Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Enzymatic harvesting of glycosyl phosphatidylinositol anchored recombinant proteins from mammalian cells Sunderji, Rumina 1994

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

Item Metadata

Download

Media
831-ubc_1994-0149.pdf [ 2.12MB ]
Metadata
JSON: 831-1.0058606.json
JSON-LD: 831-1.0058606-ld.json
RDF/XML (Pretty): 831-1.0058606-rdf.xml
RDF/JSON: 831-1.0058606-rdf.json
Turtle: 831-1.0058606-turtle.txt
N-Triples: 831-1.0058606-rdf-ntriples.txt
Original Record: 831-1.0058606-source.json
Full Text
831-1.0058606-fulltext.txt
Citation
831-1.0058606.ris

Full Text

ENZYMATIC HARVESTING OF GLYCOSYL PHOSPHATIDYLINOSITOL ANCHORED LS RECOMBINANT PROTEINS FROM MAMMALIAN CEL by  RUMINA SUNDERJI B.Sc. (Eng), University of Dar Es Salaam, 1986  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUEST FOR THE DEGREE OF MASTER IN APPLIED SCIENCE in THE FACULTY OF GRADUATE STUDIES Department of Chemical Engineering We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA February 1994 © Rumina Sunderji, 1994  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.  (Signature)  Department of  CcL  The University of British Columbia Vancouver, Canada Date  DE-6 (2/88)  9  Cl  11  Abstract Controlled release of recombinant proteins from mammalian cells enables protein product harvesting at increased concentrations and purity by separating protein expression from the protein recovery.  The chinese hamster ovary (CHO) cell line investigated was  genetically engineered to express glycosyl phosphatidylinositol (GPI) anchored human melanoma antigen p97 on the outer surface of the cell membrane. At intervals the cells were treated with a phosphatidylinositol phospholipase C (PT-PLC) harvest solution to selectively cleave the GPI anchor and recover the protein at high concentration and purity (cyclic harvesting). The growth of the recombinant CHO cells was investigated in the serum-free media  CHO-S-SFM I,  CHO-S-SFM II,  HBCHO,  DMEMJF12  and  Ham’s  F12.  CHO-S-SFM II supplemented with DNase and CHO-S-SFM I achieved single suspension 6 cells/mL). CHO-S-SFM I was selected for all cells at high densities (approximately 6x10 further investigations. A repeated harvesting technique which involved re-using the PT-PLC enzyme solution to harvest separate batches of cells was investigated.  This approach further  increased the concentration of the desired protein product after each harvest. Preliminary repeated harvesting experiments recovered 140 jig/mL p97 from 7 harvests of 1 O cells each. The first harvest recovered approximately 60 jig/mL p97, therefore theoretically 7 harvests should have recovered 420 j..tg/mL.  Since this process did not achieve the  expected concentrations, the stability of p97 and PT-PLC were investigated.  However,  p97 was found to be stable at 37 °C in the harvesting medium (PBS with I mg/mL BSA) for 24 h. Since a suitable assay for P1-PLC was not available, a new PT-PLC assay was developed. PT-PLC was stable at 37 °C for a period of 14 days. PT-PLC and p97 were also stable in the pH range of 6.0 7.9. -  Instability of the proteins involved in the  111  harvesting process was not the cause for the low product concentrations recovered from repeated harvesting. Loss of P1-PLC due to adsorption to the cells was studied. An equilibrium was established between P1-PLC in solution and P1-PLC adsorbed to the cell surface within 3 minutes. Since PT-PLC was adsorbed to each batch of cells in the repeated harvesting process, enzyme was removed with the cells after each harvest. Therefore loss of PT-PLC by adsorption was considered the cause for the reduced protein recovery. The repeated harvesting process was repeated with PT-PLC replenishment after each harvest.  Addition of 30 or 300 mU/mL PT-PLC recovered respectively 294 or  343 ig/mL p97 from 5 consecutive harvests. The estimated purity of p9’7, based on total protein, was approximately 30 %. A continuous harvesting process was also investigated. This approach involved addition of PT-PLC to the growth medium resulting in the continuous release of p97 into the medium. The continuous harvesting process was carried out simultaneously with 0, 3 and 30 mU/niL PT-PLC.  The 0 mU/niL PT-PLC control produced approximately  3.8 ig/mL p97 in a batch culture of 11 days. During the same period of time cultures with 3 and 30 mU/niL P1-PLC yielded 11.6 and 15.3 ig/mL p97 respectively. For the continuous harvesting process at 30 mU/niL PT-PLC the p97 productivity was 5. 75x 1 0 fig/cell-day or approximately 2-fold higher than that achieved by the cyclic and repeated harvesting processes.  However, the repeated harvesting process used  approximately 10 times less PT-PLC and recovered 20-fold higher p97 concentrations.  iv  Table of Contents Abstract  ii  Table of Contents  iv  List of Figures List of Tables  ix  Acknowlegdements CHAPTER 1 Introduction CHAPTER 2 Literature Review 2.1 Mammalian recombinant protein production 2.1.1 Conventional protein production processes 2.1.2 Regulated secretion 2.1.3 Enzymatic harvesting of membrane bound proteins 2.2 Application of model harvesting system 2.3 Melanotransferrin 2.4 Glycosyl phosphatidylinositol anchors 2.4.1 Structure of GPI anchor 2.4.2 Attachment of protein to GPI anchor 2.4.3 Functions of OPT anchors 2.4.4 Functions of OPT-anchored proteins 2.5 Phosphatidylinositol phospholipase C 2.5.1 Properties of P1-PLC from B. thuringiensis 2.5.2 P1-PLC assays 2.6 Protein adsorption  I  4 4 5 6 7 9 9 11 12 14 14 16 19 19 21 23  CHAPTER 3 Materials and Methods 3.1 3.2 3.3 3.4  Cell line Tissue culture P1-PLC production Production of monoclonal antibody against p97  24 24 24 25 26  V  3.5 Analytical methods 3.5.1 Cell count 3.5.2 Flow cytometry 3.5.3 Immunofluorescence assay 3.4.4 Glucose analysis 3.4.5 P1-PLC assay 3.4.5.1 P1-PLC assay based on flow cytometry 3.5.4.2 P1-PLC assay based on immunofluorescence 3.5 Experiments 3.5.1 Growth profiles for CHO cells 3.5.2 Repeated harvesting 3.5.3 p97 stability 3.5.4 P1-PLC stability 3.5.5 Adsorption of PT-PLC on cells 3.5.6 Continuous harvesting  .  26 26 27 27 29 31 31 33 33 33 34 34 36 36 37  CHAPTER 4 Results and Discussions 4.1 Growth media for CHO cells 4.2 P1-PLC assay 4.2.1 Effect of harvesting conditions on p97 removal 4.2.1.1 Incubationtime 4.2.1.2 Enzyme volume 4.2.2 P1-PLC assay based on flow cytometry 4.2.3 P1-PLC assay based on immunofluorescence 4.2.4 Reliability of P1-PLC assay 4.3 Repeated harvesting 4.3.1 p 97 stability 4.3.2 P1-PLC Stability 4.3.3 P1-PLC adsorption 4.3.4 Non-specific binding of P1-PLC 4.4 Continuous harvesting 4.4 Comparison of protein production processes  39 39 46 46 46 46 47 49 50 51 53 54 56 61 66 70  CHAPTER 5 Conclusions  73  CHAPTER 6 Future Work  75  Abbreviations  76  References  77  vi  Appendix I Growth of CHO cells in different growth media  85  Appendix 2 Variation of initial cell density  89  Appendix 3 Effect of harvesting conditions on p97 removal  92  Appendix 4 P1-PLC assay results  93  Appendix 5 Repeated harvesting process results  94  Appendix 6 p97 and PT-PLC stability  96  Appendix 7 Adsorption and desorption of PT-PLC to the cell surface  98  Appendix 8 Continuous harvesting process results  100  Appendix 9 Calculations for p97 yield on glucose and P1-PLC  102  vii  List of Figures A schematic of addition of hydrophobic sequence to a non-GPI-anchored protein  10  2.2  97 attached to the outer cell membrane by a GPI anchor p  13  2.3  A schematic model illustrating addition of protein to the GPI anchor in the endoplasmic reticulum  15  2.4  Hydrolysis of P1 into diglycerides and myo-inositols (Ikezawa, 1986)  17  3.1  Schematic showing the procedure of sample preparation for measuring the cell surface p97  28  3.2  A schematic of the reaction taking place in the wells  30  3.3  A schematic for the sample preparation procedure for PT-PLC assays  32  3.4  The schematic of the repeated harvesting process  35  4.1  Growth profile of CHO cells growing in CHO-S-SFM I. The figure shows the cell concentration (A), glucose concentration (El) and viability of the cells (V)  39  Growth profile of CHO cells growing in Ham’s F12 with 5 ig/mL of insulin, 5 p.g/mL transferrin, 10 nM sodium selenite, 50 .tg/mL BSA and 50 j.ig/mL DNase. The figure shows the cell concentration (A), glucose concentration (El) and viability of the cells (V)  40  Growth profile of CHO cells growing in Ham’s F12 with 5 j.ig/mL of insulin, 5 jig/mL transferrin, 10 nM sodium selenite, 300 .tg/mL BSA and 50 j.tg/mL DNase. The figure shows the cell concentration (A), glucose concentration (LI) and viability of the cells (V)  40  Growth profile of CHO cells growing in DMEM!F12 with 50 jig/mL DNase. The figure shows the cell concentration (A), glucose concentration (LI) and the viability of the cells (V)  41  Growth profile of CHO cells growing in DMEMIF12 with 5 p.g/mL of insulin, 5 j.ig/mL transferrin, 10 nM sodium selenite, 50 g/mL BSA and 50 .Lg!mL DNase. The figure shows the cell concentration (A), glucose concentration (LI) and viability of the cells (V)  41  2.1  4.2  4.3  4.4  4.5  viii 4.6  4.7  4.8  4.9  4.10  4.11  4.12  4.13  4.14  4.15  Growth profile of CHO cells growing in HBCHO. This figure shows the cell concentration (A), glucose concentration (LI) and viability of the cells (V) for a batch culture  42  Growth profile of CHO cells growing in CHO-S-SFM II. The figure shows the cell concentration (A), glucose concentration (LI) and viability of the cells (V)  43  Growth profile of CHO cells growing in CHO-S-SFM TI and 100 j.ig/mL DNase. The figure shows the cell concentration (A), glucose concentration (LI) and viability of the cells (V)  44  Effect of different incubation times on p97 harvest. Each sample of 6 cells was treated with 200 jiL of PT-PLC at 37 °C 4x10  46  6 cells was Effect of PT-PLC volume on p97 harvest. A sample of 4x10 treated with 50, 100, 200 and 400 iiL of P1-PLC solutions of 5 mUImL. Incubation time of 1 h was used  47  Typical standard curve obtained by PT-PLC from Boehringer Mannheim 6 cells at pH 7.5. 200 iiL of P1-PLC solution and approximately 4x10 per sample were used. A cell incubation time of 1 h at 37 °C was used to obtain this curve  48  Typical standard curves obtained from assays based on flow cytometry and immunofluorescence. The open triangles represents p97 concentration (A) and the open squares represents p97 removal (LI). 200 tL of PT-PLC solution was used to treat 4x10 6 cells. Incubation time of 1 h at 37 °C was used  49  Concentration of p97 recovered for 5 repeated harvests. Each harvest was conducted with 108 cells for 30 minutes. P1-PLC enzyme (0.5 mL) at an initial concentration of 30 mU/mL was used. Open circles (0) represent the measured values and the solid circles (.) represent calculated values which account for sampling dilution  51  Percent of p97 removal from the cells after each harvest. Approximately 6 cell sample was taken from the cell pellet and labeled with 4x10 the fluorescent antibody  52  p 9 7 stability at 37°C for a period of 48 h. Initial p97 concentration was 9.65 j.tg/mL in the immunoassay buffer. The error bars represent the standard error mean  53  ix 4.16  4.17  4.18  4.19  4.20  4.21  4.22  4.23  4.24  P1-PLC stability at 37 °C. PT-PLC from the bacterial supernatant at 52 mU/niL was used. 200 iiL of two dilutions of each sample were used 6 cells for 1 h at 37 °C to treat 4x10  54  Effect of pH on P1-PLC activity and . 97 16 mU/mL P1-PLC was p 6 cells for I h at 37 °C. PBS buffer with I mg/mL used to harvest 4x10 of BSA at pH = 6.0, 6.5, 7.0, 7.5 and 7.9 was used for these harvests. The open squares (El) stand for p97 concentration and the open triangles (t?) stand for p97 removal  55  Adsorption of P1-PLC (0) to the cell surface over a period of 60 minutes. cells/mL of P1-PLC enzyme at 18 mU/mL. A control on adsorption (•) experiment was done with P1-PLC without cells incubated at 37 °C for 1 h  57  Desorption of PT-PLC from the cell surface (0) in I mL of PBS. PT-PLC activity is less than 1 mU/mL. The cell concentration of 7 cells/mL was used 5x10  58  Adsorption equilibrium of PT-PLC at 37 °C. Different enzyme concentrations at 1 cells/mL were used to study adsorption  60  Cumulative p97 from repeated harvesting with PT-PLC replenishment. The initial harvesting was started with 0.5 mL of 300 mU/niL of PT-PLC and then 25 !.IL of P1-PLC enzyme at 3000 mU/niL was added after the 5th and 6th harvests. Incubation time of 30 mm at 37 °C was used for each harvest  62  97 removal from the cell surface during the repeated harvesting p process. The initial harvesting was started with 0.5 mL of 300 mU/mL of P1-PLC and then 25 .tL of PT-PLC enzyme at 3000 mU/mL was added after the 5th and 6th harvests. Incubation of 30 mm at 37 °C was used for each harvest  62  Percent p97 removal from the cell surface for 5 harvests lx 108 cells was monitored. Harvesting solution of 300 (0) and 30 (zX) mU/mL were used. Incubation of I h at 37 °C was used for each harvest  64  Open symbols represents cumulative p97 measured after each harvest and the solid symbols represents the calculated values of the p97 harvest which includes p97 lost due to sampling. This figure shows p97 recovered from 300 (0) and 30 (A) mU/mL of P1-PLC. Incubation of 1 h at 37 °C was used for each harvest  64  x 4.25  4.26  4.27  4.28  Continuous harvesting of CHO cells in the growth media. The solid symbols represent cell concentration and the open symbols stand for viabilities. Cultures with P1-PLC at 0 (LI) 3 (A) and 30 (0) mUImL of medium  66  Glucose concentrations during the continuous harvesting process with 0 (LI), 3 (A) and 30 (•) mU/mL of PT-PLC enzyme  66  Fluorescence due to cell surface p97 during the continuous harvesting process with 0 (LI), 3 (A) and 30 (0) mU/mL of P1-PLC enzyme  52  Cumulative p97 in the supernatant from the continuous harvesting process with PT-PLC at 0 (LI), 3 (A) and 30 (0) mU/mL of medium  67  xi  List of Tables 2.1  Functions of GPI-anchored proteins found on the cell surface  18  4.1  Comparison of P1-PLC assay based on flow cytometry and immunoflourescence assay Each sample was assayed in duplicate at 3 different concentrations  50  Stability of p97 in the harvesting solution of PT-PLC in PBS (1 mg/mL BSA). The samples were assayed at 3 different dilutions. The error bars were less than ± 1.5 .ig/mL  53  8 cells/mL of P1-PLC change in the supernatant after 15 mm. 1x10 P1-PLC solutions at different concentrations were incubated at 37°C  59  Comparison for transfected and untransfected (CHOWTB) CHO cells. Incubation time of 15 mm and I mL of P1-PLC solution was used  61  Comparison of p97 harvests from different harvesting processes  70  4.2  4.3  4.4  4.5  xli  Acknowledgments I would like to thank my supervisors Jamie Piret and Malcolm Kennard for their continuous support in completing this thesis.  I am grateful for the financial support  provided by Aga Khan Foundation, Geneva. I would also like to thank all my co-workers at Biotechnology Laboratory, especially Marta for their support and help at all times. Last but not least, I would like to thank Jurgen for his constructive criticisms and for being a good friend.  CHAPTER 1 Introduction Mammalian cells are used for producing vaccine, therapeutic and diagnostic recombinant proteins. They are able to carry out post-translational modifications such as glycosylation, acetylation and proteolytic processing that is often necessary to produce functional proteins.  However, in many processes, the desired protein is secreted at low  concentrations (i.e. 0.01 mg/mL) into the cell culture medium containing contaminating proteins at much higher concentrations (i.e. 0.1  -  1 mg/mL). These contaminants consist  of medium proteins which promote cell growth, other secreted proteins and proteins released by cell lysis. Hence, extensive downstream processing is necessary to purify the desired protein which results in losses of product and high process costs.  The use of  serum-free media has reduced the contaminating protein levels, but these media often still contain proteins in higher concentrations than the product protein. The initial purity and concentration of the desired proteins can be increased by controlled release techniques (Sambanis eta!., 1990a; Kennard eta!., 1993).  These  techniques separate the cell growth and protein expression from the harvesting of the product. A growth medium is used to provide nourishment for cell growth and protein synthesis.  In the growth medium the protein product remains cell-associated.  The  periodic replacement of growth medium with harvesting medium releases the product into a low protein containing buffer. This results in a more concentrated and purer product, hence decreasing downstream processing costs. In this thesis the controlled release technique developed by Kennard eta!. (1993) was investigated.  The model recombinant protein system studied was the human  melanoma tumor antigen p97 (melanotransferrin) expressed by chinese hamster ovary  2  (CHO) cells. The antigen p97 is a glycosyl phosphatidylinositol (GPI) anchored protein expressed on the outer cell membrane which can be harvested by a specific bacterial enzyme phosphatidylinositol phospholipase C (PT-PLC). p97 is useftul as a potential anti cancer vaccine (Hu eta!., 1988, Estin et a!., 1988; 1989),  Production of p97 by the  controlled release method represents a model system that can be applied to other GPI anchored proteins such as Leishmania gp63 and Malaria gp42 vaccines and also artificially made GPI-anchored proteins (Lin eta!., 1990; Moran and Caras, 1991; Scallon et a!., 1992). The aim of this project was to explore different harvesting techniques to improve the controlled release process developed by Kennard et a!. (1993). Before studying the harvesting processes a number of different types of serum-free media were investigated (Section 4.1) to obtain high density single CHO cell suspensions to facilitate the subsequent cell analysis. The alternative harvesting processes  investigated were  (i) repeated and  (ii) continuous harvesting. (i) The repeated harvesting technique (Section 4.3) involved multiple harvesting of cells using the same enzyme solution for up to 10 separate batches of cells.  This process increased the recovered concentration of the desired protein  product. (ii) The continuous harvesting technique (Section 4.4) involved harvesting of the product during cell growth and protein expression. PT-PLC was added to the growth medium and p97 released into the culture continuously. In this harvesting process the cells grew in the presence of PT-PLC, allowing p97 cleavage as soon as it was expressed on the outer cell membrane. This approach increased the cell specific p97 productivity. To help understand the performance of the new harvesting processes the stability of p97 (Section 4.3.1) and PT-PLC were investigated.  Available PT-PLC assays were  laborious and not sensitive enough for our needs. Hence, an improved PT-PLC assay was developed (Section 4.2).  The stability of P1-PLC (Section 4.3.2) at 37 °C was  determined. The effect of pH (Section 4.3.2) on p97 harvest was also investigated.  3  Phospholipases tend to bind non-specifically to cell surfaces (Deems et al., 1975). The PT-PLC adsorbed to the cell would be lost after each harvest thus depleting the enzyme in the harvesting solution. Hence, P1-PLC adsorption to the cells at 37 °C was studied (Sections 4.3.3 and 4.3.4). Finally, the performance of the cyclic, continuous and repeated harvesting techniques were compared (Section 4.5).  4  CHAPTER 2 Literature Review 2.1 Mammalian recombinant protein production Mammalian cells have been used to produce viral vaccines against smallpox, rabies (before 1930s), yellow fever (1930) and poliomyelitis (1949). In the 1970’s, recombinant DNA technology was developed and enabled the expression of mammalian proteins in bacterial cultures.  The simplicity of culturing bacteria provided an attractive alternative to  mammalian cell culture.  Bacterial media are usually based on single carbon energy  sources and are less costly than the media needed by mammalian cells, which contain expensive growth factors and serum. 15  -  A typical bacterial growth doubling time is  30 minutes as compared to 15 30 h for mammalian cells. -  This allows higher  productivities in bacterial cultures (approximately 100 times). However, bacterial cells are unable to carry out many of the post-translational modifications of mammalian cell proteins. These modifications include proteolytic cleavage and addition reactions such as glycosylation or carboxylation that are essential for the correct biological functions of many proteins.  For example, glycosylation could protect a protein against proteolytic  breakdown, hence maintain its structural stability (Butler, 1987).  Many recombinant  mammalian cell proteins are secreted into the medium and can be recovered and purified from the spent medium. The recovery of secreted proteins was more convenient than the commonly used recovery of protein from lysed bacterial cells. Lysing the cells resulted in extensive contamination of the protein of interest and could also release endotoxins. For these reasons, the use of mammalian cells for recombinant protein production has been expanding rapidly.  5  Most commercially produced proteins and glycoproteins are used as diagnostic, therapeutic or veterinary products and hence need extensive purification.  Products  secreted from mammalian cells include erythropoietin, tissue plasminogen activator and monoclonal antibodies. Extensive downstream processing can lead to high production costs.  Therefore, a careful study and optimization of the production processes is  important, so as to meet the demand at affordable costs.  2.1.1 Conventional protein production processes The most commonly used mammalian cell protein production processes involve secretion of proteins into the culture medium.  The cells are normally grown in stirred tank  bioreactors and the proteins recovered from the spent media. The protein productivities of mammalian cells are low and the protein is secreted at low concentrations (usually I  -  100 .tg/mL).  Spent media contains many other contaminating proteins often at  relatively higher concentrations, resulting in low purity of the desired protein. Different strategies have been implemented in attempts to increase bioreactor productivity and initial protein product concentration. For example: (i) Growing adherent cells on microcarriers. Microcarriers are able to achieve higher cell densities, which result in higher productivities. (ii) Gene amplification.  The number of gene copies per cell is increased, resulting in  increased RNA and recombinant protein production. However, there has been limited success in increasing the protein product concentrations. Since, the major use of proteins produced by mammalian cells is for medical purposes, it is important to achieve high purity. This results in extensive purification and high production costs. (Bailey and Ollis, 1986):  A typical purification process consists of the following steps  6  (i)  Removal of solid particles such as cells. This may require one of the following  operations: filtration, centrifugation or sedimentation. (ii)  The protein is then isolated and concentrated by one of the following operations:  solvent extraction, precipitation or ultrafiltration. (iii) The protein is further purified by either fractional precipitation, chromatography or adsorption. (iv) The final purification step involves recovering the product in the form required by the consumer. This usually requires crystallizing and drying. To attain higher protein concentrations and optimize the production process, controlled release techniques have been explored.  There are two types of controlled  release techniques that have been studied so far: (i) regulated secretion of intracellular proteins (Sambanis et a!., 1990a) and  (ii) enzymatic harvesting of membrane bound  proteins (Kennard eta!., 1993).  2.1.2 Regulated secretion The use of regulated secretion for protein production was first reported in 1990 by Sambanis et at.  Mouse AtT-20 cells can store secretory proteins intracellularly in  secretory vesicles and release them when stimulated by an inducer. This cell line was genetically engineered to express recombinant human insulin and growth hormone. Secretion of these proteins was induced by 8-bromo cyclic AMP. In a cyclic secretion protocol the cells were exposed alternatively to growth and secretion medium. Secretion was induced in low protein medium The secretion rates of human insulin were increased up to 6-fold during the induction phase. However, this increase was only observed for three cycles after which the cells started detaching from the culture surface and induced secretion decreased (Sambanis et at., 1990a). Induced secretion increased the production of human growth hormone approximately 4-fold.  Up to 60 % of the total cellular  7  production of human growth hormone was recovered during the induction phase. The induced secretion rates decreased after the first five cycles (Sambanis eta!., 1990b). Efforts also were made to apply this technique to a f3TC3 cell line. Signal transduction pathways were manipulated to reduce secretion of insulin during the production process by adding cyclic AMP. However, this decreased the expression of insulin (Gramp eta!., 1992).  2.1.3 Enzymatic harvesting of membrane bound proteins An alternative production process based on glycosyl phosphatidylinositol (GPI) anchored membrane proteins was developed by Kennard eta!. (1993). A CHO cell line genetically engineered to express the GPI-anchored protein, p97, on the outer cell membrane was used as a model system. p97 was harvested by PT-PLC, a specific enzyme that cleaved the GPI anchor releasing p97 into the medium.  The harvesting medium consisted of  phosphate buffered saline (PBS) containing 10 mU/mL (0.02 Ig/mL) of P1-PLC enzyme. 7 cells/mL, Over 35 jig/mL of p97 at 30 40 % purity was repeatedly recovered from 5x10 -  harvested in a cyclic fashion over a period of 44 days. The contaminating proteins are believed to be mainly other GPI-anchored proteins released from CHO cell surface. After harvesting, the cells were returned to fresh medium and they re-expressed the protein within approximately two days.  The cells could then be reharvested.  The repeated  harvesting of the cells did not affect the growth rate, viability or protein production of the cells. This controlled release technique was developed using a suspension CHO cell line. The repeated centrifugation and washing of the cells required during the harvesting process was considered impractical for industrial scale protein production. To facilitate medium changes harvesting from surface attached CHO cells was investigated (Kennard and Piret, in press).  8  Initially harvesting p97 from adherent CHO cells cultured in T-flasks was studied. Only 1.5  -  4 .tg/mL of p97 was harvested from these cells (Kennard and Piret, in press).  There were two reasons for these low p97 harvests:  (i)  relatively large volumes of  PT-PLC solution were required to cover the entire surface area containing the cells for the harvesting process. (ii) after achieving 100 % confluency, the cells tended to lift off the surface decreasing the total number of cells available for harvest. Similar culture stability problems were encountered by Sambanis eta!. (1990a; 1990b).  To overcome culture  stability problems, the adherent CHO cells were grown on porous microcarriers. Microcarriers have many advantages over conventional culture methods such as: (i) They can be used in packed-bed and fluidized-bed bioreactors; (ii) Higher cell densities can be achieved, resulting in higher productivities; (iii) Cells are protected from mechanical agitation and sparging and shear stresses; (iv) Cells can be maintained at reduced serum levels.  This system resulted in high cell densities and stable cultures.  Of particular  importance for the controlled release process, rapid sedimentation of the microcarriers facilitated easy handling when replacing medium or washing the cells for the controlled release process. The concentration of the product recovered increased to approximately 100 tg/mL at 25  -  30 % purity. Stable protein production was attained for 15 harvest  cycles over a 30 day period. Besides being expensive, the serum necessary for cell growth contributes to trace contamination of the product.  Therefore, efforts were made to  monitor the effect of reduced serum levels in the growth media on harvested p97. Decreasing serum from 10 to 0.5 % did not effect the cell specific p97 production of porous microcarrier immobilized cells (Kennard and Piret, in press).  9  2.2 Application of model harvesting system The technique of controlled release harvesting of GPI-anchored proteins could be used for a wide range of proteins of interest. There are many naturally occurring proteins such as the potential vaccines Leishmania gp63 and Malaria gp42 that have medical importance. It was determined that the GPI pre-anchor recognition sequence was contained within the 37 amino acids at the COOH terminus (Caras et at., 1987a,b). Addition of this sequence to a secretory protein resulted in GPI anchoring and targeting of the fusion protein to the plasma membrane (Caras et at., 1 987b). Recent work has shown that a pair of amino acids serving as a cleavage/attachment site positioned 10-12 residues from the 2 NH  side  of the  pre-anchor  sequence  were  required  for  GPI  anchoring  (Moran and Caras, 1991). Figure 2.1 shows the process of a non-GPI-anchored protein fused with the hydrophobic pre-anchor sequence from a naturally occurring GPI-anchored protein.  This produces GPI-anchored fusion proteins.  Human growth hormone is an  example of the secretory proteins which have been targeted to the plasma membrane as GPI-anchored proteins by using a gene fusion from GPI-anchored decay accelerating factor (Caras and Weddell, 1989; Lisanti et al., 1989).  2.3 Melanotransferrin Melanotransferrin, the model protein used in this thesis, is also known as the melanoma tumor-associated antigen p97. It is a cell surface glycoprotein (Molecular weight 97 kDa) expressed by human melanoma cells (Brown et al., 1981a), but present in only trace amounts in normal adults (Brown et at., 198 ib). The human melanoma, cell line SK-MEL 28, expresses approximately 4x10 5 molecules of p97 per cell (Brown et a!., 1981a), while 6 the CHO cell line, genetically engineered to express p97 produces approximately 2 5x10 -  10  GPI anchored protein e.g. Decay accelerating factor (DAF)  Non GPI anchored protein of interest  I  I  DAF  I  I I Hydrophobic signal sequence  I  P P DAF fusion protein -  Figure 2.1  protein  A schematic of addition of hydrophobic sequence to a non-GPI-anchored  11  molecules per cell  Although the amino acid sequence of p97 is approximately 40 %  identical to human transferrin and lactoferrin (Brown et a!., 1982), unlike these proteins it is membrane bound (GPI-anchored) and differs in amino acid residues which are involved in iron binding. It has been proposed that p97 has an intact transferrin type iron binding site in its N-terminal domain and a possibly defective site in its C-terminal domain (Baker eta!., 1987). Baker eta!. (1992) confirmed that there was only one iron binding site present on p97.  Rose et a!. (1986) reported that p97 played a role in iron  translocation while Richardson eta!. (1991) refuted this observation. Since most human melanomas express high levels of GPI-anchored p9’7, its effectiveness as an anti-cancer vaccine has been explored (HellstrOm and HelistrOm, 1969; Hu eta!., 1988). Immunization with a recombinant vaccinia virus, v-p97NY, did induce humoral and cell-mediated immunity against melanoma-associated antigens in monkeys (Estin eta!., 1988).  2.4 Glycosyl phosphatidylinositol anchors In eukaryotic cells, many other proteins are anchored to the external surface of the plasma membrane by covalently attached glycolipids containing inositol.  These anchors are  known as glycosyl phosphatidylinositols (GPI). In 1963, Slein et a!., noted that alkaline phosphatase on mammalian cell surface could be released by PT-PLC and proposed the presence of GPI anchors. Almost 15 years later, it was confirmed that mammalian cells have GPI-anchored proteins attached to the plasma membrane (Ikezawa et a!., 1976; Low & Finean, 1977). Due to low expression levels of the GPI-anchored proteins such as alkaline phosphatase and acetylcholinesterase (i.e. approximately 1,000  -  10,000 molecules per cell) it was difficult to study the structure  and release mechanisms of the anchor. In the mid 1980’s the Tiypanosonza brucei variant surface glycoprotein (VSG) was found to be anchored to the cell surface by GPT at  12  approximately  io  molecules per cell and thus could be more easily purified  (Ferguson eta!., 1985).  The structure and release mechanisms of GPI anchors were  determined by the late 1980’s (Ikezawa, 1986; Low, 1989).  2.4.1 Structure of GPI anchor The GPI anchor consists of a phosphoethanolamine and a variable glycan portion. The glycan moiety of the anchor varies with cell type.  The following is the breakdown of the  GPI anchor assembly (Figure 2.2): (i)  The protein is covalently bound to the GPI anchor via the x-carboxyl of the C-terminal amino acid.  This C-terminal amino acid is not specific for GPI  anchoring. (ii)  The ct-carboxyl group of the C-terminal amino acid is amide linked to the amino group of a phosphoethanolamine moiety.  (iii)  The phosphoethanolamine is linked to a variable glycan section that consists mainly of mannose and glucosamine.  (iv)  The glycan portion is glycosidically linked to an inositol-containing phospholipid, phosphoinositol.  (v)  Phosphatidylinositol anchors the protein to the membrane.  PT-PLC cleaves the GPI anchor at the phosphodiester bond between the phosphoinositol group and the lipid portion (Figure 2.2). The protein is released into the medium in a soluble form with the C-terminal linked to ethanolamine, the glycan moiety and phosphoinositol. Diacylglycerol is presumably left in the membrane (Ikezawa, 1986).  13 2 NH  7 p 9 NH  I CH,  Ethanolamine  CH, QE  —Q 0  Inositol  0  P1-PLC cleavage site  0  CH CH — 2  0  CH,  0  MEMBRANE (Cr (CH CH  Figure 2.2 p 97 attached to the outer cell membrane by a GPI anchor  CH  14  2.4.2 Attachment of protein to GPI anchor The attachment of GPI anchor to the protein is a rapid post-translational process (Low, 1989) occurring soon after protein synthesis.  This process occurs in the  endoplasmic reticulum, and not in the Golgi apparatus or plasma membrane (Low, 1987; Doering et at., 1990; Boivin and Delaunay, 1991). A schematic model of protein addition to GPI is shown in Figure 2.3. The protein is synthesized with a pre-anchor sequence, which contains the cleavage/signal site. The GPI anchor is synthesized independently in the endoplasmic reticulum. Part of the cleavage/signal sequence is cleaved and replaced by the pre-assembled GPI anchor (Ferguson and Williams, 1988). This replacement leaves behind  a  short  C-terminal  hydrophobic  domain  of the  pre-anchor  sequence  (Lisanti eta!, 1990; Doering et a!., 1990). The rapid addition of the anchor suggests that the removal of hydrophobic C-terminal domain and addition of GPI may be catalyzed by the same transamidase enzyme (Doering et a!, 1990).  2.4.3 Functions of GPI anchors There are no apparent common functions of GPI-anchored proteins that could help explain the presence of the anchor compared to the more common transmembrane peptide sequences. The following are a few of the possible functions of GPI anchors which have been suggested: (i) Protein motility: GPI anchored proteins have about ten-fold increased lateral mobility and diflujsion coefficients on the order of 1 (A. Ishihara et al., 1987;  Low, 1987;  -  lsec have been measured 2 4x 1 0 cm  Boivin and Delaunay, 1991).  Transmembrane  /sec (Gall and Edelman, 2 glycoproteins have diffusion coefficients of 0.5 6x 10b0 cm -  1981).  15  1  0  cc  -d 0  z  I 1  0 0  z  + r-)  I  +-c0  Figure 2.3 A schematic model illustrating addition of protein to the GPI anchor in the endoplasmic reticulum.  16  (ii) Protein removal: GPI anchors regulate the removal of the protein based on their susceptibility to mammalian phospholipase C enzymes (Boivin and Delaunay, 1991). By an endocytotic pathway, GPI-anchored proteins are recycled into compartments containing cellular P1-PLC. PT-PLC cleaves the protein, which is then either degraded by proteases or recycled back to the membrane and released into the supernatant (Ferguson and Williams, 1988). (iii) Protein secretion and sorting: GPI anchors can play a specific role in translocation of proteins across membranes. Studies have shown sorting of GPI-anchored proteins to the apical membrane of polarized epithelial cells (Cross, 1990). (iv) Cell protection: The glycan group lies along the plane of the membrane, hence it could form a diffi.ision barrier which protects the cells from foreign materials (Boivin and Delaunay, 1991). (v) Cell signaling: GPI anchors were found to be involved in signaling events evoked by proteins such as insulin (Low and Saltiel, 1988). These proteins triggered the production of GPI specific phospholipase C (mammalian PT-PLC).  Action of P1-PLC on  phosphoinosides results in diacylglycerols (Figure 2.4), an activator of protein kinase C and inositol phosphates.  Protein kinase C and inositol phosphates evoke specific  responses to insulin (Boivin and Delaunay, 1991).  2.4.4 Functions of GPI-anchored proteins During controlled release harvesting of p97, P1-PLC cleaves other GPI-anchored proteins from the cell membrane. Hence, it was important to review the functions of GPI-anchored proteins, to evaluate the potential impact of harvesting on cell function. These proteins have been identified and characterized according to their biochemical functions. biological functions of most of the proteins are still unknown.  The  I  CD  Cl)  0  Cl)  0  9  0-  Cl)  CD  0  -t  C) CD  0 0-  0  Cl) Cl)  0  Diglycerides  CH,OH  HCOCOW  CH,OCOR  ±  -  P -  PT-PLC  OH  -OCH.  myo-inositol  Phosphatidylinositol  OR  HCOCOR’  myo-inositol- 1,2-cyclic phosphate  OH  OH  —p  El  0  CH,OCOR  -  O  OH  -  OH  1-phosphate  -  0  18  Table 2.1 Functions of GPI-anchored proteins found on the cell surface  Protein  Source  Function  Reference  Alkaline phosphatase  Mammalian tissues  Hydrolase  Low and Finean, 1977; Taguchi and Ikezawa, 1978  5’ Nucleotidase  Mammalian tissues  Hydrolase  Low and Finean, 1978; Shukia eta!., 1980  Acetylcholinesterase  Mammalian blood cell  Hydrolase  Low and Finean, 1977; Low eta!., 1987  Alkaline phosphodiesterase I  Rat tissues  Hydrolase  Nakabayashi and Ikezawa, 1984, 1986  Variant surface glycoprotein  Trypanosorna Brucei  Protective coat  Ferguson et al., 1985; Low et al., 1987  Thy-I  Mammalian brain and T lymphocytes  Antigen  Low and Kincade, 1985; Tse eta!., 1985  Trehalase  Rabbit tissues  Hydrolase  Takesue et aL, 1986  Decay accelerating factor  Human blood and HeLa cells  Complement regulatory protein  Davitz eta!., 1986, 1987  gp63  Leishniania major  Protease  Bordier et at., 1986; Etges eta!., 1986  RT-6  Rat lymphocytes  Antigen  Koch eta!., 1986  Qa  Mouse T lymphocytes  Antigen  Stiernberg eta!., 1987  ThB  Mouse lymphocytes  Antigen  Stiernberg et at., 1987  T-cell activating protein  Mouse T lymphocytes  Antigen  Reiser et at., 1986  120 N-CAM  Rat, mouse and chicken brain  Cell-cell interactions  He eta!., 1986; Hemperly et a!., 1986  Heparan sulphate proteoglycan  Rat liver  Cell-cell and Cell-matrix interactions  M. Ishihara et aL, 1986  -  19  According to Turner (1990), over 50 GPI-anchored proteins are present on cell-surfaces. However, Kennard et a!. (1993) found that removal of these GPI-anchored proteins did not effect the viability and growth of the cells.  2.5 Phosphatidylinositol phospholipase C P1-PLC is produced by a number of bacteria including: Staphylococcus aureus (Doery eta!., 1965), Bacillus cereus (Slein and Logan, 1965), Clostridium novyl (Taguchi and Ikezawa, 1978) and Bacillus thuringiensis (Taguchi et a!., 1980). P1-PLC hydrolyzes phosphatidylinositol (P1) and lysophosphatidylinositol, but does  not  recognize  other  more  common  membrane  phosphatidyls  such  as  phosphatidyicholine, phosphatidylethanol amine and phosphatidylglycerol (Ikezawa and Taguchi, 1981).  This enzyme specifically recognizes the inositol-phosphate structure  present in PT, such that it does not hydrolyze even the more highly phosphorylated derivatives of PT (for example P1-4-phosphate and P1-4,5 bisphosphate) (Kuppe eta!., 1989).  P1-PLC cleaves most GPI anchors, however, some proteins with GPI  anchors are partially or completely resistant to PT-PLC (Ferguson and Williams, 1988). In this study, Bacillus subtilis transfected with PT-PLC gene from Bacillus thuringiensis was used to produce P1-PLC.  2.5.1 Properties of P1-PLC from B. thuringiensis B. thuringiensis was found to be the highest producing bacterial source of PT-PLC. PT-PLC could be purified to a homogeneous state by polyacrylamide gel electrophoresis (Taguchi eta!., 1980). P1-PLC produced by B. thuringiensis has a molecular weight of approximately 23 ± 1 kDa (Taguchi et a!., 1980; Ikezawa and Taguchi, 1981; Ikezawa, 1986).  The  20  optimum pH is between 5  -  8.5 with maximal activity at 7.5 (Ikezawa and Taguchi, 1981).  The isoelectric point (p1) is approximately 4.9 5.4 (Ikezawa and Taguchi, 1981; -  Kupke eta!., 1989; GrifflthetaL, 1991). P1-PLC stored in 20 mM Tris HC1 and pH 7.5 is stable after many freeze-thaw cycles at 0.1  -  3 mg of proteinlmL and at 37 °C during prolonged incubations  (Griffith eta!., 1991). Fresh bovine serum albumin (BSA) protects the enzyme at high temperatures. P1-PLC diluted in 0.1 % fresh BSA retains greater than 60 % of its activity when exposed to temperatures from 70 to 100 °C for 10 minutes. However, when diluted in freeze-thawed B SA, it became thermolabile and retained only 20 % of its activity in 10 minutes at the same temperatures (Kume eta!., 1992). , 2 , Mg 2 PT-PLC enzyme activity is inhibited by divalent metal ions such as Ca 2 J44fl  and  2 Zn  at  concentrations  above  10 M  (Low and Finean, 1976;  Taguchi etal., 1980). This inhibition is pH dependent (Sundler eta!., 1978) and possibly due to interactions with the substrate rather than the enzyme (Griffith eta!., 1991). Potassium chloride and sodium chloride are inhibitory at concentrations higher than 10*2 M (Sundler et al., 1978;  Taguchi et aL, 1980;  Ikezawa and Taguchi, 1981).  Other  ) at 2 compounds that completely inhibit the enzyme activity are mercuric chloride (HgC1 0.5 mM and p-chloromercuriphenyl sulfonic acid (PCMBS) at 5 mM.  The inhibitory  2 and PCMB S treated enzyme is completely restored by excess addition of activity of HgC1 the reducing agent dithiothreitol (DTT). Triton X-100 and sodium deoxycholate stimulate the activity of P1-PLC (Taguchi et al., 1980). P1-PLC cleaves GPT anchors more efficiently when treating cells from a culture at high cell density (Berridge, 1987).  21  2.5.2 P1-PLC assays The reported PT-PLC assays are based on monitoring the change in substrate or product concentrations. One unit of P1-PLC is defined as the enzyme activity that hydrolyzes 1 iimol phosphatidylinositol per minute at 37 °C and pH 7.5 to phosphoinositol and diacylglycerol:  PT-PLC PT  >  phosphoinositol  +  diacyiglycerol  (2.1)  There are 5 methods reported for determining PT-PLC activity: (1)  The determination of water-soluble inositol phosphate from radiolabelled  phosphatidylinositol (Griffith et al., 1991). This assay involves radiolabelling of rat liver H]inositol. 3 microsomes with [  Lipids extracted from these microsomes are used as a  substrate. Radiolabelled lipids are treated with PT-PLC at 37 °C and the reaction stopped after 10 minutes.  Hjinositol in the supernatant is determined and 3 Radioactivity of [  converted to PT-PLC activity. This assay is time consuming and expensive. (2)  Quantitation of phosphate released from phosphoinositol by the Eibl and Lands  method (1969).  In this assay phosphatidylinositol containing lipids from rat liver are  treated with PT-PLC at 37 °C. The reaction is stopped after 10 minutes and myo-inositol (product from the reaction above, Figure 2.4) is decomposed to phosphorous. Phosphorous is then measured by Eibl and Lands method. This is the most used PT-PLC assay. However, it is lengthy and laborious. (3) Quantitation of soluble GPI-anchored proteins from biological membranes (Ikezawa and Taguchi, 1981).  This assay is based on proteins such as GPI-anchored alkaline  phosphatase which can be quantified by their enzyme activity. In this assay, tissues from  22  rat kidneys are homogenized and treated with P1-PLC at 37 °C for 10  -  100 minutes. The  supernatant is then assayed for alkaline phosphatase. (4)  substrate,  Continuous fluorometric assay using a fluorescent  myo-inositol- 1-phosphate (Shashidhar eta!., 199 la,b).  myo-inositol- 1-phosphate  4-nitrophenyl  or  (2NIP)  2-naphthyl-  P1-PLC was added to the fluorescent substrate and the  fluorescence monitored at 403 nm wavelength. This assay is simple and gives immediate results. However, it can only measure activities above 14 mU/mL. (5)  P1 is treated with PT-PLC and diglyceride  Boehringer Mannheim assay # 5646.  produced from this reaction is reacted with lipase to produce glycerol.  A series of  reactions are then triggered with glycerol kinase, pyruvate kinase and lactate dehydrogenase.  Decrease in absorbance at 365 or 340 nm due to NADH depletion Absorbance is then converted to P1-PLC activity.  (Equation 2.6) is measured.  The  limitation of this assay was its sensitivity, it could only measure activities above 150 mU/mL.  P1-PLC PT  +  0 2 H  >  Diglyceride  +  phosphorlinosine  (2.2)  Lipase Diglyceride  +  0 2 H  >  fatty acids  +  glycerol  (2.3)  Glycerol kinase Glycerol  +  ATP  >  glycerol-3-P  +  ADP  (2.4)  pyruvate kinase PEP  +  ADP  >  pyruvate  +  ATP  (2.5)  23  Lactate dehydrogenase Pyruvate  +  NADH  +  W  >  lactate  +  NAD  (2.6)  2.6 Protein adsorption Cells adsorb proteins rapidly and non-specifically in less than a minute (Missirlis eta!., 1990; Shin eta!., 1993). Deems et at., (1975) and Shin et a!. (1993) proposed that this adsorption of proteins to cell membranes was due to hydrophobic and not electrostatic 2 binds to the surface before reacting with the substrate and interactions. Phospholipase A this binding is reversible. The enzyme then moves on the surface as an enzyme-surface complex until it collides with a phospholipid (Deems et at., 1975).  Deems et a!. also  suggest that binding studies can be applied to other phospholipases and lipases. Transferrin is also a widely investigated molecule because of its thnction as a growth stimulator. Studies on transferrin adsorption to a specific receptor resulted in a saturated state at approximately 48,000 molecules/cell (Reed, 1990).  24  CHAPTER 3 Materials and Methods 3.1 Cell line The wild type chinese hamster ovary (CHO) cell line CHOWTB (obtained from Dr. Maxfield, New York University) was co-transfected with the p97 expression vector, pSV2p97a and the G418 resistance vector, pWJ218 (Kennard et al., 1993). The entire coding region of p97 cDNA was present in the pSV2p97a vector and was driven by the SV4O promoter (Food et al., 1993).  3.2 Tissue culture The CHO cell line was maintained in suspension in serum free medium, CHO-S-SFM I 2 T-flasks (Nunc, (Gibco, Grand Island, N.Y). The cells were cultured in either 75 cm Gibco) or 150 mL and 250 mL spinner flasks (Belico, Vineland, N.J.). The cultures were 2 humidified atmosphere. incubated at 37 °C and under a 5 % CO The CHO cells also were grown in CHO-S-SFM II (Gibco), HBCHO (Irvine Scientific, Santa Ana, CA), Ham’s F12 (Gibco) and DMEMJF12 (Gibco), supplemented with 50  -  100 Ig/mL of DNase (Boehringer Mannheim, Laval, Quebec), 5 tg/mL of  insulin (Gibco), 5 jig!mL of transferrin (Gibco), 10 nM of sodium selenite (Gibco) and 50 300 g/mL of bovine serum albumin (Sigma, St. Louis, MO). -  7 cells/mL in The cells stocks were stored in liquid nitrogen at approximately 1xL0 1 mL aliqouts of a solution of 72 % CHO-S-SFM I, 20 % newborn calf serum (Gibco) and 8 % dimethyl sulfoxide (Fisher Scientific, Fair Lawn, NJ.).  25  3.3 P1-PLC production Bacillus subtilis (BG2320) transfected with the PT-PLC gene from Bacillus thuringiensis was provided by Dr. M. Low of Columbia University, N.Y. The cells were maintained in a medium containing 10 g/L of polypeptone (Becton Dickinson, Cockeysville, M.D.), 10 g/L of yeast extract (Difco, Detroit, Michigan), 0.4 g/L of potassium biphosphate (BDH, Toronto, ON), 5 g/L sodium chloride (Fisher Scientific) and 15 .tg/mL of chioramphenicol (BDH) for selection. The pH adjusted to 7.0 with 1 N NaOH (Fisher Scientific). 100 mL of medium was inoculated with Bacillus subtilis and grown overnight. This culture was then used to inoculate 2 L of medium at 2 3 % (v/v). The cultures were -  maintained in an Erlenmeyer flask at 37 °C at 150 rpm in a shaker bath (Lab-line Instruments, Meirose Park, IL.). After 12 h  (- late log phase) the culture was centrifuged  and the supernatant filtered using a 0.2 jtm membrane (VacuCap, Gelman Sciences). To partially purify the P1-PLC, 500 mL of filtered supernatant was precipitated using 600 g/L ammonium sulfate (Baker, Phillipsburg, N.J.).  The precipitate was separated by  centrifugation (Silencer, Japan) and the filtrate resuspended in 100 mL of 20 mM Tris HC1 and 3 mM EDTA buffer. The mixture was then concentrated to 10 mL using a 30,000 molecular weight cut off YM 30 membrane in an ultrafiltration cell (model 8400, Amicon). This was repeated two more times and the final 10 mL was diluted by 30 mL of the buffer solution. The enzyme solution was then aliquoted to I mL samples and stored at  -  20 °C.  For short term storage at 4 °C, the enzyme was diluted to the required concentrations in phosphate buffered saline (PBS) containing 1 mg!mL of BSA. PBS consists of 8.0 g/L , 0.2 g/L of PO 4 H 2 (Na sodium chloride (NaCl), 2.16 g/L sodium hydrogen phosphate ) O and 0.2 g/L of potassium chloride (KCI) with 4 P 2 (KH potassium hydrogen phosphate ) pH adjusted to 7.4.  26  The standard 77.8 U/mL P1-PLC solution (Boehringer Mannheim) was stored in 50 mM triethanolamine buffer with 10 mM EDTA and 10 mM sodium azide.  3.4 Production of monoclonal antibodies against p97 The 33B6E4 monoclonal antibody was provided by Dr. M. Kennard. L235 hybridoma cell line (ATCC NB 8446 L235 (M-19) was grown in roller bottles (Nunc, Gibco).  The  medium used to culture these cells was RPMI (Gibco) supplemented with non-essential amino acids, 10 g/L fetal calf serum (Gibco), 1 g/L mercaptoethanol (Sigma), 2 mM L-glutamine (Gibco), 2 mM proline (Gibco) and 0.1 mg/mL penicillinlstreptomycin. The cell culture was maintained in a batch culture until the viability decreased to 50 %. The cells were removed by centrifugation and the supernatant filtered using 0.2 tm membrane (Gelman Sciences). The monoclonal antibody was then purified using a protein G affinity column (MAbTrap G, Pharmacia LKB Baie d’Urfe, PQ) and concentrated to 1 - 2 mg/mL using 10,000 MW ultrafilter (Centricon-lO, Amicon Danvers, MA).  3.5 Analytical methods 3.5.1 Cell count The haemocytometer (Hausser Scientific, U.S.A.) and trypan blue (8 g/L trypan blue stain and 8.8 g/L NaCI, Gibco) dye exclusion method was used to monitor the cell density and viability of the cultures. Viable cells excluded the dye while the non viable ones took up the dye and were dyed reddish-blue. The haemocytometer was loaded with the sample diluted with 50 % trypan blue. The cells were then counted under the microscope (Nikon, Missisauga, ON).  27  3.5.2 Flow cytometry Immunofluorescence labeling of p97 and a flow cytometer analyzer (FACScan, Becton Dickinson, CA) were used to monitor the cell surface expression of p97. 6 cells To prepare cells for flow cytometry (FACS) analysis, approximately 4 5x10 -  per sample were spun in a 6 mL (12x75 mm) polystyrene tube (Becton Dickinson Labware, Lincoln Park, N.J.), the spent medium was removed and the cell pellet washed in 200 iiL of FACS buffer. The FACS buffer contained 0.5 % (w/v) bovine serum albumin (Sigma), 20 mM Hepes (Sigma) and 20 mM sodium azide (Baker) in Dulbecco’s Modified Eagle Medium (DMEM, Gibco).  The cells were then spun again and the cell pellet  resuspended in 100 j.tL of FACS buffer and labeled with fluorescinated (fluorescin isothiocyanate) antibody against p97 (33B6E4) obtained from Dr. M. Kennard (Section 3.4). 2 i.iL of fluorescinated antibody (4 mg/mL) was added to the cells and incubated at 4 °C for 45 minutes. The cells were then spun and the pellet washed with 1 mL of FACS buffer followed by 1 mL of PBS. After spinning the cells again, the cells were fixed in 1 mL of 1.5 % (v/v) p-formaldehyde (JBS, Pointe Claire-dorval, PQ) (Figure 3.1). 5000 events per sample were measured by the flow cytometry and average fluorescence per cell obtained.  3.5.3 Immunofluorescence assay A pandex fluorescence concentration analyzer (Idexx, Portland, ME) was used to determine the concentrations of p97 solubilized by the action of PT-PLC. The coated capture particles and the labeled antibody were provided by Dr. M. Kennard.  Carboxyl-polystyrene capture particles (0.87 tm diameter, 5 % w/v, Baxter,  Mundelein, IL) were ãoated with anti-p97 IgG (ATCC HB8446 L235 (M-19)). A second antibody to p97 IgG (33B6E4) was labeled with fluorescin isothiocyanate (FITC, Sigma).  CD CD  CD 0 CD  C  -t  Cl)  CD  -t  C C  -t  -e1 -eCD  CD  Cd,  CD C  -t  0 CD  C  -t  CD  C  Cd  0  CD  0  -  C  /  I  1000 I  U l0ocI  FACSCAN ANALYSIS  5E6 cells  wash, spin and remove supernatant  Fix the cells  wash, spin and remove supernatant  0.5 mL PBS + 0.5 mL p-formaldehyde (3 % v/v)  oç  leave overnight at4°C  spin and remove supernatant  200 jiL FACS Buffer  1 mLPBS  incubate at 4°C for45 mm.  100 p.L FACS buffer & 2 p.L fluorescinated antibody a p97 (4 mg/mL)  FLOW CYTOMETRY  wash, spin and remove supernatant  spin and remove supernatant  1 mLFACs Buffer  00  29  The assay was performed in 96 well plates (Idexx). These plates have a membrane at the bottom which enables removal of liquids from the wells using vacuum filtration. The p97 samples were diluted in immunofluorescence buffer (DMEM with 1 g/L sodium azide and 10 g/L bovine serum albumin) to within 0.15  -  1.5 tg/mL (linear range of  immunofluorescence assay), 20 iiL was added to each well and incubated with 20 iiL of the coated capture particles for 20 minutes at room temperature  (-- 20 °C).  Then a further  20 tL of the second fluorescinated antibody (50 .tg/mL) was added to each well and incubated for another 20 minutes at room temperature (Figure 3.2). The plate was then drained and washed 3X with PBS within the pandex analyser and the fluorescence in the wells read using a 485/535 nm filter at lox gain. concentrations ranging from 0.15  -  Standards with known p97  1.5 Ig/mL were used to produce a calibration curve  for each plate.  3.4.4 Glucose analysis Glucose analyser 2 (Beckman Instruments, Fullerton CA) was used to measure glucose concentrations of the cultures. approximately 0.1  - 4.5 g/L.  It can measure the glucose concentrations from  Glucose measurement is based on monitoring the decrease in  oxygen concentration. It uses an enzymatic glucose electrode to carry out the following reaction:  Glucose oxidase J3-D-glucose  +  02  >  Gluconic acid  +  0 2 H  (3.1)  All measurements were done in triplicates and the average of three taken as the representative value.  CD  CD  C) CD  0  C)  CD  CD  0  C)  CD  C)  Cl)  97 p  20 iL sample for p97 analysis p97  9J  PANDEX ANALYSIS  Measure fluorescence  \97  20 .tL capture particles (0.25 % wlv) coated with IgG CL p97  20 mm at room temperature  IMMUNOFLUORESCENCE ASSAY  20 mm.  Drain  I  1  20 tL fluorescinated antibody CL p97 (50 g/mL)  C  31  3.4.5 P1-PLC assay P1-PLC assays developed were based on cleaving p97 from the cell surface and 7 monitoring the cell surface p97 using the flow cytometry or measuring the solubilized p9 in the supernatant using the immunofluorescence assay.  3.4.5.1 P1-PLC assay based on flow cytometry  6 cells were washed with FACS buffer and incubated with Approximately 4 5x10 -  200 j.i.L of known and unknown samples of P1-PLC for 1 h at 37 °C. The cells were then washed and labeled with fluorescinated antibody according to the previously described method for preparing cells for analysis by flow cytometry. The fluorescence of the cells due to remaining cell surface p97 was measured using the flow cytometry (Figure 3.3) and the percentage removal of p97 calculated relative to positive (untreated cells labeled with fluorescinated antibody) and negative (untreated cells with no label) controls.  The  negative control gave the cell autofluorescence and the positive control gave the fluorescence due to autofluorescence plus cell surface p97. Percentage removal of p97 was calculated as follows:  % Removal ofp97  (positive control) (unknown) 1 100 4[(positive control) (negative control) J -  (3.2)  -  The percentage removal of p97 was then converted to PT-PLC concentration using a standard curve obtained from cells similarly treated with known enzyme concentrations.  —.  CD  C-)  CD  0  fIQ  + spent medium  5E6 cells 000 °°  spin and remove supernatant  200 ilL FACs Buffer wash, spin and remove supernatant  cells  /  ISP:  incubate at 37°C for 1 h.  (known or unknown sample) 200 tL of P1-PLC solution  PT-PLC ASSAY  ‘MMUNOASSJ  supemant  33  3.5.4.2 P1-PLC assay based on immunofluorescence  6 cells were washed with FACS buffer and treated with 200 iiL of Approximately 4 5x10 -  known and unknown P1-PLC samples (same as flow cytometry assay), for 1 h at 37 °C. After the incubation, the cells were removed by centrifuging and the cell free supernatant assayed for p97 concentration (Figure 3.3) as described in immunofluorescence assay section earlier. A standard curve of p97 in the supernatant against P1-PLC was obtained from samples treated with known enzyme solutions. Unknown P1-PLC activities were then obtained from the standard curve.  3.5 Experiments 3.5.1 Growth profiles for CHO cells 7 different types of serum-free media were investigated and compared to the growth profile of CHO-S-SFM I. monitored over a period of 8  Cell density, viability and glucose concentrations were -  12 days.  Media investigated are (i) Ham’s F12 with  insulin, transferrin, sodium selenite, 50 .tg/mL of BSA and 50 tg/mL DNase (ii) Ham’s F12 with insulin, transferrin, sodium selenite, 300 ig/mL of BSA and 50 .tg/mL DNase (iii)  DMEM!F12 with insulin, transferrin, sodium selenite, 50 ig/mL of BSA and  50 jtg/mL DNase  (iv)  DMEMJF12 and 50 .tg!mL DNase  (v)  HECHO  (vi) CHO-S-SFM II (vii) CHO-S-SFM II with 100 tg/mL DNase. CHO cells were grown in CHO-S-SFM I in 250 mL spinner flask to approximately 6 cells/mL. 2x10  25 mL of this culture was spun and resuspended in a mixture of two  2 T-flask. media (25 mL CHO-S-SFM I and 25 mL HBCHO) and maintained in a 75 cm The reason for using the mixture of 2 media was to adapt the cells to the new medium. Slowly, the medium was replaced by 100 % HBCHO.  150 mL spinner flask containing  34  HBCHO was then inoculated by cells from this T-flask and its growth profile monitored over a batch culture. Similar process was carried out to adapt the cells to Hams F12, DMEMJFI2 and CHO-S-SFM II.  3.5.2 Repeated harvesting 6 cells/mL. CHO cells were grown in CHO-S-SFM I in 250 mL spinner flasks to 4x10 25 mL of the culture (108 cells) in 50 mL centrifuge tube (Sarstedt, St. Laurent, PQ) was  spun and the cell pellet transferred to a 15 mL centrifuge tube (Sarstedt). The cells were washed with 5 mL PBS and then resuspended in 0.5 mL of PT-PLC solution. process, the same enzyme solution was used to harvest 5  -  In this  10 separate samples of cells.  Hence after 30 60 minutes incubation at 37 °C, the cells were spun and the supernatant -  used to treat a fresh sample of 10 cells.  A 20 50 iL sample was drawn after each -  harvest. These samples were spun and the cell free supernatant stored at 20 °C and later -  analysed for p97 using the immunofluorescence assay. For preliminary experiments, the volume of PT-PLC solution removed from sampling was replaced by PBS and for later experiments, it was replaced by concentrated P1-PLC (3000 300 mU/mL). -  This  experiment was carried out under sterile conditions.  3.5.3 p97 stability p97 harvested from CHO cells was diluted in immunofluorescence buffer to approximately 9.7 jig/mL. p97 solution was then sterile filtered by using 0.2 p.m membrane (VacuCap, Gelman Sciences). 10 eppendorfs (Sarstedt) with 1 mL of p97 solution were set up at 37 °C and one eppendorf removed at time 0, 1, 2, 3, 24 and 51 h and stored at 4 °C. These samples were then analysed for p97 concentration using the immunofluorescence assay.  a  a  CD  H  0.5mL P1-PLC Enzyme Icb  cells  Harvest# 1  ‘  LI i  mmiii  108  Wash  BIOREACTOR  Cells  Supernatant  REPEATED HARVESTING  To Harvest # 3  cells  Harvest#2  1!111111111  108  vi  36  3.5.4 P1-PLC stability a) At 37°C PT-PLC from the bacterial supernatant at 55 mU/mL was set-up in 1 mL eppendorf tubes and incubated at 37 °C. Microcentrilhige tubes were removed at different time intervals (0  -  14 days) and stored at 20 °C. The samples were then assayed for P1-PLC activity. -  b) pH 200 ILL of 16 mU/mL PT-PLC in PBS (1 mg/mL) buffer with pH of 6, 6.5, 7, 7.5 and 7.9 6 cells each. The cell surface fluorescence of the treated cells was was used to treat 4x10 measured using the flow cytometry and solubilized p97 concentrations in supernatant measured by immunofluorescence assay.  c) Freeze-thaw Since PT-PLC samples were stored at  -  20 °C and then thawed prior to use stability was  determined at 2 different dilutions (33 and 3.3 mU/mL). The enzyme was rapidly thawed in hot water (water bath at 37 °C) and frozen in the freezer  (- 20 °C).  stable. This implied that PT-PLC solution can be stored at  20 °C and thawed up to 10  -  The PT-PLC was  times without degradation In a separate experiment when PT-PLC samples were thawed in air slowly at room temperature, approximately 50 % of the activity was lost after each freeze-thaw cycle. Hence, it was important to thaw the enzyme very rapidly in hot water.  3.5.5 Adsorption of P1-PLC on cells CHO-S-SFM I was used to grow CHO cells in a 250 mL spinner flask to a density of 6 cells/mL. 4x10  25 mL of this culture was used for each sample.  The cell pellet of  37  approximately 108 cells was washed in 10 mL of PBS and then resuspended in I mL of varying concentrations (10  -  1500 mU/mL) of P1-PLC in PBS (1 mg/mL BSA) at 37 °C.  For preliminary experiments, P1-PLC in the supernatant was monitored over a period of 60 minutes.  Samples of 250 tL were drawn at different time intervals  (30 sec 60 mm). The first sample was taken as soon as the enzyme and the cells had -  been mixed. The mixture of enzyme and cells was shaken before drawing a sample, to Recovered  keep the concentration of the cells constant throughout the experiment. samples were spun at 2000 rpm for a minute and the supernatant stored at  -  20 °C. The  whole process of removing the first sample, centrifuging and storing took less than 3 minutes.  After 60 minutes, the cells were spun and washed twice in 1 mL of PBS  (1 mg/niL BSA). The cells were then resuspended in I mL of PBS and enzyme activity in the supernatant monitored over a period of 60 minutes. The samples were then assayed for PT-PLC activities. For later experiments, 108 cells were washed and incubated in 1 mL of PT-PLC solution at 37 °C.  After 15 minutes the cells were spun and the supernatant  assayed for P1-PLC activity. PT-PLC solution was incubated without cells at 37 °C and its activity monitored over a period of 60 minutes as a control. Each sample was analysed in duplicate at 3 different dilutions and the average taken as the representative value for PT-PLC activity.  3.5.6 Continuous harvesting 150 niL spinner flasks with 80 mL of CHO-S-SFM I medium were inoculated with 5 cells/mL from a common source (250 niL spinner flask). approximately 1x10 spinner flask was maintained without any P1-PLC as a control.  One  Two spinner flasks  contained 3 and 30 mU/rnL PT-PLC each. All 3 cultures were maintained for a period of 11 days (batch culture).  The cultures were sampled everyday to monitor the glucose  38  concentration, cell concentration and viability, cell surface p97 and soluble p97 concentration.  39  CHAPTER 4 Results and Discussions 41 Growth media for CHO cells CHO cells growing in suspension tend to aggregate at high cell densities. For this study it was important to maintain single cells so that the cell concentrations could be accurately determined.  Flow cytometry used to monitor cell surface protein also required single  cells, since large aggregates of cells block the flow tubes and make it impossible to analyse cells individually. Originally the transfected CHO cells were grown attached to the surfaces using Ham’s F12 supplemented with newborn calf serum (NCS). The presence of NCS causes the suspension cells to aggregate and adhere to the surface of tissue culture flasks. The suspension CHO cell line used in this work was selected (Kennard eta!., 1993) using CHO-S-SFM I, serum-free media. A culture of cells grown in CHO-S-SFM I was inoculated at lx  cells/mL and  grew to approximately 6x10 6 cells/mL of medium. The cells did not aggregate, even at the highest cell density (Figure 4.1).  CHO-S-SFM I is a proprietary medium whose  formulation is secret. A number of other serum-free media were screened in an attempt to find a less expensive alternative of known composition. Cells cultured in DMEMIF12 or Ham’s F12 media without serum formed aggregates, such that the cells could not be enumerated. The cell aggregates consisted of more than 20 50 cells each. -  Renner etal. (1993) found that aggregating of CHO  suspension cells at high densities was caused by DNA released from lysed cells and by adding 50  -  100 jigi’mL of DNase the cell aggregation could be eliminated.  Hence an  40  attempt was made to avoid the aggregating of the CHO cells by adding DNase to the media.  100  80  0 60  C•)  0) -  CD 40%.  20  0  Time (day) Figure 4.1 Growth profile of CHO cells growing in CHO-S-SFM I. The figure shows the cell concentration (A), glucose concentration (LI) and viability of the cells (V).  Ham’s F12 with DNase and 50 or 300 jig/mL of BSA were used to grow the CHO cells.  In the medium with 300 .tg/mL of BSA the cells grew to 1 .2x10 6 cells/mL  (Figure 4.2), while the cells in Ham’s F12 with 50 .tg/mL of BSA only grew to 8x1  cells/mL  (Figure 4.3).  The  cells  started  5 cells/mL. These aggregates consisted of 2 6x10  -  aggregating  at  approximately  5 cells each.  DMEMJF 12 containing only DNase or DMEMIF 12 with insulin, transferrin, sodium selenite, BSA and DNase resulted in cell densities of approximately 6 cells/mL of medium (Figures 4.4 and 4.5). 2x10 maintained  for  8  days  in  both  media.  The viability of above 90 % was  Cell  aggregates  of approximately  41  6 1.00x10  2.5  I  I  100 2.0 U)  5 7.50x10  80  0  1.5  C o  CU 500x10 5 C o  C  C 0  60-  ° U) CD  =  CD —  40  I-  o  .2 2.50x1  0.5  20  C) 0.00  0  2  I  4  I  6  I  8  I  10  .0  0.0 12  Time (day) Figure 4.2  5  Growth  ig/mL transferrin,  Figure cells  profile of CHO cells growing in Ham’s F 12 with 5 jig/mL of insulin, 10 nM sodium selenite, 50 jig/mL BSA and 50 jtg!mL DNase. The  shows the cell  concentration  (A),  glucose  concentration  (LI)  and  viability  of the  (V).  6 1.5x10  2.0  I  100 -J E  1.5  —5  o 1.OxlO  • 80  Q C  c o  60  0  1.00 CD  CU I.-  C C) o 5 C50x10 0• 0  —.  CD  40  -5  0  0.5 -  C) 0.0  0)  0  I  2  4  6  8  I  10  •0.0  20  Q  12  Time (day) Figure 4.3 Growth profile of CHO cells growing in Ham’s F12 with 5 jtg/mL of insulin, 5 p.g/mL transferrin, 10 nM sodium selenite, 300 p.g!mL BSA and 50 jig/mL DNase. The figure shows the cell concentration (A), glucose concentration (El) and viability of the cells  cv).  42  2.5x10  I  I  I  100  2  2.10  3  80  —  Cl)  Q  C-)  6 —‘i.5x10  C  C  60-  o CD  6 1.0x10  40  C)  I  C)  20  5 5.cc10  C-)  0  2  I  I  I  I  I  4  6  8  10  12  0  0  Time (day) Figure 4.4 Growth profile of CHO cells growing in DMEMIF12 with 50 j.ig/mL DNase. The figure shows the cell concentration (A), glucose concentration (Li) and the viability of the cells (V).  6 25x10  100 .1 06  C.)  ‘—‘1 5x10  3  80  Q  6  C  C) 2g CD  c.  C. .106  a)’  C) C 0 C.)  0  .4O  20  5 OxlO  0.0  0  2  I  I  I  I  I  4  6  8  10  12  -  0  Time (day) Figure 4.5 Growth profile of CHO cells growing in DMEMJF12 with 5 tg/mL of insulin, 5 jtg/mL transferrin, 10 nM sodium selenite, 50 jg/mL BSA and 50 ig/mL DNase. The figure shows the cell concentration (A), glucose concentration (LI) and viability of the cells (V).  43  5  -  10 cells  per  6 cells/mL). 1x10  aggregate  were  observed  at  higher  densities  (approximately  The increased cell growth in DMEM/F12 (Figure 4.5) compared to  Ham’s F12 (Figure 4.2) was probably due to the higher concentrations of nutrients in DMEMJF 12. Another medium that was investigated was HBCHO. In this experiment a cell 6 cells/mL was achieved (Figure 4.6). Besides attaining such a low density of about 1.5x10 density, the cells started clumping at approximately lxi 06 cells/mL. At 1. 5x 106 cells/mL, the viability of the cells had already dropped to 78 %.  Hence, this medium was not  suitable for our work.  100 1.5  -J  80  E C,)  CD C-)  0  C 0  1.0 C  .6O  C) 0 C,) CD  (U -I-,  C i) C-) C 0 C-)  CD  I  20  CD  0 0  .0  Time (day) Figure 4.6 Growth profile of CHO cells growing in HBCHO. This figure shows the cell concentration (A), glucose concentration (Li) and viability of the cells (V) for a batch culture.  Another medium that was investigated was CHO-S-SFM II. lower protein content and is also cheaper than CHO-S-SFM I. inoculated with approximately 6x1  This medium has  CHO-S-SFM II was  cells/mL and grew up to 3. 5x 106 cells/mL of medium  44  (Figure 4.7). In this medium, most cells existed as aggregates of 2 densities  (—  -  10 cells at high cell  6 cells/mL). Hence, an attempt was made to select for the single cells. 2x10  The aggregates of cells were allowed to settle down in the culture growing in the spinner flask and 10 mL of culture with single cells was decanted from the suspension and transferred to a T-flask.  Single cells that were transferred to a T-flask were grown to  6 cells/mL. approximately 2x10  At this point the cells had already started aggregating  again. Single cells were again transferred to a new T-flask and grown to high cell density and the process repeated. Although this process was repeated over five times, we did not obtain a population of cells which would not aggregate. To eliminate the problem of cell aggregating, 100 .tg/mL of DNase was added to CHO-S-SFM II.  This culture was  6 cells/mL of medium was achieved 5 cells/mL. A cell density of 7x10 inoculated at 2x10 and this culture maintained single cells throughout the batch run. (Figure 4.8).  100  80  Q 60  40  20  0  Time (day) Figure 4.7 Growth profile of CHO cells growing in CHO-S-SFM II. The figure shows the cell concentration (z\), glucose concentration (Li) and viability of the cells (V).  45  100  80  Cl)  ci  C) C C.) 0 0  60  CD  40 C)  20  C-)  0  Time (day) Figure 4.8 Growth profile of CHO cells growing in CHO-S-SFM II and 100 .tg/mL DNase. The figure shows the cell concentration (A), glucose concentration (Li) and viability of the cells (V).  Of the seven different serum-free media investigated, CHO-S-SFM II with 100 jig!mL of DNase and CHO-S-SFM I were able to achieve the highest cell density while maintaining single cells.  Since CHO-S-SFM Ills also a proprietary medium,  CHO-S-SFM I was used to grow the cells for all the investigations that followed. The experiments performed above show that DNA released into the media is not the only cause of aggregation, because addition of DNase to Ham!s F12 and DMEM]F12 did not completely eliminate the aggregating problem. However, aggregating of cells in CHO-S-SFM II was successfully eliminated by the addition of DNase. The effect of varying the initial cell density was also investigated (Appendix 2).  46  4.2 P1-PLC assay Since, available P1-PLC assays were too insensitive, too labour intensive and/or too expensive to perform routinely, it was necessary to develop a new assay.  The assay  developed was based on monitoring, in the presence of P1-PLC, either cell surface p97 or solubilized p97 in the supernatant.  4.2.1 Effect of harvesting conditions on p97 removal p97 harvest is dependent on the conditions prevailing in the harvesting process. Harvesting conditions were varied to find the suitable parameters for the P1-PLC assay. The parameters investigated were incubation time and the volume of PT-PLC solution per sample.  4.2.1.1 Incubation time  8 CHO Kennard et at. (1993) reported that in 10 mU/mL P1-PLC, the viability of i0 cells/mL did not decline up to 1 h. However, after 1 h the viability of the cells started dropping. Hence one hour incubation time was used as an upper limit. Figure 4.9 shows that p97 harvest increases with time, and that 60 minutes standard curve had the greatest range of p97 removal.  4.2.1.2 Enzyme volume  6 cells The effect of enzyme volume used per sample was investigated. Samples of 4x10 were treated with 50, 100, 200 and 400 p.L of 5 mU/mL of P1-PLC enzyme solution for  47  60  4o 0  2 I.— 0) 0.  P1-PLC (mU/mL) Figure 4.9 Effect of different incubation times on p97 harvest. 6 cells was treated with 200 iiL of PT-PLC at 37 °C. 4x10  Each sample of  1 h at 37 °C. Figure 4.10 shows that p97 harvest increases with P1-PLC volume (that is, increases with the ratio of PT-PLC to cells). At 50 !.iL of PT-PLC cleaved approximately iL of PT-PLC solution cleaved 28 % of the total 16 % of the cell surface p97, while 400 1 cell surface p97. Hence, efforts were made to keep a constant enzyme volume during each P1-PLC assay to ensure that the ratio of cells to enzyme is maintained. 6 cells Therefore for all PT-PLC assays, 200 p.L of PT-PLC enzyme solution, 4x10 and 1 h incubation time were used, unless otherwise specified.  4.2.2 P1-PLC assay based on flow cytometry The first assay developed was based on monitoring the cell surface p97, using flow cytometry (Section 3.4.5.1).  Enzyme solution (200 jiL) with a range of PT-PLC  6 cells at 37 °C for 1 h and cell surface p97 concentrations was used to treat 4x10  48  40-  •  •  •  30-  o  10-  •  0—  0  100  I  •  I  200  300  Pt-PLC volume  (p.  •  I  400  L)  6 cells was Figure 4.10 Effect of PT-PLC volume on p97 harvest. A sample of 4x10 treated with 50, 100, 200 and 400 p.L of PT-PLC solutions at 5 mU/mL. Incubation time of 1 hwas used.  measured, before and after the harvest. A standard curve of p97 removal from cells vs P1-PLC concentration was obtained with the range between 0.8 mU/mL and 80 mU/mL (Figure 4.11). Unknown P1-PLC samples were diluted to fall within this range and used to harvest cells in parallel with known standards. p97 removal of unknown samples was converted to PT-PLC activity using the standard curve obtained with the known standards. PT-PLC at 8 and 16 mU/mL were used to treat cells in triplicates. The standard error was less than 2 % showing that these results are reliable and reproducible (Figure 4.11). This assay was time consuming because of the need for repeated washing and centrifhging of the cells (Figure 3.1). Hence, an attempt was made to develop another PT-PLC assay.  49  iOo 80  60 40 20 Ui  0.1  1  10  100  P1-PLC (mU/mL) Figure 4.11 Typical standard curve obtained for P1-PLC at pH = 7.5 using 200 p.L of 6 cells per sample. A cell incubation time of 1 h PT-PLC solution and approximately 4x10 at 37 °C was used to obtain this curve.  4.2.3 P1-PLC assay based on immunofluorescence This assay made use of immunofluorescent analysis (Section 3.4.5.2) of soluble p97 concentration in the supernatant recovered from PT-PLC incubated with cells. p97 in the supernatant of the samples treated with known quantities of P1-PLC was used to obtain a standard curve of p97 recovered (jig/mL) against P1-PLC activity (mU/mL).  PT-PLC  activity of an unknown sample was determined from the concentration of p97 in the supernatant and the standard curve (Figure 4.12). Standard curves obtained from the assays based on flow cytometry and immunofluorescence were identical to each other (Figure 4.12). The immunofluorescence assay was less time consuming and involved less work.  50  12  • 100 10’ •80 _J  8  CD  E  •60  zi.  °  CD -,  •40 -20  2 .  •1  0.1  10  1  -0  100  P1-PLC (mU/mL) Figure 4.12 Typical standard curves obtained from assays based on flow cytometry and  immunofluorescence. The open triangles represents p97 concentration (A) and the open squares represents p97 removal (LI). 200 iiL of P1-PLC solution was used to treat 6 cells. Incubation time of I h at 37 °C was used. 4x10  4.2.4 Reliability of P1-PLC assay After developing two different assays, their reliability was investigated. Two unknown  PT-PLC samples at three different dilutions were tested and their average taken as the representative enzyme activity of that solution (Table 4.1). The actual PT-PLC activities of the samples according to Boehringer Mannheim were 6.2 mU/mL and 31 mU/mL respectively. The PT-PLC activities determined for the samples by flow cytometry assay were 6.23 and 34.2 mU/mL respectively.  The  immunofluorescence assay determined the samples to be 5.7 and 33.5 mU/mL respectively. PT-PLC activities obtained by both the assays were within ± 10 % of the actual values.  51 Table 4.1 Comparison of P1-PLC assay based on flow cytometry and immunofluorescence assay. Each sample was assayed in duplicate at 3 different concentrations.  Sample  Flow Cytometry Measured Calculated P1-PLC PT-PLC  Measured  Calculated  PT-PLC  P1-PLC  [mU!mLl  [mU/mLl  [mU/mLl  [mU/mLl  4  1.8  7.2  1.30  5.20  2  3.0  6  3.20  6.40  1  5.5  5.5  5.40  5.40  40  0.9  36  1.20  30.00  20  1.63  32.5  1.75  35.00  34  3.55  35.50  Dilution in PBS  1  2  10  -  Immunofluorescence  4.3 Repeated harvesting The first production process to be investigated was the repeated harvesting technique. The repeated harvesting process involves re-using the PT-PLC enzyme solution for harvesting multiple samples of cells in an attempt to further increase the product concentration. In the first experiment, 7 samples of 1  cells were harvested consecutively with  0.5 mL of 30 mU/mL of P1-PLC in PBS with 1 mg/mL BSA (Figure 4.13). After each sampling, 20 jiL of PBS (1 mg/mL BSA) was added to the harvesting solution to replace the sample volume removed to maintain a constant volume.  The cumulative p97  recovered increased for the first two harvests and then remained approximately constant. Figure 4.13 also shows the corrected values, which take into account the p97 dilution due to sampling. There also may have been dilution effects due to the liquid associated with the cell pellet. The p97 recovered from the first harvest was approximately 59.7 tg!mL. Hence, after 7 harvests theoretically 7-fold more or 417.8 tg/mL p97 could have been  52  recovered. However, there was only a 2-fold increase in the p97 concentration in the product.  These results indicated that there was loss of either P1-PLC or p97 in the  successive harvests.  -J  E  3  4  5  Harvest (#)  Figure 4.13 Concentration of p97 recovered for 5 repeated harvests. Each harvest was 8 cells for 30 minutes. P1-PLC enzyme (0.5 mL) at an initial conducted with i0 concentration of 30 mU/mL was used. Open circles (0) represent the measured values and the solid circles (.) represent calculated values which account for sampling dilution.  6 cells was washed in PBS and After each harvest, a sample of approximately 4x10 the cell surface p97 measured by flow cytometry.  p97 removal from the cell surface  decreased from 60 % to 20 % over 7 harvests (Figure 4.14). Although the recovered p97 concentration was not as high as expected, the 2-fold increase in p97 showed that the technique had potential.  Further investigations were  carried out to determine the reasons for not achieving the predicted amount of p97.  53 100-  II•I•I•II’  1  8060-  40-  20-  0-  I  .  0  1  2  3  4  5  6  7  Harvest # Approximately 6 cell sample was taken from the cell pellet and labeled with the fluorescent antibody. 4x10 Figure 4.14 Percent of p97 removal from the cells after each harvest.  4.3.1 p97 stability The p97 stability was investigated to determine if degradation of the protein could account for part of the decreased harvest yield. Since, the harvesting experiments were done at 37 °C, we investigated p97 stability at that temperature.  Firstly, purified p97  (Kennard et at, 1993) stability was studied in immunoassay buffer (storage buffer). Ten p97 harvests of 1 h each take approximately 16 h, however p97 was found to be stable in this buffer over a period of 48 h (Figure 4.15). The next experiment was done under harvesting conditions, that is in the presence of PT-PLC and possible proteases released from cell lysis. Two samples of p97 harvested from cells were incubated at 37 °C for 4 and 24 h. There was slight decline of p97 over a period of 24 h (Table 4.2). However, the decline in p97 was not enough to account for the insufficient harvesting. Hence, p97 degradation was not the cause for the low levels obtained from repeated harvesting.  54  20  .i  •  18 16 14 12  Ii  1o.  8 0) Q-  6 4.  2 0•  I  •..i  10  1  Time (h) Figure 4.15 p97 stability at 37°C for a period of 48 h. Initial p97 concentration was 9.65 ig/mL in the immunoassay buffer. The error bars represent the standard error mean. Table 4.2 Stability of p97 in the harvesting solution of PT-PLC in PBS (1 mg/mL BSA). The samples were assayed at 3 different dilutions. The error bars were less than ± 1 5 ig/mL. Time [h] 4 24  Initial Concentration [jig/mL]  Final Concentration [ig/mL]  26.4 23.2  23.3 21.3  4.3.2 P1-PLC Stability The PT-PLC enzyme stability was studied to determine if the loss of activity could explain the repeated harvesting results. P1-PLC at 52 mU/mL was incubated at 37 °C for a period of 2 weeks and samples drawn at different time intervals. These samples were stored at 4 °C and then analyzed for PT-PLC activity. PT-PLC proved to be stable over 14 day  55  period (Figure 4.16). Griffith eta!. (1991) also found that PT-PLC was stable at 37 °C for prolonged periods (they did not report the exact time).  (U.  600  0  0  0  E  P.  o -J  30-  0 ..L 200. 10O•  .......  0.01  •uI11•uI1  111111111  0.1  1  111111111  I  10  Time (day) Figure 4.16 P1-PLC stability at 37 °C. PT-PLC from the bacterial supernatant at 52 mU/mL was used. 200 p.L of two dilutions of each sample were used to treat 6 cells for 1 h at 37 °C. 4x10  The pH of the harvesting solution after 10 repeated harvests of adherent CHO cells on porous microcarriers falls to approximately 6.5 (Kennard, personal communication). Hence, effect of pH on PT-PLC activity was investigated. The volumes of PT-PLC used for the harvests in this work were too small (0.5 mL) to monitor pH routinely during the actual harvesting process. Therefore, PT-PLC enzyme diluted in PBS (1 mg/mL BSA) at 6 cells (Figure 4.17). pH = 6.0, 6.5, 7.0, 7.5 and 7.9 was used to treat 4x10  The  percentage removal of p97 based on flow cytometry analysis of cell surface p97 increased with decreasing pH. Therefore, changes in pH do effect PT-PLC activity, however it does not inactivate it. According to Ikezawa and Taguchi (1981), the optimum pH is between 5  -  8.5, with maximal activity at 7.5. However that trend did not agree with this work, in  56  contrast PT-PLC activity increased monotonically from 7.9 6 pH. Insufficient harvesting -  can not be attributed to decreases in pH. Supernatant from these harvests were then analysed for p97. Figure 4.17 shows that p97 recovered at 16 mU/mL of PT-PLC solution also decreased with increasing pH from 6.0 8.0. This trend agrees with that of p97 removal from the cell surface (Figure -  4.17). Therefore, the change in pH of the harvesting solution was not the cause for the repeated harvesting results.  100  I  I  I  80  16  12  —  6O >  40  3 r  0)  20-  0  6.0  I  I  I  I  6.5  7.0  7.5  8.0  pH Figure 4.17 Effect of pH on P1-PLC activity and p97. P1-PLC at 16 mU/mL was used to 6 cells for 1 h at 37 °C. PBS buffer with 1 mg/mL of BSA at pH = 6.0, 6.5, harvest 4x10 7.0, 7.5 and 7.9 was used for these harvests. The open squares (E) stand for p97 concentration and the open triangles (ba) stand for p97 removal.  4.3.3 P1-PLC adsorption An effort was made to investigate the loss of P1-PLC enzyme due to adsorption to the cell surface. Phospholipases are a special type of membrane interacting proteins. According  57  2 (PLA ) adsorbs to the cell surface 2 to Hendrickson and Dennis (1984), phospholipase A before reacting with the substrate.  In the repeated harvesting process the cells were  removed from the harvesting solution and replaced with a fresh set of cells after each harvest, any PT-PLC adsorbed to the cell surface would be lost at each harvest. Since, it was not possible to measure the P1-PLC adsorbed to the cell surface, PT-PLC changes in the supernatant were monitored over time and PT-PLC adsorbed to the cell surface calculated as follows:  [PIPLCJa  =  0 [PTPLC]  -  (4,1)  [PI”PLC]s  where [PI-PLC]a  =  P1-PLC adsorbed on the cell surface at time t  0 [P1-PLC]  =  Initial P1-PLC concentration in the solution  5 [P1-PLC]  =  P1-PLC concentration in the supernatant after time t  The first experiment was performed to monitor the change in PT-PLC concentration over a period of 1 h in the mixture of 18 mUImL enzyme and 108 cells/mL. Since preliminary repeated harvesting process used 108 cells per harvest, the same number of cells were used for the adsorption experiments.  The cells and enzyme were mixed  thoroughly before removing a sample, so as to maintain a constant cell concentration in the mixture. P1-PLC activity decreased rapidly to 4.2 mUImL within 3 minutes and then remained constant over a period of 60 minutes (Figure 4.18). The decrease of PT-PLC could be attributed to the rapid binding of the protein to the cell surface.  Although  samples were drawn at 30 s, the cells were centriftiged and separated from the supernatant within 3 minutes. Thus the first data point is reported at 3 minutes in Figure 4.18. As a control, P1-PLC solution without any cells was incubated at 37 °C for a period of I h. No change in enzyme activity was observed over the period of 1 h (Figure 4.18).  58  The cells were then washed twice with 2 mL of PBS and resuspended in 1 mL of PBS and incubated at 37 °C for an hour to investigate the desorption of P1-PLC from the cell surface. PT-PLC activities in the wash solutions I and II were 2.4 and 0.8 mU/mL respectively.  -J  D C) -J  0  40  Time (mm) Figure 4.18 Adsorption of P1-PLC (0) to the cell surface over a period of 60 minutes. 108 cells/mL of PT-PLC enzyme at 18 mUImL. A control on adsorption (•) experiment was done with P1-PLC without cells incubated at 37 °C for 1 h.  The ratio of volume of cells to liquid for desorption was lower than that for adsorption experiment because of the loss of cells due to sampling.  Desorption  experiment was done over a period of 2 h to investigate the effect of increased incubation time on the equilibrium between PT-PLC in solution and P1-PLC on the cell surface. In this case PT-PLC activity in the supernatant increased rapidly (within 30 sec) and then remained constant (Figure 4.19). Figure 4.19 shows p97 removal on the vertical axis instead of the P1-PLC activity, because PT-PLC activity in the desorption solution was below the linear range (less than 1 mU/mL P1-PLC) on the standard curve. Hence it was not possible to convert percentage removal of p97 to PT-PLC activities. Figures 4.18 and  59  4.19 show that adsorption and desorption of P1-PLC was at equilibrium within less than 3 minutes.  [0  0)  5  60  80  Time (mm) Figure 4.19 Desorption of P1-PLC from the cell surface (0) in 1 mL of PBS. PT-PLC activity is less than 1 mU/mL. The cell concentration of 5x10 7 cells/mL was used. Study of PT-PLC desorption was also important for the controlled release cyclic harvesting technique developed by Kennard et a!. (1993). The cyclic harvesting process involves recycling the harvested cells into the growth medium for re-expression of p97 (Figure 3.4). Presence of residual PT-PLC in the growth medium would cleave p97 during the growth cycle resulting in loss of protein to the growth medium. Since, PT-PLC will rapidly reach an equilibrium between the cells and the supernatant, two or three quick washes will remove most of the PT-PLC from the cells surface. Leaving the cells in wash solution for 5  -  10 minutes will not help the desorption process, but using large volume of  wash solutions will remove more p97 from the cell surface.  60  Enzyme activities of the samples from adsorption experiment were analyzed at 3 different dilutions and their average taken as the representative value for the activity of each sample. The adsorption experiment was repeated with different enzyme concentrations to investigate whether adsorption resulted in saturation or equilibrium. In case of saturation a fixed amount of P1-PLC would be adsorbed to the cell surface, while for equilibrium a proportional amount to initial PT-PLC concentration will be adsorbed.  Using the data  from Table 4.3 and equation 4.1, an equilibrium curve was established (Figure 4.20).  Table 4.3 P1-PLC change in the supernatant after 15 mm. 1 ü cells/mL of PT-PLC solutions at different concentrations were incubated at 37 °C.  P1-PLC at time = 0  P1-PLC at time = 15 mm  {mU/mL]  [mU/mU  2.50  0.95  9.40  5.50  11.1  9.50  38.0  15.5  80.0  30.0  128  37,5  560  156  P1-PLC added to the cells resulted into an equilibrium between PT-PLC on the cell surface and PT-PLC in solution.  Figure 4.20 shows that PT-PLC lost due to cell  adsorption was directly proportional to the enzyme concentration added to the cell pellet up to approximately 150 mU/mL. The horizontal axis on Figure 4.20 shows PT-PLC in solution after adsorption.  61  6 5x10  D  4x1O  6 3x10  -  6 2x10  Cu  0 -J 6 0... 1x10  0 0 50  75  100  125  P1-PLC in solution (mU/mL) Figure 4.20 Adsorption equilibrium of PT-PLC at 37 °C. concentrations at cells/rnL were used to study adsorption.  175  Different enzymes  4.3.4 Non-specific binding of PT-PLC Since PT-PLC selectively cleaves the GPI anchor it may be that the PT-PLC specifically binds to the anchor of recombinant proteins.  Therefore, an attempt was made to  investigate whether P1-PLC adsorption was due to specific or non-specific binding. This was done by studying P1-PLC adsorption on untransfected CHO cells (CHOWTB  -  Food et a!, 1994) in comparison to the genetically engineered cells used in this study (Table 4.4). Table 4.4 shows that both the cell lines adsorbed 1 .8x10 7 mU P1-PLC per cell, hence protein binding was not increased by the recombinant p97 on the surface of the transfected cells. In addition literature reviewed on specific binding of manganese (Suárez and Eriksson, 1993) and lactoferrin (Maneva etal., 1993) show that binding takes place slowly over a period of 30 60 minutes. -  surface was most probably non-specific.  Therefore, rapid binding of PT-PLC to cell  62  Table 4.4 Comparison for transfected and untransfected (CHOWTB) CHO cells. Incubation time of 15 mm and I mL of PT-PLC solution was used.  CHO Cell line  Cell concentration [cells/mL]  P1-PLC at 0 mm [mU/mU  PT-PLC at 15 mm [mUImL]  P1-PLC adsorbed [mU/cell]  Untransfected  7 5.5x10  32  22  7 1.82x10  Transfected  8 1x10  32  14  7 1.8x10  To improve the repeated harvesting process, one option was to add excess PT-PLC so that even after the losses due to adsorption, there is still enough P1-PLC to cleave approximately 80 90 % of the p97 from the cell surface. However, this will result -  in additional cost for P1-PLC and also increase contamination of the product by the enzyme. Hence as an alternative, the repeated harvesting process was started with low enzyme concentration and PT-PLC replenished. A repeated harvesting process was then carried out for 10 consecutive harvests with enzyme replenished after the 5th and 6th harvest to confirm the loss of P1-PLC due to adsorption. The first harvest was started with 0.5 mL of 300 mU/mL of P1-PLC and 25 iL of 3000 mU/mL PT-PLC was added to the harvesting solution after the 5th and 6th harvest (Figure 4.21) to replenish the lost PT-PLC.  The reason for using a small PT-PLC volume at high concentration for  replenishment was to minimize the dilution of the protein product. Figure 4.21 shows that p97 in the product increased 4-fold from approximately 72 .ig/mL in the first harvest to 290 tg/mU. The decrease in p97 concentrations at the 3rd and 9th harvest are believed to be due to the dilution effect due to the fluid associated  63  -J  E  8 F.— 0) 0  0  2  4  6  8  10  Harvest Number Figure 4.21 Cumulative p97 from repeated harvesting with P1-PLC replenishment. The initial harvesting was started with 0.5 mL of 300 mU/mL of PT-PLC and then 25 jiL of P1-PLC enzyme at 3000 mU/mL was added after the 5th and 6th harvests. Incubation time of 30 minutes at 37 °C was used for each harvest.  100  80 Cu  2  60 40  0) 2O  0  2  I  I  4  6  8  10  Harvest Number Figure 4.22 p 97 removal The initial harvesting was 25 iL of P1-PLC enzyme Incubation of 30 minutes at  from the cell surface during the repeated harvesting process started with 0.5 mL of 300 mU/mL of P1-PLC and then at 3000 mU/mL was added after the 5th and 6th harvests. 37 °C was used for each harvest.  64  to the cell pellet. Figure 4.22 shows that adding PT-PLC enzyme after the 5th and 6th harvests restored the percentage removal of p97 from the cell surface. Hence Figures 4.21 and 4.22 confirm that insufficient harvesting of p97 during the repeated harvesting process was caused by the loss of P1-PLC due to adsorption to the cell surface when the harvested cells were removed. Having determined that PT-PLC adsorption to the cell surface had caused reduced harvesting of p97, the next repeated harvesting experiment was carried out using 2 different initial concentrations (300 and 30 mU/mL) and enzyme supplemented at each harvest. 50 tL of the harvesting solution was drawn after each harvest for p97 analysis and replaced by 50 j.L of concentrated enzyme (Figures 4.23 and 4.24). The repeated harvesting experiment that was started with 300 mU/mL PT-PLC was supplemented with 50 p.L (approximately 300 mUImL total volume) of 3000 mU/mL PT-PLC after each harvest.  Similarly the one that was started with 30 mUImL P1-PLC was supplemented  with 50 p.L (approximately 30 mU/mL) of 300 mU/mL PT-PLC. Figure 4.23 shows that repeated harvesting process was successful and p97 concentration was increased approximately 4-fold with five harvests.  In the case of  industrial production, reduced volumes of concentrated PT-PLC can be added to minimize the dilution of the protein in the product. Also there is no need to remove a sample from the harvesting solution and this will prevent loss of p97 due to sampling. The calculated values on Figure 4.24, only account for p97 lost by sampling, it does not account for the dilution effects caused by the liquid associated with the cells. Cumulative p97 recovered from 300 and 30 mU/mL does not differ considerably. Hence, it is advisable to use lower concentrations of PT-PLC. Using less PT-PLC results in less production costs and also purer product.  65  120’ 100’  80’ &60 40  1C)  a 20  0  I  I  I  I  I  1  2  3  4  5  Harvest Number Figure 4.23 Percent p97 removal from the cell surface for 5 harvests of 108 cells was monitored. Harvesting solution of 300 (0) and 30 (z) mU/mL were used. Incubation of 1 h at 37 °C was used for each harvest.  I  I  I  I  350-J —  ‘-  300-  250  0 -20O 150  8  100  C)  50  a  0 0  I  I  I  I  I  1  2  3  4  5  Harvest Number Figure 4.24 Open symbols represents cumulative p97 measured after each harvest and  the solid symbols represents the calculated values of the p97 harvest which includes p97 lost due to sampling. This figure shows p97 recovered from 300 (0) and 30 (A) mU/mL of PT-PLC. Incubation of 1 h at 37 °C was used for each harvest.  66  4.4 Continuous harvesting The second harvesting process investigated was a continuous harvesting technique. This technique involved harvesting of p97 into the growth media and a batch run over a period of 11 days. This production process is similar to the typical protein production process by mammalian cells where the desired protein is secreted into the growth media.  An  advantage of this harvesting process was the reduction in p97 losses due to membrane turnover.  Approximately 1  -  3 % of the cell membrane is internalized per minute  (Steinman eta!., 1976; Burgess and Kelly, 1987).  This phenomena resulted in  internalization and possibly degradation of the desired membrane protein, resulting in decreased product recovery. Theoretically, PT-PLC in the growth media will cleave p97 as soon as it appears on the cell surface preventing this loss. Therefore media containing three different PT-PLC concentrations of 0, 3 and 30 mU/mL PT-PLC were inoculated with approximately 2x10 5 cells/mL. Cell viability (Figure 4.25), cell surface p97 (Figure 4.27), glucose concentration (Figure 4.26) and p97 in the supernatant (Figure 4.28) were monitored over an 11 day period. Figure 4.25 shows that PT-PLC in the growth media does not have an effect on cell concentration or the viability of the cells in the culture. The glucose consumption was also approximately the same for all three cultures (Figure 4.26). However, cell surface fluorescence of cells growing in the cultures with 3 and 30 mU/mL decreases rapidly, while the cells growing in a culture with no (0 mU/mL) PT-PLC decreases steadily (Figure 4.27) as Kennard eta!. (1993) reported over a batch culture.  67  4.0x10’  I  I  -100  I  •  -J =  -80  6.  -60  o  1)  2.0x106 c  8  6 1.0x10  -20  a)  C-) 0.00  I  I  4  6  I  •  I  •  8  10  -0 12  Time (day) Figure 4.25 Continuous harvesting of CHO cells in the growth media. The solid symbols represent cell concentration and the open symbols stand for viabilities. Cultures with PT-PLC at 0 (El) 3 (A) and 30 (0) mU/mL of medium.  I  I  •  I  •  I  •  4. -J  E3  I: •  0  I  2  •  I  4  •  I  •  6  I  8  •  I  10  12  Time (day) Figure 4.26 Glucose concentrations during the continuous harvesting process with 0 (LI), 3 (A) and 30 (•) mUIrnL of PT-PLC enzyme.  68  I  •  I  •  I  •  I  •  I  •  I  •  300  —  225  8C G)  150  0 D LL.  0  •  0  4  2  I  10  8  •  12  Time (day) Figure 4.27 Fluorescence due to cell surface p97 during the continuous harvesting process with 0 (El), 3 (A) and 30 (0) mU/mL of P1-PLC enzyme.  I  I  I  •  I  I  •  4  6  •  I  •  16 -J  14 0)  o  10  C  86 C  84  N 0•)  0 0  o  •  2  I  8  10  12  Time (day)  Figure 4.28 Cumulative p97 in the supernatant from the continuous harvesting process with PT-PLC at 0 (LI), 3 (A) and 30 (0) mU/mL of medium.  69  Figure 4.28 shows that approximately 15 tg/mL of p97 was recovered in the culture with 30 mU/mL P1-PLC. A control culture with no P1-PLC enzyme produced approximately 4 ig/mL of p97. Membrane bound p97 may have been released in the supernatant due to cell lysis. Kennard et a?. (1993) and Food et a?. (1994) reported that transformed CHO cell line used in this study also produces a secreted form of p97. Hence p97 recovered in the spent media was the result of the secreted and membrane bound p97. According to Figure 4.27 and the calculations, based on theoretical values, we have harvested all the p97 present on the cell surface by using 30 mU/mL P1-PLC. The calculation for approximate p97 production in to the media was done by the following equation:  d[p97]  =  (4.2)  dt where r  =  average rate of p97 production (jig/cell-h) number of cells  x t  =  time (h)  Integrating equation 4.3, we get: [p97]=rt where x  =  (4,3)  average number of cells  According to Kennard eta?. (1993), approximately 6x10 7 jig of p97 is expressed per cell within 48 h after P1-PLC treatment. Therefore,  r  =  [p97 per cell]  =  [6x1o Lg/cell] =  8 tg!cell-h L25x10  (4.4)  For a batch culture of approximately 11 days and an average of 3x10 6 cells/mL, substituting equation 4.4, we get:  70  [p97]  =  8 p.g/cellIh) x (3x10 (1.25x10 6 cells/mL) x (11 day) x (24 h/day). (4.5)  =  9.9 ig/mL  Total p97 attained experimentally is 15 .tg/mL, of which 4 jig/mL (0 mU/mL in Figure 4.28) is the secreted form of p97. Therefore, p97 produced due to PT-PLC action is: 15 -4  =  11 jig/mL  Taking into account the errors due to p97 analysis, p97 released from the cells due to cell lysis and the loses recovered due membrane turnover rate, we have harvested the majority of p97 that was expressed on the cell surface.  4.4 Comparison of protein production processes Three harvesting processes, cyclic (Kennard eta!., 1993), repeated and continuous harvesting techniques were compared based on their p97 productivity, yield on glucose consumed, PT-PLC used and p97 purity (Table 4.5).  The data for cyclic harvesting  process has been estimated from the work of Kennard eta!. (1993), based on harvesting 108 cells in 0.5 mL P1-PLC solution. The comparison was based on p97 harvests over 8 day period in a 250 mL bioreactor. The continuous harvesting process attained a yield of 3.4 jig p97 per mg of glucose consumed.  It also attained the highest p97 cell specific productivity  7 jig/cell-day) (Table 4.5). However, it achieved the lowest p97 concentration of (5.75x10 the 3 production processes. The repeated harvesting process carried out by using 30 mU/mL of PI-PL produced 295 jig/mL p97, while the one carried out with 300 mU/mL of PT-PLC produced 343 j.tg/mL p97.  Hence, it is not advantageous to increase the enzyme  concentration lox to increase the protein in the product by approximately 8 %. As we know from the adsorption experiments that P1-PLC lost to the cell surface will be proportional to the amount added.  Therefore, harvesting with the lower PT-PLC  71  concentrations (30 mU/mL or less) is recommended. Cyclic and repeated harvesting are very similar except with repeated harvesting the PT-PLC volume is reduced and p97 concentration is increased.  Table 4.5 Comparison of p97 harvests from different harvesting processes  Harvesting process  Cyclic 5 harvests  p97  Average p97/cell per day  p97 yield on PT-PLC  p 9 7 yield on glucose  Volume of the harvesting solution  Estimated Purity  [p.g/mL]  [tg/cell-day]  [jj.g/mU]  [ig/mgJ  [mU  [%]  62  7 1.93x10  2.07  0.39  25  30  343  7 2.14x10  0.23  0.43  5  30  295  7 1.84x10  1.97  0.37  5  30  11.5  7 5.75x10  0.38  3.40  250  3.6  30 mU/mL Repeated 5 harvests 300 mU/mL Rep eated 5 harvests 30 mU/mL Continuous lbatchculture 30 mU/mL  For calculations for Table 4.5, see appendix 9  In addition to the dilute and impure product obtained from continuous harvesting process, the quantity of P1-PLC required was high. To harvest 2875 tg of p97, 7500 mU of PT-PLC were required (ratio of 2.6:1).  The repeated harvesting process recovered  72  1474 p.g of p97 using 750 mU of PT-PLC (ratio of 0.5:1). Therefore, the yield of p97 on PT-PLC was higher for repeated harvesting process than for the continuous harvesting process. Hence, the question is to choose a harvesting process with high yield or high protein concentration.  This will depend on the relative magnitudes of the production  (including enzyme cost) and the downstream processing costs.  However, the major  motivation for this work was to achieve high initial purity and concentration to decrease extreme downstream processing. Therefore the recommended production process is the repeated harvesting technique.  73  CHAPTER 5 Conclusions The growth of CHO cells was investigated in CHO-S-SFM I, CHO-S-SFM II, HBCHO, DMEMJF12 and Ham’s F12.  CHO cells have a tendency to aggregate at high cell  densities, making it difficult to determine cell surface p97 expression. DNA released from the lysed cells triggers cell aggregation at high densities. DNase was added to these media to reduce aggregation. Addition of DNase to HBCHO, DMEMIF12 and Ham’s F12 did not  eliminate  cell  aggregation.  However,  addition  of  DNase  to  CHO-S-SFM II eliminated the aggregation problem altogether. The repeated harvesting process involves re-using the PT-PLC enzyme solution to consecutively harvest multiple samples of cells. recovered from 7 harvests of 108 cells.  Initially only 139 jig/mL p97 was  The first harvest recovered approximately  60 .tg/mL, therefore theoretically 7 harvests should have recovered 420 ig/mL. In an attempt to understand the low product recovery, the stability of p97 and P1-PLC were studied. Whereas the repeated harvesting process was complete within 16 h, p97 was found to be stable at 37 °C for a period of over 24 h. Since a suitable assay for quantifying P1-PLC was not available, PT-PLC assays based on flow cytometry and immunofluorescence analysis of p97 were developed. PT-PLC activity was found to be stable at 37 °C for a period of 14 days. Thus the low product harvests were not due to p97 or PT-PLC degradation. Loss of PT-PLC due to adsorption to the cell surface was investigated.  Cell  surface adsorption and desorption of PT-PLC was found to take place in less than 3 minutes.  An equilibrium was established between the P1-PLC in the solution and  PT-PLC adsorbed to the cell surface. Since the repeated harvesting process re-uses the  74  PT-PLC solution, adsorption to cell surfaces resulted in a loss of enzyme upon cell removal after each harvest.  This depleted the P1-PLC enzyme in the solution and decreased  subsequent p97 harvests. Repeated harvesting experiments were then done with P1-PLC replenishment after each harvest to maintain the PT-PLC concentration.  From 5 consecutive harvests of  1O cells a 4-fold increase in p97 concentration in the product was achieved. A continuous harvesting process also was investigated.  This process involved  harvesting of p97 in the cell growth medium at the same time as protein expression. From a batch culture 15.3 .tg/mL of p97 was recovered. Three production processes: cyclic, repeated and continuous harvesting were compared.  The greatest yield on glucose was obtained by the continuous harvesting  process (3.4 p.g p97/mg glucose), while the greatest yield on PT-PLC was obtained by the cyclic harvesting process (2 .tg p97/mg glucose). The continuous harvesting process also achieved the highest p97 productivity per cell (5.75x10 7 tgIcell-day), approximately 2fold higher than the cyclic and repeated harvesting processes. Increased productivity of the CHO cells in the continuous harvesting process was probably due to reduced losses due to the membrane turnover. The repeated harvesting process achieved the highest p97 concentration and its purity was comparable to that of cyclic harvesting process.  75  CHAPTER 6 Future Work 1, In this work P1-PLC adsorption to the cell surface was studied up to approximately 150 mU/mL to cover the range of enzyme concentration used to harvest the CHO cells in these production processes.  It would be interesting to develop a complete adsorption  isotherm to investigate the saturation of P1-PLC on cell surface.  2. Different GPI-anchored proteins have been reported to respond differently to cleavage by PT-PLC. Hence, it is important to apply this technique to other GPI-anchored proteins and determine if the p97 model system is representative of other proteins.  3. It would be worthwhile to do an economic analysis to compare the relative benefits of the 3 controlled release processes.  4. The controlled release harvesting process is laborious and involves multiple washing and centrifuging. These problems could be reduced by designing a specialized bioreactor to accommodate all steps of the harvesting process.  76  Abbreviations ADP  Adenosine Diphosphate  ATP  Adenosine Triphosphate  CHO  Chinese Hamster Ovary  GPI  Glycosyl Phosphatidylinositol  NAD  Nicotinamide Adenine Dinucleotide (oxidized)  NADH  Nicotinamide Adenine Dinucleotide (reduced)  NCS  Newborn Calf Serum  p97  Melanotransferrin  PBS  Phosphate Buffered Saline  PEP  Phosphoethanol Pyruvate  P1  Phosphatidylinositol  P1-PLC  Phosphatidylinositol Phospholipase C  77  References Bailey J.E., Ollis, D.F., Biochemical engineeringfundamentals, 2nd ed., McGraw Hill Publishing Co., 1986. Baker, E.N., Baker, H.M., Smith, C.A., Stebbins, M.R., Kahn, M., HelistrOm, K.E., Hellstrom, I., “Human melanotransferrin (p9’7) has only one functional iron-binding site”, FEBsLett., 298(2,3): 215-218 (1992). Baker, E.N., Rumball, S. V., Anderson, B .F., “Transferrins: Insight into structure and functions from studies on lactoferrin” Trends Biochem. Sci., 12: 3 50-353 (1987). Berridge, M.J., “Inositol lipids and cell proliferation”, Biochim. Biophys. Acta, 907: 33-45 (1987). Boehringer Mannheim Biochemica, Bulletin No. 0189, B 635.5, 1144 286. Boivin, P., Delaunay, I., “La membrane du globule rouge”, Xle congrès de la société Francaise d’Hématologie, 3(1), 125-128 (1991). Bordier, C., Etges, R.J., Ward, J., Turner, M.J., Cardoso de Almeida, M.L., “Leishmania and Trypanosoma surface glycoproteins have a common phospholipid membrane anchor”, Proc. Nat!. Acad. Sd., 83(15): 5988-5991 (1986). Brown, J.P., Hewick, R.M., HellstrOm, I., HelistrOm, K.E., Doolittle, R.F., Dreyer, W.J., “Human-melanoma associated antigen p97 is structurally and functionally related to transferrin”, Nature, 296: 171-173 (1982). Brown, J.P., Niyashama, K., HellstrOm, I., HellstrOm, K.E., “Structural characterization of human-melanoma-associated antigen p97 with monoclonal antibodies”, I ImmunoL, 127: 539-546 (1981a). Brown, J.P., Woodbury, R.G., Hart, C.E., HellstrOm, I., HellstrOm, K.E., “Quantitative analysis of melanoma associated antigen p97 in normal and neoplastic tissues”, Proc Nat!. Acad. Sci. USA. 78: 539-543 (1981b). Burgess, T.L., Kelly, R.B., “Constitutive and regulated secretion of proteins”, Ann. Rei Ce!! Biol., 3: 243-293 (1987). Butler, M., Animal cell technology: Principles and Products, Open University Press, 1987, U.S.A.  78  Caras, I.W., Davitz, M.A., Rhee, L., Weddell, G., Martin, D.W., Nusseinzweig, V., “Cloning of decay accelerating suggests novel use of splicing to generate two proteins”, Nature, 325: 545-548 (1987a). Caras, I.W., Weddell, G., Davitz, MA,, Nusseinzweig, V., Martin, D.W., “Signal for attachment of a phospholipid membrane anchor in decay accelerating factor”, Science, 238: 1280-1283 (1987b). Caras, I.W., Weddell, G., “Signal peptide for proteins secretion directing glycophospholipid membrane anchor attachment”, Science, 243: 1196-1198 (1989). Cross G.A.M., “Glycolipid anchoring of plasma membrane proteins”, Ann. Rev. Cell Biol., 6: 1-39 (1990). Davitz, M.A., Low, M.G., Nusseinzweig, V., “Release of decay-accelerating factor (DAF) from the cell membrane by phosphatidylinositol-specific phospholipase C (PT-PLC). Selective modification of a complement regulatory protein”, I Exp. Med., 163: 1150-1161 (1986). Davitz, M.A., Gurnett, A.M., Low, M.G., Turner, M.J., Nussein.zweig, V., “Decayaccelerating factor (DAF) shares a common carbohydrate determinant with the surface glycoprotein (VSG) of the African Trypanosoma brucei”, I. Immunol., 138: 520-523 (1987). Deems, R.A., Eaton, B.R., Dennis, E.A., “Kinetic analysis of phospholipase A 2 activity toward mixed micelles and its implications for the study of lipolytic enzymes”, J Biol. Chern., 250(23): 9013- 9020 (1975). Doering, T.L., Masterson, W.J., Hart, G.W., Englund, P.J., “Biosynthesis of glycosyl phosphatidylinositol membrane anchors”, J. Biol. Cheni., 265: 61 1-614 (1990). Doery, H.M., Magnusson, B.J., Gulasekharam, J., Pearson, J.E., “The properties of phospholipase enzymes in Staphylococcal toxins”, J. Gen Microbiol., 40: 283-296 (1965). Eibl, H., Lands, W.E.M., “A new, sensitive determination of phosphate”, Anal. Biochern., 30: 5 1-57 (1969). Estin, C.D., Stevenson, U.S., Plowman, G.D., Hu, S.L., Sridhar, P., HelistrOm, I., Brown, J.P., HellstrOm, K.E., “Recombinant vaccinia virus against human melanoma antigen p97 for use in immunotherapy”, Biochern., 85: 1052-1056 (1988). Estin, C.D., Stevenson, U.S., Kahn, M., HelistrOm, I., HelistrOm, K.E., “Transfected mouse melanoma lines that express various levels of human melanoma-associated antigen p9’7”,J. Natl. Cancer Inst., 81: 445-448 (1989).  79 Etges, R., Bourier, J., Bordier, C., “The major surface protein of Leishmania promastigotes is a protease”, J. Blot. Chern., 261(20): 9098-9101 (1986). Ferguson, A.J., Williams, A.F., “Cell-surface anchoring of proteins via glycosyl phosphatidylinositol structures”, Ann. Rev. Biochern., 57: 285-320 (1988a). Ferguson, M.A.J., Homans, S.W., Dwek, R.A., Rademacher, T.W., “Glycosyl phosphatidylinositol moeity that anchors Trypanosoma brucei variant surface glycoprotein to the membrane”, Science, 239: 753-759 (1988b). Ferguson, A.J., Low, M.G., Cross, G.A.M., “Glycosyl-sn-1,2-dimyristylphosphatidylinositol is covalently linked to Trypanosoma brucei variant glycoprotein”, J. Biol. Chern., 260: 14547-14555 (1985). Food, M.R., Rothenberger, S., Gobathuler, R., Haidi, I.D., Reid, G.S., Jefferies, W.A., “Transport and expression in human melanomas of a transferrin-like glycosyl phosphatidylinositol-anchored protein”, I Biot. Chern., 269: 1-7 (1994). Gall, W.E., Edelman, G.M., “Lateral diffusion of surface molecules in animal cells and tissues”, Science, 213: 902-905 (1981). Gramp, G.E., Sambanis, A., Stephanopoulos, G.N, “Use of regulated secretion in protein production from animal cells: an overview”, Adv. Biochern. Eng., Biotechnol., 46: 35-62 (1992). Griffith, O.H., Voiwerk, J.J., Kuppe, A., “Phosphatidylinositol specific phospholipase C from Bacillus cereus and Bacillus thuringiensis”, Meth. Enzyrnol., 197, 493-502 (1991). He, H.T., Barbet, J., Chaix, J.C., Goridis, C., “Phosphatidylinositol is involved in the membrane attachment of N-CAM , the smallest component of the neural cell adhesion 120 molecule”, E1’vIBO J, 5: 2489-2494 (1986). HellstrOm, K.E., Hellstrôm, I., “Cellular immunity against tumor specific antigens”, Adv. Cancer Res., 12: 167-223 (1969). Hemperly, J.J., Edelman, G.M., Cunningham, B.A., “cDNA clones of neural cell adhesion molecule (N-CAM) lacking a membrane-spanning resin consistent with evidence for membrane attachment via a phosphatidylinositol intermediate”, Proc. Nail. A cad. Sci., 83: 9822-9826 (1986). Hendrickson, H.S., Dennis, E.A., “Kinetic analysis of the dual phospholipid model for phospholipase A 2 action”, JBiot. Cheni., 259(9): 5734-5739 (1984).  80 Hu, S.L., Plowman, G.D., Sridhar, P., Stevenson, U.S., Brown, J.P., Estin, CD., “Characterization of recombinant vaccinia virus expressing human melanoma-associated antigen p97”, J. Viral., 62(1): 176-180 (1988). Ikezawa, H., “Phosphatidylinositol-specific phospholipase C”, J. Toxicol. Toxin Rev., 5: 124 (1986). Ikezawa, H., Taguchi, R., “Phosphatidylinositol-specific phospholipase C from Bacillus cereus and Bacillus thuringiensis”, Meth. Enzymol., 71, 73 1-741 (1981). Ikezawa, H., Yamanegi, M., Taguchi, R., Miyashita, T., Ohyabu, T., “Studies of phosphatidylinositol phosphodiesterase (phospholipase C type) of Bacillus cereus I. Purification, properties and phosphatase releasing activity”, Biochin. Biophys. Acta, 450: 154-164 (1976). Ishihara, A., Hou, Y, Jacobson, K., “The Thy-i antigen exhibits rapid lateral diffusion in the plasma membrane of rodent lymphoid cells and fibroblasts” Proc. Nail. Acad Sd., 84: 1290-1293 (1987). Ishihara, M., Fedarko, N.S., Conrad, H.E., “Involvement of phosphatidylinositol and insulin in the coordinate regulation of proteoheparan sulfate metabolism and hepatocyte growth”, J. Biol. Chern., 262: 4708-47 16 (1987). Kennard, ML., Food MR., Jefferies, W.A., Piret, J.M., “Controlled release process to recover heterologous glycosyl phosphatidylinositol membrane anchored proteins from CHO cells”, Biotechnol. Bioeng., 42: 480-486 (1993). Kennard, M.L., Piret, J.M., “Glycolipid membrane anchored recombinant protein production from CHO cells on porous microcarriers”, BiotechnoL Bioeng., in press (1994). Koch, F., Thiele, H.G., Low, M.G. “Release of the rat T cell alloantigen RT 6.2 from cell membranes by phosphatidylinositol-specific phospholipase C”, I. Exp. Med., 164: 1338-1343 (1986). -  Kume, T., Taguchi, R., Tomita, M., Tokuyama, S., Morizawa, K., Nakachi, 0., Hirano, J., Ikezawa, H., “The study of phosphatidylinositol-specific phospholipase C from Bacillus thuringiensis: Synthesis of homogeneous substrates, substrate specificity and other properties”, Chem. Pharm. Bull., 40(8): 2133-2137 (1992). Kupke, T., Lechner, M., Kaim, G., GOtz, F., “Improved purification and biochemical properties of phosphatidylinositol specific phospholipase C of Bacillus thuringiensis”, Eur. J. Biochem., 185: 15 1-155 (1989). -  81 Kuppe, A., Evans, L.M., McMillen, D.A., Griffith, O.H., “Phosphatidylinositol specific phospholipase C of Bacillus cereus: Cloning, sequencing and relationship with other phospholipases”, J. Bacteriol., 171(11): 6077-6083 (1989). -  Leigh, A.J., Volwerk, J.J., Griffith, OH., Keana, J.F.W., “Substrate stereospecificity of phosphatidylinositol-specific phospholipase C from Bacillus cereus examined using the resolved enantiomers synthetic myo-inositol 1-C4-nitriphenyl phosphate”, Biochern., 31: 8978-8983 (1992). Lisanti, M.P., Caras, I.W., Davitz, M.A., Rodriguez-Boulan, E., “A glycophospholipid membrane anchor acts as an apical targeting signal in polarized epithelial cells”, J. Cell. Biol., 109: 2145-56 (1989). Lisanti, M.P., Rodriguez-Boulan, E., Saltiel, A.R., “Emerging functional roles of the glycosyl-phosphatidylinositol membrane protein anchor”, I Menibr. Blot., 117: 1-10 (1990). Low, M.G., “Biochemistry of the glycosyl-phosphatidylinositol membrane protein anchors”, Biochern. J., 244: 1-13 (1987). Low, M.G., “The glycosyl-phosphatidylinositol anchor of membrane proteins”, Biochirn. Biophys. Acta, 988: 427-454 (1989). Low, MG., Finean, J.B., “The anchor of phosphatidylinositol specific phospholipases on membranes”, Biochern. J., 154: 203-208 (1976). Low, M.G., Finean, J.B., “Release of alkaline phosphatase from membranes by phosphatidylinositol specific phospholipase C”, Biochern. 1, 167: 281-284 (1977). -  Low, M.G., Finean, J.B., “Specific release of plasma membrane enzymes by a phosphatidylinositol-specific phospholipase C”, Biochirn. Biophys. Acta., 508: 565-570 (1978). Low, M.G., Futerman, A.H., Ackermann, K.E., Sherman, W.R., Silman, I., “Removal of covalently bound inositol from Torpedo acetyicholinesterase and mammalian alkaline phosphatases by deamination with nitrous acid”, Biochern. 1, 241: 615-619 (1987). Low, M.G., Kincade, P.W., “Phosphatidylinositol is the membrane-anchoring domain of the Thy-1-glycoprotein”, Nature, 318: 62-64 (1985). Low, M.G., Saltiel, A.R., “Structural and functional roles of glycosyl phosphatidylinositol in membranes”, Science, 239: 268-275 (1988b). Missirlis, Y.F., Lemm, W., “Modern aspects of protein adsorption on biomaterials”, Kinetics ofprotein adsorption, Kiuwer Academic Publishers, London, 1991.  82 Moran, P., Caras, I.W., “Fusion sequence from non-anchored proteins to generate a frilly functional signal for glycophosphatidylinositol membrane anchor attachment”, J. Cell Biol., 115(6): 1595-1600(1991). Nakabayashi, T, Ikezawa, H., “Release of alkaline phosphodiesterase I from rat kidney plasma membrane produced by the phosphatidylinositol-specific phospholipase C of Bacillus thuringiensis”, Cell Struct. Fund., 9: 247-263 (1984). Nakabayashi, T, Ikezawa, H., “Alkaline phosphodiesterase I release from eukaryotic plasma membranes by phosphatidylinositol-specific phospholipase C. I. The release from rat organs”, J.Biochern., (Tokyo), 99(3): 703-712 (1986). Reed, R., “Spurious cell surface receptors: inadequate correction for saturable, non specific binding mimics receptor binding”, Anal. Biochem., 185: 160-163 (1990). Reisser, H., Oettgen, H., Yeh, E.T.H., Terhorst, C., Low, M.G., Benacerraf, B., Rock, K.L., “Structural characterization of the TAP molecule: a phosphatidylinositol linked glycoprotein distinct from the T cell receptor/T3 complex and Thy-i”, Cell, 47: 365-3 70 (1988). cell adhesion and Renner, W.A., Jordan, M., Eppenberger, H.M., Leist, C., “Cell aggregation: influence of growth behaviour of CHO cells”, Biotechnol. Bioeng., 41: 188193 (1993). -  Richardson, D., Baker, E., “The uptake of inorganic iron complexes by human melanoma cells”, Biochirn. Biophys. Ada., 1093: 20-28 (1991). Rose, T.M., Plowmann, G.D., Teplow, D.B., Dreyer, W.J., HellstrOm, K.E., Brown, J.P., “Primary structure of the human melanoma associated antigen p97 (melanotransferrin) deduced from the mRNA sequence”, Proc. Natl. Acad. Sci., 83: 1261-1265 (1986). Sambanis, A., Stephanopoulos, G., Sinskey, A.J., Lodish, H.F., “Use of regulated secretion in protein production from animal cells: an evaluation with the AtT-20 model cell line”, Biotechnol. Bioeng., 35: 771-780 (1990a). Sambanis, A., Stephanopoulos, G., Lodish, H.F., “Multiple episodes of induced secretion of human growth hormone from recombinant AtT-20 cells”, Cytotechnol., 4: 111-119 (1990b). Scallon, B.J., Kado-Fong, H., Nettleton, M.Y., Kochan, J.P., “A novel strategy for secreting proteins: Use of phosphatidylinositol-glycan specific phospholipase D to release chimeric phosphatidylinositol-glycan anchored proteins”, Bio/Technol., 10: 550-556 (1992).  83 Shashidhar, M.S., Volwerk, J.J. Keana, J.F.W., Griffith, O.H., “A fluorescent substrate for continuous assay of phosphatidylinositol-specific phospholipase C: synthesis and application of 2-naphthyl myo-inositol-1-phosphate”, Anal. Biochem., 198: 10-14 (1991a). Shashidhar, M.S., Voiwerk, J.J., Griffith, O.H., Keana, J.F.W., “A chromogenic substrate for phosphatidylinositol-specific phospholipase C: 4-nitrophenyl myo-inositol-1phosphate”, Chem. Phys. Lipids, 60: 101-110(1991b). Shin, Y., Levinthal, C., Levinthal, F., Hubell, W.L., “El binding to membranes: Time resolved studies of spin labeled mutants”, Science, 259: 960-963 (1993). Shukia, S.D., Coleman, R., Finean, J.B., Michell, R.H., “Selective release of plasmamembrane enzymes from rat hepatocytes by a phosphatidylinositol-specific phospholipase C”, Biochem. J., 187: 277-280 (1980). 51cm, M.W., Logan, G.F. Jr., “Partial purification and properties of two phospholipases of Bacillus cereus”, J. Bacteriol. 85: 369-381 (1963). Slein, M.W., Logan, G.F., “Characterization of the phospholipases of Bacillus cereus and their effects on erythrocytes, bone and kidney cells”, J. Bacteriol. 90: 69-8 1 (1965). Steirnberg, J.. Low, M.G., Flaherty, L., Kincade, P.W., “Removal of lymphocyte surface molecules with phosphatidylinositol-specific phospholipase C: Effects on mitogen responses and evidence that ThB and certain Qa antigens are membrane-anchored via phosphatidylinositol”, J. Immunol., 38: 3877-84 (1987). Steinman, R.M., Brodie, S.E., Cohn, Z.A., “Membrane flow during pinocytosis, a steriologic analysis”, I Cell Blot., 68: 665-687 (1976). Sundler, R., Alberts, A.W., Vagelos, P.R., “Enzymatic properties of phosphatidylinositol inositol phosphohydrolase from Bacillus cereus”, J. Biot. Cheni., 253, 4175-4179 (1978). Taguchi, R., Ikezawa, H., “Phosphatidylinositol-specific phospholipase Clostridium novyi Type A”, Arch. Biocheni. Biophys., 186, 196-201 (1978).  C  from  Taguchi, R., Asahi, Y., Ikezawa, H., “Purification and properties of phosphatidylinositol specific phospholipase C from Bacillus thuringiensis”, Biochini. Biophys. Acta, 619, 48-57 (1980). Takasue, Y., Yokota, K., Nishi, Y., Taguchi, R., Ikezawa, H., “Solubilization of trehalase from rabbit renal and intestinal brush-border membranes by a phosphatidylinositol-specific phospholipase C”, FEBsLeIt., 210: 5-8 (1986).  84 Tse, A.G.D., Barclay, A.N., Watts, A., Williams, A.F., “A glycophospholipid tail at the carboxyl terminus of the Thy-i glycoprotein of neurons and thymocytes”, Science, 230: 1003-1008 (1985). Turner, A.J., “Molecular and cell biology of membrane proteins Glycolipid anchors of cell surface proteins”, Series of Molecular Biology, Ellis Horwood Publishers, Toronto, ON (1990). -  Voiwerk, J.J., Wetherwax, P.B., Evans, L.M., Kuppe, A., Griffith, O.H., “Phosphatidylinositol-specific phospholipase C from Bacillus cereus: Improved purification, amino acid composition and amino-terminal sequence”, J. Cell. Biochern., 39: 3 15-325 (1989).  85  Appendix 1 Growth of CHO cells in different growth media Tables labelled according to the corresponding figures Figure 4.1 Growth of CHO cells in CHO-S-SFM I Time  Cell Concentration  Glucose  Viability  [day]  [106 Cells/mL]  [gfL]  [%]  0 2 3 4 5 6 7 8 9 10 11 12  0.10  3.80  100.0  3.60 1.24 3.00 4.90 5.20 5.80 5.90 5.50 4.70 3.50 3.10  3.50 3.20 2.40 1.40 0.75 0.58 0.30 0.18 0.15 0.14 0.12  99.7 97.6 98.4 94.2 92.9 90.6 89.7 88.0 82.5 71.4 76.5  Figure 4.6 Growth of CHO cells in HBCHO Time  Cell Concentration  Glucose  Viability  [day]  [106 Cells/mL]  [g/L]  [%]  0 1 2 3 5 6 7  0.07 0.06 0.42 0.58 1.30 1.50 0.76  1.31 1.20 1.00 0.58 0.12 0.11 0.10  68.0 74.0 94.4 84.1 92.2 78.1 74.0  86  Figure 4.2 Growth of CHO cells in Ham’s F12 with 5 ig/mL insulin, 5 tg/mL transferrin, 10 nM sodium selenite, 50 j.tg/mL BSA and 50 .tg/mL DNase. Time  Cell Concentration  Glucose  Viability  [day]  [106 Cells/mU  [g/L]  [%]  0 1 2 3 4 5 6 7 8 9 10 11 12  0.20 0.25 0.27 0.37 0.48 0.61 0.65 0.82 0.83 0.58 0.50 0.30 0.20  1.70 1.61 1.45 1.40 1.28 1.14 1.10 0.98 0.85 0.67 0.60 0.45 0.25  87.0 96.2 96.4 90.2 92.3 95.3 81.3 79.6 80.6 65.9 50.0 33.0 23.5  Figure 4.3 Growth of CHO cells in Ham’s F12 with 5 tg/mL insulin, 5 tg/mL transferrin, 10 nM sodium selenite, 300 j.tg/mL BSA and 50 j.ig/mL DNase. Time  Cell Concentration  Glucose  Viability  [day]  [106 Cells/m.LJ  [g/L]  [%]  0 1 2 3 4 5 6 7 8 9 10 11 12  0.25 0.34 0.71 0.89 0.83 0.98 1.25 1.22 1.23 0.73 0.70 0.56 0.50  1.55 1.44 1.27 1.30 1.15 1.00 1.01 0.70 0.71 0.54 0.32 0.20 0.15  75.8 97.1 91.0 88.1 87.4 83.1 82.8 81.9 75.0 65.8 56.0 51.8 51.5  87  Figure 4.4 Growth of CHO cells in DMEM!F12 with 50 jig/mL DNase. Time  Cell Concentration  Glucose  Viability  [day]  [106 Cells/mL]  [g/L]  [%]  0 1 2 3 4 5 6 7 8 9 10 11 12  0.20 0.21 0.36 0.50 0.88 1.29 1.56 1.93 2.10 1.86 1.64 1.48 1.38  2.92 2.82 2.57 2.50 2.30 2.30 2.10 1.80 1.70 1.50 1.43 1.25 0.95  87.0 95.5 97.3 93.6 96.3 98.1 96.3 95.8 96.8 90.7 87.7 92.5 87.9  Figure 4.5 Growth of CHO cells in DMEMIFI2 with 5 pg/mL insulin, 5 tg/mL transferrin, 10 nM sodium selenite, 50 j.tg!mL BSA and 50 p.g/mL DNase. Time  Cell Concentration  Glucose  Viability  [day]  [106 Cells/mL]  [g/L]  [%]  0 1 2 3 4 5 6 7 8 9 10 11 12  0.20 0.28 0.35 0.56 0.83 1.18 1.47 1.85 1.70 1.87 1.58 1.50 1.26  2.92 2.78 2.60 2.50 2.45 2.35 2.25 1.97 1.86 1.64 1.62 1.45 1.33  87.0 96.6 97.2 94.9 96.5 96.7 97.0 97.1 94.7 9L2 85.4 86.6 81.0  88  Figure 4.7 Growth of CHO cells in CHO-S-SFM II Time  Cell Concentration  Glucose  Viability  [day]  [106 Cells/mLj  [g/L]  [%J  0 1 2 3 5 6 7 8 9 10  0.06 0.09 0.31 0.90 3.46 2.82 2.12 2.02 0.83 0.00  3.74 3.68 3.40 2.87 1.68 1.18 0.81 0.48 0.32 0.18  81.8 89.7 93.9 96.8 96.7 92.8 74.1 64.7 33.2 0.0  Figure 4.8 Growth of CHO cells in CHO-S-SFM II and 100 tg/mL DNase  Time  Cell Concentration  Glucose  Viability  [day]  [106 Cells/mU  [g/L]  [%J  0 1  0.20 0.45  3.80 3.50  100.0 97.8  2 3 4  1.10 1.54 3.00 5.00 7.00 5.20 4.60  3.35 2.95 2.50 2.20 1.60 0.98 0.58  97.4 96.9 98.4 98.0  5 6 7  8  98.2 89.4  78.0  89  Appendix 2 Variation of initial cell density The effect of varying the initial cell density on the growth profiles was investigated. The aim was to increase the final cell concentration by varying inoculation concentration. Figure 7.1 shows that cultures started with higher inoculum grew to higher cell densities. Cell culture 6 cells/mL and the 4 cells/mL of medium reached a maximum of 4x10 started with 3.5x10 culture that was started with greater lx 106 cells/mL achieved a maximum of about 6 cells/mL. There appears a limit to the high density inoculum that can be used to boost 7x10 the growth of cells.  6 cells/mL, achieved A cell culture that was inoculated with 2.2x10  6 cells/mL only. 7x10  2 E  6 7.0x10  j 6.Ox1O C-)  5.1O  6 j 4.0x10 1)  6 3.0x10  0  C)  a)  C-)  Time (day) Figure 7J Growth profile of cells growing in CHO-S-SFM I. Cultures started at different inoculum densities.  90  -J  a)  U) 0 0  CD  Time (day) Figure 7.2 Cells growing in CHO-S-SFM I. Glucose concentrations for cultures that were started at different initial cell density.  Inoculating the cells at high cell concentration is a poor scale-up technique as the inoculum culture will have to be maintained for longer periods to achieve high cell concentrations. The four initial cell concentrations that were investigated gave the following final to initial cell concentration ratios: (i) 7/1.5 (ii) 3/0.03 (iii) 7/1 4 cells/mL gave a 100-fold increase. culture inoculated with 3x10  (iv) 7/2.2. The  Starting at very low cell  4 cells/mL) will also increase the total time to achieve the densities (approximately 1x10 maximum cell density because of the initial lag time, therefore we will lose up to two days.  91  Tables labelled according to the corresponding figures  Figures 7.1 and 7.2 Effect of innoculum concentration on final cell density Time [day] A 0 1 2 3 4 5  1.48 2.76 4.20 6.48 6.88 6.50  Cell Concentration [106 Cells/mLj B C 0.35 0.44 1.60 3.90 3.00  1.06 2.38 4.86 7.00 7.04 4.60  Glucose D  A  2.16 4.35 6.70 7.00 3.00  3.69 2.22 0.84 0.32 0.14 0.12  [mg/mL] B C 3.67 3.25 2.33 1.05  3.81 2.39 1.20 0.57 0.19 0.13  D 3.88 1.85 0.75 0.30 0.14  92  Appendix 3 Effect of harvesting conditions on p97 removal Tables labelled according to the corresponding figures i)  Figure 4.9 Effect of incubation time on p97 harvest p97 removal  PT-PLC  ii)  [%J  [mU/mL]  60 mm  30mm  15 mm  10 5 1 0.5 0  52.4 318 9.5 6.2 6.2  38.9 23.2 6.1 3.6 4.5  23.9 13.6 2.7 0.9 3.6  Figure 4.10 Effect of P1-PLC volume on p97 harvest Volume of  p97  PT-PLC  removal  [jiLl  [%]  50 100 200 400  16.7 19.7 22.5 28.0  93  Appendix 4 P1-PLC assay results Tables labelled according to the corresponding figures  i)  ii)  Figure 4.11 P1-PLC assay based on flow cytometry  PT-PLC  p97 removal  Error  [mU/mU  [%J  ±  156.0 78.0 15.6 7.8 1.6 0.8 0.2  97.5 94.1 67.7 52.5 15.1 7.9 2.7  -  -  1.9 0.5 -  -  -  Figure 4.12 Standard curve based on flow cytometry and immunofluorescence  PT-PLC  p97  97 removal p  [mU/mU  [ig/mL]  [%]  156.0 78,0 15.6 7.8 1.6 0.8 0.2  10.60 9.80 9.50 8.80 4.50 2.98 1.20  98.9 97.9 87.1 74.7 36.5 25.2 10.3  94  Appendix 5 Repeated harvesting process results Tables labelled according to the corresponding figures Figures 4.13 and 4.14 Harvest  p97 removal  Cumulative  Calculated  97 p  97 p  [#]  [%j  [p.g/mLJ  [p.g/mLJ  0 1 2 3 4 5 6 7  0.0 58.1 34.3 36.6 28.0 28.0 20.4 20.4  0.00 59.7 109.0 110.5 105.5 103.6 105.8 115.3  0.00 59.7 111.4 117.2 116.6 118.9 121.1 139.0  Figures 4.21 and 4.22 Harvest  p97 removal  Cumulative p97  Calculated p97  {#]  {%]  [jig/mL]  [.tg/mL]  0 1 2 3 4 5 6 7 8 9 10  0 75.7 58.0 48.4 31.3 26.1 91.3 92.0 79.4 55.1 52.1  0.0 72.6 148.6 118.1 148.5 148.9 196.0 250.1 298.5 247.2 275.7  0.0 72.6 150.4 123.6 157.0 161.1 211.9 270.9 325.6 281.7 316.4  95  Figures 4.23 and 4.24  30 mU/mL P1-PLC  300 mU/mL P1-PLC p97  Cumulative  Calculated  p97  Calculated p97  removal  p97  97 p  [%]  [ig/mL]  [jig/mU  [%]  [jig/mL]  [jig/mU  97.1 97.5 96.5 98.2 98.1  85.2 147.6 202.7 249.5 274.3  85.2 162.4 226.0 293.7 343.3  84.9 91.9 93.3 95.2 93.1  62.1 129.7 160.3 193.4 240.3  62.1 135.9 179.5 228.7 294.8  p97  Cumulative  removal  [#J 1 2 3 4 5  Harvest  96  Appendix 6 p97 and P1-PLC stability Tables labelled according to the corresponding figures i)  ii)  Figure 4.15 p97 stability at 37°C Time  p97  Error  [hi  [jig/mU  ±  0 1 2 3 24 51  9.65 10.51 9.69 10.25 9.92 8.60  0.0 1.41 0.61 0.94 0.41 0.30  Figure 4.16 P1-PLC stability at 37 °C Time  PT-PLC  [day]  [mU/mL]  0.00 0.004 0.007 0.01 0.02 0.04 0.08 0.21 1.00 5.00 14.0  60.78 54.48 54.48 50.16 55.73 55.11 55.11 49.93 56.36 46.53 50.16  97  iii)  Figure 4.17 Effect of pH on PT-PLC and p97. P1-PLC solution of 16 mU/mL was  used. pH  6.00 6.50 7.00 7.50 7.90  97 removal p  [%J  p97 {.tg/mL]  Error ±  93.5 84.7 75.0 64.2 51.2  12.97 10.23 9.08 10.86 6.85  0.88 0.85 0.35 0.33 1.67  98  Appendix 7 Adsorption and desorption of P1-PLC to the cell surface Tables labelled according to the corresponding figures Figure 4.18 Adsorption of PT-PLC on CHO cells over a period of 60 mm. A control of i) PT-PLC incubation without cells was also done for 60 mm. P1-PLC  P1-PLC  Adsorption  Control  [mini  [mU/mL]  [mU/mL]  0 3 15.0 30.0  18 2.0 2.0 3.5 4.2  18  Time  60.0  ii)  -  Figure 4.19 Desorption of P1-PLC on CHO cells over 120 mm. Time  p97 removal  [mini  [%]  0 3 15.0 60.0 120.0  0 12.0 12.8 10.4 12.0  99  iii)  Figure 4.20 Data for the P1-PLC equilibrium curve  PT-PLC in solution [mU/mU 0.0 1.0 5.5 9.5 15.5 30.0 37.5 155.0  PT-PLC adsorbed  [  7 mU/cellj (Y 0.0 0.16 0.39 1.56 2.25 5.0 9.05 4.05  100  Appendix 8 Continuous harvesting process results Tables labelled according to the corresponding figures Figure 4.25  [day]  1 2 3 4 5 6 7 8 9 10 11  Viability  Cell Concentration [106 Cells/mU  Time  [%]  0 mU/mL  3 mU/mL  30 mU/mL  0 mU/mL  3 mU/mL  30 mU/mL  0.12 0.34 0.53 1.60 1.90 1.80 1.20 1.60 0.77 0.29 0.04  0.14 0.34 0.56 1.32 2.20 1.50 1.10 0.69 0.68 0.67 0.24  0.15 0.40 0.76 1.52 2.80 1.90 1.75 1.65 1.40 1.30 0.62  80.0 91.9 91.4 95.8 93.1 90.5 72.3 67.8 43.5 16.2 2.2  93.3 94.4 90.3 95.7 91.7 87.2 68.8 48.3 57.6 44.7 19.4  93.8 95.2 97.4 94,7 94.0 83.0 70.3 73.3 64.5 41.9 19.3  101  Figure 4.26 Time  Cell surface fluorescence  Glucose Concentration  [day]  [fluorescence/cell]  [mglmL]  0 mU/mL  3 mU/mL  30 mU/niL  0 mU/mL  3 mU/mL  30 mU/mL  PT-PLC  P1-PLC  P1-PLC  P1-PLC  P1-PLC  PT-PLC  320.0 314.4 257.6 241.2 190.2 161.3 135,7 89.8 55.5 32.1 27.7 25.8  320.0 168.1 150.7 98.4 47.4 35.6 23.7 26.5 20.0 22.6 16.2 13.6  320.0 129.1 180.8 128.5 40.5 15.3 8.50 10.9 7.5 9.2 8.2 10.6  3.90 3.70 3.33 2.88 1.88 1.55 1.32 0.95 0.92 0.63 0.66 0.68  3.90 3.40 3.37 2.87 2.07 1.50 1.33 0.98 1.04 0.93 0.87 0.77  3.90 3.72 3.35 2.62 1.91 1.34 1.23 0.83 0.50 0.38 0.17 0.13  0 1 2 3 4 5 6 7 8 9 10 11  Figure 4.27 Time  Cumulative p97  [day]  [ig/mL]  0 1 2 3 4 5 6 7 8 9 10 11  0 mU/mL  3 mU/mL  30 mU/mL  PT-PLC  PT-PLC  PT-PLC  0.3 0.3 0.4 0.7 1.4 2.3 3.0 3.1 3.1 3.2 3.5 3.8  0.4 0.4 0.8 1.5 3.4 5.8 8.5 8.6 9.2 9.3 10.5 11.6  0.4 0.4 0.8 2.3 4.5 7.9 8.8 9.7 11.5 11.5 13.4 15.3  102  Appendix 9 Calculations for p97 yield on glucose and P1-PLC i)  Cyclic harvesting  6 cells/mL Reactor Volume: 250 mL at 4x10 Total time: 8 days Number of harvests: 5 9 cells Cells per harvest: 1x10 PT-PLC concentration: 30 mUImL PT-PLC volume per harvest: 5 mL p97 recovered from single harvest (estimated from the work of Kennard et al., 1993; cells in 0.5 mL P1-PLC): 62 .tg/mL based on harvesting CHO-S-SFM I medium per harvest: 250 mL Initial glucose concentration: 4 mg/mL Final glucose concentration: 0.8 mg/mL Total glucose consumed  = =  Total PT-PLC used  (250 mL/harvest  5 harvests) *(4 0.8) mg/mL glucose -  4000 mg  (30 mU/mL)  =  *  *  (5 mL per harvest) *(5 harvests)  750 mU P1-PLC  Total p97 harvested  (62 .ig/mL/harvest) = 1550 ig  *  (5 mL/harvest)  Therefore: Yield on glucose  Yield on PT-PLC  =  (1550 ig p97)/Ql000 mg glucose)  =  0.39 i.g p97/mg glucose  =  (1550 jig p97)/(750 mU P1-PLC)  =  Average p97 /cell-day  =  2.07 jig p97/mU PT-PLC  9 cells 1550 jig p97/(1x10  *  8 days)  *  (5 harvests)  103  ii)  Repeated harvesting process  a)  Initial P1-PLC concentration 300 mU/mL  6 cells/mL Reactor volume: 250 mL at 4x10 Total time: 8 days Number of harvests: 5 9 cells Cells per harvest: 1x10 Initial P1-PLC concentration: 300 mU/mL PT-PLC replenishment after each harvest: 1500 mU Total PT-PLC volume: 5 mL Total p97 harvested: 1715 jtg CHO-S-SFM I medium per harvest: 250 mL Initial glucose concentration: 4 mg/mL Final glucose concentration: 0.8 mg/mL  Total glucose consumed  = =  Total P1-PLC used  (1250 mL medium)  *  (4 0.8) mg/mL glucose -  4000 mg  (300 mU/niL) * (5 ml) 7500 mU P1-PLC  +  (1500 mU/harvest  Therefore: (1715 .tg p97)1(4000 mg glucose) = 043 .tg p97/mg glucose  Yield on glucose  =  Yield on PT-PLC  = =  Average p97/cell-day  =  (1715 .tg p97)1(7500 mU PT-PLC) 0.23 ig p97/mU P1-PLC  9 cells * 8 days) 1715 ig p97/(1x10 2.14x10 jig p97/cell-day  *  4 harvests)  104  b)  Initial P1-PLC concentration 30 mU/mL  6 cells/mL Reactor volume: 250 niL at 4x10 Total time: 8 days Number of harvests: 5 9 cells Cells per harvest: 1x10 Initial P1-PLC concentration: 30 mU/mL PT-PLC replenishment after each harvest: 150 mU Total P1-PLC volume: 5 mL Total p97 harvested: 1474 tg CHO-S-SFM I medium per harvest: 250 mL Initial glucose concentration: 4 mg/niL Final glucose concentration: 0.8 mg/mL Total glucose consumed  Total PT-PLC used  =  (1250 mL medium)  =  4000 mg  (30 mU/mL)  =  *  (5 niL)  +  *  (4 0.8) mg/mL glucose -  (150 mU/harvest  750 mU P1-PLC  Therefore: Yield on glucose  Yield on PT-PLC  Average p97/cell-day  =  (1474 jig p97)1(4000 mg glucose)  =  0.37 i’g p97/mg glucose  =  (1474 jig p97)1(750 mU P1-PLC)  =  1.97 jig p97/mU PT-PLC  = =  1474 jig p97/(1x10 9 cells  *  7 pg p97/cell-day 1.84x10  8 days)  *  4 harvests)  105  iii) Continuous harvesting process Reactor volume: 250 mL (batch culture) PT-PLC concentration: 30 mU/mL Total time: 8 days Total p97 harvested was: 11.5 .tg/mL  *  250 mL  =  2875 ig  Average initial glucose concentration: 3.9 mg/mL Average final glucose concentration: 0.5 g/L Total glucose consumed  =  250 mL =  Total PT-PLC used  =  250 mL  *  3.4 mg/mL  850 mg glucose *  30 mU/mL  7500 mU P1-PLC Therefore: Yield on glucose  =  (2875 pg p97)1(850 mg glucose) = 3.4 jig p97/mg glucose  Yield on PT-PLC  = =  Total cells  =  250 mL  Average p97/cell-day  *  = =  (2875 jig p97)1(7500 mU PT-PLC) 0.38 jig p97/mU P1-PLC  6 cells/mL = 6.25x10 2.5x10 8 cells 2875 jig p97/(6.25x 108 cells 7 jig p97/cell-day 5.75x1W  *  8 days)  

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.831.1-0058606/manifest

Comment

Related Items