent (urea) Change in polarity (ethylene glycol) One advantage offered by the dyes is due in part to the large number available. It is possible to screen for columns that either bind the protein of interest (positive purification), or do not, but bind instead to the contaminating proteins (negative purification).In this way it is possible to devise multi-enzyme purifications which use a series of different dye columns to purify proteins selectively from an extract. Thus it proved possible to purify malate and 3-hydroxybutyrate dehydro- genases from Rhodopseudomonas sphaeroides by taking advantage of their differing, but complementary, behaviour on two dye-agarose columns. On the first, Procion Red HE3B, hydroxybutyrate dehydrogenase was eluted with KC1, but malate dehydrogenase was only eluted by NADH in the presence of KCI.On the second dye column, Procion Blue MX-4GD, hydroxybutyrate dehydrogenase w2s eluted by NADH in the presence of KC1, but malate dehydrogenase was eluted by KCI alone.7 It has also proved possible to purify three enzymes, glucokinase, glucose-6-phosphate dehydrogenase and fructokinase, from Zymomonas mobilis on three succes- sive dye-agarose columns coupled together. At pH 6.0 glucokinase and glucose-6-phosphate dehydrogenase bound to Procion Scarlet MX-G, whereas fructokinase and many other proteins did not. Contaminating proteins were further removed by passage through a Procion Brown H-5R column.Fructokinase did not bind, but was retained by a column of Procion Yellow MX-GR. The columns were separated, and glucokinase eluted from the Procion Scarlet MX-G column by increasing the pH to 7.0, whereas glucose-6-phosphate dehy- drogenase was eluted with NADP. Fructokinase was eluted from the Procion Yellow MX-GR column with ATP.8 Using dye affinity chromatography proteins from every major class have been purified. Some examples are given in Table 2, and extensive lists of such purifications have been compiled (see, e.g., references 15 and 16). How Dyes and Proteins Interact The nature of the interaction between a dye molecule and a protein is uncertain, and must vary between different types of proteins. However, for alcohol dehydrogenase, the interaction has been studied by X-ray crystallography.17 It was found that the free dye was bound to the enzyme at the NAD-binding site, with the greatest similarity in the ADP-ribose region. Most interestingly, it was found that the triazine ring portion of the dye molecule was buried in the protein, with only the144 ANALYTICAL PROCEEDINGS, MAY 1991, VOL 28 Table 2 Some proteins purified by large-scale dye affinity chromatography Enzyme Dye Glycerol kinaseg Procion Blue MX-3G Glucose kinasel0 Procion Brown H-3R Glycerol dehydrogenase" Procion Red HE-3B 3-H ydrox ybut yrate dehydrogenase7 Procion Red H-3B Malate dehydrogenase7 Procion Red H-3B Tryptophanyl tRNA Human serum albumin13 Carboxypeptidase G2I4 Procion Red H-8BN Procion Blue MX-4GD Procion Blue MX-4GD Procion Brown MX-SBR Procion Blue HB s ynthetase 12 Eluent 5 mmol dm-3 ATP 5 mmol dm-3 ATP 2 mmol dm-3 NAD 1 mol dm-3 KCl 1 mol dm-3 KCI + 2 mmol dm-3 NADH 1 mol dm-3 KCI + 2 mmol dm-3 NADH 0-700 mmol dm-3 KCI gradient 50 mmol dm-3 tryptophan 20 mmol dm-3 sodium octanoate Bind in 0.2 mmol dm-3 Zn*+ Elute with 10 mmol dm-3 EDTA anthroquinone ring exposed on the surface.This is similar to the binding of NAD, in which only the adenine ring is exposed. This, of course, has implications in the use of the dyes for purification, because the dye is attached to the matrix by the triazine ring. The site of this interaction has been further demonstrated by isolating a dyed peptide, which was obtained by reacting the enzyme with dye, digesting it with trypsin and isolating the coloured peptide.Amino acid sequence analysis of the peptide showed that the dye was attached to cysteine- 243,18 which is located within the catalytic site of alcohol dehydrogenase. Future Trends Knowledge of the interaction between a dye and an enzyme has led to the synthesis of dye analogues in which the point of attachment of the dye to the support can be controlled.19 It was found that by varying the point of attachment of the dye molecule, the degree of purification of alcohol dehydrogenase from a crude liver extract varied considerably, as shown in Table 3. The most effective point of attachment was via N-1 of the anthroquinone ring, which is analogous to N-6 of the adenine ring of NAD. NAD is an effective affinity ligand only when attached to the matrix by this atom.Table 3 Purification of liver alcohol dehydrogenase on immobilized Cibacron Blue analogues19 Immobilization point Purification factor Triazine 4.3 Anthroquinone N-1 with Triazine with 6-aminohexyl spacer 7.3 aminoethylaminoacet yl spacer 10.3 The development of dye analogues, with controlled struc- tures and sites of attachment, can be seen as the beginning of a new era in the use of dye affinity chromatography. The analogues can be expected to have higher specificity for a given protein, and by virtue of the chemistry of attachment they may also be less prone to leakage of ligand from the support. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 References Stead, C. V., J. Chem. Tech. Biotechnol.. 1987,37, 55. Haeckel, R.. Hess, B., Lauterborn.W., and Wuster, K.-H., Hoppe-Seylers Z. Physiol. Chem., 1968,349, 699. Kopperschlager, G., Diezel, W., Freyer, R., Liebe, S., and Hofmann, E., Eur. J. Biochem., 1971, 22, 40. Blume, K., Hoffbauer, R., Busch, D., Arnold, H.. and Lohr, G., Biochim. Biophys. Acta, 1971,227, 364. Ryan, L., and Vestling, C.. Arch. Biochem. Biophys.. 1974, 160,279. Easterday, R. L.. and Easterday, I. M., in Immobilised Biochemicals and Affinity Chromatography, ed. Dunlop, R. B., Plenum Press, New York, 1974, p. 123. Scawen, M. D., Darbyshire, J., Harvey, M. J., and Atkinson, T., Biochem. J., 1982, 203,699. Scopes, R. K., and Griffiths-Smith, K., Anal. Biochem., 1984,136,530. Scawen, M. D., Hammond, P. M., Comer, M. J., and Atkinson, T., Anal. Biochem., 1983, 132,413. Goward, C. R., Scawen, M. D., and Atkinson, T., Biochem. J . , 1986, 237,415. Spencer, P., Bown, K. J., Scawen, M. D., Atkinson, T., and Gore, M. G., Biochim. Biophys. Acta. 1989, 994,270. Atkinson, T., McArdell, J. E., Scawen, M. D.. Shenvood. R. F., Small, D. A. P., Lowe, C. R., and Bruton, C. J.. in Affinity Chromatography and Related Techniques, eds. Gribnau, T., Visser, J., and Nivard, R. J. F., Elsevier, Amsterdam, 1982, p. 399. More, J. E., Hitchcock, A. G., Price, S., Rott, J., and Harvey, M. J., in Protein-Dye Interactions: Developments and Applica- tions, eds. Vijayalakshmi, M. A., and Bertrand. O., Elsevier, Amsterdam, 1989, p. 265. Shenvood. R. F., Melton, R. G., Alwan, S. M., and Hughes, P., Eur. J. Biochem., 1985, 148,447. Lowe. C. R., in Topics in Enzyme and Fermentation Bio- technology, ed. Wiseman. A. R., Ellis Honvood. Chichester, 1984, p. 78. Scawen. M. D., and Atkinson, T., in Reactive Dyes in Protein and Enzyme Technology, eds. Clonis. Y. D.. Atkinson, T.. Bruton, C. J.. and Lowe. C. R., Macmillan, London, 1987, p. 51. Biellmann, J.-F.. Samama, J.-P., Branden, C. I., and Eklund, H.. Eur. J . Biochem., 1979, 102, 107. Small. D. A. P., Lowe, C . R., Atkinson, T., and Bruton, C. J., Eur. J. Biochem.. 1982, 128, 119. Burton, S. J., Stead, C. J., and Lowe, C. R., J. Chromatogr., 1990, 508, 109.

ISSN:0144-557X    

DOI:10.1039/AP9912800140

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

ANALYTICAL PROCEEDINGS, MAY 1991, VOL 28 145 Chemometrics Papers The following are three papers originally written for the newsletter of the United Kingdom Chemometrics Discussion Group. Chemometrics as an Aid to Liquid Chromatography Optimization Richard G. Brereton School of Chemistry, University of Bristol, Cantock's Close, Bristol BS8 1 TS There is a very large amount of literature on chromatographic optimization. For details the reader is referred to texts by Berridge' and Schoenmakers2 and a recent issue of J. Chromatogr.3 Most recent work on chromatographic optimi- zation is of two types, namely (1) automation, making use of fairly well established chemometric techniques to aid the experimenter choose appropriate conditions for separations of compounds and (2) extending chemometric approaches.In this paper, current directions of research in chemometric chromat- ographic optimization are discussed, with the emphasis on high-performance liquid chromatography (HPLC) . The first, and probably most important, step is to determine why there is a need to optimize the system. This, in turn, depends on what is expected from the chromatography. Chromatography is used for very diverse reasons. One of the simplest reasons is for preparative separations, for example of three or four isomers of a drug. Speed and resolution are key factors. A second reason is as an analytical method. For example, natural extracts might consist of a variety of different compounds; chromatography is used to determine which main compounds are present. A third reason is to quantify the proportion of compounds in a mixture.There has been a great deal of chemometric work in the area of HPLC, but recently, considerable interest has been shown in hyphenated techniques such as HPLC with diode array detection (DAD). Liquid chromatography-mass spectrometry has been of a lesser interest to chemometricians. Often the next stage is to define a separation criterion, often called a chromatographic response function (CRF) . Berridge4 has discussed the merits of several such functions. These generally are related to peak heights or widths in a chromato- gram. For example, a simple function might be to define the average peak height in a standard mixture of five compounds. The larger this number, the narrower the peaks, and hence the better resolved the chromatogram.Clearly, there are several other possible functions that could be determined. There are several difficulties to this approach, as follows. (1) This approach normally requires synthetic mixtures of known and reproducible standards. This is easy in some areas, such as pharmaceutical chemistry where large amounts of stable and easily characterized compounds can be mixed in known proportions; it is less useful in areas such as environmental chemistry and geochemistry where there are often significant technical difficulties in isolating pure, stable standards. (2) The approach is univariate, i.e., only one parameter is defined for the experiment. Chromatography is almost certainly a multi- variate science, in which errors should be studied over the whole system.Condition A, for example, might result in peak 1 being broad and peak 2 being sharp, whereas condition B might result in the reverse situation. A univariate index of peak widths might be unable to distinguish these two conditions. (3) The approach is not easily applicable to hyphenated techniques such as HPLC-DAD where spectroscopic purity is important, particularly if the technique is used for analytical or quantita- tive reasons. (4) Most CRFs do not take into account peak cross-overs. This means that the positions of two or more peaks swap over between experiments. The advantage, however, of using CRFs is that the subsequent optimization is easy to automate. The result of each experiment can be immediately computed and compared with the results of other experiments.This is important for on-line optimization. The next step is to choose a design. For an n-component mixture, there are n - 1 degrees of freedom, because the total proportion of solvents must add up to 100%; hence the proportion of the nth solvent is completely dependent on the proportion of the other n - 1 solvents. Often the possible proportions of solvents in a mixture are represented graphic- ally by a simplex. Fig. 1 illustrates the proportions of solvents in a three-component mixture, and the triangle within this cube represents the possible combinations of the solvents that add up to 1. This triangle is also called a simplex, which is defined as the simplest possible geometric figure in n - 1 dimensional space. A one-dimensional simplex is a straight line and a three-dimensional simplex is a tetrahedron.For a three- component mixture we wish to perform experiments over the triangle. There are several ways of arranging these experi- ments, called mixture designs.5 A common design is the simplex lattice design as illustrated in Fig. 2, in which seven experiments are taken over the triangle. These experiments can be related to the proportion of solvent in each mixture. For example, if solvent A is acetone, B is methanol and C is water, the experiment on the edge between corners A and B involves a mixture of 50% acetone and 50% water. Fig. 1 three-component mixture fit within the shaded triangle Mixture triangle. 'I'he possible combination of solvents is a One of the difficulties with the mixture design illustrated in Fig.2 is that some of the experiments are not very sensible physically. For example, there is very little point running146 ANALYTICAL PROCEEDINGS, MAY 1991, VOL 28 100% A 100% c 100% B Fig. 2 Simplex lattice mixture design. The triangle corresponds to the cross-section in Fig. 1. The seven experiments correspond to 100% of each pure constituent (three experiments at the corners of the triangle), 50% mixtures of combinations of two constituents (three experiments on the edges of the triangle) and a 33%-33%-33% mixture of all three constituents (experiment in the centre of the triangle) columns in 100% water. Fortunately, chromatographers have established methods that overcome this problem. The concept of eluotropic strength has been developed by Snyder and Kirkland6 and other workers over the last decade.Each solvent has an index of eluotropic strength; the value for water is 0. The values for other solvents differ according to different workers and the experimental procedure; values of 2.6 for methanol, 3.1 for acetone and 4.4 for tetrahydrofuran (THF) will be used here. The over-all eluotropic strength for a mixture of I solvents is given by where is the proportion of solvent i and Si the eluotropic strength for the corresponding solvent. Therefore, a 50 + 50 mixture of methanol and THF will have an eluotropic strength of 3.5. One factor that influences chromatographic separations is eluotropic strength, and more realistic designs involve running all the experiments under isoeluotropic conditions, i.e., at a constant value of y.Generally, water is used as an additional solvent. If the desired eluotropic strength is 1.5, then a proportion of methanol of 1 3 2 . 6 = 0.58 and of water of 1 - 0.58 = 0.42 represents a possible condition. Instead of the corners of the triangle in Fig. 2 corresponding to 100% of each constituent they will correspond to two-componen t mixtures of each of the solvents and water, hence all the experiments are performed at constant eluotropic strength. It is then easy to calculate all the observed experimental conditions. For example, if solvent A is acetone and B is methanol, then the conditions on the corner A of the triangle will correspond to (100 x 1.5)/3.1 = 48% acetone and 52% water and corner B 58% methanol and 42% water, hence the experiment on the middle of the edge between conditions A and B will correspond to 24% acetone, 28% methanol and 48% water.Methanol Eluotropic strength\ Fig. 3 Gradient mixture design for methanol, THF and acetone. Each triangle corresponds to one eluotropic strength Naturally, these designs can be extended to five or more solvents, but commercial HPLC systems are usually restricted to four solvent inlet systems. One weakness with these designs is that they do not take gradients into account. Most chromatographers like to change the composition of solvents during a run. In conventional two-inlet systems, the proportion of water and methanol can be altered during the chromato- graphy. There are a large number of possible ways of changing gradients, but mixture designs such as those described above do not take gradients into account.It is not really a very good idea to use chemometrics if the problem has to be so simplified that the resulting conditions are artificially restricted. Hence some workers such as Glajch and Kirkland7 have proposed designs that incorporate a gradient. This normally involves a fixed change in eluotropic strength during the course of the experiment. For example, the value of y might vary from 1 to 2. The design can be represented by a prism, each end represent- ing a mixture triangle (Fig. 3). At time zero the triangle corresponds to eluotropic strengths of 1. The seven experi- ments involve changing conditions to corresponding points on each triangle. For example, if point A corresponds to an acetone-water mixture, then the starting condition is (100 x 1)/3.1 = 32% acetone and 68% water, and the final condition 64% acetone and 36% water.Yet another possibility, being explored in Bristol, is to modify the gradient in the design. For example, we could use three levels of eluotropic strength, such as 0.75, 1.5 and 2.25, and perform three sets of seven experiments going from 0.75 to 1.5, 0.75 to 2.25 and 1.5 to 2.25. There is, however, a major drawback to many of these approaches and that is that they depend critically on the measured values of S for each constituent. Serious discrepancies exist in the literature and these can have major effects on the actual solvent composition. The chromatographer is probably best advised to measure these values himself on a range of compounds corresponding to those of direct experimental interest if he wishes to carry out innovative work on experimental design.The discussion? above omits simplex methods for optimiza- tion, covered elsewhere,”9 and also discussion of the choice of other parameters such as columns and operating conditions, often best performed using non-chemometric approaches such as expert systems. Once the experiments have been performed, the chromato- grapher needs a method of interpreting the data. Often the CRF (see discussion above) is modelled, i.e., the CRF, c, is fitted to a model of the form for a three-component mixture, where xi is the proportion of component i in the mixture. Other parameters such as y can also be employed. This polynomial can then be used to calculate the optimum (normally maximum) of the CRF.Several different CRFs can be tested and different optima established according to the definition of the CRF. An advantage of this modelling is that it is then sometimes possible , if sufficient experiments have been performed, to determine how well the model fits the data, using approaches such as an analysis of variance (ANOVA).lO An important consideration, however, is the number of degrees of freedom in the experiment. Eqn. (2) involves exactly seven terms, which is exactly equal to the number of experiments in a three-factor simplex lattice design; hence there is no information as to how well the model fits the data: this is equivalent to fitting a straight line to two experimental points; it is certainly possible to determine the slope and intercept of the straight line but no further information is available as to whether there is a genuine linear trend in the data or not.11.12 If we wished to assess the significance of the model we would need to perform additional experiments. Sometimes it is not really important to ascertain the significance of models as many chromatographers areANALYTICAL PROCEEDINGS, MAY 1991, VOL 28 147 concerned with the separation of compounds, hence this information is not really important.One problem with this apparently simple approach of fitting polynomial models is that it does not take peak cross-overs into consideration. Fig. 4 illustrates, symbolically, the position of three peaks under different conditions. Using condition B, two of the peaks overlap exactly.It is possible that condition B results in the sharpest peaks, hence a polynomial that is fitted to the mean peak width in the spectrum would select condition B as the optimum. There is no magic solution to this difficulty except to check carefully the actual chromatograms for a given application. 3 1 I 2 1 I I Condition A J \ 3 I Fig. 4 Difficulty of cross-overs. Under condition B, peaks 1 and 2 coincide, yet condition B might result in the narrowest peaks and hence the optimum of a CRF dependent on average peak widths Analysis using CRFs treats the data as univariate data. Often the factors that influence the appearance of a chromatogram are multivariate; sometimes, for example, the peak width might vary across the spectrum, or be dependent on chemical and instrumental factors.Peak shapes might change according to conditions, varying from skew to symmetrical shapes. Therefore, for more sophisticated approaches, methods such as MANOVA (multivariate analysis of variance)l3 have potential and the response function could be a multivariate one such as a principal component or a partial least-squares (PLS)14 component. There remains a great deal of work to be performed on multivariate optimization of chromatographic separations. A related reason for optimizing chromatographic conditions is to quantify peaks. A balance must be made between analysis ( e . g . , deconvolution, curve fitting and factor analysis) and improving physical separations.For example, if two peaks are completely overlapping in HPLC-DAD, unless the peaks have very different elution profiles, the only way to resolve them is to use spectroscopic information and multivariate methods. If the peaks are completely separated, quantification is usually performed by calibration against an external standard. The balance between the time spent improving separations and the time spent developing chemometric methods for quantifying components in a mixture must be determined according to the problem in hand and there is no automatic best solution. The SERC are thanked for the grant GWE73376 for support of this work. Helpful discussions with Dr. J. C. Berridge and Dr. A. Rahmani are acknowledged. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 References Berridge, J .C . , Techniques f o r Automated Optimization of HPLC Separations, WiIey, New York, 1985. Schoenmakers, P., Optimization of Chromatographic Selectivity, Elsevier, Amsterdam, 1986. J. Chromatogr., 1989, 485, Special Issue. Berridge, J. C., Chemometrics Intell. Lab. Syst. , 1988, 3, 175. Cornell, J. A., Experiments with Mixtures: Designs, Models and Analysis, Wiley. New York. 1981. Snyder, L. R . , and Kirkland, J . J., Introduction to Modern Liquid Chromatography, Wiley, New York, 1979. Glajch. J . L., and Kirkland, J. J . , J. Chromatogr., 1989,485,51. Berridge, J . C . . Chemometrics Intell. Lab. Syst., 1989, 5, 195. Burton, K. W. C., and Nickless, G., Chemometrics Intell. Lab. Syst., 1987, 1, 135. StHhle, L., and Wold, S.. Chernometrics Intell. Lab.Syst., 1989, 6, 259. Brereton, R. G . , Chemometrics: Applications of Mathematics and Statistics to Laboratory Systems, Ellis Horwood, New York, 1990, ch. 2. Allus, M. A., and Brereton, R. G . , Int. J . Environ. Anal. Chem.. 1990. 38. 279. Stbhle, L . , and Wold, S., Chemometrics Intell. Lab. Syst., 1990, 9, 127. Martens. H . , and Nxs. T.. Multivariate Calibration, Wiley, Chichester. 1989. Cluster Analysis With Infrared Spectra Michael J. Adams School of Applied Sciences, Wolverhampton Polytechnic, Wolverhampton WVI ISB Pattern recognition and cluster analysis seek to classify objects according to some set of measured parameters and assign any given object to a class of objects similar to it. Mathematically, the object is represented by a pattern vector xik denoting the value of the kth measurement on the ith object. Hence, an alloy can be classified according to its elemental composition and a protein by its amino acid constituents.For classifying and interpreting infrared spectra this is usually performed manually using structure-spectra correlation charts and it is assumed that chemists recognize discrete features and an over-all class assignment is made according to the absence or presence of such features. Computerized matching and interpretation of infrared spectra is generally undertaken in a similar manner but usually with the prerequisite that the amount of data compris- ing the spectrum is considerably reduced. A digitized infrared spectrum consists of several thousand absorption measure- ments, each made at a different frequency of incident radiation.This spectral data pattern does not constitute several thousand independent measurements and the pattern vector can be reduced by several means. An infrared spectrum is highly self-correlated, i.e. , each recorded absorption value is dependent on its predecessor. As highly correlated variables provide little additional information to the classification process, it is advantageous to employ methods which serve to reduce the correlation. One simple means of achieving lower correlation in a spectrum is to subtract each spectral value from its predecessor. The resulting first-derivative spectrum is less self-correlated and can be employed in producing efficient, error-free data compression techniques. 1 The amount of data, however, is not reduced by this method.Principal component analysis (PCA) is being employed increasingly in analytical science to reduce data and extract maximum information.148 ANALYTICAL PROCEEDINGS, MAY 1991, VOL 28 Essentially PCA involves the transformation of the original parameter axes (representing the variables) into new axes, such that these new axes lie along the directions of maximum variance. Hence, the first PC axis can provide much of the variance, i.e., information, in the original data, the second PC axis provides a smaller percentage of the total variance, the third axis a smaller percentage still, and so on. Often the number of new factors needed to describe say 90% of the sample variance is much fewer than the original number of parameters, i.e., PCA is a dimension reduction method.In addition, the PC axes are orthogonal and uncorrelated. Their use, therefore, provides efficient factors for subsequent pattern recognition and classification. Principal component analysis can reveal the underlying parameters or combination of original variables that principally determine the structure of the data, making PCA a valuable tool in interpreting and understanding subsequent classification decisions. -50 t A A A I I I I A -50 1 50 C -50 1 -100 0 100 200 300 400 PC I Fig. 1 ( a ) Loadings of the 50 infrared spectra on the first two principal components. A, OH; B, NH2; C, C-0-C; D, C=O; E, RCO2R; F, C02H; and G, NO2. ( b ) As for ( a ) but distinguishing between those containing a carbonyl group (A) and those without (B).(c) As for ( a ) but distinguishing between those containing an aromatic centre (A) and those without (B) In a simple practical example, PCA was performed o n a set of 50 infrared spectra of simple organic materials. The efficiency of PCA to reduce the dimensionality of the data is readily observed as fewer than ten components can display more than 80% of the original sample variance. The scores of the data projected on the first two principal component axes, containing more than 40% of the variance of the original data, are illustrated in Fig. l(a). A tendency is evident for the data to group or cluster according to the presence of a carbonyl functional group, Fig. l(b), and aromaticity, Fig. l(c). These results are borne out by the principal component spectra, the eigenvectors produced by PCA, Fig.2(a) and (6). The first eigenvector gives high weighting to abs,orption above 3000 cm-* compared with absorption below this value, and the second eigenvector gives emphasis to absorption at 1650-1750 cm-1, the region of the C=O stretch. Moving to the third, fourth and subsequent principal components distinguishes further features and displays other structure-spectrum correla- tions. Similar use of eigenvectors has been proposed to interpret near infrared spectra.2-3 0.3 0.2 - 0.1 6 2 g o c a W .- -0.1 -0.2 -0.3 1 i 1 I I 4000 3000 2000 1000 600 0.6 0.4 - 5 0.2 g o 0 > iij -0.2 -0.4 4000 3000 2000 1000 600 Wavenumber/cm-' Fig. 2 ( a ) First and ( 6 ) second eigenvector spectra The new uncorrelated and reduced number of factors afforded by PCA can be used as input data to cluster analysis software to demonstrate the potential for machine identifica- tion of spectra.Fig. 3 illustrates the dendrogram resulting from a hierarchical cluster analysis performed on 50 object spectra using the first ten principal component loadings as variables. I t is often the case that a group or cluster of objects as described by their selected pattern vector is not a clear, distinct and well resolved class but a fuzzy set, e . g . , Fig. l(6) and (c). There is increasing interest in classification with such fuzzy clusters. A fuzzy pattern recognition algorithm applied to an object yields the degree of membership, the so-called member- ship function, of that object in a class. A number of fuzzy clustering algorithms and techniques have been proposed in the literature.Ruspini4 introduced the notion of a fuzzy partition t o represent the clusters in a data set. In practice, the fuzzy clustering algorithm of Bezdek and Dunn' is better employed as it provides not only the membership functions of the objects but also the most representative elements of t h e data set, i . e . , the cluster centres. Starting with an estimated fuzzy partition of the data, the weighted mean (centre) of each cluster is calculated and a new partition constructed from the sum of squares of the difference in the distance between eachANALYTICAL PROCEEDINGS, MAY 1991, VOL 28 149 A A A A A A A A A L G G E B B B D B I I I G A A D E B B B C E C E C E F D D E F F E E E D D D D D D D A D D 4! 7- T L I Fig.3 Dendrogram obtained from a pair group average of the first ten principal components of the infrared spectra point and the centre. This process is repeated until subsequent changes in the partition are less than a predefined minimum. Fig. 4 illustrates the membership function of each object with respect to a group for the infrared data. An advantage of such a 0.07 0‘18 J.UY 0.12 -50 1 &ll 0.19 -100 1 I 1 0 100 200 300 400 PC I Mcmbcrship function for each object as a member of t h e Fig. 4 carbonyl clustcr. calculated using t h c fuzzy group algorithm fuzzy set representation in cluster analysis is that stray points and points isolated between clusters are classified as such and a quantitative measure of their membership of a group is calculated.One of the aims of cluster analysis is to provide a simplified, often diagrammatic, representation of multivariate data. Unf- ortunately, such simplification can distort the data and there is often no indication provided as to the degree of this distortion or the degree of fit of a classification model to a set of data. For dendrograms the cophenetic correlation coefficient can pro- vide an indication of the efficiency of a clustering technique. Briefly, by reading the apparent similarity values from the dendrogram and comparing these with the original calculated data, a linear correlation diagram can be plotted and a correlation coefficient computed. The higher this correlation coefficient the better and less distorted the dendrogram is in describing the data.References 1 2 3 4 5 Adams. M. J.. and Black, I.. A n d . Chim. Actu, 1986, 189,353. Cowe. I., and McNicol. J., Appl. Spcctrosc.. 1985.39.257. Devaux. M., Appl. Spectrosc.. 1988. 42, 1015. Ruspini, E. H.. Inf. Control, 1969. 15, 22. Bezdek, J. C., and Dunn. J . C., IEEE Trans. Comput.. 1975, C-24. 835.150 ANALYTICAL PROCEEDINGS, MAY 1991, VOL 28 0.8 0.6 0.4 0.2 2 0 - -0.2 -0.4 -0.6 -0.8 Principal Component Analysis: An Introduction - - - - - - - - I l l 1 I I I I I Gordon L. Smith Statistics for Industry Ltd., Knaresborough, North Yorkshire HG5 9EU Aims of Principal Component Analysis Principal component analysis (PCA) is one of the most widely used methods of extracting and interpreting information from multivariate data. The raw data consist of many variables, which might be absorbances at different wavelengths of a near infrared spectrum, or amounts of different constituents found in samples for analysis.Within the data set there might be intercorrelations among the variables. t sc" PCl 1 Xl -+ Fig. 1 Simple illustration Principal component analysis does the following: ( a ) trans- forms the original variables to new axes ('principal com- ponents') which are orthogonal, so that the data expressed as scores in those axes are uncorrelated with each other; (b) expresses as much as possible of the total variation in the data in only a few principal components (PCs); and (c) each successively derived PC expresses decreasing amounts of the variation. Table 1 Data Sample x1 x2 x3 x4 x5 xg x7 x8 x9 x10 1 0 0 4 90 1.5 9 4.9 16 89 490 2 21 21 9 85 3.0 22 3.1 22 105 608 3 0 10 4 12 1.6 10 5.2 21 120 517 4 21 26 22 53 4.1 12 5.9 21 118 466 5 19 28 10 49 1.0 15 2.1 11 72 302 6 18 41 16 72 2.0 9 3.8 9 95 574 7 0 9 5 98 2.0 13 5.9 10 101 472 8 0 13 7 136 2.5 14 7.6 11 86 557 9 0 12 7 94 2.0 15 6.2 21 111 561 10 3 18 9 144 4.6 12 6.2 20 47 454 11 0 22 12 111 3.6 11 5.2 11 77 531 12 3 24 24 111 3.6 10 2.1 12 76 409 Mean 7 19 11 88 2.6 13 4.8 15 91 495 SD 9.5 9.6 6.8 37.5 1.14 3.6 1.73 5.2 21.3 83.7 By converting from original variables to PCs, intercorrela- tions are removed so that each PC conveys its own unique piece of information, and most of the information is conveyed by the first few PCs, so that the multi-dimensional data (which were too vast to be plotted) can be approximately represented in perhaps two or three dimensions (which can be visualized).It is then possible to explain variation among the samples in terms of PCs and original variables. Fig. 2 1.0 I I I two PCS Derivation of Principal Components Where the data consist of p variables x l , . . . , xp, the first PC is the linear combination a1x1 + . . . + a#p which has the highest variation (measured by the variance, or squared standard deviation), subject to the normalizing con- straint a12 + . . . + ap2 = 1 In the simple example shown in Fig. 1, the two variables are replaced by PC1, the direction of which is in the direction of greatest variation in the data. Table 2 Correlation coefficients x1 x2 X 3 1 .OO 0.74 1.00 0.44 0.69 1.00 0.04 0.18 0.54 -0.41 -0.15 0.02 0.39 -0.07 -0.23 -0.50 -0.53 -0.38 0.11 -0.33 -0.18 - 0.15 -0.17 -0.08 - -0.11 -0.11 -0.23 x4 x5 1 .00 0.50 1.00 0.07 -0.01 0.33 0.19 -0.25 0.22 -0.64 -0.33 0.15 0.08 1 .oo -0.14 1.00 0.36 0.19 1.00 0.13 0.15 0.39 1.00 0.24 0.41 0.28 0.45 1.00 The second PC is the linear combination, orthogonal to the first, subject to the same constraint, with the highest variation.Successive PCs are derived in the same way, required to be orthogonal to all the previous ones. The method of derivation is to obtain the eigenvectors of the variance-covariance matrix of the xs. In order to express all the variation in the data, it requires p PCs, but each successive PC is decreasingly important. It is often considered appropriate to standardize the original variables to reduce them all to the same scale.The values of each variable are standardized by subtracting its mean and dividing by its standard deviation (SD). Each standardized variable then has a mean of 0 and an SD of 1, and the covariances then become the correlation coefficients.ANALYTICAL PROCEEDINGS, MAY 1991, VOL 28 151 An Analysis The data used to illustrate the method, from previous work on the flavour of fish,' are amounts of ten non-volatile flavorous components in 12 samples of cooked cod. The data are shown in Table I , with the correlation soefficients in Table 2. 2 1 G o a - 1 -2 I** 3* 9* 4 I * e5 I *6 I I -1 0 1 2 PC1 Fig. 3 Principal component scores The samples can be plotted on the axes of the first two PCs, remembering that the representation is only approximate as PC1 and PC2 only express 53% of the total variance.The 'scores plot' is shown in Fig. 3. Samples 5. and 6 have high positive values on PCI; hence we expect' them to have high values on xl, x2 and x3 and low values on x4. Samples 2 and 10 are at opposite extremes of PC2; hence they are likely to differ in x4 and x9. Exact comparisons of samples can only be achieved by detailed examination of the tabulated data in all the original variables, but the scores plot allows indications to be seen graphically. 1 .o 0.8 0.6 0.4 (u 2 0.2 0 -0.2 -0.4 Principal component analysis was carried out on the data after standardization. As each standardized variable has a variance of 1, the total variance is 10. Table 3 shows the eigenvalues and the amount of the total variance expressed.The loadings are shown in Table 4. For example, PC1 = 0.42~1" + 0.52~2* + . . . where each xi* is xi after standardizing. The magnitudes of the coefficients show the relative importance of each variable in the determination of a PC: PC1 is mainly dependent on xl, x2, x3 and x7; PC2 on x4 and x9, and so on. -0.6 ' 1 1 1 -0.2 0 0.2 0.4 0.6 0.8 1.0 PC 1 Fig. 4 Correlations (data unstandardized) To Standardize or Not? In some disciplines it is always necessary to standardize data as some variables might be measured on different scales (mass in grams, height in metres, other measurements in millimetres, moisture content in per cent., etc.) and the relative magnitudes of the values would be affected. In other areas, larger variation in some variables than others would genuinely indicate greater importance.~ Table 3 Eigenvalues and variance expressed Principal component Eigenvalue 1 2.95 2 2.36 3 1.73 4 1.18 5 0.89 6 0.39 7 0.34 8 0.10 9 0.0s 10 0.01 Variance (YO) 29.5 23.6 17.3 11.8 8.9 3.9 3.4 1.0 0.5 0.1 Total 10.00 100.0 Cumulative variance (Yo) 29.5 53.1 70.4 82.2 91.1 95.0 98.4 99.4 99.9 100.0 100.0 The correlations between PCs and variables are obtained by multiplying each loading by the square root of the eigenvalue for that PC. These are often plotted in PC axes, as in Fig. 2. The length of each vector (line from the origin to a point) in the two dimensions of the plot indicates how much of that variable has been explained by the PCs in the plot, as each vector would have a length of 1 in the complete ten-dimensional space of all the PCs. x2 is highly correlated with PCl, and x4 with PC2. For both these variables the vectors are close to a length of 1, but for x6 the vector is short, suggesting that it has little to do with PC1 or PC2. Reference back to Table 4 confirms that xh is mainly related to PC4. Table 4 Principal component loadings PC1 0.42 0.52 0.45 -0.12 0.10 -0.04 -0.43 -0.20 -0.15 -0.27 PC2 0.33 0.02 -0.12 -0.53 -0.33 0.27 -0.15 0.31 0.52 0.15 PC3 0.22 0.12 0.24 0.26 0.57 0.35 0.18 0.41 0.06 0.40 PC4 -0.15 0.11 0.37 -0.25 0.16 -0.68 0.30 -0.01 0.40 0.18 The plot of the correlations using PCA on unstandardized data (Fig. 4) shows how PC1 would be completely dominated by x I o , the variable that had the highest variation in Table I , whereas x4 and xy, the others with high variation, had the major contributions. In fact, PC1 expressed 78% of the total variance, and the first two PCs expressed 96%. It is generally wise to standardize to avoid such a domination. Reference 1 McGill, A. S.. Howgatc, P., Thomson, A. B., Smith. G.. and Hardy. R., in Progress in Fluvour Research. ed. Adda. J.. Elsevier Applied Sciencc Publishers, Barking. 1984.

ISSN:0144-557X    

DOI:10.1039/AP9912800145

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

152 ANALYTICAL PROCEEDINGS, Polymerase Chain Reaction-Technology Applications MAY 1991, VOL 28 and ~ ~~~~~~~~~ The following are summaries of three of the papers presented at a Joint Meeting of the Western 9egion and the Biological Methods Group held on December I l t h , 1990, in the University of Wales College of Cardiff. Molecular Archaeology Erika Hagelberg Institute of Molecular Medicine, John Radcliffe Hospital, Oxford OX3 9DU In recent years, rapid technical developments in the field of molecular biology have enabled genetic information to be retrieved from museum specimens and archaeological remains, Higuchi et al.' succeeded in cloning mitochondria1 DNA sequences from a museum specimen of a quagga, an extinct member of the genus Equus. Shortly afterwards this was followed by the cloning of repetitive human DNA sequences from an ancient Egyptian mummy by Paabo.2 This early work in molecular archaeology involved the molecular cloning of the extracted DNA into bacterial vectors, and as such was subject to serious difficulties owing to the degradation and chemical modification of the ancient DNA and to artifacts introduced during cloning.3.4 These difficulties have been overcome by the introduction of the polymerase chain reaction or PCR,S as this technique is extremely sensitive and can be used for degraded DNA samples, also being suitable for the analysis of chemically modified DNA as it selects undamaged target molecules for replication.PCR has been used to analyse a variety of archaeological remains, including a 7000 year old human brain preserved in a peat bog,6 museum skins,7 a 17-20 million year old magnolia leafs and ancient bone.9 The analysis of bone is particularly useful in archaeology owing to the relative abundance of skeletal remains compared with bog bodies or mummified tissues.The methods for the extraction and analysis of DNA from archaeological bone samples are described briefly below. Experimental DNA Extraction Bone fragments were cleaned by blasting with fine aluminium oxide in an Airbrasive unit (S.S. White Industrial Products, Piscataway, NJ, USA) and then reduced to a fine powder in a refrigerated freezer mill (Spex Industries, Edison, NJ, USA). Typical DNA extractions were 1.2-1.5 g of powdered bone in 10 volumes of a lysis buffer containing 0.5 M EDTA (pH 8.5), 100 pg ml-1 proteinase K and 0.5% N-lauroylsarcosine.*O Samples were incubated overnight at 37 "C with agitation, and then extracted with an equal volume of phenol, followed by two phenol-chloroform extractions and a final extraction with chloroform to remove trace amounts of phenol.The aqueous phase was concentrated and desalted by centrifugation-driven dialysis usigg Centricon 30 microconcentrators (Amicon). The concentrated extracts (100-200 PI) were stored at -20 "C. Polymerase Chain Reaction Amplification of DNA sequences are carried out by the method recommended by Perkin-Elmer Cetus using 2 units of Thermus aquaticus (Taq) polymerase per 25 p1 of reaction mixture. A substance inhibitory to PCR was present in most of the ancient bone extracts, but inhibition could be overcome in most instances by the addition of bovine serum albumin (160 mg ml-1) or by using less template.Oligonucleotide primers were made with an Applied Biosystems DNA synthesizer and used for PCR and sequencing without further purification. Several pairs of mitochondria1 DNA (mtDNA) primers were used successfully for the amplification of specific DNA fragments from ancient bone extracts, including the high conserved primers described by Kocher et al. ,*I which specify a 375 base pair (bp) fragment of the mtDNA cytochrome b gene and a 449 bp fragment of the mtDNA 12s rRNA gene. Other primers used were the primers A and B which amplify a 121 bp fragment of region V12 and primers for the hypervariable region of the human mtDNA genome.13 The PCR products were sequenced directly using a modifica- tion of the unbalanced PCR method of Gyllensten and Erlich14 to generate a single-stranded sequencing template.Sequencing was performed by dideoxy chain terminationl5 using Sequen- ase (US Biochemical) and the primer which was limiting in the asymmetric amplification. The sequencing products were electrophoresed on 6% denaturing polyacrylamide gels and autoradiographed for 2-4 d. Results and Discussion Fragments of mtDNA ranging from 112 to 600 bp have been amplified from extracts of archaeological human and pig bones ranging in age from 300 to over 5000 years. DNA fragments exceeding 1 kb have been amplified from forensic human bone samples buried for approximately 10 years. Very strong inhibition of PCR was observed in a number of the bone extracts but, as mentioned above, this could be overcome in many instances with bovine serum albumin or by diluting the template until the inhibition was abolished.The main point to be noted in the application of PCR to archaeological research is that the sensitivity of the technique makes the amplification of modern contaminating DNA very likely. Precautions must be taken to avoid and monitor possible sources of DNA contamination, including the use of sterile equipment and reagents and suitable DNA extraction and PCR blank controls. The amplification and sequencing of DNA from animal bones are also useful controls, as human DNA is the most likely source of contamination. We verified our techniques by the amplification and sequencing of the 375 bp fragment of cytochrome b gene from a 400 year old pig femur, and the correct pig sequence was obtained although no modern pig DNA or pig tissue had been handled in our laboratory.16ANALYTICAL PROCEEDINGS, MAY 1991, VOL 28 153 The possibility of amplifying specific DNA fragments from archaeological bone has implications for numerous areas of research.Applications in archaeology and anthropology include the study of population movements and migrations, kinship relationships and palaeopathology. Phylogenetic studies on living organisms will be enhanced by the ability to generate sequence data from extinct animals, although the possible survival of DNA in fossilized bone still remains to be determined. Although developed within the context of archaeological research, these techniques have already been applied in our laboratory to forensic cases (Jeffreys and Hagelberg, in preparation), despite the fact that until recently the analysis of DNA from skeletal remains was held to be impossible.17 References Higuchi, R., Bowman, B., Freiberger, M., Ryder, 0.A., and Wilson, A. C., Nature, London, 1984,312, 282. Paabo, S., Nature, London, 1985, 314, 644. Paabo, S., Higuchi, R. G., and Wilson, A. C., J. Biol. Chem., 1989, 264, 9709. Paabo, S., and Wilson, A. C., Nature, London, 1988,334,387. Saiki, R. K., Scharf, S., Faloona, F. Mullis. K. B.. Horn, G. T., Erlich, H. A., and Arnheim, N., Science, 1985, 230, 1350. 6 7 8 9 10 11 12 13 14 15 16 17 Paabo, S., Gifford, J. A., and Wilson, A. C., Nucleic Acids Res., 1988.16, 9775. Thomas, R. H., Schaffner, W., Wilson, A. C.. and Paabo, S., Nature, London, 1989, 340,465. Golenberg, E. M., Giannasi, D. E., Clegg, M. T., Smiley, C. J., Durbin, M., Henderson, D., and Zurawski, G., Nature, London, 1990, 344, 656. Hagelberg, E., Sykes, B., and Hedges, R., Nature, London, 1989,342,485. Maniatis, T., Frisch, E. F., and Sambrook, J., Molecular Cloning: a Laboratory Manual, Cold Spring Harbor Labora- tory, Cold Spring Harbor, NY, 1982. Kocher, T. D., Thomas, W. K., Meyer, A., Edwards, S. V., Paabo, S., Villablanca, F. X., and Wilson, A. C., Proc. Natl. Acad. Sci. USA, 1989,86,6196. Wrischnik, L. A., Higuchi, R. G., Stoneking, M., Erlich, H. A., Arnheim, N., and Wilson, A. C., Nucleic Acids Res., 1987, 15, 529. Vigilant, L., Pennington, R., Harpending, H., Kocher, T.D., and Wilson, A. C., Proc. Natl. Acad. Sci. USA. 1989,86,9350. Gyllensten, U. B., and Erlich, H. A., Proc. Natl. Acad. Sci. USA, 1988,85, 7652. Sanger, F., Nicklen, S., and Coulson, A. R., Proc. Natl. Acad. Sci. USA, 1977, 74, 5463. Hagelberg, E., andClegg, J. B., Proc. R. SOC., Ser. B , 1991, in the press. Iscan, M. Y., Yearb. Phys. Anthropol., 1988, 31, 203. Strategies for the Use of the Polymerase Chain Reaction in Forensic An a I ysis K. M. Sullivan, G. Tully, R. Hopgood, D. Lingard and P. Gill Central Research and SUDDO~~ Establishment. Home Office Forensic Science Service, Aldermaston, Reading, Berkshire RG7 4PN Since its introduction in 1985, the technique of DNA profiling has transformed forensic biology. There are, however, three areas in which the current analytical procedures could be improved : (i) sensitivity of the test: currently 100 ng of DNA or more are required for analysis with single-locus probes (SLPs); (ii) ability to type degraded samples: SLPs typically detect bands several kb in size, which are susceptible to degradation; (iii) analysis time: multiple sequential analyses with SLPs especially when characterizing weak samples can take several weeks. These problems could be alleviated by adopting PCR (poly- merase chain reaction)-based tests.Strategies for the forensic use of PCR involve the detection of either sequence or size variations of particular segments of DNA which differ between individuals. The three approaches we have investigated are dot-blot analysis of the polymorphic locus HLA DQa, amplification of variable-number tandem repeat (VNTR) loci and direct sequencing of PCR products.This paper discusses the relative merits of these approaches. HLA DQa Dot-Blot Analysis The first analytical test with a potential forensic application was developed by Cetus based on the HLA (human leucocyte antigen) DQa gene.1 In our hands this first system was prone to typing ambiguities2 and Cetus have subsequently developed an improved test (Amplitype) that detects and distinguishes between six of the eight possible DQa alleles using allele- Crown copyright. specific oligonucleotide (ASO) probes.3 In essence, the test comprises amplification, with biotinylated primers, of a 242 base pair (bp) fragment which spans the region of sequence variation.This is then characterized using a reversed dot-blot hybridization reaction in which the probe and not the amplified DNA is immobilized on a nylon membrane. Hybridization is detected by a colour reaction mediated by attachment of a streptavidin-horseradish peroxidase conjugate to the biotiny- lated PCR product. Evaluation of this system on old casework samples has given good results with blood and semen stains and with mixtures of body fluids, such as preferentially extracted vaginal swabs. The advantages of the test are that it can be performed within a working day, it will type as little as 2 ng of template DNA and it is robust in operation as it is not necessary to determine the amount of substrate DNA prior to amplifica- tion. The drawbacks are that the test is more expensive than other PCR analyses and the discrimination power (DP), depending on race, is only ca.0.93. Amplification of Variable-number Tandem Repeat (VNTR) Loci A second PCR approach that we have investigated is the amplification of small VNTR loci. Primers were used that bound to regions flanking the tandemly repeated sequence then, following PCR, the products were resolved according to size by gel electrophoresis. We have concentrated on two loci, D17S5 and DlS80, which are defined by probes pYNZ22 and pMCTll8, respectively .4,5 Amplification products from these loci have been detected by both raclioactive hybridization and direct visualization of bands. A database from a small British population was generated as follows: 1 ng samples of total154 ANALYTICAL PROCEEDINGS, MAY 1991, VOL 28 genomic DNA were amplified through 25 cycles, then resolved electrophoretically in 1.2% agarose, Southern blotted and hybridized with a radiolabelled pYNZ22 probe.The detected alleles contained between 1 and 14 repeat units (Fig. l ) , ranging in size between 169 and 1079 bp, with an over-all heterozygosity of 88%. The DP of this locus was determined to be 0.96. The advantages of amplifying VNTRs are that it is an extremely sensitive test and results are obtainable from single copies of template DNA. A drawback is that they are prone to generate spurious bands unless the reaction conditions are 0.35 1 1 1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 1 4 Allele number D17S5 PCR allele frequency data (n = 76) Fig.1 carefully optimized and the number of amplification cycles is minimized. Further, to obtain probabilities of chance associa- tion of 10-6, it will be necessary to multiplex up to six comparable VNTR loci. An alternative approach is to co- amplify several microsatellite loci, which offer similar DPs to the aforementioned VNTR loci.6 Alleles from microsatellites differ by 2, 3 and 4 bp repeats and require resolution on long acrylamide gels. The small size ranges of these loci make microsatellites ideal for multiplex PCR and for analysing highly degraded samples. The main drawback, however, is the production of sub-bands due to ‘stuttering’ of the DNA polymerase during replication, which could make the inter- pretation of results difficult, especially when analysing mix- tures.Direct Sequencing of PCR Products We are currently investigating DNA sequence differences between individuals within part of the mitochondria1 genome as a method of identification.’ Mitochondria1 DNA is present at a copy number of 1 ~ 1 O O O O per human cell. It is maternally inherited and possesses a rapidly mutating non- coding region, which could be used for determining genetic individuality and maternal relatedness. The strategy we have used is as follows: the entire non-coding region was amplified with primers which are complementary to the flanking tRNAthr and tRNAphe gene.8 The 1.3 Kb PCR product was gel purified, then amplified in a second nested asymmetric PCR reaction to generate a 464 bp product. The reaction-limiting primer was chimaeric, consisting of the M13-21 universal sequencing primer at the 5’ end coupled to a sequence complementary to a relatively invariant region of the D-loop.The mixture was then sequenced directly using fluorescently tagged M13(-21) uni- versal sequencing primers in conjunction with an Applied Biosystems 370A automated sequencer. This sequence data was verified by asymmetrically amplifying and sequencing the complementary strand. The advantage of this system is that the data are definitive and sequence comparisons between individuals within this segment are likely to confer a DP of 0.997.8 It is also ideal for the analysis of hair shafts, which contain significant amounts of mitochondrial DNA but little chromosomal DNA. The draw- backs are that it is technically demanding and relatively labour intensive. However, with the continuing improvements in instrument performance, introduction of robotics into mol- ecular biology and rapid evolution of PCR techniques, such forensic techniques, will soon be routine and highly auto- mated. References Saiki, R.K., Bugawan, T. L., Horn, G. T., Mullis, B., and Erlich, H. A., Nature, London, 1986, 324, 163. Westwood, S. A., and Werrett, D. J . , Forensic Sci. Int., 1990,45, 210. Saiki, R. K., Walsh, P.. Levenson, C., and Erlich, H., Proc. Natl. Acad. Sci. USA, 1989,86, 6230. Horn, G., Richards, B., and Klinger, K. W., Nucleic Acids Res., 1989, 17,2140. Kasai, K., Nakamura. Y., and White, R., Am. J . Hum. Genet., in the press. Weber, J., and May, P., Am. J. Hum. Genet., 1989,44, 388.Sullivan, K. M., Hopgood, R., Lang, B., and Gill, P., Electrophoresis, in the press. Orrego, C., and King, M., in YCK Protocols, a Guide to Methods and Applications, ed. Innis, M. A., Academic Press, London, 1990, pp. 416-426. Strategies of the Polymerase Chain Reaction Alan D. B. Malcolm Department of Biochemistry, Charing Cross and Westminster Medical School, Fulham Palace Road, London W6 8RF Nucleic acids have become the latest family of macromolecules to be used as analytical reagents. Compared with their predecessors (enzymes and antibodies), nucleic acids have the supreme advantage that once the primary sequence of the bases has been determined, the total three-dimensional struc- ture of the molecule can be predicted. Additionally, and even more importantly, the sequence of one strand of the nucleic acid allows the sequence of the complementary strand to be defined, which will then function as a reagent for the analyte.This particular property is unique to the nucleic acids. As for all analytical techniques, the properties being sought include: versatility (the range of similar analyses that can be carried out using essentially the same technology); sensitivity (how little one can detect); precision (how good is the discrimination between closely related analytes) and speed (which in practice means automation and hence economy). The sensitivity of any of the various analytical techniques of course depends partly on the method of detection employed. The use of the radioactive isotope 32P, detected either by scintillation counting or by autoradiography, has made DNA- based methodology among the most sensitive currently avail- able, being routinely capable of detecting as little as a few attomoles of any given sequence.While this corresponds to the amount of a single gene which can be extracted from a few millilitres of blood, it still requires between 105 and 106 copies of the particular sequence. Table 1 Optimization of PCR 1 Primer concentration 2 Template concentration 3 Template purity/template integrity 4 Mg*+ concentration 5 Annealing time and temperature 6 Extension time 7 Enzyme concentration 8 DNA denaturation timeANALYTICAL PROCEEDINGS, MAY 1991, VOL 28 Within the last five years the development of the polymerase chain reaction (PCR) for the amplification of particular DNA sequences has totally transformed this sensitivity.’ The tech- nique requires the ready availability of oligodeoxynucleotides with defined sequences.These have become readily avajlable during the last decade owing to the development of automated methods for the exploitation of the phosphoramidite coupling method. Additionally, the technique requires a DNA poly- merase molecule capable of a high rate of processing and low error incorporation, and lacking significant amounts of deoxy- ribonuclease activity. Originally the fragment produced by proteolysis of DNA pol 1 (often called the Klenow fragment) satisfied this particular condition, but an early major break- through was made with the re-discovery of purified DNA polymerases from thermophilic organisms such as Thermus aquaticus or Bacillus stearothermophilus .2,3 The existence of such thermostable enzymes able to withstand the temperatures required for the denaturation of DNA has enabled the PCR to be automated by the use of thermal cycling equipment.The PCR has therefore been used for the diagnosis, detection and characterization of inherited disease (such as sickle cell anaemia4-6 and phenylketonuria7), diseases which are a result of mutations taking place in the somatic cell line after fertilization (such as the single base changes caused by ultraviolet light, X-rays and carcinogenic chemicals), and diseases which can result in tumours such as the bladder tumour caused by the ras oncogene mutation.8.9 Additionally translocations whereby sections of chromosomes not normally adjacent to each other become rearranged can be studied. These appear to be involved in the unusual activation of particular genes found in cancers such as Burkitt’s lymphoma, etc.10 It can also be used to detect infectious agents such as virusesl1.12 including human papillomavirus13 and micro- organisms such as bacteria.14 Table 2 PCR artefacts Carryover- Plasmids Bacteria PCR products Sources- Pipetman Fingers Reagents Samples Centrifuge Speedvadvacuum bottles Microtome blade If each cycle of the PCR is working at theoretical efficiency (i.e., a 2-fold amplification per cycle), 20 cycles will lead to a 1 million-fold amplification.In practice this theoretical yield per cycle is rarely achieved, and therefore between 25 and 30 cycles are normally used in order to produce a 1 million-fold amplification.This means that a single molecule or particle can then give rise to the between 100 000 and 1 million molecules required for detection by previously developed DNA technol- ogies. It also allows for the exploitation of non-radioactive labelling methods such as fluorimetry, luminometry, biotin- streptavidin assay followed by colorimetry, and digoxigenin treatment followed by antibody detection using alkaline phosphatase as a label. In spite of the apparent simplicity of the technique, a variety of parameters require to be optimized for each gene and pair of primers. These are shown in Table 1. In practice the extraordinary sensitivity of the method can turn out to be a disadvantage as it allows the possibility of amplifying artefacts and/or impurities (Table 2).This means that several precautions and controls need to be taken in order to avoid misleading results. These are shown in Table 3. 155 Table 3 PCR precautions (a) Physical separation (b) Separate supplies/equipment (c) Disposable pipettes (d) Positive displacement pipetman (e) Aliquot reagents (f) Meticulous laboratory behaviour (g) Choice of positive controls (h) Trustworthy samples (i) Inclusion of negative controls On those occasions where the exact sequence of the DNA to be amplified is not known (because only the protein sequence is known and the genetic code must therefore be used), the problems of degeneracy can be reduced by using inosine wherever a purine (A or G) is required. Single base changes can either be identified by cleavage of the product by using a restriction enzyme or by differential oligonucleotide hybridiza- tion to the product, or by a technique known as ARMS (amplification refractory mutation system).Amplification can be made to fail by introducing the mismatch at the 3’ end of the primer. If, on the other hand, it is wished to improve the stability of the primers being used for amplification, the use of 2,6- diaminoadenine in place of adenine allows three hydrogen bonds rather than the more normal two to be formed to thymine, thereby improving the stability and hence discrimina- tion of this primer.l”l6 It is not only DNA which can be amplified. It is perfectly possible by using reverse transcriptase in a first cycle to begin with an RNA species which can then be converted to DNA and amplified as before.This type of approach has been used in order to characterize the clonal nature of the B-lymphocyte population in auto-immune diseases such as rheumatoid arthritis. 17 Within five years of its discovery, the sophistication of the PCR has become such that it is now impossible to imagine modern molecular biology without it. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 References Erlich, H. A., Gelfand, D. H., and Saiki, R. K., Nature (London), 1988,331,461. Chien, A., Edgar, D. B., and Treia, J. M., J. Bacteriol., 1976, 127, 1550. Tindall, K. R., and Kunkel, T. A., Biochemistry, 1988, 27, 6008. Saiki, R. K., Scharf, S., Faloona, F., Mullis, K. B., Horn, G. T., Erlich, H. A., and Arnheim, N., Science, 1985, 230, 1350. Embury, S. H., Scharf, S. J., Saiki, R. K., Gholson, M. A., Golbus, M., Arnheim, N., and Erlich, H. A., New Engl. J. Med., 1987, 316, 656. Saiki, R. K., Chang, C.-A., Levenson, C. H., Warren, T. C., Boehm, C. D., Kazazian, H. H., and Erlich, H. A., New Engl. J. Med., 1988, 319, 537. DiLella, A. G.. Huang, W.-M., and Woo, S. L. C., Lancet, 1988,497. Kumar. R., and Barbacid, M., Oncogene, 1988,3, 647. Farr, C. J.. Saiki, R. K., Erlich, H. A., McCormack, F., and Marshall, C. J., Proc. Natl. Acad. Sci. USA, 1988, 85, 1629. DNA Probes. Applications in Genetic and Infectious Disease and Cancer, ed. Lerman, L. S., Current Communications in Molecular Biology, Cold Spring Harbor Laboratory, 1986. DNA Technology and Rapid Diagnosis of Infection, Lancet, 1989, 14, 897. Lee, P. C.. and Hallsworth, P., Br. Med. J . , 1990,300, 1413. Ward, P., Parry, G. N . , Yule, R., Coleman, D. V., and Malcolm, A. D. B., Cytopathology, 1990, 1, 19. Olive, M. D., Atta, A. I., and Setti, S. K., Mol. Cell. Probes, 1988, 2, 47. Voss, U. B., Chollet, A., and Malcolm, A. D. B. Biochern. SOC. Trans., 1989, 17, 913. Hoheisel, J. D., and Lehrach, H., FEBS Left., 1990,274, 103. Haynes, G. A., Maini, R. N., and Malcolm, A. D. B., Ann. Rheum. Dis., 1990,49,460.

ISSN:0144-557X    

DOI:10.1039/AP9912800152

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

156 ANALYTICAL PROCEEDINGS, MAY 1991, VOL 28 Equipment News Inductively Coupled Plasma Mass Spectrometer The TS Sola is now available from Turner Scientific, a VSW Company. It can be used for direct elemental analysis of both liquid and solid samples, in which it will spectra of organics commonly confronting analytical chemists in environmental, forensic, flavours and fragrances, chem- ical process control and other investiga- tive analytical laboratory applications. It is the first large library produced using TS Sola inductivefy coupled plasma-mass spectrometer measure major and minor components. The inductively coupled plasma facility is fully automated and is controlled by menus of a 386 based desk-top computer system. The driving software is written in a window style format and is multi-task- ing.Turbo molecular pumps are used throughout the high vacuum section. VSW UK Sales, Albert Drive, Burgess Hill, West Sussex RH15 9NX. Software for Inductively Coupled Plasma Spectrometer The PS 2.0 software package for the makers’ PU7000 ICP spectrometer features an intelligent automation capa- bility. In a single key stroke the user can initiate a sequence of pre-programmed operations, and the instrument will carry out up to 240 individual analyses automat- ically. The ICP analyst will benefit from a powerful programming facility , improved data presentation and a customized wavelength library. Philips Scientific, Analytical Division, York Street, Cambridge CB1 2PX. FTIR Library A new ETIR library produced on the HP 59970B IR detector and optimally for- matted for searching with the HP 59970C IRD Chemstation features over 5000 IR high-resolution chromatography; 99% of the data have been run since January 1990.Also available is Volume 3 of The Aldrich Library of FTIR Spectra, bring- ing the number of spectra in the three- volume set to over 17200. The second edition of the Sigma-Aldrich Library of Chemical Safety Data, a reference for the safe handling of over 14500 chemicals representing 24000 Sigma and Aldrich products is also announced. Aldrich Chemical Co. Ltd., The Old Brickyard, New Road, Gillingham, Dorset SP8 4JL. Spectroscopy Software The PECSS computerized spectroscopy software package controls the Lambda ultraviolet-visible and NIR spectrometers from an industry standard PC or compat- ible. The new Version 4.0 has a number of changes which greatly enhance perfor- mance. In particular, PECSS now per- forms a virus test automatically each time the software is invoked, and it supports colour printing using a compatible printer. The PECOL colour measure- ment package accepts PECSS data and runs on most industry standard PCs equipped with a high-resolution graphics card.PSEARCH software extension module for PECSSPECOL software packages is a comprehensive package for ul traviolet-visible spectral library hand- ling and searching. PPLOT offers exten- sive plotting capabilities, and PMULT provides analysis of up to ten components for quantitative chemical analysis of mix- tures. Perkin-Elmer Ltd., Maxwell Road, Beaconsfield, Buckinghamshire HP9 1QA.Au tosampling System The Lambda 2 autosampling system is designed for unattended, routine analysis of up to 137 samples and is ideal for pharmaceutical and chemical quality assurance. The system consists of the Lambda 2 ultraviolet-visible spec- trometer and the Model AS 90 microp- rocessor-controlled, multi-purpose auto- sampler, linked to one of three compat- ible sippers. Perkin-Elmer Ltd. , Maxwell Road, Beaconsfield, Buckinghamshire HP9 1QA. Spectrophotometer The Lambda 16HP double-beam, scan- ning ul traviole t-visi ble spectrometer incorporates a high-performance pre- monochromator optical system, giving high linearity, low stray light characteris- tics and a scan speed of 1440 nm min-1. Fully PC-controlled, it uses UVCSS ultraviolet computerized spectroscopy software with PPLOT, based on PECSS software.Data can be transferred to third party software packages. Perkin-Elmer Ltd., Maxwell Road, Beaconsfield, Buckinghamshire HP9 1QA. Spectrometer The Lambda 19 ultraviolet-visible-NIR spectrometer uses the double-beam, double monochromator system of the Lambda 9, with a wavelength range of 175-3200 nm and a linearity up to 6 A. It is controlled by an industry-standard PC, and the new ultraviolet computerized spectroscopy software (UVCSS) is stan- dard. The new standard PPLOT routine allows flexibility in printout and can be used with a wide variety of plotters and printers. A range of software packages and accessories is available. Perkin-Elmer Ltd. , Maxwell Road, Beaconsfield, Buckinghamshire HP9 1QA. Magnetic Stirrer for Spectrophotometers A magnetic stirrer has been added to the range of accessories for the Uvikon 900 Series of spectrophotometers.Designed to fit neatly under the standard cellANALYTICAL PROCEEDINGS, MAY 1991, VOL 28 157 holder, it plugs directly into the accessory interface for ease of application. The stirrer is of particular value where the samples are colloidal, have suspended particles or where an even temperature distribution is required within the cell. Both sample and reference positions are stirred. Kontron Instruments Ltd., Blackmoor Lane, Croxley Centre, Watford, Hert- fordshire WD1 8XQ. GC-MS System Enhancements A PC-based GC-MS system is now avail- able with a number of significant enhance- ments, including the first mass spec- trometry software running under Micro- soft Windows 3.0.The software can run alongside other Windows 3.0 programs. Faster integration capabilities and hard- ware improvements increase analytical performance and speed. A powerful new custom report generator is included. Vari- ous system configurations are offered to meet specific requirements, with con- siderable cost savings. Hewlett-Packard S.A., 150 Route du Nant-d’Avril, CH-1217 Meyrin 2, Geneva, Switzerland. Chromatography Automation Systems Based on IBM personal computers and the multi-tasking Microsoft Windows operating environment, the AI-450 com- bines exceptional capability, flexibility and ease of use in one package. The system is available in three models: Model I, having one channel; Model 11, with two channels expandable to four; and Model 111, with two channels expandable to 12.Each channel supports one pump, one injector and one or two detectors. All models can be purchased either as a complete system or as a softwarehnter- face package for adding to an existing PS/2, PC/AT or 100% compatible com- puter. A brochure is available. Dionex (UK) Ltd., Camberley, Surrey. Peak Detector-Autozero Module PeakTrak is an on-line peak detector- autozero module easily connected to any HPLC, GC, IC or SFC system. Outputs are provided to operate a fraction collec- tor, valve or other device in response to the appearance of peaks at the detector, with programmable delay. Both visual and defeatable audio alarms signal the presence of each peak. A front panel meter shows the source or output signal in real time.Manual or pre-programmed autozero can be performed at any time. Axxiom Chromatography Inc., 11988 Challenger Court, Moorpark, CA 93021- 7122, USA. GC Analyser for Lead-free Petrol The PU4410/73 gasohols analyser can measure oxygenated compounds by capil- lary gas chromatography. The call for such a system stems from the introduction of lead-free petrol and the demand for octane-enhancing chemicals to replace lead alkyls; the most common of these are low relative molecular mass alcohols and ethers such as methanol, MTBE and TAME. Employing multi-dimensional techniques to combine the separating power of both polar and non-polar capil- lary columns, the instrument can com- plete a full analysis of the principal oxygenates in about 20 min, with good linearity and precision.Philips Analytical Chromatography, York Street, Cambridge CB12PX. Preparative Chromatography Columns A new range of columns are ideal for use in many EPA procedures. Manufactured from top quality borosilicate glass, all the columns have a minimum dead space column, a Rheodyne injector and all required connecting tubing. HPLC Technology Ltd., Wellington House, Waterloo Street West, Maccles- field, Cheshire SKll 6PJ. Silver Metal Membrane for HPLC Reliable test results are given by a silver metal membrane filter when used for HPLC. Made of pure metallic silver, having no extractables or detectable con- taminants, it requires no flushing, and it has been consistently shown to provide a clearer, steadier baseline than do nylon and PVDF filters.It is also non-adsorp- tive and non-absorptive. The membrane is currently offered in ratings of 0.2,0.45, 0.8, 1.2, 3.0 and 5.90 pm, with standard diameters ranging from 13 to 293 mm. A brochure is available. Osmonics Inc., 5951 Clearwater Drive, Minnetonka, Minnesota 55343, USA. below sintered disc and are fitted with stopcocks. The columns are available in The Series 6O0, over 150 standard combinations. hooks for safety and a choice of three Supercritical GO2 SFC and 621 SFC/GC are thoroughly optimized for SFC: a Pulseless, high-capacity Pump; Radlevs. Shire Hill. Saffron Walden. Essex CBll 3AZ. Chromatography Columns IonPac columns for the chromatography of ionic and polar compounds feature ion-exchange packings having three dif- ferent regions: an inert, non-porous, chemically and mechanically stable core, a surface-sulphonated region completely covering the core, and outer layers of permanently attached sub-micrometre ion-exchange MicroBeads where the actual chromatography takes place.A brochure describes the use of the columns in ion exchange, ion exclusion, ion pairing and ion suppression. It includes a selector guide for columns and detectors and a number of chromatograms. Dionex (UK) Ltd., Camberley , Surrey. Multi-dimensional Columns for Liquid Chromatography The OmniPac line of multi-dimensional columns combines the power of ion exchange and reversed phase and ion pairing into a single column. A brochure illustrates the use of the columns in a number of separations. Dionex (UK) Ltd., Camberley, Surrey.HPLC Detectors The 3D-HPLC System developed by Per- kin-Elmer features the TriDet tri-func- tional detector. This complete HPLC system, available at a price below that often charged for individual components, can monitor the progress of an HPLC separation using any or all of the three most commonly used detection modes (ultraviolet absorbance, fluorescence and electrolyte conductivity) simultaneously. The 3D system is supplied with the TriDet detector, the 100 Series pump, an HPLC compact high-performance oven; micro- processor controller; and easy-to-use soft- ware. A brochure is available. Dionex (UK) Ltd., Camberley, Surrey. HPLC De-gassing System The Thames Chromatography biocom- patible mobile phase handling system features the ability to be pressurized safely, a built-in de-gassing (sparging) capability, integral filtration to protect the pump from particulates, and bottles made of thick wall plastic to prevent breaking and allow safe pressurization.An integral part of the system is the gas distribution manifold, which incorporates one-way check valves, automatically preventing backflow and requiring no special attention when filling a bottle; it allows up to four bottles to be simul- taneously connected to the helium cylinder. The system is designed to be compatible with all mobile phases used in biological separations and will cut helium usage to one cylinder per year per solvent. Thames Chromatography, 16 Raymead Court, Maidenhead, Berkshire SL6 8TN. Capillary Electrophoresis System The Capillary Electrophoresis System I is designed with flexibility in mind, making it suitable for everything from basic research to methods development and automated routine analysis. It features menu-driven programming via a video display with full control of all instrument parameters from a single keyboard.Methods can be stored in non-volatile RAM, and schedules of methods can be set up so that the system will run experi- ments through the night. A brochure gives details of the various ways of inject- ing samples and the ultraviolet-visible and fluorescence detectors. Dionex (UK) Ltd., Camberley, Surrey.158 ANALYTICAL PROCEEDINGS, MAY 1991, VOL 28 Protein and Drug Testing System The Array 360 is an automated system for disease state diagnosis and monitoring. It will perform high quality protein analysis and monitor therapeutic drug levels.It stores up to 60 chemistries in its memory. A wide range of specific protein kits is available, and the selection of therapeutic kits includes the drugs phenytoin, gen- tamicin, phenobarbital, theophylline, tobramicin, quinidine and carbamaze- pine, with valproic acid, NAPA, procain- amide and amikacin available shortly. Beckman, Progress Road, Sands Indus- trial Estate, High Wycombe, Bucking- hamshire. Clinical Analyser The Synchron CX7 offers a new concept in streamlined parallel processing and random access in one integrated system. four ranges as low as 0-1 ppm. It offers full-scale accuracy of +2% and fast response (goyo of 1000-10 o0O ppm in less than 10 s). The Model 3160 is suitable for on-line monitoring of trace amounts of oxygen in nitrogen, argon, hydrocarbons, hydrogen and many other pure gases and gas mixtures.'Teledyne Analytical Instruments, The Harlequin Centre, Southall Lane, Southall, Middlesex UB2 5NH. On-line Oxygen Analyser The Model 316R monitors trace oxygen (02) contamination in nitrogen, argon, helium and many other pure gases and gas mixtures from ppm to ppb levels. It features four full-scale switch-selectable ranges: &lo, 0-100,0-1000 and 0-10000 pprn 02. Also available are optional ranges as low as &1 ppm. Beckman Synchron CX7 Over 30 chemistry checks ensure highly accurate answers, and the system processes general chemistries, electro- lytes, STATs and special chemistries. Its unique 45 s STAT interrupt makes it ideal for on-call and emergency use.With consolidation of serum, plasma, urine and CSF chemistries for routine profiles, metabolic studies, specific proteins, drugs of abuse and TDM monitoring, there are over 100 user-defined procedures avail- able. Additional QC validation data can be provided by a menu of 40 custom calculations. Beckman, Progress Road, Sands Indus- trial Estate, High Wycombe, Bucking- hamshire. Oxygen Analyser The Model 3160 trace oxygen (02) ana- lyser incorporates a microprocessor that provides versatile control of user-select- able analogue or digital outputs and high and low alarms. The 19 in rack-mounted Model 3160 features AutoRanging over Teledyne Analytical Instruments, The Harlequin Centre, Southall Lane, Southall, Middlesex UB2 SNH.Chlorine Calibration Gas Generator The Model EC4 Minigen chlorine calibra- tion gas generator combines the advan- tages of portability with ease of operation and is especially useful for field chlorine gas detector response and alarm function checking. Bromine can also be generated from the electrochemical process employed. Bedfont Technical Instruments Ltd., Bedfont House, Holywell Lane, Upchurch, Sittingbourne, Kent ME9 7HN. Hazardous Gas and Fire Monitoring System The Gasmaster permanently installed hazardous gas and fire monitoring system can be specified with any combination of detectors for toxic or flammable gases, oxygen levels and fire. Systems with one to four channels are available. Each channel can operate either one remotely sited gas detector or one or two zones of up to 20 fire detectors. Bedfont Technical Instruments Ltd., Bedfont House, Holywell Lane, Upchurch, Sittingbourne, Kent ME9 7HN.Coulometer The DL37 KF Coulometer combines the precision of Karl Fischer coulometry with the operating convenience of modern analytical instruments. The coulometric determination of water content following Karl Fischer is primarily suitable for an extremely low range (from 10 pg to approximately 100 mg of water per sam- ple). In method development, defined specific conditions for an analysis can be combined as a method and stored in the DL37. Five methods are available as standard. Method development is facili- tated by the simple, logical method struc- ture and the graphics functions of the printer integrated with the DL37.Mettler-Toledo AG, CH-8606 Greifensee, Switzerland. Filtration Membrane Medium With their hydrophobic, PTFE mem- branes in strong polypropylene housings, TF microfiltration devices offer a versatile answer to many filtration problems. PTFE is ideal for organic solvent filtration and, when pre-wetted with ethanol, will also filter aggressive aqueous solutions, including strong acids and bases. Dispos- able Puradisc 25 TF, Polydisc TF and Polycap TF devices vary in their design and style of polypropylene housing,, All are supplied ready for use, and most are available in 0.1, 0.2, 0.45 and 1.0 pm porosities. Whatman Scientific Ltd., Whatman House, St. Leonard's Road, 20/20 Maid- stone, Kent ME16 OLS. Filter Funnels A range of funnels manufactured from heat-resistant borosilicate glass consists of three standard types: conical sintered filter funnels with cone, cylindrical sin- tered filter funnels with cone and cylin- drical filter funnel without cone.Each is available in a variety of sizes, cone sizes, porosities and with optional type side- arms. In addition to the standard range of 23 options, custom models can be pro- vided. Radleys, Shire Hill, Saffron Walden, Essex CBll 3AZ. Filter Set for Balances The determination of carbon black par- ticles of engines or fuels, of particles in waste water, lubricants, oils, etc., or the particle analysis of photocopier toners is usually performed with appropriate filters and balances such as the Mettler AT20. A kit has been developed for the ATANALYTICAL PROCEEDINGS, MAY 1991, VOL 28 159 balances for the weighing of filters (up to 105 mm in diameter).It covers most applications in particle determination. The constant load disk included in the set also allows weighing with the AT201, AT261 and AT200 higher-capacity balances. Two plates are enclosed that shield against electrostatic influences. The entire kit can be mounted simply and quickly. Mettler-Toledo AG, CH-8606 Greifensee, Switzerland. Osmometers Two Gonotec osmometers are available through Roth Scientific. They are of the vapour pressure type for relative mol- ecular masses up to 50 kD and in a membrane format for the range from 10 kD to 2 MD. Both osmometers have non-refrigerated and 6K-10 refrigerated models. Both are compact, benchtop centrifuges and are high-capacity instru- ments, accommodating volumes of 6 x 500 ml or 96 x 15 ml.A full range of angle and swing-out rotors is available. V. A. Howe and Co. Ltd., Beaumont Close, Banbury, Oxfordshire OX16 7RG. Furnace The AFW 1150 laboratory furnace con- tains a modular chamber of unique design which, together with programmable con- trols, is housed in a cabinet. The furnace provides a stable maximum operating temperature of 1150°C. Additionally it offers high-speed heating and cooling, and end-loss compensation is used to effect greater uniformity of temperature throughout the chamber. Routine main- Gonotec osmometers operating temperatures from ambient to 130 "C (adjustable in 1 "C increments) and are suitable for use with a wide range of solvents. Together they cover the full range of polymer relative molecular masses.Roth Scientific Co. Ltd., Alpha House, Alexandra Road, Farnborough, Hamp- shire GU14 6BU. Vacuum Pump Exhaust Filters A full range of filters designed to remove concentrations of very fine oil droplets (below 2 pm) emitted from vacuum pumps to acceptable atmospheric levels, as required by Health and Safety Regula- tions EH 40/90, is announced. The effi- ciency of the filter range, which accom- modates flow rates as low as 5 m3 h-1 up to 1450 m3 h-1, allows vacuum pump exhaust to be discharged directly into clean work areas when the nature of the exhaust gas permits. Balston Ltd., Monckton's Lane, Maid- stone, Kent ME14 2BR. Centrifuges The Howe-Sigma 6-Series Centrifuges consists of two instruments: the 6-10 tenance is simple.Chamber capacities of 4.5, 12 and 25 1 are offered. Lenton Thermal Designs Ltd., Unit C2, Valley Way, Welland Industrial Estate, Market Harborough, Leicester- shire LE6 7PS. Measuring Cell for DSC The DSC27HP measuring cell allows DSC measurements under pressure up to 70 bar and from room temperature up to 6OO"C, thus appreciably extending the application possibilities of the makers' TA4000 thermal analysis system. Mettler-Toledo AG, CH-8606 Greifensee, Switzerland. Drying Oven The DO337 drying oven for the makers' DL37 KF coulometer enables the pre- cision of Karl Fischer coulometry to be exploited for samples that are difficult to analyse, e . g . , for plastic granules, fibrous, pasty and tacky substances, or those that react with the titrant and thus falsify results.The oven can be heated up to 300 "C, evaporating the water in a sample; a stream of purge gas leads the water through a heated tube into the titration cell. Most functions of the DO337 can be controlled by the software of the DL37. Mettler-Toledo AG, CH-8606 Greifensee, Switzerland. Multi-channel Filter Unit The VBF21M unit offers simple manual adjustment of the cut-off frequency by easy-to-read thumbwheel switches. It is based on a standard 19 in frame into which nine channels can be fitted on plug-in cards. A compact four-channel unit is also available for benchtop use. Each plug-in filter card covers a cut-off span of 9900: 1, implemented as three decades with two decimal digits of resolu- tion. An optional computer interface allows full hexadecimal control to be used, giving a 16 500 : 1 cut-off range.Base frequencies of 0.1, 1 and 10 Hz are available, giving maximum cut-off fre- quencies in manual mode of 990 Hz, 9900 Hz and 99 kHz, respectively. Kemo Ltd., 9-12 Goodwood Parade, Elmers End, Beckenham, Kent BR3 3QZ. Waveform Monitors Two versions of a highly compact wave- form monitor module designed to allow display of optical signals on an oscillo- scope are available: the Model 755, de- signed to accept optical input signals at a wavelength of 850 nm, and the Model 756, which operates at a 1300 nm wavelength. Each incorporates an FC- type input connector and a pre-amplifier capable of boosting input signals at low levels from -23 dBm. The output of the pre-amplifier is delivered via a 50 i2 BNC-R connector, at a level suitable for input to a standard oscilloscope, where it can be displayed as a voltage signal.Hakuto International (UK) Ltd., Eleanor House, 33-35 Eleanor Cross Road, Waltham Cross, Hertfordshire EN8 7LF. Microsoft Windows Based Image Analysis System An engineering company needed a PC- based image analysis for the inspection and analysis of metal alloys. In addition, they wished to de-skill their quality assur- ance procedures, to have an interactive teaching aid using images and text for junior staff, to provide annotated images with descriptive text and supporting documentation on a PC as part of a total quality management policy, to have the ability to use their microscope and high quality monitor as a stand alone or as part of the complete image analysis system, to use the image analysis system in isolation or in combination with the image and text library-archiving system, to use the image and text library-archiving system independently and to produce high qual- ity technical reports.These requirements were satisfied by supplying MicroScale TC with the makers' Picture Book Profes-160 ANALYTICAL PROCEEDINGS, MAY 1991, VOL 28 sional, a complete package for Interactive Multimedia consisting of authoring soft- ware and a PC plug-in VGA compatible, transputer based image grabbing board. Digihurst Ltd., Newark Close, Roy- ston, Hertfordshire SG8 5HL. Guardbox Pelican’s new Guardbox is made of light- weight ABS structural foam resin and is rugged and tough enough to protect equipment or samples from any potential hazard.Ideal for transporting fragile or valuable samples and instruments through the roughest terrain, it is water- tight, airtight and dustproof. Weighing 14 oz, it measures 8% X 6% X 3% in. Radleys, Shire Hill, Saffron Walden, Essex CBll 3AZ. LIMS ICI has announced that it is entering the LIMS market with a high-performance VAX-based product. The new system, ACCOMPLIS, offers the benefit of hav- ing been developed by analysts for analysts and is already in full-scale use in seven major ICI laboratories. Intended for medium and large laboratories operat- ing in the research and quality control fields, it has been designed for flexibility, efficiency and convenience. ICI Chemicals and Polymers Ltd., P.O. Box No. 1, Billingham, Cleveland TS23 1LB.Literature Two data sheets show the compatibility of Chromacol’s standard glass and plastic chromatography vials and also the new ‘Gold’ range with the Beckman 500 and Jasco 851-AS autosamplers. Chromacol Ltd., Glen Ross House, Summers Row, London N12 OLD. Supelco’s 16-page Reporter, Volume IX, Number 5, features phase ratio, or 6, a dimensionless number that expresses the ratio of the internal diameter of a capil- lary column to the thickness of the sta- tionary phase; this ratio makes the choice of column dimension easier. Also dis- cussed are the new ORBO-47 absorbent tubes for absorption and desorption of phenol and cresol isomers, the PTE-5 column for environmental analyses, and GOW-MAC gas leak detectors. Supelco Inc., Supelco Park, Bellefonte, PA 16823-0048, USA.An application data sheet entitled ‘Rapid Spectropho tome tric Determination of Tocopherol (Vitamin E) in Blood Serum’ discusses the method and details of the apparatus, reagents and solvents used. Beckman, Progress Road, Sands Indus- trial Estate, High Wycombe, Bucking- hamshire. ‘Format’, Volume 4, Issue 4, Autumn 1990, gives information on recently issued software. Personal Bibliographic Software Inc., P.O. Box 4250, Ann Arbor, Michigan 48106, USA. A brochure describes the full range of simultaneous thermal analysers offered by PL Thermal Sciences. The analysers combine the two major thermal analysis techniques, thermogravimetry and dif- ferential thermal analysis, simultaneously on a single sample. Also available is literature on TA-MS, data-handling, TG, DSC, TMA, DTMA and DETA. PL Thermal Sciences, Polymer Labora- tories Ltd., The Technology Centre, Epi- nal Way, Loughborough, Leicestershire LEll QQE. Jeol’s new Electron Microscopy catalogue stresses the use of electron microscopes not merely for high magnification obser- vations but also as instruments for elemental analysis. Analytical options include X-ray spectrometry and wavelength dispersive spectrometry. Jeol (UK) Ltd., Jeol House, Silver Court, Watchmead, Welwyn Garden City, Hertfordshire AL7 1LT. The Whatman Laboratory Products Guide is a comprehensive directory to all the laboratory products currently avail- able from Whatman. There are extensive sections on filtration, micro-filtration and chromatography. Whatman Scientific Ltd., Springfield Mill, Maidstone, Kent ME14 2LE. A brochure describes the 350 range of pulseless HPLC pumps. Three models are available from isocratic to quaternary gradient. Applied Chromatography Systems Ltd., The Arsenal, Heapy Street, Macclesfield, Cheshire SKll 7JB. A brochure gives details of a range of kits and materials for the construction of molecular models. It also describes com- puter modelling and editing programs and a database of 50000 chemicals. Aldrich Chemical Co. Ltd., The Old Brickyard, New Road, Gillingham, Dorset SP8 4JL. The 1991 Nalgene Labware Catalogue details the complete line of over 500 products and includes product specifica- tions and technical information. Filter- ware, Cryoware, Unwire test-tube racks and right-to-know safety products are highlighted. Nalge Co., 75 Panorama Creek Drive, Box 20365, Rochester, NY 14602-0365, USA.

ISSN:0144-557X    

DOI:10.1039/AP9912800156

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

ANALYTICAL PROCEEDINGS, MAY 1991, VOL 28 161 Conferences and Meetings Fifth International LIMS Conference Dr. A. Cottam (Health and Safety Executive); a wide range of companies and the industry ‘COSHH in the Bioprocessing Industries’ now has a vast experience of the approach. Speaker to be announced; ‘Health Hazards of This seminar will bring together the Total June 4-6, 1991, Egham The Fifth International LIMS Conference, or- Biotechnological Processes’ Mr. A.M. Ben- Quality experiences of both manufacturers ganized in association with the Royal Society nett (CAMR, Porton Down); ‘Monitoring and service suppliers in the chemical indus- of Chemistry,will take place at the Anugraha Equipment Performance’ Dr. G. Leaver (War- try. The seminar will have a wide appeal, Conference Centre,Egham, Surrey.It follows ren Spring Laboratory); ‘Monitoring the covering the board level business aspects of the outstanding Success of the fourth con- Workplace Environment’ Dr. D. Griffiths TQM through to implementation experiences ference which was held in Pittsburgh in 1990. (Warren Spring Laboratory); ‘Role of FMEA and the European scene. It will be organized A number of speakers from the UK, con- in Biotechnology Risk Assessment’ Dr. D. by the Royal Society of Chemistry and the tinental Europe and the USA will address a Keir (AEA Technology); ‘GMOs - Risks As- Society of Chemical Industry and held in the variety of topics organized under the follow- sociated with Deliberate or Inadvertent Re- University of Salford. ing sessions:Information as part of the Corn- lease’ Professor J.Beringer (University of The following speakers have been invited: pany’s Business; Quality by Design or Bristol); ‘Effluent Treatment for GMO-Based J. Welch (ICI Paints), S.H. Coulson (Exxon Default?; The Need to Specify Objectives Processes’ Mr. F. Popp (Boehringer Mann- Chemical International Marketing), J. Gilbert and Strategies; LIMS as part of the Corporate heim); ‘Risk Assessment in the Water Treat- (Akzo Chemicals), J.M. Askey (Harcros Body; Information-Does it Clarify or Con- ment Industry’ Dr. J. Dennis (Thames Water); Chemicals), D. Rowling (Cerestar U.K.), J. fuse? ‘Risks in Biotechnology-Perceptions and Woods (Monsanto), D. Chadwick (D.T. An exhibition will run concurrently with Realities’ Dr. M. Turner (University College, Chadwick). C.McCraight (Glaxo Pharma- the conference. The following companies will London). ceuticals), R. Thomas (Hays Chemical Dis- be exhibiting: VG Laboratory Systems, For further information contact SCI Con- tribution), K.J. van Ham (European Beckman Instruments UK, Hewlett-Packard, ference Secretariat, 14/15 Belgrave Square, Foundation for Quality Management), J.I. ICI Chemicals & Polymers, Instron, Labsys, London swlx 8ps. Evans (Dow Coming), D.K. Taylor (Hoechst Perkin-Elmer and Instem. U.K.). An accompanying commercial exhibi- For further information please contact: tion of books is being organized. Sharp, Conference Secretary, 41 Exeter Road, Davyhulme, Manchester M3 1 1 RF. The Conference Registrar, Fifth International EPA 1990-Monitoring and For more information Contact M E Car01 LIMS Conference, P.O.Box 341, High Wy- Their Impact combe, Buckinghamshire, HP12 4AS. June 26-2 7, 1991, London The Environmental Protection Act 1990 pro- vides a framework which sets out the reauire- 4th COWV-SS of the soCi&k FmnPke de Good Clinical Practice in EuroDe once again be organizing its annual event this background legislation. However, the This Congress will be held in the Congrks examining Good Clinical Practice in Europe. main emphasis of the event will be on the Louis Pasteur. The colloquia will be: The This fifth meeting in the series is being or- practical methods, techniques and instrumen- Role of Metals in Biological Processes; Ap- ganized in association with AMIFE, the tation currently available for monitoring pur- plications in Catalysis and Chemical Thera- Spanish Pharmaceutical Industry Physicians’ poses, and also the examination of new peutics; Materials of the Future; Large Association.Topics that will be covered at developments. Speakers will discuss monitor- Inorganic Molecules: Structures and Func- this year’s event include: GCP-an update on ing programmes and their effectiveness, in- tions; The Chemistry of the Hetero-Elements; EC Directives; a comparison of GCP gui- cluding cost, and the response to unexpected Reactivity in Dispersal Media; Molecule &lines across the EC; supranational auditing releases. Delegates will also have the oppor- Recognition in Biological Chemistry; Appli- of GCP compliance; adapting standard opera- tunity of seeing demonstrations from exhibi- cations of Magnetic Spectroscopy Tech- ting procedures to EC guidelines; cross rec- tors, and participating in panel discussions.niques in the Life and Materials Sciences; ognition of ethics committees; GCP in the There will be speakers from the Centre for and Perspectives in Molecular Electronics. US; financial implications of GCP; economic Environmental Technology, UK Environmen- For further information contact G. Per- consequences of GCP in drug development. tal Law Association, NAMAS Executive, reau, 250 rue Saint-Jacques, 75005 Paris, lary Pendall, Bath House (3rd Floor), 56 Hol- City Council, National Physical Laboratory, born Viaduct, London EClA 2EX. British Nuclear Fuels plc, CODEL, Rechem International Ltd. and the Stockholm Envi- Handling of EnvkOnmental and Biolo@al ronment Institute at York.Samples in Chromatography For further information please contact Hi- HMIP, National Rivers Authority, Lancaster France. Risk Assessment in Biotechnology June 20, 1991, London This one day meeting, to be organized by and held in the Society of Chemical Industry, aims to bring together scientists concerned with assessing the risks associated with the biotechnology industry. The environmental September 4-5, I 991, salford impact of recombinant organisms, hazards of biotechnology processes, process equipment performance and effluent treatment will be discussed and linked to the developing regu- latory scene. The lectures will be: ‘Risk in Biotechnology-A Regulatory Viewpoint’ tal Analytical Chemistry.Former meeting places were Lausanne (1983), Freiburg For more information on amending Of ex- September 2627, 1991, Baden-Baden, Ger- hibiting at this event contact Katie Lye, IBC many Technical Services Ltd., Bath House, 56 Hol- born Viaduct, London EClA 2EX. The Symposium will be the 5th in a series of very successful meetings initiated by the late Professor Roland Frei under the auspicies of the International Association of Environmen- Total Quality in the Chemical Industry The Total Quality philosophy has become the <1985), Palma de Mallorca (1986) and Base1 cornerstone of a drive within the chemical (1988). Continuous flow techniques, column industry towards excellence. The role of switching methodologies and robotic technol- TQM as an integral part of the over-all busi- ogy to accomplish operations such as extrac- ness plan is becoming more prominent across tion, derivatisation, enrichment or dialysis for162 ANALYTICAL PROCEEDINGS, MAY 1991, VOL 28 the handling of complex samples are among the topics which will be treated and dis- cussed.Special emphasis will be placed on techniques with automation potential and procedures suitable for the automated hand- ling of larger numbers of samples. Most of these methodologies and concepts can be ap- plied to all kinds of analytes and matrices. It is, however, the intention to focus on samples of biological (blood, urine, tissue, plant ma- terial) and environmental (water, waste water, sediments, air) origin. Consequently, drugs (pharmaceuticals) and priority pollutants, re- spectively, are the analytes of major interest). Participants interested in the presentation of a poster are asked to submit an abstract of their work.Instructions for preparing the pos- ter will be sent to the authors following ac- ceptance by the Scientific Committee. The address before and after the Sympo- sium will be Workshop Ofice IAEAC, M. Frei-Hausler, Postfach 46, CH-4123 All- schwil2, Switzerland. Laboratory Conference November 5-7,1991, London Continuing their innovations in the Labora- tory Exhibition Series, the Organizers, in conjunction with MBC Conferences, an- nounce a major international scientific con- ference that will take place simultaneously with the Laboratory Exhbition at Earls Court 2 in London. The Laboratory Exhibition and Conference will offer UK and overseas scien- tists a programme of conferences, seminars, workshops and related events.These will provide indepth coverage of the latest labor- atory practice, research, technology, tech- niques and applications. In addition, delegates will also cover general issues rele- vant to the scientific community. Not only will delegates have access to the specialist knowledge provided by eminent speakers at this comprehensive event, but they will also have the opportunity to tour the exhibition. Current plans for the conference include sessions on Fourier Transform Techniques, Laboratory Safety (including COSHH), Data Quality and Environmental Analysis. These symposia will be convened by The Royal So- ciety of Chemistry. In order to gain maximum benefit from the Laboratory Exhibition and Conference the organizers have arranged special delegate accommodation packages for those people wishing to stay overnight in London.Full de- tails together with programme information are available from The Laboratory Con- ference Registrar on 081-699-9949. Finnish Chemical Congress November 12-14, 1991, Helsinki, Finland The Congress is the largest of its kind in the Nordic countries and the Exhibition the most comprehensive mounted in Finland in the field of chemistry and chemical industry. The Congress is arranged by the Association of Finnish Chemical Societies. In 1990 the Ex- hibition covered more than 4200 m2 and there were 150 exhibitors. Over 250 presenta- tions were included in the congress pro- gramme. The lecturers were recognized experts in their field, from both Finland and abroad, and three quarters of the oral presen- tations were in English.The congress programme includes: Sym- posium on Environmental Chemistry and En- gineering: Waste Management, Soil Decontamination; Symposium on Chemical Education; Symposium on Food Science and Technology; Symposium on Clinical Chem- istry; Symposium on Analytical Chemistry: Surface Analysis, Mass Spectrometry; Com- puter Assisted Chemistry: Molecular Modell- ing, Chemometrics, Chemical Structure Databases, Retrieval of Chemical Informa- tion. Further information on the Congress is available from: The Association of Finnish Chemical Societies, Ms Eeva Kota-aho. Tel. (+358 0)408 5 11. EPRI-EPA-DOA 1991 Sulphur Dioxide Control Symposium December 3-6, 1991, Washington, DC, USA The 1991 SO2 Control Symposium is the 13th in a series that provides for the ex- change of technical and regulatory infonna- tion of flue gas desulphurization (FGD) for utility and industrial boilers.Wet and dry scrubbers, new and emerging processes, and international developments in clean coaVacid rain technologies will be the major topics. Regulatory changes, operating experiments and process testing, pilot and demonstration results, and international developments will be emphasized. The audience will include utility and industry users of the technology, process suppliers, and representatives of the government, academia, and R&D firms. The Symposium will be held in the Grand Hyatt Washington.For information contact Electric Power Research Institute, Attn: Pam Turner, Sympo- sium Coordinator, PO. Box 10412, Palo Alto, CA 94303-9743, USA. 1992 Winter Conference on Plasma Spec- trochemistry January 6-11, 1992, San Diego, CA, USA The 1992 Winter Conference on Plasma Spectrochemistry, seventh in a series of bien- nial meetings sponsored by the ICP Informa- tion Newsletter, features developments in plasma spectrochemical analysis by induc- tively coupled plasma (ICP), d.c. plasma (DCP), microwave plasma (MIP) and glow and hollow cathode discharge (GDL, HCL) sources. The meeting will convene at the San Diego Princess Convention Center in San Diego, California. Continuing education short courses at introductory and advanced levels will be presented on January 3-5.A three-day exhibition of spectroscopic in- strumentation and accessories also will be The rapid growth in popularity of plasma sources for atomization and excitation in atomic spectroscopy and ionization in mass spectrometry and the need to discuss recent developments of these discharges in spectro- chemical analysis stimulated the organization of this meeting. The Conference will bring together international scientists experienced in applications, instrumentation and theory in an informal setting to examine recent pro- gress in the field. Approximately 500 partici- pants from 25 countries are expected to attend. Approximately 200 papers describing applications, fundamentals, and instrumental developments with plasma sources will be presented in lecture and poster sessions by about 150 authors.A threeday exhibition of spectroscopic instrumentation and chemicals, electronics, glassware, publications, and soft- ware supporting plasma spectroscopy will complement the scheduled sessions with ap- proximately 30 firms participating. Introductory and advanced, four-hour short courses will also be presented. For further information contact ICP Infor- mation Newsletter, Department of Chemistry, 102 GRC Towers, University of Massachu- setts, Amherst, MA01003-0035, USA. Prep '92: Mh International Symposium on Preparative and Industrial Chromato- April 6-8, 199 I , Nancy, France The Symposium, which will be held in the Palais de Congrks, aims to reflect the most recent progress in preparative and industrial chromatography with gas, liquid or super- critical eluents.This will include: Modelling and Simulation-physicochemical interac- tions between stationary and mobile phases; Equipment Design and Optimization-new stationary phase developments; Process Inte- gration and Cost Estimation; Application and Development Case Histories. Information is available from the Prep 92 Secretary, ENSIC-LPCI, 1, rue Grandville- B.P. 451, F-54001 Nancy Cedex. graphy Cadmium 92: Seventh International Cad- mium Conference April 6-8, 1992, New Orleans, LA, USA The Seventh International Cadmium Con- ference will be held at the Intercontinental Hotel, New Orleans. Jointly organized by the Cadmium Association, London, and by the Cadmium Council, USA, and the Interna- tional Lead Zinc Research Organisation (ILZRO), the conference will be wide rang- ing and cover economic, technical, environ- mental and health aspects of cadmium. The conference will review the world scene for cadmium, with contributions covering pat- terns of production and consumption along with trends in the use of cadmium in various sectors.Preliminary information is available from Cadmium Association 42, Weymouth Street, presented. London WIN 3LQ.ANALYTICAL PROCEEDINGS, MAY 1991, VOL 28 163 NIR-92: The Fifth International Con- ference on Near Infrared Spectroscopy June 16-20, 1992, Haugesund, Noway The theme of this conference will be bridging the gap between data analysis and NIR appli- cations. The following will be considered: calibration methods; spectral interpretations; instrumentation and software; and applica- tions.The conference programme will in- clude plenary lectures, oral and poster presentations and discussions, workshops and exhibitions of NIR instrumentation and soft- ware. It will be flexible in order to include coming innovations within research and ap- plications. The invited speakers will include: D. Bertrand, France; P.R. Griffiths, USA; D. Honigs, USA; T. Isaksson, Norway; H. Mar- tens, Norway; T. Naes, Norway; E. Stark, USA; and P.C. Williams, Canada. A course in basic NIR methodology and/or advanced NIR calibration will be of- fered on June 15-1 6. For further information contact NIR-92, MATFORSK-Norwegian Food Research Institute, Oslovn. 1, N-1430 As, Norway. Chemistry, Agriculture and the Environ- ment July 16-18, 1992, Guildford The Robens Institute of Industrial and Envi- ronmental Health and Safety is organizing an international conference on Chemistry, Agri- culture and the Environment.The conference will take place at the University of Surrey, Guildford. The conference will run immedi- ately after the International Conference on ‘Risk Management of Chemicals-Can Perspective’; T.R. Roberts, Hazleton, ‘Pes- Chemicals be Used Safely?’ also run by the ticides in Water-Human Health, Agriculture Royal Society of Chemistry. and Environmental Aspects’; D. Stevenson, Management of the use of agrochemicals Guildford, ‘Operator Exposure to Agro- and related materials is a matter of growing chemical Sprays’. There will be a poster ses- concern throughout the world. The aim of sion at the meeting.These will be allowed this conference is to provide a forum for ag- prime time during the programme. Further rochemical producers, risk assessment pro- details will be available on request. fessionals, regulators, and academics, and all The conference is being organized by the those with an interest in the safe and effective Robens Institute of Industrial and Environ- use of agrochemicals. The conference will mental Health and Safety under the auspices bring together a number of authorities in vari- of The Royal Society of Chemistry and the ous aspects of chemistry, agriculture and the Federation of European Chemical Societies. environment. The invited speakers will in- The organizing committee is cochaired by Dr. clude: G.W. Aherne, Guildford, ‘Immuno- D.Stevenson (Robens Institute) and Mr. M. logical Analysis of Pesticides’; 0. Bockman Richardson (BASIC). and T. Granli, Porsgrunn, Norway, ‘Human For further information contact Dr. D. Health Aspects of Nitrate Intake from Water’; Stevenson, The Robens Institute of Health D.M. Foulkes, Schering, ‘Agrochemicals in and Safety, The University of Surrey, Guild- the Environment’; M.K. Garrett, Belfast, ‘Ni- ford, Surrey, GU2 5XH. trogen Losses from Grassland Systems Under Temperate Climatic Conditions’; Z. Gerstl, Bet-Dagen, Israel, ‘Behaviour of Organic Ag- rochemicals in Irrigated Soils’; E. Heinish, 12th Australian Symposium On Berlin, Germany, ‘Contamination of Surface Chemistry incorporating 3rd Environmen- Waters, Sediment and Biota by a Pesticide tal Maidenhead, ‘Methane and Ammonia in the tralia Atmosphere: Impact of Grassland Agricul- ture’; H. Kidd, Cambridge, ‘Information sources for Chemistry, Agriculture and the This biennial Symposium will cover most as- s. Li, Beijing, China, 6Pes- pects of Analytical and Environmental Chem- ticides, Environmental Pollution and Human istry. Eminent International and Australian Health’ ; Manno, Italy, toxicology chemists will participate. The theme will be and Risk Assessment of Pesticides’ ; J. Orme, ‘Tomorrow’s Washington, USA, ‘Health Advisories for For further information contact 12AC, Pesticides’; D. Osborne, Huntingdon, ‘Im- The Conference Office, The University of pacts of Agriculture on Wild Plants and Ani- Western Australia, Nedlands, WA, Australia mals in the UK: Chemical Impacts in 6009. Conference Producing in s.c* Jarvis, September 2&October 1, 1993, Perth, AM- Today’.

ISSN:0144-557X    

DOI:10.1039/AP9912800161

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

ANALYTICAL PROCEEDINGS, MAY 1991, VOL 28 163 Courses Update on the Disposal of Hazardous Waste July 3, 1991, Shefield The CCVE at Sheffield University are hold- ing a one day course on the Disposal of Haz- ardous Waste. This will review the legal, technical and economic aspects of hazardous waste deficiences of the system in practice. The course is intended for staff of waste pro- ducing companies who wish to update on the current scene, but in view of its general na- ture, would benefit legal, environmental, technological and managerial staff. Another course at the CCVE will be held on the safe storage of Hazardous Chemicals on October 29, 1991. This will discuss the range of hazards that can be present in chemi- cal storage and outline the legal requirements to minimise risks.The course is set at a basic level and is intended to managers, safety rep- resentatives and technical staff concerned with industrial planning who are involved with advising on the safe storage of chemi- steal a march, because of sloppy or inadequ- cals. ate protection. The framework of UK Intel- Further information can be obtained from lectual Property Law is designed to assist Mrs. K. Wainwright or Mrs. L. Rodgers at 65 businesses to protect their products, so an un- Wilkinson Street, Sheffield S20 2GJ. derstanding of intellectual property law is vital for those involved in research and devel- opment, finance and marketing. Money can A Business and Technical Manager’s also be made by exploiting IP rights by using Guide to Intekctual ~ P e r t Y - ~ ~ ~ % techniques such as technology transfers and and Exploiting your Assets technology licensing.All of these points will be covered at a seminar. to be held at the Cafe July 5, 1991, London Royal. As well as the formal talks by expert The language of the law to the non-lawyer speakers, there will be time for questions and can be confusing, especially the terminology discussion. Each delegates will receive a of intellectual property law: although the bound copy of the conference papers plus a terms are often used by non-lawyers, are they copy of the McGraw Hill publication ‘Intel- fully understood? Words such as ‘patent’, lectual Propery-A Manager’s Guide’, an in- ‘copyright’ and ‘trademark’ are frequently valuable and detailed text which will provide used, yet to a non-lawyer for a definition of a continuing source of reference. All this, any of those terms would probably yield plus refreshments and luncheon is included in widely differing interpretations. Protecting the conference fee of &325. the product legally is vital; much time and For further information contact Linda money is spent on research, development and McKay on 07 1-236-4080; Fax 07 1-489- marketing, yet it is wasted if competitors can 0849.

ISSN:0144-557X    

DOI:10.1039/AP9912800163

出版商:RSC    

年代:1991

数据来源: RSC