Chromatography in Biotechnology 9780841226692, 9780841213852, 0-8412-2669-5

Citation preview

Downloaded by 89.163.34.136 on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.fw001

Chromatography in Biotechnology

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Downloaded by 89.163.34.136 on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.fw001

ACS

S Y M P O S I U M SERIES

529

Chromatography in Biotechnology

Downloaded by 89.163.34.136 on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.fw001

Csaba Horváth, EDITOR Yale University

Leslie S. Ettre, EDITOR Yale University

Developed from two symposia sponsored by the Division of Analytical Chemistry of the American Chemical Society at the Fourth Chemical Congress of North America (202nd National Meeting of the American Chemical Society), New York, New York, August 25-30, 1991

American Chemical Society, Washington, DC 1993

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Library of Congress Cataloging-in-Publication Data Chromatography in biotechnology / Csaba Horváth, editor; Leslie S. Ettre, editor

Downloaded by 89.163.34.136 on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.fw001

p.

cm.—(ACS symposium series, ISSN 0097-6156; 529.)

"Developed from a symposium sponsored by the Division of Analytical Chemistry of the American Chemical Society at the Fourth Chemical Congress of North America (202nd National Meeting of the American Chemical Society) New York, NY, August 25-30, 1991." Includes bibliographical references and index. ISBN 0-8412-2669-5 1. Chromatographic Congresses.

analysis—Congresses.

I. Horváth, Csaba, 1930TP248.25.S47C47 660'.6'028—dc20

2. Biotechnology-

. II. Ettre, Leslie S. III. Series.

1993 93-16982 CIP

The paper used in this publication meets the minimum requirements of American National Standard for Information Sciences—Permanence of Paper for Printed Library Materials, ANSI Z39.48-1984. Copyright © 1993 American Chemical Society All Rights Reserved. The appearance of the code at the bottom of the first page of each chapter in this volume indicates the copyright owner's consent that reprographic copies of the chapter may be made for personal or internal use or for the personal or internal use of specific clients. This consent is given on the condition, however, that the copier pay the stated per-copy fee through the Copyright Clearance Center, Inc., 27 Congress Street, Salem, MA 01970, for copying beyond that permitted by Sections 107 or 108 of the U.S. Copyright Law. This consent does not extend to copying or transmission by any means—graphic or electronic—for any other purpose, such as for general distribution, for advertising or promotional purposes, for creating a new collective work, for resale, or for information storage and retrieval systems. The copying fee for each chapter is indicated in the code at the bottom of the first page of the chapter. The citation of trade names and/or names of manufacturers in this publication is not to be construed as an endorsement or as approval by ACS of the commercial products or services referenced herein; nor should the mere reference herein to any drawing, specification, chemical process, or other data be regarded as a license or as a conveyance of any right or permission to the holder, reader, or any other person or corporation, to manufacture, reproduce, use, or sell any patented invention or copyrighted work that may in any way be related thereto. Registered names, trademarks, etc., used in this publication, even without specific indication thereof, are not to be considered unprotected by law.

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

1993 Advisory Board A C S Symposium Series M . Joan Comstock, Series Editor V . Dean Adams University o f Nevada— Reno

Downloaded by 89.163.34.136 on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.fw001

Robert J. Alaimo Procter & Gamble Pharmaceuticals, Inc. Mark Arnold University o f Iowa David Baker University o f Tennessee Arindam Bose Pfizer Central Research Robert F. Brady, Jr. Naval Research Laboratory

Bonnie Lawlor Institute for Scientific Information Douglas R. Lloyd The University o f Texas at Austin Robert McGorrin Kraft General Foods Julius J. Menn Plant Sciences Institute, U.S. Department o f Agriculture Vincent Pecoraro University o f Michigan Marshall Phillips Delmont Laboratories George W. Roberts

Margaret A . Cavanaugh

North Carolina State University

National Science Foundation Dennis W. Hess Lehigh University

A . Truman Schwartz Macalaster College

Hiroshi Ito IBM Almaden Research Center

John R. Shapley University o f Illinois at Urbana—Champaign

Madeleine M . Joullie University o f Pennsylvania

L. Somasundaram Ε. I. du Pont de Nemours and Company

Gretchen S. Kohl

Peter Willett

Dow-Corning Corporation

University o f Sheffield (England)

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Foreword

Downloaded by 89.163.34.136 on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.fw001

THE A C S SYMPOSIUM SERIES was

first published in 1974 to provide a mechanism for publishing symposia quickly in book form. The purpose of this series is to publish comprehensive books developed from symposia, which are usually "snapshots in time" of the current research being done on a topic, plus some review material on the topic. For this reason, it is necessary that the papers be published as quickly as possible. Before a symposium-based book is put under contract, the proposed table of contents is reviewed for appropriateness to the topic and for comprehensiveness of the collection. Some papers are excluded at this point, and others are added to round out the scope of the volume. In addition, a draft of each paper is peer-reviewed prior to final acceptance or rejection. This anonymous review process is supervised by the organizers) of the symposium, who become the editor(s) of the book. The authors then revise their papers according to the recommendations of both the reviewers and the editors, prepare camera-ready copy, and submit the final papers to the editors, who check that all necessary revisions have been made. As a rule, only original research papers and original review papers are included in the volumes. Verbatim reproductions of previously published papers are not accepted.

M. Joan Comstock

Series Editor

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Downloaded by 89.163.34.136 on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.pr001

Preface R E C E N T A D V A N C E S IN C H R O M A T O G R A P H Y are closely associated with biotechnology's challenge to provide efficient means for the isolation, purification, and analysis of bioproducts. Because of its versatility, chromatography has met this challenge, and, as a result, chromatographic techniques and processes play a pervasive role in all levels of biotechnology. A comprehensive treatment of chromatography in biotechnology would require a multivolume treatise. This book can cover only a few segments of this fast-growing field. Most of this book is devoted to preparative and process chromatography of proteins and peptides. Industrial-scale purification processes require a departure from the linear elution mode of chromatography and the introduction of other separation schemes. Although many of the processes themselves are not new, their applications in biotechnology are. Often, progress in chromatography has been driven by the introduction of column materials that exhibit thermodynamically and kinetically superior properties. It is therefore appropriate that some of the chapters in this book deal with recent advances along these lines. The use of high-performance liquid chromatography as a multifarious instrumental technique is increasing in analytical biotechnology despite the recent emergence of capillary electrophoresis. Because many new therapeutic protein products are glycosylated, efficient methods for the analysis of complex carbohydrates are increasingly in demand. The chapters on the chromatography of glycoconjugates reflect the importance of this area of research and account for some of the new developments. This is a fertile field, and we can expect further significant advances in the chromatography of complex carbohydrates.

CSABA HORVÁTH LESLIE S. ETTRE Department of Chemical Engineering Yale University New Haven, CT 06520

December 16, 1992

ix

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Chapter 1

Chromatographic Separations in Biotechnology John Frenz

Downloaded by 89.163.34.136 on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch001

Genentech, Inc., 460 Point San Bruno Boulevard, South San Francisco, CA 94080

This review describes the role of chromatography in the development of recombinant protein pharmaceuticals. The versatility and efficacy of chromatographic techniques have made them the sine qua non in both large scale and analytical separations in biotechnology. The principal qualities of chromatographic approaches that underlie this prominent role and the main techniques employed at both production and analytical scales are discussed. A new era has dawned in technology and commerce with the advent of the new biology by which complex protein molecules can be expressed at large scale i n living cells for production of human therapeutic pharmaceuticals. The developments in molecular biology and biochemistry that underlie these production processes have been revolutionary in facilitating the production of large quantities of compounds that once were scarce in pure form. The sources of these compounds, relatively fast-growing cells altered to express the target protein, may differ markedly from their natural source. Hence, in two important ways—scale and origin~the recombinant DNA-based production methods are fundamentally distinct from the traditional approaches to the preparation of biological extracts for therapeutic use. The new manufacturing techniques thus pose challenges both to the protein pharmaceutical industry and the regulatory bodies established to police the industry (7). These challenges for the pharmaceutical industry stem from the relative newness of the technology that limits the experience available for the design and optimization of manufacturing processes. The technology for achieving high purity separations at large scale and for analytically characterizing the resulting product have been developed to meet the unique demands o f the industry. The challenges to regulatory bodies also stem from the novelty o f this class of therapeutics that differ both from small molecule drugs that are the products of organic chemical syntheses, and from blood and tissue extracts that are the traditional sources of therapeutic proteins. The purification of recombinant protein pharmaceuticals requires the separation of the product from host cell proteins and D N A and from components of the culture medium. Along with these contaminants are the stringent requirements for removal of endotoxins from products expressed in E . coli and viral particles from mammalian cell culture produced proteins. In rising to these challenges the biotechnology industry has broken ground in demonstrating that proteins can be produced and recovered at scales large enough to treat common diseases, and at a purity level unprecedented for biological products (2).

0097-6156/93/0529-0001$06.00/0 © 1993 American Chemical Society

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

CHROMATOGRAPHY IN BIOTECHNOLOGY

Downloaded by 89.163.34.136 on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch001

2

The advances i n recombinant D N A technology that gave rise to what is recognized now as the biotechnology industry began with the discovery of techniques for isolating the genes for individual proteins, expressing these proteins i n microbial or mammalian cell culture and recovering the heterologous protein in active form. The allure of these techniques is that the cells are fast growing relative to whole animals and can be engineered to produce high levels of the desired protein. Thus recombinant D N A technology provides a more practical means of production o f many human proteins for the treatment of disease than do conventional methods for extraction from blood and body parts. The commercial success of recombinant protein pharmaceuticals is suggested by the sales figures shown i n Table I. The implementation o f these techniques for the production of therapeutic proteins for T A B L E I. B I O T E C H N O L O G Y S U C C E S S E S human use, however, has World-wide sales figures adapted from reference 54. required the development o f a myriad o f new technologies and Annual Sales Product new approaches to (Billions) pharmaceutical manufacturing. $1.0 Insulin Recombinant proteins are most 0.9 Growth hormone conveniently expressed in 0.5 Interferons bacteria, a very fast growing 0.4 Erythropoietin microbe that is relatively simple Tissue plasminogen to stably transfect with the 0.2 target gene. A large number of activator proteins, however, can only be correctly expressed i n mammalian cells, either because they are incorrectly folded i n the reducing environment inside bacteria (3) or require post-translational modification such as glycosylation to maintain activity in clinical use (4). Thus methods for stable transfection o f continuous mammalian cell lines and their growth i n large scale cultures have been developed (5). A number of methods, depending on the compartment i n which the host cell deposits the protein, are e m p l o y a i for initial recovery of the protein from the cell broth (6). For example, many proteins expressed at high levels i n bacteria form inclusion bodies that can be recovered from the lysed cell by centrifugation. Other proteins in bacteria may be secreted, as are products produced in mammalian cells, so either centrifugation or filtration may be used to separate cell mass from the product The crude material recovered by these methods is then resolubilized, i n the case of an inclusion body-bound protein (7), prior to entering the process purification stream. Final purification is generally a multi-step procedure involving complementary unit operations-often predominantly chromatographic-that give a final product in an acceptable yield, cost and purity level. Since each protein has its own distinct chemical, physical and biochemical properties, the selection and ordering of unit operations to solve the final purification problem for each product is unique. The individuality of each protein also means that the assays for purity and characterization of each product are tailored to that molecule. Analytical techniques are required to guide the selection o f manufacturing conditions, from the fermentor through to the development o f the final purification train. T o supply these needs an impressive array of high performance tools-again, predominantly chrornatographic--has been developed for the analytical protein chemist (8). Within their individual disciplines, process development and analytical chemists have their own specific constraints in the selection of their tools, and in particular in their approach to implementation of chromatographic separations. Large scale chromatographic operations A m o n g the essential considerations involved in the design of processes for the large scale purification of pharmaceuticals is the adherence of the process to regulatory guidelines. The regulation of the use of protein therapeutics derived from natural

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Downloaded by 89.163.34.136 on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch001

1.

FRENZ

Chromatographic Separations in Biotechnology

3

sources has been oriented toward the assurance of a consistent manufacturing process (9). This viewpoint was adopted to recognize the heterogeneous and relatively ill-defined composition of products extracted from biological sources. A drug deemed safe and effective i n controlled clinical trials was assumed to remain safe and effective so long as the manufacturing route taken to produce the drug was invariant. Hence the manufacturing process, rather than detailed chemical characterization, defined the product. Regulatory bodies maintain this perspective for biologicals produced by recombinant D N A technology, so that manufacture o f genetically engineered proteins is also tightly controlled, and changes in the production process require clinical testing and prior approval before the drug made by the altered process can be sold. This philosophy informs many of the decisions taken in the development of the production process for a protein pharmaceutical. Since the manufacturing process must be fixed well before marketing is approved, and often relatively early i n clinical testing, important technical decisions must be made while the need for the product is still relatively low. The consequences of these decisions, however, may persist into the end stages of clinical testing or beyond, after the drug is approved and full market scale achieved. Thus, i n many cases there may be a premium placed on rapidly developing the process to provide material for initial clinical trials, while ensuring the resulting process is suitable to meet future needs should the drug continue to show promise. Chromatographic separations are of singular importance to the biotechnology industry because they deliver high purity, are relatively easy to develop at the laboratory scale and can be scaled linearly to the desired production level in most cases. Hence one reason for the ubiquity of chromatographic steps in preparative protein purification steps is that they provide a relatively efficient means to meet the major manufacturing goals of the pharmaceutical biotechnology industry. The importance of this unit operation therefore accounts for the attention focused on greater understanding of the behavior o f proteins i n chromatographic separations, through a combination of empirical study, modeling and laboratory investigations of the adsorption of proteins on surfaces and of alternative modes o f operating chromatographic columns. One of the principal features distinguishing preparative from analytical scale chromatography is the composition and morphology of the packing material in the column (10). In interactive modes of chromatography, such as ion exchange, reversed phase or hydrophobic interaction, the packing material anchors the stationary phase (77). In modes of chromatography such as size exclusion, the packing material is inert to interaction with the feed components, and serves instead to sieve the mixture. Hence the chemical characteristics o f the support material are important to its efficacy. The packing must either possess the appropriate chemical groups for its intended use or be readily modified to attach the desired chemical moieties at the surface. The packing must also be stable to buffers and cleaning reagents employed i n the purification process (10). The physical strength of the support also figures importantly in its suitability to chromatographic applications since the packed bed must be able to withstand the pressure drop associated with liquid flow without deleterious deformation, collapse or fracturing of the resin. A variety of materials have been introduced to meet these needs, including synthetic polymer (72-74), polysaccharide (75) and silica-based (77) packings. The polymeric resins have the important advantages of being stable to washing with alkaline cleaning solutions that are widely used for aseptic storage of equipment, while silica is mechanically stronger and cheaper to manufacture. The morphology of the resin is controlled by the manufacturing procedure, and is an essential characteristic determining its utility for protein separations (76). The size range of the packing particles and of the internal porosity of the resin contribute to both the efficiency and capacity of the chromatographic column. Smaller diameter resins yield increased efficiency at the expense of higher pressure drop across the column, but in any size range a narrow range of particle diameters maximizes the column efficiency. The pore size and size distribution control the extent to which the resin is accessible to components of the feed mixture, and so controls the selectivity afforded by size

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

CHROMATOGRAPHY IN BIOTECHNOLOGY

exclusion chromatography and the capacity of the column in interactive modes of chromatography. In an extreme case, the pores can be wide enough to allow convective flow through the particle that can improve the resolution afforded by large scale H P L C purification steps (77). The microstructure of the support determines the efficiency and capacity o f the column, but in interactive modes of chromatography the surface characteristics of the resin control the selectivity of the separation. One of the attractions of chromatographic approaches to protein purification is the variety of resins available that offer a broad array o f complementary selectivities (18). Table Π illustrates the use o f different combinations of steps to purify enzymes from microbial sources. The principal classes of columns include hydrophobic interaction, reversed phase, cation exchange and anion exchange media. Hydrophobic interaction chromatography (HIC) exploits the interaction between a protein and relatively hydrophobic stationary phase that is promoted by increasing the salt content o f the mobile phase (79). Reversed phase chromatography also separates according to the hydrophobicity o f the protein, but typically employs harsher conditions, including more hydrophobic columns and the use of organic solvents to elute the feed components, that can alter the conformation of the protein (20). Ion exchange chromatography resolves proteins according to the degree of interaction with surface-bound negatively charged (cation exchange) or positively charged (anion exchange) moieties (27). The strength of the electrostatic interaction between the protein and the resin surface is a function o f eluent p H and ionic strength, so these are the principal variables manipulated to effect binding to and elution from the column. A protein can present both anionic and cationic faces to the column surface under a given set of conditions, and so the binding to a column may be less influenced by the net charge of the whole protein than by the charge at particular sites that are favored for interaction with a given column. Hence, a protein may bind to both anion and cation exchangers under a given set o f conditions (22). M i x e d mode resins, such as

2 ε ft* c Enzyme

Purification step

α Aspartase from E. coli Thiolase from C. acetobutylicwn

Phosphotrans-butylase from C . acetobutylicwn

Ammonium sulfate D E - 5 2 column A X 1 0 0 0 column Crude extract Ammonium sulfate DEAE-Sephacel column Blue Sepharose column A X 3 0 0 column Crude extract Ammonium sulfate Phenyl Sepharose column DEAE-Sephacel column A X 1 0 0 column

27 2.4 0.4 1090 530 32 7 3 4235 840 102 22 6.2

ver;

tivif CO 0£

ε its/

Ο «Β

:ific

/ml

T A B L E Π. P U R I F I C A T I O N O F E N Z Y M E S F R O M M I C R O B I A L S O U R C E S Adapted from reference 55

tein

Downloaded by 89.163.34.136 on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch001

4

ο

06

w

a

0.4 4.1 58 3.1 5.8 57 111 220 7.2 28.3 182 694 1460

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

100 120 84 100 89 54 23 20 100 78 61 50 30

Downloaded by 89.163.34.136 on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch001

1.

FRENZ

Chromatographic Separations in Biotechnology

5

hydroxyapatite columns (23), that combine both anionic and cationic exchange functionalities also can provide altered useful selectivities compared to monophyletic exchangers. In addition, interactions of varying degrees of specificity have been exploited in affinity chromatographic separations (24). Affinity separations can be carried out with a wide variety of immobilized ligands, from metals to enzyme substrate analogs to monoclonal antibodies, that are chosen to specifically interact with the desired protein. In many cases this approach simplifies the purification process. W i t h all types of columns, the selectivity can vary significantly according to the precise method o f manufacturing a particular column. The manufacturing method can have a qualitative influence on the capacity or selectivity of the resin, so optimization o f a separation scheme necessarily includes screening a variety of packings for enhanced performance. The column defines the selectivity available for a given separation, but the capacity and throughput of the process is also determined by the mode of operation of the column. Among the choices made in defining the process are the steps by which the feed components are bound and eluted. The choice of operating mode i n preparative purification operations has an effect on the capacity of the chromatographic step. Since the cost of a column step increases with the size of the column, operating modes have been investigated that maximize the capacity of the column while maintaining the resolution required for the desired purity level. Figure 1 shows schematically the different operating steps employed in chromatography. The simplest operating mode is isocratic elution. In this mode aie column is equilibrated and eluted with a buffer adjusted so that the feed components transit the column at different velocities, separating as they move down the column. The operational simplicity of this approach is that no changes i n buffer composition are made at the column inlet In size exclusion chromatography this is the most common operating approach. However in many types of interactive chromatography of proteins it can be difficult to reproducibly achieve the precise buffer composition that yields the appropriate degree of interaction to allow isocratic elution or it may be impossible to find buffer conditions i n which all components of the mixture are eluted from the column. Hence, in most separations, as i n analytical chromatography of proteins, a gradient i n eluent strength is employed to elute the column (25). The change in eluent strength may be continuous or stepwise, Figure 1. Schematic chromatograms for various according to the degree of operating modes in liquid chromatography, including resolution required. Continuous (A) isocratic elution, (B) gradient elution, (C) step gradients, including linear elution, and (D) displacement. The dashed lines gradients, i n eluent strength represent ( A - C ) modifier and (D) displacer require relatively sophisticated concentrations.

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

CHROMATOGRAPHY IN BIOTECHNOLOGY

Downloaded by 89.163.34.136 on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch001

6

pumping, mixing and control equipment to deliver reproducible eluent composition profiles to the column. This approach is more practical, however, than isocratic elution for biopolymers, whose binding to stationary phases can be reversed with a small change in eluent strength (26). In order to simplify the operational mechanics o f gradient separations, stepwise elution protocols are commonly employed. In these methods, the feed mixture is bound to the resin under conditions in which, preferably, some impurities flow through the column. The buffer composition is then changed i n a stepwise fashion to elute the desired product while retaining on the column certain other impurities. Generally, stepwise elution is simpler to implement at large scale provided binding and elution conditions can be found that provide the necessary resolution. In certain cases, it may be more efficient to implement a related operating mode, frontal chromatography, in which the binding conditions are selected so that the desired product is unretained and the column binds impurities. The advantage of this approach is that none of the binding capacity of the column is taken up by the desired product, instead the column is sized for impurity removal. In the later stages of a process, where the feed stream is relatively pure, frontal chromatography thus offers a much more efficient use of the resin. Another operationally simple method is displacement chromatography (28-30), in which the feed mixture is bound to the column, and rather than be eluted, the product is displaced from the resin by a stepwise input of a solution of a compound that has a high affinity for the surface. Physically, the difference between displacement and stepwise elution mechanisms is that the displacer is more strongly retained than the feed mixture, while an eluent is less strongly retained. In practice, in displacement chromatography the displacer zone remains behind the product bands while the eluent in elution chromatography is mixed in with the product bands. Displacement separations yield purified product bands that attain a thermodynamically fixed concentration, with sharp boundaries between adjacent bands in the effluent concentration profile. This is i n contrast to stepwise elution techniques, that result i n relatively concentrated, but tailing, peaks in the chromatogram. The displacement mode can have advantages in capacity, resolution and processing simplicity compared to elution techniques, provided that the relatively narrow constraints on displacer properties can be satisfied. Protein analysis by H P L C In part, the approach to regulating the manufacture of pharmaceuticals derived from biological sources described above stemmed from the relative scarcity of purified blood- and tissue-derived products that effectively precluded in-depth chemical analysis. The low abundance of therapeutically active proteins in the complex matrices that are the classical sources of these proteins impeded large scale manufacture to extremely high purity and the degree of analytical characterization that was routine for synthetic small molecule drugs. The products of the biotechnology revolution are, however, available in relatively large quantities so not only is the production of high purity protein pharmaceuticals possible, but also the detailed characterization of the resulting drug. One aspect o f confirming the consistency of the manufacturing process has thus become the confirmation of the consistency of the resulting product by analytical protein chemistry techniques (7). The revolution in biotechnology has therefore been accompanied by a revolution i n analytical techniques, some of which are shown in Figure 2, for the detailed description of protein structures (31). The degree of complexity of the interplay of biological and biochemical processes involved in the fermentation and purification of a protein pharmaceutical is such that analytical characterization provides important evidence supporting the consistency of the process or can provide the basis for troubleshooting and improving a process. Analytical scale chromatographic techniques-especially H P L C - a r e essential tools for providing the data to meet these objectives. Hence the last decade has seen an explosion in the breadth o f applications of chromatography to analytical biotechnology and i n the availability o f columns and chromatographic systems tailored to these

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

1.

FRENZ

Chromatographic Separations in Biotechnology Property 0 U> 3 +•» Ο

s u

Analytical technique

Downloaded by 89.163.34.136 on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch001

.ο S Spectroscopic Optical MS NMR X-ray crystallography Separations

e

8

es

ε u c © U

§

s

S

CO

ε» ε

s i

eu ε

Electrophoresis HPLC Chemical Amino acid composition N-terminal sequencing C H O composition Enzymatic C-terminal sequencing Peptide mapping C H O structure Figure 2. Analytical tools employed for different aspects of the structural characterization of proteins. Adapted from reference 31.

applications. One area i n which chromatography has provided insights that are not available by any other technique is in the characterization of carbohydrate structures on glycoproteins (52). These structures can be essential to proper folding and activity o f the protein, and can strongly influence the pharmacokinetics, bioavailability and efficacy of a protein pharmaceutical, yet are usually sufficiendy heterogeneous to present a daunting analytical challenge (33). M o d e m analytical approaches nevertheless have been developed that separate and quantify the constituent monosaccharides or the individual glycoforms arrayed on the protein. This level of description o f a glycoprotein is a testament to the advances recently made in the field of analytical biotechnology that confers a new level o f meaning to the widely-coined phrase "high performance". The performance evinced by these applications clearly represents a new height. The achievement that gave birth to H P L C and underlies its continuing importance as the preeminent method in analytical protein chemistry was the adoption o f micronsized particles as supports for column packing materials (34). The use of these materials entailed the development o f costly high pressure columns and equipment i n order to effect separations, but dramatically improved the speed, resolution and sensitivity of separations compared to those available by conventional low pressure L C approaches. Continuing refinements and growth i n the technology o f H P L C has produced the current broad spectrum of columns and equipment to meet the needs of the equally diverse assortment of H P L C users (35). The variety of analytical columns and the high

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Downloaded by 89.163.34.136 on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch001

8

CHROMATOGRAPHY IN BIOTECHNOLOGY

efficiency o f analytical H P L C columns together contribute to the approach taken by analytical chemists to separations problems that is markedly distinct from that taken by designers of large scale protein purification operations. A s described above, large scale separations typically involve several columns with different selectivities that together yield high purity. Analytical separations, however, typically employ only a single column so much more reliance is placed on high efficiency of the column to provide the resolution required to ascertain the purity and composition of the sample. This distinction accounts for the widespread adoption of H P L C for analytical separations and the drive over the last two decades to increase the efficiencies of columns and instrumentation. This drive has expanded to account in large part for the allure of capillary electrophoresis that has demonstrated efficiencies in certain applications that far outpace even H P L C . Despite the improvements in instrument and column technologies that have steadily improved the efficiencies of H P L C separations, in many instances the physicochemical behavior and structural heterogeneity of proteins preclude high efficiency separations by techniques that perform well for small molecule separations. Hence, renewed attention has focused, as it has i n large scale separations, on exploiting the differences that are manifest among the rich variety o f H P L C column packing materials that are commercially available. Traditionally, the reversed phase mode has dominatedand almost become synonymous w i t h - H P L C . The denaturing conditions associated with the reversed phase mode, however, tend to mitigate the differences among proteins in a manner not unlike the denaturing conditions of S D S - P A G E (36). This mitigation has the effect, on the one hand, of narrowing the peaks obtained i n protein separations, but, on the other hand denaturation can obliterate the structural differences among proteins that provides the most useful basis for separation (37). Thus i n many cases analytical separations, like preparative chromatography, must be carried out in aqueous buffers in the size exclusion, hydrophobic interaction, ion exchange or various affinity modes of operation. Each of these modes of chromatography separates according to a different property of the analyte and so provides a different basis for selectivity of a given analytical separation, i n addition, many chemically different columns suited to each o f these modes are available from various commercial sources. The differences among columns manufactured by different routes can be sufficient to significantly alter the selectivity obtained in a separation, particularly of proteins, that interact in a myriad of subtle ways with biological and synthetic surfaces (38). Therefore, an important element of the full power of H P L C for protein separations derives from the plethora of columns available that differ in large and small ways from one another. Another aspect of H P L C that confers much of its power as an analytical tool is the variety of methods of detection that can enhance the sensitivity o f the approach as well as provide on-line characterization data on the separated components of a mixture (39). The workhorse detection method is UV-absorbance, which is convenient and relatively sensitive for compounds such as peptides and proteins. Scanning and diodearray detectors expand the utility of U V detection by providing an absorbance spectrum that yields additional information on the identity of a peak (40). Certain compounds, including tryptophan-containing proteins, fluoresce under UV-illumination, and so fluorescence detectors can provide higher sensitivity and selectivity than absorbance detection, particularly in the analysis of proteins from complex samples such as serum or cell culture broths. Electrochemical detection is also selective and so is a powerful approach for the quantitative identification of compounds that contain an electrochemically labile group. One important application of electrochemical detection is the analysis of carbohydrates released chemically or enzymatically from a glycoprotein, using pulsed amperometric detection ( P A D ) (32). Since carbohydrates are U V transparent, P A D methods provide a high sensitivity alternative to derivatization techniques for chromatographic detection. Other detection methods employed i n H P L C provide limited characterization data that can be interpreted to identify the peaks eluted from the chromatographic column. Light-scattering detection has been employed to

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Downloaded by 89.163.34.136 on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch001

1.

FRENZ

Chromatographic Separations in Biotechnology

9

determine the molecular size of monomelic and aggregated proteins i n a variety o f modes of H P L C (41). Detection by on-line mass spectrometry provides molecular mass information that can be sufficient to identify a protein or peptide following H P L C (42). The recent development of the electrospray ionization interface (43) to the mass spectrometer has revolutionized M S and L C - M S of peptides and proteins by extending both the sensitivity and mass range of high resolution instruments. The broadening acceptance of mass spectrometry i n the life sciences, particularly following the commercialization of the electrospray interface, has underscored the utility o f M S as a tool of analytical protein chemistry and the symbiosis of H P L C and M S i n many biochemical applications. The detection methods available for H P L C continue to make advances in technology to improve, i n significant ways, the sensitivity, selectivity and information content provided in applications involving proteins. These improvements have already served to entrench even further the prominent role of H P L C i n the protein chemistry laboratory. One of the most important applications of H P L C in analytical biotechnology is peptide mapping: the separation o f the mixture of peptides produced by enzymatic digestion of a protein (44). The combination of the proteolytic selectivity afforded by digestion with proteases and the high resolution obtained with H P L C yields peptide maps that are unrivaled for identification and primary structure confirmation o f a protein (45). "Restriction proteases" such as trypsin cleave a protein into peptide fragments whose small size makes them much better behaved in H P L C systems than, usually, is the intact protein. The small size of the constituent peptides is also usually more amenable to rapid characterization by a combination of amino acid analysis, mass spectrometry and N-terminal sequencing than is the entire protein in one piece. Furthermore, enzymatic mapping procedures permit the ready identification of sites of glycosylation (45) and disulfide bond formation (46) or of other post-translational modifications that are not apparent from the c D N A sequence of a protein. The ubiquity of enzymatic mapping approaches to protein characterization can thus be traced to the close match between the capabilities of standard techniques employed i n protein chemistry and the optimal analyte size for highest resolution i n H P L C . Advances in mass spectrometer and protein sequencer design are only just beginning to enlarge the potential for detailed characterization of the sequence of intact proteins, so peptide mapping can be expected to remain a standard tool in the arsenal of the protein chemist for some time to come. A second important and rapidly evolving area in which H P L C separations have become essential biochemical tools is the characterization of the carbohydrate structures found on the surfaces of glycoproteins (47). The growing usage o f recombinant glycoproteins as pharmaceuticals has intensified the appreciation of the role o f glycosylation in mediating the structural integrity, physicochemical, biochemical and biological activity o f a protein as well as its pharmacokinetic disposition in vivo. The appreciation of the impact of glycosylation on so many properties of a protein has induced an awareness on the part of manufacturers of the need to monitor these attributes in the recombinant product. Figure 3 shows the biosynthetic processing o f oligosaccharides that goes on in the rough endoplasmic reticulum (RER) and G o l g i organelles, and illustrates some of the diversity o f structures that can result. H P L C approaches have made a large impact in carbohydrate characterization efforts, i n particular the combination of high p H anion exchange chromatography and pulsed amperometric detection ( H P A E - P A D ) (48). These instruments allow either mapping of the oligosaccharides liberated by endoglycosidase digestion of a glycoprotein (49) or composition profiling of the monosaccharides released by acidic hydrolysis (50). In both cases anion exchange chromatography at high p H permits high resolution separations of closely related carbohydrate species, and P A D is a sensitive, efficient means of detecting the achromophoric sugars. The H P A E - P A D approach can also be scaled to provide preparative amounts of oligosaccharides for further structural characterization by M S , tandem M S or nuclear magnetic resonance analyses (57). The development of H P A E - P A D simplified the characterization of carbohydrates compared

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Downloaded by 89.163.34.136 on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch001

10

CHROMATOGRAPHY IN BIOTECHNOLOGY

to earlier techniques, many o f REE which relied on derivatization o f the reducing end o f the glycan to facilitate separation and detection. Since oligosaccharides are relatively easy to derivatize i n this manner, these approaches remain important i n many applications and have been the basis for powerful separations implemented i n capillary electrophoresis systems (52). Finally, the separation o f classes of glycoforms o f intact proteins have been demonstrated in ion exchange (Frenz, J.; Quan, C P . ; Cacia, J.; MacNerny, T.; and Bridenbaugh, R. In preparation.) and immobilized metal affinity chromatographic systems (55). These approaches provide an even more rapid means o f quantifying the glycosylation pattern of a protein compared to digestion methods, Figure 3. Schematic pathway o f oligosaccharide and permit the isolation o f processing on newly synthesized glycoproteins i n the homeoglycosylated proteins that rough endoplasmic reticulum ( R E R ) and G o l g i can be useful for investigating bodies. The symbols represent: P , phosphate; • , N the influence o f glycosylation on acetylglucosamine; 0, mannose; A, glucose; Δ. , protein activity and other fucose; · , galactose; • , sialic acid. Reprinted with properties. These applications permission from reference 55. demonstrate the heights attained by H P L C i n discriminating among subtle variations o f the heterogeneous ensemble of glycoforms constituting a purified glycoprotein and should become increasingly important in the higher performance of liquid chromatography i n the future. Conclusions This volume brings together papers illustrating the advances made in chromatographic techniques for producing proteins at large scale and for the characterization o f glycoproteins. The body o f papers thus illustrates the approaches to streamline the development process for large s c i e protein purification, and the high resolution of the analytical techniques that can be brought to bear to characterize the product. A t either scale the resolution afforded by chromatographic techniques is the major factor that drives the choice of separation method. The process development chromatographer deploys his arsenal of columns i n order to select a single protein i n high yield from among the thousands of compounds that may be present in a cell culture broth. The analytical chromatographer similarly has his armory to tackle the problems of resolving and defining the contributors to microheterogeneify i n a purified product. One shared goal is the demonstration of manufacturing consistency, that is a prerequisite to product approval. Meeting this goal, along with the other objectives related to the development of the fruits o f biotechnology, requires all the innovations i n tools and techniques that continue to spring from equipment manufacturers and end-users, including those described i n this volume.

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Downloaded by 89.163.34.136 on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch001

1. FRENZ

Chromatographic Separations in Biotechnology

11

Literature cited 1. Liu, D.T.-Y. TrendsBiotechnol.1992, 10, 364. 2. Garnick, R.L.; Solli, N.J.; and Papa, P.A. Anal. Chem. 1988, 60, 2546. 3. Goff, S.A.; and Goldberg, A.L. Cell 1983, 41, 587. 4. Pennica, D.; Holmes, W.E.; Kohr, W.J.; Harkins, R.N.; Vehar, G.A.; Ward, C.A.; Bennett, W.F.; Yelverton, E.; Seeburg, P.H.; Heyneker, H.L.; Goeddel, D.V.; and Collen, D. Nature 1983, 301, 214. 5. McCormick, G.; Trahey, M.; Innis, M.; Dieckmann, B.; and Ringold, G. Mol. Cell. Biol. 1984, 4, 166. 6. Cull, M.; and McHenry, C.S. In Guide to Protein Purification, M.P. Deutscher, Ed., Methods in Enzymology 182, Academic Press: San Diego, California, 1990; pp. 147-153. 7. Marston, F.A.O.; and Hartley, D.L. In Guide to Protein Purification, M.P. Deutscher, Ed., Methods in Enzymology 182, Academic Press: San Diego, California, 1990; pp. 264-276. 8. Hancock, W.S. In High Performance Liquid Chromatography in Biotechnology, W.S. Hancock, Ed., John Wiley and Sons: New York, 1990; p. 1. 9. Manohar, V., and Hoffman, T. Trends Biotechnol. 1992, 10, 305. 10. Low, D. In High Performance Liquid Chromatography in Biotechnology, W.S. Hancock, Ed., John Wiley and Sons: New York, 1990; p. 117. 11. Unger, K.K.; Lork, K.D.; and Wirth, H.-J. In HPLC of Peptides, Proteins and Polynucleotides: Contemporary Topics and Applications, M.T.W. Hearn, Ed., VCH Publishers: New York, 1991; p. 59. 12. Ugelstad, J.; Mork, P.C.; Kaggerud, K.H.; Ellingsen, T.; and Berge, A. Adv. Colloid Interface Sci. 1980, 13, 101. 13. Kao, Y.; Nakamura, K.; and Hashimoto, T. J. Chromatogr. 1983 266, 358. 14. Vratny, P.; Mikes, O.; Strop, P.; Coupek, J.; Rexova-Bendova, L.; and Chadimova, D. J. Chromatogr. 1983, 257, 23. 15. Hjertén, S. Arch. Biochem. Biophys. 1962, 99, 466. 16. Hjertén, S. In HPLC of Peptides, Proteins and Polynucleotides: Contemporary Topics and Applications, M.T.W. Hearn, Ed., VCH Publishers: New York, 1991; p. 119. 17. Regnier, F.E. Nature (London) 1991, 350, 634. 18. Henry, M.P. In High Performance Liquid Chromatography in Biotechnology, W.S. Hancock, Ed., John Wiley and Sons: New York, 1990; p. 21. 19. Shaltiel, S.; and Er-el, Z. Proc. Natl. Acad. Sci. USA 1973, 52, 430. 20. Frenz, J.; Hancock, W.S.; Henzel, W.J.; and Horváth, Cs. In HPLC of Biological Molecules: Methods and Applications, K.M. Gooding and F.E. Regnier, Eds., Marcel Dekker: New York, 1990; p. 145. 21. Regnier, F.E.; and Chicz, R.M. In HPLC ofBiological Molecules: Methods and Applications, K.M. Gooding and F.E. Regnier, Eds., Marcel Dekker: New York, 1990; p. 77. 22. El Rassi, Z.; and Horváth, Cs. J. Chromatogr. 1986, 359, 255. 23. Kawasaki, T.; Takahashi S.; and Ikeda, K. Eur. J. Biochem. 1985, 152, 361. 24. Josíc, D.; Becker, Α.; and Reutter, W. In HPLC of Peptides, Proteins and Polynucleotides: Contemporary Topics and Applications, M.T.W. Hearn, Ed., VCH Publishers: New York, 1991;p. 469. 25. Antia, F.D.; and Horváth, Cs. Ann. Ν.Y. Acad. Sci. 1990, 589, 172. 26. Kunitani, M.; Johnson, D.; and Snyder, L.R. J. Chromatogr. 1986, 371, 313. 27. Liao, Α.; and Horváth, Cs. Ann. N.Y. Acad. Sci. 1990, 589, 182. 28. Tiselius, A. Ark. Kemi. MineralGeol.1943,16A,1. 29. Horváth, Cs. In The Science of Chromatography, S. Bruner, Ed. , Elsevier: Amsterdam, 1985; p. 179. 30. Frenz, J.; and Horváth, Cs. In HPLC--Advances and Perspectives, Volume 5, Cs. Horváth, Ed., Academic Press: New York, 1988; p. 211.

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

12

CHROMATOGRAPHY IN BIOTECHNOLOGY

Downloaded by 89.163.34.136 on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch001

31. 32. 33. 34. 35.

Geisow, M.J. Biotechnol. 1991, 9, 921. Basa, L.J.; and Spellman, M.W. J. Chromatogr. 1990, 499, 205. Kornfeld, R.; and Kornfeld, S. Ann. Rev. Biochem. 1985, 54, 631. Horváth, Cs.; and Lipsky, S.R. Nature (London) 1966, 211, 748. Dorsey, J.G.; Foley, J.P.; Cooper, W.T.; Barford, R.A.; and Barth, H.G. Anal. Chem. 1992, 64, 353R. 36. Laemmli, U.K. Nature (London) 1970, 227, 680. 37. Oroszlan, P.; Wicar, S.; Teshima, G.; Wu, S.-L.; Hancock, W.S.; and Karger, B.L. Anal. Chem. 1992, 64, 1623. 38. Cacia, J.; Quan, C.P.; Vasser, M; Sliwkowski, M.B.; and Frenz, J. J. Chromatogr.; in press. 39. Yeung, E.S., Ed. Detectors for Liquid Chromatography, Wiley: New York, 1986. 40. Sievert, H.-J.P.; Wu, S.-L.; Chloupek, R.; and Hancock, W.S. J. Chromatogr. 1990, 499, 221. 41. Krull, I.S.; Suiting, H.H.; and Krzysko, S.C. J. Chromatogr. 1988, 442, 29. 42. Caprioli, R.M.; Moore, W.T.; DaGue, B.; and Martin, M. J. Chromatogr. 1988, 443, 335. 43. Fenn, J.B.; Mann, M.; Meng, C.K.; Wong, S.F.; and Whitehouse, C.M. Science 1989, 246, 64. 44. Hancock, W.S.; Bishop, C.A.; and Hearn, M.T.W. Anal. Biochem. 1979, 89, 203. 45. Chloupek, R.C.; Harris, R.J.; Leonard, C.K.; Keck, R.G.; Keyt, B.A.; Spellman, M.W.; Jones, A.J.S.; and Hancock, W.S. J. Chromatogr. 1989, 463, 375. 46. Becker, G.W.; Tackitt, P.M.; Bromer, W.W.; Lefeber, D.S.; and Riggin, R.M. Biotechnol. Appl. Biochem. 1988, 10, 325. 47. Parekh, R.B.; and Patel, T.P. Trends Biotech. 1992, 10, 276. 48. Hughes, S.; and Johnson, D.C. Anal. Chim. Acta 1981, 132, 11. 49. Maley, F.; Trimble, R.B.; Tarentino, A.L.; and Plummer, T.H. Anal. Biochem. 1989, 180, 195. 50. Edge, A.S.B.; Faltynek, C.R.; Hof, L.; Reichert, L.E.; and Weber, P. Anal. Biochem. 1981, 118, 131. 51. Spellman, M.W.; Basa, L.J.; Leonard, C.K.; Chakel, J.Α.; O'Connor, J.V.; Wilson, S.; and van Halbeek, H. J. Biol. Chem. 1989, 264, 14100. 52. Nashabeh, W.; and El Rassi, Z. J. Chromatogr. 1991, 536, 31. 53. Porath, J. J. Chromatogr. 1988, 443, 3. 54. Hancock, W.S. LC-GC 1992, 10, 96. 55. Rudolph, F.B.; Wiesenborn, D.; Greenhut, J.; and Harrison, M.L. In HPLC of Biological Molecules: Methods and Applications, K.M. Gooding and F.E. Regnier, Eds., Marcel Dekker: New York, 1990; p. 668. RECEIVED December 15, 1992

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Chapter 2

Continuous Purification of Proteins by Selective Nonadsorptive Preparative Chromatography

Downloaded by STANFORD UNIV GREEN LIBR on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch002

T. K. Nadler and F. E. Regnier Department of Chemistry, School of Science, Purdue University, West Lafayette, IN 47907

Continuous forms of chromatography are important to the biotechnology industry for the production of therapeutic proteins. Selective non-adsorption preparative (SNAP) chromatography may be used in a cross-current, simulated moving bed to produce a continuous purification system. It has the advantages of speed, efficient use of packing material, scalability, decrease cost of pumping systems, gentleness with regard to protein denaturation, and reduced dilution of the sample. An immobilized metal affinity chromatography (IMAC) column may be added in tandem with SNAP to concentrate as well as purify the protein. T h e r e i s a g r e a t need i n t h e b i o t e c h n o l o g y i n d u s t r y f o r s e p a r a t i o n t e c h n o l o g y o f i n c r e a s e d f l e x i b i l i t y , s p e c i f i c t h r o u g h p u t , and economic e f f i c i e n c y t h a t may be used i n t h e p r o d u c t i o n o f t h e r a p e u t i c p r o t e i n s . These a t t r i b u t e s a r e o f t e n a s s o c i a t e d w i t h c o n t i n u o u s p r o c e s s e s and i s t h e r e a s o n t h a t c o n t i n u o u s p u r i f i c a t i o n systems a r e now o f g r e a t interest. Methods f o r continuous purification o f p r o t e i n s by chromatographic, e l e c t r o p h o r e t i c , aqueous two-phase e x t r a c t i o n and reversed m i c e l l e separation technology are c u r r e n t l y being explored. T h i s paper f o c u s e s b r i e f l y on c o n t i n u o u s c h r o m a t o g r a p h i c t e c h n i q u e s i n g e n e r a l and p r i m a r i l y on a new c o n t i n u o u s c h r o m a t o g r a p h i c technique c a l l e d s e l e c t i v e n o n - a d s o r p t i o n p r e p a r a t i v e chromatography (SNAP).

Continuous Forms of Chromatography C o n t i n u o u s c h r o m a t o g r a p h i c s e p a r a t i o n s were f i r s t w i d e l y used i n t h e p e t r o l e u m i n d u s t r y (1,2) where work i n i t i a l l y c o n c e n t r a t e d on t h e development o f l a r g e - s c a l e b a t c h s e p a r a t i o n s . Continuous systems became f a v o r e d because t h e y were more flexible, c o u l d be r u n u n a t t e n d e d , d e c r e a s e t h e need f o r r e c y c l i n g e f f l u e n t , and u t i l i z e d s o r b e n t m a t e r i a l s more e f f e c t i v e l y . Much o f t h i s work has been a p p l i e d r e c e n t l y i n t h e development o f c o n t i n u o u s c h r o m a t o g r a p h i c methods f o r proteins (3-5).

0097-6156/93/0529-0014$06.00/0 © 1993 American Chemical Society

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

2.

NADLER & REGNIER

Continuous Purification of Proteins

15

Downloaded by STANFORD UNIV GREEN LIBR on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch002

M u l t i p l e approaches have been t a k e n i n t h e c o n s t r u c t i o n o f c o n t i n u o u s chromatography systems. Counter-current flow, c r o s s c u r r e n t f l o w , c o - c u r r e n t f l o w , c o n t i n u o u s s t i r r e d t a n k s , and s e l e c t i v e n o n - a d s o r p t i o n a r e a l l examples o f c o n t i n u o u s systems. Counter-Current Flow Separations. These systems a r e based on t h e m o b i l e phase and s o r b e n t moving i n o p p o s i t e d i r e c t i o n s as t h e name i m p l i e s . Because c o u n t e r - c u r r e n t systems g e n e r a l l y s e p a r a t e a m i x t u r e i n t o o n l y two f r a c t i o n s , s e l e c t i v i t y i s v e r y i m p o r t a n t w i t h complex mixtures. Moving beds i n which b o t h t h e s o r b e n t and s o l v e n t a r e t r a n s p o r t e d (1,2), s i m u l a t e d moving beds i n which t h e s o r b e n t i n a f i x e d bed appears t o move t h r o u g h t h e use o f v a l v e s a l o n g t h e column a x i s (6,7), a moving column approach t h a t u s e s column s w i t c h i n g t o s i m u l a t e s o r b e n t t r a n s p o r t (8,9), and f l u i d i z e d bed chromatography (10,11) a r e a l l forms o f c o u n t e r - c u r r e n t s e p a r a t i o n s . The p r o b l e m w i t h these systems r e l a t i v e t o p r o t e i n s i s t h a t t h e y a r e g e n e r a l l y i s o c r a t i c e l u t i o n systems and p r o t e i n s a r e not e a s i l y s e p a r a t e d i s o c r a t i c a l l y . Cross-Current Flow Separations. A n n u l a r chromatography i s t h e most t y p i c a l example o f a c r o s s - c u r r e n t f l o w system (12-16). Rotating a n n u l a r columns, i n which t h e s o r b e n t i s p l a c e d between two c o n c e n t r i c c y l i n d e r s t o form an a n n u l a r column, have been e f f e c t i v e i n s i z e e x c l u s i o n s e p a r a t i o n s o f p r o t e i n s ( 5 ) . Moving column systems w i t h a r o t a t i n g b u n d l e o f columns a r e a n o t h e r v e r s i o n o f t h i s approach (13,17,18). L i q u i d i n t h e moving column system i s d e l i v e r e d t o t h e columns and e l u e n t t o t h e c o l l e c t i o n f l a s k s by l i q u i d d i s t r i b u t o r s . A t the p r e s e n t t i m e o n l y i s o c r a t i c s e p a r a t i o n s o f p r o t e i n s have been a c h i e v e d w i t h t h e s e systems. I t does not appear t h a t t h e s p e c i f i c t h r o u g h p u t o f t h e s e systems i s any h i g h e r t h a n t h a t o f t h e i n d i v i d u a l columns c o m p r i s i n g t h e system. Moving Belt Separation. A v e r y unique system has been c o n s t r u c t e d by u s i n g a moving mylar b e l t on which n y l o n pouches o f immunosorbent were a t t a c h e d (19). The pouches resembled t e a bags o f immunosorbent t h r o u g h which p r o t e i n s c o u l d d i f f u s e w i t h o u t l o o s i n g t h e a d s o r b e n t . The b e l t was f i r s t d i p p e d i n t o t h e a s s o c i a t i o n chamber where i t s e l e c t i v e l y bound t h e a n t i g e n o f i n t e r e s t ( i n t h i s c a s e a l k a l i n e p h o s p h a t a s e ) . As t h e b e l t e x i t e d each chamber, t h e pouches were squeezed by a r o l l e r t o remove e x c e s s liquid. Unbound p r o t e i n was removed and antigen c o l l e c t e d by s e q u e n t i a l p a s s a g e t h r o u g h a wash chamber and d i s s o c i a t i o n chamber r e s p e c t i v e l y . Subsequent passage t h r o u g h a n o t h e r wash chamber r e c y c l e d t h e s o r b e n t pouches. A system u s i n g 5 g o f immunosorbent was c a p a b l e o f a t l e a s t a 10 f o l d p u r i f i c a t i o n w i t h 90% r e c o v e r y and a 35-40 h r c y c l e t i m e (19). Although continuous, the s p e c i f i c throughput o f t h i s system was v e r y low. I t i s q u e s t i o n a b l e whether t h i s system i s e a s i l y s c a l e d up. Continuous S t i r r e d Tank Reactors (CSTR). These systems have been used in s i t u a t i o n s where one o r two equilibrium contact stages are s u f f i c i e n t f o r a h i g h y i e l d s e p a r a t i o n (3,20,21). Each stage, or c o n t a c t o r , i s e q u i v a l e n t t o one t h e o r e t i c a l p l a t e . The CSTR must have a t l e a s t two s t a g e s , one f o r a d s o r p t i o n and a n o t h e r f o r d e s o r p t i o n o r r e c y c l i n g t o be c o n t i n u o u s . A d s o r b e n t must be s e p a r a t e d from t h e s o l v e n t i n each s t a g e . I t i s t h i s s e p a r a t i o n p r o c e s s where CSTR systems d i v e r g e i n d e s i g n . The c o n t i n u o u s a f f i n i t y - r e c y c l e e x t r a c t i o n

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

16

CHROMATOGRAPHY IN BIOTECHNOLOGY

(CARE) system o f Gordon (3) i s a good example o f a CSTR p u r i f i c a t i o n system. 6 - g a l a c t o s i d a s e from Ε, Coli was p u r i f i e d 18 f o l d w i t h 79% r e c o v e r y i n a two s t a g e CARE system u s i n g an a f f i n i t y s o r b e n t f o r 6 - g a l a c t o s i d a s e (PABTG/Agarose). A d d i t i o n o f a wash s t a g e between t h e a d s o r b i n g and d e s o r b i n g s t a g e i n c r e a s e d t h e p u r i f i c a t i o n t o 22 f o l d and recovery t o 90%. In a c o u n t e r - c u r r e n t a d s o r p t i o n d e s i g n , the purification jumped t o 170 f o l d , but r e c o v e r y fell t o 72% (3). O p t i m i z a t i o n o f t h e s e systems i s c r i t i c a l . D r a s t i c changes i n y i e l d and r e c o v e r y may be o b t a i n e d w i t h minor a l t e r a t i o n s o f system d e s i g n and o p e r a t i o n .

Downloaded by STANFORD UNIV GREEN LIBR on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch002

S e l e c t i v e N o n - A d s o r p t i o n P r e p a r a t i v e (SNAP) Chromatography C h r o m a t o g r a p h i c s e p a r a t i o n s o f p r o t e i n s a r e a c h i e v e d by d i f f e r e n t i a l adsorption at surfaces in all cases except size-exclusion chromatography. These s u r f a c e m e d i a t e d s e p a r a t i o n s a r e g e n e r a l l y b a s e d on s e l e c t i v e e l u t i o n o f s u b s t a n c e s from a s o r b e n t s u r f a c e . E l u t i o n i s a c h i e v e d i n e i t h e r an i s o c r a t i c o r g r a d i e n t e l u t i o n r e g i m e . Because p r o t e i n s adsorb t o s u r f a c e s a t m u l t i p l e s i t e s , v e r y s m a l l changes i n s o l v e n t c o m p o s i t i o n have a l a r g e impact on t h e a d s o r p t i o n / d e s o r p t i o n e q u i l i b r i u m and i s o c r a t i c e l u t i o n i s seldom used . E l u t i o n i s more commonly a c h i e v e d w i t h a s o l v e n t g r a d i e n t o f i n c r e a s i n g d e s o r b i n g strength. U n f o r t u n a t e l y g r a d i e n t e l u t i o n i s a problem i n continuous chromatography t h a t would be d e s i r a b l e t o circumvent. Protein d e n a t u r a t i o n i s a n o t h e r problem i n c h r o m a t o g r a p h i c systems. Some p r o t e i n s d e n a t u r e when t h e y a r e adsorbed t o a s o r b e n t s u r f a c e ( 2 2 ) . I t would be d e s i r a b l e t o p u r i f y p r o t e i n s w i t h o u t h a v i n g t o adsorb them t o t h e chromatography column. The f a c t t h a t p r o t e i n s a r e p u r i f i e d by a s e r i e s of d i s c r e t e p u r i f i c a t i o n steps i s yet another negative f e a t u r e of c u r r e n t s e p a r a t i o n systems. M u l t i p l e step p u r i f i c a t i o n s are both l a b o r i n t e n s i v e and d i m i n i s h r e c o v e r y . I t would be d e s i r a b l e t o combine s t e p s i n p r o t e i n p u r i f i c a t i o n . The s e l e c t i v e n o n - a d s o r p t i o n p r e p a r a t i v e chromatography (SNAP) approach d e s c r i b e d below i s an e f f o r t t o a d d r e s s t h e s e problems i n t h e c a s e o f p r e p a r a t i v e chromatography. SNAP chromatography i s u n i q u e l y d i f f e r e n t from o t h e r forms o f chromatography i n t h a t i t removes a l l o f t h e p r o t e i n s i n a sample e x c e p t t h e one o f i n t e r e s t . The t a r g e t p r o t e i n i s a l l o w e d t o p a s s through the column u n r e t a i n e d . The ion-exchange form o f SNAP chromatography was first reported by Petrilli et a l . for the p u r i f i c a t i o n o f a s p a r t a t e a m i n o - t r a n s f e r a s e ( 2 3 ) . The system s e p a r a t e s p r o t e i n s by t a k i n g advantage o f t h e o b s e r v a t i o n t h a t r e t e n t i o n o f a p r o t e i n on ion-exchange chromatography columns i s m i n i m a l when i t i s a t i t s i s o e l e c t r i c p o i n t ( p i ) . ( I t w i l l be r e c a l l e d t h a t t h e n e t c h a r g e of a p r o t e i n i s z e r o a t i t s p i . ) At pH v a l u e s above i t s p i , a p r o t e i n has a net n e g a t i v e c h a r g e and w i l l adsorb t o an a n i o n exchange s o r b e n t (Figure 1). In c o n t r a s t , t h e p r o t e i n i s p o s i t i v e l y c h a r g e d and w i l l a d s o r b t o a c a t i o n e x c h a n g i n g m a t r i x below i t s p i . Retention of a p r o t e i n on an ion-exchange column a t i t s p i can o c c u r i f i t has an uneven c h a r g e d i s t r i b u t i o n (e.g., a group o f n e g a t i v e l y charged r e s i d u e s s e p a r a t e d from t h e p o s i t i v e l y c h a r g e d r e s i d u e s ) . P e t r i l l i e t a l . u s e d an anion-exchange column f o l l o w e d by a c a t i o n - e x c h a n g e column (23), however, a s i n g l e SNAP column may a l s o be p r e p a r e d by m i x i n g anion-exchange and cation-exchange resins. The novelty of this a p p r o a c h i s t h a t by o p e r a t i n g t h e column a t t h e p i o f t h e t a r g e t

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

2.

NADLER & REGNIER

Continuous Purification of Proteins

17

Downloaded by STANFORD UNIV GREEN LIBR on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch002

p r o t e i n i t p a s s e s t h r o u g h t h e column u n r e t a i n e d w h i l e o t h e r p r o t e i n s are r e t a i n e d . SNAP chromatography o f f e r s s e v e r a l advantages o v e r c o n v e n t i o n a l g r a d i e n t - e l u t i o n chromatography. (1) S i n c e t h e p r o t e i n o f i n t e r e s t i s not a d s o r b e d t o t h e s u r f a c e , i t s s t r u c t u r e s h o u l d not be a l t e r e d by t h e media. T h i s means t h a t t h e p r o t e i n s h o u l d r e t a i n more a c t i v i t y . (2) The s e p a r a t i o n i s v e r y q u i c k because t h e p r o t e i n p a s s e s t h r o u g h t h e column u n r e t a i n e d . L o s s o f p r o t e i n from a d s o r p t i o n , p r o t e o l y s i s , and t i m e dependent dénaturâtion phenomena w i l l be d r a m a t i c a l l y r e d u c e d . (3) The e n t i r e chromatography column i s used d u r i n g t h e p u r i f i c a t i o n p r o c e s s . T h i s maximizes t h e e f f i c i e n c y o f t h e a d s o r b e n t and i n c r e a s e s the amount o f p r o t e i n p r o c e s s e d p e r volume o f a d s o r b e n t . (4) SNAP chromatography i s e a s i l y s c a l e d up f o r l a r g e r s e p a r a t i o n s by making t h e columns l a r g e r . (5) The pumping system i s l e s s e x p e n s i v e because s i m p l e r i s o c r a t i c systems a r e used i n l i e u o f more e x p e n s i v e g r a d i e n t pumping systems. (6) The sample i s not d i l u t e d by t h e s e p a r a t i o n p r o c e s s because i t i s p u r i f i e d by a form o f f r o n t a l chromatography. P e t r i l l i e t a l . s u g g e s t e d t h e name " I s o e l e c t r i c Chromatography" (23) f o r t h e i o n exchange approach, however, SNAP chromatography can be u s e d i n more t h a n t h e ion-exchange mode, a l b e i t w i t h a d i f f e r e n t s e p a r a t i o n mechanism. P r e l i m i n a r y r e s u l t s from S t r i n g h a m e t a l . (24,25) show t h a t tandem h y d r o p h o b i c i n t e r a c t i o n (HIC), p r o t e i n A a f f i n i t y , and s i z e - e x c l u s i o n chromatography (SEC) columns may be used i n SNAP s e p a r a t i o n s . SNAP s e p a r a t i o n s i n t h e a f f i n i t y and i m m o b i l i z e d m e t a l a f f i n i t y modes may a l s o be used. I m m u n o a f f i n i t y columns would be w e l l s u i t e d t o remove t r a c e c o n t a m i n a n t s . However, i t i s o f t e n d i f f i c u l t t o produce an a n t i b o d y t o each s i n g l e c o n t a m i n a n t . Some contaminants a r e more immunogenic t h a n o t h e r s , so most o f t h e a n t i b o d i e s would be d i r e c t e d a g a i n s t m a t e r i a l s o f h i g h immunogenicity and few o r no a n t i b o d i e s a r e produced a g a i n s t t h e r e s t . I t would be n e c e s s a r y t o i s o l a t e each contaminant and produce a n t i b o d i e s a g a i n s t it. T h i s i s not f e a s i b l e because many o f t h e c o n t a m i n a n t s a r e not even i d e n t i f i e d , much l e s s p u r i f i e d . Work r e p o r t e d by A n i c e t t i , a t Genetech (26), i n v o l v e d a c a s c a d e immunization scheme t o produce a n t i b o d i e s f o r a wide range o f a n t i g e n s . However, t h e i r work f o c u s e d on d e v e l o p i n g a n t i b o d i e s f o r a n a l y t i c a l methods t o d e t e c t t r a c e c o n t a m i n a n t s t h a t r e m a i n a f t e r p r e l i m i n a r y p u r i f i c a t i o n , not t h e p u r i f i c a t i o n i t s e l f . D e s c r i p t i o n o f t h e C o n t i n u o u s SNAP System. A f t e r a c e r t a i n amount o f use, a SNAP chromatography column will become saturated with c o n t a m i n a n t s and w i l l need t o be r e c y c l e d . To make t h e p r o c e s s c o n t i n u o u s , a second column must be used w h i l e t h e s a t u r a t e d column i s b e i n g c l e a n e d . The l o a d i n g and c l e a n i n g c y c l e i s d i v i d e d i n t o t h r e e p h a s e s . (1) The l o a d i n g phase, when sample i s b e i n g p u r i f i e d by t h e column, o c c u p i e s 50% o f t h e c y c l e t i m e . (2) D e s o r p t i o n u s i n g a 1.0 M s a l t b u f f e r r e q u i r e s 25% o f t h e c y c l e , and (3) r e - e q u i l i b r a t i o n w i t h the sample b u f f e r consumes t h e r e m a i n i n g 25% o f t h e c y c l e . Since d e s o r p t i o n and r e - e q u i l i b r a t i o n t a k e no l o n g e r t h a n l o a d i n g , two columns may be used i n p a r a l l e l f o r c o n t i n u o u s p u r i f i c a t i o n . One column i s d e s o r b e d and r e - e q u i l i b r a t e d w h i l e t h e o t h e r i s l o a d e d . The r e - e q u i l i b r a t e d column w i l l be r e a d y t o l o a d a g a i n when t h e o t h e r column becomes s a t u r a t e d . A t e n - p o r t v a l v e arrangement ( F i g u r e 2) was u s e d t o o p e r a t e t h e SNAP columns i n tandem i n a c o n t i n u o u s p u r i f i c a t i o n

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

CHROMATOGRAPHY IN BIOTECHNOLOGY

18

Retention Map

\

/

\

/

Downloaded by STANFORD UNIV GREEN LIBR on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch002

\ \ Cation Exchanger \ ^

/ /

Anion Exchanger

\ /

pi Buffer p H F i g u r e 1. A t pH v a l u e s below t h e p i o f t h e p r o t e i n , i t i s p o s i t i v e l y c h a r g e d and may be a d s o r b e d by c a t i o n e x c h a n g e r s , as shown by t h e i n c r e a s e i n t h e c a p a c i t y f a c t o r a t low pH. A t pH v a l u e s above t h e p i o f a p r o t e i n , i t w i l l be n e g a t i v e l y c h a r g e d and may be a d s o r b e d by a n i o n e x c h a n g e r s .

F i g u r e 2. D u a l SNAP columns were used t o p r o c e s s a c o n t i n u o u s s t r e a m o f d i l u t e d b o v i n e serum. A ten-port v a l v e switched the m o b i l e phases between serum and d e s o r p t i o n b u f f e r so t h a t as one column l o a d e d , t h e o t h e r was b e i n g r e c y c l e d .

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

2.

NADLER & REGNIER

19

Continuous Purification of Proteins

system. The v a l v e was automated w i t h a pneumatic a c t u a t o r c o m p u t e r i z e d LC system c o u l d r u n t h e p u r i f i c a t i o n u n a t t e n d e d .

so

a

Downloaded by STANFORD UNIV GREEN LIBR on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch002

The use o f p e r f u s a b l e (POROS) p a c k i n g m a t e r i a l s f o r t h e a d s o r b e n t a l l o w s h i g h f l o w r a t e s w i t h o u t i n c r e a s i n g back p r e s s u r e o r s l o w i n g mass t r a n s f e r (27,28). T h i s m a t e r i a l has been d e s i g n e d w i t h l a r g e t h r o u g h p o r e s t o improve mass t r a n s f e r a t h i g h f l o w r a t e s . L i n e a r v e l o c i t i e s i n e x c e s s o f 1 cm/sec on 250 mm χ 4.6 mm columns were used i n some experiments. SNAP chromatography, as w e l l as many o t h e r forms o f c o n t i n u o u s p r o t e i n separations, functions best with r e l a t i v e l y d i l u t e s o l u t i o n s ( c a . 1 mg/ml). T h i s means t h a t t h e p u r i f i e d p r o t e i n i s even more d i l u t e b e c a u s e i t o n l y r e p r e s e n t e d a f r a c t i o n o f t h e t o t a l p r o t e i n (5% would be 50 μ g / m l ) . I t i s o f t e n necessary t o concentrate the product solution by some means such as u l t r a f i l t r a t i o n . However, the s e p a r a t i o n p r o c e s s i t s e l f does not d i l u t e t h e sample, b e c a u s e t h e sample élûtes as a f r o n t a l c u r v e . An i m m o b i l i z e d m e t a l a f f i n i t y chromatography (IMAC) column was added i n tandem t o t h e SNAP columns t o adsorb and c o n c e n t r a t e t h e d i l u t e p r o d u c t p r o t e i n . IMAC (29-34) u t i l i z e s an i m m o b i l i z e d m e t a l i o n t o c h e l a t e h i s t i d i n e r e s i d u e s on p r o t e i n s . Copper I I was used b e c a u s e i t has t h e b r o a d e s t s p e c i f i c i t y f o r d i f f e r e n t p r o t e i n s (35). In t h i s way, t h e SNAP-IMAC t e c h n i q u e i s more g e n e r a l f o r p r o t e i n s o t h e r t h a n IgG. The p u r i f i e d p r o t e i n e l u t i n g from t h e SNAP columns a d s o r b s and c o n c e n t r a t e s on t h e IMAC column. The p r o t e i n i s t h e n e l u t e d from t h e IMAC column w i t h a p u l s e o f i m i d a z o l e t o d i s p l a c e t h e p r o t e i n . I m i d a z o l e does not s t r i p t h e column o f t h e m e t a l i o n , so t h e r e i s no need t o r e l o a d t h e IMAC column w i t h m e t a l . F u r t h e r , s i n c e i m i d a z o l e i s a s m a l l m o l e c u l e , i t may e a s i l y be removed, i f n e c e s s a r y , from IgG by using d i a l y s i s or a s i z e - e x c l u s i o n c a r t r i d g e . F u r t h e r p u r i f i c a t i o n , as w e l l as c o n c e n t r a t i o n , may be p e r f o r m e d on t h e IMAC column. By u s i n g a s t e p p e d g r a d i e n t o f i m i d a z o l e , t h e c o n t a m i n a n t s t h a t a r e more weakly adsorbed t o t h e IMAC column may be e l u t e d f i r s t w i t h a low c o n c e n t r a t i o n o f i m i d a z o l e . Then a h i g h e r c o n c e n t r a t i o n o f i m i d a z o l e may be used t o remove t h e c o n c e n t r a t e d p r o t e i n . F i n a l l y , a v e r y c o n c e n t r a t e d i m i d a z o l e p u l s e c o u l d be used t o remove any contaminant t h a t remained. The system used i n t h e s e s t u d i e s f a l l s i n t o t h e c a t e g o r y o f a c r o s s - c u r r e n t s i m u l a t e d moving bed and was used t o p u r i f y a s i n g l e p r o t e i n from a m i x t u r e . However, c r o s s - c u r r e n t systems can p u r i f y more t h a n one p r o t e i n s i m u l t a n e o u s l y . I f g r a d i e n t e l u t i o n were used t o d e s o r b t h e s a t u r a t e d column, a f r a c t i o n c o l l e c t o r a t t h e waste o u t l e t c o u l d c o l l e c t o t h e r p r o t e i n s as t h e y e l u t e d u r i n g t h e wash g r a d i e n t . However, p u r i f i c a t i o n o f m u l t i p l e c o n s t i t u e n t s s i m u l t a n e o u s l y may r e q u i r e t h e a d d i t i o n o f more v a l v e s and columns t o t h e system. Because two o r more p r o t e i n s a r e r a r e l y p u r i f i e d s i m u l t a n e o u s l y , no e f f o r t was given to the p u r i f i c a t i o n of m u l t i p l e p r o t e i n s . Observations. C o n c e n t r a t e d f e e d streams l o a d t h e SNAP chromatography columns f a s t e r t h a n t h e y may be desorbed, so t h e system works b e s t w i t h d i l u t e f e e d streams ( l e s s t h a n 1 mg/ml). I f t h e f e e d stream i s t o o

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

CHROMATOGRAPHY IN BIOTECHNOLOGY

20

c o n c e n t r a t e d , p r o d u c t w i l l a l s o be wasted. Each t i m e t h e SNAP column i s d e s o r b e d , t h e column s t i l l c o n t a i n s t h e components o f t h e f e e d s t r e a m i n i t s v o i d volume. These components i n c l u d e t h e p r o d u c t p r o t e i n as w e l l as t h e c o n t a m i n a n t s . The p r o d u c t y i e l d , o r r e c o v e r y o f t h e SNAP chromatography system i s a f u n c t i o n o f t h e column v o i d volume (V ) and t h e volume o f f e e d s o l u t i o n t h a t p a s s e s t h r o u g h t h e column (V.) b e f o r e t h e column becomes s a t u r a t e d . Optimum r e c o v e r y (R) i s r e l a t e d t o t h e s e v a r i a b l e s by t h e f o l l o w i n g e q u a t i o n : 0

Downloaded by STANFORD UNIV GREEN LIBR on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch002

*=

1

0

0

%

To o p t i m i z e r e c o v e r y , V needs t o be maximized and V m i n i m i z e d . U s u a l l y , v e r y l i t t l e c a n be done w i t h V s i n c e i t i s a f u n c t i o n o f t h e p a c k i n g m a t e r i a l . However, t h e use o f a p a c k i n g m a t e r i a l w i t h a h i g h l o a d i n g c a p a c i t y does i n c r e a s e V . A l s o , as t h e f e e d s t r e a m i s d i l u t e d , V, i n c r e a s e s because i t t a k e s l o n g e r t o s a t u r a t e t h e column. On t h e o t h e r hand, i t t a k e s e f f o r t t o r e c o n c e n t r a t e t h e p r o d u c t p r o t e i n , so t h e r e i s a compromise between d i l u t i o n and r e c o v e r y . t

0

0

t

One way t o a v o i d t h e r e c o v e r y problem i s t o u s e two s i x - p o r t v a l v e s i n s t e a d o f one t e n - p o r t v a l v e t o c o n t r o l t h e f l o w t h r o u g h t h e SNAP columns. The f i r s t v a l v e d e t e r m i n e s which column r e c e i v e s t h e f e e d s t r e a m and t h e second v a l v e d e t e r m i n e s which column i s c o n n e c t e d to the c o l l e c t i o n v e s s e l . A f t e r t h e f i r s t v a l v e t u r n s , t h e second v a l v e i s d e l a y e d e q u i v a l e n t t o t h e t i m e i t t a k e s t h e column t o p a s s t h e v o i d volume. R e c o v e r y c o u l d always be 100% as l o n g as t h e column d i d not s a t u r a t e i n l e s s t h a n one v o i d volume. S c a l e - u p o f t h e SNAP system i s s i m p l y a m a t t e r o f i n c r e a s i n g column s i z e and f l o w r a t e s . Since the separation i s i n s e n s i t i v e t o p l a t e h e i g h t , t h e u s u a l hydrodynamic problems a s s o c i a t e d w i t h l a r g e r columns a r e r e d u c e d . T h i s i s an advantage o v e r t h o s e continuous systems t h a t r e q u i r e m u l t i p l e columns w i t h e q u i v a l e n t f l o w p r o p e r t i e s . The SNAP system o n l y r e q u i r e s t h a t t h e d e s o r p t i o n t i m e o f t h e s l o w e s t column be l e s s t h a n t h e l o a d i n g t i m e o f t h e f a s t e s t column. Columns can be matched s i m p l y by a d j u s t i n g t h e f e e d r a t e r e l a t i v e t o t h e d e s o r p t i o n flow r a t e . In t h e experiments performed i n these studies, standard a n a l y t i c a l columns (4.6 mm i . d . χ 25 cm) were o p e r a t e d a t f l o w r a t e s o f 10 ml/min. T h i s means t h a t t h e l i n e a r f l o w r a t e was i n e x c e s s o f 1 cm/second. These f l o w r a t e s were p o s s i b l e b e c a u s e a p e r f u s a b l e a d s o r b e n t (POROS) was used t o pack t h e columns. H i g h t h r o u g h p u t s were then p o s s i b l e using r a p i d m u l t i p l e c y c l e s t o perform continuous s e p a r a t i o n s (See T a b l e I ) . The a d s o r b e n t i n t h e column i s used v e r y e f f i c i e n t l y i n r a p i d m u l t i p l e c y c l e SNAP chromatography. The r a t e o f p r o t e i n p r o c e s s e d f o r a g i v e n volume o f a d s o r b e n t may be g r e a t e r t h a n 1.0 mg/min/ml o f column. The IMAC column was p l a c e d i n tandem t o t h e SNAP columns ( F i g u r e 3) t o make a c o n t i n u o u s p u r i f i c a t i o n and c o n c e n t r a t i o n system. An i n j e c t i o n v a l v e was p l a c e d between t h e SNAP columns and t h e IMAC column

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

2.

NADLER & REGNIER Table

I.

D u a l Column SNAP R e s u l t s Experiment 2

Experiment 3

10

5

10

1.0

0.5

1.0

Serum D i l u t i o n

1/100

1/25

1/50

Serum P r o t e i n Cone, (mg/ml)

0.781

2.13

1.09

F r a c t i o n IgG ( r e l . area 1st peak)

8.85%

6.28%

7.47%

7.8

10.7

10.9

0.94

1.28

1.30

0.061

0.047

Experiment Flow r a t e (ml/min) L i n e a r Flow Rate (cm/sec)

Downloaded by STANFORD UNIV GREEN LIBR on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch002

21

Continuous Purification of Proteins

P r o c e s s i n g Rate (mg Serum/min) R a t e / C o l . Volume (mg/min/ml) IgG P r o t e i n Cone, (mg/ml)

0.084

1

-40%

Fold

4.52

12.8

12.7

48.9%

36.5%

54.8%

Purification

Approximate

Yield

>

80%

-95%

IgG P u r i t y ( r e l . a r e a 1 s t peak)

so p u l s e s o f a d e s o r p t i o n b u f f e r c o u l d be i n j e c t e d i n t o t h e IMAC column t o r e l e a s e t h e c o n c e n t r a t e d IgG. The v a l v e was f i t t e d w i t h a 2 ml sample l o o p f o r t h a t p u r p o s e . Once t h e IMAC column was s a t u r a t e d w i t h IgG, i t was e l u t e d w i t h 200 mM i m i d a z o l e t o y i e l d 95% p u r e IgG a t c o n c e n t r a t i o n s g r e a t e r t h a n 5 mg/ml. O t h e r methods o f c o n c e n t r a t i o n are a l s o p o s s i b l e . A f f i n i t y columns f o r t h e p r o t e i n may be p l a c e d a f t e r t h e SNAP columns t o c a p t u r e t h e d i l u t e p r o t e i n . Ion-exchange columns a l s o may be used, but t h e m o b i l e phase pH would have t o be a d j u s t e d so t h a t t h e p r o t e i n a c h i e v e d a net c h a r g e ( e i t h e r p o s i t i v e o r negative). IMAC was chosen because i t was s i m p l e t o use (no pH a d j u s t m e n t was n e c e s s a r y ) , and i t was f a i r l y g e n e r i c ( i t c o u l d be used f o r a number o f d i f f e r e n t p r o t e i n s because i t i s not as s p e c i f i c as a f f i n i t y columns). Some c o n t a m i n a n t s i n t h e p u r i f i e d IgG, were o b s e r v e d by SDS-PAGE. S e l e c t i v e e l u t i o n o f t h e s e c o n t a m i n a n t s was attempted i n order t o o b t a i n f u r t h e r p u r i f i c a t i o n as w e l l as c o n c e n t r a t i o n from t h e IMAC column. S a l t washes as h i g h as 1.0 M NaCl d i d not e l u t e any o f t h e IgG nor any o f t h e c o n t a m i n a n t s . The l o a d e d IMAC column was removed from t h e SNAP-IMAC system and e l u t e d w i t h a g r a d i e n t o f i m i d a z o l e . The IgG e l u t e d from t h e column a t a low i m i d a z o l e c o n c e n t r a t i o n ( l e s s t h a n 20 mM) w h i l e one contaminant e l u t e d a t 200 mM i m i d a z o l e .

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

CHROMATOGRAPHY IN BIOTECHNOLOGY

22

Downloaded by STANFORD UNIV GREEN LIBR on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch002

To t a k e advantage o f t h i s s e l e c t i v e e l u t i o n w i t h o u t t h e u s e o f a linear gradient, a series of pulses of increasing imidazole b o n c e n t r a t i o n were u s e d . The i m i d a z o l e p u l s e was i n t r o d u c e d v i a t h e i n j e c t i o n v a l v e between t h e SNAP and IMAC columns. A f r a c t i o n was c o l l e c t e d a f t e r e a c h p u l s e and a n a l y z e d by SDS-PAGE ( F i g u r e 4 ) . Some high-molecular-weight c o n t a m i n a n t s a r e removed by a 1 mM i m i d a z o l e p u l s e w h i l e most o f t h e IgG e l u t e s d u r i n g t h e 10 mM and 15 mM p u l s e s . A l o w - m o l e c u l a r - w e i g h t contaminant remains u n t i l a 200 mM p u l s e i s used. By u s i n g a s e r i e s o f t h r e e i m i d a z o l e p u l s e s , 5 mM t o e l u t e t h e high-molecular-weight c o n t a m i n a n t s , 15 mM t o e l u t e t h e IgG and 200 mM t o e l u t e t h e r e m a i n i n g c o n t a m i n a n t , t h e IgG i s f u r t h e r p u r i f i e d as w e l l as c o n c e n t r a t e d on t h e IMAC column. Optimization of SNAP. O p t i m i z a t i o n o f SNAP chromatography s e p a r a t i o n s has been d i s c u s s e d i n d e t a i l (24,25). To o p t i m i z e t h e SNAP system, one must choose a s a l t c o n c e n t r a t i o n j u s t h i g h enough t o e l u t e t h e p r o t e i n of i n t e r e s t . H i g h e r s a l t c o n c e n t r a t i o n s compromise t h e s e p a r a t i o n by desorbing contaminants. To f i n d t h e optimum s a l t concentration, S t r i n g h a m e t a l . (25) s t a r t e d w i t h an e q u a t i o n d e r i v e d by Snyder e t a l . (36). The e q u a t i o n r e l a t e s i s o c r a t i c and g r a d i e n t e l u t i o n . Retention t i m e ( t ) i s r e l a t e d t o average k' i n g r a d i e n t e l u t i o n (k*), column dead volume ( t ) and t h e k' a t t h e o n s e t o f t h e g r a d i e n t (k) by: R

0

t =t R

The

0

k> l o g ( 1 ^ 2 ) C +

0

(1)

a v e r a g e k' i s g i v e n by: k*=——2—

(2)

where t i s t h e d u r a t i o n o f t h e g r a d i e n t , (Δφ) i s t h e change i n volume f r a c t i o n o f t h e e l u e n t s o l v e n t (φ i s 1.0 when t h e g r a d i e n t i s r u n from 0 t o 100%) and S i s t h e change i n l o g k' f o r t h e u n i t change i n ψ i n isocratic elution. I n i s o c r a t i c systems, r e t e n t i o n i s g i v e n by: f

log(/e')=log(Jc )-S 0

Equation

(3)

(2) may be reduced t o :

When t h e e n t i r e g r a d i e n t i s r u n from 0 t o 100%. A c a p a c i t y f a c t o r o f 10 i s i n t h e s h a r p t r a n s i t i o n range d i s c u s s e d above. S u b s t i t u t i n g t h i s v a l u e i n t o e q u a t i o n 3 and r e a r r a n g i n g g i v e s : log(ic ) = 1 + 5φ 0

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

(5)

Downloaded by STANFORD UNIV GREEN LIBR on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch002

NADLER & REGNIER

Continuous Purification of Proteins

sample buffer

desorption buffer

F i g u r e 3. An IMAC column was added t o t h e d u a l - c o l u m n SNAP a p p a r a t u s . IgG was c o n c e n t r a t e d on t h e IMAC column and was e l u t e d with imidazole. A m i c r o p r o c e s s o r - c o n t r o l l e d pumping system was used t o automate t h e system.

F i g u r e 4. SDS-PAGE o f samples e l u t e d from t h e IMAC column a t v a r i o u s i m i d a z o l e c o n c e n t r a t i o n s (1 mM, 5 mM, 10 mM, 15 mM, 20 mM, 30 mM, 40 mM, 50 mM, 200 mM). IgG p r o d u c e s two bands b e c a u s e i t i s composed o f two heavy (H) c h a i n s and two l i g h t (L) c h a i n s . M o l e c u l a r weight s t a n d a r d s a r e i n d i c a t e d on t h e l e f t (68 KDa, 48 KDa, and 12.5 KDa).

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

CHROMATOGRAPHY IN BIOTECHNOLOGY

24 Substituting

(4) and

(5) i n t o e q u a t i o n

t =4?-% R

1 yields:

Clog (2 .3) +l +i*|>-log (

) ] +1

0

which s i m p l i f i e s t o : t -1.36 1 X j

yj = — „ i

y = iA...,»+i

+

(13)

1+ Σ [(σ,+ν,)^-!]^ i=2 where the mole fractions are defined as,

Xj =

yj J

^Total

%

y = i,2,...,«+i

(14)

A

F o r mono-valent ion-exchange systems without steric effects, Equation 13 reduces to the following stoichiometric exchange isotherm with constant separation factors derived by Helfferich and Klein (27):

yj =

^ Γ ι + Σ i=2

4

^

y = l,2,...,»+l

(15)

tai-i]*,

Determination of Equilibrium Parameters Moderately Retained Proteins: Log k' vs. Log Salt Method. The isotherms defined above require the determination o f three independent parameters: the characteristic charge o f the solute, v,, the equilibrium constant, K , and the steric factor, σ . Under linear conditions, well established relationships for protein ionexchange can be employed to determine the equilibrium constant and the characteristic charge (47-50). The logarithm o f the capacity factor (k) can be represented as (44), u

7

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

CHROMATOGRAPHY IN BIOTECHNOLOGY

32 Vi

logk'=log®K \ )

-

Xi

(16)

where the β is the column phase ratio. A plot o f the logarithm o f the capacity factor versus the logarithm o f the mobile phase salt concentration should yield a straight line with the following properties: slope = - v

(17a)

y ν

intercept = log ( β K

Downloaded by PENNSYLVANIA STATE UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch003

u

Λ ')

(17b)

Thus, from the linear elution behavior o f the solute under various mobile phase salt conditions one is able to determine the characteristic charge, v , and the equilibrium constant, K . The steric factor, which requires information from the non-linear portion o f the isotherm, is determined independently from a frontal experiment. The value o f the steric factor, σ , can be calculated directly from the breakthrough volume, V , using the following expression in concert with the values o f the characteristic charge and the equilibrium constant determined from the log k' analysis: ;

u

7

B

(18)

Λ - C,

where, V , is the column dead volume and the parameter Π is denned as, Q

(V Π

=

\

(19)

(f-l

Strongly Retained Displacers: Integration of Induced Salt Gradient Method. While the determination o f the characteristic charge and equilibrium constant from linear elution data works well for proteins which are moderately retained, solutes which are very strongly adsorbed are difficult to characterize in this fashion. Experimentally, these solutes exhibit regions with very strong retention or no retention at all. T o circumvent this experimental difficulty, the characteristic charge can be determined by measuring the change in the salt concentration caused by the frontal adsorption o f a solute at a concentration Cf The characteristic charge o f the solute is given by, v

v, - *·:;>•

PO)

where ACQ), is the increase in the salt concentration caused by the front o f solute i . For solutes which are very strongly adsorbed, the concentration o f solute adsorbed during the frontal procedure approaches its saturation capacity. I f the mobile phase salt concentration is sufficiently low, the concentration o f the adsorbed solute, Q can be assumed to be equal to Q™ *, the saturation capacity o f the solute. The steric factor can then be calculated directly from equation 7. i9

0

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

3.

CRAMER & BROOKS

33

Ion-Exchange Displacement Chromatography

One should be careful to only extract information about the steric factor from a frontal analysis on solutes which are retained enough to approach complete saturation o f the stationary phase. Typically, the isotherms o f highly charged polymers used as displacers in our laboratory approach square isotherms that are essentially invariant under l o w mobile phase salt conditions. Under these conditions the assumption o f complete saturation is reasonable and the frontal analysis is justified. Having determined the characteristic charge and steric factor, a second frontal experiment can then be performed at elevated salt conditions where the stationary phase concentration deviates from its maximum value. The equilibrium constant, K , can now be determined from the breakthrough volume o f the second front, V , using the following expression: u

Downloaded by PENNSYLVANIA STATE UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch003

B2

(21)

Thus, for strongly retained solutes (e.g., displacers) the determination o f equilibrium parameters requires data from two frontal experiments. The characteristic charge and steric factor are first determined form a frontal experiment performed under very low salt conditions where the solute completely saturates the stationary phase. The equilibrium constant is then determined from a second experiment performed at elevated salt conditions where the stationary phase concentration deviates from its maximum value. Calculation of Ideal Displacement Profiles Given a column o f sufficient length, displacement chromatography results in the formation o f a displacement train containing the feed components in adjacent pure zones. Under such isotachic conditions, the species velocities o f all the pure components in the displacement train are identical and equal to the velocity o f the displacer. In ion-exchange displacement chromatography, the adsorption o f the displacer causes an induced salt gradient which travels ahead o f the displacer. This increase in salt concentration is seen by the feed components traveling in front o f the displacer, and results in a depression o f the feed component isotherms. The determination o f ideal displacement profiles reduces to a problem o f determining the local salt microenvironment each feed component sees in the isotachic displacement train. The velocity o f the displacer, u , can be calculated, apriori, from the following expression: d

(22)

where u is the chromatographic velocity and the superscript, *, refers to equilibrium values in the isotachic displacement train. The ratio Q* ICf is determined from the displacer's single component isotherm (equation 6): Q

d

d

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

CHROMATOGRAPHY IN BIOTECHNOLOGY

34

ο;

" M

(co;

(

J

2

3

)

( Q ) ^ is the concentration o f salt the displacer sees (i.e., the carrier salt concentration). Under isotachic conditions, equality o f all species velocities implies,

c* -

Downloaded by PENNSYLVANIA STATE UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch003

2

c* -

- - os =

4

(

2

4

)

where 24 is the slope o f the displacer operating line. In a fully developed ideal displacement train the feed components are present in adjacent zones o f pure material. Thus, the single component equilibrium expression (equation 6) can also be used for each feed component in the ideal displacement train. Combining equations 6 and 24 yields the following expression.

where (C^)* is the local salt microenvironment seen by feed component i in the final isotachic displacement profile. It can be shown (45) that this salt concentration is given by the following expression:

=

r

(26)

Cl is the initial carrier salt concentration. &(C^) represents the increase in the salt concentration due to the introduction o f the displacer and is given by v Cf . The concentration o f feed component /, C*, in the isotachic displacement zone is given by, d

d

A - ( Q

d

{£1

Finally, the volume o f the isotachic displacement zone, V*, is given by,

V* = Vf^é

(28)

where Vf is the feed volume. Thus, once the slope o f the displacer operating line is determined (i.e., Δ\ the local salt microenvironment for each feed component can be determined from equation 26. The concentrations and widths o f the ideal isotachic displacement zones are given by equations 27 and 28, respectively.

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

3.

CRAMER & BROOKS

Ion-Exchange Displacement Chromatography

35

Downloaded by PENNSYLVANIA STATE UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch003

Experimental Materials. A Protein-Pak S P - 8 H R strong cation-exchange column (8 μΐη, 100 χ 5 mm I D . ) and a Protein-Pak Q P - 8 H R strong anion-exchange column (8 μπι, 100 χ 5 mm I D . ) were donated by Waters Chromatography Division o f Millipore (Milford, M A ) . Sodium chloride, sodium nitrate, sodium phosphate (mono and dibasic) as well as the proteins: α-chymotrypsinogen A , β-lactoglobulin A , β-lactoglobulin B , crude βlactoglobulin A - B mixture, cytochrome C and lysozyme were purchased from Sigma Chemical Company (St. Louis, M O ) . The displacers, 50kd dextran-sulfate and 40kd D E A E - d e x t r a n were donated by Pharmacia-LKB Biotechnology Inc., (Uppsala, Sweden); protamine sulfate was purchased from I C N Biochemicals (Cleveland, O H ) . Apparatus. The chromatographic system employed in this work consisted o f a M o d e l L C 2150 H P L C pump (Pharmacia L K B ) connected to the columns via a M o d e l C 1 0 W 10-port valve (Valco, Houston T X ) . The column effluent was monitored at 280 nm by a M o d e l 757 spectroflow U V - V i s detector (Applied Biosystems, Ramsey, N J ) and a M o d e l CR601 integrator (Shimadzu, Columbia, M D ) . Fractions o f the column effluent were collected with a M o d e l 2212 Helirac fraction collector (PharmaciaLKB). Procedures. A schematic o f the chromatograph system employed in the work is illustrated elsewhere (8). In all displacement experiments, the columns were sequentially perfused with carrier, feed, displacer and régénérant solutions. Fractions from the column effluent were collected throughout the displacement run and subsequently analyzed. Analytical chromatography was used for the quantification o f proteins and displacers. The concentrations o f sodium and chloride counter-ions were determined by atomic adsorption and titration, respectively (20). Results and Discussion Our laboratory has experimentally determined the S M A equilibrium parameters for a variety o f proteins and displacer compounds (20-22). In addition, we have developed a number o f ion-exchange displacement separations using an assortment o f displacers (20-23). The S M A equilibrium formalism was employed to simulate the displacement purification of α-chymotrypsinogen A and cytochrome C using a 40 kd DEAE-dextran displacer. A s seen in Figure 2, the model predicts that perfusion with the D E A E dextran displacer will result in displacement o f the proteins along with an induced salt gradient. Thus, the proteins see different salt concentrations in the fully developed displacement zones. The salt concentrations in the α-chymotrypsinogen A and cytochrome C zones are 120 and 116 m M , respectively, in contrast to the 75 m M salt present in the carrier. The S M A model can also be used to generate the adsorption isotherms (Figure 3) o f the proteins and displacer under the initial carrier salt conditions, as well as those present in the final isotachic displacement train. A s seen in the figure, the isotherm o f D E A E - d e x t r a n crosses the protein isotherms. Although the D E A E - d e x t r a n isotherm does not lie above the protein isotherms, the S M A formalism predicts that this displacement will result in complete separation of the feed proteins.

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

CHROMATOGRAPHY IN BIOTECHNOLOGY

Downloaded by PENNSYLVANIA STATE UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch003

Na

cytC a-chy

1 • . • 1 . . . . 1 .— 1

2

40 kd -DEAE- : dextran

H • . r — . — ι — . r--

3

4

5

Elution Volume (ml) Figure 2. Simulated effluent profile of 40kd DEAE-dextran displacement. (Reproduced with permission from ref. 21)

7

Mobile Phase Concentration (mM) Figure 3. Adsorption isotherms and displacer operating line for 40kd DEAE-dextran displacement. (Reproduced with permissionfromref. 21)

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

3.

CRAMER & BROOKS

37

Ion-Exchange Displacement Chromatography

The displacement experiment was then carried out to test the S M A model. A s seen in Figure 4, the experimental results are in good agreement with the predictions o f the model. B o t h the induced salt gradient and the displacement profiles match well with the theoretical predictions. In addition, an impurity in α-chymotrypsinogen A , not detectable by analytical H P L C in the dilute feed sample, was concentrated and purified in the first fraction o f the displacement zone.

150

Downloaded by PENNSYLVANIA STATE UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch003

1.6

3

4

Elution Volume (ml) Figure 4. Effluent profile of 40 kd DEAE-dextran displacement, permission from ref. 21)

(Reproduced with

This displacement differs from the traditional vision o f displacement in several important ways: the isotherm o f the displacer does not lie above the feed component isotherms; the concentration o f displaced protein exceeds the concentration o f the displacer; and the induced salt gradient produces different salt microenvironments for each displaced protein. These results demonstrate the ability o f the S M A formalism to predict complex behavior present in ion-exchange protein displacement systems. The displacement purification o f β-lactoglobulin A and Β has been reported by several investigators (15,18). W e have used this model separation to examine the efficacy o f dextran-sulfate displacers o f various molecular weights (20). The displacement purification o f β-lactoglobulin A and Β using a 50 kd dextran-sulfate displacer and the corresponding S M A simulation are shown in Figures 5 and 6, respectively. The S M A model predicts that the dextran-sulfate displacer will result in displacement o f the proteins along with an induced salt gradient. Again, the induced salt gradient produced by the displacer results in a depression o f the feed component isotherms and the proteins see different salt concentrations in the fully developed isotachic displacement zones.

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Downloaded by PENNSYLVANIA STATE UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch003

CHROMATOGRAPHY IN BIOTECHNOLOGY

1

2

3

4

5

Elution Volume [ml] Figure 5. Effluent profile of 50 kd dextran-sulfate displacement. (Reproduced with permission from ref. 21)

1.5 +

Na

ο 1.0 +

I 0.5 + 3

4

Elution Volume [ml] Figure 6. Simulated profile of 50 kd dextran-sulfate displacement. (Reproduced with permission from ref. 21)

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

3.

CRAMER & BROOKS

39

Ion-Exchange Displacement Chromatography

Downloaded by PENNSYLVANIA STATE UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch003

A s seen in the figures, the experimental results are in qualitative agreement with the predictions o f the model. Since β-lactoglobulins A and Β are known to aggregate (57), the current form o f the S M A model is not able to completely describe the adsorption isotherms o f these proteins (20). Nevertheless, the model still provides a reasonable prediction o f the displacement profile under induced salt gradient conditions. A s stated above, our laboratory is in the process o f developing a number o f dis­ placer compounds for a variety o f adsorbent systems. Figure 7 presents the displace­ ment purification o f a three component feed mixture composed o f α-chymotrypsinogen A , cytochrome C and lysozyme using protamine as the displacer (22). A s seen in the figure, this relatively inexpensive small protein (5 kd) is an efficient displacer for cation-exchange systems. The S M A simulation o f this separation is shown in Figure 8. A s seen in the figure, the experiment is well predicted by the model. The induced salt gradient caused by the displacer results in local salt microenvironments o f 74, 72 and 68 m M for the feed components α-chymotrypsinogen A , cytochrome C and lysozyme, respectively. This is in comparison to the initial carrier salt conditions o f 25 m M . Clearly, predicting the isotachic displacement profile o f this separation using the initial carrier conditions would produce erroneous results. In fact, we have shown that the induced salt gradient caused by the adsorption o f the displacer can result in the separation being transformed from the displacement regime to the elution regime (22).

3.5 80

3.0 +

.i I ί'

fi 2.5 + ο 2 2.0 fi

r ! i

Na

g 1.5 ο υ fi 1.0

60 • 50 .

>>

0.5 .

0.0

β 1 1 α

S

'S

I

70

.

.

.

1

•

•

•

•

1

•

3

4

-S 1 a --η

S 1 se J >* 1

20

,

5

Elution Volume [ml] Figure 7. Effluent profile of protamine displacement. (Reproduced from ref. 22. Copyright 1992 American Chemical Society.)

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

CHROMATOGRAPHY IN BIOTECHNOLOGY

40

S

3.0 ι

¥

2.5 }

_•• 80 •I 70 60 50

α 1.5 f ο

·': 40

Na

-I 30

Downloaded by PENNSYLVANIA STATE UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch003

-': 20 > 0.5 +

1

•I 10 2

3

4

• • •• ι 5

0 6

7

Elution Volume [ml] Figure 8. Simulated profile of protamine displacement. (Reproduced with permission from ref. 22.Copyright 1992 American Chemical Society.)

Conclusions In order to accurately describe the non-linear adsorption behavior o f proteins in ionexchange systems, steric effects as well as salt effects must be taken into account. The theoretical and experimental results presented in this chapter demonstrate the power o f the S M A formalism to predict protein adsorption and complex behavior in ionexchange protein displacement separations. Furthermore, the development o f a rapid method for obtaining ideal isotachic displacement profiles with this formalism facilitates methods development and optimization o f ion-exchange protein displace­ ment separations. Notation C, Qj C* C° (Q)* AiCj), k' K Qi Q* Q Q Q™™ x

u

l

x

mobile phase concentration solute / (per volume mobile phase) [mM] feed concentration solute /' [mM] isotachic mobile phase concentration solute i [mM] carrier salt concentration [mM] salt concentration in an isotachic zone containing solute /' [mM] change in salt concentration due to a front o f solute /' [mM] capacity factor [dimensionless] equilibrium constant [dimensionless] stationary phase concentration solute /' (per volume stationary phase) [mM] isotachic stationary phase concentration solute / [mM] stationary phase concentration, sterically non-hindered salt [mM] stationary phase concentration, sterically hindered salt [mM] stationary phase saturation capacity solute i [mM]

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

3.

CRAMER & BROOKS u w V Vfj V* V Xf y d

G

B

Q

{

Ion-Exchange Displacement Chromatography

displacer velocity chromatographic velocity breakthrough volume feed volume solute / volume (width) o f the isotachic zone containing solute / breakthrough volume o f an unretained solute mobile phase mole fraction stationary phase mole fraction solute /

41

[cm/s] [cm/s] [ml] [ml] [ml] [ml] [dimensionless] [dimensionless]

Downloaded by PENNSYLVANIA STATE UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch003

Greek letters α

separation factor

[dimensionless]

β A Λ ν σ

column phase ratio (ΐ-ε^/ε, (where ε, is the total porosity) slope o f displacer operating line stationary phase capacity (mono-valent salt counter-ions) characteristic charge steric factor

[dimensionless] [dimensionless] [mM] [dimensionless] [dimensionless]

Acknowledgments This research was funded by Millipore Corporation and a Presidential Y o u n g Investigator A w a r d to S . M . Cramer from the National Science Foundation. The gifts o f supplies and equipment from Millipore Corporation (Waters Chromatography Division, Millipore, Milford, M A ) and Pharmacia L K B Biotechnology (Uppsala, Sweden) are also gratefully acknowledged. Literature Cited

1 Regnier, F.E. In High-Performance Liquid Chromatography of Proteins and Peptides; Hearn, M.T.W.; Regnier, F.E.; Wehr, C.T., Eds.; Academic Press: New York, NY, 1982. 2 Regnier, F.E. Science, 1983, 222, 245. 3 Kopaciewicz, W.; Rounds, M.A.; Fausnaugh, J.; Regnier, F.E. J. Chromatogr., 1983, 266, 3. 4 Cramer, S.M.; Subramanian, G. In New Directions in Adsorption Technology; Keller, G.; Yang. R., Eds.; Butterworth: Stoneham, MA, 1989, 187. 5 Horváth, Cs.; Nahum, A.; Frenz, J. J. Chromatogr., 1981, 218, 365. 6 Horváth, Cs.; Frenz, J.; Rassi, Z.E. J. Chromatogr., 1983, 255, 273. 7 Horváth, Cs. In The Science of Chromatography; Bruner, F., Ed.; Journal of Chromatography Library; Elsevier: Amsterdam, 1985, Vol. 32; p179. 8 Cramer, S.M.; Horváth, Cs. Prep. Chromatogr., 1986, 215, 295. 9 Guichon, G.; Katti, A. Chromatographia, 1987, 24, 165. 10 Subramanian, G.; Phillips, M.W.; Cramer, S.M. J. Chromatogr., 1988, 439, 341. 11 Phillips, M.W.; Subramanian, G.; Cramer, S.M. J. Chromatogr., 1988, 454, 1. 12 Cramer, S.M.; Subramanian, G. Sep. Purif. Methods, 1990, 19, 31. 13 Peterson, E. Anal. Biochem., 1978, 90, 767. 14 Peterson, E.; Torres, A. Anal. Biochem., 1983, 130, 271.

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

CHROMATOGRAPHY IN BIOTECHNOLOGY

Downloaded by PENNSYLVANIA STATE UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch003

42

15 Liao, A.W.; Rassi, Z.E.; LeMaster, D.M.; Horváth, Cs. Chromatographia, 1987, 24, 881. 16 Subramanian, G.; Cramer, S.M. Biotechnol. Prog., 1989, 5, 92. 17 Jen, S.C.; Pinto, N. J. Chromatogr., 1990, 519, 87. 18 Jen, S.C.; Pinto N. J. Chromatogr. Sci., 1991, 29, 478. 19 Ghose, S.; Matiasson, B. J. Chromatogr., 1991, 547, 145. 20 Gadam, S.; Jayaraman, G.; Cramer, S.M. J. Chromatogr., (in press). 21 Jayaraman, G.; Gadam, S.; Cramer, S.M. J. Chromatogr., (in press). 22 Gerstner, J.; Cramer, S.M. Biotechnol. Prog., (in press). 23 Gerstner, J.; Cramer, S.M. (submitted to BioPharm). 24 Rounds, M.A.; Reignier, F.E. J. Chromatogr., 1984, 283, 37. 25 Drager, R.R.; Reignier, F.E. J. Chromatogr., 1986, 359, 147. 26 Cysewski, P.; Jaulmes, A.; Lemque, R.; Sebille, B.; Liladal-Madjar, C.; Jilge, G. J. Chromatogr., 1991, 548, 61. 27 Helfferich, F.G.; Klein, G. Multicomponent Chromatography: Theory of Interference; Marcel Dekker: New York, NY, 1970. 28 Helfferich, F.; James, D.B. J. Chromatogr., 1970, 46, 1. 29 Aris, R.; Amundson, N.R. Philos. Trans. R. Soc. London Ser. A, 1970, 267, 419. 30 Rhee, H.-K.; Amundson, N.R. AIChE J., 1982, 28, 423. 31 Helfferich, F.G. AIChE Symp. Ser., 1984, 233, 1. 32 Frenz, J.H.; Horváth, Cs. AIChE J., 1985, 31, 400. 33 Helfferich, F.G. J. Chromatogr., 1986, 373, 45. 34 Geldart, R.W.; Yu, Q.; Wankat, P.; Wang, N.-H. Sepn. Sci. & Tech., 1986, 21, 873 35 Katti, A.M.; Guichon, G.A. J. Chromatogr., 1988, 449, 25. 36 Yu, Q.; Wang, N.-H. Comput. Chem. Eng., 1989, 13, 915. 37 Golshan-Shirazi, S.; Guichon, G. Chromatographia, 1990, 30, 613. 38 Khu, J.; Katti, A.; Guiochone, G. Anal. Chem., 1991, 63, 2183. 39 Yu, Q.; Do, D.D. J. Chromatogr., 1991, 538, 285. 40 Golshan-Shirazi, S.; El Fallah, M.Z.; Guichon, G. J. Chromatogr., 1991, 541, 195. 41 Katti, A.M.; Dose, E.V.; Guichon, G. J. Chromatogr., 1991, 540, 1. 42 Levan, M.D.; Vermeulen, T. J. Phys. Chem., 1981, 85, 3247. 43 Velayudhan, A. Studies in non-Linear Chromatography; Thesis; Yale University: New Haven, CT, 1990. 44 Velayudhan, Α.; Horváth, Cs. J. Chromatogr., 1988, 443, 13. 45 Brooks, C.A.; Cramer, S.M. AIChE J., (in press). 46 Zemaitis Jr., J.F.; Clark, D.M.; Rafal, M.; Scrivner, N.C. Handbook of Aqueous Electrolite Thermodynamics; AIChE: New York, NY, 1986. 47 Geng, X.; Regnier, F.E. J. Chromatogr., 1984, 296, 15. 48 Velayudhan, Α.; Horvath, Cs. J. Chromatogr., 1986, 367, 160 49 Regnier, F.E.; Mazsaroff, I. Biotech. Progress., 1987, 3, 22. 50 Rahman, Α.; Hoffman, N. J. Chrom. Science, 1990, 28, 157. 51 Whitley, R.D.; Vancott, K.E.; Berninger, J.A.; Wang, N.H.L. AIChE J., 1991, 37, 555. RECEIVED October 13, 1992

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Chapter 4

Process Chromatography in Production of Recombinant Products Walter F. Prouty

Downloaded by COLUMBIA UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch004

Lilly Research Laboratories, Eli Lilly and Company, Lilly Corporate Center, Indianapolis, IN 46285

Process chromatography will continue to play a key role in industrial scale operations involving commercialization of recombinant DNA products. The strategy used in design of a commercial-scale process is dependent on whether the product is a well defined chemical entity, or represents a family of products. For example, human insulin processing involves high resolution techniques because anayltical tools can be used to define insulin as a unique chemical entity. However, protein products with considerable post-translational modifications, especially glycoproteins, are heterogeneous mixtures, just like their naturally occurring counterparts. Processing of these proteins is most efficient when group specific techniques, such as affinity methods, are used. In addition, arguments are given for why process optimization work, especially for a mature process, should focus on the latter steps of the process. Recombinant technology has made available a diverse array o f therapeutic protein products. The quality and safety o f these products depend upon a robust and reproducible purification process. Chromatographic methods provide manufacturers with the tools to make highly purified protein products reproducibly from lot to lot. Reproducibility is critical for success i n the marketplace and i n getting a product and its process approved by regulatory agencies. This article describes some o f the advantages o f using chromatography i n purification. The author also discusses his preferences for resin selection, running conditions, process control, and a strategy for directing process optimization efforts to the latter stages o f a process. Process Design Depends Heavily O n Product Source A therapeutic protein product that is characterized as a single chemical species must be treated differently than one that is a mixture o f closely related, but heterogeneous molecules. For example, recombinant proteins produced as intracellular products i n prokaryotes, such as E . coli. can be purified to a single chemical species. (That may not be an easy task as several chemical variants are produced both during synthesis by the cell and during manipulation o f the protein during purification (7,2). R e m o v a l of the initiator amino acid, methionine, or N - f o r m y l methionine, either directly or by cleavage of a leader sequence, has been described elsewhere (3,4)).

0097-6156/93/0529-0043$06.00/0 © 1993 American Chemical Society

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Downloaded by COLUMBIA UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch004

44

CHROMATOGRAPHY IN BIOTECHNOLOGY

In contrast, secreted recombinant proteins made i n mammalian cells often are subject to post-translational modifications, such as glycosylation, limited proteolysis for processing, or modification of amino acids (5,6). These modifications are not carried out uniformly on all newly synthesized proteins. Overproduction o f the recombinant protein can result i n synthetic rates that are too high to allow complete processing by the cellular machinery (6). In addition, the carbohydrate chains o f glycoproteins display considerable heterogeneity, even though the products appear to have similar or identical biological activities. The two types o f products produced by recombinant technology are also treated differently by regulatory agencies. W h e n the therapeutic agent is a single chemical entity, sensitive analytical tests can result i n complete characterization o f the product. The heterogeneous secreted product, however, is difficult, at best, to characterize completely. Thus, regulatory authorities argue the "process defines the product" (7). That is, product reproducibility and quality are assured by running a fixed process that has very tight controls on it. [ A n y change i n the process, then, raises the possibility that the product is not exactly the same as that produced before the change.] Such restrictions have important implications i n process design as they affect both the choice o f chromatographic resin and the vendor o f the resin, as w i l l be discussed later. Protein products that are single chemical entities require the use of highresolution techniques i n the purification process. F o r example, human insulin (hi) is produced from human proinsulin (hPI) by isolation o f the latter from insoluble granule fractions o f E . c o l i (8,9). Proinsulin is cleaved by enzymes to produce insulin and the product is subjected to extensive purification (10). Proinsulin can be purified to a single chemical form and, as such, can be separated from chemical variants such as those containing desamidated asparagine groups (2). Figure 1 shows an H P L C profile o f partially purified hPI amplified to illustrate the contaminants i n the process stream. A t low load (the smaller trace i n Figure 1), the product appears to be virtually pure. It is only at much higher loads that the closely eluting and chemically related contaminants are seen. Separation o f those closely related species can require high resolution techniques. Chromatography can provide that high resolution step and has been a vital tool i n bringing human insulin to the marketplace (2,10,12). It offers an excellent choice o f separation techniques that take advantage o f subtle differences i n proteins based on charge (ion exchange), size, and hydrophobicity (reversed phase and hydrophobic interaction). Since chromatographic purification involves interaction o f a part o f the protein molecule with the resin surface and usually a limited number o f points on the protein are involved i n that interaction, chromatography can be exquisitely sensitive to changes altering any o f those sites (12). The d i l e m m a for the researcher is that the selectivity o f the many different resins on the market is not readily predictable. C h o i c e o f a resin to use i n a separation must often i n v o l v e trial and error experimentation with the actual process streams. The heterogeneity inherent i n secreted protein products is illustrated i n Figure 2. A wide array of contaminant proteins with varied sizes and isoelectric points are shown i n the pre-column sample (Figure 2 A ) . In this experiment, the post-column product purity (Figure 2 B ) exceeds 90% when assayed with a specific antibody, or by biological activity, even though the stained 2 D gel still shows many spots. A l l o f those spots are product related as seen i n Western blot analysis (not shown). Generally, they are the result of heterogeneity of the carbohydrate portion o f the glycoprotein, i.e., there is a large reduction i n number o f spots on deglycosylation (13). Structural heterogeneity of the purified product implies use of high resolution techniques in the purification process is not practical. Instead, group specific methods offer a better choice. Figures 3 and 4 illustrate these two alternatives i n using chromatographic methods to purify a heterogeneous biological product: more highly resolving versus group specific procedures. Fractions from ion-exchange chromatography of a secreted product (about 90% pure by bio-assay)

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Downloaded by COLUMBIA UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch004

PROUTY

Process Chromatography and Recombinant Products

100

Elution time, sec. Figure 1. Human Proinsulin is chromatographed on a Zorbax C-8 reversed phase column (77). Buffer A was 0.1 M ammonium phosphate, p H 6.8 and Buffer Β was acetonitrile. F l o w was at 1.0 ml/rnin and the column run at 4 5 ° C . Product was detected by U V absorbance at 214 nm. The smaller tracing was achieved by a ten-fold dilution of the sample into Buffer A .

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

CHROMATOGRAPHY IN BIOTECHNOLOGY

Downloaded by COLUMBIA UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch004

46

Figure 2. 2-D polyacrylamide gel electrophoresis analysis of a recombinant product secreted from a mammalian cell (26). The pre-column sample ( A ) was prepared by filtering the serum-free broth (0.45-lm filter) to effect cell removal and 1 0 X concentration o f the m i x on a Centricon -10 microconcentrator ( A m i c o n , Beverley, M a . ) . The post column sample (B) was run as is from the column effluent.

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Downloaded by COLUMBIA UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch004

PROUTY

Process Chromatography and Recombinant Products

Time of Elution Figure 3. Ion-exchange chromatography o f a secreted mammalian c e l l product. A sample o f modified tissue plasminogen activator (mtPA) (27) was applied to a cation exchange resin ( M o n o S, Pharmacia, Upsala, Sweden) and eluted with a shallow gradient o f N a C l (0.1 to 0.157M) i n 35 m M acetic acid, p H 5, 4.9 M urea and 30% (v/v) Acetonitrile. Fractions were collected, diluted with l o w salt buffer, and those containing U V absorbing material (fractions 41 through 50) were injected onto another M o n o S column under conditions identical to the first column. The tracings from the re-injections are shown. The sample identified as " P C S " is the pre-column sample, or starting material. The various fractions elute from the chromatograph as distinct species implying structural or conformational variation. A l l fractions from 41 to 5 0 had enzymatic activity and approximately the same specific activity when measured as amidolytic activity against a small molecule substrate (S2288, K a b i Diagnostica A B , Taljegardsgatan, Sweden ) (28).

American Chemical Society library 1155 16th St.. N.W. Washington. DC. 20036 In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Downloaded by COLUMBIA UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch004

CHROMATOGRAPHY IN BIOTECHNOLOGY

Elution Volume, CV Figure 4. Affinity chromatography o f m t P A over a column o f L y s i n e Sepharose (Pharmacia). The charged c o l u m n was washed with Tris ( 5 0 m M , p H 8.0), sodium chloride (150 m M ) , Tween 80 (0.01% w/v) buffer (one column volume), then the product eluted by running a gradient to 500 m M arginine in the tris, salt, detergent wash buffer. Protein was measured as U V absorbance at 275 nm and activity measured as amidolytic units using the S2288 substrate as i n Figure 3.

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

4.

PROUTY

Process Chromatography and Recombinant Products

49

Downloaded by COLUMBIA UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch004

showed the preparation could be separated into many discrete products (Figure 3), as shown by the different elution patterns on re-injection on the analytical H P L C . Since these various fractions all contained the same specific activity i n a bioassay, such high resolution fractionation results i n large yield losses i f any o f the products is excluded from the mainstream pool. The method is valuable only i f separation is desirable. That is, it makes no sense to separate components, i f those components are going to be put back together. A better approach uses affinity techniques, w h i c h bind the protein product on the basis o f ligand interaction and w i l l be insensitive to many o f the chemical variants on the protein that are not associated with the ligand binding site. W h e n the ligand chosen for purification is related to the biological activity o f the protein, then a class o f proteins with similar biological function w i l l be isolated. Thus, the protein mixture behaves as a single entity, yielding a single peak o f activity (Figure 4) when using group specific separation techniques. Differentiating Early From Late Stages In A Purification Process Processes for isolation of recombinant proteins can be conveniently broken into an early and a late phase. In the early phase, the goal o f the process is twofold: volume reduction, since the product is usually present i n the culture broth or cell lysate i n l o w concentration, and clarification, since cell debris is a likely contaminant of product fresh from the fermentor. The early steps o f a process are targeted for high volume throughput and product capture (Table 1). H i g h resolution steps are very costly and are not a wise investment at these early processing steps. Late stages of a process may rely on high-resolution methods, i f separation o f chemically similar species is desired, especially with products o f bacterial fermentation, such as human insulin or human growth hormone. The resolution needed, however, depends upon the source o f the protein product. A s stated above, heterogeneous protein products need not depend upon high-resolution techniques and their extra expense would add undesirable costs to the purification process (Table 1). In chromatography, high resolution can be achieved by use o f small beads (14,15). The price for using small beads is that the process must be designed to handle the higher back-pressure generated i n the chromatographic columns. The small-bead columns are likely to exceed 15 psi on the production scale columns and thus require special equipment that results i n considerable added expense. A b o v e a minimal size and operating pressure (15 psi), buffer tanks must meet American Society o f Mechanical Engineers ( A S M E ) codes. Alternatively, the columns can be packed i n a short, squat geometry where the column height is much less than the diameter, or i n radial flow columns (16). However, such columns entail other challenges, such as well-designed sample distribution systems at the inlet and outlet (17). The high costs associated with high resolution may be justified by the potential for increased resolution. In a one-step process, there may be no choice, but usually processes contain multiple steps that take advantage o f different properties of the protein such as size, charge, and hydrophobicity. This multiplicity o f steps may obviate the need for high resolution techniques or, at least delay their use to the end o f the process when clean process streams would result i n a longer lifetime i n reuse of the column. Properties O f Chromatographic Resins Chromatographic resins used i n large-scale industrial processes should consist of rigid particles and a uniform size distribution. That is, the resin must neither deform at the pressures generated during the run nor contain very tiny particles that could cause excessive back- pressure (18). These properties are important i n getting a uniform and w e l l - packed c o l u m n that can be run at reasonable flow rates. Similarly, the resin should not go through cycles o f shrinking and swelling during

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Downloaded by COLUMBIA UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch004

50

CHROMATOGRAPHY IN BIOTECHNOLOGY

column runs or regeneration. Such changes would compromise the integrity o f the packed bed and it w o u l d be hard to guarantee reproducibility from run to run. Geometric constraints would prevent the shrunken material from swelling back to its original position. Another important feature o f virtually a l l resins is their ability to participate in secondary interactions. F o r example, size-exclusion resins can show weak i o n exchange properties (79), as do silica-based reversed-phase resins (20). Different ion-exchange resins from different manufacturers show different selectivity with product streams, even though the functionality may be the same (79). M i n i m a l nonspecific binding is often desirable, but on occasion, binding by multiple mechanisms can be helpful i f it provides more purification across the step (10,18,21). A g a i n , the choice o f the best resin for a product must be based upon lab experiments using the actual process streams. Economic factors must also be considered i n resin choice. Expensive resins can be made economical i f manufacturing processes involve multiple uses o f the resin i n a packed column. Thus, the resin must be hardy enough to withstand rigorous clean-in-place (CIP) procedures, such as acid and base treatment. Base treatment, i n particular, solubilizes proteins that may have fouled the resin, and it reduces bioburden (22). The choice o f manufacturer can be as important as the actual type o f resin used i n a process. The resin manufacturer must be reliable enough to supply sufficient quantities o f resin on a strict time schedule. In addition, the resin must be uniform from batch to batch to provide assurance that the purification process produces a uniform protein product. Finally, resins must not leach monomers or breakdown products into a process stream. A resin manufacturer should assist the resin user by supplying the information how the resin was made, what chemical agents could leach under certain elution conditions, and how to assay for those potential contaminants. T o maintain confidentiality, a user could refer to a D r u g Master F i l e ( D M F ) number that the resin manufacturer has submitted to the appropriate regulatory authorities. The choice of a chromatographic resin determines selectivity, but one can also manipulate the elution mode to optimize further the value of a given resin. The best elution mode depends upon where the step is i n the process (Table 2). F o r secreted products from mammalian cells, resolution is not usually an issue, so adsorption chromatography need involve only on-off mechanisms. Step-elution procedures are easier to design and subject to less equipment breakdown than the more difficult gradient mode of elution. However, manipulation o f the gradient shape during elution can provide a more pure product at better yields. A shallow gradient w i l l stretch out the peaks and provide an opportunity to make a mainstream pool that contains more o f the desired product as shown in Figure 5. Displacement chromatography is an alternative elution method (25), but to be very effective the user must carefully define conditions. Since the product i n displacement chromatography tends to elute as "square waves", contaminants should not coelute with the product. Removal o f the contaminants could result i n unacceptably low yields from the step. For example, contaminants coelute with insulin i n the displacement run shown i n the chromatogram i n Figure 6. (Note: the scale used to measure the contaminants is different from that used for the insulin and the contaminant levels are very l o w compared to the i n s u l i n , so the chromatography works quite well.) The backside o f the mainstream pool i n Figure 6 would be at fraction 39. That means some fractions (e. g. fractions #40-42) with high concentrations of insulin would not be included i n the mainstream. This is i n contrast to the gradient-mode separation shown i n Figure 5 where the contaminant elutes during a time of decreasing product concentration. One way to avoid this problem with the displacement method is to continue to search for a better displacing agent. Unfortunately, this must be done by trial and error (24).

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

4.

PROUTY

51

Process Chromatography and Recombinant Products

Table 1 Resolution Desired in Process Steps Process Stage Early

Downloaded by COLUMBIA UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch004

Late

Resolution Mammalian

Effect

Bacterial

Capture/ vol. red'n Purify

Low

Low

High

Group specific

Table 2 Elution Method Stage

Bacterial

Mammalian

Advantage

Early

Step/grad.

Step/grad.

Late

Gradient

Step/grad.

Step is easy on/off modegoal of step, e.g. cone? Desired resol'n?

0.6

Elution Volume Figure 5. Ion Exchange o f an E . coli - derived recombinant product. The protein mixture was charged to a cation-exchange resin (Fast F l o w S, Pharmacia, Upsala, Sweden) i n 7 M urea buffered with 50 m M acetic acid, p H 3.5. The product was eluted with a gradient o f N a C l as shown. Protein was monitored by U V absorbance at 280 nm while product and contaminant concentrations were monitored by an analytical H P L C assay. The fractions pooled to make product mainstream (ms pool) are indicated by the vertical lines and represent those fractions with highest specific activity.

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

CHROMATOGRAPHY IN BIOTECHNOLOGY

Downloaded by COLUMBIA UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch004

8

Fraction # Figure 6. Displacement chromatography o f human insulin (hi) (1829). Human insulin was charged to a C 8 reversed phase resin (Zorbax, DuPont, W i l m i n g t o n , De.) at 45.3 g o f product per liter o f packed resin. The starting B H I preparation (97.0% pure by analytical H P L C ) was dissolved i n 5 0 m M phosphate, p H 2.7 with 2.5% v/v methanol (Buffer A ) . The column was washed with five column volumes of Buffer A and the bound products were displaced with 10 m M cetyl trimethyl ammonium bromide i n Buffer A . Fractions were collected and analyzed for insulin (expressed as mg/ml) and percentage contaminant level (expressed as a percent o f the insulin content) by analytical H P L C . The recovery o f insulin was quantitative i n this experiment. Note the contaminants elute toward the back o f the B H I peak, but there is some overlap i n the points where contaminant concentration begins to increase and B H I is still at its peak concentration. The pre-column material ( P C S ) contained 96.8% B H I with 2.5% frontside and 0.7% backside contaminants, as measured by H P L C . The mainstream p o o l ( M S ) off o f the column contained 98.9% B H I , 0.3% frontside and 0.8% backside contaminants i n the analytical H P L C (Reprinted from Ref. 18. Copyright 1991 Marcel Dekker Inc.)

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

4.

PROUTY

Process Chromatography and Recombinant Products

53

Downloaded by COLUMBIA UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch004

A second way to improve yields is to recycle the sidestream. Recycling poses a potential regulatory problem i n that it is more difficult to assure that the recycled product is o f exactly the same quality as the mainstream product from the first pass. In addition, the recycle step probably requires a dedicated column for processing. The cost o f the recycled product, which includes the extra processing, must be weighed against the cost of merely running more material through from the beginning o f the process. The most economical approach is not always intuitively obvious. Often, it is more cost effective, concerning utilities, space, and the time it takes for both step operations and product analysis not to recover sidestreams, but to use those resources to process more single-pass mainstream material from the beginning o f the process. Where T o Spend Development Time In Process Improvement Some o f the factors that justify process improvement are: i m p r o v i n g product quality, increasing process capacity, and decreasing costs. One o f the difficulties of assessing process improvement with respect to product quality is the availability o f a suitable assay that can measure the small, incremental effects often seen with the process changes. Use o f statistical methods has become a necessity for defining when a process is or is not i n control. Statistical methods can also be used to define when variations require action or are just a part o f the random variation involved i n the evaluation. Process control charts (25) are used to track the variation i n a given process parameter, such as product purity, through a purification step (Figure 7). Variations are introduced at a process step i n many ways. F o r example, mechanical valves have tolerances that mean they may switch at slightly different times after a signal is sent to them, pumps have slight variations and the variations increase with wear, and assays have inherent variability due to both mechanical and procedural manipulations. A l l steps are also subject, o f course, to human error. These human errors involve minor differences in the way steps are carried out. Process control charts allow one to understand the degree o f variation introduced by the sum o f these individual variations without necessarily understanding the detail o f each. In the example shown (Figure 7), the purity of the product from the step is 96.31% (in runs 1 through 80) with the upper and lower control limits ( U C L and L C L , respectively) representing three standard deviations o f the mean. W h e n an individual point falls outside the U C L or L C L , a nonrandom variation is likely to have occurred i n the process and the product o f that process step must be suspect. It also warrants immediate investigation as to what caused the variation. Another use of the chart is to signal when a process is going out o f control, i.e., when a trend is developing. M a n y different criteria are used. A common one is to assume that seven consecutive points on one side o f the mean constitute a trend, and warrant careful investigation of the process step. The chart may also be use to test the effectiveness o f a process change. A review o f the product purity data subsequent to a process change (Figure 7, at run #80) shows that the process change was favorable both in purity improvement and in reduced variation i n the process step. More often, in such process changes, the new values are not changed so dramatically. Then, a large number o f runs must be made to assess whether the change resulted i n a significant improvement. When trying to optimize a process on the basis o f throughput, or capacity, a convenient method is to chart how much product the process could deliver, given defined step yields for each step i n the process (Figure 8). In the process shown, step 7 shows the lowest capacity. A n y slipping at that step means that the process w o u l d deliver less than the targeted amount o f product. L i f e saving therapeutics, such as human insulin, cannot be run below target, i f it means the market cannot be supplied. One approach to improve product throughput is to improve the process at step 7 by either increasing the yield or installing an additional unit to double step

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Downloaded by COLUMBIA UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch004

CHROMATOGRAPHY IN BIOTECHNOLOGY

- SD = 92.95

Number of Run Figure 7. Statistical process control chart for a manufacturing step (30,31). Purity across the manufacturing step is measured by analytical H P L C analysis of the product. The upper control limit ( U C L ) and lower control limits ( L C L ) represent three standard deviations from the mean. A t run #80, a process change was made; the shape of the gradient elution from an ion-exchange chromatography step was changed. A new mean, U C L , and L C L were calculated for the runs subsequent to the process change to assess whether the change had a statistically significant impact on product purity.

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

PROUTY

Process Chromatography and Recombinant Products

2.0 % capacity 160-

Downloaded by COLUMBIA UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch004

H.6

I"

2

φ

120H-2

CO

u o

Q.

Φ

υ ο

CO

80A •0.8

Φ

I ο CO

40H

hO.4 Step yield — I — 10

1—

•0.0 20

Step# Figure 8. Process capacity at each step o f the process. The graph shows the amount o f product the process can deliver i f a l l steps are running at the yields indicated. The capacity of a process step is dependent upon the number o f operating units at the given step, as w e l l as the size of those units. In chromatography, the throughput can be increased i f more material can be put on the column during each cycle (charge capacity), i f the elution rate can be increased (increase i n flow rate, or change i n steepness o f elution gradient), or i f the size o f the columns is increased. Although faster throughput is useful for economic reasons, it can also have a positive effect on product purity, especially i f there is any product instability at that step. These factors must be carefully assessed i n controlled laboratory studies. Frequently, it is necessary to run nearly full-scale model experiments to study the changes so as to better simulate production conditions.

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

CHROMATOGRAPHY IN BIOTECHNOLOGY

200 Downloaded by COLUMBIA UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch004

• % capacity

Yield Increase

180H

• Improve 10 (82-90) ϋ Improve 16 (90-95)

Φ C/>

'5 160H (0 Q. (0 ϋ Φ σ> 140-

120-

100

1

2

3

4

5

6

7

8

1II

9 10 11 12 13 14 15 16

Step# Figure 9. Effect o f yield increase at steps 10 or 16 o f the process shown i n Figure 8. If a yield increase is effected i n step 16 (an example is shown for an increase i n yield at step 16 from 90 to 95%), then the step is capable o f processing 5.6% more material. In addition, the prior step throughput capability is increased by the same factor. In reality, it means less material has to be processed through the previous step to reach the same target goal. Thus, fewer reagents, chemicals, solvents, etc, are needed for all previous steps, i f the target is not changed.

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Downloaded by COLUMBIA UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch004

4.

PROUTY

Process Chromatography and Recombinant

Products

57

capacity. F o r example, i f step 7 is a chromatographic column, a second column could be run, or, alternatively, a larger column installed. Another way to improve the process throughput, however, is to make process improvements at late steps. This results i n a positive effect on a l l earlier steps. The greater yield o f product from step 16 means step 15 and a l l prior steps need not deliver as much intermediate to get the same throughput. A y i e l d improvement means less intermediate is needed to reach target at all upstream steps (Figure 9), not just the one i n w h i c h the original improvement was made. Similarly, an improvement late i n the process can have a greater impact on costs than earlier process improvements. A l l previous steps can be run at reduced rates and still maintain the targeted capacity, resulting i n less consumption o f raw materials, labor, or utilities. The cost o f the late stage intermediate product is much higher than earlier intermediates due to the costs o f the processing steps and labor to get it there. Thus, it is more cost effective to direct cost saving efforts at increasing yields at the end of the process. One final factor that is increasingly important i n process design is the environmental impact of wastes or noxious chemicals. Improving yields or altering conditions to minimize wastes at the step generating those wastes can provide a solution to the problem. A s with throughput, improving yields at the end o f the process minimizes the amount o f waste stream production from a l l previous steps. Both approaches are usually taken. A n argument has been made for the importance o f tight control and optimization of the latter stages of a process. Regardless of the source of the protein therapeutic, the late process steps are likely to involve chromatography. These chromatographic steps can depend on high resolution, such as when the product is chemically well defined, or on batch methods, such as affinity techniques, when the product is a mixture o f closely related products. Large columns (greater than 2000 L ) can be run routinely, reproducibly and economically i n a manufacturing setting. In addition, it is important to achieve high yields at these steps, both to minimize cost and increase throughput. Finally, there is no method available that can better control quality of a therapeutic protein product while offering a wide choice of selectivity and operating variables than chromatography. Acknowledgments: T h e author thanks his many colleagues i n the B i o s y n t h e t i c D e v e l o p m e n t Department at E l i L i l l y for their many fruitful and stimulating discussions and suggestions. In particular, I thank G . K e l l y for the Cation Exchange data generated from the secreted mammalian cell product, P . Atkinson and T.Vieira-Brisson for development o f the affinity method used in Figure 4, and R . W i l k e n s for many informative discussions regarding the use o f Statistical Process Control ( S P C ) i n control of the H u m u l i n production process. Literature Cited

(1) Stadtman, E. R. Biochemistry 1990, 29(27), 6323. (2) Prouty, W. F., BioTech USA, The sixth Annual Industry conference and Exhibition. BIOTECH Proceedings, San Francisco, Calif. Oct. 2-4, 1989 (Nature Publishing Co., New York, N.Y.) p. 224. (3) Marston, F.A.O. Biochem J. 1986, 240, 1. (4) Nagai, K.; Thogerson, H. C. Methods Enzymol. 1987,153, 461. (5) Hokke, C. H.; Berwerff, Α. Α.; van Dedem, G.W.K.; van Oostrum, J., Karerling, J.P.; Vliegenthart, J.F.G. FEBS Lett. 1990, 275(1, 2), 9. (6) Yan, S - C.; Grinnell, W.; Wold, F . Trends in Biochemical Sciences (TIBS)[ 1989,14, 264. (7) Hammond, P. M.; Atkinson, T.; Sherwood, R. F.; Scawen, M. D. BioPhann 1991, 4(5), 30.

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Downloaded by COLUMBIA UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch004

58

CHROMATOGRAPHY IN BIOTECHNOLOGY

(8) Frank, Β. H.; Chance, R. E. InTherapeutic agents produced by genetic engineering Editors, Joyeaus, Α., Leygue, G., Moore, M., Roncacci, R., Schmelk, P.H. "Quo Vadis"? Symposium, Sanofi Group, May 29-30, 1985. Toulouse-Laberge, France, p. 137. (9) Frank, Β. H.; Chance, R. E. Muench. Med Wochen~chr. 1983, 125, (suppl. 1), 14. (10) Kroeff, E. P.; Owens, R.A.; Campbell, E. L.; Johnson, R. D.; Marks, Η. I. J. Chromatogr. 1989, 461, 45. (11) DiMarchi, R. D.; Long, Η. B.; Kroeff, E. P.; Chance, R. E. In High Performance liquid Chromatoeraphy in Biotechnoloegy Hancock, W. S., Ed.; John Wiley and Sons: New York, 1990; pp. 181-190. (12) Rounds, Μ. Α.; Kopaciewicz, W.; Re~er, F.E. J. Chromatogr. 1986, 362, 187. (13) Yan, S-C.B.; Razzano, P.; Chao, B.; Walls, J. D.; Berg, D.T.; McClure, D. B.; Crinnell, B. W. Bio/Technology 1990, 8(7), 655. (14) Johnson, E.; Stevenson, B. Basic Liquld Chromatography; Varian Associates: Palo Alto, Calif., 1978; pp. 15-39. (15) Rahn, P.; Joyce, W.; Schratter, P. Am. Biotechnol Lab. 1986, 4(7), 34. (16) Saxena, V.; Subramanian, K.; Saxena, S.; Dunn, M. BioPharm 1989, 2(3), 46. (17) Levine, H. L. Presented at the Conference on Frontiers in Bioprocessing, Boulder, Co., June 1987. (18) Prouty, W. F. Production-Scale Purification Processes in Drug Biotechnology Regulation: Scientific Basis and Practices: Chiu, Y-Y.H.; Gueriguian, J. L., Eds.; MlarceI Dekker: New York, 1991; pp. 221-62. (19) Regnier, R. E. J. Chromatogr. 1987, 418, 115. (20) Barford, R. A. In High Performance Liquid Chrornatoeraphy fn Biotechnology; Hancock, W. S., Ed.; John Wiley and Sons: New York, 1990; pp. 63-77. (21) Horvath, C.; El Rassi, Z. Chromatogr. Forum 1986, 1(3), 49. (22) Cleaning in Place in Perspective; Downstream, vol. 9 (a publication of Pharmacia), 1990; pp. 16-20. (23) Su6ramanian, G.; Phillips, M. W.; Cramer, S. M. J. Chromatogr. 1988, 439(2), 341. (24) Ghose, S.; Mattiasson, B. J. Chromatogr. 1990, 547, 14S. (25) Gopal, C.; Mlodozeniec, Α.; Holmes, D.S.. S. Pharm.Technol. 1991, 15(1), 32. (26) Anderson L. Two Dimensional Electrophoresis. Operation of the ISO-DALT System; Large Scale Biology Press: Washington, D.C., 1988; vol. 12, Pp. 57075. (27 Burck, P. J.; Berg, D. H.; Warrick, M. W., J. Biol. Chem. 1990, 265, 5170. (28) Wallen P.; Pohl, G.; Bergsdorf, N.; Ranby, M.; Ny, T.; Jornvall, H. Eur. J. Biochem 1983,123(3), 681. (29) Vigh, C.; Varga-Pouchon, Z.; Szepsi, G.; Gazdag M. J. Chromatogr. 1987, 386, 353. (30) Amsden, R. T.; Butler, H. E.; Amsden D. M. SPC Simpllfled: Practical Steps To Quality:; Kraus International Publications: White Plains, N.Y., 1986. (31) Ishikawa, K. Guide to Quality Control; Asian Productivity Organization, Quality Resources: White Plains, N.Y., 1990. RECEIVED October 9, 1992

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Chapter 5

Preparative Reversed-Phase Sample Displacement Chromatography of Peptides

Downloaded by NORTH CAROLINA STATE UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch005

R. S. Hodges, T. W. L. Burke, A. J. Mendonca, and C. T. Mant Department of Biochemistry and the Medical Research Council of Canada Group in Protein Structure and Function, University of Alberta, Edmonton, Alberta T6G 2H7, Canada

This report describes the early development and current applications of a novel method for highly efficient preparative-scale reversed-phase purification of peptides, termed sample displacement chromatography (SDC). This method is based upon solute-solute displacement by components of a peptide mixture following high sample loading in a 100% aqueous mobile phase, and is characterized by the major separation process taking place in the absence of organic modifier. Highly efficient use of stationary phase capacity, coupled with low flow-rates, enable the preparative purification of high sample loads on analytical columns and instrumentation with high yields of purified product concomitant with low solvent consumption. The potential value of SDC for the pharmaceutical industry was examined by its application to the purification of a five-residue synthetic peptide, representing part of the sequence of luteinizing hormone-releasing hormone. Purification of 100 mg of the peptide on a C stationary phase (300 x 4.6 mm I.D.), carried out by three variations of SDC (a multi-column approach, extended isocratic elution in the absence of organic modifier and isocratic elution in the presence of low levels of organic modifier) produced up to 97.3% yields of purified product in a short run time and with minimal solvent consumption. The simplicity of SDC, coupled with high yields of purified product, minimization of the use of potentially toxic reagents and solvents and the potential for straightforward scale-up, points to a promising future for this technique. 8

T h e growing use o f synthetic peptides i n biochemistry, immunology and i n the pharmaceutical and biotechnology industries has stimulated a concomitant increase i n requirements for rapid and efficient preparative purification methods. Considering that the impurities encountered during peptide synthesis (deletion, terminated o r c h e m i c a l l y m o d i f i e d peptides) are usually closely related structurally to the peptide o f interest, and, hence, often pose difficult purification problems, most preparative separations o f synthetic peptides take advantage o f the excellent r e s o l v i n g power and volatile m o b i l e phases o f reversed-phase chromatography ( R P C ) i n gradient elution mode (1,2). H o w e v e r , the

0097-6156/93/0529-0059$06.00/0 © 1993 American Chemical Society

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Downloaded by NORTH CAROLINA STATE UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch005

60

CHROMATOGRAPHY IN BIOTECHNOLOGY gradient elution mode o f R P C is handicapped by relatively poor utilization of the stationary and mobile phases (3,4). Thus, large-scale gradient elution separations, t y p i c a l l y i n v o l v i n g aqueous trifluoroacetic acid ( T F A ) to T F A - a c e t o n i t r i l e gradients (1,2) o f peptides closely related structurally generally require the employment o f larger column volumes i n order to maintain satisfactory levels o f sample load and product yield. T h i s , i n turn, leads to higher operating costs per unit o f purified product i n terms o f packings, equipment and solvent consumption. A perceived need for simpler, more efficient and reliable methods for preparative separations o f peptides led this laboratory to consider ways of modifying various parameters to overcome the many disadvantages frequently encountered in traditional preparative methods: (1) Column size: a more efficient utilization of stationary phase capacity would allow the employment of smaller column volumes, while maintaining high sample load; (2) Solvent consumption: consumption of expensive H P L C solvents (water, organic modifier) could be cut back considerably i f mobile phase flow-rates could be reduced; (3) Organic modifier: i f the role o f the organic modifier i n the separation process could be reduced, this would also reduce the consumption o f large volumes of frequently toxic and expensive solvents; (4) Fraction analysis: reduction o f the number o f fraction analyses required to locate the product(s) o f interest following preparative purification would save valuable time and solvents; (5) Product concentration: the large c o l u m n volumes and flow-rates typical o f traditional preparative gradient-elution methods frequently result i n dilution o f products i n large fraction volumes. T h u s , any way o f concentrating the same l e v e l o f product i n significantly smaller fraction volumes would greatly simplify subsequent product work-up; (6) Instrumentation: most researchers would probably desire to carry out laboratory-scale preparative separations, i f possible, on existing analytical equipment. W e report here the development o f a novel method for highly efficient preparative-scale reversed-phase purification o f peptides on analytical columns, termed sample displacement chromatography ( S D C ) . T h i s approach is characterized by the major separation process taking place i n the absence of organic modifier. Following a review o f the early development o f S D C (3-7), we present current applications o f the method to the preparative purification o f a pharmaceutically important synthetic peptide, underlying once again the potential o f S D C for the pharmaceutical industry. Principles of Sample Displacement Chromatography (SDC) Since peptides favour an adsorption-desorption method o f interaction with an hydrophobic stationary phase (1,2), under normal analytical load conditions an organic modifier is typically required for their elution from the reversed-phase column. However, when such a column is subjected to high loading of a peptide sample mixture dissolved i n a 100% aqueous mobile phase, there is competition by the sample components for the adsorption sites on the reversed-phase sorbent, resulting i n solute-solute displacement. T h e more hydrophobic peptide components compete more successfully for these sites than less hydrophobic components, which are displaced and quickly eluted from the column. Thus, operation i n sample displacement mode is simply employing the well-established general principles o f displacement chromatography (8-10) without using a displacer (3-7). A crude peptide mixture typically produced by solid-phase peptide synthesis may contain not only the desired product, but also hydrophilic and/or hydrophobic synthetic peptide impurities. Situations where a desired peptide product is the most hydrophilic or most hydrophobic component o f a peptide mixture offer the most convenient model system for illustrating the S D C process.

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Downloaded by NORTH CAROLINA STATE UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch005

5.

HODGES ET AL.

Reversed-Phase Sample Displacement Chromatography

Displacement of Hydrophilic Impurities by Desired Hydrophobic Peptide Component. Figure 1 demonstrates S D C o f a five-decapeptide mixture, where peptide 5 is the desired product (P5) and peptides 1 to 4 (11 to I4, respectively) represent hydrophilic impurities. T h e hydrophobicity o f the peptides increases only slightly between peptide 2 and peptide 5 (sequences shown i n Figure 1) - between 2 and 3 there is a change from an cc-H to a P-CH3 group, between 3 and 4 there is a change from a P-CH3 group to two methyl groups attached to the β - C H group, and between 4 and 5 there is a change from an cc-H to an i s o p r o p y l group attached to the α - c a r b o n - enabling a precise determination o f the potency o f preparative R P C in sample displacement mode. T h e product, P5, represented only 33% o f the total 21 m g o f the five-peptide sample mixture (analytical profile shown i n Figure 1A) loaded onto the Cs column (30 χ 4.6 m m I.D.). T h e protocol for S D C o f the sample mixture involved displacement i n water (0.05% aqueous T F A ) o f the hydrophilic impurities (I1-I4) by the more hydrophobic P5, which should remain bound to the column. Thus, following isocratic elution with 0.05% aqueous T F A at 1 m l / m i n , during w h i c h the major separation took place and all hydrophilic impurities were eluted from the c o l u m n (Figure I B ) , a gradient wash was initiated after 40 m i n ( 1 % acetonitrile/min) to remove the product from the column. F o l l o w i n g fraction analysis, Peak II (Figure I B ) was found to be pure P5 (Figure I D ) and accounted for 9 3 % o f total P5 recovered. Peak I, the unretained fraction (Figure I B ) , contained the bulk o f hydrophilic impurities I i to I4 (Figure 1C). Displacement of D e s i r e d H y d r o p h i l i c Peptide Component by H y d r o p h o b i c Impurities. Figure 2 demonstrates S D C o f a four-decapeptide mixture, where peptide 2 is the desired product (P2) and peptides 3 to 5 (I3 to I5, respectively) represent hydrophobic impurities. The product, P2, represented 50% o f the total 6 m g o f the four-peptide sample mixture (analytical profile shown in Figure 2 A ) loaded onto the Cs column (30 χ 4.6 m m I.D.). F o l l o w i n g the same S D C protocol as described i n Figure 1 ( S D C elution profile shown in Figure 2 B ) , Peak I was found to contain pure P2 (Figure 2C) and accounted for 9 9 % o f recovered P2. T h e bulk of the impurities, I3-I5, remained bound to the column as Peak II (Figure 2D) during elution with 0.05% aqueous T F A , and were only removed by the addition of acetonitrile to the mobile phase. A short (30 m m i n length) c o l u m n was used to illustrate the basic principles of S D C i n order to limit the amount of material required to saturate the hydrophobic stationary phase. T h e results o f Figures 1 and 2 demonstrate that impressive yields o f pure peptide products may be obtained even on such small columns. Development of Multi-Column Approach to S D C of Peptides T h e separations demonstrated i n Figures 1 and 2, where the desired peptide product is the most hydrophobic or h y d r o p h i l i c component, respectively, represent the simplest application o f S D C . T h e next step i n the development o f S D C as a preparative tool required the design o f a strategy to deal with the more difficult, and more realistic, separation problem o f a desired peptide product contaminated with both hydrophilic and hydrophobic impurities. In addition, scale-up to higher sample loads was a logical, and necessary development. The latter requirement also highlighted the problem o f optimizing sample load when applying preparative S D C to peptide separations on single columns. For instance, from Figure 1, i f the sample load was too high, product (P5) may appear in the breakthrough fraction, contaminated with displaced hydrophilic impurities (Il to I4); i f the sample load was too l o w , hydrophilic impurities may remain on the column. F r o m Figure 2, i f the sample load was too high, displaced impurities

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

61

Downloaded by NORTH CAROLINA STATE UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch005

62

CHROMATOGRAPHY IN BIOTECHNOLOGY

ELUTION TIME (min)

Figure 1. Principle o f reversed-phase S D C o f peptides: purification o f peptide "product" (P5) from hydrophilic peptide "impurities" ( I i , I2,13,14)· C o l u m n : Aquapore R P 3 0 0 Cs, 30 χ 4.6 m m I.D., 7-μπι particle size, 3 0 0 - Â pore size (Pierce, Rockford, I L , U . S . A . ) . H P L C instrument: V a r i a n V i s t a Series 5000 l i q u i d chromatograph (Varian, Walnut Creek, C A , U . S . A . ) coupled to an Hewlett-Packard (Avondale, P A , U . S . A . ) H P 1 0 4 0 A detection system, H P 8 5 B computer, HP9121 disc drive, H P 2 2 2 5 A Thinkjet printer and H P 7 4 7 0 plotter. Panel A : analytical separation profile o f peptide mixture; conditions, linear A B gradient (1% B/min) at a flow-rate o f 1 ml/min, where eluent A is 0.05% aqueous T F A and eluent Β is 0.05% T F A i n acetonitrile. Panel B : preparative separation profile o f peptide mixture; conditions, isocratic elution with 100% eluent A for 40 m i n at a flow-rate o f 1 ml/min, f o l l o w e d by linear gradient elution at 1% B / m i n ; sample load, 21 m g , consisting o f 7.0 g of P5 and 3.5 m g of each o f I i , I2,13 and I4 dissolved i n 500 μΐ o f eluent A . Panels C and D demonstrate analytical elution profiles (see Panel A for conditions) o f Peaks I and Π (Panel B ) , respectively. T h e subscripts of I i , I2,13,14 and P5 denote peptides 1-5, respectively. The basic sequence o f the peptide analogues is A r g - G l y - X - X - G l y - L e u - G l y - L e u - G l y L y s where X - X is substituted by G l y - G l y (peptide 2), A l a - G l y (peptide 3), V a l - G l y (peptide 4) or V a l - V a l (peptide 5). T h e peptides all contain an N acetyl group and a C - a m i d e group, except for peptide 1, which i s the same as peptide 3 save for a free α - a m i n o group. Reprinted from reference 3, with permission. a

a

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

5. HODGES ET AL.

Reversed-Phase Sample Displacement Chromatography

Product ·»|

P

2

Downloaded by NORTH CAROLINA STATE UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch005

. JLA. 3.0

Ο CM

ω o α m < 40 60 • I n i t i â t * Gradient Wash

20

0

t

Load Sampla

Paak

Π

20

ELUTION TIME (min) Figure 2. Principle o f reversed-phase S D C o f peptides: purification o f peptide "product" from hydrophobic peptide "impurities" C o l u m n and H P L C instrument: same as Figure 1. Panel A : analytical separation profile o f peptide mixture; conditions, linear A B gradient ( 1 % B/min) at a flow-rate o f 1 ml/min, where eluent A is 0.05% aqueous T F A and eluent Β is 0.05% T F A i n acetonitrile. Panel B : preparative separation profile o f peptide mixture; conditions, isocratic elution with 100% eluent A for 40 m i n at a flow-rate o f 1 ml/min, followed by linear gradient elution at 1% B / m i n ; sample load, 6 m g , consisting o f 3 m g o f P2 and 1 m g o f each of I3,14 and I5 dissolved i n 100 μΐ o f eluent A . Panels C and D demonstrate analytical elution profiles (see Panel A for conditions) o f Peaks I and II (Panel B ) , respectively. The subscripts o f and denote peptides 5, respectively. T h e sequences o f the peptides are shown i n Figure 1. Reprinted from reference 3, with permission.

(P2)

(I3,14,15).

P2,13,14

I5

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

2-

63

Downloaded by NORTH CAROLINA STATE UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch005

64

CHROMATOGRAPHY IN BIOTECHNOLOGY

may contaminate the displaced peptide product (P2) i n the breakthrough fraction; i f t i e sample load was too l o w , some peptide product may remain on the column, contaminated with hydrophobic impurities (I3 to 15). T h e situation would be even more complex with both hydrophilic and hydrophobic peptide impurities present. These concerns prompted the development o f a m u l t i - c o l u m n approach to preparative S D C o f peptides (6). Figure 3 illustrates this multi-column strategy through its application to the purification o f 100 m g o f a synthetic decapeptide. In addition to the desired peptide product, P , the crude peptide mixture also contained significant levels o f both hydrophilic ( I i ) , and hydrophobic (I2 to I5) impurities, as shown i n the analytical run (Figure 3, top profile). The 100 m g sample (in 0.1% aqueous T F A ) was applied to a multi-column set-up, consisting of ten 30 χ 4.6 m m I.D. Cs column segments (columns 1-10) i n series. Isocratic elution with 0.1% aqueous T F A (at a flow-rate o f 0.5 ml/min) was continued up to a total run time o f 50 m i n , which included the time taken for sample loading. F o l l o w i n g S D C and elution o f the i n d i v i d u a l c o l u m n segments, the distribution o f sample components through the ten c o l u m n segments was determined by analytical runs on the multi-column set-up (Figure 3, bottom profiles). T h e breakthrough fraction (0) and c o l u m n 1 (at the multi-column outlet) contained only hydrophilic impurities ( I i ) . C o l u m n 10 (at the m u l t i column inlet) contained only hydrophobic impurities (I2 to I5), while columns 8 and 9 contained a small amount o f peptide product contaminated w i t h hydrophobic impurity, I2. Columns 2-7 contained pure product, the great majority o f which was i n columns 3-7. O f the total amount o f desired peptide product loaded onto the column, 90% was recovered, with 8 1 % o f the peptide product isolated as pure peptide. The utilization of ten Cs column segments (instead of just the one segment employed i n Figures 1 and 2) enabled the application o f the substantial sample load o f 100 m g on what is effectively a standard analytical column (300 χ 4.6 m m I.D.). F r o m Figure 3, very efficient use o f the column capacity had been made during the S D C separation. In addition, only eleven fractions (the breakthrough fraction and the fractions obtained from individual elution o f the ten column segments) had to be analyzed, w h i c h greatly simplifies the fraction analysis compared to traditional preparative gradient elution chromatography. One important point to note from Figure 3 is that, although the sample l o a d was not sufficient to displace a l l hydrophilic impurities from the multisegmented column (an even higher sample load w o u l d have been required to achieve this), pure product is still obtained, since the crude sample components have been distributed along a chromatography bed divided into segments, rather than along a continuous packed bed as was the case for Figures 1 and 2. Thus, even at unoptimized (i.e., too l o w or, for that matter, too high) sample loads, significant levels of pure product may still be obtained, highlighting the flexibility of this approach. Application of Preparative Reversed-Phase S D C to the of PharmaceuticaUy-Important Peptides

Purification

In order to investigate the potential value o f the sample displacement approach to the pharmaceutical industry, preparative S D C was now applied to the purification o f a five-residue synthetic peptide [Glp-His-Trp-Ser-Tyr, denoted L H R H (1-5)], representing part o f the sequence o f luteinizing hormone-releasing hormone ( L H R H ) . In addition to its pharmaceutical significance, this peptide represented an interesting challenge due to its relatively l o w solubility i n an aqueous mobile

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Downloaded by NORTH CAROLINA STATE UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch005

5. HODGES ET AL.

Reversed-Phase Sample Displacement Chromatography

ι 28

1 ι 35 25

1 ι 38 29

1 35

Figure 3. M u l t i - c o l u m n approach to reversed-phase S D C o f peptides: purification o f 100 m g o f a synthetic decapeptide crude mixture. C o l u m n : ten 30 χ 4.6 m I.D. Aquapore R P 3 0 0 Cs (7-μπι particle size, 3 0 0 - Â pore size; Pierce, Rockford, I L , U . S . A . ) column segments i n series. The numbering of the ten columns (or fractions) starts at the column segment closest to the detector. Thus, column 1 (fraction 1) is at the multi-column outlet, while column 10 (fraction 10) is at the inlet. T h e fraction marked 0 is the breakthrough fraction. H P L C instrument: V a r i a n V i s t a Series 5000 l i q u i d chromatograph (Varian, Walnut Creek, C A , U . S . A . ) coupled to a HewlettPackard (Avondale, P A , U . S . A . ) H P 1 0 4 0 A detection system, HP9000 Series 300 computer, HP9133 disc drive, H P 2 2 2 5 A Thinkjet printer and H P 7 4 4 0 A plotter. Conditions: (1) analytical separation (top elution profile) o f the peptide mixture on the multi-column set-up [linear A B gradient (1% B / m i n and 1 ml/min), where eluent A is 0.1% aqueous T F A and eluent Β is 0.1% T F A i n acetonitrile]; (2) column equilibration, sample loading and the preparative SDC run (flow-rate=0.5 m l / m i n ; run time = 50 min) were a l l carried out by isocratic elution with 0 . 1 % aqueous T F A ; (3) f o l l o w i n g isocratic elution o f each individual column segment with 25% (v/v) aqueous acetonitrile containing 0.1% (v/v) T F A , fraction analysis o f the resulting peptide solutions was carried out by linear A B gradient elution (1% B / m i n at 1 ml/min) on the multi-column set-up). The detection wavelength was 210 n m for both analytical and preparative runs. T h e analytical elution profiles at the bottom show the peptide components retained o n each individual column segment (columns 1-10) f o l l o w i n g the S D C run. T h e peaks are a l l correctly proportioned. Ρ is the desired product; the other peaks are hydrophilic (Ii) and hydrophobic impurities. The sequence of the peptide product is A c - A r g - G l y - V a l - V a l - G l y - L e u - G l y - L e u - G l y - L y s amide, where A c denotes N - a c e t y l and amide denotes C - a m i d e . Reprinted from reference 6, with permission.

(I2-I5)

a

a

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

65

66

CHROMATOGRAPHY IN BIOTECHNOLOGY

Downloaded by NORTH CAROLINA STATE UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch005

phase based on 0.1% T F A . A l s o , T F A is not the acidic reagent o f choice since the trifluoroacetate salt o f the p u r i f i e d peptide (due to i o n - p a i r i n g o f the trifluoroacetate counterion with positively charged residues) is undesirable as a pharmaceutical product. T h u s , the mobile phase o f choice for preparative reversed-phase S D C o f L H R H ( l - 5 ) was 5% (v/v) aqueous acetic acid. Acetic acid is occasionally employed as a component o f mobile phase systems for R P C o f peptides (11), the acetate counterion being more hydrophilic than trifluoroacetate and producing acetate salts o f peptides containing positively charged groups. In addition, this particular peptide was more soluble i n 5% acetic acid. Standard Multi-Column Approach to Preparative S D C of L H R H (15). Figure 4 illustrates the results o f applying the multi-column S D C strategy (on the same ten-column set-up described i n Figure 3) to the purification o f 100 m g o f crude L H R H (1-5). Figure 4 A shows the analytical profile o f the crude mixture, where the desired peptide product, P , is contaminated by both hydrophilic and hydrophobic impurities (I). The sample (in 8 m l of 5% aqueous acetic acid) was loaded onto the column at a flow-rate o f 0.2 ml/min (i.e., 40 m i n was required for total loading o f the sample), followed by further isocratic elution at the same flow-rate with 5% aqueous acetic acid for 30 min, for a total run time o f 70 m i n . A previous study has demonstrated that l o w flow-rates enhance overall yield o f pure product (6). Figure 4 B shows the distribution o f sample components through the tencolumn segments following S D C . T h e breakthrough fraction ( B T ; inset profiles) contained only hydrophilic impurities. Column 1 ( C I ; inset profiles) at the multicolumn outlet, contained product (10.6% o f total product recovered) as well as hydrophilic impurities. Columns 9 and 10 (C9 and C 1 0 , the latter being at the multi-column inlet) contained the hydrophobic impurities and a small amount (3.2%) o f recovered product. Columns 2 to 8 (C2 to C8) contained pure product only, representing 86.2% of product recovered and underlining a very successful purification of L H R H (1-5). Peptide Elution in Preparative S D C of L H R H (1-5) in the Absence of Organic Modifier. F r o m Figure 4, the presence of desired product on column 1 ( C I ) indicated that the run was carried out with close to optimum column loading, i.e., where the product displaces totally all hydrophilic impurities into the breakthrough fraction ( B T ) , while all product remains on the column. In an attempt to avoid the use o f organic modifier to elute the product from the individual column segments, the run time was now increased, with a v i e w not only to complete the displacement of hydrophilic impurities from the m u l t i c o l u m n set-up, but also to elute the product from the c o l u m n with the 100% aqueous mobfle phase. Figure 5 illustrates the effect o f increased run time on preparative S D C o f 100 m g o f crude L H R H (1-5) (analytical profile shown i n Figure 5 A ) . The sample was loaded onto the column i n 5% aqueous acetic acid as before (Figure 4), followed by further isocratic elution with the same mobile phase for 140 m i n , for a total run time o f 180 m i n . U n l i k e the previous run (Figure 4), 2-min (i.e., 0.4 ml) fractions of the column effluent were collected as soon as sample loading commenced. F i g u r e 5 B shows the d i s t r i b u t i o n o f sample components i n the breakthrough fractions, as well as through the ten-column segments. T h e major observation here is that fully 88.8% o f recovered product has been displaced from the c o l u m n , w i t h pure L H R H (1-5), representing 83.4% o f recovered product, found in fractions 38-90 (in a total volume o f only 21.2 ml). Only 11.2% of product remained on the multi-segmented column, 10.9% recovered as pure

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

5.

HODGES ET AL.

Reversed-Phase Sample Displacement Chromatography

Downloaded by NORTH CAROLINA STATE UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch005

ε c LU ϋ Ζ


m
18%) are required for isocratic elution of the peptide off the column within a reasonable time. A t a concentration o f 15% methanol i n the mobile phase, the peptide had not been eluted from the column after 100 min i n these analytical runs. The question remained whether, under high sample loads, a much lower level of methanol could be utilized to aid the sample displacement process. In Figure 7, 100 m g o f crude L H R H (1-5) has been applied to the multic o l u m n set-up (40 m i n required for sample loading) and isocratic elution continued with 5% aqueous acetic acid for a further 30 m i n , for a run time o f 70 min. Thus, at this time, the sample components have been distributed through the breakthrough fraction and column segments i n a manner to that illustrated i n Figure 4, i.e., the major preparative S D C separation process has now taken place. Isocratic elution was now continued with 5% methanol i n 5% aqueous acetic acid for 110 m i n , for a total run time o f 180 m i n (i.e., identical to that o f Figure 5 for a meaningful comparison). F r o m the results of Figure 6, this l o w level o f methanol was deemed suitable for aiding the displacement o f product from the column without simply washing product and hydrophobic impurities off the column, and thus, running the risk o f contamination o f purified product. Comparing Figures 5 and 7, it is apparent that the presence o f methanol has indeed moved the product further along (and aided i n its displacement from)

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

69

CHROMATOGRAPHY IN BIOTECHNOLOGY

Downloaded by NORTH CAROLINA STATE UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch005

70

100

!

1

1

1

1

0

10

20

30

40

ELUTION TIME (min) Figure 6. Analytical R P C o f L H R H (1-5). C o l u m n and H P L C instrument: same as Figure 4. Conditions: (1) main profile, linear A B gradient ( 1 % B/min) at a flow-rate o f 1 ml/min, where eluent A is 0.1% aqueous T F A and eluent Β is 0 . 1 % T F A i n methanol; (2) inset profile, isocratic elution at a flow-rate o f 1 m l / m i n , where the eluent is 0 . 1 % aqueous T F A containing 15%, 18%, 20%, 25% or 30% (v/v) methanol. Absorbance at 210 nm.

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

5.

HODGES ET AL.

Reversed-Phase Sample Displacement Chromatography

Downloaded by NORTH CAROLINA STATE UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch005

CRUDE

Ε c ο τCM

—ι— 25

ι— 15

LU ϋ Ζ

Fr. 37-90 94.196 Ρ

< m

oc ο

Fr.36 2.8% Ρ

CO

m
6 0 A , < 1000A) which w i l l allow the rapid transport o f solutes to the chromatographically active surface (7). T h e ideal support's surfaces should also be amenable to chemical modification to allow the formation o f a variety o f different chromatographic supports such as reversed-phase, hydrophobic interaction and ion exchange materials. F i n a l l y , the surface should be as energetically homogeneous as possible to maximize efficiency. A s we have found, porous microparticulate zirconium oxide fulfills many o f the requirements of the ideal support for bioseparations. B a r e Z i r c o n i u m O x i d e S u p p o r t s . Over the past several years, we have been investigating the chromatographic properties o f specially formulated porous particles of zirconium oxide (7-23). These particles have a nominal particle diameter o f five microns with 300 angstrom average pore diameter. The surface area is nominally 30 square meters per gram. The surface o f zirconium oxide, like other metal oxides, is very complex. T h e major surface species have been spectroscopically identified as being Bronsted acid sites, Bronsted base sites and L e w i s acid sites (12,15 and references contained therein). These species are shown schematically in Figure 1. Bronsted acid sites arise from the acidity o f surface bound hydroxyl groups and tightly bound water molecules. Bronsted base sites arise from species such as bridging oxygen atoms. Although Bronsted base sites may also be classified as L e w i s base sites, no Bronsted base activity has been observed i n either liquid-solid or gas-solid adsorption studies. The L e w i s acid sites, which are not found on silica, arise i n a different manner than do the Bronsted sites. In the bulk o f zirconium oxide, the bonding o f metal and oxygen atoms is governed b y the coordination geometry o f the zirconium(IV) i o n . Tetravalent zirconium(IV) ion is heptacoordinate i n monoclinic zirconium oxide, the form o f zirconia used i n these studies (15). In the interior o f the material, the coordinatively bound oxygen atoms are i n resonance with the covalently bound oxygen atoms. A s a result, an extensive network o f bonding occurs which gives zirconium oxide its high mechanical strength. When this bonding continuity is interrupted, at a surface for example, the full coordination o f zirconium ions by oxygen atoms is not possible. A s a result,

American Chemical Society Library 1155 16th St.. N.W.

In Chromatography in Biotechnology; Horváth, C., et al.; Washington. DC. 20036 ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

CHROMATOGRAPHY IN BIOTECHNOLOGY

148

Downloaded by UNIV OF ARIZONA on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch011

Table I . Properties o f an Ideal Chromatographic Support* Polymeric Silica Zirconia Phases Mechanical Stability ++ H i g h Surface A r e a ++ Control o f Average Pore Diameter + + Control o f Particle Diameter ++ G o o d Solute Mass Transfer ++ Chemically Flexible ++ Energetically Homogeneous Chemical Stability Thermal Stabilitv + a. + + denotes very good performance 4- denotes good performance - denotes fair performance

a)

++ ++ ++ ++ ++ ++ -

++ ++

Η.

H y

o I

I

O-Zr-0

+ ++ + ++ + + + ++ -

Η I

O—Zr-O-Zr-0

I

ο

b)

Ο

0

0— Z r

/

N

Zr—Ο s

o'W o

Ο—Zr—Ο

(/ \>

Figure 1. Schematic depicting surface site types on monoclinic zirconium oxide. ( R e p r o d u c e d from reference 21. C o p y r i g h t 1992 A m e r i c a n C h e m i c a l Society.)

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Downloaded by UNIV OF ARIZONA on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch011

11.

CARR ET AL.

Zirconium Oxide Based Supports

149

a number o f coordination sites are exposed at the surface. These L e w i s acid sites accept electron pairs from a variety o f L e w i s bases, depending upon the composition o f the solution i n equilibrium with the surface. These coordinated L e w i s bases may be water molecules, hydroxide ions, L e w i s base solutes or other types o f L e w i s bases added to the eluent. Despite the apparent complexity o f the zirconia surface, the material is quite stable towards chemical attack. Likewise, the bonding structure makes zirconia a very physically stable support for use i n process and analytical scale chromatography. These desirable properties have prompted others to use zirconia as a cladding to improve the physical and chemical stability o f silica particles (24). Unfortunately, clad materials do not show the same stability as does homogeneous bulk zirconium oxide. The stability o f zirconium oxide toward extremes o f p H is illustrated i n Figure 2. Here, the stability o f zirconium oxide was compared to chromatographic grade alumina, which is much more alkaline stable than silica, i n a "static" stability study. In this work, the particles were placed i n a closed container and challenged with a variety o f buffers at different p H s . These conditions are far less aggressive than those used i n a chromatographic stability test where the buffer is continuously passed over the surface o f the particles. Under continuous flow conditions, it is impossible for the buffer solution to become saturated and possibly suppress further dissolution o f the support. A s shown i n Figure 2, alumina dissolves i n aqueous buffer at p H s greater than 12 and less than 3. In contrast, we could not detect zirconium at any p H from 1 to 14 using inductively coupled plasma spectroscopy as the detection method (8,12). The detection limit by this technique is 0.03 micrograms per milliliter. It should be noted that alumina is far less soluble than silica i n aqueous media at a l l p H s (12). The major challenge i n using zirconia is that the unmodified surface is not biocompatible. This is analogous to the situation with silica where proteins are not readily separated on the surface o f native silica gels. The gradient elution separation of bovine serum albumin from myoglobin, while possible, produces poor results (8,12), as shown i n Figure 3. Indeed, replicate runs on the same column show a continuous deterioration i n the performance o f the column because a very large fraction o f the protein which was injected does not elute. The main difficulty in using porous zirconium oxide for protein separations is the need to overcome the extremely strong adsorption o f proteins on the native surface. Phosphated Z i r c o n i u m O x i d e . Because o f its unique selectivity, there has been a great deal o f interest in the use o f calcium hydroxyapatite as a high performance liquid chromatographic support for the separation o f proteins (25-32). Great advances have been made i n understanding the complex adsorption process involved i n protein chromatography on calcium hydroxyapatite (25-27,33-36). Despite impressive improvements i n the physical stability and particle geometry (37-40), the hydroxyapatite materials are still chemically rather unstable. W e felt that it might be possible to induce the formation o f a layer o f zirconium phosphate on zirconium oxide, and thereby emulate some o f the properties o f calcium hydroxyapatite while retaining the physical and chemical stability o f the underlying zirconia material.

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

150

CHROMATOGRAPHY IN BIOTECHNOLOGY

Downloaded by UNIV OF ARIZONA on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch011

• • Ο • η Ε 2 σ» S υ Ε

Zr02 : pH 1-14 ΑΙ203 : pH 14 ΑΙ203 : ρΗ 12 ΑΙ203 : ρΗ 3

Ο 0

Ο

A A A — Η

-Α—I

1

A

I

1

1—Α-

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Κ Time (Days)

Figure 2. Solubility o f zirconium oxide versus aluminum oxide i n static p H stability studies.

Figure 3. Protein separation on zirconium oxide. A = bovine serum albumin; Β = myoglobin. Gradient was from 5 0 m M phosphoric acid to 5 0 m M sodium phosphate at p H 10 i n 10 minutes. F l o w rate was 1 m L / m i n at 2 5 ° C .

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

11.

CARR ET AL.

Zirconium

Oxide Based Supports

151

Downloaded by UNIV OF ARIZONA on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch011

Depending upon chemical conditions and thermal treatment, zirconium can combine with phosphate to form a wide variety o f zirconium phosphate forms (41,42). A s described elsewhere (10,11), we have been able to modify the surface of zirconium oxide by treating it with hot, concentrated phosphoric acid solutions. Such treatment results i n a surface where a significant fraction o f the surface is covered with some form o f zirconium phosphate and the rest o f the surface is covered with a layer o f adsorbed phosphate. This chemistry is illustrated i n Figure 4. After treatment o f zirconia with hot phosphoric acid, the material becomes a very useful chromatographic support for the separation o f proteins, as shown i n Figure 5. It should be pointed out that the retention times on this material are very reproducible, as are the shapes o f the chromatographic peaks (10,11,14). Further studies have shown nearly quantitative recoveries o f proteins with no denaturation. The mechanism o f retention o f proteins on phosphated zirconia supports was examined, i n part, by the experiments presented i n Figure 6. In this work (11,14), five proteins (cytochrome C , ribonuclease A , a-chymotrypsin, lysozyme and myoglobin) were studied. In a l l cases, a minimum o f 0 . 1 5 M p H 6 potassium hydrogen phosphate was present i n the eluent. The capacity factors for the various proteins are plotted versus the total potassium ion concentration o f the eluent. In the case o f the open symbols, the concentration o f potassium phosphate was varied. In the case o f the closed symbols, potassium phosphate concentration was fixed and additional potassium was added v i a the use o f potassium chloride. Under these conditions, it is only the total potassium concentration which controls the capacity factor. The concentration o f chloride or phosphate ions does not affect retention, as long as the phosphate ion concentration remains above approximately 50mM. If the buffer phosphate concentration falls below this threshhold concentration, loss o f efficiency and recovery begin to occur. These data indicate that the phosphated zirconium oxide phase is functioning as a pure cation exchange material. This conclusion is supported by the data shown i n Table II. In this work, the phosphate concentration was fixed at 2 0 0 m M and the solution p H was held reasonably constant between 6.85 and 6.90. The cation i n the mobile phase was varied from potassium to sodium to ammonium i o n . Clearly, the capacity factors of the various proteins changed significantly with the nature o f the cation. It should be further noted that only very basic proteins (e.g. lysozyme and cytochrome C ) are retained on the phosphated zirconia phase. This is i n marked contrast to the performance o f calcium hydroxyapatite, wherein both anionic and cationic proteins may be retained. Thus, the phosphated zirconia surface does not truly emulate the behavior o f hydroxyapatite. This does not mean that the phosphated zirconia material is not a chromatographically useful material. First, the sequence o f protein elution, as shown i n Figure 5, is markedly different from the elution sequence o f cationic proteins on conventional silica-based bonded phase cation exchangers. Second, it is not difficult to find practical circumstances under which the ability o f this material to adsorb positively charged proteins is useful. One such separation is

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

CHROMATOGRAPHY IN BIOTECHNOLOGY

152

Simplified Model of Zirconia Surface

Β OH ι

OH OH /\

Downloaded by UNIV OF ARIZONA on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch011

77*777

ζ/

ζ

ζ

Π ΤΠ777 77

Under "Mild" Phosphating Conditions

OH 0 ι W Ρ Ρ 0

ζ

/

0

° 0 \\|/ Ρ

H

ζ

0

H

ζ

7777 77 777 '/77777 Under "Rigorous" Phosphating Conditions

F

ο

C,

Z

r

»

\

7r » *_ Z r

/ Zr

o-p-

ο Ο

Ο

y

ΓΎΐτ\Τ\ΥΊϊΤ Figure 4. Proposed models for the surface species on various phosphate treated zirconias. A ) terminal hydroxide group; B ) bridging hydroxide group; C ) L e w i s acid site; D ) physi-adsorbed phosphate group; E ) esterified phosphate group covalently bound to the surface; F ) multi-layer zirconium phosphate region resulting from partial dissolution o f zirconia matrix.

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Downloaded by UNIV OF ARIZONA on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch011

CARR ET AL.

Zirconium

153

Oxide Based Supports

Figure 5. Chromatographic separation o f cationic proteins on phosphate modified zirconia. Linear gradient elution from 50 to 5 0 0 m M potassium phosphate buffer at p H 7.0. F l o w rate was 0.5 m L / m i n .

" 0.300

0.350

0.400

0.450

0.500

0.550

0.600

+

potassium ion concentration, [ K ] Figure 6. Isocratic protein retention as a function o f potassium i o n concentration. Proteins: ( O ) myoglobin; (Δ) lysozyme; ( • ) a-chymotrypsin; (v) ribonuclease A ; ( O ) cytochrome C . M o b i l e phase: (closed symbols) 0 . 1 5 M potassium phosphate p H 6.0 with added potassium chloride to give indicated potassium ion concentration; (open symbols) potassium phosphate buffer at p H 6.0 at the indicated concentration.

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

154

CHROMATOGRAPHY IN BIOTECHNOLOGY

Downloaded by UNIV OF ARIZONA on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch011

shown i n Figure 7. Immunoglobulin G can be produced i n animal cell tissue culture. Such cultures require the presence o f albumin and transferrin i n the growth media. A s shown i n Figure 7, bovine serum albumin and transferrin are very easily separated from I g G on the phosphated zirconia support. The separation is excellent and it is possible to heavily overload the column before the I g G peak can no longer be resolved from the transferrin peak. Z i r c o n i u m O x i d e A d s o r p t i o n M e c h a n i s m Studies. In order to carry out the separation o f proteins on bare zirconium oxide supports, it is essential to understand the surface chemistry responsible for the adsorption o f proteins on this material. A s shown i n Figure 1, we believe that the L e w i s acid chemistry o f surface zirconium ions with certain types o f solute L e w i s bases is involved i n the "irreversible" adsorption process for proteins. In order to gain a greater understanding o f the role o f L e w i s acid-base processes on zirconium oxide, we carried out a series o f studies using a wide variety o f l o w molecular weight carboxylic acids on zirconium oxide. The result o f one such study is presented i n Figure 8. Here we show the dependence o f the capacity factors o f a wide variety o f substituted benzoic acids on the p K a o f that derivative (21). This study was carried out i n two types o f buffers, acetate buffer and six similar aminosulfonate buffers with different p K a values. A s shown, there is a reasonable correlation between the capacity factors and the pKas o f the benzoate derivatives. Virtually a l l o f the data shown i n Figure 8 can be collapsed onto a single line, as shown i n Figure 9 when the effect o f p H on retention is factored out. Note that the open circles correspond to the set o f six aminosulfonate buffers, whereas the closed circles correspond to the acetate buffer data. The data indicate that an increase i n p H decreases retention. A decrease i n the p K a o f the solute also decreases the capacity factor. These results are chemically quite significant. W e do not believe that the decrease i n capacity factor as the p H is raised is due to a change i n the state o f ionization o f the benzoic acid derivative. In fact, over the entire p H range presented i n Figures 8 and 9, the benzoic acid derivatives were all essentially completely ionized (except i n acetate buffer). The decrease i n capacity factor with p H is a consequence o f the increased competition between hydroxyl ions (which are strong, hard L e w i s bases) with the carboxylate solutes for the available zirconium ion L e w i s acid sites. The variation i n retention with the p K a o f the solute is also related to L e w i s acid-base processes. The affinity o f the benzoate anions for the proton is directly measured by its Bronsted basicity. One must keep i n mind the fact that as the Bronsted acidity o f a weak acid decreases, the Bronsted basicity o f the conjugate base, that is the affinity for protons, increases. The correlation between capacity factor and p K a is easily reconciled when one considers that the proton is the quintessential hard L e w i s acid. Similarly, zirconium(IV) is an exceptionally strong, hard L e w i s acid. Since protons and zirconium(IV) ions are both hard L e w i s acids, the correlation between proton affinity and zirconium ion affinity should be high. A s mentioned above, the solutes are a l l essentially completely ionized under the chromatographic conditions employed here thus the degree o f ionization does not significantly affect retention.

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Downloaded by UNIV OF ARIZONA on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch011

11.

CARR ET AL.

Zirconium

Oxide Based Supports

155

Table II. Effect o f Various Cations on the Isocratic Elution o f Proteins* capacity factor Displacing Cation lysozyme ribonuclease A q-chymotrypsin potassium 0.77 6.27 4.72 sodium 3.17 10.28 9.99 ammonium 0J9 enr/ 18.97 S O U R C E : Adapted from reference 14. a. M o b i l e phase was 2 0 0 m M phosphate buffer with counterion as listed above at p H 6.9. Column dimensions were 50 χ 4.6 m m filled with 6-11 micron phosphated zirconia particles. F l o w rate was 0.5 m L / m i n with detection at 280 nm. b. eno = elution not observed.

BSA

30 min linear gradient from 0.05 M P 0 to 0.5 M P 0 (pH 6.0) 4

4

20/x particles

Figure 7. Separation o f immunoglobulin G from a modeled fermentation broth. Linear gradient from 50 to 5 0 0 m M sodium phosphate buffer at p H 6.0. F l o w rate was 0.5 m L / m i n .

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

CHROMATOGRAPHY IN BIOTECHNOLOGY

Downloaded by UNIV OF ARIZONA on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch011

156

"

3.25 3.50 3.75 4.00 4.25 4.50 4.75 5.00 5.25

pKa Figure 8. Plot o f log k ' o f parabenzoic acid derivatives versus their p K a i n various buffers. ( · ) acetate p H 4.8; ( O ) P I P E S p H 6.8; (v) M O P S p H 7.2; ( • ) H E P E S p H 7.5; ( • ) E P P S p H 8.0; ( • ) T A P S p H 8.4; (Δ) C H E S p H 9.3; all l O O m M . Solid lines indicate least squares fits.

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Downloaded by UNIV OF ARIZONA on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch011

CARR ET AL.

-8

Zirconium Oxide Based Supports

-6

-4

-2

0

(0.67pKa - pH) Figure 9. Correlation o f the capacity factors for parabenzoic acid derivatives with solute Bronsted acidity and eluent p H . ( O ) aminosulfonate buffers; ( · ) acetate buffer. Regressions are least squares fits.

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Downloaded by UNIV OF ARIZONA on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch011

158

CHROMATOGRAPHY IN BIOTECHNOLOGY

In addition to the effect o f p H on retention, the eluent used i n conjunction with zirconium oxide is also extremely important. A s shown i n Table III, the retention o f benzoic acid derivatives on zirconia is very strongly influenced by the presence o f L e w i s bases i n the eluent which may compete with solute L e w i s bases for the surface L e w i s acid sites. When phosphate or fluoride is added to the eluent as competing L e w i s bases, there is virtually no retention o f the benzoic acid derivatives. Both phosphate and fluoride are very strong, hard L e w i s bases towards zirconium(IV) ions, consequently they are very strong competitors for the occupation o f coordination sites on the zirconium oxide surface. A wide variety o f competing L e w i s bases have been evaluated as to their relative displacing strengths towards benzoic acid derivatives. In essence, Table III constitutes an eluotropic series o f displacement strength of the various L e w i s bases on zirconium oxide. A rough correlation exists between the logarithmic capacity factors o f the benzoic acid derivatives on the support and the logarithm o f the formation constant o f a complex between zirconium(IV) ions and a number o f L e w i s base eluents shown i n this table for which data are available. P r o t e i n C h r o m a t o g r a p h y on L e w i s base M o d i f i e d Z i r c o n i a . A chromatogram of a set o f proteins separated on the bare zirconium oxide surface when operated i n a medium containing 2 0 m M sodium fluoride is shown i n Figure 10. A 0 to 0 . 7 5 M sodium sulfate gradient produces a sequence o f sharp, extremely well resolved peaks. This chromatogram shows that, once the L e w i s acid-base chemistry o f zirconium oxide is taken into account, by adding the strong L e w i s base to the eluent (i.e. F ) it is possible to use the unmodified surface as a support for the separation of proteins. The relationship between the capacity factors o f a series o f proteins under similar elution conditions and the protein's isoelectric point are shown i n Figure 11. Here, the filled circles represent proteins at a p H below their isoelectric point. Clearly, both cationic and anionic proteins are retained on zirconia when operated i n fluoride media. This behavior is quite analogous to the behavior o f calcium hydroxyapatite. W e believe that the anionic proteins are mainly retained because o f their ability to interact with surface zirconium(IV) ion sites v i a L e w i s acid-base interactions. Another interesting but puzzling aspect o f this material is that denatured proteins are not retained. This is entirely analogous to Gorbunov's observations with hydroxyapatite (25). A s shown i n Table I V , the concentration o f sodium fluoride i n the eluent has a very strong effect on protein retention. A s anticipated, as the fluoride concentration is increased, retention decreases considerably. A l s o note i n this table that although the p H was 5.5, proteins with l o w isoelectric points, such as ovalbumin and serum albumin, are rather strongly retained. In contrast, cationic proteins, such as ribonuclease B , cytochrome C and lysozyme are less strongly retained, even i n 0 . 5 M sodium chloride media. Clearly, zirconium oxide is able to retain and separate both cationic and anionic proteins depending on the relative contributions o f ligand exchange and electrostatic interactions to retention.

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

11.

CARR ET AL.

Zirconium Oxide Based Supports

Downloaded by UNIV OF ARIZONA on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch011

Table III. Eluotropic Series Based on Benzoic A c i d Derivatives on Zirconia* 4-nitro- 4-cyano- 4-formyl- 4-chlorobenzoate benzoate benzoate benzoate Eluent Species 0.0 0.0 0.0 0.0 phosphate 0.0 0.0 0.0 0.0 fluoride 0.4 0.2 0.2 0.2 ethylphosphonate 0.6 0.2 0.2 0.3 citrate 0.5 0.3 0.2 malate 0.2 0.9 0.4 0.3 ethylenediaminetetraacetate 0.3 0.8 0.5 0.5 0.5 oxalate 1.4 0.8 0.7 0.6 succinate 1.5 0.9 0.7 0.6 glutarate 1.5 0.7 0.7 0.6 aspartate 1.7 0.9 0.7 0.7 adipate n/a 0.7 0.7 0.7 maleate 2.0 1.1 1.0 malonate 0.9 2.4 1.3 1.0 pimelate 0.9 3.0 1.8 1.6 1.5 sulfate 3.4 2.0 1.7 1.8 glycolate 5.8 2.2 1.4 2.9 tartarate 3.8 2.2 2.0 1.9 borate 5.2 2.2 2.1 1.6 nitrilotriacetate 7.6 3.4 2.2 suberate 2.0 3.4 5.9 3.0 3.0 thiosulfate 7.7 3.9 3.1 iminodiacetate 3.1 20.6 3.0 3.0 3.8 sebacate 12.7 7.6 5.6 6.3 acetate 15.6 6.7 6.0 7.6 TRIS 20.9 9.5 9.7 8.8 formate 23.5 15.3 11.1 10.8 sulfamate 34.4 18.7 15.6 butyrate 13.0 40.4 22.9 18.2 17.0 bromide 37.1 24.3 17.1 17.6 urea 27.2 eno 22.4 20.8 guanidine hydrochloride 45.2 29.6 butanesulfonate 22.8 23.5 57.6 33.5 24.7 26.0 nitrate 59.9 35.1 24.7 26.6 chloride eno 43.3 34.3 thiocyanate 32.3 eno 42.3 34.1 33.6 ethylene glycol eno eno 43.9 41.7 thiourea S O U R C E : Reprinted with permission from reference 22. Copyright 1991. a. Eluent was 2 0 m M o f the above Lewis base i n 2 0 m M M E S (morpholinoethanesulfonic acid) at p H 6.1. F l o w rate was 1.0 m L / m i n at 3 5 ° C . Injections were 10 JKL volumes o f l O m M solutions water onto a 50 χ 4.6 mm column. b. eno = elution not observed. b

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

CHROMATOGRAPHY IN BIOTECHNOLOGY

160

Downloaded by UNIV OF ARIZONA on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch011

LC

η 280,4

o-F

550,513

20

10 Τ i tae ( m i n .

J376E04R.D

)

Figure 10. Protein separation on fluoride modified zirconium oxide. A linear gradient of 0 - 0 . 7 5 M sodium sulfate i n l O O m M sodium fluoride and 2 0 m M M E S at p H 5.5 was used. F l o w rate was 0.5 m L / m i n at 3 5 ° C . Protein loadings were 4.4 /xg lysozyme, 15.4 μι α - c h y m o t r y p s i n , 13.6 μζ myoglobin and 15.4 /xg cytochrome C .

Figure 11. Correlation of capacity factor with p i for proteins on fluoride modified zirconia. Gradient elution from 0 to 0 . 5 M sodium sulfate i n 30 minutes was used. Both buffers contained 2 0 m M sodium fluoride and 2 0 m M M E S at p H 6.2 at 3 5 ° C . F l o w rate was 0.5 m L / m i n and detection was at 280 nm.

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

30

11.

CARR ET AL.

Zirconium

161

Oxide Based Supports

Downloaded by UNIV OF ARIZONA on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch011

F o r process scale separations, the loading capacity and protein recoveries are critical parameters. A s shown i n Figure 12, significant amounts o f proteins can be loaded onto the column before there is any significant change i n capacity factor. The data i n Table V show that the recovery o f protein on zirconia i n fluoride media is nearly quantitative. W e have found that when a sufficient amount o f L e w i s base is present i n the eluent, retention times on zirconia are extremely reproducible. This amount is dependent upon the adsorption properties o f the particular L e w i s base on zirconia at a particular p H . B y maintaining a nearly saturated isotherm for the eluent L e w i s base at the surface L e w i s acid sites, the solute-surface L e w i s interactions are attenuated. Increases i n eluent L e w i s base concentration serve to further reduce the

Table I V . Protein Retention i n Fluoride Media* Caoacitv Factors 0.50M NaF O . l O M N a F 0.02M N a l Protein pi eno ovalbumin 4.7 10.6 17.3 32.5 14.2 23.1 bovine albumin 4.7, 4.9 34.1 human albumin 25.3 17.0 4.6-5.3 31.2 apo transferrin 26.5 16.6 5.9 16.7 24.6 myoglobin 6.8, 7.3 3.5 15.7 24.2 hemoglobin 6.9-7.4 3.9 eno eno 21.2 alcohol dehydrogenase 8.7-9.3 13.2 18.1 a-chymotrypsin 2.3 8.8 19.7 15.9 ribonuclease A 9.3 3.3 11.1 6.8 ribonuclease Β 3.6 9.3 28.4 22.6 cytochrome C 9.4, 9.0 5.1 10.3 lysozyme 5,2 0,2 11.0 S O U R C E : Adapted from reference 15. a. Buffers consisted of: A ) sodium fluoride and 2 0 m M M E S p H 5.5 and B) buffer A with 0 . 5 M sodium sulfate added. Gradient was 0 to 100% Β i n 30 minutes then back to 0% Β i n 15 minutes with a 15 minute equilibration period. F l o w was 1.0 m L / m i n at 3 5 ° C with detection at 280 and 410 nm. Typical injections were 10 /*g protein i n 2 0 m M M E S at p H 5.5. b. Elution not observed. b

Table V . Protein Recovery* Protein Assay #1 Assay #2 lysozyme 103.9% ± 2.5% 101.2% ± 2.2% myoglobin 107.7% + 6.1% 95.7% + 4 . 1 % S O U R C E : Adapted from reference 17. a. averages o f five separate recovery experiments.

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Downloaded by UNIV OF ARIZONA on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch011

162

CHROMATOGRAPHY IN BIOTECHNOLOGY

log (protein) Figure 12. Loading study on fluoride modified zirconia. ( O ) lysozyme; ( · ) lipase. Isocratic elution at 3 5 ° C and 0.5 m L / m i n flow rate. Eluent was 0 . 3 5 M potassium chloride with 2 0 m M sodium fluoride and 2 0 m M T A P S at p H 8.4.

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

11.

CARR ET AL.

Zirconium

Oxide Based Supports

163

ligand exchange contribution to retention. Decreases i n the eluent L e w i s base concentration decrease the ability o f the eluent L e w i s base to displace the bound solute L e w i s base, thereby reducing recovery and efficiency.

Downloaded by UNIV OF ARIZONA on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch011

Acknowledgments This work was supported i n part by grants to the Institute for Advanced Studies i n Bioprocess Technology from the 3 M Company and by grants from the National Science Foundation and the National Institutes o f Health. T h e porous zirconium oxide particles used throughout this work were provided by the Ceramic Technology Center at the 3 M Company. T h e authors wish to acknowledge the contributions o f D r . E r i c Funkenbusch to this work.

Literature Cited 1. Unger, K.K. Adsorbents in Column Liquid Chromatography In Packings and Stationary Phases in Chromatographic Techniques; Unger, K.K., Ed.; Marcel Dekker: New York, NY, 1990; pp 331-470. 2. Hetem, M.; VanDerVen, L.; DeHaan, J.; Cramers, C.; Albert, K.; Bayer, E. J. Chromatogr. 1989, 479, 269-295. 3. Iler, R.K. The Chemistry of Silica; J. Wiley Interscience: New York, NY, 1979. 4. Glajch, J.L; Kirkland, J.J.;Kohler, J. J. Chromatogr. 1987, 384, 81-96. 5. Anthoni, U.; Larsen, C.; Nielsen, P.H.; Christopherson, C. Anal. Chem. 1987, 59, 2435-2438. 6. D'Ambra, A.J. Personal Communication, Department of Chemistry, University of Minneosta, Minneapolis, Minnesota. 7. Rigney, M.P.; Weber, T.P.; Carr, P.W. J. Chromatogr. 1989, 484, 273-291. 8. Rigney, M.P.; Funkenbusch, E.F.; Carr, P.W. J. Chromatogr. 1990, 499, 291-304. 9. Weber, T.P.; Carr, P.W.; Funkenbusch, E.F. J. Chromatogr. 1990, 519, 31-52. 10. Schafer, W.A.; Carr, P.W.; Funkenbusch, E.F.; Parson, K.A. J. Chromatogr. 1991, 587, 137-147. 11. Schafer, W.A.; Carr, P.W. J. Chromatogr. 1991, 587, 149160. 12. Rigney, M.P. Ph.D. Thesis, University of Minnesota, 1988. 13. Weber, T.P. Ph.D. Thesis, University of Minnesota, 1991. 14. Schafer, W. A. M.S. Thesis, University of Minnesota, 1990. 15. Blackwell, J.A. Ph.D. Thesis, University of Minnesota, 1991. 16. Blackwell, J.Α.; Carr, P.W. J. Chromatogr. 1991, 549, 43- 57. 17. Blackwell, J.Α.; Carr, P.W. J. Chromatogr. 1991, 549, 59- 75. 18. Blackwell, J.Α.; Carr, P.W. J. Liq. Chrom. 1991, 14(15), 2875-2889. 19. Blackwell, J.A.; Carr, P.W. J. Liq. Chrom. 1992, 15(5), 727-751. 20. Blackwell, J.A.; Carr, P.W. J. Liq. Chrom. 1992, 15(9), 1487-1506. 21. Blackwell, J.Α.; Carr, P.W. Anal. Chem. 1992, 64, 853-862. 22. Blackwell, J.Α.; Carr, P.W. Anal. Chem. 1992, 64, 863-873. 23. Blackwell, J.Α.; Carr, P.W. J. Chromatogr. 1992, 596, 27- 41.

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

CHROMATOGRAPHY IN BIOTECHNOLOGY

Downloaded by UNIV OF ARIZONA on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch011

164

24. Stout, R.W.; DeStefano, J.J. J. Chromatogr. 1985, 326, 63- 72. 25. Gorbunoff, M.J. Anal. Biochem. 1984, 136, 425-432. 26. Gorbunoff, M.J. Anal. Biochem. 1984, 136, 433-439. 27. Gorbunoff, M.J.; Timasheff, S.N. Anal. Biochem. 1984, 136, 440-445. 28. Gorbunoff, M.J. Meth. Enzymol. 1985, 117, 370-379. 29. Bernardi, G. Meth. Enzymol. 1973, 27, 471-479. 30. Bernardi, G.; Giro, M.G.; Gaillard, C. Biochim Biophys Acta 1972, 278, 409-420. 31. Kadoya, T.; Ogawa, T.; Kuwahara, H.; Okuyama, T. J. Liq. Chrom. 1988, 11(14), 2951-2967. 32. Kadoya, T. J. Chromatogr. 1990, 515, 521-525. 33. Barroug, Α.; Rey, C.; Fauran, M.J.; Trombe, J.C.; Montel, G.; Bonel, G. Bull. Soc. Chim. France 1985, 4, 535-539. 34. Kawasaki, T.; Niikura, M.; Kobayashi, Y. J. Chromatogr. 1990, 515, 91-123. 35. Kawasaki, T.; Niikura, M.; Kobayashi, Y. J. Chromatogr. 1990, 515, 125-148. 36. Inoue, S.; Ohtaki, N. J. Chromatogr. 1990, 515, 193-204. 37. Kawasaki, T.; Takahashi, S.; Ikeda, K. Eur. J. Biochem. 1985, 152, 361-371. 38. Hirano, H.; Nishimura, T.; Iwamura, T. Anal. Biochem. 1985, 150, 228-234. 39. Kadoya, T.; Isobe, T.; Ebihara, M.; Ogawa, T.; Sumita, M.; Kuwahara, H.; Kobayashi, Α.; Ishikawa, T.; Okuyama, T. J. Liq. Chrom. 1986, 9(16), 3543-3557. 40. Bruno, G.; Gasparrini, F.; Misiti, D.; Arrigoni-Martelli, E.; Bronzetti, M. J. Chromatogr. 1990, 504, 319-333. 41. Amphlett, C.B. Inorganic Ion Exchangers; Elsevier: Amsterdam, 1964. 42. Clearfield, A. Zirconium Phosphates In Inorganic Ion Exchange Materials; Clearfield, Α., Ed.; CRC Press: Boca Raton, FL, 1982. Received October 30, 1992

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Chapter 12

Preparative Reversed-Phase Chromatography of Proteins Geoffrey B. Cox

Downloaded by PENNSYLVANIA STATE UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch012

Prochrom Inc., 5622 West 73rd Street, Indianapolis, IN 46278

The mass-overloaded separations of model proteins under preparative gradient-elution conditions are characterized by strong solute-solute displacements. These displacements dominate the course of the chromatography, allowing much higher throughput, purity and mass recovery than would be a priori predicted. The purity and recovery of the components were not dependent upon the gradient slope. The slope does, however, influence the run time; steeper gradients allow higher production rates. The loading conditions were found to be important since too high a concentration of organic modifier prevented complete uptake of the solutes. The column efficiency was shown to be important in that a minimum number of plates (corresponding to the use of particles of 15μmdiameter or less in this case) was required to maximize the displacements. Particle sizes below 15 μm were determined to be useful only in that they allow higher mobile phase velocities and therefore higher production rates. The optimum pore diameter of the packing material is a compromise between a value large enough to allow ready access of the molecules to the interior of the particles, but small enough to give the packing material a high surface area to maximize loadability. Guidelines for the design of preparative reversed-phase gradient-elution separations of proteins are given.

Reversed-phase chromatography has been used for the analysis o f peptides and smaller proteins for approximately the last 11 years (i),(2). It was shown that the use

0097-6156/93/0529-0165$06.00/0 © 1993 American Chemical Society

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

CHROMATOGRAPHY IN BIOTECHNOLOGY

Downloaded by PENNSYLVANIA STATE UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch012

166

of packings with large pore diameters is advantageous i n protein separations. The wider pores, usually 300 or 500 Â i n diameter, aid the mass transfer o f the bulky proteins i n and out o f the particle i n order fully to utilize the available surface area o f the packing material and to maximize the column efficiency. Such materials have much lower surface area than the small-pore packings conventionally used for smallmolecule separations. M o r e recently, some workers have turned to non-porous particles for the analytical chromatography o f proteins (J). These allow rapid separations but have very low surface area. Since the loading capacity of the column is a function o f the surface area o f the packing, these two approaches may not necessarily be advantageous for preparative applications. Preparative separations o f proteins and peptides have also been performed by reversed-phase chromatography over many years. M u c h o f the early work on reversed-phase separations was directed to small-scale preparative separations. The major restraint upon the use o f reversed-phase chromatography for protein purification relates to the recovery o f active protein from the column. It is known (4),(J) that many proteins denature upon interaction with the highly hydrophobic reversedphase matrix. U p o n denaturation, the molecules unfold and can present a greater number o f hydrophobic sites which can interact with the packing. This can lead to irreversible binding o f the solute or can result i n the elution o f unfolded protein from the column. M a n y small proteins can refold and regain their activity, but, for others, inactive or partially inactivated protein is recovered from the preparative column. Thus, the application o f reversed-phase techniques is restricted to that body of proteins which either do not unfold when i n contact with the matrix or which re-fold once they have eluted from the column. This generally is comprised o f proteins with molecular weights below 20 k D , although there are a few larger proteins which may also be purified by reversed-phase chromatography. Theory and Modelling. The theory o f heavily mass-overloaded preparative chromatography has developed very rapidly over the past few years. This has been the result o f work by the groups of Snyder (6) and Guiochon (7). Most o f this work involved die isocratic separations o f small molecules. Small solutes often behave i n a predictable manner in that many follow the Langmuir adsorption isotherm model. It has been demonstrated that solute-solute interactions are important i n preparative separations and the predictions o f the displacements and tag-along effects are by now well known. It happens i n practice that many solutes do not follow the competitive Langmuir model which is universally employed i n the modelling, even though individually they may follow Langmuir adsorption isotherms. In such cases, the displacements can be much greater - or much less - than those predicted (8). M u c h less attention has been devoted to the theory and model simulation o f gradient-elution separations. F o r small molecules, the concept o f "corresponding isocratic separations" has been introduced (?) i n which the gradient separation can be predicted from model simulations relating to isocratic separations where the capacity factor o f the component is set equal to the average capacity factor ( ¥ ) i n the gradient. Other work has been performed i n which the Craig countercurrent model

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Downloaded by PENNSYLVANIA STATE UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch012

12.

COX

Preparative Reversed-Phase Chromatography of Proteins

167

has been applied to gradient separations (70). A limitation o f these approaches is that the equations and simulation programs were, as for the isocratic models, developed for solutes which follow the Langmuir adsorption isotherm and the competitive Langmuir adsorption model. Because o f the difficulties i n their measurement, the adsorption isotherms o f large biomolecules i n reversed-phase chromatography are rarely reported. It is, however, clear that such solutes do not (and should not) normally follow the Langmuir isotherm (77). This places severe restrictions upon the application o f model predictions to the preparative gradientelution chromatography o f proteins. One feature o f the modelling o f preparative gradient separations is the prediction o f solute - solute interactions with much the same effects upon peak shapes as seen i n isocratic separations. Given the applicability o f the "corresponding" separations concept, this is not surprising. M a n y differences have been found between experiment and the modelling previously carried out for isocratic separations and it was not expected that the predictions from the modelling o f the gradients would be exactly followed i n practice. The practical examination o f interactions between biomolecules i n preparative gradient-elution chromatography is reviewed i n this paper. Experimental Studies Choice of model compounds: A number o f factors were considered when solutes were chosen for studies on the preparative reversed-phase chromatography of proteins. The solutes had to have large enough molecular weights to be representative o f the proteins which may be separated by reversed-phase chromatography. This implied a desired molecular weight somewhere between 10 000 and 20 000. The solutes had to be readily available i n reasonably large quantities and i n a state sufficiently pure that no pre-purification was needed. They also had to elute reasonably closely under most gradient conditions to allow a reasonable simulation o f a "real" separation. The final criterion was that one o f the solutes had to be detectable at a unique wavelength. This was to allow the independent monitoring o f its peak profile through the overlapping zones i n the chromatograms, thus eliminating the need to collect and analyze multiple fractions for each experiment. Three solutes, cytochrome c, lysozyme and ribonuclease a ( R N A s e ) , were chosen. Cytochrome c absorbs U V radiation at a wavelength o f 475 nm, where neither o f the other proteins have absorption. Under conventional reversed-phase conditions, R N A s e elutes earlier than cytochrome c, whilst lysozyme elutes later. Using these three solutes with a H P L C system fitted with a diode-array detector, it was therefore possible to monitor peak shapes during the mass-overloaded separations. Use o f a suitable data system allowed scaling and subtraction o f chromatograms to obtain individual component envelopes. One negative factor relating to the choice o f solutes was that the three solutes differed from one another structurally and the possibility o f extrapolation o f the results o f the studies to "real" samples may be limited. W o r k designed to address this question has recently been reported [(72)]. Protein - Protein Displacements in Reversed-Phase Preparative L C : The first observation o f strong solute - solute displacements i n the reversed-phase preparative

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

CHROMATOGRAPHY IN BIOTECHNOLOGY

168

gradient-elution o f proteins was reported i n 1989 (13), during experiments designed to investigate the predictions o f model simulations. These studies were carried out using a 100 n m pore C bonded-phase silica, with a conventional gradient ranging from 0 to 70 % acetonitrile i n a time o f 20 minutes. The chromatogram arising from the introduction o f a mixture containing 10 mg each o f lysozyme and cytochrome c is shown i n Figure 1. The solid line is the chromatogram taken at 290 nm, at which wavelength both solutes absorb radiation. The dashed line is the chromatogram taken at 475 n m , where only cytochrome c absorbs. This shows a very sharp drop i n the concentration o f cytochrome c at precisely the same point at which the lysozyme begins to elute. It is believed that the rapid increase in lysozyme concentration at the peak front is sufficient to displace the cytochrome c from the packing, thus affecting a very sharp separation between the two solutes. This behavior is reminiscent o f the effects seen i n displacement chromatography. The resemblance is more striking when the load o f cytochrome c is varied whilst that o f lysozyme is held at 10 m g . The resulting series o f chromatograms is shown i n Figure 2. These traces are all taken at 475 n m , and depict the peak shape o f cytochrome c at loads varying between 0.01 mg (an "analytical" load) and 5 m g . The trailing edge o f the cytochrome c peak eluted at a constant position, which coincided with the peak front of the lysozyme band. A t the lowest load, the cytochrome c peak was found to have a bandwidth one-third that o f a peak due to an equal load when injected alone. This narrowing o f the band was due to the displacement effect. It was noted that as the sample load of cytochrome c increased, the height o f the peak remained relatively constant while the band width increased with load. Such behavior is normal i n displacement chromatography once a displacement train is set up and the result suggests that parallels exist between the displacements seen during elution chromatography and those obtained i n "classical" displacement chromatography. A second experiment was carried out with R N A s e and cytochrome c. In this case, the cytochrome c was the later eluted peak and thus the shape o f the second component could be visualised. A sharp front was observed for the cytochrome c peak, indicating that no "tag-along" effects such as are often seen i n isocratic separations occurred. Subtraction o f the chromatogram o f cytochrome c at 475 nm (after suitable scaling) from the chromatogram taken at 290 nm confirmed that the displacement effect also occurred between these components i n that the trailing edge of the R N A s e peak exhibited the characteristic drop to the baseline at the point o f elution o f the cytochrome c peak front. The chromatograms are shown i n Figure 3. Similar results have been demonstrated for a separation o f cytochrome c and lysozyme where the (scaled) cytochrome c peak, as measured at 475 n m , was subtracted from the chromatogram obtained at 290 nm. The lysozyme peak, resulting as the residual chromatogram from this operation, was seen to be undistorted and to begin at the point at which the cytochrome c peak showed the onset o f the displacement effect

Downloaded by PENNSYLVANIA STATE UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch012

8

(14). Effect o f G r a d i e n t S l o p e : Modelling and simple theories o f preparative gradient separations indicate that the separation should be independent o f the gradient slope (75). This situation should be maintained i n the presence o f displacements;

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Downloaded by PENNSYLVANIA STATE UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch012

12.

COX

Preparative Reversed-Phase Chromatography of Proteins

169

Figure 1. Chromatogram of a mixture of cytochrome c and lysozyme. Solid line, 290 nm; dashed line, 475 nm. Sample load: 10 mg each. Other conditions: C o l u m n : Zorbax P S M 1000 C 8 , 15 cm χ 4.6 m m . 0 to 70% acetonitrile i n 0.1 % trifluoroacetic acid ( T F A ) i n 20 min. F l o w rate: 1 m l / m i n .

Time (min) Figure 2. Overlays o f 475 nm chromatograms o f mixtures o f cytochrome c and lysozyme. Sample loads: 10 mg lysozyme with 0.01 mg (dotted line), 0.3 mg (long dashed line), 1.0 mg (dashed / dotted line), 3.0 mg (short dashed line) and 5.0 mg (solid line) o f cytochrome c. Conditions as Figure 1.

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

CHROMATOGRAPHY IN BIOTECHNOLOGY

170

Downloaded by PENNSYLVANIA STATE UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch012

a

b

0

1

5

1

10

Time (min)

Γ

15

Figure 3. (a) Chromatogram o f 5 mg each o f cytochrome c and R N A s e , detector wavelength 290 n m . (b) Overlays o f the 475 n m chromatogram (solid line) and the result o f subtraction o f this from the 290 nm chromatogram (dashed line). Conditions as Figure 1.

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

12.

COX

Preparative Reversed-Phase Chromatography

of Proteins

171

provided they remain strong, there should be no effect o f slope. Experiments were carried out to demonstrate this by using four gradients o f different slopes [14]. The gradient details are shown i n Table I. The recoveries for cytochrome c at a purity o f 99.5% were calculated from the resulting chromatograms; the results are shown i n Table II. These values are sufficiently close to indicate that the gradient slope had no influence upon the isolation o f the product from the separation.

Downloaded by PENNSYLVANIA STATE UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch012

Table I. Chromatographic Conditions for Gradient Slope Experiments

Column: M o b i l e Phase:

F l o w Rate: Gradients:

Detection: Sample:

Zorbax Pro-10 Protein Plus, 150 χ 4.6 m m . Solvent A : 0.1 % aqueous trifluoroacetic acid. Solvent B : 0.1 % trifluoroacetic acid i n acetonitrile. 1.0 m l / m i n . Range: 5 to 70% Β Duration: (a) 15 minutes (b) 30 minutes (c) 45 minutes (d) 60 minutes U V , wavelengths 295 and 480 n m . Cytochrome c: 3 mg. Lysozyme: 10 mg.

Table II. Purity and recovery of cytochrome c Gradient duration 15 30 45 60

Purity 99.46 99.47 99.50 99.50

Recovery 93.1 93.7 91.3 95.9

Other conditions: See Table I .

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

CHROMATOGRAPHY IN BIOTECHNOLOGY

Downloaded by PENNSYLVANIA STATE UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch012

172

Effect of Gradient Range: It is customary i n the analysis o f proteins to reduce the analysis times by using a narrow gradient range that just brackets the compounds o f interest (16). The initial solvent composition is set to a point just below that at which the solute begins to migrate and the gradient is run at the desired rate o f increase until the component is eluted, at which point a regeneration or reequilibration step is used. The situation i n preparative chromatography is a little different. W h e n the sample is injected at a high initial solvent strength, there is incomplete retention o f the solute (14). This situation is seen i n Figure 4, where a sample o f cytochrome C and lysozyme (3 and 10 mg respectively) was loaded at an initial solvent concentration o f 31% acetonitrile. M u c h o f the sample was not retained, although the displacements were still seen between the retained portions o f the solutes. This is indicated by the fact that the tail o f the cytochrome c band is eluted earlier than the retention time o f an analytical scale sample. Changing the gradient slope made little difference to the separation i n terms o f resolution or the extent o f non-retention o f the solutes. A chromatogram using a gradient slope o f 0.93 % / m i n is shown for comparison i n Figure 5. Although the distances between the peaks are increased, so are the bandwidths. This is additional evidence for the small effect o f gradient slope on the preparative separation, and indicated that the non-retention o f the solutes was not a function o f gradient slope. A series o f experiments were carried out i n which the solutes were injected individually at a load o f 10 mg for a variety o f gradients (14). Calculation o f the quantities o f solute split between the retained and non-retained peaks are shown i n Table III, along with the starting concentration o f acetonitrile and the gradient slope.

Table III. Ratio of solutes retained and unretained under various loading conditions Gradient initial slope %B 31 31 31 24 17 10

2.8 1.4 0.9 2.8 2.8 2.8

Unretained Cyto Lys

0.8 0.8 0.8 0.3 0.0 0.0

0.34 0.34 0.34 0.06 0.0 0.0

Retained Cyto Lys

0.2 0.2 0.2 0.7 1.0 1.0

0.86 0.86 0.86 0.94 1.0 1.0

Sample Load: 10 m g . (Cyto = cytochrome c; L y s = lysozyme). Other conditions as Table I .

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

12.

COX

Preparative Reversed-Phase Chromatography

of Proteins

173

Downloaded by PENNSYLVANIA STATE UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch012

2.0 H

Figure 4. Chromatogram o f a mixture o f cytochrome c (3 mg) and lysozyme (10 mg). Solid line: 290 n m , dashed line: 480 nm, dotted line: analytical load (0.05 mg each) at 230 n m . Gradient: 31 to 45% acetonitrile i n 0.1 % T F A i n 5 minutes; flow rate: 1 m l / m i n . C o l u m n : Zorbax Pro-10 Protein Plus, 15 c m χ 4.6 m m . [Figure reproduced with permission from ref 14. Copyright 1992 Elsevier Scientific Publishers.]

2.0H

AU

1 0

Time (min)

Figure 5. Chromatogram o f a mixture o f cytochrome c (3 mg) and lysozyme (10 mg). Solid line: 290 n m , dashed line: 480 n m . Conditions as Figure 4, except gradient duration = 20 min.

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Downloaded by PENNSYLVANIA STATE UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch012

174

CHROMATOGRAPHY IN BIOTECHNOLOGY

The quantity retained is much lower than the saturation capacity o f the column and thus some mechanism other than complete overloading o f the packing must be responsible for the incomplete retention o f the solute. It is believed (17) that this is due to the displacement o f adsorbed acetonitrile by the introduction o f the protein solutes. Since each molecule o f protein presumably displaces several molecules o f acetonitrile by virtue o f its larger chromatographic footprint, an introduction o f a relatively large load of protein w i l l significantly increase the local acetonitrile concentration. The retention o f a protein i n reversed-phase chromatography is very sensitive to small changes in solvent strength. This local increase i n acetonitrile concentration is enough partially to elute the protein i n the displaced solvent band as it moves through the column. A rough calculation suggests that approximately 10 moles o f acetonitrile are required to be displaced by one mole o f protein i n order for this effect to be observed. Reduction o f the initial solvent strength increases the proportion o f protein which is adsorbed. Once the solutes are introduced to the column at a low enough acetonitrile concentration that the displaced solvent no longer prematurely elutes them, then any gradient may be subsequently used. Thus, a step gradient may be used to bring the mobile phase strength to the level required for the beginning o f the elution gradient, followed by a short, rapid gradient designed to elute the solutes. Figure 6 shows the result o f such a process, allowing a rapid elution after loading at 10% acetonitrile. A s seen i n the other separations, the displacements continue to be a major feature o f the separation. Effect o f p o r e s i z e : Loadability i n the purification o f macromolecules is dependent upon the surface area available for sorption. This in turn is a function o f the fraction of the surface area which lies i n pores o f adequate dimensions to allow access o f the solute molecules. Thus, the pore diameter (and the pore size distribution) is important in determining how much sample can be loaded. Studies i n ion-exchange chromatography by frontal elution (18) have indicated a relation between the kinetics o f uptake o f solutes with the pore diameter and the particle size. The saturation capacities o f several reversed-phase packings with different pore diameters (and hence surface areas) were determined by injection o f aliquots o f a standard solution o f lysozyme and measurement o f the quantity of the solute taken up by the column (19). F r o m the known values o f the specific surface areas o f the materials, the extent o f surface utilization (in terms o f m g / m ) could be calculated. These data, shown i n Table I V , show a steady increase i n surface utilization with pore diameter. M o r e importantly for preparative chromatography, there is a maximum i n saturation capacity at an intermediate pore diameter. It should be noted that the surface utilization o f two packings with similar pore diameters but markedly differing surface areas are identical. The results demonstrate that the lysozyme molecules cannot explore the entire surface area o f the packing, even when the nominal pore diameters are considerably larger than the radius o f gyration o f the solute. This is presumably a function o f the pore size distribution, which is an important parameter in the preparative chromatogrpahy o f proteins. The maximum i n the saturation capacity occurs at pore diameters o f around 120 Â pore diameter and this value can be taken as being close to the optimum for preparative chromatography o f molecules 2

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Downloaded by PENNSYLVANIA STATE UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch012

12.

COX

Preparative Reversed-Phase Chromatography of Proteins

175

Figure 6. Chromatogram o f a mixture o f cytochrome c (3 mg) and lysozyme (10 mg). Solid line: 290 nm, dashed line: 480 n m . Conditions as Figure 4, except gradient: hold at 10% acetonitrile for 2 m i n ; step to 31 % (in 1 min); 31 to 45 % acetonitrile i n 5 min.

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

176

CHROMATOGRAPHY IN BIOTECHNOLOGY

of this and similar size. It is important to note that this pore diameter is much smaller than the 300 Â conventionally recommended for peptide and protein separations. The reduction i n surface area i n using large pore diameter packings is often more important than the small improvement in solute mass transfer between the phases that is given by the large pore diameters.

Table I V .

Downloaded by PENNSYLVANIA STATE UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch012

Surface utilization as a function of pore size Pore Diameter (Â)

70 130 120 300 500 1000

Surface Area (mVg) 300 160 98 50 30 20

Saturation Capacity (mg) 70 112 70 50 34 20

Surface Utilization [(mg/m )/g] 2

0.13 0.39 0.39 0.55 0.63 0.69

M o b i l e phase: 5% acetonitrile i n 0.1 % aqueous trifluoroacetic acid.

Effect of particle diameter: The width o f the overlap zone between two adjacent solutes i n the mass overloaded chromatograms o f proteins appears to be a function o f both the gradient parameters and the column efficiency. Experiments to vary the column efficiency while maintaining other separation parameters constant were carried out (20) using a family o f C 250 Â pore packing materials with a range o f particle diameters between 6.5 and 50 μτη. The separation conditions were as described i n Table V . Chromatograms o f the separation o f a mixture o f cytochrome c and lysozyme on the 10, 20 and 50 μπι packings are shown i n Figure 7. The insets in the figures show the peak due to cytochrome c (i.e. 490 nm) on an expanded scale. These show the effects o f band spreading on the displacement as the efficiency is decreased. Comparison o f the chromatograms indicates that as the particle size is increased from 10 through 20 to 50 μηι the band profile is progressively rounded until, at a particle diameter o f 50 /un, the distortion i n band shape due to the displacement is difficult to discern. 8

The separations on particles o f 15 ^ m and below were all closely similar [20], and for this separation it is clear that there is no need for the use o f extremely high efficiencies. This is evident when the recovery o f cytochrome c at a purity o f 99.5% is calculated. These results are shown i n Table V I . The recovery is constant for particles o f 15 μτη and below. It begins to fall when a particle diameter between 15 and 20 μπι is used; the recovery from 20 μτη particles is reduced by approximately

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

COX

Preparative Reversed-Phase Chromatography of Proteins

177

Downloaded by PENNSYLVANIA STATE UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch012

12.

Figure 7. Chromatograms o f cytochrome c (3 mg) and lysozyme (10 mg). Solid line: 300 n m , dashed line, 490 nm. Inset, expanded 490 nm chromatogram. Gradient 0 to 60 % acetonitrile i n 0.1 % T F A , duration: 15 m i n . F l o w Rate: 1 m l / m i n . C o l u m n : Matrex C 8 . Particle size: (a) 10 μ π ι . (b) 20 μΐη. (c) 50 ^ m .

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

CHROMATOGRAPHY IN BIOTECHNOLOGY

178

2 to 3 % . Use of 50 μπι particles is clearly disavantageous for the separation because of the l o w (75%) recovery obtained due to the extensive overlap o f the two peaks.

Table V . Chromatographic conditions for particle size study

Downloaded by PENNSYLVANIA STATE UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch012

Columns: M o b i l e phase: Gradient:

F l o w Rate: Detection: Sample:

Matrex 250 C 1 8 , 6.5, 10, 15, 20, 50 μ π ι . 100 χ 4.6 m m Solvents A and B , as Table 1. Range: 0 to 60% B . Duration: 15 minutes. lml/min U V , 290 and 475 n m . Cytochrome c: 3 mg; Lysozyme: 10 mg.

Table V I . The effect of particle diameter on recovery of cytochrome c Particle Size (μπι)

6.5

10

15

20

50

Recovery (%)

97

96

97

94

75

Conditions as Table V .

In order to achieve a high recovery i n the present separation one can use a particle size o f 15 μηι i n a 10 c m long column. Other configurations can, however, be adopted. Larger particles in a longer column could be employed to obtain the same efficiency. In this case caution must be exercised i n the scale-up due to the different column configuration. Use o f larger particles i n the same size column but using a lower flow rate can also give the same efficiency and thus the same separation. Figure 8 shows the chromatograms arising from the use o f 20 μηι particles at a flow rate o f 0.3 m l / m i n (20) with a gradient duration o f 45 minutes. This results i n a gradient o f equal slope (in terms o f % change per ml) to those i n Figure 7 and any difference between the chromatograms should be attributable to the higher efficiency due to the lower flow rate. Under these conditions, the column efficiency is close to that obtained with smaller (10 to 15 μπι diameter) particles. It is not surprising then, that the peak shape and recovery / purity data found i n this separation are similar to those obtained with the smaller particles at higher flow rates, although the production rate is much reduced.

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Downloaded by PENNSYLVANIA STATE UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch012

12.

COX

Preparative Reversed-Phase Chromatography

of Proteins

179

Figure 8. Chromatograms o f cytochrome c (3 mg) and lysozyme (10 mg). Solid line: 300 n m , dashed line, 490 n m . Inset, expanded 490 nm chromatogram. Conditions as Figure 7, except flow rate: 0.33 m l / m i n , gradient duration: 45 min.

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

180

CHROMATOGRAPHY IN BIOTECHNOLOGY

Downloaded by PENNSYLVANIA STATE UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch012

Another alternative is to use smaller particles. This w i l l result i n a column o f higher efficiency, when operated under the same conditions. Since the efficiency o f a column is reduced by an increase i n the flow rate, it is possible to select a flow which w i l l give the required efficiency for the protein separation. If the gradient time is adjusted to maintain a constant slope (in volume terms), the separation can then be carried out more quickly, with a corresponding increase i n production rate. Because the operating pressure is a function o f flow rate and particle size, there is obviously a limit to this strategy at which the pressure reaches the maximum allowed for the chromatographic equipment. This point w i l l correspond to the highest rate o f production. Guidelines for the design of preparative elution gradients: Sample Introduction: Protein solutes should be introduced to the column using low solvent strengths. This avoids the problems o f premature elution o f the solute by desorption o f acetonitrile (or other organic modifier). It also allows the concentration of the sample at the head o f the column, eliminating injection volume effects and allowing the injection o f somewhat dilute samples. Use o f too l o w a solvent strength should also be avoided since other work (27) has shown that i f the packing is not wetted properly by the mobile phase the protein may not be adsorbed on the column surface and w i l l again elute prematurely. Gradient Slope: Since the displacements ensure that the recovery and purity o f the solutes are independent o f gradient slope, the gradients should be as steep and short as is practical i n order to maximize the production rate. This is not true i f displacements are not observed, and it is probable i n such instances that too short and sharp a gradient w i l l result i n loss o f resolution and, i n consequence, poor purity and recovery figures. The step between loading and elution o f the sample should be short to maximise production rate. A step gradient from the loading conditions to the beginning o f the elution gradient is often the best choice. The starting point o f the elution gradient should be close to (but preferably a little below) the concentration at which the solute becomes chromatographically mobile i n order to minimize the run time. A s soon as the component o f interest is eluted, the regeneration procedure should be started i n order to minimize the cycle time. Pore Size: The pore diameter o f the packing is important i n maximizing the available surface area and therefore the column capacity for the solute. The optimum pore diameter w i l l depend upon the conformation and molecular size o f the protein, but for the range o f molecular weights between 2000 and 20000, pore diameters i n the range 100 to 200 Â are generally appropriate. In order for the solutes to maximize their use of the surface area, the pore size distribution should be as narrow as possible about the mean value. Particle Diameter: The particle diameter used depends upon a number o f factors. For many separations, a limited efficiency o f 1000 to 2000 plates, measured for the components o f interest, is sufficient to ensure narrow displacement zones and good

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

12.

COX

Preparative Reversed-Phase Chromatography of Proteins

181

recovery and purity o f products. A t conventional flow rates, particles o f 15 to 2 0 μ π ι are appropriate for this range o f efficiency. I f pressure capabilities allow, use o f smaller particles i n the columns w i l l allow a higher flow rate. This means that a higher rate o f production o f product can be obtained.

Downloaded by PENNSYLVANIA STATE UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch012

Conclusions. Recent work on the preparative H P L C o f proteins has demonstrated that many separations are controlled by displacement effects which occur between solutes adjacent i n the chromatogram. These displacements are very sharp and allow recovery o f products i n high purity and yield. The gradient slope and range do not affect the displacements but can be optimized to maximize the rate o f production o f the product. The conditions o f loading the solute into the column are important, since poor selection o f solvent strength at this point can limit the load which may be applied to the column. The parameters relating to the packing material which are o f importance are the pore diameter and particle size. There is an optimum pore size which allows the maximum sample load. This is often well below the 300 Â recommended for analytical separations. The particle size is important i n that it controls the column efficiency. There is a minimum efficiency below which the displacements are significantly degraded. Thus, there is a minimum particle size which is appropriate and preparative separations should be performed under conditions designed to give the required number o f plates. Smaller particles at high flow rates can be used to provide a higher separation speed to enhance production rate, provided there is sufficient operating pressure available. Literature Cited.

1.

Regnier, F.E.; Gooding, K.M. Anal Biochem., 1983, 103 1.

2.

O'Hare, M.J.; Nice, E.C. J Chromatogr., 1979, 171, 209.

3.

Unger, K.K.; Jilge, G.; Kinkel, J. N.; Hearn, M.T.W. J Chromatogr., 1986, 359 (1986) 61.

4.

Lau, S.Y.M.; Taneja, A.K. and Hodges, R. S. J Chromatogr., 1984, 317 140.

5.

Benedek, K.; Dong, S.; Karger, B. L. J Chromatogr., 1984,317227

6.

Snyder, L. R.; Cox, G. B. and Antle, P. E. Chromatographia, 1987, 24, 82.

7.

Guiochon, G.; Ghodbane, S. J Phys Chem., 1988, 92, 3682.

8.

Cox, G. B.; Snyder,L. R. J Chromatogr., 1989, 483, 95.

9.

Eble, J. E.; Grob, R. E.; Antle, P. E.; Snyder, L. R. J. Chromatogr., 1987, 405, 51.

10.

Snyder, L. R.; Cox, G. B.; Antle, P. E. J. Chromatogr., 1988, 444, 303.

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Downloaded by PENNSYLVANIA STATE UNIV on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ch012

182

CHROMATOGRAPHY IN BIOTECHNOLOGY

11.

Cox, G. B.; Antle, P. E.; Snyder, L. R. J. Chromatogr., 1988, 444, 1325.

12.

Cox, G. B. J. Chromatogr., submitted for publication.

13.

Snyder, L. R.; Cox, G. B.; Dolan, J. W. J Chromatogr., 1989 484 437.

14.

Cox, G. B.; Snyder, L. R. J Chromatogr, in press.

15.

Cox, G. B.; Antle, P. E.; Snyder, L. R. J. Chromatogr., 1988, 444, 325.

16.

Snyder, L. R.; Stadalius, M. A. in Horváth, Cs., Ed.; High-Performance Liquid Chromatography - Advances and Perspectives, Vol 4, Academic Press, New York, 1986, p 915.

17.

Lee, Α., personal communication.

18.

Kopaciewicz, W.; Fulton S.; Lee, S. Y. J. Chromatogr., 1987, 409, 111.

19.

Cox, G. B.; Snyder, L. R.; Dolan, J. W. J. Chromatogr., 1989 ,484, 409.

20.

Cox, G. B., paper in preparation.

21.

Cox, G. B., unpublished data.

RECEIVED

January 26, 1993

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Downloaded by 89.163.34.136 on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ix001

Author Index Blackwell, J. Α., 146 Brooks, Clayton Α., 27 Burke, T. W. L., 59 Carr, P. W., 146 Cox, Geoffrey B., 165 Cramer, Steven M , 27 Frenz, John, 1 Hodges, R. S., 59 Jacobson, Jana, 77 Kalisz, Henryk M . , 102 Mant, C. T., 59 Mendonca, A . J., 59

Miller, William J., 132 Nadler, T. K., 14 Prouty, Walter F., 43 Régnier, F. Ε., 14 Rigney, M . P., 146 Roos, P. H., 112 Schafer, W. Α., 146 Schmid, Rolf D., 102 Townsend, R. Reid, 86 Weber, T. P., 146 Y u Ip, Charlotte C , 132

Affiliation Index Bio West Research, 77 Eli Lilly and Company, 43 Genentech, Inc., 1 Gesellschaft fur Biotechnologische Forschung, 102 Institute of Physiological Chemistry I, 112 Merck Sharp & Dohme Research Laboratories, 132

Prochrom Inc., 165 Purdue University, 14 Rensselaer Polytechnic Institute, 27 University of Alberta, 59 University of California—San Francisco, 86 University of Minnesota, 146

Subject Index Biotechnology challenges, 1 chromatographic separations, 2-10 commercial success of recombinant protein pharmaceuticals, 2t cost consciousness, 82-83 importance of chromatography, 83 Biotechnology industry, interest in continuous purification systems, 14

Acid hydrolysis, release of monosaccharides from glycoproteins, 96-97 Adsorption mechanism, zirconium oxide, 160-165 Affinity chromatography, secreted mammalian cell product, 44,48/,49 Amino sugars, separation, detection, and quantification, 101-103

Biochromatographic applications, zirconium oxide based supports, 152-169

Cation-exchange chromatography, Mono S, 126,127/,128

183

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Downloaded by 89.163.34.136 on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ix002

Author Index Blackwell, J. Α., 146 Brooks, Clayton Α., 27 Burke, T. W. L., 59 Carr, P. W., 146 Cox, Geoffrey B., 165 Cramer, Steven M , 27 Frenz, John, 1 Hodges, R. S., 59 Jacobson, Jana, 77 Kalisz, Henryk M . , 102 Mant, C. T., 59 Mendonca, A . J., 59

Miller, William J., 132 Nadler, T. K., 14 Prouty, Walter F., 43 Régnier, F. Ε., 14 Rigney, M . P., 146 Roos, P. H., 112 Schafer, W. Α., 146 Schmid, Rolf D., 102 Townsend, R. Reid, 86 Weber, T. P., 146 Y u Ip, Charlotte C , 132

Affiliation Index Bio West Research, 77 Eli Lilly and Company, 43 Genentech, Inc., 1 Gesellschaft fur Biotechnologische Forschung, 102 Institute of Physiological Chemistry I, 112 Merck Sharp & Dohme Research Laboratories, 132

Prochrom Inc., 165 Purdue University, 14 Rensselaer Polytechnic Institute, 27 University of Alberta, 59 University of California—San Francisco, 86 University of Minnesota, 146

Subject Index Biotechnology challenges, 1 chromatographic separations, 2-10 commercial success of recombinant protein pharmaceuticals, 2t cost consciousness, 82-83 importance of chromatography, 83 Biotechnology industry, interest in continuous purification systems, 14

Acid hydrolysis, release of monosaccharides from glycoproteins, 96-97 Adsorption mechanism, zirconium oxide, 160-165 Affinity chromatography, secreted mammalian cell product, 44,48/,49 Amino sugars, separation, detection, and quantification, 101-103

Biochromatographic applications, zirconium oxide based supports, 152-169

Cation-exchange chromatography, Mono S, 126,127/,128

183

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

CHROMATOGRAPHY IN BIOTECHNOLOGY

Downloaded by 89.163.34.136 on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ix002

184

Chemical stability characteristic of ideal support, 152-153,154* zirconium oxide supports, 155 Chromatographic concentration control, displacement, 82-88 Chromatographic properties, zirconium oxide supports, 153,154/,155,156/ Chromatographic resins, properties, 49-53 Chromatographic separations in biotechnology large-scale operations, 2-6 protein analysis by H P L C , 6-10 Chromatographic support, optimal characteristics, 152-153,154/ Chromatography, advantages and requirements, 83 Concentration control for chromatography, displacement, 82-88 Continuous chromatography advantages, 14 continuously stirred tank reactors, 15-16 Continuous purification of proteins, S N A P chromatography, 14-24 Continuous selective nonadsorption preparative chromatography columns, 17,18/,19 dilutions, 19 dual-column system, 20-21,23/ ilow rates, 20,2 It optimization, 19-20,22,24 packing materials, 19 selective elution, 22,23/ Continuously stirred tank reactors, description, 15-16 Countercurrent flow separations, description, 15 Cross-current flow separations, description, 15

D Diethylaminoethyldextran displacement, effluent profiles, 35,36-37/ Displacement, description, 5/6 Displacement chromatography advantages, 27-28,83,85 chromatographic equipment, 83-84

Displacement chromatognphy-C ontinued differences from elution chromatography, 83 fraction analytical procedure, 84 histograms of separations, 85,88/ human insulin, 50,52/53 luteinizing hormone releasing hormone, 85,87/ operating steps, 84 operating steps for gradient elution, 84

Ε Electrochemical detection, detection method for H P L C , 8 Enzymatic hydrolysis, release of monosaccharides from glycoproteins, 97 Enzymes from microbial sources, purification steps, 4t Equilibrium parameters moderately retained proteins, 31-32 strongly retained displacers, 32-33 Escherichia coli derived recombinant product, ion-exchange chromatography, 50,51/

F Fast protein chromatography, ion exchange, microsomal cytochrome P-450 pattern analysis, 120-128

G Glucose oxidase commercial importance, 108 glucose oxidation mechanism, 108 structure vs. fungal source, 108-109 Glucose oxidase isozyme separation from Pénicillium amagasakiense by ion-exchange chromatography compositional analysis determination procedure, 110 crystallization, 110 electrophoretic procedure, 109 electrophoretic titration curve before and after purification, 110,113/

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Downloaded by 89.163.34.136 on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ix002

INDEX

185

Glucose oxidase isozyme separation from Pénicillium amagasakiense by ion-exchange chromatography-Cowi/nw^ enzyme assay procedure, 109 gradient slope vs. resolution, 110,112/ optimal resolution conditions, 114,115/ pH effect, 110,111/ properties of homogeneous vs. heterogeneous isoforms, 114r purification procedure, 109-110 salt effect, 110,111/ Glycoproteins characterization of carbohydrate structures on surfaces by H P L C , 9,10/ quantitative monosaccharide analysis by H P L C , 92-105 Gradient elution, description, 5/,6 Gradient elution chromatography advantages, 84-85 histograms of separations, 85,88/ luteinizing hormone releasing hormone, 85,86/ Gradient range, preparative reversed-phase chromatographic effect, 178i,179/,180,182/ Gradient slope, preparative reversed-phase chromatographic effect, 174,177i

Haemophilus influenzae type b importance of vaccine development, 138 monosaccharide compositional analysis of conjugate vaccine, 139-147 High-performance ion-exchange chromatography, use for protein purification, 27 High-performance liquid chromatography protein analysis, 6-10 quantitative monosaccharide analysis of glycoproteins, 92-105 Human insulin displacement chromatography, 50,52/53 H P L C profile of purified product, 44,45/ Human therapeutic pharmaceuticals, developments for production processes, 1

Ideal displacement profiles, calculation, 33-34

Immobilized metal affinity chromatography description, 128 microsomal cytochrome P-450 pattern analysis, 128-134 Ion-exchange chromatography Escherichia coli derived recombinant product, 50,51/ secreted mammalian cell product, 44,47/49 separation of glucose oxidase isozymes from Pénicillium amagasakiense, 108-115 Ion-exchange displacement chromatography of proteins adsorption isotherms, 35,36/ apparatus, 35 effluent profile of DEAE-dextran, 37/ effluent profile of β-lactoglobulins, 37,38/39 effluent profile of protamine, 39/ equilibrium parameter determination for moderately retained proteins, 31-32 equilibrium parameter determination for strongly retained displacers, 32-33 ideal displacement profile calculation, 33-34 multicomponent equilibrium theory, 30-31 salt effect, 27 simulated effluent profile of DEAE-dextran, 35,36/ simulated effluent profile of β-lactoglobulins, 37,38/39 simulated effluent profile of protamine, 39,40/ steric mass-action ion-exchange equilibrium theory, 28,29/30 Ion-exchange fast protein liquid chromatography, microsomal cytochrome P-450 pattern analysis, 120-128 Isocratic elution, description, 5/

β-Lactoglobulin displacement, effluent profiles, 37,38/39 Large-scale chromatographic purification of pharmaceuticals, 2-5 Lewis base modified zirconia capacity factor vs. isoelectric point, 164,166/

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

Downloaded by 89.163.34.136 on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ix002

186

CHROMATOGRAPHY IN BIOTECHNOLOGY

Lewis base modified zircoma-Continued protein chromatography, 164,166-169 protein loading, 167,168/,169 protein retention, 164,167* protein separation, 164,166/ Light-scattering detection, method for H P L C , 8-9 Liquid chromatographic separation of microsomal cytochromes P-450, techniques, 118 Luteinizing hormone releasing hormone displacement chromatography, 85,87/ gradient elution chromatography, 85,86/ purification by preparative reversed-phase sample displacement chromatography, 66-67,70-80 M Media of different pore and particle diameters, characteristic of ideal support, 153,154* Metal affinity chromatography, immobilized, microsomal cytochrome P-450 pattern analysis, 128-134 Methanolysis, release of monosaccharides from glycoproteins, 93,95/96 Microsomal cytochrome P-450 pattern analysis with fast protein liquid chromatography cation-exchange chromatography on M o n o S , 126,127/128 detergent selection, 119 gradient shape vs. resolution, 121,122-123/ inducer effects, 121,123-125/126* isozyme identification, 121,126 optimization of resolution, 120*,121 sample load vs. resolution, 121*, 124/ sample preparation for analytical fractionation, 119-120 undesired effects of sample pretreatment, 119-120 Microsomal cytochrome P-450 pattern analysis with immobilized metal affinity chromatography behavior of separated proteins, 134 buffer selection, 129 chromatographic protocol, 128 detergent vs. gradient shape, 129,130*,131-132/

Microsomal cytochrome P-450 pattern analysis with immobilized metal affinity chromdtogrzphy-C ontinued eluting component selection, 129 isozyme identification, 130,133/134* metal ion selection, 128,129* Moderately retained proteins, equilibrium parameter determination, 31-32 Mono S, cation-exchange chromatography, 126,127/128 Monosaccharide(s) quantitative analysis of glycoproteins with H P L C , 92-105 structures, 93,94/ Monosaccharide compositional analysis of Haemophilus influenzae type b conjugate vaccine immunoelectron micrograph of vaccine, 139,140/ outer membrane protein complex, 139,141-145 polyribosylribitol phosphate, 145,146/ polyribosylribitol phosphate-outer membrane protein complex conjugate vaccine, 145,147/ Monosaccharide derivatives postcolumn derivatization, 100 precolumn derivatization, 97-100 Moving belt separation, description, 15 Multicomponent equilibrium, theory, 30-31

Ν Neutral sugars, separation, detection, and quantification, 101-103

Ο On-line mass spectrometry, detection method for H P L C , 9 Operating modes, large-scale chromatographic purification of pharmaceuticals, 5/ Optimum recovery, determination, 20 Outer membrane protein complex, monosaccharide compositional analysis, 139,141-145

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

187

Downloaded by 89.163.34.136 on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ix002

INDEX

Packing material, large-scale chromatographic purification of pharmaceuticals, 3-4 Particle diameter, preparative reversed-phase chromatographic effect, 181-186 Pénicillium amagasakiense, separation of glucose oxidase by ion-exchange chromatography, 108-115 Peptide(s), preparative reversed-phase sample displacement chromatography, 59-80 Peptide mapping, use of H P L C , 9 pH stability, zirconium oxide supports, 155,156/ Polyribosylribitol phosphate monosaccharide compositional analysis, 145,146/ use in Hib vaccine, 139 Polyribosylribitol phosphate-outer membrane protein complex conjugate vaccine, monosaccharide compositional analysis, 145,147/ Pore size, preparative reversed-phase chromatographic effect, 180,181f Postcolumn derivatization, monosaccharide derivatives, 100 Precolumn derivatization, monosaccharide derivatives, 97-100 Preparative elution gradient design guidelines, 186-187 Preparative reversed-phase chromatography of proteins applications, 172 elution gradient design guidelines, 186-187 gradient range effect, 178i,179/,180,182/ gradient slope effect, 174,177r model compound selection, 173 modeling, 172-173 particle diameter effect, 181-186 pore size effect, 180,1811 protein-protein displacement procedure, 173-174,175-176/ theory, 172-173

Preparative reversed-phase sample displacement chromatography for purification of pharmaceutically important peptides comparison of approaches, 77,79i,80 low levels of organic modifier effect, 74-78 peptide elution in absence of organic modifier, 67,72-73/74 peptide studied, 66-67 standard multicolumn approach, 67,70-71/ Process chromatography in production of recombinant products differentiation of early from late stages in purification process, 49,5It environmental impact of wastes, 57 process capacity for each step, 53,55/57 process control chart for manufacturing step, 53,54/ product source effect on product design, 43-49 properties of chromatographic resins, 49-53 worthwhile process improvements, 53,54-56/57 yield increase vs. step, 56/57 Proinsulin, H P L C profile of purified product, 44,45/ Properties of chromatographic resins elution mode, 50,51r/,52/,53 uniform size distribution of particles, 49-50 Protamine displacement, effluent profiles, 39-40/ Protein(s) continuous purification by S N A P chromatography, 14-24 ion-exchange displacement chromatography, 27-40 preparative reversed-phase chromatography, 171-187 Protein analysis by high-performance liquid chromatography, 6-10 Protein chromatography, Lewis base modified zirconia, 164,166-169 Protein loading, Lewis base modified zirconia, 167,168/169 Protein pharmaceutical industry, challenges, 1 Protein product source, process design effect, 43-49

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

CHROMATOGRAPHY IN BIOTECHNOLOGY

188

Downloaded by 89.163.34.136 on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ix002

Protein retention Lewis base modified zirconia, 164,167* zirconium oxide, 157,159/ Protein separation Lewis base modified zirconia, 164,166/ zirconium oxide, 157,159/160,161/ zirconium oxide supports, 155,156/ Purification of proteins, continuous, S N A P chromatography, 14-24 Purification methods, influencing factors, 60 Purification process, differentiation of early from late stages, 49,51/

Quantitative monosaccharide analysis of glycoproteins by high-performance liquid chromatography acid hydrolysis for monosaccharide release, 96-97 comparison of reported monosaccharide compositions, 104,105* detection of neutral and amino sugars, 101-103 detection of sialic acids, 103-104 enzymatic hydrolysis for monosaccharide release, 97 methanolysis for monosaccharide release, 93,95/96 monosaccharide structures, 93,94-95/ postcolumn derivatization of monosaccharide derivatives, 100 precolumn derivatization of monosaccharide derivatives, 97-100 quantification of neutral and amino sugars, 101-103 quantification of sialic acids, 103-104 separation of neutral and amino sugars, 101-103 separation of sialic acids, 103-104 R Recombinant D N A based production methods origin and scale, 1 Recombinant D N A technology, role in development of biotechnology industry, 2

Recombinant products, role of process chromatography in production, 43-57 Recombinant protein pharmaceuticals, large-scale chromatographic operations, 2-6 Resins chromatographic, properties, 49-53 selectivity effect on protein purification, 5 Reversed-phase chromatography advantages, 84 applications, 84,171-172 Reversed-phase chromatography in gradient elution mode, disadvantages for peptide purification, 59-60

Salt, protein purification effect, 22,24 Sample displacement chromatography application to purification of pharmaceutically imported peptides, 66-67,70-80 development of multicolumn approach, 61,66,68-69/ displacement of desired hydrophilic peptide component by hydrophobic impurities, 61,64-65/ displacement of hydrophilic impurities by desired hydrophobic peptide component, 61 principles, 60-61,62-65/ Secreted mammalian cell product affinity chromatography, 44,48/49 ion-exchange chromatography, 44,47/49 Secreted protein products, heterogeneity, 44,46/ Selective nonadsorption preparative chromatography adsorption to anion-exchange sorbent above isoelectric point, 16-17,18/ continuous system ,17,18/19 dual-column system, 20-21,23/ flow rates, 20,21* optimization, 19-20,22,24 selective elution, 22,23/ separation methods, 17

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.

189

Downloaded by 89.163.34.136 on August 2, 2012 | http://pubs.acs.org Publication Date: May 24, 1993 | doi: 10.1021/bk-1993-0529.ix002

INDEX Sialic acids separation, detection, and quantification, 103-104 structures, 93,95/ Sodium fluoride, protein retention effect, 164,167* Source of protein product, process design effect, 43-49 Step elution, description, 5/6 Steric mass-action ion-exchange equilibrium protein multipoint attachment and steric hindrance, 28,2Sy theory, 28-30 Strongly retained displacers, equilibrium parameter determination, 32-33 Surface site types, zirconium oxide supports, 153,154/155 Surface species models, zirconium oxide, 157,158/ Synthetic peptides disadvantages of purification by reversed-phase chromatography in gradient elution mode, 59-60 factors affecting purification, 60 preparative reversed-phase sample displacement chromatography, 59-80 requirements for purification methods, 59

Τ Therapeutic protein products, purification process effect on quality and safety, 43-49

U Ultraviolet absorbance, detection method for H P L C , 8

Ζ

Zirconium oxide adsorption mechanism, capacity factors vs. ρΚ , 160,162-165 Zirconium oxide supports chemical stability, 155 chromatographic properties, 153,154/155,156/ pH stability, 155,156/ phosphate addition, 155,157-161 protein separation, 155,156/ surface site types, 153,154/155 Λ

In Chromatography in Biotechnology; Horváth, C., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1993.