UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Cell separations by immunoaffinity partition Stocks, Susan Jill 1989

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1989_A1 S76.pdf [ 10.76MB ]
Metadata
JSON: 831-1.0059445.json
JSON-LD: 831-1.0059445-ld.json
RDF/XML (Pretty): 831-1.0059445-rdf.xml
RDF/JSON: 831-1.0059445-rdf.json
Turtle: 831-1.0059445-turtle.txt
N-Triples: 831-1.0059445-rdf-ntriples.txt
Original Record: 831-1.0059445-source.json
Full Text
831-1.0059445-fulltext.txt
Citation
831-1.0059445.ris

Full Text

CELL SEPARATIONS BY IMMUNOAFFINITY PARTITION By SUSAN JILL STOCKS B.Sc, Reading University, England, 1982 M.Sc, The Univ e r s i t y of B r i t i s h Columbia, 1986 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES DEPARTMENT OF CHEMISTRY We accept t h i s thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA SepteWt>er,^' 1 9 8 9 © Susan J i l l Stocks ? 1989 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of LWw\x*>V The University of British Columbia Vancouver, Canada Date OcX . iau DE-6 (2/88) ABSTRACT The p a r t i t i o n of c e l l s i n an aqueous polymer two-phase system composed of dextran T500, polyethylene g l y c o l 8000 and buffer was studied. The e f f e c t of various immunoaffinity ligands on erythrocytes and lymphocytes was examined. Separations of rabbit and human erythrocytes were achieved using a combination of monoclonal mouse anti-NN glycophorin IgG with trypan blue-derivatized sheep anti-mouse F c fragment IgG as well as with a polyacrylamide-derivatized rabbit anti-human erythrocyte IgG (PAA-rochrbc). The PAA-rahrbc was able to completely resolve the two erythrocyte species i n a countercurrent d i s t r i b u t i o n (CCD) of 20 transfers. The lymphocyte separation problem was of two sub-lines of a transformed mouse lymphocyte, MBL-2(4.1) and MBL-2(2.6), which d i f f e r i n the surface d e n s i t i e s of an antigen recognized by a rat monoclonal IgG, YE1.48.10. Binding studies showed that at saturation MBL-2(4.1) bound 2.4 x 10 6 molecules of YE1.48.10. per c e l l whereas MBL-2(2.6) bound 8 x 10 5 molecules of YE1.48.10. per c e l l . This represented an extremely stringent separation problem compared to previous immunoaffinity erythrocyte separations. Attempts to separate the c e l l s by immunoaffinity p a r t i t i o n using the following combinations of ligands were not s u f f i c i e n t l y successful to achieve useful separations: (a.) PEG 1900-derivatized YE1.48.10. (b.) PEG 1900-derivatized RG7/11.1, a mouse monoclonal IgG s p e c i f i c f o r rat F c fragment, and YE1.48.10. (c. ) b i o t i n - d e r i v a t i z e d YE1.48.10. and PEG-derivatized avidin. However polyacrylamide grafted onto YE1.48.10. produced an e f f e c t i v e immunoaffinity ligand, separating MBL-2C4.1) and MBL-2(2.6) on the basis of t h e i r antigenic d i f f e r e n c e s . A separation was achieved i n 60 CCD transfers. i i Table of Contents Page Abstract. i i Contents. i i i L i s t of Tables. vi L i s t of f i g u r e s . vii Acknowledgements. x Dedication xi Chapter 1. Methods of C e l l Separation. 1 1. Sedimentation. 1 2. Single c e l l s o r t i n g and multiparameter analysis. 3 3. Magnetic separation. 4 4. S o l i d phase a f f i n i t y f r a c t i o n a t i o n . 6 5. Electrophoresis. 7 6. B i o l o g i c a l separations. 8 7. P a r t i t i o n i n g i n aqueous polymer two-phase systems. 8 a. History of the development of p a r t i t i o n i n g i n aqueous polymer two-phase systems. 11 b. Recent uses of p a r t i t i o n i n g i n aqueous polymer two-phase systems. 12 i . Proteins, n u c l e i c acids and s u b - c e l l u l a r p a r t i c l e s . 12 i i . Mammalian c e l l s . 14 c. Immunoaffinity p a r t i t i o n . 16 i . Immunoaffinity p a r t i t i o n - an approach to bone marrow purging. 20 i i . Immunoaffinity p a r t i t i o n - an approach to f e t a l i s l e t c e l l separation. 23 8. An o u t l i n e of the model separation problem. 24 Chapter 2. Theoretical Aspects of P a r t i t i o n i n g . 26 1. Phase separation. 26 2. Molecular a f f i n i t y p a r t i t i o n . 29 3. P a r t i c l e p a r t i t i o n . 32 4. P a r t i c l e a f f i n i t y p a r t i t i o n . 34 5. The theory of counter current d i s t r i b u t i o n . 35 Chapter 3. Background to the Separation Problem. 39 1. Hi s t o r y of the c e l l l i n e used i n the separation. 39 2. The structure of immunoglobulin G. 41 3. The a v i d i n - b i o t i n i n t e r a c t i o n . 43 4. Polyacrylamide-protein g r a f t copolymers as ligands. 44 Chapter 4 . Materials and Methods. 46 1. Fragmentation of mouse IgG. 46 2. Production of sheep anti-mouse F c fragment IgG. 46 3. Monoclonal antibody production via a s c i t e s f l u i d . 47 i i i Page 4. Sodium dodecyl sulfate polyacrylamide gel electrophoresis. 47 5. Hemagglutination assay. 49 6. Radiolabelling of proteins with 1 2 5 I . 49 7. Trypan blue-derivatization of protein. 50 8. Synthesis of 2(alkoxypolyethyleneglycoxy)-4,6 dichlorotriazine (PCO 50 9. Determination of the hydrolyzable chlorides of PCC. 51 10. Reaction of PCC with protein. 51 11. Degree of modification of the protein with PEG. 52 12. Enzyme-linked immunosorbent assay. 52 13. Culture of hybridoma and transformed lymphocyte cel l s . 53 14. Supplemented serum-free medium (SSFM). 54 15. Culture of YE1.48.10. and RG7/11.1 cells in SSFM. 54 16. Harvesting and purifying the monoclonal antibodies. 54 17. Measurement of the binding isotherms of YE1.48.10. and RG7/11.1 antibodies to MBL-2 cells. 55 18. Fluorescent staining of cells. 56 19. Trypsin digestion of MBL-2 cells. 57 20. Preparation of two-phase systems. 57 21. Partitioning of cells and proteins. 58 22. Biotin-derivatization of protein. 58 23. Determination of avidin, biotin and degree of protein derivatization with biotin. 59 24. Polyacrylamide-derivatization of proteins. 60 25. Radiolabelling of MBL-2 cells. 61 26. 5 1Cr labelling of erythrocytes. 61 27. Counter current distribution of cells. 62 Chapter 5 . Results and Discussion. 63 1. Separation of erythrocytes using a trypan blue-derivatized second ligand. 63 2. Separation of MBL-2(4.1) and MBL-2(2.6) cells by immunoaffinity partition. 74 a. Production and purification of YE1.48.10. and RG7/11.1 monoclonal antibodies. 74 b. Characterization of the MBL-2 cells. 83 c. Analysis of the binding experiments. 84 d. Partitioning studies on MBL-2 cells. 93 i . The effect of phosphate and chloride ions. 93 i i . The effect of PEG-linoleate. 97 i i i . The effect of PEG-YE1.48.10. 98 iv. The effect of PEG-RG7/11.1 and YE1.48.10. 100 v. The effect of biotin-YEl.48.10. and PEG-avidin. 105 3. The effect of polyacrylamide-derivatized antibodies on erythrocyte and MBL-2 c e l l partition. 115 Chapter 6. SUMMARY. 130 iv Page Glossary of terms. 134 Glossary of symbols and abbreviations. 136 Appendices. 1. P a r t i a l d e r i v a t i v e s used to calculate errors i n v and v/L. 139 2. T h e o r e t i c a l CCD p r o f i l e s f o r the c e l l separations i n t h i s study. 142 Bibliography. 145 v L i s t of Tables Page 1. E x t i n c t i o n c o e f f i c i e n t s of 4-hydroxyazobenzene-2'-carboxylic a c i d and i t s complexes with avidin. 59 2. A comparison of the e f f e c t of PEG and Trypan Blue on BSA as a f f i n i t y ligand modifying agents. 64 3. Results of p a r t i t i o n i n g experiments using Trypan Blue-modified sheep anti-mouse F c as a second a f f i n i t y ligand. 68 4. The e f f e c t of Trypan Blue-sheep anti-mouse F on the p a r t i t i o n of other mouse IgGs. 73 5. Summary of the information obtained by extrapolation of the Scatchard pl o t s i n Figs. 25-27. 87 6. The e f f e c t of d i f f e r e n t phase systems and phosphate on the p a r t i t i o n of MBL-2 c e l l s . 94 7. The e f f e c t of PEG-YE1.48.10. on the p a r t i t i o n of MBL-2 c e l l s . 99 8. The p a r t i t i o n of MBL-2 c e l l s i n the presence of PEG-RG7/11.1 and YE1.48.10. antibodies. 104 9. The e f f e c t of b i o t i n - d e r i v a t i z a t i o n , PEG-avidin and av i d i n on the p a r t i t i o n of bovine serum albumin. 109 10. The e f f e c t of reaction r a t i o of N-hydroxysucclnimidoblotin on the degree of IgG modification with b i o t i n and the p a r t i t i o n c o e f f i c i e n t . 110 11. The r e s u l t s of p a r t i t i o n i n g experiments using biotin-YE1.48.10.and MBL-2 c e l l s . I l l 12. The e f f e c t of PEG-avidin and a v i d i n on the p a r t i t i o n c o e f f i c i e n t of biotin-YEl.48.10. i n the presence of MBL-2 c e l l s . 114 13. A summary of polyacrylamide-protein g r a f t copolymers synthesized with d e t a i l s of reaction r a t i o s , v i s c o s i t i e s and p a r t i t i o n c o e f f i c i e n t s . 117 14. Hemagglutination assays with polyacrylamide-derivatized ligands. 118 15. The p a r t i t i o n of human and rabbit erythrocytes i n the presence of polyacrylamide-derivatized rabbit anti-human erythrocyte IgG. 121 vi List of Figures Page 1. A t y p i c a l c e l l p a r t i t i o n i n g experiment. 10 2. Schematic diagrams of approaches to immunoaffinity 18 p a r t i t i o n of c e l l s . 3. General phase diagram f o r a PEG/dextran/water phase system. 28 4. A diagrammatic representation of f i v e CCD transfer steps. 36 5. Differences between the d i s t r i b u t i o n type i n l i q u i d - l i q u i d and l i q u i d - i n t e r f a c e CCD. 38 6. The structure of the IgG molecule. 42 7. The structure of b i o t i n . 44 8. The p a r t i t i o n of trypan blue dye. 65 9. V i s i b l e spectrum of trypan blue-derivatized and native sheep anti-mouse F c IgG. 66 10. A schematic diagram of the binding of trypan blue-derivatized sheep anti-mouse F c IgG to mouse anti-NN glycophorin which i s bound to an erythrocyte. 67 11. Fast protein l i q u i d chromatography (FPLC) p r o f i l e of the papain digest of mouse IgG. 69 12. Sodium dodecyl s u l f a t e polyacrylamide gel electrophoresis (SDS-PAGE) of mouse IgG and F c fragment. 70 13. FPLC p r o f i l e of sheep-anti mouse F c fragment preparation. 71 14. FPLC p r o f i l e of monoclonal mouse anti-NN glycophorin IgG preparation. 72 15. A schematic diagram of the binding of a MBL-2 c e l l by YE1.48.10. monoclonal antibody (MAb) and PEG-derivatized RG7/11.1 MAb. 75 16. Growth curves f o r YE1.48.10. and RG7/11.1 hybridomas i n supplemented serum-free medium (SSFM). 78 17. SDS-PAGE of hybridoma supernatants. 79 18. FPLC p r o f i l e of SSFM YE1.48.10. hybridoma culture supernatant. 80 19. FPLC p r o f i l e of SSFM RG7/11.1 hybridoma culture supernatant. 81 vi i page 20. SDS-PAGE of FPLC f r a c t i o n s of YE1.48.10. hybridoma cu l t u r e 82 supernatant. 21. SDS-PAGE of FPLC-purified RG7/11.1 and YE1.48.10. MAbs. 82 22. Fluorescein isothiocyanate (FITC)-labelled MBL-2 c e l l s . 85 23. Scatchard p l o t f o r independent, i d e n t i c a l binding s i t e s . 86 24. A d i r e c t p l o t of the binding of YE1.48.10. MAb to MBL-2 c e l l s . 89 25. A scatchard p l o t of the binding of YE1.48.10. MAb to MBL-2 c e l l s . 90 26. A scatchard p l o t of the binding of b i o t i n - d e r i v a t i z e d YE1.48.10. MAb to MBL-2 c e l l s . 91 27. A scatchard p l o t of the binding of YE1.48.10. Mab to human erythrocytes and mouse lymphocytes. 92 28. CCD p r o f i l e of MBL-2 c e l l s . 95 29. The e f f e c t of PEG-1inoleate on the p a r t i t i o n of MBL-2 c e l l s . 96 30. Binding curve of a t y p i c a l ELISA of YE1.48.10. 98 31. Scatchard p l o t f o r the binding of RG7/11.1 MAb to MBL-2 c e l l s . 101 32. A schematic diagram of an enzyme-linked immunosorbent assay (ELISA) and modified ELISA used to assay d e r i v a t i z e d antibodies. 102 33. Binding curves f o r native and PEG-derivatized RG7/11.1. 103 34. The e f f e c t of PEG-avidin and avidin on the p a r t i t i o n of b i o t i n . 106 35. The e f f e c t of PEG-derivatization of avidin on the binding of 2(4'-hydroxyazobenzene) benzoic acid (HABA). 107 36. The e f f e c t of b i o t i n on the p a r t i t i o n of HABA i n the presence of PEG-avidin. 108 37. The e f f e c t of PEG-avidin and avidin on the p a r t i t i o n of YE1.48.10. Mab. 112 38. The e f f e c t of a v i d i n and PEG-avidin on the p a r t i t i o n of MBL-2 c e l l s i n the presence of biotin-YEl.48.10. Mab. 113 39. The p a r t i t i o n of human and rabbit erythrocytes i n the presence of polyacrylamide (PAA)-derivatized rabbit a n t i -human erythrocyte IgG (rahrbc). 119 viii page 40. The e f f e c t of PAA-rochrbc concentration on the p a r t i t i o n of rabbit and human erythrocytes. 120 41. CCD p r o f i l e of rabbit and human erythrocytes i n the presence of PAA-YE1.48.10. 122 42. The e f f e c t of PAA-YE1.48.10. concentration on the p a r t i t i o n of MBL-2 c e l l s . 125 43. CCD p r o f i l e f o r MBL-2 c e l l s i n the presence of PAA-YE1.48.10. 126 44. CCD p r o f i l e f o r MBL-2 c e l l s i n the presence of PAA-YE1.48.10. (repeat of Fig. 43) 127 45. CCD p r o f i l e f o r MBL-2 c e l l s using MBL-2 c e l l s i n an e a r l y stage of growth and a d i f f e r e n t PAA-YE1.48.10. ligand. 128 46. CCD p r o f i l e f o r MBL-2 c e l l s using MBL-2 c e l l s i n an e a r l y stage of growth and a d i f f e r e n t PAA-YE1.48.10. ligand. (repeat of F i g . 45) 129 ix Acknowledgements I would l i k e to take t h i s opportunity to than everybody i n the lab f o r t h e i r invaluable advice, assistance and making e x i s t i n g i n an a i r l e s s , windowless dungeon bearable. In p a r t i c u l a r thanks to Raymond, Rosemarie, John and Nancy for t h e i r constant good humor and endless enthusiasm f o r p a r t i e s . Thank you to Johan for h i s help with the error a n a l y s i s . I am e s p e c i a l l y g r a t e f u l to Dave for h i s help with the error a n a l y s i s and a l l that followed, including the a l l night photocopying session and much more! Most of a l l thank you to Don Brooks f o r h i s advice, encouragement, suggestions and i n p a r t i c u l a r patience! Without Don t h i s study would not have been accomplished, nor would I have met somebody who I hope w i l l remain my f r i e n d . The f i n a n c i a l support of the Science and Engineering Research Committee, England i s recognized and appreciated as well as the support of the MRC and the use of f a c i l i t i e s and equipment i n the chemistry and pathology departments. To Alan, Jose and Lynn. Without whom t h i s would never have been accomplished. CHAPTER 1 METHODS OF CELL SEPARATION The present need f o r methods to separate, i d e n t i f y and characterize d i f f e r e n t types of c e l l s i s recognized amongst immunologists, hematologists, c e l l b i o l o g i s t s , c l i n i c a l pathologists and biomedical researchers. Unless c e l l s e x h i b i t i n g d i f f e r e n t functions or at d i f f e r e n t stages of d i f f e r e n t i a t i o n are separable i t w i l l be d i f f i c u l t to study some, of the mechanisms involved i n c e l l recognition, s p e c i a l i z a t i o n , c y t o t o x i c i t y and transformation. Some of the immediate p r a c t i c a l b enefits obtained by applying c e l l separation techniques are the c l i n i c a l diagnosis of disease and the use of i s o l a t e d c e l l s f o r therapeutic (e.g. immunotherapy) or s u r v i v a l (e.g. transfusion) purposes. A number of techniques for c e l l separation e x i s t based on the physical or b i o l o g i c a l properties of the c e l l s and a b r i e f d i s c u s s i o n of these methods follows: 1. Sedimentation. The technique of v e l o c i t y sedimentation was f i r s t described by Lindahl (1948) who termed i t counter streaming c e n t r i f u g a t i o n . However l i t t l e i n terest was shown i n the method u n t i l the l a t e I960's when several adaptations of the technique such as sedimentation at unit g r a v i t y (Peterson and Evans, 1967; M i l l e r and P h i l l i p s , 1969), e l u t r i a t i o n (McEwen et al, 1968; M e i s t r i c h et a i , 1977; 1981), sedimentation i n an i s o k i n e t i c gradient (Pretlow, 1971; Pretlow et a i , 1975) and c e n t r i f u g a t i o n using a reori e n t i n g gradient zonal rotor (Wells et al, 1977a, b) increased the e f f i c i e n c y and p o p u l a r i t y of sedimentation techniques. V e l o c i t y sedimentation can be c a r r i e d out using a continuous or discontinuous density gradient. In the case of the discontinuous gradient the chamber i s f i l l e d with successively le s s dense solutions r e s u l t i n g i n in t e r f a c e s at which the s o l u t i o n density changes abruptly. This has been l a r g e l y superseded by the use of continuous gradients, p a r t i a l l y to avoid the problem of the c e l l s c o l l e c t i n g at the i n t e r f a c e s where they tend to aggregate. However one frequently used discontinuous gradient method employs a Ficoll/Hypaque (sodium d i a t r i z o a t e ) density 1 gradient to remove erythrocytes from blood ( A u i t i et al, 1974). The more common gradient forming polymers include F i c o l l , albumin and P e r c o l l ( c o l l o i d a l s i l i c a ) . The sedimentation of c e l l s i n a c e n t r i f u g a l f i e l d i s described by (Pretlow and Pretlow, 1982): 2 2 , a (D - D )w r ,. ^ dr _ c m (1) dt kr) where r i s the distance of the c e l l from the center of revolution, t i s time, a i s the diameter or radius (depending upon the value of k) of the c e l l , D c i s the density of the c e l l , D m i s the density of the gradient at the l o c a t i o n of the c e l l , u i s the angular v e l o c i t y (the speed of the ce n t r i f u g a t i o n ) , TJ i s the v i s c o s i t y of the gradient at the l o c a t i o n of the c e l l , and k i s a constant. V e l o c i t y sedimentation uses both c e l l density and diameter as the c r i t e r i a f o r separation since i n s u f f i c i e n t c e n t r i f u g a t i o n time i s used fo r the c e l l s to reach their respective buoyant d e n s i t i e s . For t h i s reason i t i s generally considered superior to isopycnic sedimentation where s u f f i c i e n t force i s applied f o r s u f f i c i e n t time to allow the c e l l s to a r r i v e at t h e i r respective buoyant d e n s i t i e s therefore making c e l l density the only c r i t e r i a f o r separation (Pretlow and Pretlow, 1982). D i f f e r e n t i a l sedimentation at unit g r a v i t y or d i f f e r e n t i a l c e n t r i f u g a t i o n does not make use of a density gradient. The less r a p i d l y sedimenting c e l l s remain i n suspension (Anderson, 1966). This generally gives poor r e s o l u t i o n and may need to be repeated several times r e s u l t i n g i n the loss of some of the desired c e l l s . Sedimentation at unit g r a v i t y i s sometimes combined with electrophoresis where the applied f i e l d i s across the density gradient to give a 2-dimensional separation (Platsoucas, 1983). Ce n t r i f u g a l e l u t r i a t i o n balances the outwardly d i r e c t e d i n e r t i a l ( c e n t r i f u g a l ) forces acting on a c e l l against an inwardly d i r e c t e d hydrodynamic force created by pumping f l u i d through a chamber i n the c e n t r i f u g a l rotor. This i s usually arranged such that the hydrodynamic force i s s l i g h t l y dominant causing the c e l l s to move inward, i . e . those with low sedimentation rates move inwards f a s t e r . The main disadvantages of v e l o c i t y sedimentation techniques are 2 that the basis for separation is not directly involved with c e l l function and care must be taken to maintain an isosmotic gradient to avoid selective shrinking of the cells. The capacity is small, usually 7 in the order of 1-5x10 cells although large rotors have been used for up to 10 1 0 cells (Wells and James, 1972). If the capacity is exceeded in any one band the cells in that particular band w i l l tend to sink lower in the gradient. This is a result of the concentration of sedimented substances sufficiently altering the density of the part of the gradient where they are located to exceed the density of the centrifugally located gradient. The denser solution then sediments rapidly as a large bolus to a denser part of the gradient. There are problems with aggregation especially at the wall of the tube where ce l l s tend to collect since the force vectors radiate from the center of the rotor. Swirling during acceleration and deceleration can be a problem as well as the streaming of cells l e f t standing ln the gradient for extended time periods. 2 . Single c e l l sorting and multiparameter analysis. This technique sorts ce l l s individually according to the signal generated when fluorescently labelled or optically characterized cells interact with a light beam as they pass by in suspension in single f i l e . The development of high speed flow systems for single c e l l sorting combined with automated analysis has made i t possible to resolve homogeneous c e l l populations from heterogeneous populations. Sorting is rapid, usually based on electrostatically charging drops containing cells of interest. Typically 105 c e l l s per minute may be analyzed and the data displayed as frequency histograms. Since many cells may be examined the s t a t i s t i c a l precision is high and small sub-populations can be detected and separated with confidence e.g. 1-5% of the total population. The beginnings of flow cytometry occurred in 1934 when Moldaven reported a bright-field photometric method for counting individual c e l l s in liquid suspension flowing through a capillary tube located en e microscope stage (Moldaven, 1934). With the development of the Coulter counter and size analyzer in the mid 1950's contributing significantly, the f i r s t system capable of automatically measuring multiple parameters was described by Kamentsky et al (1965), who sorted cells on the basis i 3 of uv light absorption and visible light scatter as they flowed through a narrow channel under a microscope. Modern multiparameter c e l l sorters combine both optical and electrical sensing techniques to discriminate according to preselected parameters relating to c e l l volume and surface area, total or two color fluorescence from stains bound to specific biochemical features and light scatter related to internal and external cellular characteristics. Other parameters measured include axial light extinction, time of fl i g h t and membrane polarization. The main disadvantage of single c e l l sorting is that the apparatus is very expensive, large c e l l numbers cannot be handled and problems may be encountered in disaggregating the cells and keeping them monodisperse during fixing, staining and analysis. In addition only two or three isolated sub-populations result from the process, limiting i t s capacity for separation in comparison to the analytical capability of the method. For further information see Steinkamp et al (1974); Mullaney et al (1976) and Preffer and Colvin (1987). 3. Magnetic separation. In this method cells are made weakly magnetic either on the basis of intrinsic magnetization, e.g. erythrocytes, or by specific labelling to make them magnetizable for the purposes of the separation, e.g. by attaching paramagnetic materials to lectins or antibodies. The separation of magnetic from non-magnetic ce l l s in a suspension is carried out by a flow system through permanent magnets (low gradient magnetic f i l t r a t i o n ) or by high gradient magnetic f i l t r a t i o n . In the latter case a column is loosely packed with fine magnetic stainless steel wire and placed between the poles of an electro- or super-conducting magnet. As the ce l l suspension is passed through the column the magnetizable cells become bound to the wire. The small diameter of the wire ensures that the distance over which the induced magnetic f i e l d vanishes is small. This means that a large f i e l d gradient is generated providing the necessary attractive force. The column must be sufficiently long to ensure that each c e l l w i l l come close enough to the wire to be attracted i f magnetizable. Loose packing of the column means that few problems due to clogging are encountered (Oder, 1976; Owen and Liberti, 1987). Some methods which are used to generate paramagnetic c e l l s include: 4 a. Phagocytosis. C e l l s which ingest magnetic compounds may be removed from those incapable of rapid phagocytosis e.g. the incubation of spleen c e l l s with carbonyl i r o n i s used to deplete macrophages from c e l l suspension (Hudson and Hay, 1976). Several t r i v a l e n t lanthanide ions such as erbium and -dysprosium have been incorporated into c e l l s rendering them magnetic but no s e l e c t i v e uptake has been demonstrated (Graham and Se l v i n , 1982). b. P r e - e x i s t i n g magnetic compounds. Erythrocytes are i n t r i n s i c a l l y non-magnetic but the i r o n present i n the hemoglobin may be rendered paramagnetic by the removal of r e v e r s i b l y bound oxygen. This may be done by the a d d i t i o n of sodium d i t h l o n i t e or the use of an a r t i f i c i a l lung machine. The low e f f i c i e n c y of erythrocyte removal (60-70%) makes t h i s method more appropriate f o r the removal of erythrocytes from blood p r i o r to- a more rigorous separation procedure (Paul et al, 1978). M a l a r i a l p arasites are known to cause i n t r a c e l l u l a r degradation of hemoglobin. Some of the breakdown products are paramagnetic which may prove useful i n malaria diagnosis (Paul et al, 1981). Another method of rendering erythrocytes paramagnetic i s the formation of methemoglpbin by treatment of oxy-hemoglobin with sodium n i t r i t e (15-30 mM). The loss of one electron from each heme-iron r e s u l t s i n a stable paramagnetic erythrocyte. Due to the r e l a t i v e i n e f f i c i e n c y of the procedures discussed i n a and b magnetic separation techniques are not used extensively to debulk red blood c e l l s from whole blood, but targeted c e l l s have been studied i n greater d e t a i l . c. Rosetting. When an excess of anti-immunoglobulin or protein-A coated sheep erythrocytes are centrifuged with immunoglobulin-coated lymphocytes, the erythrocytes w i l l adhere to the surface of the lymphocytes forming rosettes (Goding, 1983). If sheep red blood c e l l s are treated with n i t r i t e to form methemoglobin then used to rosette T-lymphocytes, the rosetted T-lymphocytes may be magnetically removed from the suspension (Owen, 1982). 5 d. Magnetic particles. Coating magnetic spheres with targeting agents such as antibodies has produced a variety of particles which can bind to specific types of cells, thus rendering them magnetic. The f i r s t magnetic particles were agarose-polyacrylamide beads (Guesdon and Aurameas, 1977) but since then microspheres of starch (Mosbach and Schroder, 1979), dextran and agarose (Molday and MacKenzie, 1982; Schroder and Mosbach, 1983), cellulose (Forrest and Landon, 1978), polystyrene (Ugelstad et al, 1983; Treleaven et al, 1984), styrene-divinyl benzene (Ugelstad et al, 1984), methacrylate (Molday et al, 1977) and heat-denatured albumin (Widder et al, 1981) containing Fe^^ have been used. The most monodisperse and uniform microspheres are those made of styrene-divinyl benzene (Ugelstad et al, 1983). Magnetic microspheres coated with antibody have been used to remove tumor c e l l s from bone marrow (Treleaven et al, 1984). The use of sheep anti-mouse IgG coated spheres means that the same beads may be used for separations of different malignancies where the marrow is pretreated with monoclonal IgGs of differing s p e c i f i c i t i e s . These beads have also been used for the separation of T- and B-lymphocytes (Schroder et al, 1986). The main disadvantage of this method, as with any solid-phase a f f i n i t y technique, is the problem of non-specific absorption discussed later in the text. In addition high magnetic fields can be required, making the method expensive, particularly when liquid helium-cooled magnets are employed. 4. S o l i d phase a f f i n i t y f r a c t i o n a t i o n . This technique involves the specific binding of cells directly to a solid phase support. Often the support is modified by the attachment of ligands. Various ligands have been used; antibodies provided the f i r s t c e l l separation by this method (Wigzell and Anderson, 1969) but avidin (Basch et al, 1983), protein A (Langone, 1982) and several lectins (Sharon, 1984) have also been used. There are two serious drawbacks to this technique. Fir s t a substantial fraction of the c e l l s often become non-specifically bound to the matrix. Supports of hydrophilic polymer gels such as dextran (Schlos.sman and Hudson, 1973), agarose (Carton and Nurden, 1979), polyacrylamide (Baran et al, 1982), poly(2-hydroxyethyl methacrylate) (Tlaskalova-Hogenova et al, 1981) and a copolymer of acrylamide and N-acryloyl-2-amino-6 2-hydroxymethyl-l-3-propane diol (Bonnafous et al, 1983) have been reported to show less non-specific absorption than conventional matrices such as glass and poly(methyl methacrylate). Another type of a f f i n i t y c e l l selection uses nylon strings to which ligands are physically adsorbed. The fibers are incubated with the c e l l suspension and cells bound to the fibers are recovered by plucking each fiber with a needle (Edelman et al, 1971). Another "panning" method uses a plastic petri dish as the support (Wysocki and Sato, 1978). A group of block and graft copolymers have been found to eliminate adsorption and subsequent contact-induced activation of platelets (Akemi et al, 1986). These polyamine graft copolymer matrices have been used to separate T- and B-lymphocytes (Kataoka et al, 1988). 5. Electrophoresis. Electrophoretic methods of c e l l separation are based on differences in the density and distribution of the surface charges on a c e l l . The electrophoretic mobility is a complex function of the number of charge-bearing l i p i d , protein and glycoprotein species on the surface, the depth and charge distribution of this surface-bound material and the ionic strength of the suspending medium (Levine et al, 1983; Sharp and Brooks, 1985). The more common electrophoretic methods of c e l l separation are: a. Free-flow electrophoresis. This may be done in a vertical or horizontal, rotating column. The cells are introduced into a flowing buffer curtain and deflected to varying degrees within the buffer curtain. They are collected at the end of the column into tubes depending on their position within the buffer (Hannig, 1972). Problems with this method include sedimentation of cells due to gravity and the creation of thermal convection currents. b. Dielectrophoresis. In this case an alternating current is applied which causes the charged cells to oscillate around a mean position whereas the polarization of neutral particles means that they w i l l migrate to regions of high f i e l d intensity (Pohl, 1977). c. Density gradient electrophoresis. This is done in a vertical column 7 containing a density gradient, usually formed with ficoll/sucrose (Platsoucas, 1983). The density gradient helps to avoid sedimentation problems but the very low ionic strength buffers that must be employed in order to minimize electric heating, as this results in mixing of the cells by- convection currents, may cause c e l l damage (Seaman, 1975). 6 . Biological Separations. These methods are applicable to ce l l s which may be maintained in culture for limited or extended time periods. The techniques include differential outgrowth from tissue fragments attached to a culture dish (Lechner et al, 1981) or differential attachment which is frequently used to separate fibroblasts from epithelial cells (Kasten, 1973). Differential detachment using enzymes and/or chelating agents is commonly used (Rheinwald, 1980) as well as di f f e r e n t i a l digestion with trypsin or collagenase (Owens et al, 1976). Perfusion has been used to isolate rat liver parenchymal cells (Seglen, 1973). Cloning is the most rigorous separation procedure available but a considerable fraction of the culture l i f e span can be used up in the isolation procedure (Ham, 1972). The use of selective reagents exploits the differences between dividing and non-dividing cells, normal and neoplastic c e l l s and specific c e l l types, e.g. nucleic acid components and analogs may be used to selectively block metabolic pathways of unwanted c e l l types as in HAT selection for hybridoma cells ( L i t t l e f i e l d , 1964). Deficiencies in particular amino acids may be used to suppress growth of certain c e l l types (Leffert and Paul, 1972) or substitution of some amino acids may be selective, e.g. the replacement of L-valine by D-valine w i l l select out epithelial cells (Gilbert and Migeon, 1975). In some cases selective media may be used. This usually involves the absence of serum with or without the addition of hormones and growth factors (Sato, 1975). Optimization of the basal medium may also give some degree of selectivity (Jennings and Ham, 1983; Lechner et al, 1982; Kaighn et al, 1983). 7. Partitioning in aqueous- polymer two-phase systems. Despite this variety of c e l l separation methods with varying advantages and disadvantages, most c e l l separations are based on properties not 8 d i r e c t l y associated with c e l l function, i.e., s i z e , density, shape or charge of the c e l l . C e l l sorting and s o l i d phase a f f i n i t y techniques are more s e l e c t i v e because of their use of s p e c i f i c ligands. These methods, based on c e l l surface biochemistry, are l i k e l y to be associated with c e l l f u n c t i o n or dysfunction and to be more di s c r i m i n a t i n g . However i n d i v i d u a l c e l l s o r t i n g i s expensive and impractical for bulk separations and the use of s o l i d phase supports generally poses problems i n c e l l removal from the s o l i d support and non-specific adsorption. One method of c e l l separation which depends p r i m a r i l y on c e l l surface properties i s p a r t i t i o n i n g i n aqueous polymer two-phase systems (APTS) and i t i s with t h i s technique that t h i s t hesis i s concerned. P a r t i t i o n i n g i n an APTS involves the unequal d i s t r i b u t i o n of c e l l s between the i n t e r f a c e and one of the phases of a two phase system composed of two incompatible polymers, usually polyethylene g l y c o l (PEG) and dextran (dx), and an aqueous buffer. A t y p i c a l procedure, F i g 1., involves adding the c e l l s to the APTS, mixing the phases then allowing them to s e t t l e , followed by sampling and analysis. It i s analogous to a t r a d i t i o n a l solvent extraction. As the separation i s performed i n s o l u t i o n there i s no problem with non-specific adsorption and the method has been scaled up almost 40,000 times with no loss i n r e s o l u t i o n when used f o r enzyme p u r i f i c a t i o n (Kroner et al, 1982). It i s a well established f a c t that a two phase system can be formed from two incompatible neutral polymers i n a common solvent (Flory, 1953) and while many other types of two and multiphase systems e x i s t , the former i s the only type used to any degree f o r c e l l separations. Since the common solvent i s generally water, these two phase systems may be buffered and made isotonic, making them compatible with b i o l o g i c a l materials and l i v i n g c e l l s . P a r t i t i o n i n g i n aqueous polymer two phase systems has proven to be a most v e r s a t i l e separation technique, having been used to separate amino acids, n u c l e i c acids, membrane fragments, organelles, micro-organisms and c e l l s , amongst others. The p a r t i t i o n c o e f f i c i e n t , K, i s defined as the r a t i o of the concentration of solute i n the upper phase to the concentration i n the lower phase ( i . e . K = C /C ) and i n the case of c e l l s or p a r t i c l e s as the r a t i o of the number T B of c e l l s or p a r t i c l e s i n the upper phase to t o t a l number at the i n t e r f a c e and i n the lower phase ( i . e . K = n /(n + n ). If the s i n g l e 9 step partition process is repeated by exposing each phase to fresh complementary phases in a continuous manner, a multistep procedure of increased resolution can be developed known as a countercurrent distribution (CCD). This is the discrete analog of, and precursor to, chromatography. LET SETTLE SAMPLE, ANALYZE Fig. 1. A typical c e l l partitioning experiment is illustrated. The phase system i s composed of polyethylene glycol (PEG), dextran (dx} and an aqueous buffer. From Walter (1969). 10 a. History of the development of p a r t i t i o n i n g i n aqueous polymer two-phase systems. The use of APTS for the separation of b i o l o g i c a l material was developed by Albertsson (1956) who used a two-phase system formed by mixing PEG and phosphate to i s o l a t e chloroplasts from green algae. He went on to pioneer the use of two incompatible h y d r o p h i l i c polymers as the APTS forming agents which enabled separations at p h y s i o l o g i c a l pH and t o n i c i t y . By the use of CCD he demonstrated that the p a r t i t i o n c o e f f i c i e n t observed i n an APTS composed of two polymers was a r e v e r s i b l e , thermodynamic p a r t i t i o n r a t i o rather than an adsorption process f o r both p a r t i c l e s and c e l l s (Albertsson and Nyns, 1959; 1961; Baird et al, 1961; Albertsson and Baird, 1962). Albertsson also studied the r e l a t i o n s h i p between the molecular weight or surface area of proteins and viruses and t h e i r p a r t i t i o n c o e f f i c i e n t (Albertsson, 1958, 1959), i s o l a t e d ribosomes from rat b r a i n microsomes (Albertsson et al, 1959) and concentrated and p u r i f i e d v iruses (Frick and Albertsson, 1959). Antibody-antigen binding was also found to a f f e c t the p a r t i t i o n c o e f f i c i e n t (Albertsson and Philipson, 1960). A l l t h i s information was published as a book (Albertsson, 1960) which i s now i n i t ' s t h i r d e d i t i o n (Albertsson, 1986). The development of a t h i n layer CCD apparatus which decreased s e t t l i n g time compared to the t r a d i t i o n a l Craig apparatus increased the r e s o l u t i o n over a s i n g l e p a r t i t i o n step (Albertsson et al, 1965) and enzyme p u r i f i c a t i o n s by p a r t i t i o n using salt-PEG systems were accomplished (Okazaki and Kornberg, 1964). Salt-PEG systems are now used extensively for i n d u s t r i a l scale enzyme separations (Hustedt et al, 1985). It was known that small changes i n the i o n i c composition of the APTS had a large e f f e c t on the p a r t i t i o n of proteins, n u c l e i c acids or c e l l s (Albertsson, 1986). Studies by Johansson (1970a) showed that s a l t s p a r t i t i o n e d unequally i n an APTS and the e l e c t r i c a l p o t e n t i a l created by t h i s unequal p a r t i t i o n was estimated using electrodes (Reitherman et al, 1973; Johansson, 1974). A l i n e a r r e l a t i o n s h i p between the log of the p a r t i t i o n c o e f f i c i e n t and p r o t e i n net charge was observed i n some cases (Johansson, 1970b; 1971) which was used to determine i s o e l e c t r i c points by c r o s s - p a r t i t i o n (Albertsson, 1970) and estimate net charge by p a r t i t i o n (Blomquist, 1976). Cross p a r t i t i o n i s a technique which makes use of the f a c t that at the i s o e l e c t r i c point (pi) of the p r o t e i n or 11 particle the partition is not determined by electrostatic effects such as the potential difference, ionic strength or electrostatically induced conformation changes. Thus a plot of K as a function of pH or net charge in two systems with different potential differences due to their salt compositions should cross at the pi. b . i . More recent uses of partitioning in APTS (proteins, nucleic acids and sub-cellular particles). Over the past twenty years partitioning in APTS has been applied to the study and separation of a wide range of solutes and particles. The partition coefficients for a variety of proteins have been measured (Albertsson, 1958, 1960) and several general rules describing the partition behavior of proteins were established. Theoretical descriptions of partitioning described later in the text agree with these observations: 1. The higher the concentration of phase forming polymer, the more extreme the partition of the protein (providing no precipitation occurs). 2. With constant PEG/dx concentrations ( or constant tie line length as defined later in the text) lowering the molecular weight of one of the polymers increases the partition of the protein into the phase in which that polymer predominates. 3. Protein dissociation may alter the partition of a protein as demonstrated by the pH dependence of the partition coefficient. 4. The partition may be influenced by the addition of salt, the effect depending on the pH, ionic composition and concentration of the salt. For a negatively charged protein the partition coefficient is often successively decreased in the series: P03~ > S02~> F_> CH COO- > Cl" > ' 4 4 3 Br" > I" and Li +> NH+> Na+> K+. 4 5. The partition coefficient is independent of the concentration of the partitioned material up to 50g/L. One of the more useful applications of partitioning in APTS has been the separation of proteins from nucleic acids e.g. the purification of RNA polymerase a peptide from E. Coli/bacteriophage (Goff, 1974). For a complete l i s t of protein/nucleic acid separations see Johansson (1985). The incorporation of charged PEGs such as trimethylamino-PEG (TMA-PEG) and PEG-sulfonate (PEG-S) into the APTS causes the partition 12 of proteins to become extremely pH sensitive. These charged systems have been applied to the fractionation of baker's yeast (Johansson, 1985). Many other proteins have been partitioned including milk proteins (Igarashi et al, 1974), histones (Axelsson and Shanbhag, 1976; Bidney and Reeck, 1977), cellulases (Tjerneld, 1985), various albumins, ovalbumins and lactalbumins, hemoglobulins, serum globulins, interferon and various enzymes (For details see Johansson, 1985). A more specific effect on protein partition may be obtained by including a PEG-bound ligand in the APTS. Trypsin was extracted into the upper phase of an APTS by the addition of diaminodiphenylcarbamoyl-PEG to the system (Takerkart et al, 1974). Similarly S-23 myeloma protein was extracted with dinitrophenyl-PEG (Flanagan and Barondes, 1975). Fatty acid derivatized PEGs and APTS have been used for assessing the degree of exposure of hydrophobic surfaces on proteins (Axelsson, 1978). PEG and dx derivatized procion dyes are now used industrially to purify glycolytic yeast enzymes (Johansson and AnderssOn, 1984; Johansson and Joelsson, 1985; Johansson, 1984). A novel use of an APTS was the separation of free and bound labelled ligand in a binding assay which investigated interactions between concanavalin A and various glycoproteins and carbohydrates (Mattiasson and Ling, 1980). Partitioning has also been applied to the purification of nucleic acids. As with proteins the partition coefficient depends on the electrolytes present in the APTS, the structure and size of the nucleic acid and the presence of ligands. Separations include native and denatured DNA (Alberts, 1967), ribosomal genes from sea urchin sperm (Patterson and Stafford, 1970), plasmid DNA from bacterial lysates (Ohlsson et al, 1978a), DNA from rat tumors (Furnica, 1975) and the fractionation of chromosomal deoxyribonucleoproteins (Turner and Hancock, 1974). The development of chromatographic techniques whereby one of the phases is bound to a support has proven useful for nucleic acid separations (Miiller, 1985). The partitioning of sub-cellular particles has also received attention and is reportedly the only practical means for biomedical characterization of organelles and plasma membranes (Flanagan, 1985). Membranes have been isolated by partitioning for the past 20 years (Brunette and T i l l , 1971; Johansson, 1986). One example is the 13 separation of plasmalemma from human mammary carcinoma (Leung and Edgington, 1980). Three phase systems were used to p a r t i t i o n hydrophobic membrane components (Albertsson, 1973) and p a r t i t i o n i n g has also been used f o r the separation of post-synaptic density structures (Gurd et al, 1982). The separation of right-side-out and inside-out plasma membrane v e s i c l e s by p a r t i t i o n i n g i s used a n a l y t i c a l l y and preparatively (Walter and Krob, 1976a). A f f i n i t y p a r t i t i o n i n g using a TMA-PEG ligand has proven useful for the i s o l a t i o n of n i c o t i n i c c h o l i n e r g i c receptors (Flanagan, 1976). Many organelles including peroxisomes (Suga et al, 1979) and microsomes (Ohlsson et al, 1978b) have been p a r t i t i o n e d (Flanagan, 1985). A f f i n i t y p a r t i t i o n i n g of organelles includes the separation of c a l f brain synaptosomes with a PEG-procion dye a f f i n i t y ligand (Muino Blanco et al, 1986). Other uses of APTS include phase transfer c a t a l y s i s (Harris et al, 1982), the separation of a c t i n l d e s (widely occurring terpene a l k a l o i d s , Sotobayashi, 1977) and the immunological q u a n t i f i c a t i o n of b a c t e r i a l c e l l s (Ling et al, 1982). i i . The p a r t i t i o n i n g of mammalian c e l l s . One i n c r e a s i n g l y used a p p l i c a t i o n of p a r t i t i o n i n g i n APTS i s the separation and study of mammalian c e l l s . It i s with t h i s aspect of p a r t i t i o n i n g that t h i s t h e s i s i s most concerned. The p a r t i t i o n of i n t a c t c e l l s , as f o r a l l the previously mentioned solutes and p a r t i c l e s , depends on the composition of the APTS, i.e., on the pH and s a l t and polymer type.and concentration as well as on the surface properties of the c e l l (Walter and Krob, 1976b). Separations may be on the basis of the native or ligand exposed c e l l p a r t i t i o n . The majority of work done to examine the determinants of c e l l p a r t i t i o n to date has been c a r r i e d out on erythrocytes, probably due to t h e i r a v a i l a b i l i t y as s i n g l e c e l l s i n large q u a n t i t i e s . Erythrocyte p a r t i t i o n c o e f f i c i e n t s have been shown to c o r r e l a t e with e l e c t r o p h o r e t i c m o b i l i t y (Brooks et al, 1971) and separations of s p e c i e s - s p e c i f i c erythrocytes have been accomplished by CCD (Walter et al, 1967). Removal of s i a l i c a c i d by neuraminidase was found to e i t h e r lower or r a i s e the p a r t i t i o n c o e f f i c i e n t , depending on the buffer (Walter and Coyle, 1968; Walter, 1985). The age of the i n d i v i d u a l erythrocyte i n the same species 14 has an e f f e c t on the p a r t i t i o n (Walter et al, 1981) as does the age of the donor (Seaman et al, 1980). The i n c l u s i o n of PEG-fatty a c i d esters i n the APTS has a large e f f e c t on erythrocyte p a r t i t i o n . The f a t t y a c i d part of the ligand i s presumed to insert into the erythrocyte membrane thereby coating the erythrocyte with PEG and causing i t to p a r t i t i o n into the PEG-rich upper phase of the APTS (Raymond and Fisher, 1980; Van A l s t i n e and Brooks, 1984). It has been observed that ghosts and membrane v e s i c l e s have highly d i f f e r e n t p a r t i t i o n c o e f f i c i e n t s to the erythrocytes from which they were prepared (Walter and Krob, 1976c). An e f f o r t to c o r r e l a t e erythrocyte p a r t i t i o n i n g behavior with diseases associated with erythrocyte surface changes has been made. Rats treated to become highly anemic show a large increase i n numbers of c i r c u l a t i n g r e t i c u l o c y t e s . These "stress" r e t i c u l o c y t e s have d i s t i n c t l y lower p a r t i t i o n c o e f f i c i e n t s than mature erythrocytes (Walter et al, 1972). This observation was also made i n patients s u f f e r i n g from r e t i c u l o c y t o s i s . Deposition of complement protein 3Cb -on abnormal erythrocytes was used to separate sub-populations of erythrocytes from patients with paroxysmal nocturnal hemoglobinuria (PNH, Pangburn and Walter, 1987). Small but s i g n i f i c a n t differences i n p a r t i t i o n behavior between normal erythrocytes and those from multiple s c l e r o s i s p atients have been observed i n the presence of PEG-fatty acid esters (Van A l s t i n e and Brooks, 1984). Erythrocytes from rats bearing subcutaneous Leydig c e l l tumor F344 have a lower p a r t i t i o n c o e f f i c i e n t than normal rat erythrocytes (Gascoine et al, 1983). However no d i f f e r e n t i a l p a r t i t i o n behavior has been observed i n the erythrocytes of s i c k l e c e l l anemia, patients (Walter, 1985) or erythrocytes of chronic a l c o h o l i c s despite increased membrane ch o l e s t e r o l l e v e l s i n these erythrocytes (Walter et al, 1979a). P a r t i t i o n i n g of mammalian c e l l s other than erythrocytes has also been studied i n systems containing no ligands. These include the separation of lymphocytes and polymorphonuclear c e l l s by CCD (Walter et al, 1969) and observations of differences between mouse leukemic c e l l s during lag and exponential growth phases i l l u s t r a t e d by CCD p r o f i l e s (Gersten and Bosmann, 1974). In f a c t changes associated with c e l l d i f f e r e n t i a t i o n may be traced by p a r t i t i o n i n g (Stendahl et al, 1982). Walter et al (1979b) were able to separate T-lymphocytes, B-lymphocytes 15 and F c receptor bearing c e l l s by CCD and Malmstrom (1980a,b) showed that natural k i l l e r (NK) c e l l s were located i n f r a c t i o n s with higher G values than the T- and B-lymphocytes and NK c e l l s . Differences i n the p a r t i t i o n of normal and transformed f i b r o b l a s t s have been observed (Sherbet and Lakshmi, 1981) and monocytes with high phagocytosing capacity have been found to p a r t i t i o n d i f f e r e n t l y to those with lesser phagocytic properties (Walter et al, 1980). There has been some success at d i s t i n g u i s h i n g c e l l s with high metastatic p o t e n t i a l from those with low metastatic p o t e n t i a l (Bosmann et al, 1973; Miner et al, 1981; Van A l s t i n e et al, 1986) and the e f f e c t s of some drugs on c e l l surface hydrophobicity has been investigated by p a r t i t i o n i n g (Kessel and McElhinney, 1978). Other c e l l s studied by p a r t i t i o n i n g include bone marrow c e l l s (discussed l a t e r i n the text), rat l i v e r c e l l s (Walter et al, 1973a), i n t e s t i n a l e p i t h e l i a l c e l l s (Welser, 1973; Walter and Krob, 1975) and i r r a d i a t e d c e l l s which appear to show only small, i f any, change i n p a r t i t i o n behavior even af t e r extremely high l e v e l s of i r r a d i a t i o n (Gersten and Bosmann, 1975; Niepokojczycka et al, 1982). Although p a r t i t i o n i n g of c e l l s i n the absence of ligands may achieve the desired separation i n some cases, i t i s often necessary to separate c e l l s on the basis of more disc r i m i n a t i n g f a c t o r s than the o v e r a l l surface properties of the c e l l . The incorporation of a s u i t a b l e a f f i n i t y ligand into the system may increase the r e s o l u t i o n of the separation. The most s p e c i f i c type of a f f i n i t y ligand i s an antibody. These provide a powerful class of ligands f o r use i n p a r t i t i o n i n g , provided they have been modified such that they p a r t i t i o n predominantly into one of the phases c. Immunoaffinity p a r t i t i o n . Immunospeciflc separations of high r e s o l u t i o n can be obtained using antibodies i n a v a r i e t y of separation techniques, f o r instance a f f i n i t y chromatography. By applying these same p r i n c i p l e s to p a r t i t i o n i n g i n a technique known as immunoaffinity p a r t i t i o n , the separation of populations of v i a b l e i n t a c t c e l l s on the basis of t h e i r surface antigens i s possible. An immunoaffinity ligand i s generally an antibody which has been modified such" that i t p a r t i t i o n s into one of the phases of an APTS, e.g. a PEG-derivatized antibody w i l l tend to p a r t i t i o n into the upper phase of a PEG/dx system. The ligand 16 simultaneously binds to the solute or p a r t i c l e to be separated while p a r t i t i o n i n g into one of the phases, thereby increasing the p a r t i t i o n of that solute or p a r t i c l e into that phase. The approaches taken i n c e l l immunoaffinity p a r t i t i o n have u t i l i z e d a primary antibody a f f i n i t y l igand (Sharp et al,. 1986; Karr et al, 1986) and a second antibody a f f i n i t y ligand (Stocks and Brooks, 1988). In the primary antibody a f f i n i t y ligand case, a PEG-modified antibody recognizing a c e l l surface antigen has been incorporated, along with the c e l l mixture to be separated, into an aqueous two phase system containing PEG and dx. The PEG-antibody on binding to the c e l l surface e f f e c t i v e l y coats i t immunospecifically with PEG, thereby increasing the p a r t i t i o n of that c e l l into the upper, PEG-rich, phase (Fig. 2.a). The f i r s t studies of c e l l separations by immunoaffinity p a r t i t i o n were of species s p e c i f i c erythrocytes (Sharp et al, 1986; Karr et al, 1986). Although primary a f f i n i t y ligands c l e a r l y demonstrate immunospecific e f f e c t s on c e l l p a r t i t i o n , there are disadvantages to the use of a d i r e c t l y - m o d i f i e d primary antibody as an a f f i n i t y ligand, the most s i g n i f i c a n t being the need f o r a d i f f e r e n t modified-antibody ligand f o r each c e l l separation problem. This e n t a i l s time consuming optimization of the modification chemistry as well as separate a n a l y s i s and c h a r a c t e r i z a t i o n of the ligand f o r each c e l l separation. This disadvantage may be overcome by using a second antibody or other modified reagent as the a f f i n i t y p a r t i t i o n i n g ligand i n combination with a native primary antibody. One approach i s to rai s e a monoclonal antibody (MAb) which s p e c i f i c a l l y binds to the F c region of a primary antibody. The F £ region of an antibody of any one clas s i s constant i r r e s p e c t i v e of antibody s p e c i f i c i t y (Goding, 1983). This antibody may be modified such that i t p a r t i t i o n s into one of the two phases e.g. by attachment of PEG or trypan blue dye (Fig. 2c). Trypan blue p a r t i t i o n s strongly into the upper phase (see l a t e r i n discussion). Thus the same second ligand may be applied to any separation problem f o r which a primary antibody i s av a i l a b l e (Stocks and Brooks, 1988). Another advantage of s e l e c t i n g the second antibody to bind to the F c region of the primary antibody i s that the attached PEG i s remote from the s i t e of cel l - p r i m a r y antibody binding. Since some s t e r i c interference by PEG i n the c e l l - a n t i b o d y binding i s thought to occur when using the primary 17 A-PEG PEG polyethylene glycol B biotin PAA polyacrylamide A avidin M trypan blue or PEG Fig. 2. Schematic diagrams of approaches to immunoaffinity partition of c e l l s , a. PEG-derivatlzed primary ligand (Sharp et al, 1986; Karr et al, 1986) b. Polyacrylamide (PAA)-derivatized primary antibody (this work), c. PEG- or trypan blue-derivatized second antibody (Stocks and Brooks, 1988; this work). d. Biotin-derivatized primary antibody and PEG-derivatized avidin (this work). 18 antibody technique (Sharp et al, 1986), t h i s i s a d i s t i n c t advantage. Monoclonal antibodies are of p a r t i c u l a r use i n t h i s context since they can be i s o l a t e d as a pure molecular species, an i n f i n i t e constant supply i s p o t e n t i a l l y a v a i l a b l e and they may be chemically modified i n a reproducible way. The use of a MAb as a primary ligand i s the only way i n which the r e s o l u t i o n of immunoaffinity p a r t i t i o n can approach that required f o r the more stringent c e l l separation problems discussed l a t e r . Another approach to the second ligand technique i s the use of PEG-derivatized Staphylococcal protein A i n combination with a primary IgG ligand (Karr et al, 1987). Staphylococcal protein A binds to the F c region of IgG with a high a f f i n i t y and s p e c i f i c i t y (Goding, 1983). However there i s considerable v a r i a t i o n i n the strength of binding among the d i f f e r e n t IgG sub-classes i n each species, some of which are not bound at a l l . In some cases IgA and IgM w i l l also be bound and multivalent p r o t e i n A i s l i k e l y to cause greater aggregation than a divalent antibody. Another method i s to d e r i v a t i z e the primary antibody with b i o t i n and use PEG-derivatized a v i d i n or native s t r e p t a v i d i n as the second ligand (Fig. 2.d.). S t r e p t a v i d i n p a r t i t i o n s into the upper phase (Flanagan, 1985). This has s i m i l a r disadvantages to the d i r e c t l y modified primary antibody but has advantages i n that i t uses les s l a b i l e reagents and an exceptionally strong binding reaction (Stocks and Brooks, 1987). There are few reports of c e l l separations by immunoaffinity p a r t i t i o n and f o r the most part these separations have been of species s p e c i f i c erythrocytes using the primary antibody technique. For the technique to be of c l i n i c a l use i t must be capable of separating c l o s e l y r e l a t e d c e l l types with only small differences i n surface antigens. The main object of t h i s study i s to apply immunoaffinity p a r t i t i o n to a more stringent and p o t e n t i a l l y useful c e l l separation, while simultaneously examining novel a f f i n i t y ligands and optimizing p a r t i t i o n conditions f o r an e f f i c i e n t separation of viable 1 c e l l s . Two p a r t i c u l a r l y i n t e r e s t i n g p r a c t i c a l problems to which immunoaffinity p a r t i t i o n could be applied are the purging of bone marrow p r i o r to r e i n f u s i o n during autologous bone marrow transplants as leukemia therapy and the separation of f e t a l 19 i s l e t of Langerhans cells for implantation in patients with insulin-dependent diabetes mellitus. i . Immunoaffinity p a r t i t i o n - An approach to bone marrow purging. Numerous studies have demonstrated that bone marrow transplants from histocompatible donors in conjunction with ablative therapy can be an effective treatment for acute leukemia and lymphoma (Zwann et al, 1982; Tutschka et al, 1980). Although considerable controversy remains regarding the relative merits of bone marrow transplant and intensive chemotherapy as treatment for acute leukemia (Jehn and Grunewald, 1988) most reports favor bone marrow transplants for other types of cancer (Williams et al, 1989). Only 40% of patients have histocompatible donors, however, and the relevance of HLA compatibility to sustained marrow engraftment is well known. A far more immunosupressive regime is required in cases where the donor is not an HLA-ldentical sibling (Anasetti et al, 1989). Autologous bone marrow transplant, in combination with purging of the tumor cells, has been successfully used and does not require a bone marrow donor (Ritz et al, 1982; Wells et al, 1979). A similar treatment for severe combined immunodeficiency disease has also been applied; this involved the removal of immunocompetent T lymphocytes by MAb (Reinhertz et al, 1982). Other diseases which have been treated by bone marrow transplants include Hodgkin's disease (Gribben et al, 1989), thalassemia and other inherited metabolic diseases (Levinsky, 1989). Bone marrow transplants have also been used to obtain donor-specific unresponsiveness in kidney allograft recipients. A course of anti-lymphocyte serum at the time of transplant is followed by transfusion of donor bone marrow (Barber et al, 1989). A typical procedure for autologous bone marrow transplant in the treatment of leukemia or lymphoma commences with the induction of remission by chemotherapy, followed by harvesting of the bone marrow. The bone marrow is processed in some way to remove tumor cell s , then cryopreserved. Meanwhile the patient undergoes ablative chemotherapy and f i n a l l y total body irradiation (TBI). Approximately 12 hours after TBI, the cleansed bone marrow is reinfused into the patient. The purging of the bone marrow is a crucial step in the treatment. As a result, considerable research is currently underway in the areas of 20 long term storage and culture of bone marrow cells and methods of enriching the bone marrow white cells while eliminating the tumor ce l l s from the bone marrow in vitro. Several methods have been applied to the removal of tumor cells from normal cells while maintaining the latter in a viable condition, including: 1. Treatment with MAbs and complement (Linker-Israeli et a i , 1981; Wells et a i , 1979; Economu et a i , 1978; Netzel et al, 1980). 2. Long term marrow cultures which favor the production of multipotential stem cells while disfavoring the maintenance of tumor ce l l s (Chang et al, 1986). 3. Incubation with MAbs, followed by exposure to anti-mouse antibody bound to magnetite-containing microspheres. Exposure of the ce l l s to a series of magnets results in the removal of the tumor cells recognized by the primary antibody (Treleaven et a i , 1984). A more recent method employs polystyrene magnetic beads (Morecki and Slavin, 1988). 4. Short incubation with 4-hydroperoxycyclophosphamide (Sharkis et al, 1980). Although cyclophosphamide derivatives are the most commonly used drugs others such as Ditercalinium (a pyridocarbazole derivative) and cis-platinum have shown potential, the latter in particular since i t exhibits low myelotoxicity (Benard et al, 1988). 5. Fractionation on discontinuous albumin gradients (Dicke et al, 1979). This method has been semi-automated using a ficoll-hypaque gradient and transfused cells prepared by this method engrafted within 20 days (English et al, 1989). A l l of the above approaches suffer from drawbacks of varying severity. Complement lysis is not 100% efficient due to the presence on most c e l l surfaces of enzymes which inactivate the complement complex before c e l l lysis occurs (Ritz and Schlossman, 1982). The selectivity of long term marrow cultures for the normal population w i l l vary depending on the type of tumor and condition of the donor. The use of solid phase supports in the magnetic bead-based separation suffers from the same problem as occurs in immunoaffinity chromatographic separations of cells , namely the non-specific adsorption of non-target cells to the beads. Similarly, separations based oh size and density differences are never specific since the relevant properties vary so widely through the normal and tumor c e l l populations and change depending on the stage in 21 the c e l l cycle in which the members of each population occur. Hence, i t seems clear that alternate techniques for bone marrow purging ought to be explored. A converse approach is the positive selection of stem cells rather than the removal of tumor cells. MAbs directed against a human progenitor c e l l antigen have been used for autologous stem c e l l selection in primates and these selected cells were able to f u l l y reconstitute hemopoietic function following TBI (Levinsky, 1989). Partitioning in APTS is one method with potential for the fractionation of populations of bone marrow cells. There have been a few studies of the CCD profiles of bone marrow cell's but only the hemoglobin-containing cells have been studied in any detail. In erythroid-stressed rats (due to phenyl hydrazine injection) a large proportion of the bone marrow cells (947.) are reticulocytes which have a lower partition than normal erythrocytes located in the bone marrow in appropriate phase systems (Walter et al, 1973b). This was also observed in peripheral erythrocytes (Walter et al, 1972). X-ray irradiation in vitro (10 Gray ) does not alter the CCD pattern of bone marrow ce l l s (Walter et al, 1974). The shift in hemoglobin synthesis from embryonic to adult hemoglobins in the bone marrow cells of adult rats has been assigned to erythroid cells with different CCD profiles at different stages of their development (Weiser et al, 1976). The development of MAbs specific for tumor-associated c e l l surface antigens (TAAs) is extensively documented (Hawkey et al, 1986; Robinson et al, 1986; Shipman et al, 1983; Springer, 1985; Linker-Israeli et al, 1981). Such monoclonal antibodies have been used for the identification (Dhokia et al, 1986a; b), separation (Treleaven et al, 1984; Kemshead and Ugelstad, 1985) and selective destruction of malignant cells (Nadler et al, 1980; Thierfelder et al, 1977). For example, a MAb which reacts with the cells of patients with acute nonlymphoblastic leukemia (ANLL) has been shown capable of indicating periods of remission and relapse up to three months prior to the event (Levy et al, 1985). The same MAb has been used to label and selectively k i l l ANLL-associated tumor c e l l lines using an antibody-hematoporphyrin conjugate and laser irradiation.(Mew et al, 1985). To date the principal use of monoclonal antibodies against TAAs has 22 been in the identification of malignant cells, although the potential development of "magic bullets" to destroy target c e l l s following antibody binding has received considerable attention. Less effort has gone into applying anti-TAA MAbs in the separation of malignant from normal cells, fluorescence-activated c e l l sorting (FACS, Herzenberg, 1977) and a recently developed technique using MAb-coated magnetic microspheres (Treleaven et al, 1984) being the principal approaches used. The poss i b i l i t y of applying immunoaffinity partition to c e l l types relevant to the bone marrow purging problem is investigated in this thesis. The approach taken is the development of a second, MAb a f f i n i t y ligand to be used in combination with one of the aforementioned panel of MAbs to various TAAs in an APTS to separate tumor cells recognized by the MAb. In order to model to a degree the bone marrow purging problem, a separation of two sub-lines of a transformed mouse lymphocyte was attempted. The problem was selected to test more stringently the limits of immunoaffinity partition since the lymphocyte sub-lines selected were qualitatively identical in surface characteristics, only the surface concentration of a membrane antigen differing in the two c e l l types. i i . Immunoaffinity partition - An approach to transplantation of fetal i s l e t tissue for Type II diabetes treatment. The ultimate goal of pancreatic i s l e t transplantation is to reinstate glucose and insulin homeostasis in the insulin-dependent patient, thus preventing the long-term complications of the disease. The main problem with i s l e t transplantation is not only immune rejection but the fact that the islets of Langerhans constitute only 1-2% of the mass of the pancreas. The fetal pancreas has many potential advantages as a donor organ for the treatment of insulin-dependent diabetes mellitus due to i t s capacity for growth and development (Clark and Rutter, 1972). Also, removal prior to exocrine development avoids problems due to immune responses to acinar c e l l s (Brown et al, 1976), the small size of the organ means that i t is easily cryopreserved (Mazur et al, 1976) and possibly unlimited supplies are available, subject to public opinion. However, although -transplantation of fetal pancreas effectively reverses diabetes in rodents (Brown et al, 1984; Federlin and Bretzel, 1984) the transplant 23 of human fetal pancreas has had only marginal success (Farkas and Karacson, 1985; Groth et al, 1980). The detrimental effect of transplanted acinar tissue on graft function and v i a b i l i t y may be avoided i f the i s l e t c ells can be positively selected. The selection procedure involves disruption by mechanical distension, collagenase perfusion or mechanical or enzymic digestion (Scharp, 1984) followed by purification by sedimentation with or without density gradient, hand-picking, elutriation, FACS, electrophoresis or a f f i n i t y techniques using anti - i s l e t antibodies. An auto-analyzer which employs ce l l culture to form aggregates of predominantly i s l e t c ells and elutriation has been used with some success for dog transplants (Scharp, 1984). A similar automated method employing a f i c o l l gradient is currently in c l i n i c a l t r i a l s (Ricordi et al, 1989). Other relevant developments in is l e t c e l l isolation include the development of MAbs to human islet cells (Soon-Shiong et al, 1988), the discovery of lectins that bind i s l e t cells (Peterson et al, 1986), the fact that addition of nicotinamide to i s l e t c ells in culture increases i s l e t DNA replication (Sandler et a l , 1989) and the encapsulation of i s l e t cells in agarose resulting in a longer lifetime on transplantation (Iwata et al, 1989). The most commonly used method to enrich i s l e t c e l l s i s long-term culture which forms aggregates of i s l e t c e lls which can be separated by density gradients (Sandler et al, 1985). Since MAbs to the islet cells exist i t seems that this separation may be feasible using immunoaffinity partition. A model separation such as that outlined below of cells distinguished by low surface density antigens recognized by monoclonal antibodies is relevant to this problem as well as to the bone marrow purging problem. 8. The model separation problem. The separation problem chosen for this study was of two sub-sets of a Moloney-virus transformed T lymphocyte c e l l line, MBL-2 (2.6) and MBL-2 (4.1) (Takei, 1983; Chan and Takei, 1986). The only known difference between these two c e l l lines is the surface density of an antigen recognized by the rat monoclonal antibody, YE1.48.10. Prior to the development of this antibody these two sub-sets of MBL-2 were unknown and the only currently available method of separating these two sub-sets is by FACS. The MBL-2 cells and YE1.48.10. 24 hybridoma were gif t s from Dr. F. Takei, Terry Fox Laboratory, Vancouver. Thus the specific objective of this study is to examine a variety of novel immunoaffinity ligands and to apply them to separate populations of MBL-2 (2.6) and MBL-2 (4.1) cells by immunoaffinity partition using the rat monoclonal antibody YE1.48.10 as the primary a f f i n i t y ligand. The dye trypan blue, a system u t i l i z i n g biotin, avidin and streptavidin and polyacrylamide-MAb ligands are a l l examined. The trypan blue and polyacrylamide derivatized antibodies are f i r s t used on an erythrocyte separation problem and the polyacrylamide antibody is then applied to the separation of the MBL-2 lymphocytes. A schematic diagram of the approaches studied in this thesis is shown in Fig. 2. 25 CHAPTER 2 THEORETICAL ASPECTS OF PARTITIONING 1. Phase Separation. The s t a t i s t i c a l mechanical mean f i e l d theory of phase separation f o r any number of polymer species i n a common solvent i s based on the approach f i r s t taken independently by F l o r y (1941) and Huggins (1941). The method i s to calculate the free energy of mixing polymer and solvent molecules, AG , from the sum of the enthalpy and m entropy of mixing (AH and AS ). The theory i s based on a l a t t i c e i n m m which s i t e s can be occupied by a solvent molecule or a segment of a polymer molecule. In the case of two polymers and a solvent, the expression turns out to be (Flory, 1953): AG = kT [n In* + n In* + n In* + m 1 r l 2 2 3 3 (n + n P +-n P ) (* * x + <f> <t> X + <f> <t> X ) 1 (2) 1 2 2 3 3 r l r 2*12 r l r 3 * 1 3 r 2 r 3 * 2 3 where subscript 1 denotes the solvent and 2 and 3 denote the two polymer species (characterized by t h e i r molecular volumes), * i s the volume f r a c t i o n of component i (* = n P /[n + n P + n P ] ) , n i s the number of M 1 1 1 2 2 3 3 1 molecules of i on the l a t t i c e , P i s the number of segments of volume equal to a solvent molecule per polymer molecule and £ i s the F l o r y i n t e r a c t i o n parameter between i and j . xi represents the maximum i n t e r a c t i o n energy a segment of molecule i can possess i n a mixture, i . e . when completely surrounded by segments of molecule j . By d i f f e r e n t i a t i n g AG m with respect to n at constant temperature and pressure f o r any of the species the chemical p o t e n t i a l of that species may be calculated, i . e . (u - u°) = N (SAG / 3n ) where u i s 1 1 A m I n 1 J the chemical p o t e n t i a l of i when i t s volume f r a c t i o n i s $ , i s the standard state chemical p o t e n t i a l of i when *= 1 and N a i s Avogadro's number. If two phases are to be present at equilibrium then the chemical p o t e n t i a l , u, must be i d e n t i c a l i n either phase. The c r i t i c a l conditions f o r the appearance of two phases are described by (Flory, 1953): 5u = 3 2u = 0 (3) l ' l a*2 a*2 26 The character of the s o l u t i o n may be seen f o r a s i m p l i f i e d case as follows. Solving t h i s expression for c r i t i c a l values of 4>l a n ^ X{^ i . e . , those which w i l l produce phase separation i n a two polymer, sin g l e solvent system, assuming that P^= P , i.e. both polymers have the same number of segments per molecule, and that both polymers are equally soluble i n the solvent (implying * = xi3) r e s u l t s i n : <f> = <p = (1-0 )/2 (4) 2C 3C 1C ^23C 2 2C Where c denotes c r i t i c a l values. These expressions describe the general features of phase separation i n an APTS. That i s , phase separation w i l l occur r e a d i l y since P 2 i s large fo r high molecular weight polymers. Hence, phase separation i s the rule rather than the exception implying only small p o s i t i v e (unfavorable) value of * 2 3 c i - s necessary to produce separation i n mixtures of polymers. The ease of separation w i l l increase with increasing molecular weight, and since a; and ^ are absent from these expressions, only polymer-polymer int e r a c t i o n s are important i n determining phase separation i n systems of equally soluble polymers. The range of concentrations above which a two phase system w i l l a r i s e as defined above may be represented by a phase diagram (Fig. 3). The curved l i n e d i v i d i n g the two areas of the phase diagram i s termed the b i n o d i a l and a l l compositions of, in t h i s case, PEG and dx above the b i n o d i a l w i l l produce two-phase systems. Points along the b i n o d i a l or nodes (e.g. B,B' , C,C ) can be joined to form t i e l i n e s (eg.B-C, B ' - C ) . Any point along the t i e l i n e w i l l give r i s e to a two-phase system with the same phase composition but the phase volumes w i l l d i f f e r . The t i e l i n e length (TLL) becomes zero at the c r i t i c a l point, K, and t h i s defines the minimum polymer concentrations which w i l l phase separate. The %w/w r a t i o of bottom to top phase i s equal to the r a t i o between those parts of the t i e l i n e defined by the t o t a l system composition (eg.AC:AB). 27 Dx (%w/w) F i g . 3. General phase diagram f o r a PEG/dx/water phase system. The l i n e s BC or B'C connect the p o i n t s r e p r e s e n t i n g the composition of the two phases at e q u i l i b r i u m i . e . B represents the bottom phase c o m p o s i t i o n and C the top phase comp o s i t i o n . The t o t a l composition i s g i v e n by A and the r a t i o of AC:AB w i l l g i v e a weight r a t i o of bottom to top phase. K i s the c r i t i c a l p o i n t . 28 The process of phase separation does not occur instantly but takes a significant time depending on density and viscosity differences between the phases and the time taken for droplets to coalesce into larger drops (Raymond and Fisher, 1981). It has been described in terms of the movement of complex microphases and is similar to the upward and downward creaming of emulsions (Becher, 1965). The rate of phase separation increases with TLL for moderate concentrations which seems to relate to density differences, interfacial tension and viscosity differences which in turn depend on. the polymer composition. Systems close to the c r i t i c a l point (K) are slow to separate because of their small density difference. 2. Molecular A f f i n i t y Partition. If the Flory-Huggins expression (Eq. [2] ) for the free energy of mixing is applied to a four component system, the fourth component considered to be the partitioned molecule and a l l the components are considered equally soluble, then the chemical potential, u , may be calculated. At equilibrium u must be identical in 4 4 both upper and lower phase so by equating u for each phase and assuming <p to be small so second order terms can be dropped, an expression for 4 K. the partition coefficient of the fourth component, is obtained 4 (Brooks et al, 1985): K = exp P [(**-/) (1-* ) + (^V)(l/P -X ) + (£V)(1/P -> )] (6) 4 4 1 1 14 2 2 2 24 3 3 3 34 T B where K = <p /</> and T and B denote top and bottom phases. 4 4 4 This illustrates several features of partitioning which are experimentally observed. That i s , the partition coefficient depends exponentially on the properties of the phase system and the partitioned material and will become more one-sided with increasing molecular weight of the partitioned material or increasing difference in polymer concentration between the phases. The partition coefficient w i l l be determined by the balance of the interaction energy of the partitioned material with the phase polymers and the solvent. If the molecular weight of one of the phase polymers is decreased, then the partitioning into that phase w i l l be increased. 29 It must be recognized, however, that the above concentrated s o l u t i o n Flory-Huggins theory i s limited i n that i t assumes there i s no volume change on mixing nor any entropy change (other than c o n f i g u r a t i o n a l ) associated with segment-solvent contacts. There are also general, model-independent, thermodynamic expressions which can be used to describe p a r t i t i o n behavior (Brooks et al, 1985). Consider the chemical p o t e n t i a l of an uncharged solute molecule of species i , , i n a s i n g l e phase, treated as a continuum with the average properties of the s o l u t i o n : Hi = u° + RT l n a i (7) where u° i s the standard state chemical p o t e n t i a l i n that phase, R i s the gas constant and a i s the a c t i v i t y of i i n that phase (a = f c ) & 1 J ^ l l i where f i s the a c t i v i t y c o e f f i c i e n t and c i s the concentration of i J l species i ) . At equilibrium must be the same i n e i t h e r phase, therefore: U 0 T+ RT l n a T = u° B+ RT l n a B (8) l l l i If s u f f i c i e n t l y d i l u t e solutions are considered then the a c t i v i t y c o e f f i c i e n t approximates to one and the above expression may be considered i n terms of concentration rather than a c t i v i t y . Solving f o r T B the p a r t i t i o n c o e f f i c i e n t , K = c /c , r e s u l t s i n : v i l l v , 01 OB, Ki = exp -(u t -Uj ) (9) RT By equating t h i s with Eq.[6], the F l o r y Huggins expression f o r , i t i s apparent that i n t h i s i n t e r p r e t a t i o n the standard state chemical p o t e n t i a l d i f f e r e n c e of the p a r t i t i o n e d molecule i n the two phases contains contributions from the energies of i n t e r a c t i o n between the solute and a l l three components of the phase systems. If a ligand associates strongly with another molecule and i f the ligand has a d i f f e r e n t p a r t i t i o n c o e f f i c i e n t from the molecule, the binding w i l l change the p a r t i t i o n c o e f f i c i e n t of the complex. This i s the basis of a f f i n i t y p a r t i t i o n (Flanagan and Barondes, 1975) and the use of phase 30 systems in studying association reactions (Miiller and Gautier, 1975; Hustedt and Kula, 1977). It can be considered simply as a way of OT OB altering u j and thereby changing K^ . A more detailed treatment of a f f i n i t y partition can be provided, as follows (Brooks et al, 1985). If a macromolecule has n equivalent, independent binding sites for a ligand, then the binding of ligand to the macromolecule can be analyzed by considering the sequential addition of ligand up to saturation. If denotes the concentration of molecules which have i of their sites occupied by ligand at a given equilibrium concentration, L, then the total concentration of macromolecule, M is t o t n the sum of a l l the species of M , that is M = Z M . This may be i t o t 1=0 l J evaluated by applying the binomial theorem (Cantor and Schimmel, 1980) resulting in: n Z M = M (1 + k L ) n (10) 1=0 1 0 a where k is the microscopic association constant for ligand binding to any site. n n Since K = Z M1/ Z MB, then K = K (1 + k T L T ) n (11) m 1=0 1 1=0 1 m 0 a (1 + k V ) n a T B where K is M /M , the partition coefficient of empty macromolecules. o o o T B k ,k are the microscopic association constants in top/bottom a a phase. T B L , L are the equilibrium concentrations of ligand in top/bottom phase. If the ligand concentration is very high, a l l molecules w i l l be saturated and Eq.[11] wi l l simplify to (Flanagan and Barondes, 1975): K = K K n (k T/k B) n (12) m 0 L a a In the case where the ligand is covalent ly bound, as in a PEG-derivatized protein, the association constant becomes inf i n i t e in 31 both phases, therefore K = K K.N. These expressions (11) and (12) m 0 L predict a very strong effect of ligand binding on partition coefficients providing K l is not equal to 1. 3. P a r t i c l e p a r t i t i o n . The above theory for molecular partition cannot be applied to c e l l and particle partition since the concept of chemical potential for a particle as large and slowly diffusing as a c e l l does not really apply. Moreover, i t is generally observed that c e l l s distribute between the interface and one of the phases rather than between the two bulk phases. Therefore, numbers of c e l l s at the interface and in the phases rather than concentrations are the more appropriate measure of c e l l number. One possible starting point for developing a theory for c e l l and particle partition is the Boltzmann equation (Brooks et al, 1985). K = n /n = exp (-AG°/kT) (13) 1 2 This relates the probabilities of a particle being in either of two compartments, designated 1 and 2, to the standard state change in free energy required to move the particle between the compartments, AG°. The use of kT in the expression assumes that the particle diffuses freely and is distributed by thermal energies. For large particles such as cell s , however, a significant reduction in free energy can result from adsorption at the interface. If a particle is located at the liquid-liquid interface, the interfacial area is reduced by the cross-sectional area of the particle. The decrease in area produces a proportionate decrease in free energy equal to the interfacial tension times the area lost, thus stabilizing particles adsorbed at the interface. If a particle has equal a f f i n i t y for both phases, then adsorption at the interface w i l l be significant when the free energy associated with adsorption is of the order of the average thermal energy of a particle. For example, a typical phase system with an interfacial ~*3 2 tension 5x10 erg/cm would adsorb particles with a diameter over 320A at the interface. The greater the surface tension, the smaller the particles that w i l l adsorb at the interface. The work done on moving a particle from the interface into the top 32 phase, A G t i , i s the sum of two components: (i) the transfer of part of the particle surface of area A from the bottom phase to the top phase, B with a net energy change of -A Ay; ( i i ) the change in surface area of B the interface by A , with a net energy change of A y . Thus: TB TB TB AG = A r - A Ar (14) T I TB TB B where is the surface free energy at the interface, Ay is the particle c e l l surface free energy difference in top and bottom phase, A r is a contribution from the net energy change associated with the TB TB change in interfacial surface area, and A iy is a contribution from the B energy required to transfer a portion of the particle surface area, A B , from bottom phase to top phase. Equation [14] has been tested v i a the microscopic measurement of the contact angle, 6, formed between the interface and the tangent to the c e l l surface at the contact line (Adamson, 1976). In terms of the contact angle, the expression for particle partition is as follows (Brooks e t al, 1985): In K = -AG /kT (15) T I In K = -7 7 r a 2 ( l - cos e ) 2 (16) TB kT The contact angle is related to the free energies in the system by cos 0 = Ay/y , where A ? is the difference in particle surface free energy TB in top and bottom phase, r j B is the interfacial tension between top and bottom phase and a is the particle radius. When typical values are substituted into this equation, i t is found that A G t i is several orders of magnitude larger than kT, suggesting essentially no particles should partition into the upper phase. Experimentally this is not observed. However, i t would s t i l l be expected that particle partition would depend on the surface properties of the particle, their size and area, the temperature and the interfacial tension between the phases. In fact an exponential dependence of K on A G ^ has been observed (Sharp, 1985). 33 4. Particle a f f i n i t y partition. The following is taken from Sharp and Brooks (1989). An expression for the effect of an a f f i n i t y ligand binding to a particle on the surface free energy difference, Ay, of the particle may be obtained by integrating the Gibbs equation: dy = -Z r dut ,(17) where r is the surface excess of the ith component and du^ is the change in chemical potential of the ith component. As eqn [17] applies in both phases then the surface free energy difference between a particle in top and bottom phase may be written as: dAy = z r B du B - z r T du T (18) l l l l If i t is assumed that the binding of ligand does not significantly alter the particle area and the ligand is the only significant contribution to the integral of eqn [18] then: dAy = T B du B - T T du T (19) L L L L where subscript L refers to the ligand. The chemical potential of the ligand in the top phase is u T = u°T+ kT ln c T and the differential i s : du T = kT (20) T C T The surface excess of ligand in the upper phase, r , is the amount of ligand bound per unit area. The simplest binding isotherm which may be used to describe this is the Langmuir isotherm for n identical, independent binding sites per unit area (Cantor and Schimmel, 1980): n = nc k (21) ~, T, T 1 + c k T where k is the association constant for the binding reaction. The integral of the 2nd term in equation [19] i s : T T c c T , T R T T = j. nkTk dc o o ( 1 + C k ) 34 / - / = nkT In (1 + c V ) L o (23) Repeating t h i s f or the lower phase and s u b s t i t u t i n g into eqn [19] gives: Ay - Ar = nkT In (1 + c V ) (24) L o (i + c Y ) Using eqn [21] to substitute for the terms i n parentheses i n eqn [24]: Ar - Ar = nkT In ( n T k B ) (25) L o ( n B k T K L ) Experiments using PEG-palmitate and erythrocytes were consistent with t h i s theory (Sharp, 1985). 5. Counter current d i s t r i b u t i o n (CCD). CCD i s a method for carrying out repeated p a r t i t i o n steps and i s useful f o r separations i n which one p a r t i t i o n i n g step does not s u f f i c i e n t l y resolve the components. It was o r i g i n a l l y designed f o r aqueous-organic or organic-organic two-phase systems (Craig, 1960) but was modified by Albertsson (1965) who developed a thin-layer CCD apparatus with f a s t e r s e t t l i n g times. This made the technique more appropriate f o r viscous, polymer-containing phases and b i o l o g i c a l separations. A CCD apparatus consists of multiple chambers i n a c i r c u l a r arrangement. By r o t a t i o n of the top h a l f of the apparatus i n a clockwise d i r e c t i o n while the bottom h a l f remains stationary, the upper phase i s transferred to the chamber immediately to the r i g h t as shown i n F i g . 4, while the upper phase i n the chamber to the l e f t replaces the upper phase that has been transferred. The three parameters useful in p r e d i c t i n g CCD curves are P, the f r a c t i o n of the t o t a l amount of material i n the upper phase, K, the p a r t i t i o n c o e f f i c i e n t , which i s the r a t i o of the concentrations of the d i s t r i b u t e d material and G, the d i s t r i b u t i o n r a t i o , which i s the r a t i o of the masses of the material of i n t e r e s t i n the phases. These are r e l a t e d as follows: 35 G = KV / V (26) t b P = G/ (G + 1) (27) A schematic diagram i n F i g . 4 depicts stages i n the generation of a f i v e transfer CCD s t a r t i n g with 1000 units of a material with a p a r t i t i o n c o e f f i c i e n t (K) of 1. Materials with K other than 1 w i l l produce asymmetric d i s t r i b u t i o n curves. It i s r e l a t i v e l y straightforward to predict t h e o r e t i c a l CCD curves providing the d i s t r i b u t i o n r a t i o , G, of the solute or p a r t i c l e i s known. 500 500 250 250 250 250 125 250 125 125 250 125 63 188 188 63 63 188 188 63 31 125 188 125 31 31 125 188 125 31 F i g . 4. A diagrammatic representation of 5 CCD transfer steps f o l l o w i n g the loading of 1000 units of a soluble sample (K=l, G=l, P=0.5). From T r e f f r y et al, 1985). The p a r t i t i o n i n g of solutes occurs between the two phases but the p a r t i t i o n of p a r t i c l e s i s between the interface and one of the phases (Fig. 5). In order to apply these parameters as defined i n (26) and (27) 36 to predict CCD curves; the physical separation of the phases must be at the interface. This is possible when solutes are being partitioned but not for particles when a small volume of, in the case of an upper phase partition, upper phase is left behind to ensure that no particles adsorbed at the interface are carried over. Thus the amount transferred is C (V-v), i.e., the number in the upper phase minus that which is not transferred, and the amount remaining stationary is a + Ctv, i.e., the number adsorbed at the interface and the number in the upper phase that is not transferred, (see Fig. 5). Then the distribution coefficient, G, w i l l be: G = C t(V t-v) (28) a + C v t For n transfers the distribution of solute is described by (1+G)n or (1-P) n 1 which may be expanded using the binomial distribution (Hecker, 1955). From this P may be calculated and the CCD curve predicted. If n distributions are carried out the fraction of the total population appearing in the rth cavity is given by: F(r) = n! P r ( l - P ) n _ r (29) r!(n-r)! The peak of the distribution w i l l be defined as those locations in which two adjacent cavities contain equal amounts I.e. F(r ) = F(r ). m ra+1 Applying (29) to this equation and solving gives: r = nP (30) ID Hence, from eqn [27]: r = nG/ (G+l). (31) m 37 L i q u i d - l i q u i d D i s t r i b u t i o n L i q u i d - i n t e r f a c e D i s t r i b u t i o n K = C / C t b G = C V t t C V b b V (moving layer) b (stationary layer) G = C (V -v) a + C v t V -v (moving layer) (stationary layer) where a i s the number of p a r t i c l e s adsorbed at the interface, C i s the concentration in the upper phase, V i s the volume of the upper phase, v i s the volume of upper phase not transferred i n p a r t i c l e p a r t i t i o n , G i s the d i s t r i b u t i o n r a t i o and K i s the p a r t i t i o n c o e f f i c i e n t . F i g . 5. Differences between the d i s t r i b u t i o n type i n l i q u i d - l i q u i d CCD (a) and l i q u i d - i n t e r f a c e CCD (b). To the l e f t no s i g n i f i c a n t i n t e r f a c i a l adsorption takes place and the whole top phase i s transferred at each step. To the rig h t , the d i s t r i b u t i o n takes place between the top phase and the interf a c e . The bottom phase, the in t e r f a c e and a small layer above the interface together form the stationary layer which i s not transferred. From Albertsson and Baird (1962). 38 CHAPTER 3 BACKGROUND TO THE SEPARATION PROBLEM. 1. History of the c e l l lines used for the separation. A variety of cultured c e l l lines and products thereof form the basis of this study. The rat monoclonal antibody specific for MBL-2 cells, YE1.48.10., used as the primary ligand throughout this study was raised against a mouse lymphocyte line resulting from the fusion of a Con-A activated normal mouse spleen c e l l and a transformed mouse T c e l l (EL-4BU). A "transformed" c e l l is one that has undergone a stable heritable change that usually enables i t to grow into a tumor in an appropriate recipient animal. As the animal from which a c e l l is derived may not be available for tumorgenicity testing in vivo other tests have been developed although nude mice are s t i l l frequently used. One test is the a b i l i t y of transformed cells to grow in the absence of substratum (i.e. in culture) or the rapid metabolism of glucose exhibited shortly after transformation. Some other changes commonly observed on transformation are growth to an unusually high c e l l density, lowered requirement for growth factors in serum and they are less anchorage dependence. There are also abnormalities in adherence and plasma membrane related properties. Transformed cells generally have a different morphologic appearance than normal cells in culture. They are often more rounded and are usually covered with microvilli and lamellopodia. It is possible to transform almost a l l the cells in a culture into cancer cells within a few days with RNA tumor viruses. Cells can also be transformed with chemical carcinogens. Concanavalin A (con-A) is a mitogen commonly used to activate lymphocytes. When confronted with a stimulus such as con-A, the resting lymphocyte undergoes changes in virtually every aspect of cellular metabolism culminating in DNA synthesis and mitosis. In vivo this only occurs in cells having specific mechanisms for antigen recognition but in vitro a large number of cells can be stimulated in this way. It appears that mitogens mimic the action of antigens on lymphocytes. However, while an antigenic determinant stimulates only a very small proportion of a l l lymphocytes, mitogens stimulate a large number of 39 them. The mechanism of mitogenic stimulation is thought to involve aggregation of con-A receptors to form small clumps on the c e l l surface. These receptors are proposed to be transmembrane-1inked to the cytoskeletal system which results in cytoplasmic changes in the c e l l leading to changes in metabolism (Golub, 1977). The fusion of EL-4BU and the con A activated spleen c e l l resulted in ECA17.9.8, a doubly cloned hybrid. The technique for fusing cells with polyethylene glycol is also used to produce hybridoma lines secreting monoclonal antibody. Aqueous solutions of PEG at a sufficiently high concentration will fuse almost a l l c e l l types . The fusing concentration of PEG is usually above 30%w/w and the optimum concentration is determined by the balance between increasing fusion efficiency and minimizing c e l l damage. The mechanism by which PEG causes fusion is not known with certainty. The concentration dependence seems to be related to the a b i l i t y of PEG to decrease the water activity of the solution since solutions inducing maximum fusion have vi r t u a l l y no water unassociated with the polymer. Many other polymers and viruses, in particular Sendai virus which was used in early hybridoma technology, w i l l induce c e l l fusion. The primary ligand YE1.48.10., was a rat MAb raised against ECA17.9.8. cells (Takei, 1983). The spleen cells of a rat immunized with ECA17.9.8. cells were fused with rat myeloma Y3 cells (Galfre and Milstein, 1981). The hybrid culture supernatants were selected for ECA17.9.8. specific clones with an indirect binding assay using radio-iodinated (F* ^) fragment of anti-rat IgG. A l l positive hybrids were doubly cloned in soft agar. Cloning ensures that the hybridoma line is descended from a single c e l l . In this case dilute concentrations of cells were seeded into medium solidified by agar and c e l l s forming colonies were transferred individually to culture dishes. The MAbs produced by the resulting hybridomas were tested for their reactivities with various lymphoid cells. One of the antibodies, YE1.48.10., an IgG2b, reacted strongly with two transformed T c e l l lines EL-4 and MBL-2. EL-4 is a chemically induced T-leukemia c e l l " line of B6 origin (B6 refers to the mouse strain). A 5-bromo-2-deoxyuridine sub-line of EL-4, EL-4BU was one of the fusion partners in the hybrid c e l l line used to generate the YE1.48.10. hybridoma. MBL-2 is a Moloney 40 virus-induced T-leukemia c e l l line, also of B6 origin. The antigen recognized by YE1..48.10. has been immunoprecipitated from EL-4 cel l s and is a dimer of 90kD (non-reduced). There are approximately 1.5-3 x 105 antigen molecules per EL-4 c e l l as measured by binding of radiolabeled YE1.48.10. The antigen i s also expressed on normal T-cells but is hidden, not exposed as on EL-4 and MBL-2 cells. The antigen may be exposed on normal T-cells by non-ionic detergent solublization so i t is likely to be hidden by by membrane-associated molecules rather than chemical modifications such as glycosylation (Chan and Takei, 1986). The MBL-2 cells were resolved into two sub-populations, MBL-2(2.6) and MBL-2(4.1) by FACS on the basis of a higher surface antigen density on MBL-2(4.1) cells as defined by YE1.48.10. (Takei, 1986). 2. Structure of Immunoglobulin G (IgG). IgG (Fig. 6) is a symmetrical molecule made up of two identical glycosylated heavy chains (MW = 50 000-75 000) and two identical non-glycosylated light chains (MW « 25 000). The heavy chains are joined by two disulfide bonds in the non-variable or F c region and each light chain is joined to a heavy chain by a disulfide bond. Enzymic digestion by papain s p l i t s IgG into one F c and two F & b fragments. The F a b fragment consists of one light and part of a heavy chain and includes the variable region (F ) which binds antigen, thus varying in antibodies of different s p e c i f i c i t i e s (Goding, 1983). The F c fragment consists of only heavy chain and is constant,in any one class of IgG irrespective of antibody specificity as i t is not involved in antigen binding. Papain digestion produces fragments with retention of biological activity making i t preferable to chemical degradation. 41 Fig. 6. Structure of the IgG molecule. Reduction of the d i s u l f i d e cross links w i l l s p l i t the molecule into heavy and light chains. The enzyme papain w i l l cleave the molecule into F a f a and F c fragments during limited digestion. 42 3. A v i d i n - B i o t i n reaction. One of the ligand combinations u t i l i z e d in this study is a biotin-derivatized antibody and PEG-avidin (Fig. 2.d.) Biotin (Fig. 7, vitamin H) was f i r s t encountered as early as 1916 when i t was observed that diets high in raw egg white were toxic to rats (Batemann, 1916). Later the symptoms of egg white injury in rats were described as muscle incoordination, dermatitis, hair loss and nervous problems (Boas, 1927). It was also noted that cooked egg white did not have any effect and that liver, yeast and other foods protected rats against raw egg white toxicity. The protective substance was called vitamin H (Gyorgy et al, 1940). Previously a potent growth stimulant for yeast had been identified and named biotin (Kogl and Tonnis, 1936). In 1941 the structure of biotin was elucidated, synthesized and confirmed to be vitamin H (Gyorgy et al, 1941). At the same time the presence of avidin in egg white, a basic glycoprotein with anti-vitamin H ac t i v i t y was confirmed. Dietary deficiencies of biotin are rarely observed, however, due to it s synthesis by intestinal bacteria (although the daily requirement is estimated at 0.01 to 0.2 mg). A deficiency of biotin results in reduced activity in urea synthesis, purine synthesis, carbamylation and tryptophan catabolism. This is not accounted for by any known biotin enzymes nor is the above effect observed on administration of avidin. Biotin is covalently linked to the enzyme (e.g. a carboxylase) through the e-amino group of a lysine residue resulting in a f a i r l y long, flexible attachment (Fig. 7). The biotin-lysine peptide has been isolated by hydrolysis of biotin-containing enzymes and is known as biocytin. The most important function of biotin is as a coenzyme for a variety of carboxylatlng enzymes. One of the most important enzymes for which biotin is a coenzyme is pyruvate carboxylase which interconverts pyruvate and oxaloacetate. The biotin becomes carboxylated at N only when acetyl CoA is bound to the pyruvate carboxylase. This allosteric activation is an extensively studied physiological control mechanism. 43 Fig. 7. The structure of biotin. Other enzymes for which biotin is a cofactor are acetyl CoA carboxylase, which is involved in fatty acid biosynthesis, and propionyl CoA carboxylase which metabolizes propionyl CoA, a breakdown product of fatty acids with odd numbers of acyl carbons. The reactions involved are summarized below. ATP, co -» ADP + p CH CO-SCoA - -> HO CCH CO-SCoA 3 b i o t i n , A c e t y l CoA 2 2 . i . c a r b o x y l a s e ,. , , _ , Acetyl CoA ' Malonyl CoA ATP, CO -» ADP + P CH CH CO-SCoA - ! > HO CCH(CH )C0-SCoA 3 2 b i o t i n , p r o p i o n y l CoA 2 3 carboxy1ase Propionyl CoA Methyl Malonyl CoA Biotin contains fused imidazole and thiophene rings as illustrated in Fig. 7. and is bound tightly by avidin (k=1021 M - 1). This strong biotin-avldin interaction is used extensively in c e l l - l a b e l l i n g procedures and various immunoassays. 4. Polyacrylamide-protein g r a f t copolymers as ligands. Polyacrylamide (PAA) is highly hydrophilic compared to most non-ionic water soluble polymers. For instance, i t has a partition coefficient of 18 in favor of the water-rich phase in a water/phenol two phase system (Dobry, 1956). Before this work was carried out its behavior in dx/PEG systems was not known but i t seemed l i k e l y that PAA would partition in favor of the 44 dx-rich phase. This supposition is based on the use of PAA-coated chromatographic supports to bind stationary dx-rich phase for use in liquid-liquid partition chromatography (Muller, 1986). Only small surface coverage of the bead with PAA was required to immobilize large volumes of stationary phase. In the work to be described, polymerization of acrylamide was carried out via a free radical mechanism. It is known that some eerie salts such as nitrate and sulfate form effective redox systems in the presence of organic reducing agents such as alcohols, thiols, glycols, aldehydes and amines. This produces cerous ions and free radicals capable of in i t i a t i n g vinyl polymerization. The eerie salts form complexes with alcohols and glycols and the disproportionation of these complexes is the rate determining step (Duke and Forist, 1949; Duke and Bremer, 1951). The mechanism of the ini t i a t i o n reaction for alcohols can be written generally as: Ce I V + RCH OH , B • Ce" 1 + H+ + RCHOH 2 IV where Ce represents the aqueous eerie complex, B the ceric-alcohol complex and RCHOH a free radical which initiates polymerization in the presence of a vinyl monomer. If the reducing agent is a macromolecule and the reduction is carried out in the presence of a vinyl monomer, the free radical produced on the macromolecule initiates polymerization to produce a polymer graft. This yields predominantly grafted polymers since the free radicals are formed preferentially on the backbone. In the absence of a reducing agent, eerie salts w i l l i n i t i a t e acrylamide polymerization at a very slow rate. For instance, in the case where polyvinyl alcohol was the reducing agent the i n i t i a l rate was 3.0%/minute compared to 0.04%/min in the absence of reducing agent (Mino and Kaizerman, 1958). This means that there w i l l be a small amount of free polyacrylamide (PAA) in the PAA-macromolecule solution. In the present work polyacrylamide was grafted onto the primary and secondary hydroxyls of IgG to form a lower phase partitioning immunoaffinity ligand. 45 CHAPTER 4. METHODS AND MATERIALS. 1. Fragmentation of mouse IgG. This is an adaptation of the methods of Utsumi (1969) and Porter (1959). Mouse IgG (Sigma, 1 mg/mL, 1 mL) in Na HPO / NaH PO buffer (0.01 M, pH 8.0) containing ethylene-2 4 2 4 diaminotetraacetate (EDTA, 2 mM) and dithiothreitol (1 mM) was incubated with activated mercuripapain (BDH, 0.1 mg/mL, 0.1 mL, 15 minutes, 37°, buffer as above) for one hour. The reaction was stopped with lodoacetamide, (final concentration 20 mM) and kept in the dark at 0° for one hour. The mixture was dialyzed against sodium acetate (NaAc, 10 mM, pH 5.5) at 4°, then fractionated on a mono Q anion exchange column coupled to a Pharmacia Fast Protein Liquid Chromatography (FPLC) apparatus with an elution gradient of 0.01-1 M NaAc (pH 5.5). The fractions were identified by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). 2. Production of sheep anti-(mouse F c fragment) IgG (s a mFc)- A sheep was injected subcutaneously with mouse F c fragment (50 jug), obtained as described above, in complete Freunds adjuvant 1 (1:1 emulsion). Two subsequent injections at two week intervals were given in incomplete Freunds adjuvant 1 After one week, blood was drawn and subsequent blood drawing was preceded by a booster injection one week beforehand. The blood was allowed to clot at room temperature and the serum obtained by centrifugation and decantation. A solution of saturated (NH ) SO was 4 2 4 added over 10 minutes to the slowly stirred serum at room temperature. At about 40% v/v saturated solution, most of the immunoglobulins were precipitated (Goding, 1983). This was confirmed by SDS-PAGE. The 1An adjuvant is a substance which augments immune responses in a non-specific manner. Freund's complete adjuvant is a water in o i l emulsion in which k i l l e d and dried M. tuberculosis bacteria are suspended in the o i l phase, Freund's incomplete adjuvant omits the bacteria (Herbert, 1973). 46 suspension was stirred for a further 30 minutes. This was followed by centrifugation at 10 000 g for 10 minutes at 4° and the pellet washed three times with 40% saturated (NH ) SO solution. The pellet was 4 2 4 ^ dissolved in phosphate buffered saline (PBS, 16.7 mM Na2HP04> 3.3 mM NaH2P04> 130.4 mM NaCl, pH 7.2), dialyzed against water, the precipitate discarded and the solution redialyzed against Na HPO / NaH P0 (10 mM, . ° 2 4 2 4 pH 8.0). The solution was chromatographed on a mono Q anion exchange column using FPLC with a 0-1 M NaCl ionic strength elution gradient. SDS-PAGE was used to identify the fractions. 3. Monoclonal antibody production via ascites f l u i d . This method (Goding, 1983) was used to produce mouse anti-NN glycophorin (mctNN glyc) from the hybridoma c e l l line NN-5 (American Tissue Culture Collection HB 8476). BALB/c mice were injected intraperitoneally (i/p) with 2,6,10,14-tetra-methylpentadecane (pristane, 0.5 mL per mouse) and 3-4 days later injected with 106 hybridoma cells i/p in 1 mL of serum-free medium. After 7-10 days, the ascites f l u i d was drained by inserting a needle i/p and this was repeated every 2-3 days until death of the mouse. Each mouse typically yielded 5-6 mL of f l u i d containing 5-15 mg/mL of antibody. The ascites was centrifuged and the supernatant stored at -20° until purification. The mctNN glyc was purified by (NH ) SO precipitation as described earlier followed by FPLC on a mono 4 2 4 Q column with a 0.01-0.3 M Na HPO / NaH P0 , pH 8.0 ionic strength 2 4 2 4 ^ elution gradient. The peaks were identified by SDS-PAGE. 4. Sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE). The method used for slab gels was an adaptation of the methods of Davis (1964) and Ornstein (1964). The stock solutions were made up as follows: 1. Acrylamide:bis (30:0.8). Acrylamide solution (60 g in 200 mL of water) was heated to 60°, then maintained at 56° for 30 minutes in the dark. The solution was f i l t e r e d through Whatman #1 f i l t e r paper and N,N'-Methylene-bis-acrylamide (bis) (1.6 g per 200 mL) added. The solution was stored at 4° in the dark. 2. 1.875 M [Tris(hydroxymethyl)aminomethane, pH 8.8 (trls-base). 3. 1.0 M [Tris(hydroxymethyl)aminomethane hydrochloride], pH 6.8 47 (tris-HCl). 4. Sample preparation solution (10% SDS). The following were combined: SDS (0.41 g), tris-HCl (1.28 mL), water (12.3 mL) and bromophenol blue dye to color. 5. Running buffer. Glycine (57.6 g) and SDS (4 g) were added to 4 L of tris-base (0.05 M), pH 8.3. 6. Coomassie stain. Coomassie b r i l l i a n t blue dye (0.4 g) was dissolved in isopropanol (30 mL), water (130 mL) and acetic acid (30 mL). 7. Destain solution. An aqueous solution of 35% v/v ethanol and 10% v/v acetic acid was used. Acrylamide:bis (10 mL), tris-base (6 mL), EDTA (0.2 M, 0.3 mL) and water (13.4 mL) were mixed and degassed under vacuum. N,N,N',N'-tetramethylethylenediamine (TEMED, 0.015 mL), SDS (10%, 0.3 mL) and freshly made (NH ) S 0 (10%, 0.3 mL) were added and this, the 4 2 2 8 resolving gel, poured into a Bio-Rad Mini-PROTEAN II gel electrophoresis apparatus. Water was layered over the surface and polymerization was complete within 30-40 minutes. This procedure was repeated using acrylamide:bis (2.5 mL), tris-HCl (1.88 mL), EDTA (0.2 M, 0.15 mL), water (10.3 mL), TEMED (0.0075mL) and (NH) S 0 (10%, 0.15 mL) to 4 2 2 8 prepare the stacking gel. The water was removed from the surface of the running gel, a sample comb inserted and the stacking gel poured. Polymerization was complete within 30-40 minutes. The quantities quoted are for a 10% resolving gel and 5% stacking gel. Each protein sample (10-20 ug) was mixed with an equal volume of sample preparation solution and reduced with mercaptoethanol (5%) at 100° for 3 minutes in the case of reduced samples. The samples were applied to the gel (10-20 uL), the apparatus was f i l l e d with a 4:1 dilution of running buffer, air bubbles removed, and a constant current of 130 mA applied until the tracking dye reached the end of the gel. The gel was stained for 2 hours, destained over several hours followed by overnight soaking in 3% glycerol then dried for 3 hours on a Bio-Rad 543 gel drier. When rod gels were run the resolving gel was cast in 125 x 5 mm inside diameter tubes which were mounted in a Bio-Rad 150A electrophoresis chamber and run at an i n i t i a l current of 0.5 mA per tube 48 u n t i l the sample entered the gel at which point the current was increased to 8 mA per tube. When the tracking dye reached the end of the gel i t was marked with India ink and the gels f i x e d and stained as before. 5. Hemagglutination assay. S e r i a l d i l u t i o n s of antibody s o l u t i o n (50 uL) were made up ln PBS in a m i c r o t l t e r p l a t e (a PVC pla t e containing m u l t i p l e w e l l s ) . An equal volume of 1% hematocrit f r e s h c e l l suspension ( u s u a l l y erythrocytes) was added to each well. The suspensions were mixed and examined af t e r four hours. In the wells containing sub-agglutinating concentrations of antibody the c e l l s were r o l l e d down the sides of the well and s e t t l e d as a button at the bottom of the wells whereas i n -the wells containing a g g l u t i n a t i n g concentrations of antibody the c e l l s were prevented from r o l l i n g by adhesion to t h e i r neighbors and uniformly d i s t r i b u t e d over the well. 125 6. R a d i o l a b e l l i n g of proteins with I. This was done using iodobeads which are N-Chloro-benzenesulfonamide (sodium s a l t ) - d e r i v a t i z e d , uniform, non-porous, polystyrene beads (Markwell, 1982). Of a l l the amino acids that are capable of rea c t i n g with iodine i n a protein , tyrosine i s of over r i d i n g importance because of the r e l a t i v e s t a b i l i t y of the bond between the iodine and the pr o t e i n (Krohn et al, 1977). This r e a c t i o n i s summarized below. P r o t e i n s o l u t i o n (0.5 mL, 0.5 mg/mL i n PBS) was added to 2-3 iodobeads 125 (Pierce) and Na I (2 uCi), then s t i r r e d end over end f o r 40 minutes. A small a l i q u o t (^  luL) of the reaction mixture was added to a mixture of t r i c h l o r o a c e t i c a c i d (20%, 1 mL) and bovine serum albumin (BSA, 0.5 mg/mL, 1 mL). The p r e c i p i t a t e and supernatant were separated, counted R R R 49 and the fraction of bound I calculated. When the reaction was over 90% complete, usually within one hour, the free iodine was separated from the labelled protein with a G-150 Sephadex column using PBS as the elution buffer. The column fractions (1 mL) were sampled (1 uL) for counting and the f i r s t , labelled protein-containing peak was broken into aliquots and stored at -20°. The labelled stock protein solution was added to unlabelled protein to obtain the required specific activity for binding or partitioning studies. The specific activity of the stock solution was measured by absorbance at 280 nm (A 1 % = 13.5, 3 l-cm Kirschenbaum, 1973) and y counting on a LKB 1282 Compugamma y counter. In a l l cases proteins were labelled prior to derivatization with PEG, trypan blue or biotin. 7. Trypan blue-derivatization of protein. Radiolabelled protein solution (5-10 mg/mL in 0.1M NaAc, pH 4.8, 4 mL) was mixed with trypan blue solution (50 mg/mL in 0.1M NaAc, pH 4.8, 1 mL). l-Ethyl-3(3-dimethylaminopropyl)carbodiimide-HCl (EDC, 200 mg/mL in 0.1M NaAc, pH 4.8, 0.5 mL) was added and the pH of the mixture adjusted to 4.8. The mixture was rotated end over end for 4 hours. Free dye was removed by dialysis against NaCl solution (1 M) and the amount of dye bound to the 125 protein was estimated from the ratio of I to dye absorbance at 610 nm. The validity of this estimate is discussed later. 8. Synthesis of 2(alkoxypolyethyleneglycoxy)-4,6 dichlorotriazine (PCC). The usual approach to preparing protein-PEG conjugates is to synthesize an "activated" PEG which w i l l readily couple with a protein functional group. In this study PEG 1900 was activated using cyanuric chloride. The resulting product reacted mainly with protein primary amines (Abuchowski et al, 1977). The solvents were dried as follows: Benzene was d i s t i l l e d from CaH2 onto 4A molecular sieves, acetone was kept over 4A molecular sieves and petroleum ether was predried over MgS04 then d i s t i l l e d from CaH2 onto 4A molecular sieves. To synthesize PCC, NaH (0.18g) was suspended in anhydrous benzene (50 mL) and the system purged with N^ PEG-1900 (lOg) in anhydrous benzene was added over fifteen minutes. The reaction mixture was stirred for one hour under N flow and one hour 50 under a stationary Ng atmosphere. A four-fold molar excess of cyanuric chloride (3.9 g) was added and the mixture refluxed (80°) for 12 hours. The reaction mixture was poured into petroleum ether (b.p. frac. 60-80°, 150 mL) at 0° and the precipitated product collected by f i l t r a t i o n on Whatman #42 paper. The product was redissolved in a minimum of anhydrous acetone and reprecipitated with two volumes of petroleum ether at 0°. The reprecipitation was repeated three times. The PCC was vacuum dried and stored at -70° under N^. It was stable for up to 6 months or over 2 years when stored under N2- The reaction efficiency was determined by assaying the hydrolyzable chlorides of the PCC. In a l l cases PEG-1900 was used to modify proteins and PEG 8000 was used to form APTS. 9. Determination of the hydrolyzable chlorides of PCC. The Buchler Cotlove Chloride Titrator applies the coulometrlc principle of titration. Silver ions are generated at the anode which react with the chloride ions and precipitate. When a l l the free chloride has reacted, the increase in current is detected as the end point. Four samples of PCC (approx. 40 mg) were accurately weighed and two of them hydrolyzed in Na B 0 (0. 1M, pH 10) for at least 2 hours. The other two samples 2 4 7 were dissolved in the same buffer, pH 9, immediately before t i t r a t i o n . A solution of HN0 (0.1M) and CH C00H (10'/., 4 mL) and gelatin (0.62%, 0.2 3 3 mL) was added to each sample, a blank and a NaCl standard. The free chlorides were titrated using the chloridometer. 10. Reaction of the PCC with protein. Protein solution (2-8 mg/mL in Na 2B 40 7 > 0.1M, pH 9) was mixed with PCC to give a PCC: lysine molar ratio of 1:1, 3:1 or 5:1, assuming a lysine content of 124 lysines per IgG molecule (Edelman et al, 1969). After 40 minutes at room temperature the mixture was either used immediately, stored at -20° or the unbound PCC 51 and PEG removed by u l t r a f i l t r a t i o n through an Amicon PM 10 membrane or dialysis. The degree of protein primary amine modification was estimated by reaction with 4-phenylspiro[furan-2H(3H),1'-phthalan]-3,3'dlone (fluorescamine). CI CI 11. Degree of modification of the protein. The reagent 4-phenylspiro-tfuran-2H(3H),1'-phalanx]-3,3' dione (fluorescamine) reacts with primary amines to form fluorophors. The resulting fluorescence is stable over several hours and excess reagent Is hydrolyzed within a minute. Moreover, significant quantities of ethylene glycol-based detergents have no effect on the resulting fluorescence (Stocks et al, 1986). The amount of protein-bound PEG was estimated by assaying the protein primary amines with fluorescamine before and after reaction with PCC. Protein solutions were made up in sodium phosphate buffer (0.1M, pH 8), each test tube containing between 0 and 2.5 ug of protein made up to 1.5 mL. Fluorescamine (0.3 mg/mL in acetone, 3 mL) was added to each test tube while vortexing. After a minimum time of 7 minutes the fluorescence in the solution was measured on a Turner Model 430 spectrofluorometer with an excitation wavelength of 390 nm and emission detected at 475 nm. (For details see Stocks et al, 1986). 12. Enzyme-linked immunosorbent assay (ELISA). A 96 well PVC microELISA plate was coated by adsorption with antigen solution (1 pg/mL in 15 mM Na2C03> 35 mM NaHC03> pH 9.6, 200 uL per well) at 4° overnight. The plate was washed three times with PBS-Tween (0.5% Tween 80) manually or in a SLT EAW8/12 ELISA plate washer. An aliquot of a serial dilution of antibody solution was added (200 uL per well) and incubated for one hour 52 at 37°. The plate was washed three times with PBS-Tween. A solution of BSA (0.5%) in PBS-Tween was added (200 uL per well) as a blocking agent and incubated 30 minutes at 37°. The plate was washed three times and a solution of horseradish peroxidase (HRP) conjugated antibody (Sigma) specific for the primary antibody was added (1:8000 dilution of a 0.5 mg/mL stock solution), followed by incubation at 37° for 2 hours. In some cases where the primary antibody was derivatized, the HRP-conjugated antibody was specific for the antigen. The unbound conjugate was removed by washing three times and O-phenylenediamine solution (OPD, 0.04% in 25 mM c i t r i c acid, 51 mM Na HPO , pH 5.0) 2 4 containing H ^ (0.12%) added immediately before use was added to each well (200 uL per well). The color was allowed to develop in the dark until the reaction was stopped by the addition of H SO (4 M, 50 uL per 2 4 well) and the absorbance of each well measured at 492 nm with a 690 nm reference beam on a SLT EAR 400 AT ELISA plate reader interfaced to an Epson LX-800 printer. 13. Culture of hybridoma and transformed lymphocyte c e l l s . Freezing and thawing of cells: The cells were frozen by resuspending approximately 5 x 106 c e l l s in freezing medium (90% fetal bovine serum (FBS), 10% DMS0, 1 mL), followed by i n i t i a l freezing at -70° for 24 hours. They were stored in liquid N2 indefinitely. The cells were thawed over 5 minutes at room temperature, then 5 min at 37°, added to cold (4°) serum-free medium and mixed by inversion. An aliquot (100 uL) was stained with erythrosin B dye and the v i a b i l i t y of the cells estimated by counting live and dead cells on a hemocytometer. Only dead cells are stained by erythrosin B. If the v i a b i l i t y was over 80%, the cells were centrifuged (300 g, 2 minutes), the medium aspirated off and the cells resuspended in growth medium containing FBS (10-15%) to give a fi n a l concentration of 105-106 cells per mL. Incubation was at 37°, 100% humidity and 5% C02. The cells were passed at the appropriate c e l l density for each c e l l line. A l l c e l l lines were gradually transferred into 5% FBS-containing medium within several weeks. The NN-5, RG7/11.1 and MBL-2 cells were a l l cultured in RPMI-1640 (Cancer Control Agency, Vancouver, B.C.) medium but YE1.48.10. cells were started in DMEM (Gibco) then transferred into 53 RPMI-1640. 14. Supplemented serum-free medium (SSFM). RPMI-1640 medium (Gibco) containing L-glutamine (0.3 mg/mL, Gibco) and penicillin-streptomycin (50 units/mL and 50 ug/mL respectively, Gibco) was supplemented (Murakami et al, 1982) by the addition of bovine insulin (5 ug/mL, Sigma), human transferrin (35 ug/mL, Sigma), sodium selenite (2.5 mM, Sigma) and ethanolamine (20 pM, Sigma). Insulin stock solution (2 mg/mL) was prepared in d i s t i l l e d water adjusted to pH 2.5 by addition of HC1 (1 M) and transferrin stock solution (2 mg/mL) was prepared in PBS. The protein stock solutions were stored at -20° and the other stock solutions at 4°. The SSFM was made immediately before use or stored at - 4°. 15. Culture of YE1 48.10. and RG7 11.1 cel l s in SSFM. Rapidly growing cells (30 mL) adapted to growth in 5% FBS-containing medium were centrifuged (300 g, 2 minutes) and the serum-containing medium aspirated off. The cells were resuspended in SSFM (100 mL) in an 850 cm2 r o l l e r bottle (Fisher) and incubated overnight, upright, in an atmosphere of 1007. humidity and 5% CO^ at 37°. The following day, the volume of SSFM was made up to 700 mL and the bottle allowed to re-equilibrate with the 5% C0 2 > 100% humidity atmosphere. It was then sealed and rotated at 1 rpm at 37°. Each day the bottle was re-equilibrated with the 5% C02, 100% humidity atmosphere until the cells were confluent, usually approximately 10 days later. Cell numbers were estimated by counts using a hemocytometer and v i a b i l i t i e s by erythrosin B dye uptake. (For details see Stocks and Brooks, 1989). 16. Harvesting and purifying the monoclonal antibodies. The contents of the roller bottles were centrifuged (300 g, 5 minutes, 4°) in a Sorvall RC-5 centrifuge equipped with a GSA rotor and the supernatant decanted and f i l t e r e d through a 0.22 pm f i l t e r . The supernatant was then concentrated 50 fold by u l t r a f i l t r a t i o n using an Amicon XM100A or PM10 f i l t e r (Stocks and Brooks, 1989). The concentrated supernatant (2 mL) was injected onto a Pharmacia mono Q anion exchange column linked to a 54 FPLC apparatus. The elution buffer was lOmM sodium phosphate with a pH gradient from pH 8.5 to 7.0 and a 0-1 M sodium chloride ionic strength gradient (Gemski et al, 1985). The IgG peak, identified by SDS-PAGE was concentrated and stored at -70°. 17. Measurement of the binding isotherms of YE1 48. 10. and RG7 11.1 antibodies to MBL-2(4.1) and MBL-2(2.6) cell s . A l l binding studies were performed at 4° in PBS containing NaN^  (3 mM). The ce l l s were washed three times in PBS, then resuspended at a measured hematocrit (H, 30-50%) and kept at 0° for a minimum of 30 minutes or until required. Eppendorf tubes were incubated with BSA solution at 4° overnight, washed three times with PBS, dried and weighed (w ). Antibody of a known specific activity (S) and buffer to make the volume up to 1 mL were added and the eppendorf tube reweighed (w^). The antibody solution was incubated for 1 hour at 4° with end over end stirring, followed by the removal of 100 uL for y counting ( s^ ) and reweighing of the eppendorf tube (w3). Up to 0.3 mL of c e l l suspension of hematocrit H was added and the eppendorf tube reweighed (w ), followed by a further 1 hour 4 incubation with end over end stirring at 4°. The cells were centrifuged (300 g, 4 minutes) and an aliquot of supernatant (200uL) removed for counting (^2)- The eppendorf was reweighed (wg). If a wash analysis was required, the remaining supernatant was removed and the cells washed three times with PBS, each of the discarded washings being counted. The f i n a l pellet was. resuspended, weighed, counted and the hematocrit of the resuspended pellet measured. In this case a comparison could be made between the equilibrium binding and that bound after washing the cel l s . The following equations were used to calculate the bound and free ligand concentrations: [y (w -w )/(w - w )] - [y (w - w ) - (w - w )H]/(w - w ) V = 1 3 1 2 3 2 4 1 4 3 4 5 (w - w ).H.S 4 3 v {[y (w -w )/(w - w )] - [y (w - w ) - (w - w )H]/(w - w )}(w - w ) - = 1 3 1 2 3 2 4 1 4 3 4 5 4 5 L (w - w ).H.y 4 3 2 55 where v = surface concentration of bound ligand (molec./ g of packed c e l l s ) . L = free ligand concentration at equilibrium (molec./ g of supernatant). wi = weight of empty eppendorf tube .(g). w2 = wi + weight of radiolabeled antibody solution (g). w3 = w2 ~ weight of sample y for counting (g). w4 = w3 + weight of ce l l suspension added (g). w5 = w4 ~ weight of sample y 2 for counting (g). y = y counts of approx. 100 uL of supernatant before adding cells (cpm). y 2 = y counts of approx. 100 uL of supernatant after adding cells (cpm). H = mass fraction occupied by cells in added c e l l suspension, assumed equal to the hematocrit, or volume fraction of cells, in 9 the added c e l l suspension. The mass of the packed cells was 10 ce l l s per g. S = specific activity of antibody solution (cpm/ molec). Error estimates for these measurements were made from the partial derivatives of the above expressions for v and v/L and are shown in Appendix I. The surface area per MBL-2 c e l l was estimated from a photomicrograph and this information used to calculate the number of ligands bound per c e l l . 18. Fluorescent staining of cells. Approximately 5 x 106 cells were washed 3 times in PBS containing NaN (3.0 mM). The cells were incubated 3 in YE1.48.10. culture supernatant at 4° for 30 minutes with end over end sti r r i n g , then centrifuged. The supernatant was aspirated and the cells washed three times in PBS containing NaN3. -Fluorescein isothiocyanate-conjugated goat anti-rat IgG (FITC-garlgG) was added (0.5 mL, 0.05 mg/mL) and the incubation repeated as before. The cells were centrifuged, washed three times then resuspended at -approximately 50% hematocrit and placed on a slide. A l l cells except erythrocytes were 56 fixed in 95% ethanol and mounted under glycerolrPBS (90:10). Photofluorescence microscopy was carried out using a Zeiss Photomicroscope II equipped with a Zeiss Neofluor 40/0.75 Oel o i l immersion lens and Kodachrome 400 ASA film with an approximately 30 second exposure. Excitation was at 450-500 nm with a 510 interference f i l t e r and a high pass f i l t e r of greater than 528 nm. 19. Trypsin digestion of MBL-2 c e l l s . MBL-2 cells (approximately 10 1 0) were washed 3 times in PBS, resuspended in serum-free RPMI 1640 containing trypsin (10 mL, 0.25 mg/mL) and incubated 30 minutes at room temperature with end over end stirring. The cells were centrifuged, washed 3 times with PBS and used for fluorescent staining in some cases. Phenylmethyl sulfonyl fluoride (PMSF,- 0.02 mM fi n a l concentration) was added to the supernatant which was stored at -70° and used to coat ELISA plates. 20. Preparation of the two phase systems. Dextran T500 (Mw=494 000, M = 181 200, lot 26066) was obtained from Pharmacia and PEG 8000 (M = 8000) w was obtained from Union Carbide. The following stock solutions were prepared: 1. Dextran. Dextran T500 (22g) was made up to lOOg with water and stirred until dissolved. The %w/v was determined with a Steeg and Reuter 25 polarimeter on a 10 fold dilution of the stock solution using [a] = D +199° followed by conversion to %w/w using the partial specific density of dextran. 2. PEG. PEG 8000 (30g) was made up to lOOg with water. The %w/v was determined by measuring the refractive index of a 10 fold dilution of the stock solution using a Bausch and Lomb "Serum Protein" Meter then converted to %w/w using the partial specific density of PEG. 3. NaCl (0.6M). 4. Na HPO / NaH P0 buffer (214 mM / 67 mM, mixed to give pH 7.2). 2 4 2 4 5. Sorbitol (0.6 M). The stock solutions and water were mixed to give the required composition, mixed well and allowed to separate either at room temperature or 4° or by centrifugation (200 g, 10 minutes). The phases 57 were stored separately at 4 or -20. The systems used in the study were designated by the abbreviation X,Y,Z where X denotes the % w/w dx, Y denotes the % w/w PEG and Z is the buffer composition which may be I, II, III or S. I is 110 mM Na HPO / NaH PO , II is 96 mM Na HPO / NaH PO 2 4 2 4 2 4 2 4 and 50 mM NaCl, III is 10 mM Na HPO / NaH PO and 130 mM NaCl and S is 2 4 2 4 10 mM Na HPO / NaH P0 and 100 mM sorbitol. An example of a phase 2 4 2 4 composition description would be 5,4,11 which would be 5% dx, 4% PEG, 96 mM Na2HP04/ NaH2P04 and 50 mM NaCl. The phase systems used in this study were 5,4,Z where Z is I, II, III or S; 5.3.4.Z where Z is I, II, III; 5,3.5,1; 5,3.6,1; 5,3.7,1;. 5,3.8,1; 5,3.9,1; 5,5,1; 5,4,S and 7,4,S. 21. Partitioning of c e l l s and proteins. Protein solution (up to 100 uL) was added to 2 mL of phase system. If the resulting dilution of the phase system formed one phase, then stock PEG 8000 (up to 10 uL, approx. 30% w/w) was added to reform the phases. The phases were mixed by 20 inversions and allowed to settle before sampling and assay of the phases for protein concentration with Coomassie blue dye or radioactive counts or in the case of biotin or avidin with 4-hydroxyazobenzene-2' -carboxylic acid (HABA). For c e l l partitioning, erythrocytes were collected into 0.38% sodium citrate (final concentration) by venipuncture and cultured cells were removed from the medium by centrifugation. A l l cells were washed three times with PBS and resuspended in upper phase. The load mix concentration was determined by impedance c e l l counting on a Particle Data Inc. Electrozone Celloscope, 8 6 usually approximately 2 x 10 cells/mL. Approximately 2 x 10 ce l l s were added to 2 mL of phase system, the system mixed by inversion 20 times and upper, lower and total phases sampled. The cells were counted by impedance counting or radioactive counts. 22. Biotin-derivatization of protein. N-hydroxysuccinimidobiotin (BNHS, 25 mg/mL in dimethylformamide, Sigma) was added to protein (1-50 mg in 0.1 M Na2HP04/NaH2P04, pH 8.0) to give molar reaction ratios ranging from 1:10 lysine:BNHS. The mixture was stirred end over end for 4 hours then dialyzed overnight against the appropriate buffer, usually 10 mM Na HPO /NaH P0 , pH 8.0. The biotin content of the conjugate was assayed 2 4 2 4 K J e J 58 with HABA as defined and described below. 23. Determination of avidin, biotin and degree of protein derivatization with biotin. The assay of avidin and biotin is based on the use of the dye 4-hydroxyazobenzene-2'-carboxylic acid (HABA) which binds to avidin. The binding of HABA by avidin is accompanied by spectral changes summarized in Table 1. Table 1. Extinction coefficients of 4-hydroxyazobenzene-2'-carboxylic acid and i t s complexes with avidin. From Green (1970). X (nm) £ 282 350 500 Avidin 282 25 000 0 0 HABA 350 2 800 20 500 480 Avidin-HABA complex 500 - 2 500 34 500 HABA is not bound by the avidin biotin complex and since the - 2 1 dissociation constant of the latter is so low (10 M) the dye is stoichiometrically displaced by the biotin. HABA was recrystallized from aqueous methanol and dissolved in one equivalent of sodium hydroxide. a. Assay for avidin. HABA (10 mM, 50 uL) was added to avidin solution (0.1 - 1 mg, 2mL) in sodium phosphate buffer (0.1, pH 8) in a 1 cm cuvette. The absorbance at 500 nm (A ) was measured then biotin added 5 0 0 (2 mM, 50 uL) and the A remeasured. The number of binding sites and 500 the concentration of the original solution were calculated using the following expressions. [binding sites] = AA mM 6 500 34 avidin (mg/mL original solution) = 2.05 mL x 16.2 ^ 2~ mL 34 500 = 0.49 AA 500 59 b. Assay for biotin. A solution of avidin-HABA complex (0.2 - 0.4 mg avidin/mL, 0.25 mM HABA in 0.1M sodium phosphate buffer, pH 8, 2mL) was pipetted into a 1 cm cuvette and the absorbance at 500 nm was measured (A i). A known volume of biotin solution (v mL) was added and the absorbance remeasured (A ). The biotin concentration was calculated as 2 follows. [biotin] = A - A (v + 2)/2 M • l 2 mM 34 c. Assay of degree of protein modification by biotin. When solutions of biotinyl proteins are added to avidin-HABA complex, the dye is displaced more slowly than by the free biotin. When about 30% of the dye has been displaced the reaction becomes very slow and does not go to completion (Green, 1970). For this reason the protein-biotin complex was digested prior to assay for biotin. The modified protein was dialyzed against 2x4L of sodium phosphate buffer (0.1M, pH 8), then digested with pronase (Sigma, 0.2 mg/mL, 2 hours, 37°) and assayed for biotin as described above. 24. Polyacrylamide-derivatization of proteins. Protein solution in water (25 mL, approximately 2 mg/mL), in a three neck 250 mL flask equipped with a condenser, a dropping funnel and a N 2 inlet through an empty gas washing bottle, was degassed by bubbling N 2 through the solution for three hours under reflux at room temperature. The N2 exited the apparatus through a Dudley bubbling tube. (NH ) [Ce(N0 ) ] (54 mg/mL in 4 2 3 6 1 M HN03> 0.1 mL) in the dropping funnel was degassed by means of a Y junction in the N 2 line, then added to the degassed protein mixture. The reaction mixture was stirred minimally in order to mix the reagents but no further s t i r r i n g was required throughout the reaction. It has been shown that a higher degree of polymerization is obtained in the absence of s t i r r i n g (Muller, 1986). The reaction continued for three hours with constant N2 bubbling through. The relative extent of reaction was estimated by the increase in solution viscosity. A Cannon-Manning Semi-Micro viscometer was f i l l e d by suction and allowed to equilibrate at 25°. The liquid was pumped up to the upper mark, allowed to flow to 60 the lower mark and the time taken measured. The mean of three measurements was taken. The density of the solution was measured using a 2 mL pycnometer which was f i l l e d and weighed accurately twice. The viscosity was calculated using the following expression. Tf) = T) x s 'H o 2 where TJ = sample viscosity (cp). s 7) = viscosity of water (0.8904 at 25°). H 0 2 time (s) and time (H2O) are the measured times for the sample and water. m(H o) and m(s) are the densities for water and the sample (m(H o) 2 2 = 1.0015 g/mL at 25°). The derivatives were concentrated three fold and free acrylamide was removed by washing with 10 volumes of water and u l t r a f i l t r a t i o n through an Amicon XM100A f i l t e r . time (s) time (H O) 2 m(H o) 2 m(s) 7 25. Radiolabelling of MBL-2 cell s . Approximately 10 cells (10 mL) close to 100% v i a b i l i t y at approximately 106 c e l l s / mL (i.e. underconfluent) 125 were incubated overnight with 6 uCi of 5-[ I] Iodo-2'-deoxyuridine (Amersham) or 1 4C amino acid mixture (Amersham high specific activity mixture CFB.104). The cells were washed three times in PBS and the specific activity determined by c e l l impedance counting and radioactivity counting in a LKB Compugamma y counter or a Ph i l l i p s PW 7 4700 liquid s c i n t i l l a t i o n counter . Typical specific activities were 10 7 cpm/mL or 10 dpm/mL. 26. 5 1Cr labelling of erythrocytes. Erythrocytes (0.5 mL packed cells) were washed three times in PBS and incubated with [ 5 1Cr] NaCr 0 (0.5 2 7 7 mL, 10 Bq/mL, Amersham) for 30 minutes. Free label was removed by washing four times in 10 volumes of PBS. Typical specific a c t i v i t i e s 7 8 were 10 - 10 cpm/mL of cells. 61 2 7 . Counter current distribution (CCD) of cells. CCDs were run on a 60 cavity thin-layer CCD rotor (CCD 1240C, Buchler Instruments, Fort Lee, N.J.). Cavities were loaded with 0.6 mL of lower phase and 0.7 mL of upper phase when a lower phase partition was expected and vice versa for an upper phase partition. In the case of polyacrylamide (PAA)-derivatized antibodies a lower phase partition was anticipated and the concentrations of derivatized-antibody were 0.08 mg per cavity for polyacrylamide-derivatized rabbit anti-human erythrocyte IgG (PAA-rahrbc) and 0.2 mg per cavity for polyacrylamide-derivatized YE1.48.10. (PAA-YE1.48.10.). The appropriate number of cavities were loaded with phase system containing ligand or control phase systems without ligand except for the f i r s t cavity to which the load mix of 7 cells was added (approx. 2 x 10 total cells ) . The required number of transfers was performed, with the systems being mixed for 30 seconds and allowed to settle for 5 minutes. Erythrocyte CCDs were performed at room temperature and MBL-2 c e l l CCDs at 4°. After the run was finished, 1 mL of PBS was added to each cavity to dilute the phase system and the contents collected in tubes. The c e l l concentrations were determined by impedance counting or radiolabel counts. 62 CHAPTER 5 RESULTS AND DISCUSSION. 1. Separation of Erythrocytes using a Trypan Blue-Derivatized Second Ligand. Trypan blue (3,3'-[(3,3'-Dimethyltl,1"-biphenyl]-4,4'-diyl) bis (azo)bis[5-amino-4-hydroxy-2,7-napthalenedisulfonic acid] tetrasodium salt) is a blue dye (MW = 961) which partitions strongly into the upper phase of a PEG/dx two phase system (Fig. 8). The maximum partition coefficient measured for trypan blue was 10 at 0.5 ug/mL which decreased to 7.5 at 5 ug/mL. The partition coefficient decrease at higher concentrations was probably due to aggregation. As trypan blue is approximately half the molecular weight of PEG 1900 i t seemed possible that antibodies modified by covalent attachment of trypan blue may be less deactivated than those modified by covalent attachment of PEG 1900. Trypan blue was linked to bovine serum albumin (BSA) using l-ethyl-3-(3-dimethylaminopropyl)-carbodiimide (EDO at various trypan blue concentrations. It was confirmed that the trypan blue and BSA were linked by a red spectral shift of 6 nm in the visible spectrum of trypan blue. This is shown for IgG in Fig. 9. PEG 1900 was also linked to BSA using cyanurlc chloride activated PEG 1900 to make a comparison of the two a f f i n i t y modifying agents. The degree of modification of BSA with trypan blue was estimated from the ratio of absorbance at A to x J ^ 596 counts of radiolabelled BSA of a known specific activity. However since the trypan blue was bound to the protein as indicated by the spectral shift, i t would be anticipated that the extinction coefficient of the dye-protein conjugate would be different to that of the free dye. In order to estimate the number of trypan blue molecules bound to each protein molecule the value of the extinction coefficient for the free dye was used to estimate the concentration of trypan blue in the dye-protein conjugate. There is no basis for assuming the extinction coefficient of the free dye is the same as that of the bound dye. In fact changes in extinction coefficient on dye-protein binding are used in this study to assay avidin and biotin concentrations. However measurements of the absorbance of trypan blue at A before and after max the dye was coupled to the protein were similar at the same dye 63 concentration. As this was the only simple method available for estimating the amount of dye bound to the protein i t was used in this study. Although the absolute values of dye molecules bound to the protein may not be correct, i t gives an indication of the relative amount of dye bound in the conjugate. The amount of PEG 1900 bound per BSA molecule was estimated with fluorescamine (Stocks et al, 1986). A comparison of the partition of trypan blue and PEG 1900-derivatized BSA in various phase systems is made in Table 2. It must be realized, however, that these are only estimates of bound trypan blue as discussed above. Trypan blue derivatized BSA requires a higher degree of modification than PEG 1900-derivatized BSA to obtain an equivalent partition coefficient. As i t is possible to obtain an equally high partition coefficient using trypan blue as the modifying agent as with PEG and because of the possibility that trypan blue modification would have a less deactivating effect on the antibody than PEG 1900 modification due to i t s lower molecular weight, the use of trypan blue as an a f f i n i t y ligand modifying agent was investigated. Other studies suggest that the molecular weight of the modifying agent is more Table 2. A comparison of the effect of PEG 1900 and trypan blue on bovine serum albumin as af f i n i t y ligand modifying agents. Ligand DOM1 % Upper Phase Partition in the System Specified 5,3.4,1 5,3.4,11 5,3.4,III 5,4,1 5,4,11 5,4,III native BSA 0 52 52 53 53 49 42 PEG-BSA 16 100 100 100 100 100 100 PEG-BSA 34 100 100 100 100 100 100 PEG-BSA 46 100 100 100 100 100 100 BSA + TB 0 40 43 TB-BSA 1 42 45 TB-BSA 12 56 58 TB-BSA 36 83 85 TB-BSA 60 100 100 DOM1 degree of modification (molec. of PEG 1900 or TB per BSA molec). TB - trypan blue. BSA - bovine serum albumin. 64 s i g n i f i c a n t with respect to antibody a c t i v i t y decreases than the coupling chemistry employed to modify the protein (Sharp et al, 1986). A model erythrocyte separation was examined to investigate the p o t e n t i a l of trypan blue-modified antibodies as second a f f i n i t y ligands. The problem used to test t h i s approach was to separate human and rabbit erythrocytes using a monoclonal mouse anti-NN glycophorin IgG (mocNN glyc) as the primary ligand and a trypan b l u e - d e r i v a t i z e d polyclonal sheep anti-mouse F c fragment IgG (TB-samFc) as the second ligand. This i s depicted schematically i n F i g . 10. Fi g . 8. Trypan blue dye in a 5,4,111 phase system at t o t a l concentrations from l e f t to right of 5, 2.5, 1.25, 0.75 ug/mL. The corresponding p a r t i t i o n c o e f f i c i e n t s are from l e f t to r i g h t 7.5, 9.0, 9.8 and 10.5. Mouse F c fragment was prepared from mouse IgG by papain d i g e s t i o n followed by FPLC. The f r a c t i o n s were i d e n t i f i e d and the p u r i t y assessed by SDS-PAGE. The F c fragment was the second peak eluted by FPLC (Fig. 11). The SDS-PAGE gel of the IgG digestion products i s shown i n F i g . 12. In lane 2, F i g . 12, the crude IgG digest, there are three major bands. The reduced F c fragment has a MW of less than 30 000 and any IgG l i g h t chain or F ^ fragment, which would also have a molecular weight less than 30 000, would also be included i n t h i s band. The other two major bands are probably p a r t i a l l y digested F c fragments defined as IF and s F c (MW = 48 000 and 40 000 respectively, Utsumi, 1969). A small amount of heavy chain at over 70 000 suggests that the d i g e s t i o n was close to completion and that only small amounts of l i g h t chain are present i n 6 5 200 400 600 800 WAVELENGTH (nm) Fig. 9. Visible spectrum of trypan blue and sheep anti-mouse F c (samFc> ) and trypan blue-derivatized samFc ( - - - ) . The conjugate shows a 6 nm red spectral shift. 66 m a N N glyc NN Glycophorin A Fig . 10. A schematic diagram of the binding of trypan blue-derivatized sheep anti-mouse F^ (samFc) to mouse anti-NN glycophorin (mctNN glyc) which is bound to an erythrocyte, thus immunospecifically coating the erythrocyte with trypan blue. the digest.These p a r t i a l l y digested products, IF and s F c > have also been observed in the papain digest of rabbit IgG (Stocks and Brooks, 198S) and in other mouse F c fragment preparations (Clezardin et al, 1985). The FPLC puri f ied F c fragment, shown in Fig . 12, lane 3, is s l i gh t ly contaminated with a band at approximately 50 000 but this preparation is considerably purer than the commercial F c fragment preparation, also shown in Fig . 12, lane 4, which has contaminating bands at 40 000 and 48 000 corresponding to the putative p a r t i a l l y digested F c fragments, l F c and sF c_ The absence of F ^ fragment in the F c preparation was confirmed by ELISA using an a n t i - F ^ enzyme-conjugated antibody. The F c preparation was used to raise sctmFc by standard techniques and puri f ied from sheep serum by ammonium sulfate precipi tat ion and FPLC (Fig. 13). The f i r s t two peaks eluted are both IgG; i t is l ike ly that these represent different classes of IgG. SDS-PAGE gel of the pure samFc is shown in Fig . 12, lanes 7 & 9, which shows the samF to be 67 homogeneous. The second antibody, maNN glyc MAb, was purified from ascites by ammonium sulfate precipitation and FPLC (Fig. 14). The IgG was eluted in the f i r s t and only IgG-containing peak reflecting the monoclonal origins of the antibody. This also produced a homogeneous IgG preparation as shown by SDS-PAGE (Fig. 12, lanes 6 & 8). It was shown by hemagglutination assays that the MAb, maNN glyc had no specificity for NN erythrocyte agglutination over that for MM or MN erythrocyte agglutination. In this case since the antibody was required to distinguish between human and rabbit erythrocytes, this was not important. Table 3. Results of partitioning experiments using trypan blue-modified sheep anti-mouse F IgG as a second a f f i n i t y ligand. Cell Type 1° Ligand 1 2° Ligand 1 PC %UP 2 Control PC %UP TB-samF 5 83 0.30 23 mlgG TB-samF 3. 4 77 0.60 37 human rbc maNN glyc TB-samF 2. 6 70 0. 10 18 rabbit rbc maNN glyc TB-samF c 0. 25 20 0.09 8 human rbc TB3 - 21 - 18 human rbc TB4 - 12 - 18 A l l antibody concentrations were 0.25 mg/mL and at a modification ratio of trypan blue:samF of 28:1. 2 The control partition is measured in the presence of primary ligand only. In the case of native TB the control is the partition in phase system only. 3 4 At a TB concentration of 1 ug/mL and 10 ug/mL. PC = partition coefficient, TB = trypan blue. %UP = the % of total cells located in the upper phase. mlgG = non-specific mouse IgG. The results of partitioning experiments in the model system outlined are summarized in Table 3. A l l experiments using this ligand combination 68 Fig. 11. Fast protein liquid chromatography (FPLC) pr o f i l e of the papain digest of mouse IgG on a Mono Q column. The ionic strength elution gradient (- - -) was 0.01-1M NaAc, pH 5.5. The second peak was identified as F £ fragment by SDS-PAGE (see Fig. 12). 69 1 2 3 4 5 6 7 8 9 Fig. 12. 10% (lanes 1-4) and 3% (lanes 5-9) SDS-PAGE gels of IgGs and F c fragment used i n studies of the e f f e c t of trypan bl u e - d e r i v a t i z e d second a f f i n i t y ligand on erythrocyte p a r t i t i o n . Lanes 2-5 and 8-9 are reduced, the others are native. 1. Molecular weight markers. 2. Reduced papain digest of mouse F c fragment. The reduced F c fragment has a MW of less than 30 000. The two other major bands represent p a r t i a l l y digested IgG (see text). 3. Reduced FPLC-purified F c fragment (MW s 30 000) from mouse IgG digest i n lane 2. 4. Reduced commercial mouse F c fragment (Sigma). 5. Molecular weight markers. 6. Native monoclonal mouse anti-NN glycophorin IgG. 7. Native pol y c l o n a l sheep anti-mouse F c fragment IgG. The band at approximately 300 000 i s l i k e l y a dimer of IgG. 8. Reduced monoclonal mouse anti-NN glycophorin IgG as i n 6. 9. Reduced pol y c l o n a l sheep anti-mouse F c fragment IgG as i n 7. The putative dimer i n 7 i s reduced and no longer observed. 70 TIME (min) Fig. 13. FPLC profile of sheep anti-mouse F c fragment (samFc) preparation. The elution gradient (- - -) was 0-1M NaCl in 0.01M NaH PO / Na2HP04 buffer, pH 8. Both IgG peaks were identified by SDS-PAGE and ELISA. (Fig. 12, lanes 7 & 9). IgG2 is more basic than lgGy and is therefore eluted f i r s t (Goding,- 1983). The two IgG peaks, representing different classes of IgG, were combined and were indistinguishable by SDS-PAGE. 71 TIME (min) F i g . 14. FPLC p r o f i l e of monoclonal mouse anti-NN glycophorin (maNN g l y c ) . The e l u t i o n gradient (- - -) was 0-1M NaCl i n 0.01M NaH PO / 2 4 Na 2HP0 4 buffer, pH 8. The IgG peak was i d e n t i f i e d by SDS-PAGE and ELISA. (Fig. 12, lanes 6 & 8). Only one c l a s s of IgG was detected by FPLC r e f l e c t i n g the monoclonal o r i g i n of the antibody. 72 were in 5,4,1 (100 mM phosphate buffer) systems at 4 . Trypan blue modification of the samF resulted in an increased partition coefficient c which increased with the degree of modification. The modification of IgG with trypan blue required a lower degree of modification than BSA to obtain a similar partition coefficient suggesting that the trypan blue bound to IgG is more exposed to the phase system than that bound to BSA. Two different modified antibody preparations were examined, namely a 15:1 and 28:1 trypan blue to IgG mole ratio. The degree of modification was estimated from the ratio of dye absorbance to concentration of radiolabelled protein. However, the 15:1 conjugate had a partition coefficient of only 1.3 and did not have a significant effect on the partition of either antibody or erythrocytes. At higher concentrations of both native and antibody-conjugated dye, a decrease in the partition coefficient of erythrocytes was observed; most lik e l y this was caused by aggregation of the dye and erythrocytes. A l l experiments were carried out at 0.25 mg/mL of trypan blue-derivatized antibody, the maximum TB-samFc concentration at which no decrease in the partition coefficient was observed. Although the TB-samFc and maNN glyc combination was able to increase the partition of human NN erythrocytes from 18% to 70% this was not as specific as the PEG-derivatized ligand (Stocks and Brooks, 1988) since the same ligand combination also increased the partition of rabbit erythrocytes from 8% to 20%. This is probably due to non-specific binding of the trypan blue-derivatized antibody to the rabbit erythrocytes. It is Table 4. The effect of TB-socmFc on the partition of other mouse IgGs. 3° Ligand 1 2° Ligand 2 1° Ligand 3 PC %UP Control %UP PC TB-samF mlgG 2.8 74 0.2 17 c 4 TB-samFc marIgG . rlgG 1.4 58 0.9 47 TB-samFc mapili 3.2 76 0.3 23 1 2 3 Concentrations of ligands are 0.15 mg/mL, 0.10 mg/mL and 0.05 mg/mL. 4The partition of rabbit IgG (rlgG) in the presence of mouse anti-rabbit IgG (marlgG) only. 73 unlikely that the maNN glyc cross-reacts with the rabbit erythrocyte since i t did not agglutinate rabbit erythrocytes in a microtiter assay. Since one of the main advantages of using a second antibody as the a f f i n i t y ligand is the potential to use the same second ligand with different primary ligands for different separation problems, the effect of the trypan blue-derivatized antibody on other potential primary ligands was examined. These were polyclonal non-specific mouse IgG (mlgG), monoclonal mouse anti-rabbit F c fragment IgG (marFc) with rabbit IgG (rlgG) and monoclonal mouse a n t i - p i l i IgG (mapili). The results are summarized in Table 4. In a l l cases the trypan blue-derivatized antibody was able to alter the partition coefficient of the primary antibody, suggesting that this primary antibody could be used as a general ligand. 14 The effect of TB-samFc and mouse a n t i - p i l i IgG on the partition of C labelled p i l i was examined. No partition change was observed but this may have been due to the availability of only small amounts of a n t i - p i l i antibody. The TB-samFc and malgG were also used to increase the partition of a third antibody, non-specific rabbit IgG. In summary trypan blue dye may be used as an a f f i n i t y modifying agent and is comparable to a PEG-derivatized ligand in effectiveness. However a disadvantage is aggregation of the dye and non-specific binding of the dye to proteins and cells as illustrated by the effect of the free dye on human erythrocyte partition and a non-immunospecific increase in rabbit erythrocyte partition. 2. Separation of MBL-2(4.1) and MBL-2(2.6) ce l l s by Immunoaffinity Partition. The i n i t i a l approach taken to the separation of MBL-2(4.1) and MBL-2(2.6) cells is summarized in Fig. 15. The MBL-2 c e l l i s immunospecifically coated with PEG through the binding of YE1.48.10., a rat MAb specific for MBL-2 cells and the binding of PEG-derivatized RG7/11.1, a mouse MAb specific for the F c fragment of rat IgG2b, to the YE1.48.10. The f i r s t step in the separation was to generate and purify large quantities of MAb from both YE1.48.10. and RG7/11.1 hybridoma lines. a. Production and purification of YE1.48.10. and RG7/11.1 monoclonal antibodies. The method used to purify the m a NN glyc for the previous 74 separation was ammonium sulfate precipitation of antibody-containing ascites f l u i d followed by FPLC (Stocks and Brooks, 1987; 1988). Ascites refers to a non-solid tumor produced intraperitoneally ln rats or mice by injection of hybridoma c e l l s after priming with pristane. It is rich in MAb (approximately 2mg/mL). However losses in MAb a c t i v i t y due to ammonium sulfate precipitation have been observed (Bruck et al, 1982; McGregor et al, 1983) and losses in activity previously observed by the author may well be due to this step of the purification (Stocks, 1986). For this reason u l t r a f i l t r a t i o n was used to concentrate the antibodies rather than ammonium sulfate precipitation. Fig. 15. Schematic diagram of the binding of a MBL-2 c e l l by YE1.48.10. MAb and PEG-derivatized RG7/11.1 MAb to immunospecifically coat the MBL-2 c e l l with PEG. MAbs have traditionally been generated in rat or mouse ascites (Goding, 1983) but this i s not practical for large quantities of MAb because of the large numbers of animals required. Also MAbs of ascites origin w i l l be contaminated with non-specific mouse or rat antibodies. The alternative is to generate large quantities of c e l l culture medium but the addition of f e t a l bovine serum (FBS) to provide necessary growth factors also adds bovine antibodies to the medium and the large amount of bovine protein present in 107. FBS-containing medium makes the purification of a MAb of a typical concentration of 10 pg/mL d i f f i c u l t . Therefore both hybridoma c e l l lines were adapted to growth in supplemented serum free medium (SSFM, Murakami et al, 1982; Stocks and Brooks, 1989). They continued to secrete antibody at the same rate as when plated in serum-containing medium and became confluent at a maximum \PEG P E G RG7/11.1 YE1.48.10. 75 c e l l density of 1-2 x 106 cells/mL within 10 days of i n i t i a l l y plating 4 at 1-2 x 10 cells/mL. Typically they exhibited a lag of approximately three days prior to exponential growth when i n i t i a l l y plated into SSFM (Fig 16). Cells maintained in SSFM exhibited a much shorter lag of approximately one day but they were not routinely maintained in SSFM as they began to deteriorate after several weeks in SSFM. Cell v i a b i l i t y as estimated by erythrosin B dye uptake was close to 100% in SSFM until the cells became confluent. U l t r a f i l t r a t i o n was found to be an extremely effective method for both concentration and partial purification of the supernatant at 4°. The SDS-PAGE gels shown in Fig. 17 compare culture supernatants concentrated through different f i l t e r s and assess the purity of insulin and transferrin added to the medium. It can be seen that both the transferrin and the insulin are not contaminated by any proteins detected by Coomassie blue-stained SDS-PAGE gels (Fig. 17, lanes 6 & 7). The addition of serum to the culture medium creates a more complicated separation problem as is shown by comparison of lanes 3 & 4, Fig. 17, where the unconcentrated serum-supplemented medium has similar protein concentrations as in the 20 fold concentrated SSFM. The large amount of IgG in lane 3, Fig. 12 is mostly bovine IgG originating from the serum. This is another advantage of using SSFM as i t avoids the problem of quantitating the amount of bovine IgG relative to monoclonal IgG in a serum-containing preparation. U l t r a - f i l t r a t i o n through an Amicon XM100A f i l t e r which has a molecular weight cut-off of 100 000 removed a major protein band of approximate molecular weight 67 000 (Compare lanes 4 & 5, Fig. 17). This 67 kD band has a molecular weight corresponding to albumin and must originate from the hybridoma ce l l s since the transferrin and insulin, the only additions to the medium, are seen to be relatively pure. The putative albumin may be that bound to the c e l l surface during culture in serum-containing medium which could dissociate from the c e l l surface in the SSFM. Transferrin was retained by the Amicon XM100A f i l t e r despite having a molecular weight of 77 000, significantly less than 100 000. The concentrated SSFM was chromatographed on a mono ' Q anion exchange FPLC column and the elution profile is shown in Fig. 18. From the corresponding SDS-PAGE gels shown in Fig. 20 i t can be seen that the 76 IgG is eluted in fractions 9 & 10 and the transferrin in earlier fractions. The two bands (MW ~ 25-30 kD) visible in lanes 9 & 10 (Fig. 20) are a consequence of the secretion of K light chain, as well as the A light chain which forms the active antibody, by the YE1.48.10. hybridoma. This a b i l i t y to synthesize K chain originates from the myeloma half of the fusion partners (Galfre and Milstein, 1981). The peak eluted in fractions 17 & 18, Fig. 18 did not contain any protein when examined by SDS-PAGE and is likely due to elution of the phenol red dye in the medium. The FPLC elution profile for the RG7/11.1 MAb is shown in Fig. 19. It is essentially the same as for the YE1.48.10. MAb; transferrin was eluted in fractions 2,3 & 6 whereas IgG was eluted mainly in fraction 10 (Fig. 19). The SDS-PAGE gels of the purified YE1.48.10. and RG7/11.1 are shown in Fig. 21. RG7/11.1 is seen to exist as a dimer which dissociates under reducing conditions (Fig. 21, lanes 4 & 5). Yields of antibody were approximately 30 ±5 mg per l i t e r of SSFM for YE1.48.10. MAb and 15 ±5 mg per l i t e r for RG7/11.1 MAb. In summary this is an effective method to process one l i t e r of culture supernatant per day, typically requiring 5-8 FPLC runs, each chromatographing approximately 5 to 10 mg of total protein. FPLC was the only step performed at room temperature as i t only required 40 minutes and no loss in activity was detected by ELISA. For a more labile antibody this step could be performed at 4°. Both antibodies retained high activities, e.g. assuming the MAb concentration in SSFM to be in the typical range of 5-10 ug/mL (Goding, 1983), the minimum concentration at which MAb activity was s t i l l detectable in SSFM was 2-4 ug/mL, i.e. a 2-3 fold dilution of the antibody-containing SSFM. For the same antibody purified by this procedure the corresponding actual MAb concentration was similar:11 ug/mL, whereas the same antibody purified by ammonium sulfate precipitation showed a significant loss in activity: the minimum concentration at which MAb activity was detected was 330 ug/mL. These relatively high minimum concentrations at which activity was detected do not necessarily indicate a low a f f i n i t y antigen as the surface antigen density on the ELISA plate was unknown. These values are only useful in terms of the comparison made here rather than an indication of absolute activity. For further details see Stocks and Brooks (1989). 77 \ 1 5 — A I 5 S t DAYS medium added F i g . 16. Growth curves f o r YE1.4S.10. (A) and RG7/11.1 (<>) hybridoma c e l l s . The c e l l s were i n i t i a l l y resuspended at approximately 10 4 cells/mL i n lOOmL of supplemented serum-free medium (SSFM) and a f u r t h e r 60G mL of SSFM added at day 2. The v i a b i l i t y of both c e l l l i n e s was estimated by e r y t h r o s i n B dye uptake. 78 Fig. 17. 10% SDS-PAGE of serum-containing and serum-free hybridoma supernatants reduced with 5 mM 2-mercaptoethanol. lane 1. Molecular weight standards. lane 2. RPMI-1640 containing 5% f e t a l bovine serum (FBS). The major band at approximately 67 500 i s probably albumin. lane 3. C e l l culture supernatant containing 5% FBS. Again the major band at 67 500 i s albumin and the bands at 43 000 and 27 000 correspond to heavy and l i g h t chains of IgG. The two l i g h t chains are due to the synthesis of K chain by the hybridoma c e l l (Galfre and M i l s t e i n , 1981). lane 4. Supplemented serum-free medium (SSFM) c e l l c u l t u r e supernatant concentrated 50 f o l d through an Amicon UM10 f i l t e r (MW c u t - o f f = 10 000). The putative albumin (see text) and IgG bands are present as described i n 2 and 3. lane 5. SSFM c e l l culture supernatant concentrated 50 f o l d through an Amicon XM100A f i l t e r (MW cut-off = 100 000). The IgG and t r a n s f e r r i n (MW = 77 000) are seen but the putative albumin (see text) i s absent due to the u l t r a f i l t r a t i o n . lane 6. T r a n s f e r r i n (5 pg, Sigma, MW = 77 000). This i s free of major contaminants. lane 7. Insul i n (5 ug, Sigma, MW = 5 800). This i l l u s t r a t e s the p u r i t y of the i n s u l i n . 7 9 TIME (min) Fig. 18. FPLC pr o f i l e of supplemented serum-free medium (SSFM) YE1.48.10. hybridoma culture supernatant concentrated by u l t r a f i l t r a t i o n through an Amicon XM100A f i l t e r . The flow rate was 1 mL/minute and 1 mL fractions were collected. The IgG peak is eluted in fractions 9 & 10. 80 TIME (min) Fig. 19. FPLC profi l e of supplemented serum-free medium (SSFM) RG7/11.1 culture supernatant concentrated by u l t r a f i l t r a t i o n through an Amicon XM100A f i l t e r . The flow rate was 1 mL/minute and 1 mL fractions were collected. The IgG peak is eluted in fraction 10 while transferrin is eluted in fractions 2, 3 & 6. 81 MW. x103 94" 6 7 — 43-30 — 20.1 1 4 4 T 3 4 7 9 10 18 19 F i g . 20. 10% SDS-PAGE of FPLC f r a c t i o n s of YE1.48.10. hybridoma c u l t u r e supernatant as shown i n F i g . 18. Lane 1 i s molecular weight standards and the other lane numbers correspond to the FPLC f r a c t i o n as shown i n F i g . 18. The bands at 77 000 are t r a n s f e r r i n and those at 55 000 and l e s s than 30 000 correspond to IgG heavy and l i g h t c h a i n . The two l i g h t chains are due to s e c r e t i o n of K chain by the hybridoma (Goding, 1983). MW. x 10 3 3 0 0 — 1 5 0 — 5 5 — 1 F i g . 21. 3% SDS-PAGE of FPLC f r a c t i o n s of RG7/11.1 and YE1.48.10. pure MAb. Lane 1 i s molecular weight standards, 2 & 3 are non-reduced and reduced YE1.48.10. and 4 & 5 are non-reduced and reduced RG7/11.1. The band at 300 000 i n lane 4 i s a dimer of IgG and bands at approximately 150 000 are whole IgG. The heavy chains of IgG are at approximately 66 000 and the l i g h t c h a i n i s w e l l below 66 000. 82 b. Characterization of the MBL-2 Cells. The known distinction between the two sub-lines of MBL-2, MBL-2(4.1) and MBL-2(2.6), is that MBL-2(4.1) has a higher surface density than MBL-2(2.6) of an antigen defined by the binding of rat MAb YE1.48.10. (Takei, 1986). Fluorescence microscopy of MBL-2 cells incubated with YE1.48.10. followed by staining with fluorescein-conjugated goat anti-rat IgG confirmed that more YE1.48.10. MAb molecules bound to MBL-2(4.1) than MBL-2(2.6) cells as shown in Fig. 22. The MBL-2(4.1) cells showed greater fluorescence than MBL-2(2.6) cells (compare 1 & 2, Fig. 22) whereas a lymphocyte control A6, a mouse hybridoma (Stocks and Brooks, 1988),showed no fluorescence (7, Fig. 22). Short incubation of either MBL-2 sub-line with trypsin almost completely removed the binding site for YE1.48.10. (3 & 4, Fig. 22) but Triton X, a non-ionic detergent, only slightly reduced the binding of YE1.48.10. (5 & 6, Fig. 22), although the antigen recognized by YE1.48.10. has been purified from EL4-BU cells by detergent solubilization (Takei, 1983). In order to assess the f e a s i b i l i t y and d i f f i c u l t y of the separation problem i t was necessary to measure the equilibrium binding of YE1.48.10. MAb to MBL-2 cells as a function of MAb concentration and c e l l surface area. The cells were characterized by equilibrium binding 125 studies using I labelled YE1.48.10. antibody, and the surface area per c e l l was measured by digitization of a 69000-fold magnification photomicrograph of the two c e l l types. In measuring the c e l l surface area, three points on the c e l l circumference were digitized and used to describe a cross section of the sphere assumed to represent the 2 lymphocyte. Using the expression A = 4nr , the surface area of a sphere corresponding to the two-dimensional photomicrograph was found to be 615 2 ±30 pm . However accurate measurements of lymphocyte surface area have shown that the folding of the membrane generally gives rise to a total surface area 130% that of the corresponding sphere (Schmid-Schonbein et al, 1980). This would estimate the area of an MBL-2 c e l l at 800 ±39 pm2, but i t i s unlikely that a l l of this area would be available for ligand binding. The former measurement may be considered a lower limit and the latter an upper limit. The following calculations assume the lower value and thus the corresponding numbers of ligands bound per unit surface area represent maximum values. The density of MBL-2 cel l s was found to 83 be 1.0 g/mL by w e i g h i n g a c c u r a t e volumes o f c e l l s u s p e n s i o n a t v a r i o u s h e m a t o c r i t s up t o 60%. The number o f packed c e l l s p e r u n i t volume was 9 x 1 0 8 c e l l s / m L a f t e r c e n t r i f u g a t i o n a t 300 g f o r f o u r minutes; T h i s c a l i b r a t i o n was used to e s t i m a t e c e l l numbers from the h e m a t o c r i t and was o n l y used t o e s t i m a t e the number o f c e l l s i n the p e l l e t . I t c o u l d n o t be c o n s i d e r e d a v a l u e f o r the c e l l volume s i n c e i t does not t a k e i n t o a c c o u n t t h e t r a p p e d volume o f s u p e r n a t a n t o r c e l l d e f o r m a t i o n . c. A n a l y s i s of B i n d i n g E x p e r i m e n t s . E q u i l i b r i u m measurements o f l i g a n d b i n d i n g gave i n f o r m a t i o n on the moles o f l i g a n d bound p e r gram o f c e l l s a t a known h e m a t o c r i t and f r e e l i g a n d c o n c e n t r a t i o n . In o r d e r t o c o n v e r t t h i s i n f o r m a t i o n t o the number o f moles o f l i g a n d bound a t s a t u r a t i o n ( i . e . t h e number o f b i n d i n g s i t e s ) and the m i c r o s c o p i c a s s o c i a t i o n c o n s t a n t (k ) o f the l i g a n d - b i n d i n g s i t e i n t e r a c t i o n the f o l l o w i n g S c a t c h a r d a n a l y s i s was employed (C a n t o r and Schimmel, 1980). Assume a r e a c t i o n o f the type: P + L = PL where P = a b i n d i n g s i t e on the c e l l s u r f a c e . L = l i g a n d . PL = the l i g a n d - b i n d i n g s i t e complex. k = i m t l and * = [ P L ] [PL] 1 [P]+[PL] where k^ = the m i c r o s c o p i c d i s s o c i a t i o n c o n s t a n t . d> = t h e f r a c t i o n a l s a t u r a t i o n o f s i t e i . . [ P ] ( [ P L ] / [ P ] ) 1 [ P ] ( 1 + [ P L ] / [ P ] ) s i n c e [ P L ] / [ P ] = [ L ] / k d [ L ] / k <p = — (25) 1 l + [ L ] / k d 84 5 6 7 F i g . 22. MBL-2 c e l l s stained with YE1.48.10. MAb and fluorescein-conjugated goat a n t i - r a t IgG. 1. MBL-2(4.1) c e l l s 2. MBL-2(2.6) c e l l s 3. Trypsinized MBL-2(4.1) c e l l s 4. Trypsinized MBL-2(2.6) c e l l s 5. T r i t o n X treated MBL-2C4.1) c e l l s 6. T r i t o n X treated MBL-2(2.6) c e l l s 7. Lymphocyte control c e l l s (A6, Stocks and Brooks, 1988) 85 A similar expression may be written for each site and assuming that a l l the sites are identical, the sum of a l l the species w i l l give the total moles of ligand bound per unit surface area, v, at equilibrium, i.e. 1^ = v. n[L]/k d nk [L] Therefore v = or in terms of k v = — - (26) l+[L]/k a 1+k [L] d a This is a convenient expression to relate v and k (k is normally a referred to as k when discussing antibody-antigen binding) when there are a maximum of n independent identical sites per unit area since the values for k and n may be calculated using various plots. A direct plot of v vs [L] is one method to evaluate these parameters but requires data at high values of [L] which is d i f f i c u l t to measure and not practical when large quantities of costly MAb are required. Another possible plot Is an inverse plot of 1/v vs 1/[L] but this is d i f f i c u l t to analyze s t a t i s t i c a l l y since the smaller, less certain measurements dominate the linear regression f i t s . The plot used to analyze this data was a plot of i>/[L] vs v (Fig. 23) This allows a linear extrapolation to give n and does not weight the data at low values of v or [L]. 0 v Fig.23. Scatchard plot for independent identical binding sites. 125 Scatchard plots were constructed for the binding of I-labelled biotin-derivatized and native YE1.48.10. MAb to MBL-2 cel l s as well as to erythrocytes and a mouse hybridoma control. A direct plot of the binding of YE1.48.10. to MBL-2(4.1) and MBL-2(2.6) is shown in Fig. 24. 86 Both MBL-2 sub-lines became saturated at 0.3 mg/mL which happens to be a convenient MAb concentration to use in an APTS. The Scatchard plot corresponding to Fig. 24 is shown in Fig. 25. The errors in several of the points were estimated by taking the partial derivatives of v and v/L with respect to each measurement and multiplying by the sum of the errors of each measurement. The partial derivatives used in the calculation are summarized in Appendix 1. As can be seen in Fig. 25 the errors in each point are similar which made i t possible to use linear regression analysis to f i t the best line to the data. From the X-intercept and the slope of the graph i t was possible to obtain values for v and k as illustrated in Fig. 23. Similar plots were constructed for the two MBL-2 sub-lines using biotin-derivatized YE1.48.10 as shown in Fig. 26. Error analysis was carried out as-before and the errors were sufficiently similar to allow linear regression analysis. The control Scatchard plots constructed using human erythrocytes and a mouse hybridoma line are shown in Fig. 27. No binding was observed within the limits of experimental error to trypsinized MBL-2 cells of either type. A summary of the information obtained from the plots i s given in Table 5. Table. 5. Summary of the information obtained by linear regression of the Scatchard plots in Figs. 25-27. Molec. of Molec. of Cell type Ligand ligand bound ligand bound k (M ) per c e l l . per cm  MBL-2(4.1) YE1. 2. ,36 ±0, ,28 xlO 6 3, ,83 ±0. .23 xlO 1 1 xlO 1 1 xlO 1 1 xlO 1 1 9.0 ±0. 5 xlO 8 xlO 8 xlO 8 xlO 8 MBL-2(2.6) YE1. 7. ,96 ±1, ,61 xlO 5 1, .29 ±1. .30 8.6 ±0. 3 MBL-2(4.1) B-YE1. 1. ,30 ±0, , 16 xlO 6 2, . 10 ±0. . 13 2.5 ±0. 2 MBL-2(2.6) B-YE1. 6. ,67 ±1. ,37 xlO 5 1, .08 ±1. .11 3.2 ±0. 08 Human-erythrocyte YE1. 2. ,47 ±0. ,61 xlO 3 *1. 76 ±0. 43 xlO 9 2.73±0. ,06 xlO 9 Mouse-xlO 1 0 xlO hybridoma YE1. 8. ,20 ±1, .87 xlO 4 1, .37 ±0, .94 4.4 ±0. ,5 YE1. = YE1.48.10. MAb. B = biotin * -6 2 , assuming the surface area of an erythrocyte to be 1.4x10 cm (Evans and Fung, 1972). 87 The YE1.48.10. MAb is quite specific for the MBL-2 cells with respect to the hybridoma control, the binding to the mouse hybridoma being only 4% of that to MBL-2(4.1) cells and 11% of that to MBL-2(2.6) 2 cells on a per cm basis. The erythrocytes bind very small amounts of the YE1.48.10. MAb, i.e., less than 1% of that bound to MBL-2(4.1) cells 2 and 1.4% of that bound to MBL-2(2.6) cells on a per cm basis. The surface antigen difference between the MBL-2 sub-lines is only a factor of 3. This small antigenic difference and low surface antigen density makes this a stringent separation problem. However tumor associated antigens (TAAs) are often of this order of surface density (Pimm et al, 1982; Price et al, 1982). One of the objects of this work was to define and extend the limits of immunoaffinity partitioning sensitivity with respect to antigen density difference necessary for separation. Previous immunoaffinity partition separations of species-specific erythrocytes have used antibodies raised against the whole c e l l (Sharp et al, 1986; Karr et al, 1986) and in these cases large antigenic differences would be expected. A smaller surface antigen density was used as the basis for separation when rabbit anti-NN glycophorin was used as the primary ligand (Stocks and Brooks, 1988) but even in this case there was s t i l l a large surface 11 2 antigen density difference of 0 and 5 xlO molecules per cm as the basis for the separation. A measure of the sensitivity of the partition coefficient to the number of PEG chains bound per unit area was obtained by Sharp (1985). He measured the binding of PEG 8000-palmitate to the erythrocyte surface and the corresponding partition coefficient. The binding of 11 2 approximately 6 xlO molecules/cm of PEG-palmitate was required to increase erythrocyte partition from 0 to 20% in the upper phase. The 11 2 binding of 1 xlO molecules/cm would have increased the upper phase partition from 0 to 6%. This implies that an immunospecific increase in the partition of even MBL-2(4.1) cells is approaching the limits of the 11 2 technique as applied to date as this only binds 3.83 xlO molecules/cm of YE1.48.10. MAb. Also MBL-2 cells are much larger than erythrocytes 2 2 (615 pm compared to 140 pm , Evans and Fung, 1972) and wil l have a larger free energy of adsorption at the interface. Another factor in the comparison is that the PEG 8000 ester of palmitate was used by Sharp 88 . 24. A direct plot of the binding of YE1.48.10. MAb to MBL-2(4. ) and MBL-2(2.6) ( • ) ce l l s . 89 4 - i - > 0 5 10 15 2 0 2 5 v Cx IO 1 4 molec./"g c e l l s ? Fig. 25. A Scatchard plot of the binding of YE1.48.10. MAb to MBL-2(4.1) ( 0 ) and MBL-2(2.6) (*) ce l l s . The error bars were calculated as described in Appendix 1. 90 0.6 ^ v (x 10 molec./g ce l l s ) Fig. 26. A Scatchard plot of the " binding of biotin-derivatized YE1.48.10. MAb to MBL-2(4.1)(0) and MBL-2(2.6) (*) ce l l s . The error bars were calculated as described in Appendix 1. 91 Fig. 27. A Scatchard plot of the binding -of YE1.48.10. MAb to human erythrocytes (*) and mouse lymphocytes (CD- The error bars were' calculated as described in Appendix 1. 92 whereas the MAbs in this study are modified using the smaller PEG 1900 to preserve MAb activity meaning that the ligand partition coefficient w i l l be lower. Offsetting this, however, is the fact that each IgG molecule is bound by approximately twenty PEG 1900 molecules in a typical a f f i n i t y ligand compared to only one PEG 8000 in the case of the palmitate ester, although their effect is much lower than is theoretically expected for this level of derivatization (Sharp et al, 1986). Hence on the basis of past experience the MBL-2 separation problem seems to be on the limit of the resolution of immunoaff inity partition. Binding studies were also done on the biotin-derivatized YE1.48.10. MAb. The biotin-YEl. 48. 10. bound to the cells at a sl i g h t l y lower surface density than the native MAb and had a lower a f f i n i t y for the MBL-2 cells. This is reflected in the k which is three fold less for a the biotin-YEl.48.10. than the native YE1.48.10.(Fig. 26 and Table 5). If this is a real difference in binding i t is possible that the increase in hydrophobic character of the MAb due to biotin-derivatization weakens the binding due to some kind of interaction with hydrophilic areas on the c e l l surface. The glycocalyx is known to be hydrophilic. Another possibility is that the sites are not identical; although the Scatchard plot is linear, the differences in binding energy may not be sufficient to affect the linearity. d. Partitioning Studies on MBL-2 Cells. i . The effect of phosphate and chloride. The partition of MBL-2 cel l s in the absence of ligands was examined in several phase systems with varying phosphate:chloride ratios at isotonic ionic strength and pH 7.2. The results are summarized in Table 6. 93 Table 6. The effect of different phase systems and phosphate on the partition of MBL-2 (4.1) and MBL-2 (2.6) cells at room temperature. % upper phase partition with 10 mM NaP added (uL) System 0 10 20 50 250 Cell type 4.I 1 2.62 4. 1 2.6 4.1 2.6 4.1 2.6 4. 1 2.6 35,4,I 2 2 7 9 3 2 4 3 6 7 5,4,11 0 0 2 2 - - 1 1 4 2 5,4,III 2 3 4 4 1 2 2 2 2 3 5,3.4,I 31 32 63 61 51 58 60 62 74 78 5,3.4,II 27 27 23 25 16 19 20 21 59 52 5,3.4,III 7 5 3 5 7 8 3 4 8 9 45,3.4,I 71 80 5,3.5,I 54 52 5,3.6,I 40 35 5,3.7,1 7 10 5,3.8,I 9 13 5,3.9,I 5 9 5,5, I 0.5 1 V l = MBL-2(4.1) cells and 22.6 = MBL-2(2.6) cells. A l l experiments were in 2 mL of phase system. 3 & 4These two sets of data were obtained with different stock polymer solutions and MBL-2 cells at different stages of growth. The results of partitioning experiments shown in Table 6 il l u s t r a t e a problem encountered with MBL-2 c e l l partition throughout the study. The value of the % of cells returning to the upper phase varied widely in experiments carried out using cells harvested at different stages of growth with slightly different v i a b i l i t i e s and different batches of phase system. MBL-2 cells are seen to have different partition coefficients in identical phase systems in experiments carried out at different times (3 & 4). Although the absolute value of K varied widely, the two c e l l types consistently had similar K values at similar stages of growth and the ligands a l l had a reproducible effect on c e l l partition. In this respect c e l l partition is very different from molecular partition in which the molecules do not change with time and repeated measurements o f K wil l give consistent results. Control partitions were carried out with every experiment to be certain of the 94 F i g . 28. A t y p i c a l CCD p r o f i l e of MBL-2(4.1) c e l l s ( ) and MBL-2(2.6) c e l l s (- - -) i n a 5,4,1 phase system at room temperature. 95 \ 100 F i g . 29. The e f f e c t of PEG-1inoleate on the p a r t i t i o n of MBL-2(4.1) (*), MBL-2(2.6) c e l l s ( <> ) and dead MBL-2C4.1) c e l l s (0) i n a 5,4,1 phase system at 4°. 96 effect of the ligand and the two MBL-2 ce l l lines were thawed simultaneously and used in approximately the same passage. They did not always grow at the same rate and could not be described as identical with respect to their age distribution. It was found that in general the partition coefficient tended to increase with increasing age of the culture as well as did heterogeneity, reflected by the spread of the CCD (Fig. 28). However the rate of increase in K or the increase in spread of the CCD with number of c e l l passages was not consistent with the number of passages but varied from batch to batch of thawed cell s . As expected both MBL-2 sub-lines are sensitive to phosphate and phase composition, the upper phase partition increasing with phosphate concentration and decreasing with increasing tie line length (TLL). No significant differences between the partition of MBL-2(4.1) and MBL-2(2.6) were observed. The c e l l partition in the 5,4 systems was less sensitive to increasing phosphate concentration while c e l l partition in 5,3.4,1 and 5,3.4,11 systems was the most sensitive to phosphate concentration. A CCD of the two MBL-2 sub-line showed slight differences between the two ce l l lines as shown in Fig. 28. However this small difference i s likely a reflection of the heterogeneity of the populations than differences between the cells. The difference is insufficient to separate the cells. i i . The effect of PEG-linoleate. The addition of PEG-linoleate to the system had the expected effect of increasing the upper phase partition of both MBL-2 c e l l sub-lines (Fig. 29). No differential partition behavior between the two MBL-2 sub-lines was observed although the ester was less efficient at increasing the partition of dead ce l l s . For this reason a l l subsequent experiments were done using cells close to 100% vi a b i l i t y . As i t seemed unlikely that the cells could be separated by their native partition characteristics or the use of fatty acid esters, the only reasonable alternative appeared to be immunoaffinity partition. The effect of various types of immunoaffinity ligands on the MBL-2 ce l l s was Investigated. 97 i i i . The effect of PEG-YE1.48.10. This is the directly modified ligand approach as depicted in Fig. 2.a. In order to assay YE1.48.10. antibody activity by ELISA i t was necessary to remove the antigen from MBL-2 cells by tryptic digestion and use the digest to coat the ELISA plate. Several l i t e r s of MBL-2(4.1) cells ( ~10 1 0cells) were digested, the digest aliquoted and stored at -70°. The same digest was used throughout the study to coat ELISA plates at 10 pg/mL of total protein as this concentration gave the highest readings with a low background. The results of an ELISA using pure YE1.48.10. are shown in Fig. 30. The minimum concentration of YE1.48.10. at which binding was detected was 11 pg/mL. The f i r s t immunoaffinity experiments were carried out with antibodies derivatized with cyanurlc chloride-activated PEG 1900 (PCC). 0.6 - i 0.6 -B C © H < y 0.4 CQ OS o W m < 0.2 -0.0 11 i i i 11 i 111 i i i«11 »1111111 »«»111 11 11 »i i 11 5 10 15 20 LOG DILUTION FACTOR 2 Fig. 30. Binding curve of a typical ELISA of pure YE1.48.10. The i n i t i a l YE1.48.10. concentration was 22 mg/mL. 98 Previous studies (Sharp et al, 1986) have shown the 1:3 lysine:PCC ratio to be the best compromise between increasing the partition coefficient and minimizing the antibody activity loss; The same batch of PCC was used throughout this study. It was stored at -70° under nitrogen and was found to be 26% modified by ti t r a t i o n of the hydrolyzable chlorides. YE1.48.10. was modified using PCC at a modification ratio of lysine:PCC of 1:1, 1:3 and 1:5 which resulted in the modification of 4, 10 and 17 lysines per IgG molecule as estimated with fluorescamine and assuming 124 lysines per molecule (Edelman et al, 1966). At the two higher modification ratios the ligands partitioned strongly in to the upper phase. However none of these modified YE1.48.10. antibodies had any effect on the partition of either MBL-2 c e l l as shown in Table 7. Table 7. The effect of PEG-YE1.48.10. on the partition of MBL-2 (2.6) and MBL-2 (4.1) cells at 4°. No. of PEGs Partition % c e l l UP System per IgG molecule coefficient MBL-2(4.1) MBL--2(2.6) of PEG-YE1. no IgG _ 7 5 5,3.4,III 4 5.7 6 6 l l I I 0 1 6 5 I I I I no IgG - 3 3 5,4,1 4 5.5 3 2 l l I f 0 1 2 1 I I I I no IgG - 5 6 5,3.4,III 10 10 6 6 I I I I 0 1 5 5 f t I I no IgG - 2 2 5,4,1 10 10 2 2 l l I I 0 1 2 2 H l l no IgG - - 4 4 5,3.4,III 17 14 5 4 I I l l 0 1 4 5 I I I I no IgG - 2 2 5,4,1 17 15 2 2 I I l l 0 1 1 2 I I I I A l l partitioning was done at 2mg/ml ligand concentration. 99 iv. The Effect of PEG-RG7/11.1 and YE1.48.10. The use of a second MAb RG7/11.1, which was directed against rat IgG and bound the primary 2 b MAb, YE1.48.10., was examined as the ligand. Binding studies on MBL-2 cell s saturated with YE1.48.10. were carried out and the Scatchard plots are shown in Fig. 31. The underivatized antibody bound only weakly and at saturation only 6-12% of the cell-bound YE1.48.10. was bound by RG7/11.1. In the case of MBL-2(4.1) 1.3 x 105 molec./cell of RG7/11.1 were bound compared to 2.36 x 106 molec./cell of YE1.48.10. The 4 5 corresponding numbers for MBL-2(2.6) cells are 9.6 x 10 and 7.96 x 10 molec./cell respectively. Unfortunately this was the only anti-rat F c(IgG 2 b) MAb available. As with the primary antibody described previously, a number of modification ratios were used and the degree of modification assayed with fluorescamine. It is d i f f i c u l t to assess the activity of a PEG-derivatized antibody by an ELISA since a decrease in binding of the second, enzyme-conjugated, anti-MAb antibody could be due to decreased binding of the primary antibody or due to interference by the bound PEG in the binding of the second antibody. In order to avoid this problem an experiment was performed in which the inhibition of the binding of a second, enzyme-conjugated polyclonal mouse anti-rat F c to YE1.48.10. by the PEG-derivatized or native primary antibody, RG7/11.1, was assayed. Thus there was an inverse relationship between the amount of RG7/11.1 antibody bound and the colorimetric enzyme-catalyzed reaction. The ELISA plate was coated using YE1.48.10., then native or PEG-derivatized RG7/11.1 was added. The second enzyme-conjugated antibody, polyclonal anti-rat F c IgG, was added after the RG7/11.1 and BSA blocking agent. The concept is described in Fig. 32. The PEG 1900 derivatized YE1.48.10. retained approximately half i t s original activity estimated by this assay (Fig.32), at the 10 PEGs per IgG level of derivatization, judged by the observation that at saturation i t is about half as effective at inhibiting the binding of the HRP-conjugated antibody. This is a high retention of activity compared to previous results (Stocks and Brooks, 1988) but PEG-enzymes have retained higher activities (Abuchowski et al, 1977; Beauchamp et-al, 1983). 100 0.20 -i v (x I O 1 4 molec./g c e l l s ) Fig. 31. Scatchard plots for the binding of RG7/11.1 MAb to MBL-2(2.6) (*) and MBL-2C4.1) ( • ) cells coated with YE1.48.10. MAb. The error bars were calculated as described in Appendix 1. 101 PVC m i c r o t i t e r p l a t e Substrate COPD) C c o l o r l e s s } \ Fig. 32. Schematic diagram of an ELISA (top) and modified-ELISA (bottom) used to assay the binding of PEG-derivatized antibodies. 102 0.2 1 0 11 u I I I I I J ^ I I I I I I I I 1^1 I I I I I I 1 1^1 I I 1 I I I I 1^1 I I I I I I I i j LOG DILUTION FACTOR 2 Fig. 33. Binding curves for native (+) and PEG 1900-derivatized RG7/11.1 (•) to YE1.48.10. assayed by an ELISA as described in Fig. 32. The PEG-RG7/11.1 was modified by the attachment of 10 PEGs per IgG molecule. Increasing values of absorbance indicate a decrease in ligand binding. Partitioning experiments were carried out on c e l l s which had been preincubated with YE1.48.10., then washed twice or with both antibodies present in the phase system and no preincubation period. The former practice was an attempt to economize on MAb as experiments with washing radiolabeled c e l l s had shown that the YE1.48.10. antibody did not dissociate s i g n i f i c a n t l y under these conditions. However, as can be seen in Table 8, none of the antibody combinations had any effect on the partition of MBL-2 c e l l s . 103 Table 8. The p a r t i t i o n of MBL-2 (4.1) and MBL-2 (2.6) c e l l s i n the presence of PEG-RG7 11.1 and YE1 48.10. antibodies at 4°. no. of PEGs system ligand concentration % UP per RG7/11.1 1[PEG-RG7] 2[YE1.48.10] 3(4.1) 4(2.6) 5,3.4,1 - - 70 72 0 " 0.45 mg/ml preincub. 71 73 3 " " " " 72 73 10 " " " " 72 73 21 " " " " 68 71 23 " " " " 68 68 0 5,3.4,11 0.22 mg/ml 0.15 mg/ml 51 42 3 " " " 48 42 10 " " " 43 34 21 " " " 37 30 23 " " " 53 43 5,4,111 - - 1 0 0.3 mg/ml preincub. 1 ^ II n n II ^ 10 " " " " 2 21 " " " " 2 23 " " " " 1 5,4,11 - - 11 8 0 " 0 . 3 mg/mL preincub. 10 7 7 " " " 5 4 1 9 " " 3 4 42 » " " 2 4 46 " " " 3 4 1[PEG-RG7] = PEG-derivatized RG7/11.1 MAb concentration. 2[YE1.48.10.] =YE1.48.10. MAb concentration. 3(4.1) = MBL-2(4.1) c e l l s . 4(2.6) = MBL-2(2.6) c e l l s . 104 v. The effect of Biotin-YEl.48.10. and PEG-avidin. The partition 14 coefficient of biotin was measured using C labelled biotin and found to be 1.4. The derivatization of avidin with PEG-1900 increased the partition of avidin from 0.6 to 11 in the case where 20 PEG molecules were bound per avidin. This PEG-avidin derivative was able to increase the partition of biotin from 1.4 to 15, an 11 fold increase, whereas native avidin decreased the partition of biotin from 1.4 to 0.31. These results are shown in Fig. 34. The binding of the dye 2(4'-hydroxyazobenzene)benzoic acid (HABA) by an excess of avidin leads to an increase in e from 480 to 34 500. J 500 The avidin-HABA complex may be dissociated by the addition of 4 moles of biotin per mole of avidin (Green, 1965). This reaction was used to measure the partition coefficient of biotin-derivatized proteins, avidin and PEG-avidin. The PEG-derivatization of avidin affected i t s binding capacity for biotin. This loss in binding a b i l i t y could be estimated from the amount of biotin required to displace HABA from the avidin-HABA complex. There is evidence that HABA and biotin bind at the same site (Green, 1970) so this probably gives a good indication of the loss in binding a b i l i t y caused by PEG-derivatization. In Fig. 35 an example of the decrease in a f f i n i t y for avidin at two different PEG-modification ratios (i.e. 11 & 16 PEG 1900 per avidin) is given. From the slopes of the graphs i t was estimated that 36% and 60%, respectively, of the biotin binding sites had been destroyed by PEG-modification. The partition coefficient of HABA was increased by including PEG-avidin in the phase system and this could be inhibited with biotin as is shown in Fig. 36. The effect of biotin-derivatization and PEG-avidin or avidin on the partition of BSA was also investigated and the results are summarized in Table 9. 105 1.5 q [ P E G - a v i d i n l or Iavidin) mg/mL Fig. 34. The effect of PEG-avidin (*) and avidin (•) on the partition of biotin in a 5,4,1 system at room temperature. The biotin concentration was 0.05 nM. 106 - 0 0.2 0.4 0.6 0.8 IPEG-avidin] or Ia v i d i n ! mg/mL 1 J 0 a v i d i n O PEG-avidin O PEG-avidin C36>S modified) C60X modified) F i g . 35. The e f f e c t of PEG 1900 modification on the a b i l i t y of a v i d i n to bind 2(4'-hydroxyazobenzene) benzoic acid(HABA) 107 1 . 6 -i BIOTIN ADDED (mM) Fig. 36. The effect of biotin on the partition of 2(4'-hydroxyazobenzene) benzoic acid (HABA) in the presence of PEG-avidin (0.4 mg/mL). The concentration of HABA was 0.5 mM. Insert shows HABA in a 5,4,1 system in the presence of PEG-avidin (0.4 mg/mL) and biotin concentrations from right to l e f t of 0.3, 0.2, 0.1 and 0.05 mM. 108 Table 9. The effect of blotin-derivatization, PEG-avidin and avidin on the partition of BSA. mol biotin/ % of biotin-BSA % with PEG-avidin % with avidin mol BSA in upper phase included included (0.5 mg/ml) (0.4 mg/ml) (0.4 mg/ml) 0 50 53 52 42 84 100 22 37 77 85 20 26 72 80 20 21 70 71 17 • 15 65 67 14 7 55 57 14 A l l experiments were in a 5,4,1 system at 4° and the mole ratio of biotin:BSA was assayed with fluorescamine. Biotin-derivatization caused the BSA to partition in favor of the upper phase and the effect was further increased by the addition of PEG-avidin while native avidin decreased the effect. The larger the number of moles of biotin bound per BSA, the greater the partition into the upper phase. This illustrates the effect of increasing the molecular weight of the partitioned material since the biotin-BSA has a greater partition coefficient than native biotin. Biotin-derivatized polyclonal sheep anti-mouse F c IgG (samFc) also partitioned into the upper phase as shown in Table 10. It was found that a large excess of N-hydroxysuccinimidobiotin was required to achieve a high modification ratio. This is also true of the PCC-protein reaction and is attributable to the.competing hydrolysis reaction. Even at these high modification ratios described in Table 10, no loss in activity of the samFc was detected with an ELISA assay. However this is a polyclonal antibody and this w i l l not necessarily be true for a MAb. In fact a Scatchard plot of biotin-YEl.48.10. showed some loss in YE1.48.10. activity at a lower modification ratio (Fig. 26). This may be because an ELISA is less sensitive than a binding assay or because of the polyclonal origins of the samF antibody. 109 Table 10. The e f f e c t of reaction r a t i o of N-hydroxysuccinimidobiotin on the degree of modification of IgG with b i o t i n and the p a r t i t i o n c o e f f i c i e n t . r eaction r a t i o product r a t i o biotin:IgG (mol) 7. B N H S reacted A c t i v i t y by ELISA (7. of o r i g i n a l ) K BNHS:IgG (mol) 54 119 272 388 581 36 36 98 111 209 46 21 25 19 24 100 100 100 100 100 5.3 5.4 6.0 7. 1 7.3 BNHS = N-hydroxysuccinimidobiotin The next approach to the MBL-2 separation was the combination of b i o t i n - d e r i v a t i z e d YE1.48.10. and PEG-avidin as shown i n F i g . 2.d. The b i o t i n - d e r i v a t i z e d YE1.48.10. had a lower a f f i n i t y f o r MBL-2 c e l l s than native YE1.48.10. as was discussed e a r l i e r . The biotin-YEl.48.10. p a r t i t i o n e d into the upper phase at modifications ranging from 17, 27 and 63 b i o t i n s per IgG molecule (K = 6.7, 11.5 and 19 re s p e c t i v e l y ) and the ELISA at a l l these l e v e l s of modification did not show any loss i n a c t i v i t y despite the binding assay i n d i c a t i n g some decrease i n the amount bound. The addit i o n of PEG-avidin i n a l l cases was able to move the remaining biotin-YEl.48.10. into the upper phase while a v i d i n s h i f t e d i t a l l into the lower phase. An example of the dependence of the p a r t i t i o n of biotin-YEl.48.10. (27 b i o t i n s per IgG molecule) on PEG-avidin and a v i d i n concentration i s shown i n Fig. 37. Although the MAb responded appropriately to blotin/PEG-avidin modification, the combination of these two ligands was unable to a l t e r the p a r t i t i o n of MBL-2 c e l l s , although small but s i g n i f i c a n t changes i n p a r t i t i o n were obtained with the biotin-YEl.48.10. alone. The r e s u l t s of p a r t i t i o n i n g experiments with both MBL-2 c e l l s and YE1.48.10. are shown i n Table 11. A l l the modified YE1.48.10. MAbs were able to increase the p a r t i t i o n of both c e l l l i n e s and the p a r t i t i o n c o e f f i c i e n t of MBL-2(4.1) c e l l s was approximately twice that, of MBL-2(2.6) c e l l s i n every case. However these di f f e r e n c e s are small compared to the p a r t i t i o n of native c e l l s and a separation on the basis of these p a r t i t i o n c o e f f i c i e n t s would require a CCD of 200 transfers. The addition of PEG-avidin to the 110 system a c t u a l l y lowered the MBL-2 c e l l p a r t i t i o n s l i g h t l y with increasing PEG-avidin concentration. Even native a v i d i n d i d not decrease the MBL-2 c e l l p a r t i t i o n , although the di f f e r e n c e s may not be s i g n i f i c a n t from PEG-avidin. This i s shown i n Fig. 38. Table 11. Results of p a r t i t i o n i n g experiments using biotin-YEl.48.10. and MBL-2 c e l l s . C e l l type System YE1.48.10. Biot i n s K % UP K [ce cone.(mg/mL) per IgG (B-YE1) ( c e l l s ) MBL-2(2.6) 5,4,111 0.20 0 1.0 45 0 .8 MBL-2(4.1) 5,4,111 0.20 0 1.0 45 0 .8 MBL-2(2.6) 5,4,111 0.20 17 6.7 48 0 .9 MBL-2(4.1) 5,4,111 0.20 17 6.7 63 1 .7 MBL-2(2.6) 5,4,111 0.20 10 4:2 50 1 .0 MBL-2(4.1) 5,4,111 0.20 10 4.2 62 1 .6 MBL-2(2.6) 5,4,11 0.15 0 1.0 30 0 .4 MBL-2(4.1) 5,4,11 0.15 0 1.0 31 0 .4' MBL-2(2.6) 5,4,11 0. 15 27 11.5 27 0 .4 MBL-2(4.1) 5,4,11 0.15 27 11.5 46 0 . 9 The s l i g h t e f f e c t of the PEG-avidin and a v i d i n i n lowering the p a r t i t i o n of MBL-2 c e l l s was not examined i n d e t a i l . If the e f f e c t i s re a l i t i s d i f f i c u l t to explain. The amount of PEG added to the system would not change the i n t e r f a c i a l tension of the system s u f f i c i e n t l y to a l t e r the c e l l p a r t i t i o n . In fact the phase diagram of the system with a l l the ligands was measured and was almost i d e n t i c a l to that without ligands. If the e f f e c t were due to agglutination of the c e l l s by the av i d i n binding to more than one b i o t i n , then the native a v i d i n would be expected to have a much greater e f f e c t as i t has more b i o t i n binding s i t e s a v a i l a b l e . No agglutination of the c e l l s was observed at 40-times magnification. The f a c t that some increase i n c e l l p a r t i t i o n i s observed with the biotin-YEl.48.10. alone suggests that the PEG-avidin and a v i d i n i n t e r f e r e with the binding of biotin-YEl:48.10. to•the MBL-2 c e l l as the biotin-YEl.48.10. has no e f f e c t i n the presence of these ligands. Table 12 shows the p a r t i t i o n c o e f f i c i e n t s of B-YE1.48.10; with and without c e l l s i n the presence and absence of avidin or PEG-avidin. The p a r t i t i o n c o e f f i c i e n t of the biotin-YEl.48.10. i s e s s e n t i a l l y the same i n the 111 - 3 -i i i i i i i \ j » i i i i i i i i ^ i i i i i i i i i ^ I PEG—avi din! or [avidin] mg/mL Fig. 37. The effect of PEG-avidin (#) and avidin (•) on the partition of YE1.48.10.MAb. (0.3mg/mL). 112 1 I P E G - a v i d i n ] or t a v i d i n ] mg/mL K C b i o t i n - Y E l . 48. 10. D w i t h P E G - a v i d i n © K CMBL-SC4.13 c e l l s w i t h b i o t i n - Y E l . 4 8 . I O . D and P E G - a v i d i n K CMBL—EC4.13 c e l l s w i t h b i o t i n - Y E l . 48. IO. D and a v i d i n Fig. 38. The effect of avidin and PEG-avidin on the partition of MBL-2 cell s in the presence of biotin-YEl.48.10. (0.3 mg/mL). 113 Table 12. The effect of PEG-avidin and avidin on the partition coefficient of biotin-YEl.48.10. in the presence of MBL-2C4.1) cell s . without MBL - 2 U . 1 ) cells with MBL-2(4.1) cells PEG-avidin avidin K log K K log K cone. cone. (B-YE1) (B-YE1) (B-YE1) (B-YE1) (mg/mL) (mg/mL) % UP MBL-2 (4.1) 0.41 0.83 1.24 1.65 1.82 2.97 6.7 211.5 35.5 10 20 0.8 1.6 0.7 1.0 5.3 9.8 4.5 18 24 0.7 1.0 0.65 1.3 63 62 46 50 44 31 21 17 0.41 0.83 1.24 1.65 1.82 2.97 0.02 0.02 0.02 0.02 0.02 0.02 -1.7 0.02 0.01 0.01 0.01 0.01 0.01 -1.7 -0.2 48 40 45 45 56 % UP = % of cells partitioning into the upper phase. A l l experiments used 0.2 mg/mL B-YE1.48.10. b i o t i n s per YE1.48.10. = 17 2Biotins per YE1.48.10. = 27 3Biotins per YE1.48.10. = 15 presence or absence of cells plus PEG-avidin or avidin. If the biotin-YEl.48.10. and PEG-avidin or avidin were binding to the cel l s and the ligands were not strong enough to alter the c e l l partition, a decrease in K (biotin-YEl.48.10.) in the presence of cells compared to that in the absence of cells would be seen as the ligand became bound to cells at the interface. This was not observed. However as seen in the f i r s t three rows of Table 12, in the case where the partition is done in the absence of PEG-avidin or avidin, the K (biotin-YEl.48.10. ) is decreased in the presence of MBL-2 cells relative to in the absence of cells . In the the presence of PEG-avidin and avidin however, the K (biotin-YEl.48.10.) is unaffected by the presence of MBL-2 cel l s 114 suggesting that under these circumstances the YE1.48.10. Is not binding to the c e l l . In summary the avidln-biotln system works well on a molecular level but did not affect MBL-2 c e l l partition. It seems possible that this is due to the binding of PEG-avidin or avidin to biotin located In or close to the antigen binding site of the antibody thus interfering with the cell-antibody binding; As none of the upper phase partitioning ligands were able to alter the partition of MBL-2 cells by a sufficient amount i t was decided to try a lower phase partitioning ligand. Since most of the MBL-2 ce l l s are located at the interface In a l l the systems used and i t is easy to optimize the systems for a lower phase partitioning ligand, c e l l s partitioning Into the lower phase may be separated from those at the interface. 3. The effect of polyacrylamide-derivatized antibodies. As discussed in the introduction, polyacrylamide has a strong a f f i n i t y for the lower phase of a PEG/dx APTS. It was decided to graft polyacrylamide (PAA) onto antibody molecules, the advantages being that the antibody would partition more strongly into the lower phase than a PEG-ligand would into the upper phase. This supposition is based on the fact that dextran has a much larger lower phase partition than the corresponding PEG upper phase partition. For example in a 5,4 system the upper phase partition of PEG is 4.0 whereas the lower phase partition of dx is 28.3 (Albertsson, 1986). This means less polymer would need to be bound to the MAb thus decreasing the activity loss. Also because the reaction i s specific for alcohols the sites of attachment are limited to serine hydroxyls and possibly threonine. The oligosaccharide hydroxyls are capable of i n i t i a t i n g polymerization but are insignificant (approx. 8 per IgG) compared to the numbers of serine and threonine hydroxyls. The optimization of the reaction conditions for grafting PAA onto protein was done using BSA and non-specific bovine IgG. A drawback of the polymerization is that since almost a l l eerie salts are insoluble, a buffer which did not precipitate the initiator could not be found and a l l polymerizations were carried out in water only. A minimum volume of (NH ) [Ce(N0 ) ] in 1M HNO was added to avoid protein precipitation and 4 2 3 6 3 115 denaturation. While i t was possible in the cases of the two antibodies used in this study to maintain antibody activity this may not be true for a l l proteins. The increase in solution viscosity after polymerization was measured as this was a simple method of indicating that a reaction had occurred. PAA was grafted onto BSA, bovine IgG, rabbit anti-human erythrocyte IgG (rahrbc) and YE1.48.10. The viscosity increases at various reaction ratios and partition coefficients for these derivatives are quoted in Table 13. One of the acrylamide:protein mole ratios was sufficiently high to produce a gel. The PAA-protein derivatives were not characterized. Nitrogen analysis, the usual method of characterizing PAA bound to a solid support, was not feasible due to the nitrogen present in the protein, a radiolabelled acrylamide monomer was unavailable and an amino acid analysis encountered problems in the hydrolysis of the sample. Hemagglutination assays (Table 14) using rabbit and human erythrocytes with a l l the PAA-rahrbc" derivatives did not show any decrease in agglutinating a b i l i t y due to PAA derivatization nor did they agglutinate rabbit erythrocytes. The most highly modified rahrbc antibody (1:100) showed a slight increase in agglutinating a b i l i t y , probably due to the larger amounts of free PAA present. High molecular weight PAA is known to agglutinate cells non-specifically although in the present case this was not entirely non-specific since the rabbit cells were not agglutinated by the PAA-rahrbc. In the case of the MBL-2 cells and PAA-YE1.48.10. , again no loss in activity was observed at the 1:17 reaction ratio level as measured by microtiter but the gelled antibody had no agglutinating abil i t y . The gel was resuspended by pipetting in order to carry out a hemagglutination assay. The PAA-YE1.48.10. did not agglutinate human erythrocytes or a control mouse lymphocyte. The PAA-YE1.48.10. agglutinated MBL-2 cells at much lower concentrations than PAA-rahrbc, however. This is reflected in the partitioning experiments as no aggregation problems were encountered when using the PAA-rahrbc but aggregation was a definite problem when using the PAA-YE1.48.10. This aggregation is antibody-mediated since none of the control cells: aggregated as they would i f this were an effect only of the PAA. 116 Table 13. A summary of polyacrylamide-grafted proteins synthesized with details of reaction ratios, viscosities and partition coefficients in a 5,4,1 system. Protein Hydroxyl:Acrylamide Viscosity (cp) K K reaction ratio unreacted reacted (native) (PAA-protein) (mol:mol) BSA 1: :32 0.94 1. .34 2. . 1 0. 38 b IgG 1: :5 0.98 2. ,31 1. .5 0. 16 b IgG 1: :3 0.92 1. ,23 1. .5 0. 25 b IgG 1: : 1.6 0.91 0. .99 1. .5 0. 28 rahrbc 1: : 100 0.91 1. 36 1. . 1 0. 18 rahrbc 1: :25 0.90 1. ,04 1, . 1 0. 21 rahrbc 1; :20 0.90 0. 99 1. . 1 0. 19 YEl.48. 10. 1: : 17 0.90 0. .97 1. .2 0. 051 YEl.48. 10. 1: :260 0.91 gel 1. .2 BSA = bovine serum albumin b IgG = bovine IgG rahrbc = rabbit anti-human erythrocyte IgG PAA = polyacrylamide 1measured in a 7,4,S phase system The PAA-rahrbc was an extremely effective ligand (Figs. 39 & 40, Table 15) and obtained the most efficient immunoaff inity separation of erythrocytes reported to date. Even at concentrations as low as 0.08 mg/mL PAA-rahrbc almost a l l the human erythrocytes were in the lower phase whereas the rabbit erythrocyte partition was unaltered. As shown in Fig. 40 a ligand concentration of 0.08 mg/mL increased the human erythrocyte partition from 8 to 86% whereas the rabbit erythrocyte concentration remained unaltered at 18%. Although a good separation could be obtained in a single partition step, a CCD was used to separate the cells with 100% purity. A 5,4,1 system was found to give an efficient separation and was used for the CCD runs. Theoretical CCD profiles for the values of G of 0.9 and 0.22 for rabbit and human erythrocytes are shown in Appendix II. Theoretically, a CCD of 15 transfers would be sufficient to achieve a separation of 100% pure c e l l 117 Table 14. Hemagglutination assays of polyacrylamide-derivatized ligands. Antibody Hydroxyl:acrylamide Cell Minimum agglutinating reaction ratio (mol:mol) concentration (pg/mL) rahrbc hrbc 2.0 PAA-rahrbc 1: :25 hrbc 2.0 PAA-rahrbc 1: : 100 hrbc 0.4 rahrbc rrbc none PAA-rahrbc 1: :25 rrbc none PAA-rahrbc 1: : 100 rrbc none YE1.48.10. - MBL-2(4.1) 0.001 PAA-YE1.48. 10. 1: : 17 MBL-2(4.1) 0.001 PAA-YE1.48. 10. 1: :260 MBL-2(4.1) none YE1.48.10. - hrbc none PAA-YE1.48. 10. 1: : 17 hrbc none YE1.48.10. - m lymphocyte none PAA-YE1.48. 10. 1: : 17 m lymphocyte none PAA = polyacrylamide rahrbc = rabbit anti-human erythrocyte IgG hrbc = human erythrocyte rrbc = rabbit erythrocyte m lymphocyte = mouse lymphocyte control 1No agglutination at the highest concentration of antibody (0.5 mg/mL for rahrbc and 1 mg/mL for YE1.48.10.). populations but in order to allow for c e l l heterogeneity i t was decided to use 20 transfers. Human erythrocytes were labelled using 5 1Cr and the CCD profile of a mixture of rabbit and human erythrocytes in the presence of PAA-rahrbc obtained in a CCD run and the corresponding theoretical profile is shown in Fig. 41. The rabbit and human erythrocytes were almost completely resolved. The broadening of the peaks compared to the theoretical peaks presumably reflects c e l l heterogeneity. 118 F i g . 39. The p a r t i t i o n of human erythrocytes i n a 5% dx, 4% PEG, 110 mM sodium phosphate buffer system showing (top) from l e f t to ri g h t 0, 0.04, 0.08, 0.12, 0.16, 0.2 mg/ml polyacrylamide-derivatized rahrbc. The bottom row shows the p a r t i t i o n of rabbit erythrocytes i n the same conditions with (a) and without (b) 0.2 mg/mL polyacrylamide-derivatized rahrbc. 119 [PAA-r a hrbc] mg/mL Fig. 40. The partition of human (*) and rabbit erythrocytes (+) in the presence of increasing concentration of polyacrylamide-derivatized rahrbc antibody in a 5,4,1 phase system at room temperature. 120 Table 15. The partition of human and rabbit erythrocytes in the presence of polyacrylamide-derivatized rahrbc. Ligand Cell type % LP in the system specified 5,3.4,1 5,3.4,111 5,3.4,1 5,4,111 hrbc 12 33 11 48 PAA-rahrbc hrbc 93 99 93 85 rrbc 16 22 31 37 PAA-rahrbc rrbc 19 39 29 46 PAA-rahrbc = polyacrylamide-derivatized rahrbc. LP = lower phase. A l l PAA-rahrbc concentrations were 0.2 mg/mL. The same approach was applied to the separation of the MBL-2 sub-lines. The partitioning of MBL-2 cells with PAA-YE1.48. 10. was less straight forward than the erythrocyte partition. The MBL-2 cells tended to aggregate in a 5,4,1 system even at extremely low ligand concentration. The resulting stabilization of the emulsion formed when the two phases were mixed meant that the separations were taking in excess of 45 minutes. The aggregation made i t impossible to count the cel l s by impedance counting. In order to avoid aggregation a system with the tonicity maintained using sorbitol (5,4,S) was used. Sorbitol is non-ionic and less l i k e l y to promote aggregation than high ionic strength phosphate buffer due to electrostatic repulsion between cells. The PAA-YE1.48.10. ligand had no effect on the partition of either MBL-2 ce l l in a 5,4,S system but changing to a 7,4,S system had a large effect. The two MBL-2 c e l l lines partitioned quite differently in this system, as shown ln Fig. 42. The cell s partitioned increasingly to the lower phase with increasing PAA-YE1.48.10. concentration. At a PAA-YE1.48.10. concentration of 0.2 mg/mL, 27% of the MBL-2(2.6) and 55% of the MBL-2(4.1) cells were in the lower phase. At this point the partition of MBL-2 cells preincubated with PAA-YE1.48.10. then washed three times in PBS was measured. Under these conditions the number of cells in the lower phase was 7 and 14% for MBL-2(2.6) and MBL-2(4.1) cells respectively. The native lower phase 121 60 -a FRACTION Fig. 41. CCD profil e of rabbit (•) and human erythrocytes (*) in the presence of PAA-r a hrbc (0.08 mg/mL) in a 5,4,1 phase system at room temperature. G values are 0.9 and 0.22 for rabbit and human erythrocytes respectively. The dashed line (- -) represents the theoretical CCD plot for these G values. 122 c e l l partition was less than 2%. If the antibody had remained sufficiently bound during washing to cause the cells to partition strongly in to the lower phase, i t would be possible to economize on costly MAb. However, this implies that the removal of the ligand from the c e l l s after the separation would be relatively easy. The separation was not as efficient as that of the human and rabbit erythrocytes, likely reflecting the higher antigen density on the human erythrocytes. It was necessary, therefore, to run a longer CCD to separate the two MBL-2 cel l types. Theoretical CCD profiles for the appropriate G values are shown in Appendix II. A 60 transfer CCD was run and is shown, along with the corresponding theoretical plot, in Fig. 43. As expected the peaks are somewhat broader than the theoretical CCD due to c e l l heterogeneity but the two populations were resolved with 95% purity and by discarding overlapping cavities 14 to 24, the cells could be completely separated i f this were more important than yield. The smaller area under the MBL-2(2.6) peak (Fig. 43) implies that fewer of these cells were present but in fact equal amounts of both cells were in the load mix. The MBL-2(4.1) cells were labelled with 1 4C amino acids and the numbers of MBL-2(2.6) cells were obtained by impedance counting. Aggregation would decrease the numbers of cells counted by impedance counting but not the numbers of radiolabeled cells. The CCD was 1 4 repeated with both c e l l lines labelled, MBL-2(4.1) with C and 125 MBL-2(2.6) with I and is shown in Fig. 44. The profile is essentially the same as that obtained earlier (Fig. 43) but the areas under the peaks are similar, supporting the idea that the apparent loss of MBL-2(2.6) cells in the previous CCD was due to c e l l aggregation. The effect of increasing dextran concentration is not easily explained in the context of current particle a f f i n i t y partition theory. Increasing the dx concentration from 5 to 7% would increase the 3 3 2 interfacial tension (y ) from 3.8x10 to 13.5x10 erg/cm (Albertsson, TB 1986). This would result in a greater free energy of interfacial adsorption, i.e. the opposite effect to that which is observed in this case. The other effect of increasing the dx concentration is to increase the values of K for both polymers. In this example the K is increased PEG from 4.0 to 7.0 while K, only changes slightly. This favors an increase dx in the lower phase partition of both the cells and the ligand as is 123 observed. However as can be seen from eqn 16, the interfacial tension (y ) in principle plays a stronger role in determining c e l l partition T B than the difference in ce l l surface free energy (Ay). In order to test the reproducibility of this separation, the entire procedure was repeated. The YE1.48.10. MAb was purified from a different batch of culture supernatant and the PAA-derivatization was carried out at the same 17:1 hydroxyl:acrylamide reaction ratio. The cells used for this separation had only been passed twice. The results of two CCDs performed on the radiolabelled MBL-2 c e l l lines are shown in Figs. 45 and 46. The G values for MBL-2(4.1) and MBL-2(2.6) cells were 0.19 and 0.41 respectively. The peaks are located in approximately the same fractions as in the previous separations shown in Figs 43 and 44 but they are much narrower, likely-a reflection of less heterogeneity in the MBL-2 cells as a result of their being in only their second passage. As these c e l l lines are changing constantly as illustrated by the inconsistent partition, i t was important to ensure that they were, in fact, the same with respect to binding of the YE1.48.10. MAb. Large amounts of the original c e l l line were frozen in early stages of growth. These vials of original MBL-2 cells are immunologically similar to the MBL-2 cells used in the binding analysis. Binding assays which estimated the antigen density from c e l l pellet counts did not show evidence of any change in surface antigen density. In this repeat experiment (Figs. 45 and 46) efforts were made to reduce the variation in c e l l surface properties and to ensure that they were antigenically similar to the original c e l l line used in the binding analysis by using recently thawed cells . 124 60q [PAA-YE1.48.10.] mg/mL Fig . 42. The effect of increasing concentration of polyacrylamide-derivatized YE1.48.10. on the part i t ion of MBL-2(4.1) (*) and MBL-2(2.6) ( O ) c e l l s in a 7,4,S system at 4 ° . 125 25-3 FRACTION Fig. 43. CCD profi l e for MBL-2(2.6) ( * ) and MBL-2(4.1) c e l l s ( + ) in a 7,4,S system at 4°. The G values are 0.22 and 0.44 for MBL-2(4.1) and MBL-2(2.6) c e l l s respectively. The MBL-2(4.1) ce l l s were labelled with C and the MBL-2(2.6) cells were counted by impedance counting. The dashed line (- -) illustrates the theoretical CCD plot for these G values. 126 Fig. 44. CCD profile for MBL-2C2.6) ( -) and MBL-2(4.1) ce l l s ( ) in a 7.4.S system at 4°. The G values are 0.22 and 0.44 for MBL-2(4.1) and MBL-2(2.6) c e l l s respectively. (MBL-2(4.1) ce l l s were labelled with 1 4C and MBL-2(2.6) c e l l s with 1 2 5 1 . 127 Fig. 45. CCD profile for MBL-2(2.6) (- - -) and MBL-2U.1) ce l l s ( ) in a 7,4,S system at 4°. The G values are 0.18 and 0.41 for MBL-2(4.1) and MBL-2(2.6) ce l l s respectively. (MBL-2(4.1) cells were labelled with 1 4C and MBL-2(2.6) c e l l s with 1 2 5 I . This CCD profile is different to those shown in Figs. 43 and 44 as i t uses a different PAA-YE1.48. 10. preparation and MBL-2 ce l l s in an earlier stage of growth. 128 Fig. 46. CCD profile for MBL-2(2.6) ( ) and MBL-2(4.1) c e l l s ( ) in a 7,4,S system at 4°. The G values are 0.18 and 0.41 for MBL-2(4.1) and MBL-2(2.6) c e l l s respectively. (MBL-2(4.1) cells were labelled with 1 4C and MBL-2(2.6) c e l l s with 1 2 5 I . This is a repeat of T i g . 45. 129 SUMMARY C H A P T E R 6 Two d i f f e r e n t c e l l separation problems were accomplished i n three ways by means of immunoaffinity p a r t i t i o n . Rabbit and human erythrocytes were separated with a trypan blue-derivatized sheep anti-mouse F c IgG and monoclonal mouse anti-NN glycophorin as well as with a polyacrylamide g r a f t copolymer of rabbit anti-human erythrocyte IgG. The other separation was of two sub-lines of a transformed mouse lymphocyte l i n e , MBL-2(2.6) and MBL-2(4.1). Trypan blue was examined as an a f f i n i t y ligand modifying agent. When compared with PEG 1900 for the modification of BSA, i t was necessary to use a larger number of attached trypan blue molecules per IgG to achieve s i m i l a r p a r t i t i o n c o e f f i c i e n t s but the smaller molecular weight of trypan blue compared to PEG 1900 suggested that trypan blue modification may be less deactivating. However, a model separation of rabbit and human erythrocytes using trypan b l u e - d e r i v a t i z e d sheep anti-mouse F c was quite e f f i c i e n t , r e s u l t i n g in the increase from 18% to 70% of human erythrocytes in the upper phase. The upper phase p a r t i t i o n of rabbit erythrocytes under the same conditions was also increased from 8% to 20%. Trypan blue modification seems to be le s s d e a c t i v a t i n g than PEG-modification as a s i m i l a r experiment using a PEG 1900-modified antibody only increased erythrocyte p a r t i t i o n from 10% to 30% i n a system with a lower i n t e r f a c i a l tension (Stocks and Brooks, 1988). However i t i s less s p e c i f i c because the rabbit erythrocyte p a r t i t i o n was also increased from 8% to 20% while i n the PEG-antibody experiment the rabbit erythrocyte p a r t i t i o n was unaffected. The trypan b l u e - d e r i v a t i z e d antibody was used i n several other molecular a f f i n i t y p a r t i t i o n experiments proving i t to be a general ligand, a major advantage of the second antibody immunoaffinity technique. In an e f f o r t to achieve a c l i n i c a l l y more useful separation by immunoaffinity p a r t i t i o n and, more s p e c i f i c a l l y , to model two c l i n i c a l c e l l separation problems, namely bone marrow purging i n combination with autologous bone marrow transplantation as leukemia therapy and the separation of f e t a l i s l e t tissue f or implantation as diabetes treatment, a lymphocyte separation was attempted. This was of two sub-lines of a 130 transformed mouse lymphocyte, MBL-2(4.1) and MBL-2(2.6), which d i f f e r e d i n the surface density of an antigen recognized by the rat MAb, YEl.48.10. Binding studies showed that at saturation 2.4 x 10 6 molecules of YEl. 48. 10. were bound to a MBL-2(4.1) c e l l and 8 x 10 5 molecules of YEl.48.10. were bound to a MBL-2(2.6) c e l l . This small d i f f e r e n c e i n surface antigen density represents an extremely stringent separation problem. Previous separations by immunoaffinity p a r t i t i o n have been of species s p e c i f i c erythrocytes and the most stringent separation to date has been on the basis of a surface antigen d i f f e r e n c e of 0 and 5 x 1 0 U 2 molecules of IgG per cm (Stocks and Brooks, 1988). Experiments with erythrocytes and a PEG-palmitate ligand (Sharp, 1985) suggested that the MBL-2 separation would be close to the l i m i t s of immunoaffinity p a r t i t i o n . The MBL-2 c e l l s were indi s t i n g u i s h a b l e by t h e i r native p a r t i t i o n or with a PEG-linoleate ligand, a f a t t y a c i d ester commonly used to a l t e r c e l l p a r t i t i o n . PEG-derivatized YEl.48.10. d i d not have an e f f e c t on eit h e r MBL-2 c e l l despite ELISA assays suggesting that the antibody was s t i l l r e l a t i v e l y active compared to previous. PEG-antibody a f f i n i t y ligands (Stocks and Brooks, 1988; Sharp et al, 1986). A second MAb, RG7/11.1, which bound to the F c region of rat IgG was obtained. Unfortunately binding studies on RG7/11.1 showed that t h i s MAb bound only approximately 10% of of the YEl.48. 10. bound to the MBL-2 c e l l at saturation. As expected i n l i g h t of the poor a f f i n i t y of RG7/11.1 f o r YEl.48.10., a combination of PEG-derivatized RG7/11.1 and YEl.48.10. d i d not a l t e r the p a r t i t i o n of eit h e r MBL-2 c e l l . The next system studied was a b i o t i n - d e r i v a t i z e d YEl.48.10. i n combination with PEG-avidin. The b i o t i n - d e r i v a t i z e d YEl.48.10. remained quite a c t i v e ; at saturation 1.3 x 10 6 and 0.67 x 10° molecules per c e l l were bound to MBL-2(4.1) and MBL-2(2.6) c e l l s respectively. B i o t i n d e r i v a t i z a t i o n alone a c t u a l l y caused the YEl.48.10. to p a r t i t i o n into the upper phase and the addition of PEG-avidin moved almost a l l the ligand into the upper phase. Although the PEG-avidin, biotin-YEl.48.10. system worked well on the molecular l e v e l i t did not achieve a separation of the MBL-2 c e l l s . The b i o t i n - d e r i v a t i z e d YEl.48.10. alone di d e f f e c t a small d i f f e r e n t i a l p a r t i t i o n of the MBL-2 c e l l s but a di f f e r e n c e of t h i s magnitude would require 200 CCD transfers to resolve the two MBL-2 sub-lines. The combination of PEG-avidin and b i o t i n -131 YE1.48.10 actually caused a slight decrease in MBL-2 c e l l partition. Due to the inability of any of the upper-phase partitioning ligands to discriminate between the MBL-2 sub-lines i t was decided to try a lower-phase partitioning ligand. APTS can easily be optimized such that a l l the cells are located at the interface thus a lower-phase partition is equally useful as an upper-phase partition. Polyacrylamide graft copolymers of rabbit anti-human erythrocyte IgG were used for the separation of rabbit and human erythrocytes. This effected a dramatic immunospecific separation at low antibody concentrations. The partition of human erythrocytes was altered from 12% to 93% lower phase partition while the partition of rabbit erythrocytes was unaffected. Larger differences were achieved in other systems but the results quoted also gave a low rabbit erythrocyte partition. This is the most efficient immunoaffinity partition separation of cells to date. Karr et al (1986) increased the partition of human erythrocytes from 14% to 58% with no effect on sheep erythrocyte partition with a PEG-derivatized antibody and Sharp et al (1986) increased human erythrocyte partition from 15% to 28% with no effect on rabbit erythrocytes. These separations required CCDs of 30 and 60 transfers respectively to resolve the cells completely whereas the polyacrylamide ligand achieved this in 20 transfers. The same method was used to synthesize a polyacrylamide graft copolymer of YE1.48.10. Significant problems due to aggregation of MBL-2 cell s were encountered in systems buffered by phosphate. This was avoided with the use of sorbitol. No effect of the ligand was observed in a 5,4,S system but increasing the dextran concentration to 7% optimized the system for this ligand. The polyacrylamide-derivatized YE1.48.10. increased the partition of MBL-2(2.6) from 8% to 26% and MBL-2(4.1) from 8% to 55% in the lower phase. This was sufficient to allow resolution of the MBL-2 cells in a CCD with 60 transfers. This result was surprising in light of the current particle a f f i n i t y partition theory which predicts that effect of the increased surface tension in this system would dominate over the effect of the increased partition coefficient of PEG resulting in less not more ce l l s partitioning in to the lower phase. One useful outcome of the study was the development of a new method of producing and purifying large quantities of MAb by means of culture 132 in supplemented serum-free medium, u l t r a f i l t r a t i o n and FPLC (Stocks and Brooks, 1989). This method produces antibodies with a high retention of activity compared to those purified by ammonium sulfate precipitation. Overall this study has shown that cells may be separated on the 11 11 basis of small antigenic differences such as 3.83x10 and 1.29x10 2 molecules/cm . A polyacrylamide-derivatized ligand has a much stronger effect on c e l l partition than a ligand derivatized by trypan blue, PEG or biotin. As TAAs recognized by MAbs are within this order of antigen density this study supports the potential of immunoaffinity partition for use in bone marrow purging and fetal i s l e t c e l l isolation. 133 GLOSSARY OF TERMS Acinar c e l l - a secretory c e l l in an acinous gland. Agarose - a linear polymer of alternating D-galactose and 3,6-anhydro-L-galactose. Avidin - a basic glycoprotein (M ~ 66,000) that binds biotin. Biotin - 2' -oxo-3,4-imidazollne-2-tetrahydrothiophene-n-valeric acid (vit H), is bound tightly by avidin. Complement system - 23 blood protein components activated by antibodies to bind and thereby eliminate foreign substances. Concanavalin A (Con A) - a lectin and mitogen used to activate lymphocytes. Confluent - a term which describes a cultured c e l l population which occupies a l l the available growth surface. Dextran - poly a(l,6)-D-glucose. F fragment - a region of the IgG molecule defined by papain digestion which is constant irrespective of antibody speci f i c i t y (MW « 50 000 - 75 000). F i c o l l - a synthetic copolymer of sucrose and epichlorohydrin. Fluorescamine - 4-phenylspiro[furan-2H(3H),1'-phthalan]-3,3' dlone. HAT selection - a method of selecting hybridoma c e l l lines by growth in medium containing hypoxanthine, aminopterln and thymidine which only supports hybridoma growth. Hypaque - sodium diatrizoate (3,5-Bis[acetylamino]-2,4,6-tri-iodobenzoic acid). Hybridoma - a continuously growing c e l l line formed by the fusion of a malignant and normal c e l l . Lectin - a group of antibody-like proteins found in plants which tend to bind to specific c e l l surface glycolipids or glycoproteins. Leukemia - progressive proliferation of abnormal leukocytes. Lymphoma - abnormally proliferative, generally neoplastic disease of the lymphoid system. Mitosis - nuclear division in the somatic c e l l s of eukaryotes. Myeloma - a tumor consisting of a malignant form of plasma c e l l . Neuraminic acid - 5-amino-3,5-dideoxy-D-glycero-D-galactonoulsonic acid. 134 Passage - the dilution of c e l l cultures which have become confluent in order to maintain a viable culture. Polyethylene glycol (PEG) - a polymer of ethylene glycol. Polyacrylamide (PAA) - a linear polymer of acrylamide. Protein A - a protein made by most strains of Staphylococcus aureus whi binds IgG. S i a l i c acid - N and 0 acetyl derivatives of neuraminic acid. Stem c e l l s - usually immature, undifferentiated cells capable of rapid division and differentiation T-, B-lymphocyte - thymus-dependent, thymus independent colorless nucleated blood cells. Thalassemia - an inherited disorder of hemoglobin metabolism. 135 GLOSSARY OF SYMBOLS AND ABBREVIATIONS A particle surface area A absorbance at 500 nm 5 0 0 a the activity of i l ANLL acute non-lymphoblastic leukemia APTS aqueous polymer two-phase system BNHS N-hydroxysuccinimidobiotin BSA bovine serum albumin CCD counter current distribution DMEM Dubellco modified Eagle medium dx dextran 18AO-PEG PEG-linoleate ELISA enzyme-linked immunosorbent assay f molal activity coefficient FACS fluorescent activated c e l l sorting FBS fetal bovine serum FITC fluorescein isothiocyanate FPLC fast protein liquid chromatography G distribution ratio in a CCD fraction = KV /V where K = partition t b coefficient and v t / v b is the volume ratio of upper to lower phase. H hematocrit (%v/v packed cells) HABA 4-hydroxyazobenzene-2'-carboxylic acid HLA .. human lymphocyte antigens HRP horseradish peroxidase IgG immunoglobulin G k Boltzmann's constant, 1.36 x 10 1 6 erg/molec.K° K partition coefficient or c r i t i c a l point on a phase diagram K L partition coefficient of free ligand K partition coefficient of unbound macromolecules o k or k microscopic association constant a krf microscopic dissociation constant L ligand m density of i l J M the concentration of molecules which have i of their sites occupied 136 number average molecular weight M weight average molecular weight w MAb monoclonal antibody MBL-2 a transformed lymphocyte c e l l line which has two sub-lines, MBL-2(4.1) and MBL-2(2.6), defined by monoclonal antibody YE1.48.10. MW - molecular weight m a NN glyc. monoclonal mouse anti-NN glycophorin IgG N Avogadros number A P number of polymer segments (the ratio of polymer molecular volume . to solvent molecular volume or a binding site on a c e l l surface or fraction of total amount of material in the upper phase of a CCD cavity PAA polyacrylamide PBS phosphate buffered saline PCC 2(alkoxypolyethyleneglycoxy)-4,6 dichlorotriazine ("activated PEG") PEG-S polyethylene glycol sulfonate PEG X polyethylene glycol, mol. wt. approx. X g/mole R molar gas constant, 8.314 x 107 erg/mol.K° r a hrbc rabbit anti-human erythrocyte IgG RG7/11.1 a mouse monoclonal antibody specific for rat F c fragment RPMI-1640 a defined c e l l culture medium S specific activity of a radiolabelled antibody or c e l l s a mFc sheep anti-mouse fragment IgG SDS-PAGE sodium dodecyl sulfate polyacrylamide gel electrophoresis SSFM supplemented serum-free medium TAA tumor associated antigen TB trypan blue TLL tie line length on a phase diagram TMA-PEG trimethylamino polyethylene glycol YE1.48.10. a rat monoclonal antibody specific for an antigen expressed on MBL-2 cells w^  weight of sample i X Flory interaction parameter A E t i free energy of particle interface attachment AG free energy of mixing m Ay surface free energy difference between phases y interfacial tension between the phases 137 Tf surface free energy 9^ radioact ive counts of sample i r surface excess of the i t h component 7)^  v i s c o s i t y of i u°, u standard state chemical po ten t ia l , chemical potent ia l * volume f rac t ion <b f r ac t iona l saturat ion of s i t e i v molecules of l igand bound per unit surface area 138 Appendix I The error in a calculated quantity is estimated as the root mean square of the errors introduced by the basic measured quantities. The contribution to the f i n a l quantity y of error in measuring a basic quantity x is given by the partial derivative 6y/5x. Thus the total error is given by: Ay = Sx l 1 5y 5x Ax )' 2 Sy Sx AX r n or Ay = S (|X ' Ax ) 2 1=1 5x i l where 5y/5x are the partial derivative, Ax^  the measurement error and Ay the error in the calculated value (Mendenhall et al, 1981.) Partial Derivatives for v: [(* (w -w )/(w - w )] - [(y [(w- w ) - (w - w )H]/(w - w )] \^  1 3 1 . 2 3 2 4 1 4 3 4 5 (w - w ) .H.S 4 3 Let [(y (w -w )/(w - w ) ] - [ (y [ (w - w ) - (w - w )H]/(w - w )] = [N] 1 3 1 2 3 2 4 1 4 3 4 5 5v 5H -vS + y /(w - w ) 2 4 5 S-.H 8S -v/S 8v 5y l (w - w ) .H.S 4 3 (w - W ) 3 1 (w - W ) 2 3 139 8v Sy 2 (w - w ) .H.S 4 3 (w - w ) - (w - w )H 4 1 4 3 (w - W ) 4 5 Sw l (w - w ) .H.S 4 3 + (w - w ) (w - w ) 2 3 4 5 Sw (w - w ) .H.S 4 3 -y (w - w ) 1 3 1 (w - w )' 2 3 5v Sw [N] (w - w ) .H.S 4 3 (w - w ) .H.S 4 3 (w - W ) 2 3 1 + (w - w ) 3 1 (w - W ) 2 3 - *2 H (w - W 4 5v Sw -IN] (w - w ) .H.S 4 3 (w - w )(w - w )H.S 4 3 4 5 1 - H -(w - w ) 4 1 (w - W 4 (w r w ) 4 5 - * 2[ ( y V - ' v w 3) H ] 5 W 5 (w- w ).H.S (w -w ) 2 4 3 4 5 P a r t i a l Derivatives f o r v/L: v _ {(y (w -w )/(w - w )] - t(y [ (w - w ) - (w - w )H]/(w - w )}(w - w ) ' 1 3 X 2 3 2 4 1 4 3 4 5 4 5 L (w - w ).H.y 4 3 2 V 1 L H y (w - w ) (w - w ) (w - w ) 1 3 1 4 5 4 1 2 (w - W ) (W - W ) (w - W ) 2 3 4 3 4 3 J + 1 140 5{v/L) 6H 6{u/L) 1 (w 3- W i ) ( w 4 - w5) Hr (w - w )(w - w ) 2 2 3 4 3 8{v/L) " y i ( V w i ) ( V w 5 ) H> (w - w ) (w - w ) 2 2 3 4 3 3(iVL) Sw H(w - w ) 4 3 1 -y (w - w ) 1 4 5 y (w - w ) 2 2 3 5(iVL) 5w H(w - w ) 2 3 Vy w 5 ) ( v w i ) Vy w 3 ) ( v w3} 5(iVL) 5w H(w - w ) 4 3 Vy V Vy w 3 ) (w -1 + (w -<3(iVL) Sw H(w - w ) 4 3 y i ( W 3 ~ W l } Vy V i -(W (w 8(v/L) 5w - 1 H(w - w ) 4 3 Vy V Vy V a ( i v u SS = 0 141 Appendix I I T h e o r e t i c a l counter current d i s t r i b u t i o n p r o f i l e s f o r the c e l l separations by CCD in t h i s study. 1. Human ( -, G = 0.22) and rabbit ( , G = 0.9) erythrocytes i n the presence of polyacrylamide-derivatized rabbit anti-human erythrocyte. a. n = 10 40-1 w o 30-20-10-b. n = 15 40' 30- / / \ 142 1 4 3 144 REFERENCES Abuchowski, A., McCoy, J.R., Palczuk, N.C., Van Es, T. anri Davis, F.F. (1977) E f f e c t of covalent attachment of polyethylene g l y c o l on immunogenicity and c i r c u l a t i n g l i f e of bovine l i v e r catalase. J. Biol. Chem. 252, 3582-35S6. Adamson, A.W. (1976) Physical Chemistry of surfaces, 3rd ed. Wiley, New York. Akemi, H., Aoyagi, T, Shinohara ,1, Okano, T, Kataoka, K. and Sakurai, Y. (1986) Synthesis of a new antithrombic block copolymer containing f l u o r i n a t e d segments poly(nonafluorohexyl acrylate-/3-styrene) Hakromol. Chem. 187 (7), 1627-1638. Alberts, B.M. (1967) F r a c t i o n a t i o n of nucl e i c acids by dextran-polyethylene g l y c o l two-phase systems. Neth. Enzymol. 12A, 566-581. Albertsson, P-X. (1956) Chromatography of c e l l s and c e l l fragments. Nature (London) 177, 771-774. Albertsson, P-X. (1958) P a r t i t i o n of proteins i n l i q u i d polymer-polymer two-phase systems. Nature (London) 182, 709-711. Albertsson, P-X. (1959) The behavior of p a r t i c l e s and proteins i n l i q u i d polymer-polymer two-phase systems. In "Symposium uber Makromolekule i n Wiesbaden" Int. Union Reine Angew. Chem., Vol. 5, C2. Verlag Chemie, Weinheim. Albertsson, P-X. (1960) P a r t i t i o n of C e l l P a r t i c l e s and Macromolecules. Almqvist & Wiksell, Stockholm; Wiley, New York. Albertsson, P-X. (1965) Thin layer countercurrent d i s t r i b u t i o n . Anal. Biochem. 11, 121-125. Albertsson, P-X. (1970) P a r t i t i o n of c e l l p a r t i c l e s and macromolecules i n two-phase systems. Adv. Protein Chem. 24, 309-341. Albertsson, P-X. (1973) A p p l i c a t i o n of the phase p a r t i t i o n method to a hydrophobic membrane protein, phospholipase Al from Escherichia coli. Biochem. 12, 2525-2530. Albertsson, P-X. (1986) P a r t i t i o n of C e l l P a r t i c l e s and Macromolecules, 3rd ed. Wiley (Interscience), New York. Albertsson, P-X. and Baird, G. (1962) Counter-current d i s t r i b u t i o n of c e l l s . Exp. Cell Res. 28, 296-322. Albertsson, P-X. and Nyns, E.J. (1959) Counter-current d i s t r i b u t i o n of proteins i n aqueous polymer phase systems. Nature (London) 184, 1465-1468. Alber tsson, P-X. and Nyns, E.J. (1961) P a r t i t ion of proteins i n an aqueous phase system of dextran and polyethylene g l y c o l . Influence of the 145 e l e c t r o l y t e content. Ark. Kemi. 17, 197-206. Albertsson, P-X. and Phillpson, L. (1960) Antigen-antibody in l i q u i d two-phase systems: A method for studying immunological reactions. Nature (London) 1S5, 3S-40. Albertsson, P -X. , Hanzon, V. and Toschi, G. (1959) I s o l a t i o n of ribonucleoprotein p a r t i c l e s from rat brain microsomes by a l i q u i d two-phase system. J. Ultrastruct. Res. 2, 366-372. Anasetti, C. , Amos, D., Beatty, P.G., Appelbaum, F.R., Benslnger, W., Buckner, CD., C l i f t , R. , Doney, K. , Martin, P.J. and Mlckelson, E. (1989) E f f e c t of HLA-compatability on engraftment of bone marrow transplants i n patients with leukemia or lymphoma. New Eng. J. Ned. 320, 197-204. Anderson, N.G. (1966) An introduction to p a r t i c l e separations i n zonal centrifuges. Natl. Cancer Inst. Monogr. 21, 9-13. A u i t i , F. , C a r o t t i n i , J . C , Coombs, R.R.A., Cooper, M. , Dickler, H.B., Froland, S.S., Fundenberg, H.H., Greaves, M.F., Grey, H.M., Kunkel, H.G., Natvig, J.B., Preud'homme, J.L., Rabellino, E. , R l t t s , R.E.', Rowe, D.S., Seligmann, M., Sle g a l , F.P., Stjernsward, J., Terry, W.D. and Wybran, J. (1974) I d e n t i f i c a t i o n , enumeration and i s o l a t i o n of B- and T-lymphocytes from human peripheral blood: Report of a WHO/IARC-sponsored workshop on human T- and B - c e l l s , London, 15-17 July 1974. Scand. J. Immunol. 3, 521-532. ' Axelsson, CG. and Shanbhag, V.P. (1976) Histone-hydrocarbon i n t e r a c t i o n . P a r t i t i o n of histones i n aqueous two phase systems containing polyethylene glycol-bound hydrocarbons. Eur. J. Biochem. 71(2), 419-423. Axelsson, C G . (1978) Changes i n structure and hydrophobic surface properties of /3-lactoglobulin determined by p a r t i t i o n in aqueous two-phase polymeric systems. Biochim. Biophys. Acta 533, 34-42. Baird,, D., Albertsson, P-X. and von Hofsten, B. (1961) Separation of ba c t e r i a by counter-current d i s t r i b u t i o n . Nature (London) 192, 236-239. Baran, M.M., All e n , D.M., Russ e l l , S.R., Scheetz, I.I. and Monthony, J.F. (1982) C e l l s o r t i n g using a u n i v e r s a l l y applicable a f f i n i t y chromatography matrix: solid-phase a n t i - f l u o r e s c e i n isothiocyanate antibody. J. Immunol. Meth. 53, 321-324. Barber, W.H., Diethelm, A.G., Laskow, D.A., Deierhoi, M.H., J u l i a n , B . A . and C u r t i s , J.J. (19S9) Use of cryopreserved donor bone marrow i n cadaver kidney a l l o g r a f t r e c i p i e n t s . Transplantation 47, 66-71. Basch, R.S., Berman, J.W. and Lakow, E. (1983) C e l l separation using p o s i t i v e immunoselective techniques. J. Immunol. Meth., 56, 269-280. Bateman, W.G. (1916) Egg proteins. J. Biol. Chem., 26, 263. Becher, P. (1965) Emulsions: Theory and pract i c e , 2nd ed. 169. Van 146 Nostrand-Rheinhold, Princeton. Beauchamp, CO., Gonias, S.L., Menapace, D.P. and Pizzo, S.V. (1983) A new procedure for the synthesis of polyethylene g l y c o l - p r o t e i n adducts; e f f e c t s on function, receptor recognition and clearance of superoxide dismutase, l a c t o f e r r i n and a^-macroglobulin. Anal. Biochem.. 131, 25-33. Benard, J., Bettan-Renaud, L., Gavoille, A., Pico, J.L., Beaujean, F., Lopez, M. and Riou, G. (19S8) In vitro chemical e r a d i c a t i o n of small c e l l lung cancer: A p p l i c a t i o n i n autologous bone marrow transplantation. Eur. J. Cancer Clin. Oncol. 24, 1561-1566. Bidney, D.L. and Reeck, G.R. (1977) Association products and conformations of s a l t - d i s s o c i a t e d and acid-extracted histones. A two-phase procedure for i s o l a t i n g s a l t - d i s s o c i a t e d histones. Biochem. 16, 1844-1849. Blomquist, G. (1976) Cross p a r t i t i o n and determination of net charge of the Isoenzymes of enolase. Biochim. Biophys. Acta 420, 81-86. Boas, M.A. (1927) E f f e c t of dessicat i o n upon n u t r i t i v e p r o p e r t i e s of egg white. Biochem. J., 21, 712-724. Bonnafous, J . C , Dornand, J., Favero, J. , Sizes, M. , Boschetti, E. and Mani, J.C (1983) C e l l a f f i n i t y chromatography with ligands immobilized through cleavable mercury-sulfur bonds. J. Immunol. Meth., 58, 93-107. Bosmann, H.B., Bieber, G.F., Brown, A.E., Case, K.R., Gersten, D.M., Kimmerer, T.W. and Lione, A. (1973) Biochemical parameters c o r r e l a t e d with tumor c e l l implantation. Nature (London) 246, 487-489. Brooks, D.E, Seaman, G.V.F. and Walter, H. (1971) Detection of di f f e r e n c e s i n surface-charge-associated properties of c e l l s by p a r t i t i o n i n two-polymer aqueous phase systems. Nature (London), New Biol. 234, 61-62. Brooks, D.E., Sharp, K.A. and Fisher, D. (1985) Theoretical aspects of p a r t i t i o n i n g . In " P a r t i t i o n i n g i n Aqueous Two-Phase Systems" (H. Walter, D.E. Brooks, D. Fisher, eds.) 11-84, Academic Press, New York. Brown, J., Clark, W.R., Molnar, I.G. and Mullen, Y.S. (1976) F e t a l pancreas transplantation for reversal of streptozotocin-induced diabetes i n r a t s . Diabetes 25, 56-64. Brown, J., Danilovs, J.A., Clark, W.R. and Mullen, Y.S. (19S4) F e t a l pancreas as donor organ. World J. Surg. 8, 152-157. Bruck, C , P o r t e l l e , D. , Glineur, C. and Bolton, A.J. (1982) One-step p u r i f i c a t i o n of mouse monoclonal antibodies from a s c i t i c f l u i d by DEAE A f f i - G e l blue chromatography. J. Immunol. Meth. 53, 313-319. Brunette, D.M. and T i l l , J.E. (1971) A rapid method for the i s o l a t i o n of L - c e l l surface membranes using an aqueous two-phase polymer system. J. Nembr. Biol. 5, 215-224. 147 Cantor, CR. and Schimmel, P.R. (1980) Biophysical Chemistry, Part III. 849-886, Freeman, San Francisco. Cartron, J.P. and Nurden, A.T. (1979) Galactosyl transferase and membrane glycoprotein abnormality i n human p l a t e l e t s from Tn-syndrome donors. Nature (London), 282, 621-623. Chan, P-Y and Takei, F. (1986) Expression of a T C e l l r e c e p t o r - l i k e molecule on normal and malignant murine T c e l l s detected by Rat monoclonal antibodies to nonclonotypic determinants. J. Immunol. 136, 4, 1346-1353. Chang, J., Morgenstern, G., Deakin, D., Testa, N.G., Countinho, L., Scarffe, J.H., Harrison, C. and Dexter, T.M. (1986) Re c o n s t i t u t i o n of haemopoietic system with autologous marrow taken during relapse of acute myeloblastic leukaemia and growth i n long-term culture. Lancet i , 294-295. Clark, W.R. and Rutter, W.J. (1972) Synthesis and accumulation of i n s u l i n i n the f e t a l rat pancreas. Dev. Biol. 29, 468-4S1. Clezardin, P., McGregor, J.L., Manach, M., Boukerche, H. and Dechavanne, M. (1985) One-step procedure for the rapid i s o l a t i o n of mouse monoclonal antibodies and t h e i r antigen binding fragments by fa s t p r o t e i n l i q u i d chromatography on a mono Q anion-exchange column. J. Chromatog. 319, 67-79. Craig, L.C. (1960) P a r t i t i o n . In "A Laboratory Manual of A n a l y t i c a l Methods of Pr o t e i n Chemistry vol 1" (P. Alexander and R.J. Block, eds.) 122-160, Pergamon, Oxford. Davis, B.J. (1964) Disc electrophoresis - II. Method and a p p l i c a t i o n to human serum proteins. N.Y. Acad. Annuals. 121, 404-427. Dicke, K.A., Zander, A.R., Spitzer, G., Verma, D.S., Peters, L.J., Vellekoop, C., Thomson, S., Stewart, D., Hester, J.P. and McCredie, K.B. (1979) Autologous bone marrow transplantation i n relapsed adult acute leukemia. Exp. Hematol., 7 (suppl. 5), 170-187. Dhokia, B., Pectasides, D. , S e l f , C., Habib, N.A., Hershman, M., Wood, C.B., Munro, A.J. and Epenetos, A.A. (1986a) A low pH enzyme linked immunoassay using two monoclonal antibodies f or the s e r o l o g i c a l d e t ection and monitoring of breast cancer. Br. J. Cancer 54, 885-889. Dhokia, B., Canney, P.A., Pectasides, D., Munro, A.J., Moore, M. , Wilkinson, P.M., S e l f , C. and Epenetos, A.A. (1986b) A new immunoassay using monoclonal antibodies HMFG1 and HMFG2 together with an e x i s t i n g marker CA125 for the s e r o l o g i c a l detection and management of e p i t h e l i a l ovarian cancer. Br. J. Cancer 54, 891-895. Dobry-Duclaux, A. (1956) Partage des substances macromoleculaires entre deux solvents. Makromol. Chem. 18/19, 317-321. 148 Duke, F.R. and F o r i s t , A.A. (1949) The theory and k i n e t i c s of s p e c i f i c oxidation. III. The cerate -2,3-butanediol r e a c t i o n i n n i t r i c a c i d s o l u t i o n . J.A.C.S. 71, 2790-2792. Duke, F.R. and Bremer, R.F. (1951) Formal p o t e n t i a l of cerium (IV)-cerium (III) couple i n p e r c h l o r i c acid. Anal. Chem. 23, 1516. Economu, J.S., Shim, H.S., Kaiser, H., Santos, G.W. and Schron, D.S. (1978) Bone marrow transplantation i n cancer therapy: I n a c t i v a t i o n by antibody and complement of tumor c e l l s i n mouse syngeneic marrow transplants. Proc. Soc. Exp. Biol. Med. , 158, 449-453. Edelman, G.M., Cunningham, B.A., G a l l , W.A., G o t t l i e r , P.D., Rutishauser. U. and Waxdal, M.J. (1969) The covalent structure of an e n t i r e yG g l o b u l i n molecule. Biochem. 63, 78-S5. Edelman, G.M., Rutishauser, U. a n d M i l l e t t e , C F . (1971) C e l l f r a c t i o n a t i o n and arrangement on f i b e r s , beads and surfaces. Proc. Natl. Acad. Sci., USA, 68, 2153-2157. English, D., Lamberson, R., Graves, V., Akard, L.P., McCarthy, L.J. and Jansen, J. Semiautomated processing of bone marrow g r a f t s f o r transplantation. (1989) Transfusion 29, 12-16. Evans, E.A. and Fung, Y.C (1972) Improved measurements of the erythrocyte geometry. Microvasc. Res. 4, 335-347. Farkas, G. and Karacson, S. (1985) C l i n i c a l t r a n s p l a n t a t i o n of f e t a l human pancreatic i s l e t s . Biomed. Biochem. Acta 44, 155-159. , Fe d e r l i n , K.F. and B r e t z e l , R.G. (19S4) The e f f e c t of i s l e t t r a n s p l a n t a t i o n on complications i n experimental diabetes of the rat. World J. Surg. 8, 69-178. Flanagan, S.D. (1976) A f f i n i t y p a r t i t i o n i n g of membranes: Cholinergic receptor-containing membranes from Torpedo californica. J. Biol. Chem. 251, 858-865. Flanagan, S.D.(1985) P a r t i t i o n i n g of animal membranes and organelles. In " P a r t i t i o n i n g i n Aqueous Two Phase Systems" (H. Walter, D.E. Brooks and D. Fisher, eds. ) 453-495, Academic Press', New York. Flanagan, S.D. and Barondes, S.H. (1975) A f f i n i t y p a r t i t i o n i n g : A method fo r the p u r i f i c a t i o n of proteins using s p e c i f i c polymer-1igands i n aqueous polymer two-phase systems. J. Biol. Chem. 250, 14S4-14S9. Flory, P.J. (1941) Thermodynamics of high polymer solutions. J. Chem. Phys. 9, 660-661. Flory, P.J. (1953) " P r i n c i p l e s of Polymer Chemistry," Chapter 13. Cornell Univ. Press, New York. Forrest, G.C and Landon, J.R. (1978) F u l l y automated continuous-flow radio-immunoassay system employing magnetizable p a r t i c l e s . In "Automated 149 Immune-analysis 2" (R.F. Richie, ed.) 567-594, Dekker, New York. F r i c k , G. and Albertsson, P-X. (1959) Bacteriophage enrichment In l i q u i d -l i q u i d two-phase system with subsequent treatment with "Freon" 113. Nature (London) 183, 1070-1072. Furnica, M. (1975) Biochemical mechanisms implied i n r a d i o s e n s i t i z a t i o n by chemical agents to the a c t i v i t y of r a d i a t i o n and c y t o s t a t i s t i c s . VI. The d i s t r i b u t i o n of DNA from tumors of rats r a d i b s e n s i t i z e d by iodoacetamide i n a biphasic system of PEG-dextran polymers. Rev. Roum. Biochim. 12 (2), 91-97. G a l f r e , G. and M i l s t e i n , C. (1981) Preparation of monoclonal antibodies: S t r a t e g i e s and procedures. Meth. Enzymol. 73, 3-46, Gascoine, P.S., Dix, C.J., and Fisher, D. (19S3) Hydrophobic a f f i n i t y c e l l p a r t i t i o n detects a l t e r a t i o n s i n surface properties of erythrocytes i n r a t s bearing a subcutaneous Leydig tumor. Exp. Cell Biol. 51, 322-326. Gemski, M.J., Doctor, B.P., Gentry, M.K., Pluskal, M.G. and S t r i c k l e r , M.P. (1985) Single step p u r i f i c a t i o n of monoclonal antibody from murine a s c i t e s and tissue culture f l u i d s by anion exchange high performance l i q u i d chromatography. Biotechniques 3, 378-385. Gersten, D.M. and Bosmann, H.B. (1974) Behavior i n two-phase aqueous polymer systems of L5178Y mouse leukemic c e l l s . II. The lag and exponential phases of growth. Exp. Cell Res. 88, 225-230. Gersten, D.M. and Bosmann, H.B. (1975) Surface properties of plasma membranes following i o n i z i n g r a d i a t i o n exposure. Exp. Cell Res. 96, 215-223. G i l b e r t , S.F. and Migeon, B.R. (1975) D-Valine as a s e l e c t i v e agent for normal human and rodent e p i t h e l i a l c e l l s i n culture. Cell 5, 11-17. Goding, J.W. (1983) Monoclonal antibodies: P r i n c i p l e s and prac t i c e . Academic Press Inc., New York. Goff, C.G. (1974) Chemical structure of a modification of the Escherichia coli r i b o n u c l e i c aci d polymerase a polypeptides introduced by the bacteriophage T i n f e c t i o n . J. Biol. Chem. 249, 6181-6190. Golub, E.S. (1977) "The C e l l u l a r Basis of the Immune Response," Sinauer Associates Inc., Sunderland, Massachusetts. Graham, M.D. and Sel v i n , P. (1982) Separation of lanthanide binding c e l l s . IGEE Trans. Magn. 18, 1523-1525. Green, N.M. (1965) A spectrometric assay f o r a v i d i n and b i o t i n based on binding of dyes by avidin. Biochem. J. 94, 23-24. Green, N.M. (1970) Spectrophotometric determination of a v i d i n and b i o t i n . Meth. Enzymol. ISA, 418-421. 150 Gribben, J.G., Linen, D.C., Singer, C.R., McMillan, A.K., J a r r e t t , M. and Goldstohe, A.H. (1989) Sucessful treatment of r e f r a c t o r y Hodgkin's disease by high-dose combination chemotherapy and autologous bone marrow trans p l a n t a t i o n . Blood 73, 340-344. Groth, C.G., Andersson, A., Bjorken, C., Gunnarsson, R. , Hellerstrom, C., Lundgren, G., Petersson, B., Swenne, I. and Ostman, J. (1980) Transplantation of f e t a l pancreatic microfragments v i a the p o r t a l v e i n to a d i a b e t i c patient. Diabetes 29 (Suppl. 1), 80-83. Guesdon, J.L. and Aurameas, S. (1977) Magnetic s o l i d phase enzyme-immunoassay Immunochem. 14, 443-447. Gurd, J.W., Gordon-Weeks, P. and Evans, W.H. (1982) Biochemical and morphological comparison of post-synaptic d e n s i t i e s prepared from rat, hamster and monkey brains by phase p a r t i t i o n i n g . J. Neurochem. 39, 1117-1124./': Gyorgy, P., Rose, C.E., Hofmann, K., M e l v i l l e , D.B. and du Vigneaud, V. (1940) A furth e r note on the i d e n t i t y of vitamin H with b i o t i n . Science 92, 609. J ' Gyorgy, P., Rose, C.E., Eakin, R.E., S n e l l , E.E. and Williams, R.J. (1941) Egg-white i n j u r y as the r e s u l t of nonabsorption or i n a c t i v a t i o n with a v i d i n . Science 93, 477-478. Ham, R.G. (1972) Cloning of mammalian c e l l s . Methods Cell Physiol. 5, 37-74. Hannig, K. (1972) Free flow electrophoresis. In "Methods of Microbiology v o l . 5B" (J.R. Norris and D.W. Ribbons, eds.) 513-548, Academic press, New York. Ha r r i s , J.M., Hundley, N.H., Shannon, T.G. and Struck, E.C. (1982) Polyethylene g l y c o l s as soluble, recoverable, phase-transfer c a t a l y s t s . J. Org. Chem. 47, 4789-4791. Hawkey, C.J., Holmes, C.H., Smith, P.J., Austin, E.B. and Baldwin, R.W. (1986) Patterns of r e a c t i v i t y of the monoclonal-antibody 791T/36 with d i f f e r e n t tumor-metastases i n the l i v e r . Br. J. Cancer 54, 871-875. Hecker, E. (1955) Verteilungsverfahren im laboratorium. In "Angewandten Chemie und Chemie-Ingenieur-Technik vol 67" Verlag Chemie, Weinheim. Herbert, W.J. (1973) M i n e r a l - o i l adjuvants and the immunization of laboratory animals. In "Handbook of Experimental Immunology" (D.M. Weir, ed.), Vol. 3, Chapter 2. Blackwell, Oxford. Herzenberg, L.A. Analysis and separation using the fluorescence a c t i v a t e d c e l l s o r t e r (1977). In "Handbook of Experimental Immunology, vol.2" (D.M. Wier, ed.) chap. 22, 1-21, Blackwell S c i e n t i f i c Publications, Oxford. Hudson, L. and Hay, F.C. (1976) In " P r a c t i c a l Immunology. Phagocytosis of ir o n powders" chap. 1., 20, Blackwall S c i e n t i f i c Publications, London. 151 Huggins, M.L. (1941) Solutions of long-chain compounds. J. Chem. Phys. 9, 440. Hustedt, H. and Kula, M.-R. (1977) Studies on the i n t e r a c t i o n between aminoacyl-tRNA synthetase and transfer ribonucleic acid by eq u i l i b r i u m p a r t i t i o n . Eur. J. Biochem. 74, 191-198. Hustedt, H., Kroner, K.H. and Kula, M.-R. (1985) Applications of phase p a r t i t i o n i n g i n biotechnology. In " P a r t i t i o n i n g i n Aqueous Two Phase Systems" (H. Walter, D.E. Brooks and D. Fisher, eds.) 529-587, Academic press, New York. Igarashi, Y., Saito, Z. and Sato, N. (1974) P a r t i t i o n of milk proteins in water polymer two phase systems. Hirosaki Daigaku Nogakubu Gakujutsu. Hokoku 23, 60-65. Iwata, H., Amemiya, H., Matsuda, T., Takano, H., Hayashi, R. and Akutsu, T. (1989) Evaluation of microencapsulated i s l e t s i n agarose gel as b i o a r t i f i c i a l pancreas by studies of hormone-secretion in c u l t u r e and by xenotransplantation. Diabetes 38 (Suppl. 1), 224-225. Jehn, U. and Grunewald, R. (19SS) Post-remission treatment of acute leukemia i n adulthood: A l l o g e n i c bone marrow transplantation or chemotherapy? Klinische Wochenschrift 66, 614-623. Jennings, S.D. and Ham, R.G. (1983) Clonal growth of primary c u l t u r e s of rabbit ear chondrocytes i n a 1ipid-supplemented defined medium. Exp. Cell Res. 145, 415-423. Johansson, G. (1970a) P a r t i t i o n of s a l t s and their e f f e c t s on p a r t i t i o n of proteins i n a dextran-polyethylene g l y c o l system-water two-phase system. Biochem. Biophys. Acta 221, 3S7-390. Johansson, G. (1970b) Studies on aqueous dextran-polyethylene g l y c o l two-phase systems containing charged PEG. 1. P a r t i t i o n of Albumins. Biochem. Biophys. Acta 222, 381-389. Johansson, G. (1971) E f f e c t s of d i f f e r e n t ions on the p a r t i t i o n of proteins i n a dextran-polyethylene glycol-water two-phase system. In "Proceedings of the International Solvent Extraction Conference, 1971" (J.G. Gregory, B. Evans and P.C. Weston, eds.) Vol. 2, 928-935. Soc. Chem. Ind., London. Johansson, G. (1974) E f f e c t s of s a l t s on the p a r t i t i o n of proteins i n aqueous polymeric biphasic systems. Acta Chem. Scand., Ser. B B28, 873-SS2. Johansson, G. (1984) A f f i n i t y p a r t i t i o n i n g . Meth. Enzymol. 104, 356-364. Johansson, G. (1985) P a r t i t i o n i n g of proteins. In " P a r t i t i o n i n g i n Aqueous Two Phase Systems" (H. Walter, D.E. Brooks and D. Fisher, eds.) 161-226, Academic Press, New York. 152 Johansson, G. (1986) Time-dependant e f f e c t of the p a r t i t i o n i n g of ele c t r o p l a x membranes i n aqueous biphasic systems using counter-current d i s t r i b u t i o n technique and hexaethonium-polyethylene g l y c o l . J. Chromatog. 361, 131-140. Johansson, G. and Andersson, M. (19S4) Parameters determining a f f i n i t y p a r t i t i o n i n g of yeast enzymes using polymer-bound t r l a z i n e dye ligands. J. Chromatog. 303, 39-51. Johansson, G. and Joe l s s on, M. (19S5) Preparation of Cibacron blue F3G-A polyethylene g l y c o l i n large scale for use in a f f i n i t y p a r t i t i o n . Biotechnol. Bioeng. 27, 621-625. Kaighn, M.E., Kirk, D., Kigawa, S., Verner, G. and Narayan, S.K. (1983) R e p l i c a t i v e c u l t u r e s of normal human bladder e p i t h e l i a l c e l l s i n serum-free medium. In Vitro 19, 252-253. Kamentsky, L . A . , Melamed, M.R. and Derman, H. (1965) Spectrophotometer: New instrument f o r u l t r a rapid c e l l analysis. Science 156, 1364-1367. Karr, I J. , Harris, J.M., Shafer, S.G., Van A l s t i n e , J.M. and Snyder, R.S. (1986) Immuno-affinity p a r t i t i o n of c e l l s i n aqueous polymer two phase systems. J. Chromatog. 354, 269-282. Karr, L.J., Harris, J.M., Shafer, S.G., Van A l s t i n e , J.M. and Snyder, R.S. (1987) PEG-modified protein A for use in immunoaffinity phase p a r t i t i o n i n g . In "Advances i n separations using aqueous phase systems i n biology and biotechnology" (D. Fisher and I. Sutherland, eds.) Plenum Press, London. In press. Kasten, F.H. (1973) Mammalian myocardial c e l l s . In "Tissue Culture Methods and a p p l i c a t i o n s " (P.F. Kruse, Jr., and M.K. Patterson, J r . , eds. ) 72-81 Academic Press, New York. Kataoka, K., Sakurai, Y. , Maruyama, A. and Tsuruta, T. (1988). Separation of B- and T-lymphpcytes by c e l l u l a r adsorption chromatography using poly(2- hydroxyethyl methacrylate)/polyamine g r a f t copolymer as column adsorbent. J. Biomedical Materials Res. 22, 555-571. Kemshead, J.T. and Ugelstad, J. (1985) Magnetic separation techniques: t h e i r a p p l i c a t i o n to medicine. Molec. Cell. Biochem. 67, 11-18. Kessel, D. and McElhinney, R.S. (1978) E f f e c t of d i t h i o c a r b a n i l a t e s on some b i o l o g i c a l and biophysical properties of leukemia L1210 c e l l membranes. Mol. Pharmacol. 14, 1121-1129. Kogl, F. and Tonnis, B. (1936).Plant-growth substances. XX. The bios problem. I s o l a t i o n of c r y s t a l l i n e b i o t i n from egg yolk. Z. physiol. Chem. , 242, 43-73. Krohn, K.A., Knight, L.C., Harwig, J.F. and Welch, M.J. (1977) Differences i n the s i t e s of iodination of proteins f o l l o w i n g four methods of r a d i o i o d i n a t i o n . Biochim. Biophys. Acta 490, 497-506. 153 Kroner, K.H., Schutte, H., Stach, W. and Kula, M.-R. (1982) Scale-up of formate dehydrogenase by p a r t i t i o n . J. Chem. Technol. Biotechnol. 32, 130-137. Langone, J . J . (1982) Applications of immobilized p r o t e i n A in immunochemical techniques. J. Immunol. Neth. 55, 277-296. Lechner, J.F., Haugen, A., Autrup, H., McClendon, I.A., Trump, B.F. and Harris, C C . (1981) Clonal growth of e p i t h e l i a l c e l l s from normal adult human bronchus. Cancer Res. 41, 2294-2304. Lechner, J.F., Haugen, H., McClendon, I.A. and P e t t i s , E.W. (1982) Clonal growth of normal adult human bronchial e p i t h e l i a l c e l l s i n a serum-free medium. In Vitro 18, 633-642. L e f f e r t , H.L. and Paul, D. (1972) Studies on primary cultures of d i f f e r e n t i a t e d f e t a l l i v e r c e l l s . J. Cell Biol. 52, 559-568. Leung, J.P. and Edgington, T.S. (1980) Sub c e l l u l a r l o c a l i z a t i o n of the. sedimentable form of mammary tumor glycoprotein to the tumor c e l l plasmalemma. Cancer Res. 316-321. Levine, S., Levine, M., Sharp, K.A. and Brooks, D.E. (1983) Theory of the e l e c t r o k i n e t i c behavoir of human erythrocytes. Biophys. J. 42, 127-135. Levinsky, R.J. (1989) Recent advances i n bone marrow transplantation. Clin. Immunol. Immunopath. 50, 124-132. Levy, J.G., Logan, P.M., Whitney, S. , Malcolm, A.J., Naiman, S. , Powles, R. and Crawford, D. (1985) Demonstration of a leukemia-associated antigen (CAMAL) i n perip h e r a l blood leucocytes of bone marrow transplant patients p r i o r to relapse. Transplantation, 40, 167-173. Lindahl, P.E. (1948) P r i n c i p l e of a counter-streaming centrifuge for the separation of p a r t i c l e s of d i f f e r e n t s i z e s . Nature {London) 161, 6 4 S - 6 4 9 . Ling, T.G.I., Ramstorp, M. and Mattiasson, B. (1982) Immunological q u a n t i t a t i o n of b a c t e r i a l c e l l s using a p a r t i t i o n a f f i n i t y ligand assay: A model study on the quantitation of Streptococci B. Anal. Biochem. 122, 26-32. L i n k e r - I s r a e l i , M., B i l l i n g , R.J., Foon, K.A. and Terasaki, P.I. (1981) Monoclonal antibodies reactive with acute myelogenous leukemia c e l l s . J. Immunol. 127, 2473-2477. L i t t l e f i e l d , J. (1964) Sele c t i o n of hybrids from matings of f i b r o b l a s t s in vitro and t h e i r presumed recombinants. Science 145, 709-710. Malmstrom, P., Jonsson, X., Hallberg, T. and Sjogren, H.O. (1980a) Counter- current d i s t r i b u t i o n of lymphocytes from human peripheral blood i n an aqueous two-phase system. Separation into subsets of lymphocytes bearing d i s t i n c t i v e markers. Cell Immunol. 53, 39-50. Malmstrom, P.,Jonsson, X. and Sjogren, H.O. (1980b) Countercurrent 154 d i s t r i b u t i o n of lymphocytes from human peripheral blood i n an aqueous two-phase system. II. Separation into subsets of lymphocytes with d i s t i n c t i v e functions. Cell Immunol. 53, 51-64. Markwell, M.A.K. (1982) A new s o l i d - s t a t e reagent to iodinate proteins. Anal. Biochem. 125, 427-432. Mattiasson, B. and Ling, T.G.I. (19S0) P a r t i t i o n a f f i n i t y ligand assay (PALA). A new approach to binding assays. J. Immunol. Meth. 38, 217-223. Mazur, R., Kemp, J.A. and M i l l e r , R.H. (1976) Survival of f e t a l rat pancreases frozen to -78 and -196°C. Proc. Natl. Acad. Sci. USA. 73, 4105-4109. McEwen, C.R., S t a l l a r d , R.W. and Juhas, E. (1968) Separation of b i o l o g i c a l p a r t i c l e s by c e n t r i f u g a l e l u t r i a t i o n . Anal. Biochem. 23, 369-377. McGregor, J.L., Brochier, J., Wild, F. , Fo l l e n , G. , Trzeclak, M.C, James, E., Dechavanne, M., McGregor, L. and Clemetson, K.J. (1983) Monoclonal antibodies against p l a t e l e t membrane glycoproteins. C h a r a c t e r i z a t i o n and e f f e c t on p l a t e l e t function. Eur. J. Biochem. 131, 427-436. Mendenhall, W., Scheaffer, R.L. and Wackerly, D.D. (19S1) Mathematical S t a t i s t i c s with Applications, Wadsworth Inc., Belmont. Me i s t r i c h , M.L., Grdina, D.J., Meyn, R.E. and Barlogie, B. (1977) Separation of c e l l s from mouse s o l i d tumors by c e n t r i f u g a l e l u t r i a t i o n . Cancer Res. 37, 4291-4296. Me i s t r i c h , M.L., N e l l , L.J. and Richie, E.S. (1981) Separation of AKR mouse thymus lymphoma from normal thymic c e l l s by c e n t r i f u g a l e l u t r i a t i o n . J. Immunol. Meth. 41, 289-301. Mew, D. , Lum, B. , Wat, CK., Towers, G.H.N. , Sun, C.H.C, Walter, R.J.. Levy, J.G. and Berns, M.W. (1985) A b i l i t y of s p e c i f i c monoclonal antibodies and conventional a n t i s e r a conjugated to hematoporphyrin to lab e l and k i l l selected c e l l l i n e s subsequent to l i g h t a c t i v a t i o n . Cancer Res. 45, 4380-4386. M i l l e r , R.G. and P h i l l i p s , R.A. (1969) Separation of c e l l s by v e l o c i t y sedimentation at unit gravity. Nature (London) 214, 824-825. Miner, K.M., Walter, H. and Nicolson, G.L. (19S1) Subfractionation of malignant varients of metastatic murine lymphosarcoma c e l l s by countercurrent d i s t r i b u t i o n i n two-polymer aqueous phases. Biochem. 20, 6244-6250. Mino, G. and Kaizerman, S. (1958) A new method for the preparation of g r a f t copolymers. Polymerization i n i t i a t e d by eerie ion redox systems. J. Polym. Sci. 31, 242-243. Moldaven, A. (1934) A ph o t o e l e c t r i c technique for counting microscopical 155 c e l l s . Science, 80, 188. Molday, R.S., Yan, S.P.S. and Rembaum, A. (1977) Ap p l i c a t i o n of magnetic microspheres i n l a b e l l i n g and separation of c e l l s . Nature {London), 268. 437-438. Molday, R.S. and MacKenzie, D. (1982) Immunospecific ferromagnetic iron-dextran reagents for the l a b e l l i n g and magnetic separation of c e l l s . J. Immunol. Meth. 52, 353-367. Morecki, S. and Slavin, S. (1988) Combination of magnetic and chemocytotoxic cancer c e l l depletion f o r autologous bone marrow transplantation. Israel J. Med. Sci. 24, 488-493. Mosbach, K. and Schroder, U. (1979) Preparation and a p p l i c a t i o n of magnetic polymers for targeting drugs. FEBS Lett. 102, 112-116. Muino Blanco, M.T., Cebrian Perez, J.A., Olde, B. and Johansson, G. (1986) E f f e c t of dextran- and polyethylene glycol-bound Procion Yellow HE-3G on the p a r t i t i o n of membranes from c a l f brain synaptosomes within an aqueous two-phase system. J. Chromatog. 358, 147-158. Mullaney, P.F., Steinkamp, J.A., Crissman, H.A., Cram, L.S., Crowell, J.M., Salzman, G.C. and Martin, J.C. (1976) Laser flow microphotometry for rapid analysis and so r t i n g of mammalian c e l l s . Ann. N.Y. Acad. Sci. 267, 176-190. Muller, W. (1985) P a r t i t i o n i n g of nu c l e i c acids. In " P a r t i t i o n i n g in Aqueous Two Phase Systems" (H. Walter, D.E. Brooks and D. Fisher, eds.) 227-266, Academic Press, New York. Muller, W. (1986) New phase supports for l i q u i d - l i q u i d p a r t i t i o n chromatography of biopolymers i n aqueous poly(ethyleneglycol)-dextran systems. Eur. J. Biochem. 155, 213-222. Muller, W. and Gautier, F. (1975) Interaction of heteroaromatic compounds with n u c l e i c acids. A T - s p e c i f i c non-intercalating ligands. Eur. J. Biochem. 54, 385-394. Murakami, H., Masui, H., Sato, G.H., Sueoka, N., Chow, T.P. and Kano-Sueoka, T. (1982) Growth of hybridoma c e l l s i n serum-free medium -ethanolamine i s an e s s e n t i a l component. Proc. Natl. Acad. Sci. USA 79, 1158-1162. Nadler, L.M., Stashenko, P., Hardy, R., Kaplan, W.D., Button, L.N., Kufe. D.W., Antman, K.H. and Sehlossman, S.F. (19S0) Serotherapy of a patient with a monoclonal antibody d i r e c t e d against a human lymphoma-associated antigen. Cancer Res. 40, 3147-3154. Netzel, B., Haas, .R.J., Rodt, H., Kolb, H.J. and Th i e r f e l d e r , S. (1980) Immunological conditioning of bone marrow for autotransplantation i n childhood acute lymphoblastic leukaemia. Lancet i , 1330-1332. Niepokojczycka, E., Budzicka, E. and Szumiel, I. (1982) The e f f e c t of 156 combined treatment of hypoxic L5178Y-R and L5178Y-S c e l l s with l i d o c a i n e and X-rays. Neoplasma 29, 605-611. Oder, R.R. (1976) High gradient magnetic separation theory and appl i c a t i o n s . IEEE Trans. Magn. 12, 42S-435. Ohlsson, R., Hentschel, C.C. and Williams, J.G. (1978a) A rapid method fo r the i s o l a t i o n of c i r c u l a r DNA using an aqueous two-phase p a r t i t i o n system. Nucleic Acids Res. 5, 583-590. Ohlsson, R., J e r g i l , B. and Walter, H. (1978b) F r a c t i o n a t i o n of microsomal membranes on the basis of the i r surface properties. Biochem. J. 172, 189-192. Okazaki, T. and Kornberg, A. (1964) Enzymatic synthesis of deoxyribonucleic acid. XV. P u r i f i c a t i o n and properties of a polymerase from Bacillus Subtilis. J. Biol. Chem. 239, 259-268. Ornstein, L . (1964) Disc electrophoresis. I. Background and Theory. Ann. N.Y. Acad. Sci. 121, 321349. Owen, C S . (1982) High gradient capture of red blood c e l l s ; J. Appl. Physics, 53, 3884-3887. Owen, C S . and L i b e r t i , P.A. (19S7) Magnetite-protein conjugates for the separation of c e l l s by high gradient magnetic f i l t r a t i o n . In " C e l l Separation: Methods and Selected Applications vol. 4" (T. G. Pretlow and T.P. Pretlow, eds.) 259-275, Academic press, New York. Owens, R.B., Smith, H.S., Nelson-Rees, W.A. and Springer, E.L. (1976) E p i t h e l i a l c e l l cultures from normal and cancerous human ti s s u e s . J . Natl. Cancer Inst. 56, 843-849. Pangburn, M.K. and Walter, H. (1987) S p e c i f i c deposition of complement protein C3b on abnormal PNH erythrocytes permits t h e i r separation by p a r t i t i o n i n g . Possible general approach for i s o l a t i o n of s p e c i f i c c e l l populations. Biochim. Biophys. Acta 902, 278-286. Patterson, J.B. and Sta f f o r d , D.W. (1970) Sea urchin s a t e l l i t e deoxyribonucleic acid. It's large scale i s o l a t i o n and h y b r i d i z a t i o n with homologous ribosomal r i b o n u c l e i c acid. Biochem. 9, 1278-1283. Paul, F., Roath, S. and M e l v i l l e , D. (1978) D i f f e r e n t i a l blood separation using a high gradient magnetic f i e l d . Br. J. Haematol. 38, 273-280. Paul, F., Roath, S. and M e l v i l l e , D. (1981) Separation of malarial infected erythrocytes from whole blood: use of s e l e c t i v e high gradient magnetic separation technique. Lancet ii, 70. Peterson, E.A. and Evans, W.H. (1967) Separation of bone marrow c e l l s by sedimentation at unit gravity. Nature (London), 214, 824-825. Peterson, P., Reimann, B. and Hammermeister, M. (1986) S p e c i f i c l e c t i n binding on pancreatic tissue and i t ' s a p p l i c a b i l i t y i n the i s o l a t i o n of 157 i s l e t c e l l s . Z. Exp. Chir. Trans. Kunst. 19, 262-266. Pimm, M.V., Embleton, M . J . , Perkins, A.C, Price, M.R., Robins, R.A., Robinson, G.R. and Baldwin, R.W. (1982) In vivo l o c a t i o n of a n t i -osteogenic sarcoma 791T monoclonal antibody i n osteogenic sarcoma xenografts. Int. J. Cancer 30, 75-S5. Platsoucas, CD. (1983) Separation of c e l l s by preparative density gradient electrophoresis. In " C e l l Separation: Methods and Selected Applications vol 2" (T.G. Pretlow and T.P. Pretlow, eds) 145-178, Academic press, New York. Pohl, H.A. (1977) Dielectrophoresis: a p p l i c a t i o n s to the c h a r a c t e r i z a t i o n and separation of c e l l s . In "Methods of C e l l Separation vol 1." (N. Catsimpoolas, ed) 67-165, Plenum press, New York. Porter, R.R. (1959) The hydrolysis of rabbit y - g l o b u l i n and antibodies with c r y s t a l l i n e papain. Biochem. J. 73, 119-126. Pr e f f e r , F.I. and Colvin, R.B. (1987) Analysis and s o r t i n g by flow cytometry: a p p l i c a t i o n s to the study of human disease.In " C e l l Separation: Methods and Selected Applications vol 5."(T.G. Pretlow and T.P. Pretlow, eds) 311-347, Academic press, New York. Pretlow, T.G. (1971) Estimation of experimental conditions that permit c e l l separations by v e l o c i t y sedimentation on i s o k i n e t i c gradients of f i c o l l i n tissue culture medium. Anal. Biochem. 41, 24S-255. Pretlow, T.G., Wier, E.E. and Zettergren, J.G. (1975) Problems connected with the separation of d i f f e r e n t kinds of c e l l s . Int. Rev. Exp. Pathol. 14, 91-204. Pretlow, T.G. and Pretlow, T.P. (1982) Sedimentation of c e l l s : an overview and discussion of a r t i f a c t s . In " C e l l Separation: Methods and Selected Applications v o l . 1" (T.G. Pretlow and T.P. Pretlow, eds.) 41-57, Academic press, New York. Price , M.R., Pimm, M.V., Dawood, F.A., Embleton, M.J. and Baldwin, R.W. (1982) C h a r a c t e r i s t i c s and tumor l o c a l i z i n g properties of anti-human osteogenic monoclonal antibodies. In "Tumor Progression and Markers" (K. Lapis and A. Jeney, eds.) Kugler Press, Amsterdam. Raymond, F.D. and Fisher, D. (1980) E f f e c t of polyethylene g l y c o l -palmitate on the p a r t i t i o n of c e l l s in aqueous two-phase polymer systems. Biochem. Soc. Trans. 8, 118-119. Raymond, F.D. and Fisher, D. (19S1) C e l l surface properties and c e l l p a r t i t i o n i n two polymer aqueous phase systems. In " C e l l E l e c t r o p h o r e s i s in Cancer and Other C l i n i c a l Research" (A.W. Preece and P.A. Light, eds.) 65-68. E l s e v i e r / North Holland Biomedical Press, Amsterdam. Reinherz, E.L., Geha, R. , Rappeport, J.M., Wilson, M. , Penta, A.C, Hussey, R.E., F i t z g e r a l d , K.A., Daley, J.F., Levine, H., Rosen, F.S. and Schlossman, S.F. (1982) Reconstitution a f t e r transplantation with 158 T-lymphocyte depleted HLA haplotype-mismatched bone marrow f o r severe combined immunodeficiency. Proc. Natl. Acad. Sci. U.S.A., 79, 6047-6051. Reitherman, R., Flanagan, S.D. and Barondes, S.H. (1973) Electromotive phenomena i n p a r t i t i o n of erythrocytes i n aqueous polymer two-phase systems. Biochem. Biophys. Acta 297, 193-202. Rheinwald, J.G. (19S0) S e r i a l c u l t i v a t i o n of normal human epidermal ' keratinocytes. Meth. Cell Biol. 21A, 229-254. R i c o r d i , C., Lacy, P.E. and Scharp, D.W. (1989) Automated i s l e t i s o l a t i o n from human pancreas. Diabetes 38 (Suppl. 1), 140-142. R i t z , J. and Schlossman, S.F. (19S2) U t i l i z a t i o n of monoclonal antibodies i n the treatment of leukemia and lyphoma. Blood, 59, 1, 1-11. Ri t z , J ., S a l l a n , S.E., Bast, R.C., Lipton, J.M., C l a v e l l , L.A., Feeney, M., Hercend, T., Nathan, D.G. and Schlossman, S.F. (1982) Autologous bone marrow tr a n s p l a n t a t i o n i n CALLA-posltive acute lymphoblastic leukemia a f t e r i n vitro treatment with J5 monoclonal antibody and complement. Lancet i i , 60. Robinson, A., Williams, A.R.W., P i r i s , J., Spandidos, D.A. and Wyllie, A.H. (1986) Evaluation of a monoclonal antibody to ras peptide, RAP-5, claimed to bind p r e f e r e n t i a l l y to c e l l s of i n f i l t r a t i n g carcinomas. Br. J. Cancer. 54, 877-883. Sandler, S., Andersson, A., Schnell, A., Mellgren, A.,•Tollemar, J., Borg, H. , Petersson, B. , Groth, CG. and Hellerstrom, C. (1985) Tissue culture of human f e t a l pancreas. Development and function of B - c e l l s in vitro and trans p l a n t a t i o n of explants to nude mice. Diabetes 34, 1113-1119. Sandler, S., Andersson, A., Korsgren, 0., Tollemar, J., Petersson, B., Groth, CG. and Hellerstrom, C. (1989) Tissue culture of human f e t a l pancreas. E f f e c t s of nicotinamide on i n s u l i n production and formation of i s l e t l i k e structures. Diabetes, 38 (Suppl. 1), 168-171. Sato, G.H. (1975) The r o l e of serum in c e l l culture. In "Biochemical Actions of Hormones vol 3." (G. Litwack, ed.) 391-1396, Academic Press, New York. Scharp, D.W. (1984) I s o l a t i o n and transplantation of i s l e t t i s s u e . World J. Surg. 8, 143-151. Schlossman, S.F. and Hudson, L. (1973) S p e c i f i c p u r i f i c a t i o n of lymphocyte populations on a digestable immunoabsorbent. J. Immunol. 110, 313-319. Schmid-Schonbein, G.W., Shih.Y.Y. and Chien, S. (1980) Morphometry of human leukocytes. Blood 56, 866-875. Schroder, U. and Mosbach, K. (1983) Magnetic p a r t i c l e s f or i n t r a v a s c u l a r administration: comprising magnetic material i n carbohydrate matrix. 159 Patent Cooperation Treaty Int. Appl. W0S3 91738. Schroder, U. , Segren, S., Gemmefors, C. , Hedlund, G., Jansson, B. . Sjogren, H.O. and Borrebaeck, C.A.K. (1986) Magnetic carbohydrate nanoparticles f or a f f i n i t y c e l l separation. J. Immunol. Meth. 93, 45-53. Seaman, G.V.F. (1975) E l e c t r o k i n e t i c behavoir of red c e l l s . In "The Red Blood C e l l , 2nd ed., v o l . II" (D.M. Surgenor, ed.) 1188-1195, Academic Press, New York. Seaman, G.V.F., Tamblyn, C.H., Colvin, N. and Walter, H. (1980) Physicochemical pr o p e r t i e s of erythrocytes from rats of d i f f e r e n t ages. Fed. Proc. Fed. Am. Soc. Exp. Biol. 39, 1917. Seglen, P.O. (1973) Preparation of rat l i v e r c e l l s . III. Enzymatic requirements for t i s s u e dispersion. Exp. Cell Res. 82, 391-398. Sharkis, J . J . , Santos, G.W. and Colvin, M. (1980) E l i m i n a t i o n of acute myelogenous leukemic c e l l s from marrow and tumor suspensions i n the rat with 4-hydroperoxycyclophosphamide. Blood, 55, 3, 521-523. Sharon, N. (1984) Use of l e c t i n s for separation of c e l l s . In " C e l l Separation: Methods and Selected Applications v o l . 4." (T.G. Pretlow and T.P. Pretlow, eds) 13-52, Academic press, New York. Sharp, K.A. (1985) Theoretical and experimental studies on erythrocyte p a r t i t i o n i n aqueous polymer two-phase systems. Ph.D. Thesis. U n i v e r s i t y of B r i t i s h Columbia. Sharp, K.A. and Brooks, D.E. (1985) Ca l c u l a t i o n of the el e c t r o p h o r e t i c m o b i l i t y of a p a r t i c l e bearing bound p o l y e l e c t r o l y t e using the nonlinear Poisson-Boltzmann equation. Biophys. J. 47, 563-566. Sharp, K.A. and Brooks, D.E. (1989) C e l l a f f i n i t y p a r t i t i o n i n aqueous polymer two phase systems. A the o r e t i c a l model and i t s a p p l i c a t i o n to erythrocyte p a r t i t i o n with PEG 8000-palmitate. In Press. Sharp, K.A., Yalpani, M., Howard, S.J. and Brooks, D.E. (1986) Synthesis and a p p l i c a t i o n of a poly(ethylene glycol)-antibody a f f i n i t y ligand f o r c e l l separations i n aqueous polymer two phase systems. Anal. Biochem. 154, 110-117. Sherbet, G.V. and Lakshmi, M.S. (1981) Membrane surface properties of normal and malignant c e l l s : P a r t i t i o n i n aqueous two-polymer phase systems. Exp. Cell Biol. 49, 267-271. Shipman, R., Malcom, A. and Levy, J.G. (19S3) P a r t i a l c h a r a c t e r i z a t i o n of a membrane antigen which exhibits s p e c i f i c i t y f o r c e l l s of patients with acute myelogenous leukemia. Br. J. Cancer 47, 849-852. Soon-Shiong, P., Heintz, R. and Terasaki, P. (1988) Ch a r a c t e r i z a t i o n of monoclonal antibodies CBL3 and TT62 with d i f f e r i n g antigenic determinants s p e c i f i c f o r i s o l a t e d human i s l e t s . Biochem. Biophys. Res. Commun. 150, 7-75-780. 160 Sotobayashi, T., Suzuki, T. and Kudo, H. (1977) L i q u i d - l i q u i d e x t r a c t i o n of some a c t i n i d e s with polyethylene g l y c o l and i t ' s d e r i v a t i v e s . J. Radioanal. Chem. 36, 145-152. Springer, T.A. Hybridoma Technology i n the Biosciences and Medicine, Plenum Press, New York (1985). Steinkamp, J.A., Romero, A. and Van D i l l a , M.A. (1974) Multiparameter a n a l y s i s and sor t i n g of mammalian c e l l s . Exp. Cell Res. 84, 15-23. Stendahl, 0., Dahlgren, C. and Hed, J. (1982) Physicochemical surface changes on phagocytic c e l l s during d i f f e r e n t i a t i o n i n r e l a t i o n to chemotaxis and phagocytosis. In "Biochemistry and Function of Phagocytes" (F. Rossi and P. Pa t r i a r c a , eds.) 531-537, Plenum Press, New York. Stocks, S.J., Jones, A.J.M., Ramey, C.W. and Brooks, D.E. (1986). A fluo r o m e t r i c assay of the degree of modification of pro t e i n primary amines with polyethylene g l y c o l . Anal. Biochem. 154, 232-234. Stocks, S.J. and Brooks, D.E. (1987) Second immunoaffinity ligands for c e l l separation. In "Advances i n separations using aqueous phase systems i n biology and biotechnology" (D. Fisher and I. Sutherland, eds.) Plenum press, London. In press. Stocks, S.J. and Brooks, D.E. (1988) Development of a general ligand for immunoaffinity p a r t i t i o n i n g i n two phase aqueous polymer systems. Anal. Biochem. 173, 86-92. Stocks, S.J. and Brooks, D.E. (1989) Production and i s o l a t i o n of large q u a n t i t i e s of monoclonal antibody using serum-free medium and fa s t p r o t e i n l i q u i d chromatography. Hybridoma, 8 (2), 241-247. Suga, T., Horie, S., I s h i i , H. and O r i i , H. (1979) Studies on peroxisomes by aqueous polymer two-phase systems. In " S e i t a i Seibun no Bunseki Kagaku Shinpojumu, 4th" (K. Yoshishu, ed.) 37-40, Nippon Yakugakki, Tokoyo. Takei, F. (1983) Two surface antigens expressed on p r o l i f e r a t i n g mouse T lymphocytes defined by rat monoclonal antibodies. J. Immunol. 130, 2794-2797. Takei, F. (1986) MBL-2(4.1) c e l l s have a greater surface density of an antigen recognized by YEl.48.10. than MBL-2(2.6) c e l l s . (Personal communication) Takerkart, G. , Segard, E. and Monsigny, M. (1974) P a r t i t i o n of tryp s i n in two phase systems containing a diamino-cx,w-diphenylcarbamyl polyethylene g l y c o l as competitive i n h i b i t o r of tryp s i n . FEBS Lett. 42, 218-220. T-hierfelder, S. , Rodt, H. and Netzel, B. (1977) Transplantation of syngeneic bone marrow incubated with leucocyte antibodies. I. Supression of lymphatic leukemia of syngeneic donor mice. Transplantation 23, 459-463. 161 T j e r n e l d , F., Persson, I. and Albertsson, P-X. (1985) Enzymic h y d r o l y s i s o f c e l l u l o s e i n aqueous two phase systems. I. P a r t i t i o n of c e l l u l a s e s from Trichoderma reesei. Biotechnol. Bioeng. , 27(7), 1036-1043. Tlaskalova-Hogenova, H., Coupek, J., P o s p i s i l , M., Tuckova, L., Kaminkova, J . and Mancal, P. (1981) A f f i n i t y chromatography of human lymphocytes on Sepheron-immunoadsorbent columns. J. Polym. Sci. Polytn. Sympc 68, 89-95. T r e f f r y , T.E., Sharpe, P.T., Walter, H. and Brooks, D.E. (1985) Thin-layer countercurrent d i s t r i b u t i o n and apparatus. In " P a r t i t i o n i n g i n Aqueous Two-Phase Systems" (H. Walter, D.E. Brooks and D. Fisher, eds.) 131-148, Academic Press, New York. Treleaven, J.G., Ugelstad, J . , P h i l i p , T., Gibson, F.M., Rembaum, A. and Caine, G.D. (1984) Removal of neuroblastoma c e l l s from bone marrow with monoclonal antibodies conjugated to magnetic microspheres. Lancet 8368, 70-73. Turner, G. and Hancock, R. (1974) F r a c t i o n a t i o n of chromosomal deoxyribo-nucleoproteins by p a r t i t i o n i n two-phase aqueous polymer systems. Biochem. Biophys. Res. Commun. 58, 437-445. Tutschka, P.F., Elfenbein, G.J., Sensenbrenner, L.L., Saral, R., Order, S.E., Kaizer, H., Beschorner, W.B., Farmer, E. and Santos, G.W. (1980) Preparative regimens for marrow trans p l a n t a t i o n i n acute leukemia and a p l a s t i c anemia. Am. J. Ped. Hematol. Oncol. 2, 363-370. Ugelstad, J., Soderberg, L., Berge, A. and Bergstrom, J. (1983) Monodisperse polymer p a r t i c l e s - a step forward for chromatography. Nature {London) 303, 96-98. Ugelstad, J., Rembaum, A., Kemshead, J.T., Nustad, K., Funderud, S. and Schmid, R. (1984) In "Microspheres and Drug Therapy. Pharmaceutical, Immunological and Medical Aspects" (S.S. Davis, L. Ilium, J.G. McVie and E. Tomlinson, eds) 365-382, E l s e v i e r , Amsterdam. Utsumi, S. (1969) Stepwise cleavage of rabbit immunoglobulin G by papain and i s o l a t i o n of four types of b i o l o g i c a l l y a c t i v e F fragments. Biochem. J. 112, 343-355. C Van A l s t i n e , J.M. and Brooks, D.E. (1984) C e l l membrane abnormality detected i n erythrocytes from patients with multiple s c l e r o s i s by p a r t i t i o n i n two- polymer aqueous-phase systems. Clin. Chem. (Winston-Salem, N.C.) 30, 441-443. Van A l s t i n e , J.M., Trust, T.J., Webber, T.J., Greig, R., Poste, G.T. and Brooks, D.E. (1986) Heterogeneity i n the surface properties of B16 melanoma c e l l s from sublines with d i f f e r i n g metastatic p o t e n t i a l detected via -two polymer aqueous phase p a r t i t i o n . Expt. Cell Res. 164, 366-374. Walter, H. (1969) In "Red C e l l Membrane: Structure and Function" (G.A. Jamieson and T.J. Greenwalt, eds.) 368-370, Lipp i n c o t t , Philadelphia. 162 Walter, H. (19S5) P a r t i t i o n i n g of red c e l l s . ' I n " P a r t i t i o n i n g i n Aqueous Two Phase Systems" (H. Walter, D.E. Brooks and D. Fisher, eds.) 327-376, Academic Press, New York. Walter, H. and Coyle, R.P. (1968) E f f e c t of membrane mod i f i c a t i o n of human erythrocytes by enzyme treatment on their p a r t i t i o n i n aqueous dextran-polyethylene g l y c o l two-phase systems. Biochim. Biophys. Acta 165, 540-543. Walter, H. and Krob, E.J. (1975) Alterations i n membrane surface properties during c e l l d i f f e r e n t i a t i o n as measured by p a r t i t i o n i n aqueous two-polymer phase systems. Rat i n t e s t i n a l e p i t h e l i a l c e l l s . Exp. Cell Res. 91, 6-14. Walter, H. and Krob, E.J. (1976a) Hydrophobic a f f i n i t y p a r t i t i o n i n aqueous two-phase systems containing polyethylene g l y c o l - p a l m i t a t e of rightside-out and inside-out v e s i c l e s from human erythrocyte membranes. FEBS Lett. 61, 290-293. Walter, H. and Krob, E.J. (1976b) Membrane surface properties other than charge involved i n c e l l separation by p a r t i t i o n in polymer, aqueous two-phase systems. Biochem. 15, 2959-2964. Walter, H. and Krob, E.J. (1976c) P a r t i t i o n in two-polymer aqueous phases r e f l e c t s d i f f e r e n c e s between membrane surface properties of erythrocytes, ghosts and membrane v e s i c l e s . Biochim. Biophys. Acta 455, S-23. Walter, H., Selby, F.W. and Garza, R. (1967) On the countercurrent d i s t r i b u t i o n of red blood c e l l s . An addendum. Biochim. Biophys. Acta 136. 148-150. Walter, H., Krob, E.J. and Ascher, G.S. (1969) Separation of lymphocytes and polymorphonuclear leukocytes by countercurrent d i s t r i b u t i o n i n aqueous two-polymer phase systems. Exp. Cell Res. 55, 279-283. Walter, H., M i l l e r , A., Krob, E.J. and Ascher, G.S. (1972) Membrane surface properties of r e t i c u l o c y t e s rendered severely anemic with phenylhydrazine as determined by p a r t i t i o n i n aqueous phase systems. Exp. Cell Res. 73, 145-153. Walter, H., Krob, E.J., Ascher, G.S. and Seaman, G.V.F. (1973a) P a r t i t i o n of rat l i v e r c e l l s i n aqueous dextran-polyethylene g l y c o l phase systems. Exp. Cell Res. 82, 15-26. Walter, H., K rob , E.J. and Ascher, G.S. 1973b) Surface pr o p e r t i e s of rat bone marrow c e l l s as determined by p a r t i t i o n i n two-polymer aqueous phase systems. C e l l s containing hemoglobin. Exp. Cell Res. 79, 63-72. Walter, H., Tung, R., Krob, E.J. and Swingle, K.F. (1974) Membrane surface properties of red blood c e l l s from x - i r r a d i a t e d r a t s as measured by p a r t i t i o n i n two-polymer aqueous phase systems. Radiat. Res. 59, 614-628. Walter, H., Krob, E.J., Webber, T.J., Ascher, G.S. and Morin, R.J. 163 (1979a) E f f e c t of membrane cholesterol enrichment or de p l e t i o n on the p a r t i t i o n behavior of human erythrocytes i n dextran-polyethylene g l y c o l aqueous phases. Biochim. Biophys. Acta 550, 138-144. Walter, H., Webber, T.J., Michalski, J.P., McCombs, C.C., Moncla, B.J.,Krob, E.J. and Graham, L.L. (1979b) Subfractionation of human peripheral blood lymphocytes on the basis of the i r surface p r o p e r t i e s by p a r t i t i o n i n g i n two-polymer aqueous phase systems. J. Immunol. 123, 1687-1695. Walter, H., Graham, L.L., Krob, E.J. and H i l l , M. (1980) C o r r e l a t i o n between phagocytic and membrane surface properties r e f l e c t e d by p a r t i t i o n i n g of human peripheral blood monocytes i n two-polymer aqueous phases. Biochim. Biophys. Acta 602, 309-322. Walter, H., Krob, E.J. and Ascher, G.S. (1981) Aging of erythrocytes r e s u l t s i n a l t e r e d red c e l l surface properties i n the rat, but not i n the human. Studies by p a r t i t i o n i n g i n two-polymer aqueous phase systems. Biochim. Biophys. Acta 641, 202-215. Weiser, E., Yeh, C.-K., Lin, W. and Mazur, A. (1976) Non-uniform biosynthesis of multiple haemoglobins i n the adult rat and guinea pig- J. Biol. Chem. 251, 5703-5710. Weiser, M.M. (1973) I n t e s t i n a l e p i t h e l i a l c e l l surface membrane glycoprotein synthesis. I. An indicator of c e l l u l a r d i f f e r e n t i a t i o n . J. Biol. Chem. 248, 2536-2541. Wells, J.R. and James, T.W. (1972) C e l l cycle analysis by cu l t u r e f r a c t i o n a t i o n . Exp. Cell Res. 75, 465-474. Wells, J.R., Opelz, G. and Cline, M.J. (1977a) C h a r a c t e r i z a t i o n of f u n c t i o n a l l y d i s t i n c t lymphoid and myeloid c e l l s from human blood and bone marrow I. Separation by a bouyant density gradient technique. J. Immunol. Meth. 18, 63-77. Wells, J.R., Opelz, G. and Cline, M.J. (1977b) Ch a r a c t e r i z a t i o n of f u n c t i o n a l l y d i s t i n c t lymphoid and myeloid c e l l s from human blood and bone marrow II. Separation by v e l o c i t y sedimentation. J. Immunol. Meth. 18, 79-93. Wells, J.R., B i l l i n g , R., Herzog, P., Feig, S.A., Gale, R.P., Terasaki, P. and Cline, M.J. (1979) Autotransplantation a f t e r in vitro immunotherapy of lymphoblastic leukemia. Exp. Hematol. 7 (suppl. 5), 164-169. Widder, K.J., Senyei, A.E., Ovadia, H. and Paterson, P.Y. (1981) S p e c i f i c c e l l binding using staphylococcal protein A magnetic microspheres. J. Pharm. Sci. 70, 387-389. Wigzell, H. and Andersson, B. (1969) C e l l separation on antigen-coated columns. El i m i n a t i o n of high rate antibody-forming c e l l s and immunological memory c e l l s . J. Exp. Med. 129, 23-36. 164 Williams, S.F., Bitran, J.D., Hoffman, P.C, Robin, E. , Fullem, L. , Beschorner, J., Golick, J. and Golomb, H.M. (1989) High-dose, mu l t i p l e -a l k y l a t o r chemotherapy with autologous bone marrow r e i n f u s i o n in patients with advanced non-small c e l l lung cancer. Cancer 63, 238-242. Wysocki, L.J. and Sato, V.L. (1978) "Panning" f o r lymphocytes: a method for c e l l s e l e c t i o n . Proc. Natl. Acad. Sci., USA, 75, 2844-2848. Zwann, F.E., Jansen, J., Colpin, G.G.D. and Simonis, R.F.A. (1982) G r a f t i n g f o r acute-leukemia during remission: r e s u l t s i n 23 patients. Exp. Hematol.,10 (suppl. 10), 87. 165 Stocks, S.J. and Brooks, D.E. (1989) Immunoaffinity p a r t i t i o n of erythrocytes and lymphocytes using polyacrylamide-derivatized antibodies. J. Immunol. Submitted for publication. Stocks, S.J. and Brooks, D.E. (1989) Production and i s o l a t i o n of large qu a n t i t i e s of monoclonal antibody using serum-free medium and f a s t p r o t e i n l i q u i d chromatography. Hybridoma, 8 (2), 241-247. Stocks, S.J. and Brooks, D.E. (1988) Development of a general ligand f o r immunoaffinity p a r t i t i o n i n g i n two phase aqueous polymer systems. Anal. Biochem. 173, 86-92. Brooks, D.E., Sharp, K.A.S. and Stocks, S.J. (1988) Immunoaffinity separations of c e l l s i n two polymer aqueous phase systems. Nakromol. Chem. Makromol. Symp. 17, 387-399. Stocks, S.J. and Brooks, D.E. (1987) Second immunoaffinity ligands f o r c e l l separation. In "Advances i n separations using aqueous phase systems i n biology and biotechnology" (D. Fisher and I. Sutherland, eds.) Plenum press, London. Stocks, S.J., Jones, A.J.M., Ramey, C.W. and Brooks, D.E. (1986). A fluorometric assay of the degree of modification of protein primary amines with polyethylene g l y c o l . Anal. Biochem. 154, 232-234. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0059445/manifest

Comment

Related Items