AD LARGE SCALE FERMENTER GROWTH OF ANIMAL CELLS FOR VIRUS VACCINE PRODUCTION: CONTROL OF SEED STOCK AND GROWTH CONDITIONS (U) FINAL REPORT by Donald J. merchant, Ph.D. Professor, Department of Microbiology and. Lloyd L. Kempe Professor, Departments of Chemical and Metallurgical Engineering and of Microbiology December 1969 (For the period 1 January 1969 to 31 August 1969) Supported by: U. S. Army Medical Research and Development Command Office of The Surgeon General, Washington, D. C. 20315 Contract No. DADA-17-67-C-7073 The University of Michigan Ann Arbor, Michigan 48104 DOD DISTRIBUTION STATEMENT Each transmittal of this document outside the Department of Defense must have prior approval of Commanding General, U. S. Army Medical Research and Development Command. The findings in this report are not to be construed as an official Department of the Army position unless so designated by other authorized documents.

SUMMARY Using equipment previously constructed on the program L-M mouse cells have been grown in a 5-liter fermenter in medium 199 supplemented with peptone and serum. Changes in pH, O-R potential, dissolved oxygen and cell number and viability have been continuously monitored. Cells also have been grown under conditions of controlled pH and O-R potential and the relationship of O-R potential to growth rate has been determined. Work has been completed on a computer program for analysis of chromosome changes in aneuploid cell populations. The method also is applicable to study of diploid cells. Automated methods combined with computer analyses permit a rapid and sensitive monitoring of chromosomal shifts in cell population. Using the computer program, as well as more conventional methods of chromosomal analysis, data is presented on the effects of several types of environmental selection pressures on the chromosomal complement of L-M mouse cells and BHK21PD4 cells. Detailed studies of the amino acid nutrition of L-M mouse cells in chemically defined medium is presented. Combining the simplicity of Eagle basal medium with the sensitivity and accuracy of an amino acid analyzer, it has been possible to obtain critical evidence concerning the rates of use of individual amino acids, the role of imbalance of amino acids in utilization patterns and the effects of supplementing nonessential amino acids. Finally, studies of the nutrition of BHK21PD4 cells have led to the development of a very efficient medium (schedule 8) for this cell\ line. The comparative costs and yield is compared with more conventional media. Improved methods of storage and of freezing seed stocks of these cells also are presented.

FOREWORD In addition to the authors the following persons have contributed to this work: Gregory Baumann Richard Giles William Kerr, Jr. Edward Lunk Jens Pedersen Richard Platte, Jr. Howard Stockdale, Ph.D. Gary Stoner iv

TABLE OF CONTENTS Page LIST OF TABLES viii LIST OF FIGURES x INTRODUCTION 1 THE POPULATION DOUBLING TIME OF STRAIN L-M MOUSE CELLS IN SUSPENSION CULTURE AS A FUNCTION OF CONSTANT, CONTROLLED REDOX POTENTIAL-Jens U. Pedersen 3 Abstract 4 Introduction 5 Materials and Methods 5 Cell strain 5 Medium 5 Medium preparation 6 Common experimental conditions for the controlled fermenter runs 7 Procedure for a run 8 Cell count 10 PCV 10 Toxicity 10 Results 11 Discussion 29 Definitions 30 References 31 CHROMOSOME MEASUREMENT: ORGANIZATION AND APPLICATION OF A WORKING SYSTEM —Gregory Baumann, Richard Giles, Edward Lunk, and Richard Platte, Jr. 33 Introduction 34 Equipment and Methods 34 The DDP-24 system 36 Extraction and analysis of measurement data 42 Chromosome 46 Use of computer output 52 Results 58 Accuracy of chromosome measurement 58 Characterization of cell populations 58 Comparison of populations 58 Measurement of human cells 61 VT

TABLE OF CONTENTS (Continued) Page Conclusions 61 References 64 Related Resources 65 Appendix I-Chromosome Procedure 73 Appendix II-Film Processing Technique 74 Appendix III —Sample Session on the DDP-24 75 Appendix IV-DDP Magnetic Tape Track Record Organization 78 Appendix V —Chromosome Data Tape Records 79 Appendix VI —Sample Session Using "SEARCH" 80 Appendix VII —Source Program Listing for "SEARCH" 82 Appendix VIII —Source Program Listing for "CHROMOSOME" 90 Appendix IX-Sample Session Using "CHROMOSOME" 99 Appendix X-Supporting Programs 108 THE EFFECTS OF ENVIRONMENTAL SELECTION PRESSURE ON THE CHROMOSOMAL COMPLEMENT OF L-M MOUSE CELLS AND BHK21PD4 CELLS-Richard Giles, Gregory Baumann, Edward Lunk, and Richard Platte, Jr. 109 Summary 110 Materials and Methods 110 Results 114 L-M cells 114 BHK21PD4 cells 120 Discussion 130 References 131 GROWTH OF STRAIN L-M MOUSE CELLS IN CHEMICALLY DEFINED MEDIUM — Gary Stoner 133 Introduction 134 Materials and Methods 134 Results 135 A. Growth of strain L-M cells in "(2X Eagle" basal medium (containing only the thirteen essential amino acids) 135 B. Growth of strain L-M mouse cells in "(2X) Eagle" basal medium supplemented with nonessential amino acids 135 C. Growth of strain L-M mouse cells in "(2X) Eagle" basal medium containing (4X) cystine 140 D. Spontaneous decomposition of glutamine in uninoculated medium incubated at 37~C for 250 hours 142 References 145 vi

TABLE OF CONTENTS (Concluded) Page STUDIES OF BHK21PD4 CELLS-Howard Stockdale 147 Growth of BHK21PD4 Cells in Schedule 8 and its Modification 148 Growth of BHK21PD4 Cells in Low-Serum Media 149 Growth Response to Serum Concentration 151 The Effect of Varying Glucose Levels on Growth of BHK21PD4 Cells 151 Freezing of BHK21PD4 Cells for Suspension Culture 153 Survival of BHK21PD4 Cells at 40C 154 Uptake of Amino Acids by BHK2 1PD4 Cells During Growth in Schedule 8 Medium 157 DISTRIBUTION LIST 162 vii

LIST OF TABLES Table Page IA. Differences in Experimental Conditions and Systems in Two Fermenter Runs, NB-11 and NB-15 9 IIA. Run NB-11: Population Doubling Times Determined from Total and Viable Counts at Selected, Constant Levels of the Redox Potential for Eight Consecutive Growth Curves (Conditions Described in Section "Common Experimental Conditions" and in Table IA) 12 IIIA. Run NB-15: Population Doubling Times Determined from Total and Viable Counts at Selected, Constant Levels of the Redox Potential for Eight Consecutive Growth Curves (Conditions Described in Section "Common Experimental Conditions" and in Table IA). (Included are Mean Values of pH and Dissolved Oxygen.) 13 IB. Criteria Used to Judge a, Chromosome Spread Before Using it for Extensive Analysis Procedures 35 IIB. The Types of Errors in Chromosome Measurement Data Which Can be Detected by SEARCH 45 IIIB. A List of the Information Provided for Each Spread of Chromosomes by the Program CHROMOSOME 48 IVB. Catagories of Ranges of Lengths Used in the Construction of Arm and Chromosome Length Frequency Distribution 49 VB. Variables Being Compared Using the Wilcoxon TwoSample Test 55 VIB. Means and Standard Deviations for the Values Obtained for Each Characteristic Examined in the Cell Lines Listed 59 VIIB. Values Obtained for aY in the Comparison of a Number of Populations Using the Wilcoxon Two-Sample Test 62 viii

LIST OF TABLES (Concluded) Table Page IC. Frequency Distribution of Total Chromosome Numbers and Numbers of Biarmed Chromosomes in 128 L-M, NB 7/24 L-M, 2817 L-M, 2832 L-M, 2860 L-M, 2884 L-M, P3071 L-M, 170 SL-M Series 3, 170 SL-M Series 2, 101 L-M T1C3 Clone 2. 118 a IIC Frequency and Distribution of the Marker Chromosomes in 128 L-M, NB 7/24 L-M, 2832 L-M, 2860 L-M, 2884 L-M, 1% SL-M Series 2, 1% SL-M Series 3, P3071 L-M. 121 IIIC. Frequency Distribution of Total Chromosome Numbers and Numbers of Biarmed Chromosomes in BHK21 PD4 Suspension No. 49, NB 8 BHK21PD4 Special Medium, 72 BHK21PD4 MEM TP10 CS5, 80 BHK21PD4 MEM TPo1CS1, BHK21PD4 sus- spin. MEM TPlo CS1, BHK21PD4 sus. spin. Special Medium (6). 124 IVC. Frequency and Distribution of mT and Q Chromosomes in 72 BHK21PD4 (monoo), 80 BHK21PD4 (mono.), BHK21PD4 (Spinner) MEM TPloCS1, BHK21PD4 (spinner) Special Medium. 128 ID. Micromoles of Essential Amino Acids Available/ml of Medium at Various Times Throughout Growth of Strain L-M Cells in "2X Eagle" Basal Medium 136 IID. Comparison of the Production and Utilization of Nonessential Amino Acids by Strain L-M Cells in (2X) Eagle Basal Medium With and Without Added Nonessential Amino Acidso 140 IE. Schedule 8 Medium For Spinner Culture 148 IIE. MEM 1% Calf Serum Medium 150 IIIE. Cost of Media for BHK21PD4 Cells 150 IVE. Survival of BHK21PD4 Cells at 4~C 156 VE. Uptake of Free Amino Acids by BHK21PD4 Cells During Growth 158 ix

LIST OF FIGURES Figure Page 1A. Run NB-11: PDTT vs. E. 14 2A. Run NB-ll: PDTV vs. E. 16 3A. Run NB-15: PDTT vs. E. 18 4A. Run NB-15: PDT vs. E o 20 5A. Run NB-15: Growth curve no. 1: Total and viable cell count and PCV vs. time. 22 6A. Run NB-15: Growth Curve no. 1: pH, E, D.O. (mean and range) vs. time. 26 7A. Run NB-15: Growth Curve no. 1: P. vs. time. 28 1B. A photograph of the X-Y coordinate digitizer used to make chromosome measurements. 37 2B. Pictorial development of the method used to describe chromosomes as bubble chamber events. 39 3B. Drawing of the console used for user/machine communications in the DDP-24 system. 40 4B. Sketches of a number of chromosomes showing the location of points which might be recorded, and the course which would be taken in locating the cross hairs on each of these points. 41 5B. Basic organization of the chromosome measurement system (from Baumann and Merchant) (11). 43 6B. Flow chart for the program "SEARCH." 44 7B. Flow chart for the program "CHROMOSOME." 47 8B. CALCOMP plot of chromosome length vs. arm ratio for four BHK cells. 56 9B. Sample frequency distribution plots for chromosome and arm length in two cell lines. 57 x

LIST OF FIGURES (Continued) Figure Page lOB. Karyotype drawn by the CALCOMP plotter to check the accuracy of the measurement system (from Baumann and Merchant) (11). 60 liB. Plot of arm ratio for a single human cell (male) showing how the chromosomes fall into pairs and groups. 63 1Co Marker chromosome of the L-M strain and substrains. 111 2C. Origin of L-M monolayers and suspension cultures. 112 3Co Origin of the BHK21PD4 monolayers and suspension cultures. 113 4C. Karyotype of a cell from 128 L-M grown on modified medium 199 plus 0.5% peptone and 2.5% fetal calf serum in monolayer culture. 116 5C. Karyotype of a cell from NB 7/24 L-M grown on modified medium 199 plus 0.5% peptone and 5% fetal calf serum in fermenter culture. 117 6C. Karyotype of a cell from 2884 L-M grown on modified medium 199 plus 0.5% peptone and 2.5% fetal calf serum in spinner culture. 122 7C. Karyotype of a cell from NB8 BHK21PD4 grown on Schedule 8 spinner medium in fermenter culture. 125 8C. Karyotype of a cell from 72 BHK21PD4 grown on Eagle MEM (supplemented with 10% Tryptose Phosphate Broth) plus 5% calf serum in monolayer culture. 126 9C. Karyotype of a cell from 72 BHK21PD4 grown on Eagle MEM (supplemented with 10% Tryptose Phsophate Broth) plust 5% calf serum in monolayer culture. 127 10C. Karyotype of a cell from 80 BHK21PD4 grown on Eagle MEM (supplemented with 10% Tryptose Phsophate Broth) plus 1% calf serum in monolayer. 129 xi

LIST OF FIGURES (Concluded) Figure Page 1D. Utilization of glutamine and production of ammonia by L-M cells in monolayer cultures in 2X Eagle basal medium. 137 2D. Production and utilization of nonessential amino acids by L-M cells in monolayer culture in 2X Eagle basal medium. 138 3D. Production of the nonessential amino acids glycine and serine by monolayer cultures of L-M cells in 2X Eagle basal medium. 139 4D. Protein production by L-M cells in monolayer cultures of L-M cells in 2X Eagle basal medium with and without added nonessential amino acids. 141 5D. Protein production by L-M cells in monolayer cultures in 2X Eagle basal medium with and without 4X cystine. 143 6D. Glutamine decomposition at various time intervals in uninoculated 2X Eagle basal medium incubated at 370C. 144 lEo Cell yield of BHK21PD4 and pH decline against initial glucose level. 152 2E. Spinner growth of BHK21PD4 cells from frozen stock. 155 3E. Supernate levels of serine and glycine during growth of BHK21PD4 cells. 160 4E. Serine uptake during growth of BHK21PD4 cells. 161 xii

INTRODUCTION The pruposes of this project were (1) to determine the degree of drift in cell populations during continuous cultivation and during build-up from seed stocks to fermenter growth, (2) to determine the major causes of drift, and (3) to explore control procedures for minimizing these variables. L-M mouse fibroblasts and BHK21PD4 hamster cells were used as models. The former is heteroploid, the latter is near diploid. Growth rates, nutritional requirements and metabolic behavior were examined. Markers used to follow population changes have been (1) chromosomal (number, kind, markers, etc.), (2) alkaline phosphatase activity, and (3) colony type. The project had to be terminated prematurely when Dr. Merchant left Ann Arbor to accept the position of Director of the W. Alton Jones Cell Science Center, Lake Placid, New York. 1

THE POPULATION DOUBLING TIME OF STRAIN L-M MOUSE CELLS IN SUSPENSION CULTURE AS A FUNCTION OF CONSTANT, CONTROLLED REDOX POTENTIAL Jens U. Pedersen

ABSTRACT Strain LM mouse fibroblasts was grown in suspension culture in a modified, 5-liter New Brunswick fermenter in antibiotic free medium 199P SP FCS5 under constant, controlled redox potential, pH = 6.80 ~.06, and temperature = 35.0 +.5~C. Also kept constant were the agitation rate = 145 rpm, influent gas flow rate to the overlay = 300 ml/min, and carbon dioxide concentration in the influent = 5% v/v. Dissolved oxygen was monitored, The redox potential was controlled via variation of the partial pressure of oxygen in the overlay. The population doubling time during the logarithmic growth phase was determined as a function of the redox potential in two separate fermentations of 48 to 46 days' duration each comprising eight growth curves. During the first fermentation, with pH control via o32N HCL and.32N NaOH, the redox potential was varied in the range Es = 5 - 146 mV, where Es is the potential measured against the silver-silver chloride-saturated potassium chloride reference electrode, and the resulting population doubling time determined from each growth curve varied from 47.5 hr (from total count or 45.0 hr from viable count) at E* = 29 mV through a minimum of 29.0 hr (or 27.5 hr) at Es = 56 mV to 46.0 hr (or 45.3 hr) at E5 = 119 mV. During the second fermentation, with pH control via.3: NaHC03 and gaseous carbon dioxide, the redox potential was varied in the range Es 29 - 211 mV, and the population doubling time varied from 66.5 hr (or 66.5 hr) at Es = 35 mV through a minimum of 24 hr (or 23 hr) at Es = 176 mV to 65.0 hr (or 70.0 hr) at Es = 202 mV. *See Definitions on page 30.

INTRODUCTION A fermentation run (NB-15) was carried out in the 5-liter New Brunswick fermenter with the objective to confirm the results of run NB-11 reported in the previous progress report (Merchant, 1969). Due to severe toxicity problems and occurrences of oxygen probe membrane ruptures three unsuccessful runs were attempted. The toxicity occurred without any changes in procedure from previous runs, but on the basis of the observations described below (under Toxicity), it was decided to introduce some changes in the set-up used for pH control before attempting a subsequent run. The differences between NB-11 and NB-15 are described below. MATERIALS AND METHODS CELL STRAIN The cells used in the fermenter experiments are strain LM mouse fibroblasts and have been carried continuously in monolayer and spinner cultures for 1-1/2 years on medium 199P SP FCS5. MEDIUM The medium used in designated 199P SP FCS5, the meaning of which appears below: 1. 199: 199 refers to the experimental medium no. 199 derived by Morgan et al. (1950), and modified by Salk et al. (1954), in both method of preparation and composition of base solution. The method of preparation from the single components is described in Merchant et al. (1964), and the base solution is Hanks balanced salt solution (using.05% sodium bicarbonate) instead of the original Earle balanced salt solution. The term M150 has also been used (Morgan et al., 1955) to signify medium 199 with Hanks base solution. Helen J. Morton (1968) advocates the term M150. In this report the designation medium 199 is used as a continuation of usage in previous reports. 2. P (Peptone): The medium contains.5% w/v (weight/volume) Bacto Peptone (Difco), lot no. 1446035. 3. SP (Special): A growth performance similar to that of the complete medium 199 was found when the following components were omitted: 5

D - ribose Ascorbic acid D - deoxyribose Vitamine A Glutathione Adenosine Triphosphate L-cysteine. HC1 Penicillin Streptomycin This special medium, 199 SP, has been used throughout in the fermenter experiments. It was purchased as instant powdered medium from Grand Island Biological Company, N. Y., corresponding to their formula E12 (with Hanks base and glutamine, without sodium bicarbonate) but without the nine components listed above (lot nos. 674537, 685482, 685630, 930077, 930143). 4. FCS_ (Fetal Calf Serum, 5% v/v (volume percent)): The fetal calf serum was obtained from Hyland, Division of Travenol Laboratories, Inc., Los Angeles, California, and stored at -20~C until used. (Lot nos. 3135D061C1, -71C1, and -68C1 have been used.) MEDIUM PREPARATION A normal procedure for dissolving the powdered medium 199 SP and adding the bicarbonate, peptone, and fetal calf serum was followed. However, in order to eliminate or minimize variations in medium component concentration, osmolality, pH, ORP (D.O.), and other not known factors influenced by incubation and storage, the following points were observed: 1. Glass distilled water of high resistivity was used (~1 Mohm-cm). 2. To eliminate excess oxygen, the water was flushed with nitrogen or freshly boiled or autoclaved and cooled to ambient temperature before use. 5. The powder was dissolved in a partial volume of water such that the concentration was lower than 5x. Above this concentration all constituents in the medium will not dissolve. 4, The peptone was added as a freshly autoclaved and cooled 5x solution. The autoclaving of peptone was found to improve growth. 5. After adding all ingredients the pH was routinely checked before filtration and required to be 6.8 ~ 0,07 as an indication of consistency in preparation. 6. Filter sterilization took place under nitrogen, again to avoid medium oxidation, at 2-5 psig through an absolute filter or at 10-15 psig through a.22. membrane filter. 6

7. Sterility check of the medium was carried out by incubation at 3536~C routinely for 3 days and then it was stored at 4'C until used. The medium for monolayers and spinner flasks was kept in tightly closed, screw-capped prescription bottles to avoid oxidation (increase in ORP) and increase in pH by equilibration with air. The medium to be used in the fermenter was kept in receiving flasks (in essence, Erlenmeyer aspirator flasks with an upper hoseadapter) where the medium was allowed to equilibrate with the surrounding air through a cotton or glasswool filter. Conditioning (ORP - D.O. and pH) of the medium took place in the fermenter within 1 to 2 hr after charge. COMMON EXPERIMENTAL CONDITIONS FOR THE CONTROLLED FERMENTER RUNS The following conditions for a fermenter run were decided upon prior to the entire experimental series: 1. Temperature: 35TC was chosen on the basis of previous growth at that temperature in monolayers and spinners. It gives a reasonable high growth rate and diminishes the risk of impairment caused by accidental temperature fluctuations above 39~C, which is detrimental to the cells. The temperature was maintained at a constant, controlled level of 35.0 + 0.5~C by means of an external water bath. 2, Agitation Rate: 145 rpm has proven to provide adequate but gentle agitation to keep the cells suspended and to provide for sufficient gas transport across the surface of the medium and adequate mixing for pH control. 3, Gas Flow Rate: The influent gas flow rate from the gas mixing set up to the overlay in the fermenter was chosen as 300 ml/min. This flow rate provides a fast enough renewal rate in the overlay to accomplish the desired ORP controlling action, without causing excessive medium evaporation. 300 ml/min is the lower limit for accurate flowmeter settings within the ranges used for the partial flows: 15 ml/min to 285 ml/min. It was decided not to sparge in order to avoid foaming, denaturation and unnecessary stripping of volatiles. 4. pH: 6.8 was chosen as the initial level for constant, controlled acidity under the assumption that lactate production, i.e., glycolysis, would be low (Willmer, 1966). Further, a carbon dioxide concentration of 5% v/v is in equilibrium with the bicarbonate concentration of medium 199P SP FCS5 at pH = 6.8. 5a C02: Carbon dioxide is an important intermediate metabolite in the TCA cycle and has been used widely in concentrations corresponding to 7

a 5% v/v gas mixture. 6. ORP: The redox potential of the medium and suspension culture was measured with a platinum electrode versus a reference electrode having an Ag/AgCl/AgClSAT, KC1sAT internal element and solutions and a palladium tip junction. The redox potential is given as ES in millivolts at any existing physicochemical conditions in the suspension, where the subscript S refers to the silver-silver chloride (etc.) reference electrode. The reported values, ES, are uncorrected for the normal electrode potential EO,Ag/AgCl, etc,, which at 25~C is approximately 185 mV for the specific reference electrode used in the fermenter experiments. This correction has varied somewhat during the period reported but not more than about 15 mV. Hence, to obtain the redox potential in terms of Eh (relative to the hydrogen scale) all ES values should be increased by 185 ~- ~8 mV. A more precise correction is being calculated. During growth experiments the ORP was maintained at selected constant, controlled levels initially suggested by the results of monitored runs and ultimately by run NB-ll. 7. D.O.: The dissolved oxygen was monitored during the fermenter experiments and is expressed as po2 in mm Hg, the partial pressure of oxygen in the gas phase which is in equilibrium with the suspension. The po2, which corresponds to 100o saturation with air, is a function of the barometric pressure and the water vapor pressure above the suspension. The normal barometric pressure at the elevation where the experiments were conducted is 736 mm Hg, Under the assumption that the gas overlay in the fermenter is saturated with water at 35~C (PH0o = 42 mm Hg) the normal total pressure of atmospheric air is Ptot = 694 mm Hg and P2 = 0.2095 x 694 = 145 mm Hg. The basic conditions for the controlled fermenter runs as described above were common for both runs NB-11 and NB-15. Due to toxicity problems mentioned elsewhere in this report, it was decided to change the pH control set-up. This resulted in the following differences between NB-ll and NB-15 experimental conditions (see Table IA). PROCEDURE FOR A RUN The platinum-reference electrodes were calibrated with a poising standard solution: 1. 1 M Fe(CN)Fe(CN)6/K4Fe(CN)6 in N M KC1 300 10 which has a potential, Eh = 430 mV at 25~C. This was done before assembling the fermenter and again after termination of the run to observe possible drift of the reference electrode potentials 8

TABLE IA. Differences in Experimental Conditions and Systems in Two Fermenter Runs, NB-11 and NB-15 Fermenter Run Condition or System NB-11 NB-15 Base addition 0-32N NaOH 0o32N NaHC03 Acid addition 0o32N HC1 C02 added via influent gas Buffer system Phosphate: initial.012% Phosphate: initial.o012% Bicarbonate:.05% Bicarbonate: normal range during run: 0.05-0.16% Fermenter content during autoclaving 1.5 1 glass distilled 1.5 1.5% peptone water water After autoclaving the D.O. probe was calibrated with nitrogen and air at 35~C with glass distilled water or.5% peptone still in the fermenter. The fermenter content was expelled, save for 100 ml and the fermenter was charged with 199P SP FCS5 to 1500 ml giving a.93 x strength medium. This was conditioned to the desired redox potential and pH via the control system. A grab sample measured with a laboratory pH meter established standardization of the fermenter pH. Inoculation from a spinner flask with a culture in midlog phase (-1.5-2.0 x 106 cells/ml) brought the volume to about 1600 ml with an estimated population density of 2.5 x 105 cells/mli Samples were taken at varying intervals dependent upon the actual growth rate. Sampling and Distribution of Sample Volume A sample volume of 19-2Oml was drawn with a sterile syringe and special needle with pointed tip through the silicone rubber diaphragm sampler. About 1 ml was discarded to avoid contamination with Dakinfs solution used to sterilize the sampler before each sampling. Two ml was delivered to each of three silicone* treated viability tubes' Silielad, Clay-Adams, Inc,, New York. 9

for later cell count and viability determination. Three ml was delivered to a narrow test tube kept in a 355C water bath for pH measurement with a thin probe combination electrode immediately after the distribution of the sample volume. One ml was delivered to a l-dram vial with rubber lined screw cap kept in an ice bath. One was frozen at -20~C and stored for later glucose and lactate analyses. The other was stored at 4~C for subsequent inorganic phosphate (Pi) analysis. The sampling took 20-30 sec and the distribution of the sample volume was carried out within 1 min. The pH measurement was done within 3-4 min from sampling time. Cell Count From each of the viability tubes was taken 1/2 ml aliquot for counting with a Coulter counter* in a conventional manner. Each aliquot was counted six times giving a total of eighteen determinations for the cell concentration. Viability was determined by counting in a hemocytometer about 600 cells of the suspension mixed with Erythrosin B. Cells that appeared red were considered nonviable. PCV The packed cell volume was read from the thrombocytocrit tube. The packed cells were transferred to a screw capped test tube and washed with.85% saline and centrifuged for 5 min at 540 G. The supernate was discarded and.5 ml of distilled water was added to the pellet. The tube was sealed and frozen at -20~C for eventual nucleic acid, protein, and phospholipid determination. The supernate from the thrombocytocrit tube may be used for osmolality measurements. TOXICITY The following sources are the most likely causes of the observed growth impairment. While not immediately discernible, prolonged or repeated exposure to corrosive action by base, acid, and autoclaving gradually developed the problem. *Coulter Electronics, Hialeah, Florida. 10

1. Cu: Faulty construction material in a luer-lok needle used as inlet for acid and base to the fermenter: While ordered as 316 st. the luer-lok was delivered as a standard nickel plated brass part. The canula itself was stainless steel as ordered. The plating was corroded by prolonged exposure to hydrochloric acid and sodium hydroxide and copperhydroxide was formed. 2. Fe: Incompatibility of different kinds of stainless steel in the fermenter re-electrolytic corrosion and exposure to corrosive compounds (e.g., Cl, 02, HC1, NaOH) and extreme low pH during autoclaving resulting in Fe formation and precipitation of Fe(OH)3. 3. Ag: Electrolytic corrosion of the silver anode of the dissolved oxygen probe due to probe membrane rupture and to electrical potential differences between the anode and the grounded parts in the fermenter. RESULTS The experimental findings of run NB-11 are compiled in Table IIA and graphically represented in Figures 1A and 2A. The population doubling time was determined from both total and viable counts. Refer to the presentation of run NB-15 for a description of how the values and graphs were derived from the basic growth curves, measurements, and recorder traces. The experimental findings of run NB-15 are compiled in Table IIIA and graphically represented in Figures 3A and 4A, Below follows an outline of the construction of the graphs presented in Figures lA-4A. As an example is used growth curve no. 1 of run NB-15. This growth curve is representative as for the precision with which the PDT and Es are obtained. Other growth curves yield narrower or wider ranges and spreads for PDT and Ese It was decided to present both PDTT and PDTV vs. Es since the viability varied during some growth curves. 1. Figure 5A. From this growth curve was determined the slope of a straight line that could be drawn through the points that represent the logarithmic growth phase. When the points were few or had a deviation from a straight line a spread for the value of the slope was obtained, The population doubling times, PDTT and PDTV, were determined from PDT = Slope Slope 11

TABLE IIA. Run NB-1l: Population Doubling Times Determined from Total and Viable Counts at Selected, Con stant Levels of the RedoxPotential for Eight Consecutive Grwth Curves (Conditions Described n Section "Common Experimental Conditions" and in Table IA) Redox Potential PDT Based on Start/End PDT Based on in Terms of EsGrowth SatE Total Count Viable Count Length of E Curve ~sEs Curvot LntoMost Range Most Range Ee Growth Curve Most Rang e N. Grwt Cre Prob. Lo/Hi Prob. Lo/Hi Start/End hr Pr edom. hr mV o.,o/ 131.1 1 00151 42.3 42.1/42.5 42.5 42.1/42.5 82 91/ 68.5 111/56 151.1 131.1/ 276.1 2 15.1 261 46.0 45.8/46.2 45.3 45.1/45.5 119 137/105.5 146/95 145.0 276.1/ 380,0 2. 3276.1/ 380.0 29.5 29,0/30.0 28.5 28.5/29.0 56 61/ 52.0 80/40 103.9 Ro 380o0/ 539.3 4 500 35.0 34.8/35.2 35.0 34.8/35.2 61 75/ 55.0 91/42 159.5 559.3/ 714o,4 5 559.5/ 714.4 47.5 47.5/48.0 46.0 45.8/46.2 29 42/ 15.0 50/05 175.1 v.4.4/8 8o84 6 144.0 31.0 ~~29,o/>1.o 3o.o 29.5/>0.0 5o 75/ 38.0 -Loo/5o 858.4/1019.8 7 858.4/1019.8 32.5 31.5/33.0 32.6 31.5/52.6 41 70/ 28.0 82/10 161.4 1019 ~8/1164.3 8 1019.8/1164.5 29.0 28.0/29,5 27.5 27.3/27- 7 56 64/ 47.0 74/35 144.5

TABLE IIIA. Run NB-15: Population Doubling Times Determined from Total and Viable Counts at Selected, Constant Levels of the Redox Potential for Eight Consecutive Growth Curves (Conditions Described in Section "Common Experimental Conditions" and in Table IA). (Included are Mean Values of pH and Dissolved Oxygen.) Redox Potential.... DTI Based- on DO Start/Etnd in Terms of Es Mean Gro Legtth fTotal Count Viable Count Mean GrwhLength of E Cu~~~a~~e MoLst....e Mest a Curve Growth Curve Most Range Most Range s s E Value Mean No. oProbCProbe Most Range No. hProb. Lo/Hi Prob. Lo/Hi Start/EndMa i Start/EM hr hr Predom. Max/Min hr MVmm mmg mV o0./ 164.3 6.81.1 O~o/ 164-3 34.0 33-5/35.5 34.0 33-5/35.5 53 59/ 46 68/ 35 $15/7-5 164.3 ~.06 16453/ 2505 6.79 2 86.2 28.5 28.0/28.5 28.0 28.0/28.5 119 122/113 133/104 3/44 86,2 ~t~~~~~~~~~~~~~~~~~~~~~,02 250-5/ 418.4 6~30 3 25075/ 48 66.5 64.0/68.5 66.5 64.0/68.5 35 41/ 33 46/ 29 6/2.2 167.9.+,04 418o4/ 514,7 44 96.4 54 29,5 28.0/29.5 29.0 ~ 29 87 94/ 84 94/ 78 2/22677 96.3 ~+.03 514,7/ 625.5 6.8o ~5 514- 655 24.0 24.0/26.0 24.0 24.0/26.0 168 177/155 182/155 21/>145.0 llO,8 -+.O5 625.5/ 754,3 67 6 6255/ 754. 65.0 64.0/66.0 70.0 70.0/72.0 202 208/197 211/187 197/494 6 128,8. 4 754,3/ 970.8 7*:/970.8 46o0 ~ 46 36,0 3550/37,0 119 120/117 12/114 -.82 216,5+.o 970.8/1107.7 6.81 8 156.9 177 24.0 24.0/24.5 23.0 23.0/23.5 176 180/170 184/160. 136.9

Figure 1A. Run NB-11: PDTT vs0 Es. The population doubling time of strain LM mouse fibroblasts in suspens sion culture based on the total count as a function of constant, cointrolled redox potential in medium 199 special supplemented with.5% peptone and 5% fetal calf serum (conditions described in section "Common Experimenrtal Conditions" and in Table IA). Legend: #1: Values obtained from growth curve no. 1 * Coordinates (Es, PDTT most prob.) i.e., most predominant Es during pertinent part of growth curve, most probable kDT during log phase _///_///__/_' Width: Range of Es during pertinent part of growth curve Height: Spread of PDTT as determined from slope of log plot of growth curve during log phase: Horizontal line, entire range of Es encountered during growth curve: Line between points: Most likely relationship between PDTT and Es

48 0 44 40 36 32 28 0 50 100 150 ESPmV Figure lA.

Figure 2A, Run NB-11: PDTV vs EO The populationdoubling time of strain LM mouse fibroblasts in sspension culture based on the viable count as a function of constant, controlled redox potential in medium 199 special supplemented with 5% peptone and 5% fetal calf serum (conditions described in section "Common Experimenrtal Conditions" and in Table IA). Legend: #1: Values obtained from growth curve no. 1 *: Coordinates (Es, PDTV, most prob.) i.e., most predominant E during pertinent part of growth curve, most probable PDT during log phase V/////_/l///' Width: Range of Es during pertinent part of growth curve Height: Spread of PDTV as determined from slope of log plot of growth curve during log phase ~ Horizontal line, entire range of E encountered during growth curve: Line between points: Most likely relationship between PDTV and Es

OI 48 ~ 44F - _ 40 32 — - 28-50 0 50 100 150 Es, mV Figure 2A.

Figure 3Ao Run NB-15: PDTT vso EO The population doubling time of strain LM mouse fibroblasts in suspen sion culture based on the total count as a function of constant, controlled redox potential in medium 199 special supplemented with.5% peptone and 5% fetal calf serum (conditions described in section "Common Experimental Conditions" and in Table IA). Note: Point no. 7 does not belong to the family of the other points due to a higher CO2 content in the gas phase during the growth. It is included only as a comparison. (See comment at bottom of Table IIIA.) Legend: #1: Values obtained from growth curve no. 1: Coordinates (Es, PDTT, most prob.) i.e., most predominant Es during pertinent part of growth H curve, most probable PDT during log phase ////////// ~: Width: Range of Es during pertinent part of growth curve Height: Spread of PDTT as determined from slope of log plot of growth curve during log phase: Horizontal line, entire range of E encountered during growth curve s Line between points: Most likely relationship between PDTT and Es D.O.: The range of dissolved oxygen from start of growth curve till the time of the last sample used to determine the PDT, mm Hg

70 (~) D.O.:0. 6 —2. 2 ~ D. 0.:197-494 60 D.( 0. ( 530 20 0 50 100 150 20 0 50 100 150 200 Es, mV Figure 3A.

Figure 4Ao Run NB-15: PDTV vso E.. The population doubling'time of strain LM mouse fibroblasts in suspension culture based on the total count as a function of constant, controlled redox potential in medium 199 special supplemented with.5% peptone and 5% fetal calf -serum (conditions described in section "Common Experimental Conditions" and in Table IA.) Note: Point no, 7 does not belong to the family of the other points due to a higher C02 content in the gas phase during the growth. It is included only as a comparison. (See comment at bottom of Table IIIA.) Legend: #1: Values obtained from growth curve no, 1: Coordinates (Es, PDTV, most prob.) i.e., most predominant Es during pertinent part of growth curve, most probable PDT during log phase ro 0 _________|: Width: Range of Es during pertinent part of growth curve Height: Spread of PDTV as determined from slope of log plot of growth curve during log phase: Horizontal line, entire range of E encountered during growth curve Line between points: Most likely relationship between PDTV and Es D.O, The range of dissolved oxygen from start of growth curve till the time of the last sample used to determine the PDT, mm Hg

70 60 "r~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ C~~~~~af O~~~~~~o. 2 22 D. 0..3 30~~~~~~~~~~~~~~~~~~~~~~0 20 t- -~-100.F.1gre 4A,.

Figure 5A. Run NB-15: Growth curve no. 1: Total and viable cell count and PCV vso time. Legend: L: Total cells/ml O: Viable cells/ml: PCV, % giA: Estimated total cells/ml on basis of a previous count in connection with dilution of medium*' Estimated viable cells/ml on basis of a previous count in connection with dilution of medium*, Estimated PCV, %, on basis of a previous count in connection with'-,,' dilution of medium*: Vertical broken line: dilution of culture. Signifies start of new growth curve or termination of run. *~See also last entry. 22

107 1 | 1. 0 - Total, No. /ML - Viable, No. I/ML O PCV, I lf6. 1 10 -- o105!1O5 1 I~~~ I ~ I o.01 0 50 TIME HR 100 150 Figure 5A. 23

or by measuring the im rcrement of the abscissa corresponding to a doubling of the ordinate. Also the PCV points were plotted and the slope of the line through those points were used in deciding the logarithmic phase and hopefully supporting the PDT. (For this figure the PCVDT = 34 ~ 1 hr was comparable to the PDT = 34 + 1.5/0.5 hr.) The values so obtained were tabulated' in Tables IA and IIA. 2. From the continuous recordings of Es, D.0., and pH were plotted a condensed history of the mean and range of these variables. The recorder traces were corrected for drift of analyzers, recorders and, where possible, electrodes. The pH electrode was standardized at each grab sample from the fermenter. The Es and D.O. electrodes could be calibrated only before and after each fun. It was found that the ORP analyzer was stable to ~ 2 mV over 1-1/2 years. As mentioned under "Common Experimental Conditions," the reference electrode drift amounts to about 15 mV over 1 year despite several autoclavings. The reported values for Es have not been corrected for electrode drift. The D.O. electrode's service life was in one instance long enough to recheck the calibration after a run: A drift of - +5% was observed. The D.O. values must therefore be considered reliable. An Average mean (= Es) was then determined from the Es vs. time plot (Figure 6A) for the period from where the variable reached the control range until or close to the final sample used to determine PDT. This value, Es, is considered the most predominant during the growth. However this concept is still subject to interpretation. Likewise was done for pH. The range of D.O. from start to final determining sample was tabulated in Table IIIA and indicated as a variable parameter on Figures 3A and 4A. 3e All the values for Es, Es, and Es and PDT, i.e., most predominant Es, most probable PDT including the spread and range about these values were plotted in a PDTT vs. Es and PDTV vs. Es graph. 4. Also Pi concentration in the cell free medium vs. time was plotted (Figure 7A). Note the constancy of (Pi) during the initial lag phase where the population density and PCV increase slightly. Further (Pi) seemed to decrease linearly between 30 and 80 hr where after exponential decrease occurred. 24

Figure 6A. Run NB-15. Growth Curve no. 1: pH, E, D.O. (mean and range) vs. time. Legend:: Mean trace of variable * Upper and lower range of variable: Time of sample no. III: NB-15 Vertical broken line: dilution of culture. Signifies start of new growth curve or termination of run 26

7. 2 7.0 6. 8 6. 6 200 180 160 140 120 100 680 40 20 20 - E Ei \ 140 120 100 80 60 40 20 O0 0 50 100 150 TIME, HR Figure 6A. 27

40 38 36 34 I 32 I E I 30 28 26 24 22 0 50 100 150 TIME, HR Figure 7A. Run NB-15: Growth Curve no. 1: Pi vs. time. Legend: O: Pi, inorganic phosphate concentration in the cell free medium, mg/liter

DISCUSSION There appears to be an overall qualitative similarity between the two plots of PDT vso ES for the two runs NB-11 and NB-15. The curve appears concave in the approach to the indicated PTDmin and convex toward higher ES values. (Note: even though NB-15 does not definitely confirm this, it certainly does not negate convexity.) The quantitative deviation is attributed to at least two factors: 1. Change in conditions are listed in Table IA. 2. Possible change in the cell population during the time elapsed between NB-11 and NB-15, despite a uniform procedure in maintaining the cells for inoculation into the fermenter. Within one run comprising eight growth curves it was possible to establish the relationship between PDT and ES as shown and to essentially duplicate a point on the "curve": growth curves nos. 3 and 8 in run NB-11 and growth curves nos, 5 and 8 in run NB-15. The exactitude of the ES points nos. 3 and 8 in NB-11 of course may be subject to interpretation: should they both be 56 mV? However, there is a strong indication of a minimum PDT at about 55 mV and it certainly could be anywhere from - 51 to 55 mV yielding maybe an even lower PDT. For run NB-15 a minimum is indicated between 176 and somewhat below 202 mV. The data reported does not allow any conclusion whether a lower PDTV than 23.0 hr exists in this run. The constancy of PDT versus constant ES during a growth curve was maintained with an in general exponentially increasing DO. The concept of a constant, controlled redox potential in tissue culture as a practical and useful tool in manipulating the growth rate to obtain constancy over an extended period seems strongly indicated by the reported experiment s. I interpret the quantitative difference between the two runs to indicate the necessity of establishing the PDT vso ES relationship for each particular system in question, but when this is done the data would enable a prediction of growth rates as needed. 29

DEFINITIONS ES: Redox potential as measured against the silver-silver chloride/sat. silver chloride, sat. potassium chloride/palladium junction reference electrode at any existing physico chemical conditions in the suspension, mV ES: (also ES mean) the mean value of ES over a period of time where ES varies in a cyclic manner (stemming from control action) and exhibits a reasonably constant trend. Also shorter, aperiodic fluctuations of ES are considered valid as a basis for ES, mV ES: The most predominant value of the ES during the part of the growth curve where ES has affected the growth rate on which is based the determination of the population doubling time, mV PDT: Population doubling time, hr PDTT: Population doubling time based upon total count, hr PDTV: Population doubling time based upon viable count, hr Pi: Inorganic phosphate ORP: Oxidation reduction potential. Also, redox potential, mV 199M: Short term for 199P SP FCS5 = medium no. 199 special supplemented with.5% peptone,.05% sodium bicarbonate, and 5% fetal calf serum, (Refer to section on "Medium" for explanation of "Special" and a discussion of this medium.) PCV: Packed cell volume, % PCVDT: Doubling time of the culture based on PCV (- packed cell volume doubling time), hr 5o

REFERENCES 1. Merchant, D. J. "Large Scale Fermenter Growth of Animal Cells for Virus Vaccine Production: Control of Seed Stocks and Growth Conditions (U)." Ann. Progr. Rep., U. S. Army Medical Research and Development Command, Contract No. DADA-17-67-C-7073, February 1969. 2. Merchant, D. J., R. H. Kahn, and W. H. Murphy. "Handbook of Cell and Organ Culture." Burgess, 1964. 3. Morgan, J. Fo, M. E. Campbell, and Helen J. Mortcn. "The Nutrition of Animal Tissues Cultivated in Vitro. I, A Survey of Natural Materials as Supplements to Synthetic Medium 199." J. Nat. Canc. Inst. 16, 557 (1955). 4. Morgan, J. F,, Helen J. Morton, and R. C. Parker. "Nutrition of Animal Cells in Tissue Culture. I, Initial Studies on a Synthetic Medium." Proc. Soco Exper. Biolo and Med. 73, 1 (1950) o 5. Morton, Helen J. in Tissue Culture Association Newsletter, Vol. 1 (No. 4), p. 2, September 1968. 6. Salk, J, E., J. S. Younger, and E. N, Ward. "Use of Color Change of Phenol Red as the Indicator in Titrating Poliomyelitis Virus or Its Antibody in a Tissue-Culture System," Amer, J, Hyg. 60, 214 (1954). 7. Willmer, E. N, (Ed.). "Cells and Tissues in Culture." Vol. I, p. 258, 2nd printing, Academic Press, New York, 1966. 31

CHROMOSOME MEASUREMENT: ORGANIZATION AND APPLICATIONS OF A WORKING SYSTEM Gregory Baumann, Richard Giles, Edward Lunk, and Richard Platte, Jr. 33

INTRODUCTION A number of investigators have reported on the application of various measurements to the study of chromosome morphology, particularly with reference to the development of automated techniques for chromosome analysis of diploid cells (1-5). However, such systems are generally quite complex, require highly sophisticated computer programming, and, at this time, are extremely expensive to operate. In addition, these systems do not provide adequate descriptions of chromosomes in mixoploid cells due to the fact that any analysis of a mixoploid population must take into account the widely varying chromosome numbers and morphologies which are present in different cells within the population. Our efforts have been directed at developing an easily applied, sensitive system for providing a quantitative description of a cell population based on the measurement of chromosome length in a small sample of cells using semiautomatic measuring equipment and digital computer techniques. One goal in this system has been to reduce the importance of karyotyping and other techniques which are primarily descriptive of single cells and to stress those features which appear to be characteristic of the population as a whole. So, although this system may be used to study diploid cells (especially in following the processes of transformation which these cells undergo), it is most useful in the study of mixoploid cells where differences between single cells are of much less importance. The specific methodology and computer programming may apply only to the exact equipment configuration being used here, but the types of equipment used are widely available, and most aspects of the logical design and methodology may be applied to any system. Chromosome measurements are made from the projected images of 35-mm negatives using an X-Y coordinate digitizer on-line to a small computer. This equipment produces magnetic tapes containing digitized data on chromosome spreads which are further processed on a large computer system. Final output is obtained from the machine through a high speed printer, teletype terminal, and CALCOMP plotter. The entire system for preparing, digitizing, and processing chromosome data will be described. EQUIPMENT AND METHODS Metaphase chromosome preparations are made using a modification of the technique described by Giles, Merchant, and Masselink (6) in which cell cultures in the logarithmic growth phase are treated with colcemid to stop growth in metaphase and then are expanded in a hypotonic solution of sodium 34

citrate and fixed in an acetic acid-methanol mixture (see Appendix I). Microscope slides are prepared by flame drying of the fixed cell suspension and are stained with Giemsa stain.1 Inasmuch as the computer methods to be described are being used at this time as an adjunct to standard chromosome analysis procedures, all microscope slides are first scanned for high quality metaphase plates using a Zeiss photomicroscope.2 Only those spreads fitting the criteria given in Table IB are used for extensive analysis although some spreads of lesser quality may be counted. TABLE IB. Criteria used to judge a chromosome spread before using it for extensive analysis procedures. 1. Chromatid coiling, the centromeric constriction and any secondary constrictions which are present must be clearly visible. 2. If any chromosomes are found to be overlapping in the spread, the overlap must not conceal the centromere or chromatid ends of the chromosomes involved. 3. The spread must be confined to a sufficiently small area of the microscope slide for the technician to rule out the loss of chromosomes in spreading or the presence of chromosomes belonging to other spreads. The standard procedure used to study acceptable chromosome spreads consists of making differential chromosome counts (i.e., counting separately the numbers of biarm chromosomes,telocentric chromosomes, total number of chromosomes, and total number of arms), visually identifying any marker chromosomes present, and noting the presence of abnormal chromosome morphologies. This procedure is carried out on 50-100 metaphase plates from each population examined, and those of highest quality are photographed for further study. 1Fisher Scientific Company, Fair Lawn, New Jersey. Code number SO-G-28. Carl Zeiss, Germany. 35

All photographs of mitotic figures used for measurement studies are produced using a 100 x oil immersion objective, an Optovar factor3 of 1.25 and a projective factor3 of 3.2. This optical configuration produces an image on the 35-mm film surface 400 times the size of the original object. Using the automatic exposure feature of the microscope, black and white negatives of consistently high quality are easily produced on Kodak high contrast film.4 Exposed films are developed according to standard dark room techniques (see Appendix II). Although prints are occasionally made for the production of karyotypes or visual presentation of data, all subsequent analysis is carried out using the projected images of the film negatives. Lengths are measured from these images using an X-Y coordinate digitizer. This equipment, which is used to automatically record the X and Y coordinates of any point on a planar measuring surface, was designed and built in the Department of Physics at The University of Michigan where it is used to study photographs of bubble chamber tracks (see Figure 1B). Equipment of this type enables one to easily translate analqg data on a photographic image into a digital form.5 THE DDP-24 SYSTEM Six X-Y coordinate digitizers are connected on-line to a, 5K DDP-24 computer6 in The University of Michigan system. A data point is recorded by manually positioning a set of movable cross hairs over the point and then depressing a foot pedal. Movement of the cross hairs is transmitted through a system of steel bands and pulleys to a rotary pulse incoder which actuates a, bidirectional counter providing the digital X-Y location of the point to the DDP-24 computer through an eight-channel multiplexor. Supervisor and control programs in the system are written in FORTRAN and assembly code and provide for interaction with the user while the measurement is being carried out by means of a small console next to each digitizer. These programs are set up to provide real-time job scheduling with five different priority levels. Of course, the system has been designed specifically 3The terms Optovar factor and projective factor refer to optical equipment on the Zeiss photomicroscope which provide additional magnification in the light path to the photographic film. 4 Eastman Kodak Company, Rochester, New York. Code number M-135. Similar equipment is available commercially and can even be built with little difficulty (7). Honeywell Inc., Computer Control Co. Div., Farmington, Mass.

X-Y COORDINATE DIGIT IZER Figure 1B. A photograph of the X-Y coordinate digitizer used to make chromosome measurements. 57

for bubble chamber studies, and it is, therefore, necessary to describe each chromosome as if it consisted of charged tracks created by atomic particles all originating from a single point of vertex. An atomic event of this type is termed a neutral induced event and is described as having a number of prongs equal to the number of charged tracks seen to originate from the vertex (8). Figure 2B shows how chromosome structures can be redefined to fit the structure of a neutral induced event. An additional problem was encountered due to the fact bubble chamber tracks are treated as three-dimensional objects and are viewed from three different views on different films. The computer is programmed to automatically switch between films during the measurement of single events. Because chromosome photographs are normally taken from only one view it was necessary to change the system to allow the user to skip the program sequence which changes views. The Department of Physics did this and modified the program to omit a number of checks normally made on bubble chamber data. When applying this system, the user is prompted for the information needed by the computer through four display windows on the communication console (see Figure 3B). Replies to the machine are entered by depressing various combinations of the ten numerical keys and four termination keys or by depressing the foot pedal. The illumination of a large red light notifies the user of the existence of some error condition. It is possible to check and modify previously entered information through an alter sequence and to return to an earlier-step in the program sequence by sending a numerical entry followed by the "C" terminator. The user signals the machine that he has completed the requested information entering sequence by depressing the "P" terminator which signals the machine to "proceed." Although a sample session using the DDP-24 is given in Appendix III, it may be useful to describe such a session. The computer first requests a variety of items of set-up information including the user's identification number, the experiment number and roll number, each of which consists of a two-digit numerical entry followed by the "P" terminator, Then in order to alter the program sequence so that it will not switch between films, the alter sequence is used. The alter button ("A") is depressed, then the numerical g9" and, finally, the alter button is pressed again. The user is next requested to enter a frame number, event number (a number between 1 and 100, assigned to each chromosome measured in increasing order), and event type (either "2N" for a telocentric chromosome or "4N" for a, biarm chromosome, as described in Figure 2B). At this point the computer will request that the user measure the vertex (centromere) of the first track followed by from two to thirteen points located along the track by locating the cross hairs and depressing the foot pedal. The idea here is to select enough points along each track so that the total length can be approxima'ted by accumulating the distances between a series of short straight line segments (see Figure 4B)a When the end of a chromatid has been reached, the "'P" terminator is depressed, and the computer will request the vertex of 38

c. Because the incoming particle was invisible........... and 4 outgoing tracks / \are visible the event a. An incoming, / is termed a "neutral invisible,, induced 4 prong event" neutral par- s (abbreviated "N4"). ticle. b. At this point (the vertex) an interaction involving the neutral particle caused the emission of a number of charged tracks (or beams). d. Although an event like this might never be seen, it could also be described as an N4 /_ -____event. e. Notice the similarity between the structure in (d) and the biarm chromosome shown here. The centromere can be assumed to be the point of origin (vertex) of the chromatids which can be looked at as charged tracks. A biarm chromosome is described as an N4 event. f. Similarly, a telocentric chromosome can be described as an N2 event. Figure 2. Pictoijal development of the method used to describe chromosomes as bubble chamber events. Figure 2B. Pictorial development of the method used to describe chromosomes as bubble chamber events. 39

MESSAGE MESSAGE DISPLAY DISPLAY NUMERICAL NUMERICAL DISPLAY DISPLAY Numerical- Entry Keys Terminator Keys B C I I1 Figure 3B. Drawing of the console used for user/machine communications in the DDP-24 system. 40

the second track. Requests for vertices and points will continue until all the tracks for a chromosome have been measured, at which time the machine will again request a frame number. Although the system sounds complex, an experienced operator can easily measure 65 chromosomes in a single spread in 30 min, about half the time required to make the measurements by hand. Figure 4B. Sketches of a, number of chromosomes showing the locations of points which might be recorded, and the course which would be taken in locating the cross hairs on each of these points. The projected image of the negative from which all measurements are made is ten times the size of the negative itself. In the image plane the digitizer can discriminate between two points separated by a, minimum distance of 3 x 10-3 in. Therefore, if it were possible optically and photographically, the system would be capable of detecting two points separated by a distance (minimum) of 0.02 t on the original microscope slide. This is much greater accuracy than is actually necessary for chromosome measurement studies. Processed data from the DDP-24 computer is written on a magnetic tape through a Kennedy incremental tape drive.7 Since output is written after each track is measured, track records from different measuring machines are completely intermixed on the tape and must be sorted out in further processing. These tapes contain 24-bit words which must be translated for processing on machines using different word sizes. The format for magnetic tape, track records is outlined in Appendix IV. Kennedy model 140 OR. 41

Before magnetic tapes are made generally available for further processing, they are copied. The copy tapes are written as binary tapes on a seven-track tape drive at a density of 556 bytes/in. using odd parity. EXTRACTION AND ANALYSIS OF MEASUREMENT DATA All further analysis of measurement data is done using the IBM 360/67 computer at The University of Michigan Computing Center. Because this computing facility allows for operation from a remote terminal or in batch mode, many of the programs being used have been written to allow man/machine interaction. Basically, two programs are used. The first, called SEARCH, scans the DDP tapes for chromosome measurement data, checks this data for errors and then rewrites the information on a second magnetic tape. CHROMOSOME, the second program, reads the material produced by SEARCH and provides the desired information on each chromosome spread examined. Both are written in FORTRAN IV (SEARCH has one subroutine in assembly code), and both include a number of features only compatible with The University of Michigan Terminal System ('"MTS'), a supervisory system developed at the University. Being a modification of the program used to sort out measurement data in the Department of Physics, SEARCH (see Figure 5B), although it has been changed slightly in our laboratory, was originally written in that department. It is the goal of this program to scan the DDP tapes with a minimum amount of data processing and extract all chromosome data. Some difficulty arises due to the fact that chromosome data is only a small fraction of the total information on a tape, and, yet, at least part of every record must be translated from the 24-bit words used by the DDP machine to IBM compatible 32bit words. The program first translates the experiment number found in word two of the logical track record (see Appendix IV and Figure 6B). The entire record is only translated when the experiment number matches that used for all chromosome work. Next the program determines which machine was used to measure the track and transfers control of the program to a specific subroutine which handles data from that machine only. This provides the mechanism used to sort data which is completely intermixed because it was written in realtime into a meaningful form in which data is separated according to which machine it came from. Each new track record is first checked to determine if it represents data from a new chromosome or if one or more tracks have already been read. If the current track has originated from a new chromosome, all data for the preceding event is written on the output tape. Each event is also checked to make sure it is described as coming from a structure having an N2 or N4 configuration. Thus, the program is capable of detecting a number of errors in the data as it is being sorted. These errors, which are identified by the typing of two lines of information describing the event containing the erroneous information, are listed in Table IIB. 42

35 mm NEGATIVES OF CHROMOSOME. SPREADS MAGNETIC TAPE X-Y COORDINATE HONEYWELL DDP-24 DIGITIZER COMPUTER PROGRAM: "SEARCH" PROGRAM: "CHROMOSOME" TAPE R CORD PRGTRANSMTE ID I REA T1FE R UEND OF INPT Igu_ Bc on —t1 c m st ( DESCRIBINGUTd YES..L....M DSTAN IZE NO S NEW FRAME 7 AND CALEULT DATA ITRANS LATE ET -.ATISTICAL VALUES NEOUESER RCORD FROMDP _,,,, DXIS~ CSOM PUTER.... IBM 360/67 Fiur 5. asc rgniaio o te hrmoom eaurmet ysem(fo? LENGTBa-umann'and. Merchant) (11).A LAS

BAD (ST RTA r 1 EVENT START - - I~~~ 1 STORE ALL PROTOTYP1 /TRAC K ~ TRACK INFO. WHAT FILE.MACHINE I <FROM N2 or A ND X,Y DOES n SUBROU- I N4 EVE PAIRS INPUT TINE \ START IN? I I 0X no /ALL ON READDFILE CALL "STORE' ___m? NUMBER es - -- -- SALL'D, _____ I ST esTROM N AND "SKIP" \ RETURI\ STORED VENT TO POSITION _no INPUT TAPE \ - - I 4 L____________________ END HOW MANY FILES AND EVENTS ARE TO BE READ READ # OF eof 1 FILE 100,000 FILES,EVEN'S EVENTS "..... -es,ALL "IBUFI~"~iERo'. L IF ~eof(end of file) LES TO > TO READ TAP \ RECORD INTO RE \ BUFFER / no 6ALL "IBMDDPe' /CALL \ / ~ " A /\INRECD / TRANSLATE "R \/ AL' EXP. WORD T GET RECORD \ IBM WORD - FROM BUFFE MM MACH.NO Figur 6B lwc chromosome work assigned yesF' exp. 8 THERE AL. EACH A n {GO TO MM 6 ALL"IBMDDP~ ~CH ROUTIGE R~~GOUTINE\ TRANSLATE <TOOUTPUT r, p BAD 3 5 ENTIRE RECDLST E EVE TO IBM WDS/ ALL " ff PEM' CALL CALL /ETURN TO ~MACHl' 11fv1"ACH6'1 MTS CALL 0 CALL " S'MACH2'MC5 no CALL \ / CALL EVETyes END FMACH3 f tMACH."

TABLE IIB. The types of errors in chromosome measurement data which can be detected by SEARCH. 1. Invalid event type (every chromosome must be described as a neutral induced, two or four prong event). 20 Event does not have two or four tracks o 3o More than 16 points were recorded on a. single track. 4. Not all tracks were measured. Output from SEARCH is placed on a 9-track magnetic tape at a density of 1600 bytes/in. using odd parityo Each chromosome is described separately on the output device by a sequence of records (see Appendix V). The first record contains the roll, frame and event numbers for the chromosome in question as well as the number of coordinate pairs describing the event, the number of tracks, and the date and time of measurement. Each of the following records contains the track number, number of points on the track and a single X-Y coordinate pair. This program makes reference to four logical input/output device numbers and is designed to be run from a remote terminal. Unit 8 is used to refer to the input magnetic tape (from the DDP system). Writing of the output tape occurs through device number 3 which must be preset to the desired starting point by the user before running the program. Units 5 and 6 are used as input and output devices, respectively, for conversational interaction with the computer. As can be seen in the sample session using SEARCH in Appendix VI, the machine requests information as to which tape file is to be scanned first and then asks how many consecutive files are to be read and how many chromosomes are present (a maximum number)~ By sending a. logical. end of file' the user causes the program to default to scanning one file for a maximum of 100, 000 chromosome s Application of the program is invoked by the appropriate MTS "run" commands to provide for mounting of the necessary magnetic tapes and initiation of the program itself. The use of these commands is shown in the sample session (Appendix VI). A complete listing of the source coding of this program is given in Appendix VIIo

CHROMOSOME Although it makes use of a number of MTS and IBM library subprograms, CHROMOSOME itself was written entirely in our laboratory. This program reads the magnetic tape produced by SEARCH, calculates the length of each chromosome measured, and then produces a variety of information based on the lengths of all the chromosomes in a measured spread. It may be helpful to refer to Figure 7B, which contains a similified flow chart of CHROMOSOME, while reading this description. When the program begins execution in the computer, the subroutine CHECK is called. This program sets the input tape to the desired starting location and, if the program is being run in "batch" mode, sets up the required storage areas for constructing plots for printing on the line-printer. The main program then begins to read input data. As the first record for each chromosome is read, it is checked to see if it represents data from a new roll or frame. Because it is possible for a new frame number to be written on the tape once accidentally, the program checks to see that all chromosomes from the preceding frame have been processed. This is done by spacing the tape ahead to see if the next event contains the same new chromosome number. If it does not, the single chromosome with an incorrect frame number is skipped, but, if it is in reality a new frame, the tape is back spaced to the beginning of the data for the new frame. When a new frame number or end of file mark are read off the tape, all the data for the preceding frame is printed. This includes those items listed in text Table IIIB (also see Appendix IX). In addition, when an end of file mark or new roll number are encountered it is assumed that all data from an entire population sample has been examined and some summary information is printed~ This summary consists of two plots. The first is a plot of arm ratio vs. relative length for each biarm chromosome. In this plot every chromosome is plotted using the same character and there is no way to differentiate data from different cells.8 The second plot is a frequency distribution of relative chromosome and arm lengths found in all frames measured. To construct this plot each length is placed in a category with other chromosomes having similar lengths. Twenty-six different groups are used as outlined in text Table IVB. The final plot shows the distribution of chromosome lengths (plotted with an asterisk) and arm lengths (plotted with an X) as the fraction of the total number of chromosomes or arms plotted against the length (see Appendix IX). Both plots are produced using the line printer plot subroutines which are a part of the MTS library. A program described in Appendix XI plots this same data using a different character for each frame on the CALCOMP plotting system.

TTR, i \ /~C~L~ ~ALL PLOTR /\/- \batch /SET UP PLOT CALL "CHECKS ATCH OR TE1> < MATRIX terminal D I /c ) cE AEA! A CA L REWINA FIRST AND SKIP -J ~~ eof RECOR \ TO POSITION SPACE TAPE BACKSPACE READ yes TAPE TO LA: EW RMEATO NEXT CHRO.AME F MOSOMEAME? TAPE CHROMOSOME MOSOME ~~RECOR #? no no n TAPE RECORD =PTS ON HROM PRINT CALC PTS/ CALL ILAPS #L TDIRACK TRACK,AVE CALCULAT YESRAI 0 - CALCULATE CHROM, & ARM MESRN RAT f~.~. Mo \! ~' ~L s/R T=2? STAM RELATIVE LENGTHS LENGTHS _. W _ } ALL ORDER I=! lyes /TO ARRANGE \ I>N''\CHROMS BY LENGTH PT= no READ LA(PASRT) I PRINT I/LL PLOTR ALL SY c~~j~ —A ~~ ~ n Co CALDIST I ALCULATE DISTANCE BE- IR,CKN TRACK? RACKS ON LAS AND SHORT RS nCHROMOSOME A RACKS CHROM? EU no SHORT AR LOG ARM FgTORE AS LONG A:TOTAL/2 — TOTAL/2 ENTRO- TOTAL/ HERE 2 REVERSE LONG AND SHORT ARMS IF NECESSARY ARM RATIO= LONG ARM/ SHORT ARM Figure 7B. Flow chart for the program "CHROMOSOME."

TABLE IIIB. A list of the information provided for each spread of chromosomes by the program CHROMOSOME. 1. Roll and frame numbers. 2. Time and date at which measuring of the spread was started and completed. 3. The elapsed measuring time. 4. Total number of chromosomes in spread. Number of telocentric chromosomes present. 6. Number of biarm chromosomes present. 7. Number of tracks measured. 8. Average number of points recorded on each track. 9. Total length of all chromosomes in spread. 10. Average length of a chromosome. 11. Average arm length. 12. Separate lists of data on biarm and telocentric chromosomes (in order of decreasing total length). a. Location of the centromere (the first vertex recorded for the chromosome). b. Lengths of the long and short arms in relative length units (absolute length * 1000 / average chromosome length). c. Arm ratio (long arm length/short arm length). d. Total length of each chromosome in relative length units. 13. Considering the biarm chromosomes as one group and the telocentrics as a second, independent group, the following values are calculated for each group. a. Mean relative length. b. Variance and standard deviation of the mean. 48

TABLE IVB. Catagories of ranges of lengths used in the construction of arm and chromosome length frequency distribution. GROUP RANGE OF LENGTH 1 0 - 94 2 95 - 194 3 195 - 294 4 295- 39) 5 395 - 494 6 495 - 594 7 595- 694 8 695 - 794 9 795 - 894 10 895 - 994 11 995 - 1094 12 1095 - 1194 13 1195 - 1294 14 1295 - 1394 15 1395 - 1494 16 1495 - 1594 17 1595 - 1694 18 1695 - 1794 19 1795 - 1894 20 1895 - 1994 21 1995 - 2094 22 2095 - 2194 23 2195 - 2294 24 2295 - 2394 25 2395 - 2494 26 2495 -.,o 49

The number of objects falling into each of the classifications shown in Table IVB is also printed so that information on a number of plots can easily be combined manually in the event that a single plot does not contain the data for a complete population sample. One of the procedures which must be used in preparing data for printing when an entire frame has been processed is the ordering of chromosome data according to length. This is done by the subroutine ORDER using a successor list. A successor list is an array containing the subscripts defining the order in which values should be obtained from other arrays. Once such a list has been constructed the program first looks at the "header," a variable containing the index of the first value to be printed in the final list. This index also locates the location in the successor list array containing the index-value to be printed. One simply repeats this operation until all values have been printed. The advantage of this method is that all the data associated with each chromosome (centromere locations, long-arm lengths, shortarm lengths, etc.) can be ordered at the same time by merely modifying the contents of the successor list. It is unnecessary to rearrange all the data for each chromosome. After all the summary data and lists of data described have been printed, the program procedes to process data for the next spread. All the bookkeeping information required by the machine is found on the first card image for each chromosome (see Appendix V). The event number is stored to be printed in the final print-out for the spread, and the total number of biarm or telocentric chromosomes, whichever is appropriate, is incremented. The number of tracks is used to increment the total number of tracks measured in the frame and to indicate to the machine whether or not an arm ratio is to be calculated. Using the total number of coordinate pairs for the current chromosome, the program initiates a loop to read in the appropriate number of X-Y coordinate pairs and stores this number for use in calculating the average number of points per track, The time and date are only used as they are obtained from the first and last chromosomes in a spread to calculate the elapsed measuring time (via the subroutine ILAPSE). Within the coordinate pair reading sequence the program reads a card image and, if the data is the first record on a track, stores the length of any preceding tracks, zeros the necessary accumulator, and stores the coordinate as the last point measured. Succeeding points on a track (read from the card images which follow) are first stored as the current X-Y pair, used to calculate the distance between the current point and the last point, and then stored as the last point before the next card image is read. Distances between successive points are accumulated. When the total lengths of two tracks have been calculated in this way, they are averaged and the result is stored as the length of the long arm (in absolute units). If the program then finds points for two more tracks, their averaged length is stored as the short arm length and, after reversing these values (if necessary) an arm ratio is calculated. The first coordinate pair read for 5o

a chromosome is stored as the centromere location. Distances between points are calculated using the Pythagorean Theorem. To use this method, one simply assumes the two points whose coordinates are (X1,Y1) and X2,Y2) to be the two vertices of the acute angles of a right triangle. A line drawn between these two points is then the hypotenuse of the triangle. The distance between the points is calculated using Equation (1)o C = (X1-X2)2 + (y1-_y)2 (1) Where C = the length of a line between the two points and the length of the hypotenuse of the triangle formed by the two points. A one-line function in the main program is used to calculate C. As was mentioned above, all chromosome lengths are stored in absolute length units. The calculation of relative lengths does not take place until the program has learned that all data, for an entire spread has been read. Relative length calculation occurs in the subroutine MODIFY which places the resulting values in the storage locations originally containing the absolute lengths. Equation (2) describes the calculation used. R = (L * 1000)/A (2) Where R = the calculated relative length. L = the absolute length of a chromosome arm. A - the average length of a'chromosome in the current spread. Multiplication of the absolute length by 1000 allows the computer to perform most arithemtic operations in integer mode without losing significant figures. Any errors discovered while running the program are handled by the subroutine IFAIL which terminates execution of the program after printing the comment; "ERRORo.. RETUPR CODE = nnn' 51

where "nnn" is a numerical return code describing the error which occurred (see Appendix X). Of course, other errors causing termination of program execution may occur. These are errors detected by the supervisory system or by the program interupt code contained in the program status word which is printed in the event of an abnormal program termination. "CHROMOSOME" makes reference to four logical input/output device numbers. Device 5 is assigned to the magnetic tape containing chromosome data. The location at which the program is to begin reading data off the tape is set by the program after the user is interrogated via unit 6 and replies with the starting file and record number on device 3. In addition to questions to the user, all program output which requires little time to print is assigned to unit 6. Long lists of output information are assigned to logical device 7. This type of device assignment provides for the program to be easily run in either batch or terminal mode. When operating through a terminal, the program omits the printing of plots and, if unit 7 is assigned to a dummy device,9 the printing of long output is also omitted. Terminal operation allows the user to check the operation of the program and to examine the contents of his magnetic tapes. In batch mode it is generally sufficient to allow 3 pages of printed output and 8 sec of central processing unit time for each frame processed under the MTS system. The cost of running the program is about $.62 per frame. USE OF COMPUTER OUTPUT As can be seen in the preceding discussion, the system which has been developed provides a substantial quantity of information on the chromosome complement of each cell examined. However, as was mentioned, our interest is not specifically in the characteristics of single cells but in the inference which might be drawn to the makeup of the population as a whole. In examining the results of a number of population analyses made with this system some experience has been gained in determining which is the most useful data for comparing populations. This is of great significance for the reason that a variety of problems can be reduced to population comparisons. Following a population over a period of time, observing the transformation of diploid cells, comparing two apparently unrelated populations of cells, comparing a cell population to normal cells from the tissue or origin, and searching for abnormal populations of cells are all examples of comparisons between populations. The fact that chromosome measurement provides quantitative data is also helpful because such information lends itself to statistical interpretation. In the MTS system *DUMMY* is used as such a device. Any material written on such a device is simply lost. 52

To date, the test which we have found most helpful for comparing characteristics is the Wilcoxon two-sample test. This is a, ranking test used to determine if the values obtained for some variable in one population are lower than the values from the other population with sufficient consistency for one to say the differences between the values are probably not due to chance alone. An arbitrary value for a, the probability that the results are due only to chance, of 0o01 is currently being used as the maximum acceptable value for saying the two populations are different on the basis of the single characteristic considered in the calculation of a. This in itself is not, however, taken to mean the populations are truly different. The number of variables providing values less than 0.01 is also considered. To apply the Wilcoxon two-sample test to the results obtained for some variable in the computer output, the values from both populations are first ranked in ascending order (Equation 3). Values from population 1 Scale of possible values for X II. (3) Values from population 2 Then a rank is assigned to each value along the linear scale. Population 1 1 2 6 8 9 Value for X.I. (4) Population 2 3 5 7 Next, a test statistic, W1, is calculated as the sum of the ranks obtained from population 1l W1 - 1 + 2 + 6 + 8 + 9 = 27 (5) Under the null hypothesis, HO, all the possible combinations of rank orderings have equal probabilities of occurring, and the test statistic, W1, is distributed according to the normal distribution with expectation E(W1) and a variance, Var(W1), calculated according to Equations (6) and (7). 53

E(W1) = sN + 1) (6) Var(Wl) s(N - s) (N + 1) () 12 where s = the number of values for the variable obtained from population 1. N = the total number of values for the variable obtained from both populations. High or low values for W1 are considered as evidence against the null hypothesis. The value for O is calculated using the values for W1, E(W1), Var(W1), and the normal distribution. First z is calculated. W- - E(W1) (8) \[Var(W1) Then the area outside the region between -z and +z on the normal distribution is obtained using the appropriate tables. In this way the Wilcoxon two-sample (9) is used as a two-tailed test to give the probability that one might, due to only chance, obtain any value for W1 as extreme as or more extreme than the observed value. The variables which have been compared thus far using this test are listed in Table VB. It was hoped that, just as a karyotype provides a good description of a diploid cell, a list of quantitative characteristics could be provided by this analysis system which would be descriptive of the population examined. A first attempt at this consists of simply listing the mean and standard deviation of the values for each variable obtained from the entire population sample examined. Of course, these values are not as easily visualized as a karyotype. Attempts have also been made to visually display the results obtained to aid in the interpretation of results. This is being done through the use of the two plots already mentioned, arm ratio vs. length and chromosome length frequency distributions. Execpt in the study of diploid cells, the plots of arm ratio vs. length are only of questionable value. An example of such a 54

Table VB. Variables being compared using the Wilcoxon two-sample test. 1. Mean length for all biarm chromosomes in each spread examined. 2. Standard deviation of (1). 35 Mean length for all telocentric chromosomes in each spread examined. 40 Standard deviation of (3). 5, (1) - (2). 6. Standard deviation of the mean length of all chromosomes in each spread. 7. Total number of arms in each spread examined. 8. Total number of chromosomes in each spread examined. 9. Values obtained by dividing the average chromosome length by the average arm length in each spread examined. plot produced for a number of frames of data from BHK cells (BHK 21 Clone 4) on the CALCOMP plotter is shown in Figure 8B.10 On the other hand, chromosome length frequency distribution plots appear to be quite useful. These plots are prepared by manually combining all the plots produced by the computer for arm length and chromosome length vs. frequency constructed for a single population sample. To combine them, the numbers of items falling into each of the classifications in Table IVB are added and the totals in each group are divided by the total number of items (chromosomes or arms) being considered. A smooth curve is drawn through the resulting points. Although it is apparent in the examples in Figure 9B that visual comparison of these plots is not difficult, the method is not very quantitative To correct this, plans are being drawn up to write a supporting program which 10 Supporting programs including the one used to construct these CALCOMP plots are listed in Appendix XI. 55

8. O 7.0 6.0 o 5.0 < 4.0 x 3.0 + + x+ 2.0 + 2.0 0, ~+,, $Z,.X / + x x X i. O 0 +A 0x " + 0.20.40. 0.80 1.00 1. 20 1.40 1.60 1.80 2.00 2. 20 2.40 LENGTH Figure 8B. CALCOMP plot of chromosome length vs. arm ratio for four BHK cells.

x BHK susp. HS 49 0.18 -\ o- Chromosomes L) "'E --- x- Arms L/) / o 0. 12 -') -,_ X0.06 r,_) 0.0 0.4 0.8 1.2 1.6 2.0 2.4 RELATIVE LENGTH x 10-3 128 LM 199 P FCS2.5 - Ch romosomes 0. 18 x Arms LU= /) o 0.12 -r O0 z 0.06 0.0 0.4 0.8 1.2 1.6 2.0 2.4 RELATIVE LENGTH x 10-3 Figure 9B. Sample frequency distribution plots for chromosome and arm length in two cell lines. 57

will fit a line to the points using the lowest power polynomial regression resulting in a line varying no more than + 10% from each point. The calculated formula will then be used to draw this line on a CALCOMP plot. RESULTS ACCURACY OF CHROMOSOME MEASUREMENT Previous analysis of data obtained by this system (11) has shown that the results obtained are an improvement over manual measurement techniques with respect to reproducability and the magnitude of sample variance obtained. It is felt that the errors observed are of a small enough magnitude that useful results can be obtained from chromosome measurement studies using this system. To check the accuracy of the manual process of translating photographic images into digitized information, a FORTRAN IV program has been used to draw karyotypes using the X-Y coordinate data on the CALCOMP plotting system. An example is shown in Figure lOB. By comparing this drawing to the original photograph it is possible to determine, visually, the accuracy of the digital data. This has been done only periodically to check the operation of the system. CHARACTERIZATION OF CELL POPULATIONS To date, our attempts at quantitative characterization of the chromosome complements of a population sample consist simply of providing the plots which have been described and listing the means and sample standard deviations of the values being used in population comparison studies. Such a listing is shown in text Table VIB. More experience must be obtained before it will be possible to ascertain the value of having such a description. COMPARISON OF POPULATIONS Use of the Wilcoxon two-sample test as a two-tailed test for similarity in values from the populations used has proved very useful., Of course, more experience will be necessary before it will be possible to determine which of the nine items being compared are the most useful, but the advantages of making comparisons in multiple categories are already apparent in that they allow the introduction of a third dimension into the process. That is, the difference between populations can be described on the basis of how many categories show differences of significance as well. When more studies have been carried out it will be possible to dispose of categories which are not useful and to attach weights to the remaining ones so that a final single value can be provided on which the acceptance or rejection of the null hypothesis can be based.

TABLE VIB. Means and standard deviations for the values obtained for each characteristic examined in the cell lines listed. CHARAC- CELL LINE TERISTIC BHK HS-49 128 L-M* 90 L-Ma* 246 L-M* HUMAN MALE # Chromo- 4350+1~0 63.4+1.0 60.4+1.9 63.5+2.1 46.0+0.0 somes # Arms 76.4~1.9 76.0+1.6 74.2~1.7 76.6~2.5 87.1+3~ 8 Xbiarms 1117~34 1884+166 1654+65 1728+80 1056+44 Cbiarms o09+66 504+94 250+42 252~45 412~23 Xtelos 604+51 779~44 806+27 811+24 433+199 CY telos 112+49 404+593 144+12 157+20 106+82 telos X - Xt 51381 1115+215 851+98 927~98 619+166 -whole cell 5369+39 526+78 381~+59 429~52 whole cell Xchrom 1078+04 1.20+0.1 1.21+.06 1.21+.06 1.91+.08 X arm All L-M cells are grown in medium 199 supplemented with peptone. The medium of the 128 L-M cells also contained 2.5% fetal calf serum. 59

246 LM. 12/31/68. FRRME 12 (Z-10) V V 11 v Y Figure 10B. Karyotype drawn by the CALCOMP plotter to check the accuracy of the measurement system (from Baumann and Merchant) (11).

Sample comparisons between a. variety of populations are shown in Table VIIB. MEASUREMENT OF HUMAN CELLS Initial studies on the measurement of human chromosomes have shown that within a single spread all chromosomes can be paired and grouped according to the Denver classification on the basis of measurements and within a population sample of normal cells criteria can be set up for placing each chromosome in its correct group. These grouping decisions are made on the basis of relative length- and arm-ratio calculations~ It is also believed that the numerical characteristics currently being examined for cell populations may be advantageous in studies of human cells for the purpose of locating abnormalities without detailed karyotypic studies. Of course, to develop such a technique it would be necessary to use some completely automated system for collecting chromosome data. The difference between this technique and others developed recently is that it depends on numerical features of the chromosome complement other than those normally observed in systems dependent on karyotype construction. An arm ratio-length plot showing how human chromosomes from a single spread fall into groups and pairs is shown in Figure llB. CONCLUS IONS.. A system which has been developed for the description and comparison of the chromosome complements of cell populations has been described in detail A semiautomatic, X-Y coordinate digitizer system is used to provide taped descriptions of chromosome spreads for analysis on the IBM 360/67 computer system at The University of Michigan. This system provides a number of numerical values which may be used as a mechanism for comparing or simply describing populations on the basis of results from a small sample of cells. Results obtained to date have demonstrated the potential of the system, but some additional experience will be necessary to completely define the procedure for analyzing results. 61

TABLE VIIB. Values obtained for a in the comparison of a number of populations using the Wilcoxon two-sample test. Basis for. Comparison Cells Cel Chroms arm s X bi s b avg. chr om Compared Cbiarms biarms telos telos bt Gwhole cell v avg. arm BHK HS-49 VS. 0.00014 0.52709* 0.00014 0.00147 0.00014 0.00014 0.00032 0.00245 0.07186 128 LM 128 LM VS. 0.00022 0.04338* 0.00067 0.0oolo 0.04338 0.00010 0.00067 0.00047 o.log60 90 LMa 90 LMa VS. o.moo8i 0.01928* 0.00905 0.00511 0.09894 0.43539 0.14429 0.21498 0.00010 246 LM 128 LM** vs. 0.32709* 0.70394* 0.0511 0.03573 0.02321 0.00010 0.00633 0.00052 0.22628 246 LM 128 LM** 0.52709* 0.70394* 0.00511 0.03557 0.02521 0.00010 0.00633 0.00052 0.22628 Human VS. 0.00047 0.00047 0.00047 0.00047 0.00614 0.55526 0.25014 0.00147 BHK Human vs. 0.0047 0.00047 0.00047 o.ol685 0.00047 0.00047 0.01314 0.00047 128 LM Complete culture designations: 128 L-M 199 P fetal calf serum 2.5% 246 L-M 199 P 90 L-Ma T1C3 Clone 2 199 P BHK 21 Clone 4 suspension HS 49 *These values support other results because they ipdicate samples taken were not different with respect to chromosome and arm number to start with. **These populations cannot be distinguished on the basis of enumeration alone.

8. 0 7.O0 6. 0 5. 0 <C D C3 1 4.0 15 3. 0 IIXII 22 12 11 + 5 B 2.0 Y' E C~~~~~1~~6 6~F 3A 20 19 Al 1. 2 0 0.4 0.8 1. 2 1. 4 1.8 2.2 RELATIVE LENGTH Figure 1lB. Plot of arm ratio for a single human cell (male) showing how the chromosomes fall into pairs and groups.

REFERENCES 1. Gofman, J. W., J. L. Minkler, and R. K. Tandy, "A specific common chromosomal pathway for the origin of human malignancy. Electronic scanning and computer chromosome measurement," UCRL-50356, November 20, 1967. 2. Stroud, A., J. Butler, and M. Butler, "The combination of microscopes and computers for the analysis of chromosomes," Microscope, Vol. 15, July, 1967, p. 68. 3. Neurath, P. W. "Human chromosome analysis by semiautomatic method," TID21641. 4. Ruddle, F. H. "Quantitation and automation of chromosomal data with special reference to the Hampshire pig," Symposium of the International Society for Cell Biology, Vol. 3, 1964, p. 273. 5. Rutovitz, D. "Automatic chromosome analysis," British Medical Bulletin, Vol. 24, 1968, p. 260. 6. Giles, R. E., D. J. Merchant, and E. Masselink, "Chromosomes of L-M mouse cells and variants," Journal of the National Cancer Institute, Vol. 37, No. 5, Nov. 1966, p. 663. 7. Siegel, L. "Digitizing graphic records for computer analysis," IEEE Transactions on Bio-Medical Engineering. BME-14, Jan. 1967, p. 7. 8. Chapman, J. Physics Department, The University of Michigan, Personal communication. 9. Diem, K., ed. Documenta Geigy, Scientific Tables, Geigy Pharmaceuticals: Ardsley (New York), sixth edition, 1962, p. 191. 10. Eastman Kodak, Processing Chemicals and Formulas, sixth edition, Eastman Kodak Co.: Rochester, 1965. 11. Baumann, G. W. and D. J. Merchant "A computer system for the quantification of chromosome analysis data from mixoploid cell lines," Proceedings of the Seventh Annual Bio-Medical Sciences Instrumentation Symposium, Imagery in Medicine, 1969 (in press). 64

RELATED RESOURCES G. Gallus, N. Montanaro, and G. A. Maccacro. A Problem of Pattern Recognition in the Automatic Analysis of Chromosomes: Locating the Centromere. Computers and Biomed. Res, 2:187-197 (1968). H. S. Frey. An Interactive Program for Chromosome Analysis. Computers and Biomedo Reso 2:274-290 (1969). Human Intervention for Error Correction and Overlaps, Probablistics Approach to Group Assignment. J. L. J. Gaillard, Van Duign and A. Shaberg. Photometric Determination of DNA in Chromosomes, Exp. Cell Res. 53:417-431 (1968).. Fuelgen Staining, Photograph, Elute Dye from Color Photo: Conclude DNA Better to Place in Karyotype. A. Robinson. A Proposed Standard System of Nomenclature of Human Mitotic Chromosomes. JAMAo 174:159 (1960). I. H. Porter. Heredity and Disease. McGraw Hill Book Co. Inc. New York: 3 (1968) Denver Classification. A'Proposed Standard System of Nomenclature of Human Mitotic Chromosomes. Lancet. 1:1063 (1960). Mo Crippao The Mouse Karyotype in Somatic Cells Cultured in Vitro. Chromosoma. 15:301-311 (1964). Js W. Gofman, J. L. Minkler, R. K. Tancy. A Specific Common Chromosomal Pathway for the Origin of Human Malignancy - Electronic Scanning and Computer Chromosome Measurement. UCRL-50356: (1967). A. Stroud, Jo Butler, M. Butler. The Combination of Microscopes and Computers for the Analysis of Chromosomes. The Microscope 15:68 (1967). M, W. Shaw and R. S. Krooth. Chromosomes of the Tasmanian Rat-Kangaroo (Potorous Tridactylis Acicalis).~ Cytogenetics. 3:19-33 (1964).* Wakonig-Vaarataja and J. Read. Measurements of Allium Cepa Chromosomes. Exp. Cell Res. 38:264-271 (1965). B. Chierellio Statistical Analysis of Relationship Between Genome Length and Chromosome Number. Experentia. 23:672 (1967). P. W. Neurath. Human Chromosome Analysis by Semiautomatic Method. TID-21641:().

K. H. Rothfels and L. Siminovitch. The Chromosome Complement of the Rhesus Monkey (Madaca Mulatta) Cetermined in Kidney Cells Cultured in Vitro. Chromosoma. 9:163-175 (1958). A. Levan and T. C. Hsu. The Human Idiogram. Fereditas. 45:665-674 (1959). T. C. Hsu and M. T. Zenzes. Mammalian Chromosomes in Vitro XVII - Idiogram of the Chinese Hamster. JNCI. 32:857 (1964). A. Levan, T. C. Hsu, and H. F. Stich. The Idiogram of the Mouse. Hereditas (Lund). 48:677-687 (1962). A Carrel and A. H. Eberling. Pure Culture of Large Mononuclear Leukocytes. J. Exp. Med. 36:365 (1922). O. Von H. Owens, M. K. Gey, and G. 0 Gey. A New Method for the Cultivation of Mammalian Cells Suspended in Agitated Fluid Medium (Abstract). Proc. Am. Assoc. Cancer Res. 1:41 (1953). J. C. Bryant, E. L. Schilling and W. R. Earle. Massive Fluid-Suspension Culture of Certain Mammalian Tissue Cells I - General Characteristics of Growth and Trends of Population. JNCI. 21:331-348 (1958). A. Hellman, J. D. Regan and D. H. Martin. Large Scale Cultivation of Mammalian Cells in Vitro. Applied Microbiol. 15:201-202 (1967). R. J. Kulcher and D. J. Merchant. Growth of Tissue Cells in Suspension. The University of Michigan Med. Bul. 24:200-212 (1958). W. F. Mclimans, F. E. Giarcinello, E. V. Davis, C. J. Kucera and G. W. Rake. Submerged Culture of Mammalian Cells: The Five Liter Fermentor. J. of Bact. 74:768-777 (1957). W. A. Rightsel, H. Mccalpin and I. W. Mclean. Studies on Large-Scale Methods for Propagation of Animal Cells. J. Biochem and Microbiol Tech and Engin. 2:313-325 (1960). D. W. Ziegler, E. V. Davis, W. J. Thomas and W. F. Mclimans. The Propagation of Mammalian Cells in a 20 Liter Stainless Steel Fermentor. Applied Mibiol. 6:305-310 (1958). M. Harris. Cell Culture and Somatic Variation. Holt Rinehart and Winston. Chicago (1964). T. C. Hsu and C. J. Merchant. Mammalian Chromosomes in Vitro, XIV - Genotype Replacement in Cell Populations. JNCI 26:1075-1083 (1961). 66

So A. Keefe, D. J. Merchant and W. H. Kelsey. Alkaline Phosphatase Activity of L-M Mouse Cells and Variants. Proc. Soc. Exp. Biol. and Med. 118: 1031-1037 (1965). C. R. Eidam and D. J. Merchant. The Plateau Phase of Growth of the L-M Strain Mouse Cell in Protein-Free Medium, I - Patterns of Protein and Nucleic Acid Synthesis and Turnover. Exp. Cell Res. 37:132-139 (1965). J. C. Bryant, V. J. Evans, E. L. Schilling and W. R. Earle. Effect of Chemically Defined Medium NCTC 109 Supplemented with Method and of Silicone Coating the Flasks on Strain 2071 Cells in Suspension Cultures. JNCI 26:239-252 (1961). R. E. Giles, D. J. Merchant and E. Masselink. Chromosomes of L-M Mouse Cells and Wariants. JNCI 37:663-676 (1966). D. J. Merchant, R. H. Kahn, and W. H. Murphy Jr. Handbook of Cell and Organ Culture. 2nd Ed. Burgess Publishing Co. Minneapolis (1964). M. A. Bender, Mo A. Kastenbaum and Jo B. Davidson. Metric Analysis of the Human Karyotype. CRNL-3999. 97-98 (). A. Do Glinos and D. D. Hargrove. Interrelations Among Chromosome Number, Type and Size in L-Strain Cells. Exp. Cell Reso 39:249-258 (1965). P. M. Labauveo A Digitized Comparator for Karyotype Analysis. LADO-6449: (1964). E. H. Kleisner De Galano Numerical and Structural Interpretation of Normal Human Chromosomes. Mammalian Chromosomes Newsletter. No. 18:146 (1965). F. H. Ruddle and R. S. Ledley. Consideration of Metaphase Chromosome Parameters Amendable to Digital Computer Analysis. In Vitro. 1:21 (1965). T. C. Hsu and M. T. Zenzes. Mammalian Chromosomes in Vitro XVII, Idiogram of the Chinese Hamster. JNCI. 32:857 (1964). J. J. Clausen and J. T. Syverton. Comparative Chromosomal Study of 31 Cultured Mammalian Cell Lines. JNCI. 28:117-145 (1962). M. Crippa. The Mouse Karyotype in Somatic Cells Cultured in Vitro. Chromosoma. 15:301-311 (1964). A. Fjelde, A Levan and Ro Rask-Nelson. The Chromosomes of Four Transplantable Murine Plasma Cell Leukemias Characterized by Varying Pathological SerumProtein Changes and/or Amyloic Formation. Fereditas. 48:630-644 (1962). L. S. Penroseo A Note on Mean Measurements of Humn Measurements of Human Chromosomes. Ann of Human Genetics (London). 28:195 (1964). 67

F. H. Ruddle. Quantitation and Automation of Chromosomal Data with Special Reference to Hampshire Pig. Symp. Int. Soc. Cell Biol. 3:273 (1964). A. Levan, W. W. Nichols, M. Peluse and L. L. Coriell. The Stemline Chromosomes of Three Cell Lines Representing Different Vertebrate Classes. Chromosoma (Berl.). 18:343-358 (1966). A. N. Stroud, J. Butler, and M. Butler. The Application of Computers for the Analysis and Pairing of Chromosomes. Argonne Natl. Lab. Report, Biol and Med Research Div Annual Report. 7316:136 (1965). R. S. Ledley. High-Speed Automatic Analysis of Biomedical Pictures. Science 146:216-223 (1964). R. H. Kahn, B. L. Baker and D. B. Zanotti. Factors Modifying the Stimulatory Action of Norethynodrel on the Mammary Glanid. Endocrinology 77(1):162-168 (1965).* R. H. Kahn and B. L. Baker. Prolactin Content of the Rat Hypophysis Following Treatment with Norethynodrel. Acta Endocrinologica. 51:411-414 (1966). SC Injection to Cvariectomized Rats at Certain Dose Levels Caused Increase in Prolactin Content.* B. L. Baker, R. H. Kahn and D. B. Zanotti. Influence of Norethynodrel on the Adrenal Cortex of Rats. Endocrinology. 77(1):155-161 (1965).* R. H. Kahn and B. L. Baker. Effect of Norethynodrel Alone or Combined with Mestranol on the Mammary Glands of the Adult Female Rat. Endocrinology 75(5):818-821 (1964). B. L. Baker, R. H. Kahn-and D. Besemer. Ovarian Histology After Treatment of Rats with Norethynodrel. Proc Soc for Exp Biol and Med. 119:527-531 (1965).* H. Stockdale, C. W. Ribbons and E. A. Dawes. Occurence of Poly-Beta-Hydroxybutyrate in the Azotobacteriaceae. J. of Bacti. 95(5):1798-1803 (1968).* Nichols and Levan. Chromosome Preparations by Blood Culture Technique in Various Lab Animals. Blood. 20:166 (1962). Sherz. Simple Method for Making Chromosome Slides. Am J. of Clin Pathology 40:222-225().* J. W. Butler, Et al. Automatic Classification of Chromosomes, I, II, III. Data Acquisition and Processing in Biology and Medicine. Vols 3, 4, 5. Pergamon Press, N. Y.: (). Wright. Physiological Centics - Ecology of Population and Natural Selection. Perspect. In Biol and Med. 3:107 (1959). 68

K. W. Petersen, et al. Colcemic Effect on Meiotic Analysis. Mammalian Chromo. Newsletter. 9(1):42 (1968). In Vivo Colcemic Exposure for 1-12C Hrs at 1-5 MG/KG, Minced Tests Used for Preps. L. E. Schneider. The Microscopic, Three Dimensional Viewing of Mammalian Chromosomes. Exp. Cell Res. 47:658 (1967). 3-D Morphology Shown Using Oblique Incident Light Microscopic Technique. R. Hancock and H. Amos. Nuclear Binding of Exogenous Histones by L Cells at Low Ph. J. Cell Biol. 36(l):01(1968). Exogenous Histones Bind at Ph < 4 to Nuclei. F. Gianelli and R. M. Howlett. The Identification of the Chromosomes of the E-Group (16-18 Denver): An Autoradiographic and Measurement Study. Cytogenetics. 6(6):420 (1967)o B. A. Kihlman and B. Hartley. Sub-Chromatic and Folded Fiber Model of Chromosomes Structure. Herecitaso 57(1-2):289 (1967). M. S. Ramanna and R. Prakken. Structure and Homology Between Pachytene and Somatic Metaphase Chromosomes of the Tomato. Genetica. 38(2):115 (1967). L. Wilson and M. Friedkin. The Biochemical Events of Mitosis II. The In Vivo and I Vitro Binding of Colohicine in Grasshopper Embryos and its Possible to the Inhibition of Mitosis. Biochemistry. 6:3126 (1967). Ko E. Hampel and A. Levan. Breakage of Human Chromosomes Induced by Low Temperatures. Hereditas. 51:315-343 (1964). J. J. Maio and L. Decarli. Distribution of Alkaline Phosphatase Variants in a Heteroploid Strain of Human Cells in Tissue Culture. Nature (Lond). 196:600-601 (1962). M. A. Ferguson-Smith and S. Do Handmaker. Observations on the Satellited Human Chromosomes. Lancet, (1):638-640 (1961). M. A. Vogt. A Study of the Relationship Between Karyotype and Phenotype in Clonal Lines of Strain Hela. Genetics. 44:1257-1270 (1959). K. H. Rothfels and R. C. Parker. The Karyotypes of Cell Lines Recently Established from Normal Mouse Cells. J Exp. Zool. 142:507-520 (1959). W. 0. Rieke, et al. The Culture and Karyotype of Rat Lymphocytes Stimulated with PHA. Anat Rec. 150: 383-390 (1964). M. Crippa. The Mouse Karyotype in Somatic Cells Cultured in Vitro. Chromosoma 15:301-311 (1964). 69

H. Willard, et al. Mouse Leukocytes in Culture. Proc Soc Exp Biol and Med. 118: 993-996 (1965). A. A. Hadjiolov. An Endomessenger Hyposthesis on the Mechanism of Information Transfer in Animal Cells. J Theoretical Biol. 16(2):229 (1967). C. E. Ford and P. A. Jacobs. Human Somatic Chromosomes. Nature. 181:1565 (1958). Make Preparations from Sternal Puncture Speciman of Bone Marrow. T. C. Hsu, et al. Mammalian Chromosomes in Vitro, XV, Patterns of Transformation. JNCI. 27:515-541 (1961). Two Patterns Seen; Dipli. J. A. Huberman and G. Attaidi. Isolation of Metaphase Chromosomes from Hela Cells. J. Cell Biol. 31(1):95 (1965). PH 3 Stabilizes Against Breaks. L. Carlson, et al. Application of Quantitative Cytochemical Techniques to the Study of Individual Human Chromosomes. Exp Cell Res. 31:589 (1963). Used Scanning Interferometer to Determine Relative Areas. M. R. Zelle. Biological Effects of Ultraviolet Radiation. IRE Trans on Med Electronics.:130 (1960).* Fry and Fry. Fundamental Neurological Research and Human Neurosurgery Using Intensive Ultrasound. IRE Trans on Med Elec.:173(1960).* D. E. Hughes and W. L. Nyborg. Cell Disruption by Ultrasound. Science. 138: 108-114 (1962).* W. T. Szymancwski and R. A. Hicks. Biologic Action of Ultrahigh Frequency Currents. J. of Infect Dis. 50(1):().* Soviet Microbiological Research Utilizing Radiation and Ultrasonic Techniques OTS #60-21924.: (1959). Some Organisms Lose Virulence But Retain Ability to Produce Immunity. A. Levan. Chromosomes in Cancerous Tissue. Ann N Y Acad Sci. 63:774-792 (1956). T. C. Hsu. Variability Among Progenies of a Single Cell. Univ Tex Publication. 5914:129-134 (1959). Isolation and Study of Mutants From Mammalian Cells in Vitro. Proc Nat Acad Sci. 47:867-872 (1959). Yerganian and Leonard. Maintenance of Chromosome Features in Long Term Tissue Cultures. Science 133:1600-1601 (1961). Hampar. Chromosome Abberations Induced by Virus. Nature. 192:145-147 (1961). 70

J. A. Book. Chromosome Nomenclature. Ann of Human Genetics (Lond). 24: 319-325 (1960). A. B. Griffin. Chromosome Identification. J. Cell Comp Physiol. 56 Suppl 1:113-121 (1960). T. CO Hsuo Chromosome Breakage in Cell Populations of Chinese Hamster. Canad Cancer Conf. 5:117 (1963). E. Chu. Mammalian Chromosome Cytology. American Zoologist. 3:3-14 (1963). Gibson and Scandlyr. Slide Processing for Mammalian Meiotic Chromosomes. Stain Technol. 37:1-5 (1962). Ko S. Lion and B. S. Gould. Effect of High Frequency Electric Fields on Some Microorganisms. 221 (1949).* Hu Ming-Kuei. Visual Pattern Recognition by Moment Invariants. IRE Trans It.:179-187 (1962).* R. S. Ledley, et alo Pattern Recognition Studies in the Biomedical Sciences. AFIPS Conf Proc. 28:411-430 (). F. L. Alto Digital Pattern Recognition by Moments. J. Assoc. Comp. Mach. 9:240 (1962). B. Perry and M. L. Mendelsohn. Picture Generation with a Standard Line Printer. Com. of the Acm. 7:311-313 (1964). J. R. Clark. Information Capacity of Photographic Films. J. Opt Soc of Amer. 51:1159 (1961). Lo L. Hundleyo A Flying Spot Interference Microscope. Ann N Y Acad Sci. 97(2): 514-515 ()0 H. Gay. Chromosome Structure and Function. Carnegie Inst Yrbk. 63:608-609 (1964)o A. Falek, P. W. Neurath and T. Warma. Analysis of Chromosomes by Computer. Mammalian Chromosomes. 1.5:101 (1965). J. G. Hoffman. et al. Digital. Computer Studies of Cell Multiplication by Monte Carlo Methods. J N C I. 17:175-188 (). G. S. Sebestyen. Recognition of Membership in Classes. IRE Trans on Information Theory II-7. (1):44-50 (1961). 71

P. Montgomery (Ed). Scanning Techniques in Biology and Medicine. Ann NY Acad Science 97:329 (1962). R. S. Ledley. Use of Computers in Biology and Medicine. McGraw-Hill, NY.: (1965). U. R. Konigsberg and H. M. Nitowsty. Studies of the Karyotype of Clonal Strains of Chang Liver Differing in Alkaline Phosphatase Activity. JNCI. 29:699-709 (1962). L. Decarli, et al. Alkaline Phosphatase Activity and Chromosome Variation in Human Cells in Culture. JNCI. 31(6):1501 (1963). Loss of 7-8 Chroms With Loss in Enzyme Activity. M. Wand, et al. Tritiated Thymidine:Effect of Decomposition by Self Radiolysis on Specificity as a Tracer in DNA Synthesis. Science. 157:436 (1967). J. H. Taylor. Autoradiography with Tritium Labled Substances. Advan in Biol and Med Physics. 7:107 (1960). C. P. Stanners and J. E. Till. DNA Synthesis in Individual L-Strain Mouse Cells. Biochem and Biophys Acta. 37:406-419 (1960). N. Branchi, et al. A Technique for Removing Silver Grains From Tritium Autoradiographs of Human Chromosomes. Hereditas 51:207-211 (1964). 72

APPENDIX I CHROMOSOME PROC EDURE 1. Add 0.1 ml of Colcemid solution (Grand Island Biological Co., Catalog number 521) (25 ptg/ml) for every 5 ml of growth medium to a culture in middle log phase and incubate for 3-13 hours (depending on the doubling time of the cell being used). 2. Harvest cells, centrifuge at 500-800 RPM for 5 min. and discard medium. Resuspend in 3 mls of 1% sodium citrate in distilled water warmed to 370C (prepare fresh each time). Let stand 6-8 min. at 370C, then add one drop of fixative (1:8, acetic acid:methanol, prepared fresh each time) and centrifuge for 5 min. (at 500-800 RPM). Discard supernatent. 3. Resuspend cells in 0.5 ml fixative (for every 5 mls of original culture fluid) by adding fixative very sloply (a drop at a time). Allow to stand at room temperature for 10 min. Centrifuge, change fixative two more times allowing the suspension to stand 10 min after each change. 4. Dip a clean slide in 70% methanol, immediately drop one or two drops of cell suspension on the slide. Ignite slide by passing it through a Bunsen burner flame. Let stand 15 min. to complete drying. 5. Permanent slides are prepared as follows: a. Stain 40 min in dilute Giemsa solution (diluted 1:15 in distilled water before use). b. Rinse rapidly in distilled water (10 sec). c. Rinse in two changes of acetone. d. Rinse 3 times in acetone:xylol (2:1), 3 times in acetone: xylol (1:2), and 10 minutes in fresh xylol. e. Mount using a xylene soluble mounting medium. PROBLEMS: 1. Clumping may be reduced by adding fixative more slowly or increasing the concentration of acetic acid in the fixative. 2. Cell breakage may be reduced by decreasing the concentration of acetic acid in the fixative. 3. The number of mitotic figures present may be increased by increasing the incubation time with Colcemid, but the problem may be that cells are not mitotically active. 73

APPENDIX I I FILM PROCESSING TECHNIQUE All photographs are taken on Eastman Kodak high contrast copy film (HC-135) and developed as follows: 1. Develop film for the time specified by the manufacturer for the room temperature where the prepared developer has been stored. Agitate film for 10 seconds out of each minute. Eastman Kodak developer, D-19, is used. 2. Replace the developer with Eastman Kodak stop bath, SB-5a, for 30 seconds. The stop bath solution is made 10 up in our laboratory according to the following formula Water 500 ml 28 % Acetic Acid 640 ml Sodium sulfate (desicated) 45.0 gr Water to make 1000 ml 3. Pour off stop bath and fix the film for 5 min in Kodak fixer (general purpose, hardening fixer). 4. Wash films for 30 min. in running water. 5. Rinse films in a solution of Kodak photo-flo. 6. Hang films to dry.

APPENDIX III SAMPLE SESSION ON THE DDP-24 Each line in the session which follows describes the users response to the display which appears on the communications console. All of the responses described are carried out on that console. KEYBOARD STEP MESSAGE DISPLAYED KEYBOARD TERMINATOR EXPLANATION ENTRY 1 ENTER EXP 1 C Returns machine to signon sequence. 2 ENTER ID 66 P Enter user ID no. 3 ENTER EXP A Enter alter sequence to change program parameters so film 4 any message 9 A v b view being used doesn't change 5 any message P during measurement. 6 ENTER EXP 8 P Enter experiment no. assigned for chromosome measurement. 7 ENTER ROLL 40 P Enter 4 digit (max) roll number. 8 ENTER FRAME 2 P Enter 4 digit (max) frame number. 9 ENTER EVENT 1 P Enter event number. 10 ENTER TYPE 1 2 or 4 N Enter event type (2N or 4N only). 11 ENTER TYPE 2 P Leave event type loop. 12 MEASURE VERTEX 1 none,just foot Record centromere for locate cros peddle first track. hairs 13 MEASURE TRACK 1 1 " Locate points along track. 14 MEASURE TRACK 1 2 15 MEASURE TRACK 1 n P Signal that end of track has been reached. 16 Program automatically returns to (12) until all vertices and tracks have been measured. 17 Program automatically returns to (8) 75

PROBLEMS 1. Frame number entered in step 8 too low causing the red error indicator to light. a. If the entered frame number was correct (this is checked by entering the number again and pressing the "P" terminator - red light should remain on), the number as frame number for the preceding chromosome was incorrect. Enter 1-C and start over at step 1. Only the preceding chromosome must be remeasured (the one having an incorrect frame number associated with it). b. If the entered frame number was incorrect, enter the correct frame number followed by the "P" terminator. 2. Red light goes on after event number is entered in step 9. The event number entered was not greater than the one entered for the preceding event in the current frame. Enter a higher number followed by the "P" terminator. 3. Incorrect event type entered. Continue through the sequence until step 12 (or any other step in which "MEASURE" is displayed) is reached. Enter 6-C. This will return the program to step 10. 4. Error made in locating cross hairs, ie. a point was recorded incorrectly. Enter 1-C to return to the first point of the current track or enter 0-C to step the measurement sequence back one point at a time or enter 6-C to begin the event over again (this returns the program to step 10). 5. Measurement of a film completed. Enter 3-C to return the program to step 7 ("ENTER ROLL"). 6. User desires to check previously entered information. a. Depress the "A" terminator. b. Depress the "C" terminator repeatedly to display all the values entered for the current event. c. Depress the "P" terminator to return to the normal program sequence 76

7. Measurements completed for the current session. a. Enter 1-C to return the program to step 1. b. Enter 1-P. This has the effect of signing off the current user. 77

APPENDIX IV DDP MAGNETIC TAPE TRACK RECORD ORGANIZATION The contents of each word produced by the DDP computer in writing track records are described by the following list (8). WORD NO. CONTENTS 1 Number of words in record 2 Experiment number 3 Roll number. 4 Frame number 5 Event number 6 Track number * 1000 + maximum track number 7 Not important 8 Scanner (user) ID * 1000 + measuring machine number 9 Date (MMDDYY) 10 Time (HHMMSS) 11 1 12 Event type word 13 Not important 14 Not important 15 Number of coordinate pairs in view 16 X coordinate of point 1 17 Y coordinate of point 1 18 X coordinate of point 2 78

APPENDIX V CHROMOSOME DATA TAPE RECORDS Each chromosome is described independently on magnetic tape by a series of tape records. Record 1 (in format 7 I 10) FIELD CONTENTS 1 Roll number 2 Frame number 3 Event number 4 Total number of coordinate pairs in event (N = number of coordinate type records to follow, always greater than 2) 5 Number of tracks (2 or 4) 6 Date (MMDDYY) 7 Time (HHMMSS) Records 2 - N+l (in format 3 I 10) FIELD CONTENTS 1 Track number * 1000 + number of points on track 2 X coordinate 3 Y coordinate While being simple to use, this type of organization is highly wasteful of magnetic tape storage space and reading time. A new, more flexible system is being worked out. 79

APPENDIX VI SAMPLE SESSION USING "SEARCH" The following is a complete teletype session showing how the program "SEARCH" is run. All lines entered by the user on the teletype keyboard are underlined, and control characters are indicated as small letters. MTS (LA47-0250) WHO ARE YOU? UM MICBIOAAAs #NEXT EXPECTED SYSTEM SHUT' DOWN IS 5 30 PM #$SIGNON SF40 PW= s #**LAST SIGNON WAS: 16:09.54 06-18-69 # USER "SF40" SIGNED ON AT 13:26.08 ON 06-18-69 #$RUN *MOUNTs #EXECUTION BEGINS *MOUNT is used to mount ENTER MOUNT PARAMETERS magnetic tapes on the tape drives at the Computing Center. P026 7TP,*TAPE*,MODE=50F,RING=OUT, SIZE=8000,'395L's DONE ENTER MOUNT PARAMETERS G113 9TP,*TAPE2*,MODE=1600,RING=IN,SIZE=80,' SF40-1's DONE ENTER MOUNT PARAMETERS Both the input and output tapes have now been mounted. cs END OF FILE #EXECUT ION TERMINATED #$COPY FROM *SOURCE* TO *TAPE2*4CCs >REWs This sequence of steps >FSFs is used to space the output >FSFs tape to the desired >$ENDFILEs starting point. #$RUN ]SEARCH 3=*TAPE2* 8=*TAPE*s #EXECUTION BEGINS UNITS 3 AND 8 MUST BE MAGNETIC TAPE ENTER NUMBERS IN FLOATING POINT ONLY INPUT STARTS IN WHAT FILE NO. (UNIT 8)?? 2.Os 80

OUTPUT TAPE (UNIT 3) READY? IF NOT, TERMINATE! HOW MANY FILES SHOULD BE READ?? FOR DEFAULT SEND EOF 1.0s HOW MANY EVENTS?? 415.s DO 1 FILES. STOP AT EVENT 415 BEGINNING FILE 1 ROLL 31,FRAME 8,EVENT 5 ** TRACKS,FAIL CODE = 60 ROLL 31,FRAME 8,EVENT 9 6 TRACKS,FAIL CODE = 60 ROLL 31,FRAME 9,EVENT 14 6 TRACKS,FAIL CODE = 60 ROLL 31,FRAME 12,EVENT35 5 TRACKS,FAIL CODE = 31 ROLL 31,FRAME 13,EVENT 6 6 TRACKS,FAIL CODE = 60 ROLL 31,FRAME 13,EVENT10 6 TRACKS,FAIL CODE = 60 ROLL 31,FRAME 13,EVENT19 2 TRACKS,FAIL CODE = 10 ROLL 31,FRAME 13,EVENT57 6 TRACKS,FAIL CODE = 60 ROLL 31,FRAME 14,EVENT 6 6 TRACKS,FAIL CODE = 60 ROLL 31,FRAME 14,EVENT12 3 TRACKS,FAIL CODE = 20 THERE ARE 271 GOOD N2 EVENTS AND 134 GOOD N4 EVENTS. THERE ARE 10 BAD EVENTS. #EXECUTION TERMINATED #$SIGNOFFs #OFF AT 13:38.42 #ELAPSED TIME 753.676 SEC. #CPU TIME USED 85.72 SEC. #STORAGE USED 7266.43 PAGE-SEC. #DRUM READS 22 #APPROX. COST OF THIS RUN $7.26 #FILE STORAGE 589 PG-HR. $.15 81

APPENDIX VII SOURCE PROGIAM LISTING FOR "SEARCH" CCNNCN/CCLNT/I EVCT,N,lN EBE, N.CCC ( 2 ), NTAPE ( 2 ) CIENS ICi. I (25C0), IA(2C0),MSW(7 C UNIT 5'IS FCR It\PLT CF EAT FPARPIETERS CC 7 k=1,7 7 PSIo1g)=1 f.NN= 1 NTAPE (1)=3 NTAPE (2)=3 90C FCR'VT ('lOUNIT7S 3 pAC E'LST BE PACNETIC TAPE'/' ENTER NUMBERS IN F ~2LCATING POINT CNLY'"/'eCINPLT STARTS I.N WHAT FILE NO. (UNIT 8)??') 5 I'ITE (6,gO) 9C1 FCRMAT (FlC.l) REtC ( 5,901, ENC='Sc,ERR=S9 )G NC 1=G 1-1 REV'INC 8 IF (NG1) lC,2C,3C SC2 FCRMAT ('OC...NC NEhATIVE NUMPERS CR ZERCS... ) 10 A PITE(6, C02) CC TC 5 3C CALL SKIP(NGlIC,8) SC3 FCPM7AT(' OUTPLT TAPE ('NIT 3) PEACY? IF NOT, TERMINATE') 2C WRITE (6,903),C4 FCRMAT(' HCW MANY FILES SI-CULE EE REAC?? FCR CEFAULT SEND EOF') 80 WR ITE (6,904) REAC(ES,SC1 ENC=SC,ERR=S9) GR1 IFCT=CR1 IF ( IFCT ) 7C,7C, 1 70 WiRI'TE(6,9C2) CC TC 80 905 FCRMAPT(' HCW t MANY EVENTS?? ) 81 WRITE (6,9C5) REtE (5,901, EN0=5C, ERR=-9)GR1 VXN=RC 1 IF (N ) 85, 85, 8 85 WRITE(6,902) CC TC 81 9C IFCT=1 X N= 100 C 0 C O C GC TC 98 906 FCRPAT('CINPtT ERCRO') 99' RITE(6,906) CALL SYSTEP 98 NGCC( 1 =0 NCCC:E (2 )=0 NEAC=C 907 FCFMT('CCO',12,' FILESS. STCF'T EVENT',I7) WRITE. (6,r9C7)IFCTMXN IEVCT=O CC 100C IFC=1,IFCT'RITE(6,100) CC C FCPMT 1F1 ) hiFIT(6,110) IFC 82

11C FCFMT (1F,3CX,14f-BEEG1 IING ILEI3) N'6=0 L= IELFIN (25CC,C) 120 IF(N ) 14C,13Ctl5C C REtC PHYSICAL TAPE RECCRD 1 C N~: IPtFIN( I 1) IF(N%) 14C,1CCC,14C C TRANSFER LCGICAL TPACKX RECCRC FRCM BUFFER 14C N.:=IN RECI, I. ) CC TC 120 15C'f=I 1(8 )-( I (8 )/IC) *1C IF (L I IT MM, 1,7))' 16C 14C 16C 16C I S'=:SW (M) GC'TO (200,210,22C,23CS,240,2C26C) tMM 2CC CA/LL MACI(IA,'It S I) CC TC 300 2 1C CALL:ACH2 ( I,!S ) GC TC 30-C 22C CALL MAC-:3 I( A,ISt) GC TC 3CC 23C CALL ACH4( IA, IS) GO TO 3C0 24C CALL MACH5( I A, PI<) CG TO 300 25C CALL MACf6(I,:,I%.) GC TO 30C 260C fCLL NACF7( IA, IS') 3CC MS (M ) =2 IF (I EVCT —MXN) 1 4C, 1CC, 15C 1CCC CONTINUE C N'hi TERMINATE. NtLST CALL EA.CF tVACHINE RCUTINE TO OUTPUT C LAS7 ST-CREC EVENT. 1-CC IS'h=3 DC 2CCO MM=1,7 IF (MS)'(MM).-2) 2CCC 16C, 2C00O 16CC G- TC (161C,1620,1l63C164C,16b0,166CO1 670),MM 161C CALL PACH1(IAOIS) CO TO 2000 162C CALL PACF2IUtISh ) GC TC 20C0 163C CALL MACF3I A,'ISh) G- TO 2CCC 164C CALL MIACF4(1IA, IS'.) GC TC 2000 16%C CPLL YACH5{IA,ISh) CO TO 2000 1660 CALL PE/F 6(IP-IS1:) GC TC 2CCC 167C CALL VACh7(IAISlh) 2CCC I ( 6 t l ) = 1 WRITE(6,2CiC) (NGCC' (J),J=1,2) 2C1C FCRMAT(l1HOFTFEPE A RE,14,19F- GCCC N2 EVENTS AND,I4, 1 16F GCCOn 4 EVENTS.) WPITE(6,2020) NPAE 202C FCPRAT(1 lFOJ-IFEPE R EI4,12Ft PEA EVENTS.) CALL SYSTElf 85

SLBRCLTINE STORE( IA, I S WISET, ITSTIXIYLAST) CCMMCN/CCUNT/ICT,LCWERITCPEACNGO CD( 2 ),NTAPE( 2 DIMENSION IX(4,16),IY(4116) I (2CC0), ISET(20) tITST(4),LAST(4) C ISET ARRAY' C (1) FAIL FLAG C 0 GCCC EVE T C 10 NET PREPER EVENT TYPE C 20 NCT N2 CR N-4 EVENT TYPE C 3C MAX TRACK C.T. 4 C 31 TRACK # G.T. MAX TRACK # C 32 # PCINTS CN TRACK NCT BETWEEN 1 AND 16 C 45 ALL TRACKS NCT PEASLRED C 60 SAME AS =2C C (2) EXP NC C (3) ROLL NC C (4) FRAME NO C (5) EVENT NC C (6) EVENT TYPE (2 CR 4 CNLY) C (7) MACFINE NC C (8) MVAX NC CF TRACKS C (9) CATE (1MCCYY) C (10) TIME (FFPMMSS) C (11-2C) UNIMPCRIANT CC TC (5,20,111),ISW 5 CC 15 K=2,5 15 ISET(K)=IA(K) ISET( 1)=0 DC 17 K=8,20 17 ISET(K)=0 2G eD 25 K=27,5 IF(IAIK)-ISET(K)) 1'1.1,25,t111 25 CONTINUE IP=IA(11)+11 IF(IP-12) 28,2S,2E 28 ISET(1)=10 GO TO 100 2S CE 3C K=%9 lP 3C ISET(K)=IA(K) ISET(7)=IA(8)-(IA(8)/1C)*10 ISET(6)=IA (12) IF ((ISET(6)-2)*( ISET(6)-4) ) 31,34,%31 C CHECK TO SEE IF EVENT TYPE IS N2 OR N4 31 ISET(1)=20 34 ITR=IP(6)/1000 INTR=IA (6)- ITR* 1CO IF(LIMIT(INTR,1,4)) 4C,35-40 35 ISET(1)=30 INTR=4 4C- IF i(LIMIT ( ITR, 1 INTR )) 50,45,5C 45 ITR=INTR ISET(1)=31 5C ISET(8)=INTR IP=IP+3 L T ( I TR)= IA ( IP ) NPS=LAST (ITR) IF(LIPIT(NPS,1,16)) 65~6C965 6C ISET(1)=32 GC TC 80 84

65 CC 75 NPT=1,\PS IX( I'RNPT )=I ( IP+ 1) IY ( ITR,N PT )=I A(I+2) 75 P= I P +2 EC CC 85 J=ITP,INTR 85 I1 1 T(J)= ITT ( ITR)=1 1CC RETURN C NC' GIUPUT CLC EVENT —NEW E-ENT IS NCN IN IPA ARRAY 111 INTR=ISET(8) M..N=I SET (7) I 71= I C T + 1 IF (L I U'IT ( ICT i LCER, IT'P ) ) 12 5,350,125 125 IF(INTR) 14C,14C t13 13C CC 135 J=1, INTP IF(IIST(J)-1) 500,135,50C 135 CCNT INLE 14C 1v=ISET(6)/ 2 IF(L I ITP 1,', 2)) 142141, 142 141 ISEI(1)=60 142 IF(ISET(1) ) 6CC -145,6CC 145 NGCCD( M )=NGCCE (I + 1 N=NTAPEIM) 15C CALL CCNVFT( ISET (), IET (10 ) NPS=O CC 16C J=lIhlTR 16C hP=NPS+LAST (J) C FIRST CLTPLT RECCPE IS RCLLtFRWMEEVENTI #,TGTTL # COORD PAIRS IN EVENT,;C # TRACKS, DATE, TI IE URITE(M,17C) ( I-El(J),J=3,5),hPS ISET(6),ISET (9) ISET( 10) 17C FCRMAT(7 1 ) CE 2CC J=1, I'TF I TK=J*1000+LAS T( J) PS=LAST (J) IF(LIIT(NPS, 1,16)) 175,2GC, 175 175 DC 19C K=1,tPS C TRACK RECECRDS ARE (IRK #)*1CCC+ # POINTS CN TRACKX COORD, C Y CCORC WRITE(M, 8C) IK,IX(JK),IY(JK) 18C FCFMAT(3I10) 1C CC NIN LE 20C CCNTINUE 35C GC TC (5,5,1CC), IS 50C ISET(1)=45 62C FCPMAT (' RCLL',I41,',FRAEE',I4,',EVENT',I2/ 2'',I2,' TRACKS, FAIL CCCF =,1I2) 60C IRITE(6,62C) (ISETJ),J=3,6)tISET(1) NEPC=NBAC+1 C-'C IC 350 ENL F IIE 85

FUNCTION I I UFIN( N ) EXTERNAL ICR CCMMCN /PICK/ IAT CIfENSIGN IA(1) INTEGER *2 NP IF(N) C10C, 1CC, 2CC 1CC NeF= MIAl) IST=l GC TO 400 2CC CC 3CC <=l,NPR 3CC IP(K)=O CALL SICERR(ICR) CALL REAC( IANECLNBP,8, 65CO) 35C ITI=O 4CC IPLFIN=I ST RETURN 50C IST=O GC TC 400C ENC FUNCTION INRECC(IA,IP) CCMPCN/P ICK/ I AT DIMENS.ICN I11), IE(1) L=IBMCDP( IA(IAT+1)) IF(L-200) 5C,1C,1CC 5C IF(L) 100C5CC,1C 1CC L=2CC IC KX='IAT+2, IEXP=IBMDDP(I(KX})) IF{IEXP-8) 2C,3C,2C 20 I 4T=,I.AT+L I N RE C C=- RETCRN 3C CC 3CC K=I,L KX=IA'T+K 3CC IPK)=IB8MDCP(IaIKX)) IAT'=I T+L 5CC INRECC=L RETURN EN C SIteRCLTINE ICR IRITE (6, 10)

10 FCRMAT(' TAPE ERRGR ICNCPED') RETURNEND. S tROUTINE P:CHl(IAIS1W ) DIMENSION IA(2CC) R. ETURN END SLBROLTINE MACH2 (IA ISW) D-IMENSION A(2C00) RETURN SLERGUTINE MACH3(IAlSh) DIMENSION IA(200) RETURN~ ENC SUBRCUTINE MACH4(IA, ISW) DIMENSION IA(200) RETURN ENC SUEROUTINE M/CH5(IPASW) DIMENSION IA(2C00)t ISET(20),ITST(4) IX(4,16 ) IY(4,16) tLAST(4) CALL STORE (IA, ISh, ISET, ITSTI IX IYtLAST) RETURN END SUBROLUTINE F ACH6(IAISW) CIM.ENSION IAf200),ISET(20)- ITST(4)tIX(4,16),IY(4,16),LAST(4) CALL STORE(IAISiISSETITST, TXIYLAST) RETURN ENE SUBROUTINE MPCH7(IA, IS) D IENSICN IA(20C) RETURN END 87

SLHRCLTINE CCNVPT(ICAI, ITI) IF (IT IW-24(CCC) 1CCG 1CC, 1 1 IT.= ITTP -24CCCC C RESET TIE AFTER NICNITE LCAY= ICAl/CC/1 L YfR=' IT - L r A V 1CC LCNTF-=LEAY/ICC l C'Y Lt Y -C N T h* 1 IF(LIFIT(LCNTH, 1,12) ) 5,1CC,5 C NCh CENVERT CATE 5 GC TC (20,4C,2C,6C,20,6E,20,2C,fiC,20,60,80 ), LPCNTF 2C IF(LDCPY-31) cC.87FE7 4C IF(LCEY-28) qCE7,87 6C IF(LECtY-3C) cC,F7,87 EC IF(ItEPY-1 ) _cC, 5,F5 E5 LEAR=LYEAPR+1 L MCN'T- = 1 CE TO nP E7 LM[N TH=LCN TH+ 1 e8 LCAPY= CC TC 95.C LEC:A=LC/Y+1 T ICAT= (LMCENT* I 1CC+LCAY lCC+LYEAR IF ( IT UT-24COCC ) ICC, 1CC, 1 C REPEAT IF NECESSaR~ 1CC PETURN EN C FLU CTIC\ L I tT ( I Ch t C,LF ) I T S T = 0 IF (I l-LCh) 2CC,15C, 11C 11C IF ( IT-LI) 15C,5C,20C 15C ITST=I 2CC LIWIT=ITST RE ILR C SUBROUTINE "IBMDDP" IBvcCP CSECT LUSING *4115 L 1,(C,1) L 1,C(C,1) L COZERC SL 1,2(0) SLCL C,6(C) SL L 1,2(C) SLCL C,6C) St!_ 1,2(C) _ SLECL C,6(C) _ __ SLL 1,2(C) 88

SLCL C,6(C) C tNEG eL KXT S O NEG LNP 0,0 KXT BR 14 NEG DC X'008COOOC' ZERO DC X'CCCCCCCOC END 89

APPENDIX VIII SOURCE PROGRAM LISTING FOR "CHROMOSOME" IMFLICIT INTECER(P-K,P-CS-Z),REAL (R')LOGICA L*1L) CIM ENS ION IDA (TA7), XCCCPC( 100 ), INCEX(100), 2YCCP lO(100)I,SIl0C),'EACD(2) XBAR(2), SIGMA(2) REAL SIGMA ECUIVeLENCE (TRACKS, ICATA(5)),(TICATAf4)) CC~PPC ILCNG(l00), SHCRT(0l0)t, RATIO(100) INTEGER RTCT ICIST (Xl,Y1-,X2Y2)=S:RT (1.*({(X-X2)**2+ (Yl-Y2)**2)). CALL CHECK (LPLCT.,LSUPM) 1C LFIRPS=. TRtE. IND=O LA SIT=.FALSE. LRCLL=.FALSE. LSTCP=.FALSE.' LWRCNG=.FALSE. 90C - FCRMAT (7110) 20 REC (5,9CC,END=9C000) (ICATA( I}),I=17) C LFIRST=TRUE FCR FIRST lATA SET IN FRAME C IND=CATA INDEX C ICAIA=FIRST INPUT RECCRC FCR EVENT C 1= ROLL 2= FRAME C 3=EVENT 4=#CCCIRE PtIRS IN EVENT C 5=*TRACKS (20R4) 6=DATE IMMDDYY) C 7-=TIME (HHMMSS) C LSICP=TRUE FCR EN OCF FILE C UNIT 5=INPUT CATA IF(LFIRST)CC TC 24 IF(FRAME-ICATA(2)) 3C,4C,30 30 T SVE=T+'2 CC 35 I=1l, 35 REAC{5,902)T,ATtXY REAC('5,900END=9000)(IDATA( I ), I=1,7) IF(FRAME-ICATA(2))'31,40,31 31 LAST=.TRUE. CC 32 1=1,TSAVE 32 EPCKSPACE 5 RE {E 5 -,90CC )('ID/PA I ), I =1,7)' C LAL.ST=TRUE EEANS END CF FPRME C LRCLL=TRUE MEANS END CF RCLL AND FRAME C 4C IF( IRCLL-IDATA(l) )50,(C,5C 50 LRCLL=.TRUE. L S'T=. TRUE GC TO 60 24 LFIRST=.FALSE. 25 IRCLL=IDATPl)i el ARM=O FRAME=ICATA (2) DATE= IOATA(6) ETCT=0 90

RTCT=C IIME=ICATA (7') PPT=O TR=C C C RTCT=TCTAL LErNCTF CF SFREAC C PP1=TCTAL #PAIRS IN F-RAME C 901 FCRMAT.('1MEASUREMENT ANALY'SIS'/' ROLL',1I3,' FRAME',13. 2' STARTEE AT',17,' CN',I7) IRITE (6,'SC1) IRCLL,FFAPE,TIME,CATE GC TO 100 9CCC LSTCP=. TRUE. GC TO 50 C EXPLANATICN CF LSTCP AECVE 60 IF(LAST) GC TC 7CCO C ENO CF FRAME SEQUENCE IS 7CCC 70 IF(LRCLL) CC TE 8000 C ENG CF SAMFLE SECUENCE IS 800C 100'INC= IND+1 C INCREMENT CATA INEEX IF(1CO-IND) lClICltlC2 101 CALL IFAIL (10) 1C2 INDEX (IND) = ICETA(3) CATE2 = IOATA (6) TIME2=ICATA(7) PPT=PP'T+T TR=TR+TRACI<KS C C T=#CCCRD PAIRS IN EVENT C PPT=TCTAL PAIRS IN FRAME C TR=ICITAL TRACKS IN FRAME C BIARM=TCTAL #eIAPRM CHPCFNS C IF (TRACKS-2) 11C,12C,11C 120. RATIC(INC)=C. SHCRT (INC)=O GC TO 121 11C PIARM=RMeIRM+1 121 ACCUMp=O CTN=O C C CX,CY-CLC X ANC Y C ACCUP=LEINGTH ACCUFULATCR C CTN=CURRENT TRACK NC. C NPCT=NUMPER PEINTS CN TFACK C CPCT=CURRENT PCIlNT CN TRCK C C-TI=CLC TRACK nUMEER C TCAT=TRACK#*1OCC+#P'TS [N TRACK C C'C 15C J=1,T S02 FORMAT (311C) REEC (5,902, END=14C,ERR=14C) TAT,,Y CTN=TCAT/1000 IF(CTN-CTN) 122,123,122 122 CINT=CTN CPCT=1 NPCT=fCD(TCAT,1CCC ) CX=X CY=Y 9i

125 XCCCRC (INEJ=X YCccRc INC- )}= GC TC 15C 123 CFCT=CPCT+l ACCUP=ACCU+ ICD IS1T X,Y,CX CY) CX=X C'YCEv=Y IF (NPCT-CPCT 1'aC,13C,15C 13C IF(CTN-2) 132,131,132 131'ACCLt=ACCUL/2 C FINISI-ED WITH ONE ARV ILCNG ( INC) = - C CCU RTC T=-PTT+t CCUV ACCL.M=0 CC IC 150 132 IF {( CTN-4) 15C,133,11 133 ACCUL=ACCLM/2 RIC T=PRTOI + CC SHCRI ( INOD:=M INC ( I LCNG (IN ),ACC UM) ILCG ( INC )=MAXC ( ILCNG INC ) ACCUM IF.(SHCPT(IND ) 16,160,134 1.34 CCNTINLUE RA'TIC (INC)=(ILCNC( INC )*1.0)/SHCRT(IND) 16C IF(J-T) 161,20,161 161 CALL IFAIL(3C) 150 CCNTINLE GC TC 20 C3- FCPMAT ('CTRACK =',I3,', PC IT =',I 3,'t EVENT =',I4) 14C WRITE(6,903) CTN,CPCTIINE CALL IFAIL (2C) C C ENC FRAME SECUENCE 904 FCRMAT ('OFINISHEE AT',17?,' ChNI7) 70CC WRITE 16,9C4) TIIE2,CIATE2 CALL ILAPSE (CTE EATE 2, [A IET E2,TE 2 ) C R=PCINTS/ARM R=PPT/(TR*1.0) J=I N-PI AR' C J= #TEL S 905 FC:RMAT ('0',I3,' CFIRCPfCSCNES,',I3',' TELCCENTRICS,',I3,' BIARMS'/ 4'','5,1 TRACKS MEASUPEC,',F5.2,' PCINTS/TRACKI) WR ITE (6,905) INC, J E I A PS. T 7, R C R=AVE. L ENGTF-/CHREPCS CIE C R1=AVE. LENGTF/ARP R I=RTCT*1.C R = R 1/ I N R 1=Rl *2/TR 906 FCRMAT ('OL ENCGIF S... CTAL =',I 1C,',A VE/CHRC M=',F 8.3,t AVE/ARM= It 2FE.3) hRITE (6,9$06)R ICT,R 1 CALL MCDIFY(PRR1, INC) CALL CRDER (IHECtESINCLWRCNG) IF (LWRONG) CALL IFAIL(4C) EC 72C0 J=1,2 X PR (J ) = 0 S I C PA(J)= CTN=O 92

C CTIN IS LICP CFECK HL= EC ( J IF(HL)720C,72CC,7120 907 FORMAT ('OCODE INCEX CCC~RD YCECRD LARM SARM', 2'RATIC LENGTFI'/') 712C WRITE(7,907) 7160 CTN=CTN+1 TCT= I LCNC F-L )+SICRT L XPAR (J)=XXBAP (J )+TCT SIGMA(J)=SIGMP (J)+TCTtTCT gce8 FCRMAT ( v',616,F6.2, I6) WRITE (7,908) Ch IN-DEX (FL),XCCCRC ( FL ),YCOCRC (L ), ILONG(HL)t 2SHCRT (HL )RAT IC~4 ) TCT IF(S(FL) )72CC72 CC, 714C 714C HL=S(HL) IF(IN'-CTN)715C,715C,7 160 7150 CALL IFAIL(60) 72CC CCNTINUE 730C. CTN = INO-PIARP IF (IARM. EC.-C ) E I RP 1 IF(CTN EQ.C)CTN=1 SIGMA (1 )=( EIAPl*~ ICMA (1 )-XEA.R ( 1)*2 ) /(B IARM*P IARM) SIGMA ( 2 =( CTN*S I-A (2 )-XEAR(2 )**2 )/( C'TN*CTN) R= SQR T i( S IGM A ( -1 ) *1.C) RI=SQPRT(SIGMA (2)*1.0) XBAR(1)=XBfAR'.1)/E I'APM XB.AR (2 )=XBPR (2) /CTN S 909 FCP M AT ('O P R MS' ) 910 FCRMAT ('OTELCCENTRICS') 911 FCPMAT(' N =',I3/- XBAR =',I5/' VAR =',F8.0/' S.D. =-',F5.0) WRITE (6,t90) WPRITE l6,911)EIRMBAR( 1),SIGMA( ),R.WRITE'(6,91-) WRITE(6,911)CTN, XBAR2),SIGMA(2),Ri IF(LSLMM) CALL STORE2(XEARSICMA,EIARMCTN) IF(LPLCT) CALL FLCTR1 (INC,EIARt,LFRAM) LAST=.FALSE. INC=0 IF(LRC- LL) GC-'TC 8CO0 GC TO 25 C END PC-PULATICN CR ENC RUN SECUENCE 80CC IF(LPLCT) CALL PLCTR2 IF(LSLMM) CALL S'TCRE3 IF LSIOP) CPLL EXIT FRMP0 INC=O LRCLL=.F ALSE. IF (LSTRE)CALCL STCRE1 GC TC 25 ENC9 975

SLP RCLT I:E IL APSE (C1 E2 tl1,t72) IFLIC IT INTECER ( -Z) F(DL-D2) 1C,2Ct1C SCC F[RMPT ('OMEtASU INC- TCCK NIRE ThAN. 1 CAY') IC hRITE(6,9OC) E TLR N 20 kHl=TI/lCOCC tF2=T2/ 1 CCCC T 1=T 1-HHl* CCCC 12=T2-FH2* 1CCC FFI lI=H-2-HH 1 NN 1-1T /1CC lvl=Tl2/lC IF (MN l-MV2) 3C,30,21 21 IP'2=MP2+6C FF.I=FF1-I 30 1Ml: 2-VM1 SSl=C.F ( Ti1,1 CC S S2.=vf' { (2, 100 ) IF I 1-ScS2 )4C,4C C,3 39 SC2=SS2+60 f M 1 = t F 1- 1 4C SSl=SS2-SS1 901 FCRM.T {' ELAPSEF TIME=',1I3,' RS.',13,' MIN.',I3,' SEC.'I) RETLRN SLERPUTINE MCOIFV(R Rl, I [N) IFFLICIT INTECER (A-G,S- 7, REAPL(R) C RELA'T IE ME S. EETERI INED U S ING A E. CHREM. LENGTH C RETUR \iPVALUE IS REL. LENCTElCO'O0 CCIvCN LUCNG lCC), SECRT(ilCC ), RATIC(1QO) EC 10 I=lI INC LCNC ( I )= (LC(C ( I 1CCC /R iF (SHRT {(I) 2C,1C,2C 20 SIFRT ( I )= (SCT (I }'I CC) /R 1C CCNTINUE RETURN FEN StLR(OL, TI:NF IF It( I 9CC F C F A T ('OER C R...RET URI CD E =', I4) 94

R ITE { 6 C ) I CALL EXIT EkE SLtECUTINE CRCER (t-EAC,S"INO, SWRCNG) II.FLICIT INTEGER(P-GS-Z), REAL (R) CCPiEN LARMI1LC), SARP(100), RATIC(100) DIIENSION -(I ),{ NE),E (I2) LCG ICAL SWRPCNC* 1 HEAP( 1 )=C F E EC { 2 )= P=C ICC Iv=M+ 1 IF(M.GCT.INE) R ET RN S(M)=0 J=2 IF (RAT I (M).G.C.C )J= 1 IFF(HEAC(J).NE.C)CC TC 110 ~E/C."(J) =K GC TO 100 11C LSI:=C I-=HEAP ( J) 20C IFt (LARM(M)+SARMI}) ).LT.(LARM( I )+SARM(I ) ) )GO TO 21C IF(LA ST.EQC.0)GC TC 205 S (LAST)= GC TO 10l 2C5 S(M)=I GC TO 100 21C IF(S I).IE.O)G 7C 22C S(I)=~ GC TC 10L 22C L AST=I I=S(.I) IF (C.LT. I.AND.I.L T. N.LAS'T.I E.I)GC TO 200 SCRGNC=. TRUE. 9CC FCRPMT("O ORDER/LPCF EFRCR - CFRCt.OSCME'E,I5) WRITE (6, 9C0)M RETURN. END SUPRCLTINE CFECK (LPLCTLSUVM) IPFLICIT INTECER (A-K,1-Y),LCCICAL*1(L),REAL(Z) REAL R EtTA PT,YN/P /'Y','T',~''N'/ PREt INC 5 9Cc FCRPAT ('OBATC- CR TEf-INAL??') iRITE (6,9C00) SCI FCRMtT (Al) 5 REAC(3,9C1,ENC=88E). IF(A-E)10,20,10 10 IF (A-T) 3C,4C,3C SC2 FCRMAT (' INVAPLIC WCRE.C.ENC B CR T') 30C hRITE(6,9C2) GO IC 5 20 LPL CI= TRUE. 95

L SMM=. TRUE,. CALL ~STORE1 CALL PLCTR3 C - BATCH DEFAULTS TC ALL FPSSIBLE PLCTS AND SUMMARIES 903 FORMAT ( F. 1 ) C BATCH REQUIRES NEXT TWC CARES HAVE FILE AND RECORD IN FLOATING POINT 90.4 FORMAT(''WHAT FILE OCES DATA START IN (FLOATING PT.)?I' 50 WRITE (690C4) REtD ( 3,S03 tEKD=8 8,EP-R='777)R NF=R-1. 905 FEIRP-AT {' WfAT RECCRC') } WRITE (6,905) READ(3,9C3,END==8E8ERR=777)R NR=R-1. CALL SKIP (NFNR',5) RETURN 888 CALL IFAIL(8C) 777 CALL IFAIL(5C) 906 FCPRMT( WANT SUFPARY?' ) 40 1RITE6,'906) C TERMINAL DEFA.LTS TC NC PLOTS REDC (3,t901 ENC=EC,,ERR=777 )A IF A-Y) 7560,775 6C LSUMM=.TRUE. C-ALL STORE1 7C LPFLT=.FALSE. gC TO 50 75 IF(A-N)9C,8C,9C 907 FORMAI(' INVALID ANSkER....YES OR NO?') 90 WRITE(6,907) GSC TO, 40 EC LSLMt=.FALSE. GCO TO 70 EN. SUBRCU'TINE ST CRE1 ENTRY STCRE2 ENTRY. STORE3 RETURN StBRCLTI NE PLCTPR CCPMCN LONG(lCO).SHCRT(10C),RATICOl00) INTEGER A(52),C(52).,IN'CeBIARMN, LINGtSHORTIMAGE(2784), IDATA(6) DIMENSION X(52), IRAT(3) LCGICAt FRAME*1 DPTP IDATA /'FRAC''TICh',' CF',1CHRO','MOSOt''MES I/ 2,IRAT/'ARM'tRATlItC I/ C A, iS STCRAGE FCR ARM LENGTH C ISTR'IBUTION, C IS FOR C CHRC~CSOME LENGTH CISTRIBUTIChN X IS FOR VALUES OF X C CCORD INAT E 96

c LEN(JE)=(JP+5)/5C+1 C LEN CCMPU'TES SUPSCRIPT CC I 1=1, 52 1 X ( X(I)=( -1)*,005 5 CC 10 I=1,52 Af I )-C 10 C(I)=c CHREC=0. C CHRCM ANOC ARM ACCUI:ULATE NUMBER OF C'BJECTS BEING CONSIDERED C IN THEIR ASSCCIATED DIS'TRIBUTICNS CVER=C CALL PLOT2 (IVAGEB,2.5 C.,4.0, 1.C) RETURN ENTRY PLCTR1 {INCeIAFI,FRAE) CC 30 I=1,INC IF ( SHORT( I )2 }9 31,32 C 29 NEANS ERRCR - NECATIVE SHCRT ARM LENGTH 29 CALL IFAIL47C) 31 JC=LEN(LCNG(I)) F ( JC-52 ).34,34,3 3 33 JC=52 34 A ( J C ) = A ( JC )+ 1 C(JC)=C( JC)+l GCC TO 30 32 JC= SCRT ( I )+LCNC ( I X2=JC/1000.C. IF(4.C-RATIC( I)) CC, lC, lC1 ICC CVER = OVER + 1 GC Ti 102 101 CALL FLLCT3(' X,X2,RATIC( I ),lt4) 10 C 2 CCNTINUE JC=LEN(JC) IF ( JC-52 ) 326,335 35 JC=52 36 C(JC)=C(,JC)+1 JC=L E(SRT ( I ) IF(JC-52)3E,3E,3 39 JC=52 3E A(JC)=A(JC)+l JC=LE ( LCNG (I)) IF(JC-52)41,41,4C 4C JC'=52 41 A (.JC)=A(JC)-*1 30 CCKTINUE ARP=AR+IkNC+BeIAR C-PCM=CHRCM+ INCRETILRN ENTRY PLCTPR2 s03 FCFPA'T 1 1RATICS > 4.0...',I3) NRITE(7,903) EVER CALL PLC-T4 {9,IRP'( 1)) WRITE(7,901 ) CALL PLOT2 ({IACE,2.5,C.,C.3,C.) CC 8C I=1,51,2 Y= (I ) (I+1) (I)=Y'Y=Y/APM 97

CALL FLOT3 ('I'X)(I),Y,1,4) Y=C( I)+C I+1) C (I )=Y Y=Y/CHRCM CALL FPLOT3('*',X(I),Y,1,4) 8C CCNTINUE 9CC FCRMAT('LFREQUENCY DISIRIBUTICN CF ARM (X) AND CHRCMOSOME (*4 LENG 2THS'/'',F4.O,' ARMS MEASURECD'/''F4.0t' CHRCMOSOMES MEASURED') WRITE 6,9CO C)A PRMCH-Rc' JC= (52 )+C (52) 902 FCPMRT (',13,' VALUES > 2.5') WRITE(6,902) JC CALL PLO'T4(23,ICATA(~1)) sCl FCRMAT ('O' 20X,'RELATIVE LENGTH') WRITF (6,901)'04 FiCRMAT(' ARMS'26I13/' Cf-PCMS',:2I3) GC TO 5 ENC 98

APPENDIX IX SAMPLE SESSIONS USING "CHROMOSOME" Although "CHROMOSOME" is usually run in batch mode, it can be run from a remote teletype terminal to check data or program operation. Operation of the program depends slightly on the mode being used as can be seen in the examples which follow. OPERATION FROM A REMOTE TERMINAL: All lines entered by the user are underlined, control characters used are in lower case type. MTS (LA20-0036) WHO ARE YOU? UM MICBIOAAAs #$SIGNON SF40 PW= s #**LAST SIGNON WAS: 09:24.40 07-15-69 # USER "SF40" SIGNED ON AT 09:38.14 ON 07-15-69 #$RUN *MOUNT PAR=G113 9TP,*TAPE*,MODE=1600,SIZE=80,'SF40-l's #EXECUTION BEGINS DONE #EXECUT ION TERMINATED #$RUN CHROMO 5=*TAPE* 7=*DUMMY*s BATCH OR TERMINAL?? TERMINALs WANT SUMMARY? The summary option is NOs not operational at this time. WHAT FILE DOES DATA START IN (FLOATING PT.)? 1.s WHAT RECORD? 1.s MEASUREMENT ANALYS IS ROLL 40, FRAME 1 STARTED AT 2757 ON 70969 FINISHED AT 12116 ON 70969 ELAPSED TIME= 0 HRS. 53 MIN. 19 SEC. 47 CHROMOSOMES, 1 TELOCENTRICS, 46 BIARMS 186 TRACKS MEASURED, 5.96 POINTS/TRACK LENGTHS...TOTAL = 11634,AVE/CHROM= 247. 532,AVE/ARM= 125.097 99

B IARMS N = 46 XBAR = 1014 VAR = 148 390. S.D. = 385. TELOC ENTRICS N = 0 XBAR = 0 VAR = -0. S.D. = 0. #EXECUTION TERMINATED #$S IGNOFF OPERATION IN BATCH MODE: Printed lines which were read off data cards are underlined in the batch session which follows. $RUN *MOUNT PAR=G:13 9TP! *TAPE* MODE=1600 SIZE=80, SF40-1' EXPECUTTON BEGINS DONE EXECUTION TERMINATED PRUN, CHIR.CR. 9=*T:TAPE* EXECUT TCN B EGINS PATCH CR TERMINAL??? Three data cards are WHAT FILF DOES PATA START IN (FLCATING PT.)? read in at this point WHtT FFCRD? but are not printed. They contain the data requested. 100

MEfISUREPENT ANALYS'T! RCLL 9, FRA.MF I STARTEC AT 211202 Cl: 5'186S FINISHED AT 214634 CN 51869 ELPSEOC TME-= 0 HRS. 34 MIN. 32 SEC. 64 CHf.CMf0SOlESv, 50'TELCCFFNTRICS, 14 Bl IARS 156 TRACKS P.ASURED, 6.57 PCINTS/TRACK LEtGTF-S..TOTAL = 9317,AVfE/CfPCM= 14.5'78, AVE/ARM- 119.449 COCE INMFX XCCRD YCOGOPC LARP SRP PRATIC LENGTF 1 46 -12%50 1. 17C 1449 13C5 1.11?754 2 6.4 -1573 62 1552 1C30 1.51 258E 3 18 -1018 852 1538 989 1,56 2527 4 4 -18 446 1243 12C8 1.C3 2451 5 331 -637 272 11IC 1016 1.10 2135 f6 57 -1870 - 163 1C64 1.0C: 1.Cl 2114 7 47 -1130 -82 1408 597 2-.3 20C5 p 36 -412.-2S4 1016- 892 1. 14 1SC8 9 22 -1227 123S 954 748 1.28 17C2 10 15 -768,11.98 1078 487 2.21 1565 11 56 - 17 1 1 -252 776 769 1.0 1 154-5 12 9 -206 725 1078 2e8 3.74 1366 13 48 -1125 -4P7 789 535 1.z7 1324 14 50 -1340 -5c3 686 350 t.96 1036 C Oa INDEX XCC.RC YCCCRD LARPM SAR. R. PA TIC L EN G'T H 1 26 -1507 1032 12002 C 0.0 12C2 2 24 -1483 1.224 1 fC C cO.0 1 160'3 7 36 718 1133 C 0.c 1133 4 12 -587 1?60 1126 C C.C 1126 5. 5- 5 -56 -5 4 0 C C 954 6 1 470 26 947 0 C C 947 7 40 -873 425 934 C C.C 934 98 53 -1434 -739 927 c C.0 27 9 14 -9.10 1487.2C C C.C 920 10 23 -1227 989 90-6 C C.C 906 11 2 53 -73 F92 0. C.0 89 2 12 45 -1C65 6-6.6 886 C C.C 886 12.3 30 -566 553 P86 C0c. 8P6 14 16 -732 1C'8 872 C C.C 872 15 29 -468 5'7 P:65 C C 0 865 16 49 -1040 -650 831 C C.C 831 17.41 -1057 20C6 81 0 C.c 8311 18 42 -917 115 F17 C C.C 817'1C. 11 -59. 172 8 03 C C.C 8C3 20 3?2 -600 1.44 789 0. C,.C 789 21 25 -1589 1472 789q C C.C 789 22 10 -267 1315 776 C C.0 776 23 21. -9s5 1125 744 C C.C 748 24 37 -350 -39Q 714 C C.C 714 2; 65 -1667 -68 70n 0 C.0 7C0 26 51 -1375 -8C0 686 C C.C 686 27 35 -73C -15C 686 C.C 686 28 33 -51 75 6C 0 O C 68C 29 5-4 -1569 -447 673 C.C 6'73 30 34 -64? -0 666 C 0.C 666 101

31 28 -1772 8=5 659 C C.C 6 3? 60 - 4193 46 645 C C,C 64-5 33 62 -1549 3c3 638 O CC 638 34 43 -869 -26 638 C C.C 638 35 63 -1629 2 E P 63 C C 6 31 36 61 -1472 482 631 C C.C 631 37 8 -13C 782 6631 C C C 63 38 52 — 1490 -c76 604 Q C.C 604 39 20 -1C84 96' 7 6C4 C CC 6C 4 40 13 -6P5 1303 604 C CC -6C4 4 ], 3 52 35. 6 C4 C C.C 6C4 42.44 -976 -70C 597 0 C.C 5c7 43' 27 -1756 988 583 0 C.C E 3 44 5' -1893 -54S 549 0Q C.0O 544.5 39 -801 562 480 0 0.C 4 F 46 59 -2040 13 1 -473 0 C.C 473 47 19 -.1C 15 775 43s C C.C 49 48 38 -752 -511 432 C C.C 432 49 6 -6C =C9 364 C C.C 3(4 5C 17 -922 10.2 336 C C.C 336 N = 14 XPAtR = 19-29 VAR = 259419. S.D. 509. TFLCCFNTRICS N = 5C XP.AR = 738 VAR — = 3 479 S-.'Co = 196. 102

ME. SUREME.NT ANALYSIS RPLL c, FFRA.E 2.ST/TF! AT 2149C4 CN 5186' \ F INTSHF PAT.2 2 22' 1.F'N 51869 ELAPS' TMF- 0 - C PS. f?4 MIN. 12 SFC. 64 CHR0C'SCOME, 52 TFtCCFNTPTCICS 12 BTAP. S 1592 TRA'CKS MEASUREOt 7..44 Pr.INTS/TRACK LFNT-TFS...TCTL - 11318,AVE/rFPCR.= 176.8z44, V'FE/,APM= 148.921 rCOE E TINEX XC-OP,. YCCCPD LAPR. StRP RPTi Lf f t\GTF 1 13 — 4 C 5 59c9'175?2 1 02 1. 71 27'75 2 69 -792 166 1.0 1 24'4 1. C5 2544 3 93 -1.412 55-2 1 8 83' 1. 22 21EF. 4.10 8 - 1. 2.1 -214 1. C Ic..14 2 1 1S 5 5?21 2(C 1057 1]COC 1. C 2057 b 95 -1603 1077 1543 424 3.64 19.7 7 26 -885 745 1C80 86.29 lq 16 8 77 -962 -9 33 7 7 16 21 1 7C?. q 75 -E3- -3 1 1159 497 2.33 1656 10 92 -1220C 969 791 74 C 1 C7 15'1 1 81 -1010 -I 14 65,0. 22 2.c45 87C 12 89.-1254 2c2 5-42 282 1,92 824 C C.E INDEX XCCROD YCCRPO!L,PRM S.aPR. PAT.TIf LENGTF 1 105 -167C 381 1453 C C.0 1t43 2 25 -6S7 702 1424 C 0.C 1424 3 20 -69-3 1.41 3 1249 0 C0.C 124S 4 83 -1285 -607 1t5'. C C.0 1159 5 17 -244.16CR 11.42 C C.C 1142 E 91 -1187 -65 1136 C C.0. 116 7 10.6 -.1576 -75 1C8.; C CC 10C85 8 98 -1246 125 1051 C C.C 10f1 107 -1,592 -4C C 2 1C 102 C C.C 10 12 1i 99 -1621 12 27 10CC- C C.C O1CC i1 79 -612 -q94 10-00 C C.C 10CC 12 73 -610 -38 995 C C.C 995 13 87 -1521 -557 S7P C C.C'78 1.4 9 -110 -625 97E8 C'C-.O 78 15 23 -954 1 384 C72 0 C.C 9'72 16 18 -294 1350 955 CC. C 9; 17 78 -272 -820 9 16 0 C.0 916 1 8 24 -1026 153 99 0 C. C 899 19 1 9 - 528'38= q3 C C,. 893 2C 74 -610 -2EC R87 C0 C. e87 21 8 12 1C3 84 C C C 848 22 94 — 187 931 825 C C.C 825 23 80 -996 -10c8 825 C C.C 829 24 7 278 21 9. 02 C C.C 0.C2 25 12 -32. Cf C 797 C CC 7c7 26 3 91 823 769 C C.C 76c 27 104 -1846 327 757 C C.C 757 28 102 -1868 565 746 C C.C 746 29 86 -15C4 -4CC 723 C CC 723 30 1 257 55 7C6 C C(.C 7C06 31 11 -2CF 759 701 C C0C 7C1 32 10 -147 424 6.95 C C.C 695 105

~3 4 -44.96C 689 C C.C 689. 34 76 -794 -456 678 c C.C 678 35 S7 -1428.1292 672' 0 C.C 672 36 98 -1467 1, 75.2 667 C C.C 667 37 27 -837 519 667 C C.C 667 38 22 -672 1150 667 c C.C 667 39 2 148. -722 -61 C C.C 661 40 100 -18 845 655 0 C.C 655 41 84 -1254'-1S6 655 0 CO 655 42 90 -1389S 274 65c 0 C.C 6 5(C 43 101 -2C96 672 61C C C.O 61C 44 15 -18S 1351 53 C C.C 593 45 103 -2C84 482 576 0 C.C 576 46 82 -1204 -980 531 C C0C. 531 47 16 -22C 1440 514 C C0, 514 48 96 -1537 126-.4-5 C C.C 458 49 14 -33 8 72 424 c C0O 424 50 28 -786 480 373 C C.C 373 51 85 -1345 -22C 361 C C.C 361 52 21 -'73 4 132'2 339 C C.C 339 BIARMS N = 12 XBAR = 1845 VPR = 312844. S.o. - 559. TELCCENTRICS N = 52 XPBR. 804 VAR = 61409. S.C. = 248. 104

RfiTICS > 6.(} o.. C 6.000 +- 4 + t 4 ~...........t, ~ i + I I I I I I I I I I I I I I I I I! I'I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I! I I I I I I I I I'I I I' I I I I I I I I I I I ~ I I I 5.000 +..........I-.........+, +- t i +........'-4 +,. I I I I I I I I I I I I I I'I I I I I I I I I I I I I.I I I I I I I I! I I I'I I I I I I I I I I I I I I! I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I "4.000 + -+ 4 ~,~ 4 -, + +..........+ A I I I I I I I I I I R I I I I I I ~I iI I I ~, I I I I I I X I I I I I I I I I I I I'I XI I' I R I I I I I I I I. I I I 0~. ^ I I I I I I I I I I I' k.n T I I I I I I I I I I I I I! I I I'I I I I I I 0 I! I I I I I I I I I?.000 + — + 4 ~ t.~ ~, + ~. + I I I I X I I I I I I I I I I I I I I I I I I I I I I I' I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I X I I I I I I I I I X I I I I I I I I I I I X I I I I I I I I I I I I I I I 2.000 + —' -, ) + t-X +.........4 +.........+.........+....... —— + I I I I X.... I I I I I I I I I I I I! I I I I I I I I I I I. I' I I I I I I I I I I I I I I I I I I I I I X I I I I I I I I I I I I I I I I I I I I I'. I I X I X I I I I' I I I I.I.' I X I I X I I I I I I I'! IX I X I X X I I 1 ~ 000 +.........4 + + +-' +- X.......+........+ —— X....4 X-+.. O. 000.2~0.500.750 I'-000 1.250 1. 500 1. 750 2.00C 2.250 2.500

FRFQUFNCY DISTRIBUTICN CF ARM (Xi ANC CPPRCM0SCME (4) LENGTHS 154. ARMS MEASURED 12E. CHRCMOSCMES MEtSUREO 300 ----------------— + —--- + —-- ---— + - ----- -------- I I I I I I I j I I I I I I _ I I _I I I I i T I I I I I I I I I I I I I I I I I I I) I I I i1 I 1 1f 1 _ _ _ _ _ _ _ _ I I _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ r _ _ _ _ _ _ _ I T I I I I I I i I I I I I T IIII I I I I I I I I I I I,2240 ---------- --------------------------- + —----— + —------ - ----------— + —-----— + —-----— + —------- I I I I I I I I I I I I 1 I I' i I I _ _ _ _ _ _ Ir _ _ _ _ I _ _ _ _ _ _ I _ _ _ _ I _ _ _ I I1111 R I I I I I I I I I I R I I x I I I I I I I A I I I I I I I I I I I C I I I I I I I I I I I T I I I I I I I I I I I o 10 — 1 —------------------------- ----------------------- --------- --- + —------ ---------- + N I I I I I I I I I I _ _ _ _ _ I I I _ _ _ _I I I I I I I I F ~ ~ ~ ~ ~ ~C I I I I I I I I I I I F I TI I I II I I I I I I I I I I I C I _ _ _ _ _ _ I __ _ _ _ _ _ I I I I ~I I I H I I I I I I I I I I R 1T ____1___ I XI I I I I I O I I I I I~ I I I I I I M * 120 —-------------------------- ----------- O3 j 1 1 I X I I I I I I S I ______ I ______ I I X X I I I I I I 0, T *I I I I I I N I I I I * I I I I I I I F I I I I: I I I I I I I S I I *I I I I I I I I I I I j I'-I I I I I ~ I I I I I I I I ____I I I I I I I f 1 1 I. I I I. 060 + ~ ~ - -+ ~-~ —~X- ~ -~~~ + ~~ —~ + —- X —--- + ~~ —~I+ ~~- ~ -+ —------— +- 4 ~ + ~ -4 I I X I * 1 1 I I 1 1 * 4- 1. 1 I 1 1 I I I 1 14- I I I I I I I 1 I TX I I I xI I I I I _ __I I I I x * I I I I I I I X * I * I I I I XT I I I 4-4 - X4I * I I I I I I I IX I I I I I I I I I I I 1 * XI 1 1 4- I 0.000 * — 4 —0-+ —-- * + -— +- ------— + —I —-— X~+-4 —-X —-X —- — *+*- - RELATIVE LENGTH

APPENDIX X ERROR RETURN CODES USED BY "SEARCH" AND "CHROMOSOME" Error return codes from "SEARCH" (none of these are fatal) CODE ERROR 10 Not proper event type 20 Not an N2 or N4 event type 30 Maximum track number greater than 4 31 Track number greater than maximum track number 32 Number of points on track not between 1 and 16 45 All tracks not measured 60 Not an N2 or N4 event type Error returns from "CHROMOSOME" (all are fatal) CODE ERROR 10 More than 100 chromosomes in one frame 20 End of file mark or error encountered in track record read sequence 30 Number of coordinate pairs on tape not same as declared number 40 Error (loop) encountered in ordering chromosome data 50 Error encountered in attempt to read information in subroutine "CHECK" 60 Unrecoverable loop discovered in data output sequence 70 Negative short arm length discovered 80 Logical end of file mark encountered while attempting to read control information 107

APPENDIX XI SUPPORTING PROGRAMS These programs are not yet completely documented. 108

THE EFFECTS OF ENVIRONMENTAL SELECTION PRESSURE ON THE CHROMOSOMAL COMPLEMENT OF L-M MOUSE CELLS AND BHK21PD4 CELLS Richard Giles, Gregory Baumann, Edward Lunk, and Richard Platte, Jr. 109

SUMMARY L-M cells, normally grown in modified medium 199 plus.5% Bacto peptone, in monolayer were subjected to the selective effects of supplementation of the growth medium with fetal calf serum, growth in spinner culture, and growth in a 1.5-liter fermenter (with automatic control of pH and OR potential). The chromosomes of several populations of L-M cells, which had been exposed to one or a combination of environmental changes, were analyzed to determine if any detectable alterations had occurred in the number or type of chromosomes, or the frequency distribution of characteristic marker chromosomes. Population samples were selected to investigate the stability and reversibility of detected changes. BHK21PD4 cells, were exposed to environmental changes similar to those which challenged the L-M cells, in order to observe the effects of various selection pressures on a near diploid permanent cell line, It was of particular interest to discover whether or not a cell population was significantly altered in chromosomal pattern during the process of going from small scale monolayer cultures to large scale fermenter cultures via spinner culture. MATERIALS AND METHODS Chromosome preparations were made according to the procedure reported previously (1). Colcemide exposure was varied from 4-8 hr for BHK21PD4 cells, to 12 hr for L-M cells. Metaphase cells were selected for analysis based on favorable fixation and staining of well spread chromosomes. Spreads were selected which appeared to be complete; i.e., were oval and/or had a visible haze of precipitated cytoplasm. Spreads with complexes of overlapping chromosomes were rejected. The number of chromosomes, number of biarmed chromosomes, number of telocentric chromosomes, and number of chromosome arms were counted. The frequency and distribution of two marker chromosomes characteristic of L-M cells (2), and of an unusual morphology (see Figure 1C), which have been previously used to follow changes in the chromosomal pattern of L-M cells on a finer scale than chromosome number (1), were tabulated. Chromosomes, which appear quite similar to either the E or F marker (see Figures lC-3C), were scored as E or F derivatives (E' or F'). This was done because of the fact that variations in coiling, fixation, staining, and the physical orientation of the chromosome can make the identification of all marker characteristics difficult in a particular cell. The E marker will often not show the secondary constriction nearest the centromere for the above reasons, and was scored as an E derivative in this case, F derivatives were of two major types. The first type of F derivative corresponds to the E' type of chromosome; i~e, a long subtelocentric chromosome is observed with one constriction 110

MARKER CHROMOSOMES OF L-M AND SUBLINES _ a E G H - -_ -_ F Figure 1C. Marker chromosome of the L-M strain and substrains. 111

128 L-M 199PSP FCS2.* (mono.) 5/8/68 27% L-M 199PSP FCS5 (sus. spin.) 5/6/68 FDL3 L-M 199PSP FCS5 sus. NB 7/24 L-M 199PSP FCS5 sus. 2817 L-M 199PSP FCS5 (sus. ferm.) 8/19/68 4 spin. (sus. ferm.) 6/13/68 spin. * (sus. spin) 7/17/68 sus. spin. sus.{ spin. 2899 L-M 199PSP FCS5 2832 L-M 199PSP FCS5 (sus. spin.) 10/24/68 (sus. spin. ) 8/11/68 sus.| spin. 147 SL-M 199PSP FCS5 2860 L-M 199PSP FCS5 (mono.) 10/28/68 (sus. spin.) 9/10/68 mono. sus. spin. 164 SL-M 199PSP FCS5 2884 L-M 199PSP FCS, (mono.) /17/69 (sus. spin.) 10/10/68 P2994 L-M 199PSP FCS5 (sus. spin.) 3/17/69 165 SL-M 199PSP FCS5 (mono.) 3/29/69 mono. 1 170 SL-M 199PSP FCS5 P3071 L-M 199PSP FCS5 170 SL-M Series 2 (mono.) 5/16/69 (sus. spin.) 6/18/69 199PSP FCS5 4 (mono. 5/16/69 Figure 2C. Origin of L-M monolayers and suspension cultures. 112

BHK21PD4 MEM TP1 o CS5 (mono.) BHK21 PD4 mono. Suspension #49 (sus. spin.) 7/13/68 NB8 BHK21PD4 72 BHK21PD4 MEM TPlo CS5 Special Medium (mono.) 6/17/69 (ferm.) 7/25/68 mono. BHK21PD4 EME TPlo CS1 80 BHK21PD4 MEM TP CS BHK1PD4 Speical Medium (sus. spin.) 7/23/69 (mono.) 7/23/69 (sus. spin.) 7/23/69 Figure 3C. Origin of the BHK21PD4 monolayers and suspension cultures.

more or one less than normally seen in the majority of F chromosomes. The second type of F derivatives are subtelocentric chromosomes, which appear to have resulted from a structural modification of the F marker, and always involve a more extensive difference than the presence or absence of a constriction. A marker and an apparent derivative are only observed at a low frequency (2%) in the same cell (1). The history of the L-M cell and the medium and culturing procedures have been reported elsewhere (1,3). Figure 2C presents a diagramatic picture of the history of the L-M cell samples selected for analysis. The number immediately preceding the strain designation of monolayer cultures is the passage number, which indicates the number of subcultures. The growth medium is indicated after the strain designation. This is modified medium 199 (4) plus 0.5% peptone (199P or 199PSP) supplemented with fetal calf serum (FCS). The concentration of serum is denoted by the subscript. The date the culture was initiated is included, as is the mode of culture (sus. = suspension, ferm. = fermenter, spin. = spinner, and mono. = monolayer). Two types of fermenters were employed, with the specific fermenter used for a particular culture indicated by a code preceding the strain designation. NB 7/24 stands for the seventh experiment run in the 5-liter New Brunswick fermenter, equipped for continuous pH and oxidation-reduction potential control, while FDL3 indicates the third experiment run in the 15-liter (L) Fermentation Design fermenter. The number preceding the strain designation of a spinner suspension culture (5) is a culture identification number, not a passage number. Labels on the arrows indicate the mode of intervening culture on the same growth medium (Figure 2C). The BHK21PD4 cell line was obtained from Dr. Robert Brackett, Parke-Davis and Company, Detroit, Michigan, the PD4 signifying Parke-Davis Clone 4. A history of this line as carried in this laboratory appears in Figure 3C. The nomenclature used is basically the same as that used to designate L-M cell cultures, with the following abbreviations applicable: MEM = Eagle's minimum essential medium; CS = calf serum; Sched 8 = Schedule 8; Spinner medium = Special Medium; and TP = Tryptose Phosphate Broth. RESULTS L-M CELLS A previous study on the chromosomal complement of L-M cells grown in monolayer on modified 199P showed this strain to have a modal range of 64-65 chromosomes per cell, with other chromosome numbers present at significantly lower frequencies (1). The modal number of biarmed chromosomes fell in the range of 12-14. Two marker chromosomes, the E (a submetacentric with two secondary constrictions) and the F (a subtelocentric with 3-4 secondary constrictions) were noted to occur frequently (in more than 50% of the cells), 114

and often in combination. The modal number of chromosomes in the L-M cell grown in modified 199P appeared quite stable in monolayer culture over a fouryear period. At the end of this period the percentage of modal cells had increased significantly (from 35% to 49%), while the marker chromosome frequency remained nearly the same, Transfer to chemically defined medium (2x concentration of Eaglets basal medium in Hanks BSS) as well as animal passage in C3H/An mice, were accompanied by shifts in the chromosomal pattern. It was of interest to learn if population selection could be detected, via chromosomal analysis, as a result of changing the mode of culture. The possibility that selection occurs when a cell strain is transferred from a monolayer system to a suspension system, and finally to a large scale fermenter culture, could have a critical effect on the ability of a large scale culture system to produce uniform results, important for such applications as virus-vaccine production, Baumann and Merchant (6) studied the chromosomal pattern of 128 L-M 199PSP FCS2 5 (Figure 4C) and NB 7/24 L-M 199PSP FCS5 (Figure 5C) to determine the effect, on the chromosomal pattern of the L-M cell, of transfer from monolayer culture to fermenter culture, via spinner culture. In addition, the effect of transferring L-MaTlC3 Clone 2 cells from monolayer to spinner suspension culture was also studied (the L-Ma cell line was derived from in vivo passage of L-M 199P cells (3)), Because of the relevance of this data to the chromosomal analysis of subsequent populations (Figure 2C) the results of this analysis are repeated in the context of the current report. The chromosomal pattern of 128 L-M l99PSP FCS2 5 underwent a significant change as a result of its period in spinner and fermenter culture. The number of chromosomes and number of chromosome arms showed an increase of one in modal number coupled with an approximate doubling in the frequency of modal cells (see Table IC). The telocentric chromosomes, which presented a modal range of 48-53 in 128 L-M, had altered to depict a sharp mode of 52 in the NB 7/24 L-M fermenter culture. The modal number of biarmed chromosomes remained unchanged at 13 during this selection pressure, though the frequency of modal cells of this category also roughly doubled, A shift of one chromosome in the modal number of chromosomes in a mixoploid cell line is probably not significant by itself, However, this shift, accompanied by the dramatic increase in the frequency of modal cells, indicates a definite selection in. the cell population, The L-MaTlC3 Clone 2 line showed small variations in modal number and the frequency of modal cells between monolayer culture (101 L-MaTiC3 Clone 2) and spinner suspension culture (2827 L-MaT1C3 Clone 2), Because of the magnitude of the variations observed in the L-Ma cell system, it appears that the selection pressure of transferring L-Ma cells from monolayer to spinner culture is small, when compared to the selection observed for the L-M cell transferred from monolayer to the fermenter system. This may be due to the continuous adjustment of the pH and oxidation-reduction potential in the fermenter system, or the fact that the LZMa cells were only exposed to suspension culture for approximately half the time that the L-M cells were in suspension. 115

ff~~~~~i2 1M AcPSe l jii0A /6 62 lCRQOM Vigure ha. Kryotype of a cell from 128 T,-M' grown on modified imedium 199 pius O~-~/ peptono and 2.5w fetal calf setrum in monolayer culture. 16

...................................................................................:................................................................................................................................................................................................................................................................................................................................................................................................................................. ~...............................~.............~........~.........,:,:,.............................:,,:,.......................................................................................,:,:,:,:,,:,:,:::,,::,,:...........................:,:..................~...........~.................~.........~......,:,........,:::,:,.....................:............................................................................................................................:::,:,::,:,.:,,,:,:,:,,,:::::,,:,:,:::,:,,,............~..........,::,::,,:,:,:,:,:,,:::,:,:,:::,:,:,::,:,,:,,::,:,,:,:,::::::::,:,:,.,:::::,:::::::,:,:::,,,::::,:::,:::::,::::::::,,:::,::,::,::,:,.............:::::,:::,,:,:::,:,:,:,:,::,:,,,:,,,:,:,::,:,:::,,:........,:,:::,::,:,::::,:::::,,::,::::::,:,,,:,::::,,:...........,:,:,:,:,:,:,.......,,:,:,:,:,::,:,:,::::,:,::::,.,:::,:,,:,::::,:::::,.:::,:::,.........~....................... i!~11~j+~:[i)i [~:! ~i;~i~i ~i!!~~ iiii:~:ij)i~~i: ~~11~~i ~d~~!!~ ii~~i ~:i~: ~ii~iiii ii~~: ~ili![:::i i:~::i~~;i~:? i:iC:~1!i i;~~i~:i~i::~i!!![i~i[iiiiiii~i~iiiiiiiiiiiii~!ii::11iiii~ili:ii:~:i~~ iii~~11~: i:iii:i!i~ii!j:iiiii~! i~[iiiii:: ~ii ii[!~ii:~::i:~ ii} )!ii~~ i; i:;~i! ii~~i~i ii~~~:~[i:i!~i;::i i~i:!~i:C!:~i!ii~:j:[i i~i:~:;ili:!ii?:~i::; ~;~ i~:ii ~ii~: ~:: ~::i i~!~i ~:i:!:i:iiij!i~ij:i!iiiii~i!i~i!iii~ii~!i~i~;:i~i!ii~i~i!!:i!ij!::!~iii:i~i[i~iiiii~iii~i!i~iii)i~ii...........................................................................................................................................,:,:.......... ~........ ~...... ~............,,,,,,,,,,,,,,,.,,.......:............,,,,,............................:,,,,:,,,,,,,,:,:,,,:,,,,,,,,,,:,,,,,,:,,,,:,,,:,:,,:,,,,,,,:,,,,,,,...........,:,,,,,:,,,.:,,:,,,:,,.,,,,:,:,,,,,,:,,,:.,:,:,,:,:,:,,:,,:,,:,,,,,,:,:,:,:.,:,,,,:.,,,.,,,,,:,:,:,:.:.:,:,:.,.:,:.,:.: ~i~!~,'~ i, ~,ii: ~,i:i~i,(i,!j,!,:i~': ~ ~:'::j' >!:': i i~, ~!,~i ~'!,i::!~i,':~ i'!:i~,?ii~,,!~: ili',i'ii:~i ~:i~:::i:~,i~i~i,~'i~i~ii'i ~ i~i~i:~il i:~iiji:i~i i!i i~i i~i'iifi:i ~ ili,!ii ii!:i:!},i!i:i!ili:i!~,i~ii!:: i:i ilii (i~i'!,!,~i~i ~ i:~j!,!:,',!,: ii!i,:!j ~'i i!: i~:, i!:,:i'~:!,,:,, ~"i' ~'/!: ~:'!,: ~:i: ~:,',,',~}~,,, ii,,!/,,:~ ~:! ~i', ~ i),: ~::: ~:':,,:,::~,~ ~,,~:~," ~:,~,~i~:A~!f~::!::~:~:~':'~(~i~:~b~,~:~:~'(~,~,~!~', ~:::~: ~'~,~:~:~....:,,,::,:,..............,:,,:.::,,............ ~:::,:,,,:,,:,::::,..........:,,:,:,:,:,,:,:............... ~........................... ~...................... ~.................... ~.................................,::,,::,:,:,,.........,,:,,:,,,,,::,,.... ~.........................,,:,,::,,,:,:,................. ~.................................................................................,,:,,,:,:::,,,:,:,,::,,...................................... ~i i'i,! ii,~:~i~:ii:?i (i,~,,!i': ili,,:,i~ i',~,:~ i': ~i~:~:~!i:i:~ii:~! iii~i, i'!,i'~[: ~i'i i~ i'i:i~:i:::!:~:~:i'~!, ~ iil!'i~i~i iii:ii~ i:(i~!~: ili!i~i:i!ii,!~ii iii'~ ~:ij~i i'i!~:i,~ ili~i':,i,: i!ii ii "iii~:/i:i: i!i ~:i,i~ )~!,~ ~,i(::~'i'~ ),,j:i,,',,'~,,:~i: i" ~!~,,~,~'i,~ii' ili'~,~::i!i ~,!i:,:ii!i~,(:i'!:,i:ili:!',, i~ i'ii:~i'f i,~:~!i:::!iiiii:!fiji ~i!ili iiiii~ii i:i~ ii:i:i:i~ii~,~:ii~i/i/! i~i:i~i~i~i:ili i!!ii'i~i~i:i:i:i:i~i:ii~ ~ ~ ~??C+~i~i~'~'~!~i~i~i~ ~!~!~ 7~?~ i~'~!~ i~~'~ ~!~'~'~!~?~ i~?~~ i~'~'~'~ ~ ~ ~ ~ ~ Y~ ~ ~!C'~ i~!~!~!~!~i~'~!~'~!~!~!~!~!~!~ i~!~!~!~!~!~ iC!~!~!~ ~'~i~!~!71~!~ Y?!~!~'~!~!~!C ~i~!~'~ ~ iC!C!!~!~!~i~!~!~!~!~!~!~!~!~! ~ i,':".....,,::': ~ i ~ ~ ~ (,:',,,11.................................. Fi.gure 5C Karyotype of a cell from NB':':'f/2t~ L-M grown on modified lnedium 199 plus O.5~ peptone and 5~ fetal.. ca.l,.f serum:t.n fermenter culture.

TABLE IC. Frequency distribution of total chromosome numbers and numbers of biarmed chromosomes in 128 L-M, NB 7/24 L-M, 2817 L-M, 2832 L-M, 2860 L-M, 2884 L-M, P3071 L-M, 170 SL-M Series 3, 170 SL-M Series 2, 101 L-MaT1C3 Clone 2, and 2827 L-Ma T1C3 Clone 2. The cells were grown in modified medium 199 supplemental with 0.5% peptone and 5% fetal calf serum, with the exception of 128 L-M, which has the fetal calf serum reduced to a concentration of 2.5%. Number of Chromosomes _____ Cell Line end Date hom s6 8 51 531 54 551 5 7 58 59 60 61 62 6314 65 656 67 /'8 /-'9 128 L-M 199PSP FCS2., 1 4 5|1012 6 2 1 5/8/68 50 cells NB 7/24 L-M 199PSP FCS5 1 1 4 2 14 26 1 1 6/13/68 50 cells 2817 L-M 199PSP FCS5 1 1 2 1 1 3 1 9 8 16 5 2 7/17/68 50 cells 2832 L-M 199PSP FCS5 1 1 7 12 18 10 1 8/11/68 50 cells 2860 L-M 199PSP FCS5 2 2 7 6 10 15 7 1 9/10/68 50 cells 2884 L-M 199PSP FCS5 3 41 111 16 5 10/10/68 50 cells P3071 L-M 199PSP FCS5 2 2 4 6 81319 35 25 24 6 3 3 6/18/69 150 cells 170 SL-M Series 3 199PSP FCS5 1 1 1 1 5 9 13 9 7 3 5/16/69 50 cells 170 SL-M Series 2 199PSP FCS5 1 1 1 3 3 8 21 8 3 1 5/16/69 50 cells 101 L-MaT1C3 C/2 199PSP FCS5 2 1 2 1 2 3 3 2 6 7 11 6 2 1 1 7/9/68 50 cells 2827 L-MaTIC3 C/2 199PSP FCS5 1 1 1 3 4 8 11 12 8 1 7/26/68 50 cells Number of Biarmed Chromosomes g 0 lo11 12 1 11 151 17 18 19 20121 2212324 125 2l - 128 L-M 199PSP FCS5 1 4 9 22 13 1 5/8/68 50 cells NB 7/24 L-M 199PSP FCSs 4 40 5 1 6/13/68 50 cells 2817 L-M 199PSP FCSs 1 1 2 14 18 13 7/17/68 50 cells 2832 L-M 199PSP FCS5 2 4 12 22 5 2 2 1 8/11/68 50 cells 2860 L-M 199PSP FCSs 1 5 4 24 14 1 1 9/10/68 50 cells 2884 L-M 199PSP FCS5 2 2 11 16 14 4 1 10/10/68 50 cells P3071 L-M 199PSP FCS5 1 1 2 11 57 29 14 7 8 5 4 8 2 1 6/18/69 150 cells 170 SL-M Series 3 199PSP FCSs 1 20 16 10 3 5/16/69 50 cells 170 SL-M Series 2 199PSP FCS5 1 4125 8 9 3 5/16/69 50 cells 101 L-MaT1C3 C/2 199PSP FCS5 2 6|20 18 3 1 7/9/68 50 cells 2827 L-MaTsCs C /2 199PSP FCS5 31 7|15 20 3 1 7/26/68 50 cells

From NB 7/24 L-M, a continuous series of spinner suspension cultures was initiated, of which 2817 L-M, 2832 L-M, 2860 L-M, and 2884 L-M are representative samples taken over a three-month period. A striking observation evident on an examination of the chromosomal analysis of these samples is the constant repetition of 13 (the mode in 128 L-M and NB 7/24 L-M) as the modal number of biarmed chromosomes (Table IC). It appears that this characteristic is remarkably stable with respect to the selection pressure of suspension culture. It is also noteworthy that the high frequency of modal cells (biarmed chromosomes) observed in NB 7/24 L-M was not maintained in the spinner culture series. The modal number of chromosomes for each sample of the spinner culture series flucturated from 64-65, a range previously reported for L-M cells grown in monolayer culture on serum free medium (1). Based on changes in the modal numbers of chromosomes from this series and the cited reference, it appears that there is little selection pressure in continuous spinner suspension culture, but strong selection pressure in fermenter culture, However, an examination of the marker frequency and distribution (Table IIIC) of the E and F markers and their respective derivatives, appear to indicate that there was actually a distinct selection as a result of continuous spinner suspension culture, The great difference between the frequency of the F + F' category found in 128 L-M and NB 7/24 L-M compared to the frequency observed in spinner culture, seems to indicate selection on a finer level than chromosome number, With the decrease in frequency observed between 2860 L-M and 2884 L-M (.78 to.58), some doubt is introduced as to the long-term stability of this marker frequency change, A second major series of cultures (see Figure 2C) was obtained through the subculture of NB 7/24 L-M in several spinner cultures, passage in a 15liter fermenter culture (FDL3 L-M), followed by further spinner subculture, and monolayer culture (147 SL-M 199PSP FCS5 10/28/68)o The monolayer cultures were given the S prefix at this point to indicate that the cells had been previously cultured in suspension. 164 SL-M was used to continue, unbroken, the current monolayer lineage (SL-M Series 2), as well as to initiate a new suspension culture series, from which monolayers were subsequently planted (SL-M Series 3). Samples were selected from this series to investigate: (1) the reversibility of changes resulting from selection in suspension culture (170 SL-M Series 2 vs. 128 L-M); (2) the selective effect of periodic suspension culture separated by periods of monolayer growth (170 SL-M Series 2 vs. 170 SL-M Series 3); and (3) the effect of continuous suspension culture of cells previously grown in suspension and returned to monolayer culture (P3071 L-M vs. 2817 L-M through 2884 L-M), "170 SL-M Series 2 vs. 128 I-M": In 170 SL-M Series 2 the modal number of chromosomes showed a decrease of one from the value observed for 128 L-M (64). This was accompanied by a definite rise in the fraction of cells having the modal number of chromosomes, The biarmed chromosome mode remained at 13, with a high percentage of cells in both populations possessing this mode (44% and 50%)4 The modal number of chromosomes shown by 170 SL-M Series 2 was two less than the strong mode of 65 observed for NB 7/24 L-M. Apparently, there 119

was a strong selection away from the chromosomal pattern displayed by NB 7/24 L-M in either subsequent suspension culture or monolayer culture. The frequency of the F + F' category was observed to be.48, a value noticeably higher than that detected in 128 L-M, but quite a bit lower than the frequency observed for cells carried continuously in suspension culture (2817 L-M through 2884 L-M). The overall chromosomal pattern of 170 SL-M Series 2 is different from both 128 L-M and NB 7/24 L-M, but whether this series is gradually returning to the pattern of the parent monolayer or is establishing a new chromosome pattern, is not evident. "170 SL-M Series 2 vso 170 SL-M Series 3": SL-M Series 3 had a modal number of chromosomes of 62 and a low frequency of modal cells compared to Series 2, Both populations had a high frequency of the biarmed chromosomes mode of 13. Interestingly, the marker frequency (Table IIC) of the F + F' category for 170 SL-M Series 3,.78, is quite consistent with the frequencies noted for 2817 L-M, 2832 L-M, and 2860 L-M, suggesting, again, that the high frequency of this category results from selection in the transfer of L-M cells from monolayer to spinner suspension culture. A further argument in support of this contention is that the comparatively low frequency of this category in Series 2, which was not selected a second time in spinner suspension culture, is due to the fact that the absence of selection pressure allowed the frequency of the F + F' category to drift toward a pattern characteristic of the parent monolayer. The frequency of the E (E') category observed in the parent monolayer (128 L-M), while the slightly higher value noted for this category in Series 3 is closer to the range seen in suspension culture (see Table IIC). We have an indication, on the basis of marker frequency, that suspension culture results in a population selection which may be reversible. "P3071 L-M vs. 2817 L-M through 2884 L-M": P3071 L-M, which resulted from the spinner suspension culture used to initiate 170 SL-M Series 3, had a modal number of chromosomes of 61, with a significant number of cells having 62 and 63 chromosomes (see Table IC). The spinner suspension culture series 2817 L-M through 2884 L-M, characterized by modal numbers of chromosomes from 64 to 65, appears to have a significantly different population structure based on chromosome number, although all these samples had a modal number of 13 biarmed chromosomes (Figure 6C). This downward shift in the modal number of chromosomes observed for P3071 L-M may be a result of two factors: (1) the total time P3071 L-M was exposed to the selection pressure of suspension culture was roughly 7-1/2 months compared to 4 months for 2884 L-M; and (2) an intervening period of monolayer culture in the history of P3071 L-M of approximately 5 months. In any case, it is obvious that variations in culture method, such as switching between monolayer culture and suspension systems causes significant flucturations in the chromosomal pattern. BHEK21 PD4 CELLS The chromosomal pattern of BHIK21PD4 cells closely resemble the normal 120

TABLE IIC. Frequency and distribution of the marker chromosomes in 128 L-M, NB 7/24 L-M, 2832 L-M, 2860 L-M, 2884 L-M, 1% SL-M Series 2, 1% SL-M Series 3, P3071 L-M. Marker Frequency and Distribution (?,) Expected Observed E or E' E or E' Cell Line and Date None E or E' F or F' E or E' E or + + F or F' F or F' 128 L-M 199PSP FCS2.5 22 78 20 15.6 20 5/8/68 50 cells NB 7/24 L-M 199PSP FCS5 10 90 20 18.0 20 6/13/68 50 cells 2817 L-M 199PSP FCS5 2 80 76 60.8 58 7/17/68 50 cells 2832 L-M 199PSP FCS5 2 94 74 69.6 70 8/11/68 50 cells 2860 L-M 199PSP FCS5 4 84 78 65.5 66 9/10/68 50 cells 2884 L-M 199PSP FCS5 2 92 58 53.3 52 10/10/68 50 cells P3071 L-M 199PSP FCS5 23 82 88.7 73.5 72.7 6/18/69 150 cells 170 SL-M Series 3 199PSP FCS5 2 94 78 73.3 74 5/16/69 50 cells 170 SL-M Series 2 199PSP FCS5 14 78 48 37.4 40 5/16/69 50 cells 121

::~ ~ ~ ~ ~ 28 PA 1991 SP —- [C 10110/ 8: CIIR():- $- OPALS-i- ~:~: ~ ~:: ~:::;: ~:.~::~: ~ _::::i~~:~:::~::::: ~: _~: ~::: ~::::~; ~::::~i~~~~~i~:::~;~::~::::~::::~? ~:~:~::~: ~;~: ~:::::: ~:: _ ~: ~ ~ ~ ~ ~: ~::....:: ~: ~:: ~:~ ~: ~::: ~:::: ~::~::l I-::::::;: -:::::::::::::;::-::-:::::::1::::;:: 1::;:I:::::- -::?:::::::;1:1::::: 1-::.:-:;::::-::;:-:::::-;F11gure 6C. Karyotype of a cel. fron 288LSM P4 grown on modifled mediur:l plus 0.5% peptone and 2.5% fet al calf serum i.n spinner culture. 1.22

pattern presented by the Syrian or Golden Hamster (7), which has four pairs of telocentric chromosomes, a pair of very small subtelocentrics (mT), a pair of very small metacentrics (sM), and recognizable X and Y chromosomes, with the remaining 15 pairs all biarmed chromosomes of intermediate lengths and ratios, BHK21 cells (Clone 13) have been reported to be uniformly male (8). The fact that there are three pairs of chromosomes and one group of four pairs, which may be readily detected and monitored to follow genotypic changes during in vitro culture, makes this near diploid continuous cell line particularly well suited to studies where it is desired to follow population selection based on karyotypic changes, BHK21PD4 suspension no, 49 had a high frequency of cells with 43 and 44 chromosomes (see Table IIIC), with the diploid number predominating. The modal number of biarmed chromosomes was 36, which is equal to the number of biarmed chromosomes in the normal Syrian hamster (scoring the mT chromosomes as biarmed), NB 8 BHK21PD4 showed a bimodal pattern of chromosome number, with males of 42 and 44. Figure 7C shows a pseudo diploid cell with 3 mT chromosomes, the normal number of telocentrics, and a normal pair of XY chromosomes, but the cell is missing one biarmed chromosome. The modal number of biarmed chromosomes was 34, as a result of grouping the mT chromosomes with the telocentrics, which was done in this ans subsequent analysis of BHK21PD4 cells. Apparently the transfer of BHK21PD4 cells to fermenter culture results in population selection somewhat different from that observed for the L-M cell (i.e., increased uniformity for the L-M cell system versus increased variability for the BHK21PD4 cell system). 72 BHK21PD4 also showed a bimodal pattern of chromosome number with modes of 44 and 45 (76% of the cells fell in this range), and a biarmed mode of 33. The variability in the modal number of chromosomes was due to variations in the number of telocentric chromosomes. Comparing the karyotype of a cell with 44 chromosomes (Figure 8C), to a cell with 45 chromosomes (Figure 9C), one notes several interesting differences, Figure 8C shows a cell with only slight deviations from the normal Syrian hamster karyotype, i.e., a missing biarmed chromosome and an extra mT chromosome, In Figure 9C we see a cell with the normal (hamster) number of biarmed chromosomes (excluding mT), but with an extra telocentric as well as an extra mT chromosome. In addition, the size variations in the two chromosomes labelled sM may not be entirely due to differential coiling, but may represent a structural rearrangement. Another cell with 45 chromosomes, showed an extra pair of telocentrics and a missing pair of biarmed chromosomes in addition to the typical extra mT chromosome (Table IVC). It is apparent that BHK21PD4 cells vary more than is shown by chromosomes' number alone, 123

TABLE IIIC. Frequency distribution of total chromosome numbers and numbers of biarmed chromosomes in BHK21PD4 suspension no. 49, NB 8 BHK21PD4 Special Medium, 72 BHK21PD4 MEM TP10 CS5, 80 BHK21PD4 MEM TPlo CS1, BIHK21PD4 sus. spin. MEM TPo1 CS1, BHK21PD4 s pin. Special Medium (6). Cell Line Number of Chromosomes Number of Biarmed Chromosomes and Date 353638 39 40414243 4445 5657 58 222425 2728 2930 31 32 33 34 3536 37 BHK21PD4 1 31422 8 1 2 51127 4 sus. spin. #49 7/13/68 NB8 BHK21PD4 1 1| 1 612 614 5 1 3 1 1 2 1 5117 9 Special Medium Ferm. 7/25/68 72 BHK21PD4 1 7 18 20 4 1 2 8 28110 1 MEM TPo1 CS5 6/17/69 80 BHK21PD4 3 5 9 1911 3 4 7 13517 9 MEM TP1o CS1 7/23/69 BHK21 PD4 1 1 41011 11 9 0 3 1 1 1 1 1 4 8 9 14 2 3 MEM TP10o CS1 (sus. spin.) 7/23/69 BHK21PD4 1 1 21 1 2 1110 12 8 2 0 2 1 61521 3 1 1 1 Special Medium (sus. spin.) 7/23/69

Figure 7C. Karyotype of a eell from NB 8 BW% E1}D4 grown on Schedule 8 spinner nedium in fermenter cu.tue..,t Cell i psetudodiploid

61,A In C1 rx*S................................ ~ ~ ~z:::................:ia...:... fzliziii!:zz00t0ii0 zliii0izlzii0i0i..............................................0 Fiture 8C Karyotype of a cell from 72 BHEIK2PD4 growan on Lagie IME3M (supplemented with IO% Tryptose Phosphate Breoth) pwus 5%. 2lf serum in monolayer culture. Cell is pseudosdipaloid. 126

Figutre 9C'* Karyotyje of o cell fro tlr 7 BlK 1.$4 grown on bEa4o Mt1Ph/ (supploe mont d witth IsOj Tryptos o ItosJphato IBoth) plit 57 calf c-nna in monol yor oultu~r* Coil nS nax dipo.+id. 127??iiii...... ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~i.......... mented1111' wih10iiTrpoe hshaeBot)iiicl sriinTinoae culture. Cell is near diploid,~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~iiiiiiiii!

TABLE IVC. Frequency and distribution of mT and Q chromosomes in 72 BHK21PD4 (mono.), 80 BHK21PD4 (mono.), BHK21FD4 (spinner) MEM TP10 CS1, BHK21PD4 (spinner) Special Medium. Marker Frequency (%) Cell Line and Date 5 or More mT Q 72 BHK21PD4 MEM TP1o CSs 86 6/17/69 50 cells 80 BHK21PD4 MEM TP1o CS1 100 40 7/23/69 50 cells BHK21PD4 (sus. spin.) MEM TPo0 CS1 76 14 7/23/69 50 cells BHK21PD4 (sus. spin.) Special Medium 70 0 7/23/69 50 cells Methocel added. 80 BHK21PD4, grown on a medium with a reduced concentration of calf serum (1%), had a pronounced modal number of chromosomes of 44, with a modal number of biarmed chromosomes of 33 (mT scored as a telocentric). Three or more mT chromosomes were present in all of the cells examined, indicating that all the modal cells were pseudo-diploid. In addition, a new marker chromosome was observed, which had two secondary constrictions and a light staining (sometimes "stretched") region in the short arm adjacent to the centromere. This new marker, designated the Q marker (see Figure 10C), is quite likely a derivative of the Y chromosome, and was found in 40% of the cells examined, The appearance of the Q marker at this point in the culture history may be a consequence of the reduced calf serum concentration. The 1% level of calf serum did not produce any dramatic effect on the modal number of chromosomes, though it is noteworthy that, in comparison with 72 BHK21PD4, the fraction of the population having 45 chromosomes was reduced (see Table IIIC). Two spinner suspension cultures were studied at the same time 80 BHK21BD4 was analyzed. One culture was grown on Eagle's MEM plus 10% TP and 1% CS (hereafter referred to as Spinner 2). Both cultures showed a biarmed chromosome mode of 335 The surprising observation was the wide range in modal number of chromosomes shown by both cultures (42-45). Of further interest was the frequency of the Q marker, which was 0 for Spinner 2 (high concentration of calf serum) and 14% for Spinner 1 (low concentration of calf serum), an additional indication that this marker may be selected by low calf serum concentration. From a comparison with BHK21PD4 suspension no. 49 7/13/68, it is evident that either or both of the following have occurred: (1) the stock BHK21D14 cell line has changed over a one-year period to the extent that 128

F'igaue l.OC. Karyotype of a cellI from 80O BiiK21It)4 grown on fEagJ:e MEM (supp..... mented with 10O Tryp)tose Phosphate Broth) plus 1% cnalf7 sorum In monolayer. Cell,. is neare diploid.

it responds differently to the selection pressure of spinner culture; (2) the spinner culture system is significantly different, DISCUSSION From the data presented on the L-M cell, it appears that fermenter culture exerts a strong selection pressure on the cell population. Several additional points which are evident are: (1) switching between various modes of culture produces changes in the genotypic constitution of the cell population; (2) the chromosome pattern observed in suspension culture is significantly effected by the prior modes of culture; (3) some changes resulting from population selection may be reversible over a short period of time (i.e., marker frequency and distribution), while other changes (modal number of chromosomes), may not be quickly reversible; (4) the L-M cell can be maintained in spinner suspension culture without major changes in the chromosomes pattern over a three-month period; and (5) that marker frequency appears to be a useful characteristic in evaluating the selective effects of various modes of culture, particularly spinner suspension. It appears that a biarmed chromosome mode of 13 is an unusually stable characteristic of the L-M cell grown on l99PSP and fetal calf serum, as this value did not change under a variety of selection pressures. The consistency with which modal values fluctuated closely above the previous finding reported for the L-M cell on 199P (1) indicates that the line is probably of greater stability now than when dealt with in 1959 (9). It is certainly indicated that investigators desiring to produce large-scale cultures of consistent properties with respect to virus susceptibility, growth requirements, etc., must consider the selective effects of the sequence of different types of culture on the gernotypes and phenotypes present in any given population. Data on the BHK21PD4 cell line has shown that, in comparison to the L-M cell, selection in large scale fermenter culture may not take the same direction in different cell lines (i.e., L-M cell showed increased population uniformly, while the BHK21PD4 population showed increased variability). Karyotypes of several BHK21PD4 cells indicated that the variability of this cell line is somewhat greater than is indicated by a diploid or near diploid modal number of chromosomes. The appearance of the Q marker in monolayer culture and the different frequencies observed, when this marker was scored in spinner cultures of BHK21PD4 cells in two different medias, may prove to be a useful system for evaluating population selection in this near diploid line, 130

REFERENCES 1. Giles, R. E., D. J. Merchant, and E. Massilink. Chromosomes of L-M Mouse Cells and Variants. J. Nat. Can. Inst. 37: 663-676, 1966, 2, Hsu, T. C. and D. J. Merchant. Mammalian Chromosomes In Vitro. XIV. Genotypic replacement in cell populations, J. Nat. Can. Inst. 26: 1075-1083, 1961, 35 Keefe, S. A., D, J. Merchant, and W. H. Kelsey. Alkaline phosphatase activity of L-M mouse cells and variants, Proc. Soc. Exp, Biol, Med. 118: 1031-1037, 1965. 4. Eidam, C. R. and D. J. Merchant. The plateau phase of growth of the L-M strain mouse cell in a protein free medium, I. Patterns of protein and nucleic acid synthesis and turnover, Exp Cell Res 37: 132-139, 1965. 5. Merchant, D, J., R. H. Kahn, and W. H. Murphy, Jr. Handbook of cell and organ culture, 2nd ed,, Burgess Publishing Co., Minneapolis, 1964, 6, Merchant, D, Jo Large scale fermenter growth of animal cells for virus vaccine production: Control of seed stocks and growth conditions (4). Progress Report 08621-2-P, U. S. Army Medical R and D Command, Office of the Surgeon General, D/A, Washington, D. C. 20315, 7. Hsu, T. C., ed, An atlas of mammalian chromosomes, 8, American Type Culture Collection. Registry of animal cell lines certified by the cell culture committee, 3rd supplement. American Type Culture Collection, Cell Repository, 12301 Parklawn Drive, Rockville, Md. 20852, 1968, 9. Hsu, T. C. Mammalian chromosomes in vitro XI. Variability among the progeny of a single cell. Biological Contributions, The University of Texas (Austin), Pub. Aro 5914, Fall 1959, 131

GROWTH OF STRAIN L-M MOUSE CELLS IN CHEMICALLY DEFINED MEDIUM Gary Stoner 133

INTRODUCTION The minimum essential amino acid requirements for maintenance of strain L mouse fibroblast cells were reported by Eagle (1). Thirteen amino acids; viz., arginine, lysine, tryptophan, histidine, leucine, isoleucine, valine, cystine, methionine, glutamine, phenylalanine, tyrosine, and threonine were found to be essential for cell maintenance. Since Eagle's original report, strain L mouse fibroblast cells have been adapted to grow in a variety of chemically defined media, and analyses of amino acid utilization by the cells in these media have been studied. However, most of these media are very complex and contain nonessential amino acids. Merchant and Hellman reported growth of strain L-M mouse fibroblast cells in Eagle's basal media at (2X) concentration containing only the thirteen essential amino acids (2). This medium is the simplest medium ever described for continuous propagation of mammalian cells. This report will contain data obtained on the kinetics of amino acid utilization and production by strain L-M mouse fibroblast cells in "(2X) Eagles" medium. MATERIALS AND METHODS Medium was prepared so that the only variables to be studied were the amino acids themselves. The medium was incubated for 48 hr at 37~C. Glutamine was stored separately at -20oC and added to the medium after incubation to prevent spontaneous decomposition of glutamine at 370C. Medium was not changed throughout the growth curve. Monolayer cultures in 30 cm2 Falcon flasks were inoculated at an original concentration of 3.0 x 105 cells/ml. Three replicate cultures were harvested at 24-hr intervals; medium was stored at -20~C for subsequent amino acid analysis, and after washing with Hanks BSS, cells were stored at -70~C for subsequent protein analysis by the Eagle-Oyama modification of the Lowry method. A final cell count was taken at the end of the growth curve. Amino acid analyses were performed with a Beckman Model 120C amino acid analyzer. Both short and long columns were internally standardized so that any malfunction of the machine could be detected. The short column was standardized by the addition of a standard amount of a-amino-p-guanidinoproprionic acid and the long column with a standard amount of L-norleucine. Chromatogram peak heights and widths were determined; the values were converted by a computer to the concentration of amino acids in micromoles/ml medium. The computer program accounted for the dilution of the medium by addition of the standards. It was necessary to run two long column analyses because glutamine occluded the threonine-serine peaks. Glutamine could be removed completely by hydrolysis with an equal volume of 2NHCL for 1-1/2 hr at 100~C. 134

RESULTS A. GROWTH OF STRAIN L-M CELLS IN "(2X) EAGLE? BASAL MEDIUM (CONTAINING ONLY THE THIRTEEN ESSENTIAL AMINO ACIDS) Table ID indicates the kinetics of utilization of the thirteen essential amino acids by the cells, Throughout the growth curve, it is apparent that the extent and rate of depletion of the individual amino acids varies, For instance, arginine and cystine are utilized very rapidly, whereas lysine is not utilized until the later stages of growth (perhaps along with proline for the synthesis of collagen). Isoleucine, leucine, and methionine are nearly depleted throughout growth and glutamine is entirely depleted. Figure 1D indicates that at 72 hr, ammonia production nearly parallels glutamine utilization indicating that the cells are removing the amide nitrogen of glutamine resulting in the production of ammonia, This suggests, as indicated by Eagle (3), that the amount of glutamine utilized as such for cellular biosynthetic processes is relatively small; glutamine is either converted to glutamic acid for subsequent utilization by the cells or is converted to pyrollidone carboxylic acid, Figures 2D and 3D indicate that a number of nonessential amino acids are produced into the medium during growth, Alanine, proline, and to some extent, serine, appear to be reutilized by the cells, Glycine production occurs most significantly at the later stages of growth indicating that it may be released by plateau and decline phase cells, There is no evidence for the release of asparagine. Figure 5D indicates the production of protein by strain L-M mouse fibroblast cells in "(2X) Eagle" medium, In this system, 130 ig protein is equivalent to 1,2 x 106 cells/ml medium, It has repeatedly been observed that protein synthesis ceases within 24 hr after glutamine is depleted, suggesting that glutamine is growth-limiting. If glutamic acid is added to the medium at a concentration of 4 mM and glutamine at 0.4 mM, growth ceases within 24 hr after glutamine is depleted, suggesting that glutamic acid will not substitute for glutamine. Also, glutamic acid does not contribute to the production of ammonia, A good correlation can be obtained when computing the conversion of essential amino acid nitrogen into protein nitrogen, nonessential amino acid nitrogen and ammonia, B. GROWTH OF STRAIN L-M MOUSE CELLS IN "(2X) EAGLES" BASAL MEDIUM SUPPLEMENTED WITH NONESSENTIAL AMINO ACIDS Nonessential amino acids were added to the initial medium at a concentration approximately equal to that at which they were produced at the maximum level of protein synthesis in (A) above, It can be observed in Table IID that the cells produced much smaller quantities of nonessential amino acids 135

TABLE ID. Micromoles of Essential Amino Acids Available/mi of Medium at Various Times Throughout Growth of Strain L-M Cells in "2X Eagle" Basal Medium Time, hr Amino Acid 0 24 48 2 96 120 144 168 Pm/ml pm/ml pm/ml Mm/ml,pm/ml p/ml m/ml pm/ml Lysine 0.310 0.310 0.310 0.300 0.300 0.290 0.260 0.200 Histidine 0.080 0.080 0.070 0.060 0.040 0.040 0.040 0.030 Arginine 0.210 0.170 0.150 0.110 0.090 0.070 0.070 0.060 Tryptophan 0.030 0.020 0.020 0.010 Trace Trace Trace Trace Threonine 0.260 0.2412 0.240 0.235 0.230 0.230 0.210 0.210 Valine 0.350 0.320 0.280 0.240 0.200 0.140 0.100 0.030 Cystine 0.190 0.130 0.090 0.070 0.060 0.055 0.050 0.050 Methionine 0.070 0.060 0.060 0.050 0.030 0.020 0.010 Trace Isoleucine 0.350 0.260 0.220 0.190 0.160 0.090 0.050 0.010 Leucine 0.340 0.250 0.210 0.170 0.130 O.L70 0.030 0.010 Tyrosine 0.180 0.140 0.130 0.120 0.110 0.100 0.090 0.080 Qalanine 0.170 0.150 0.150 0.130 0.120.0110 0.100 0.080 Glutamine 4.020 3.150 2.210 1.300 0.310 Trace 0 0 136

4. 00 3U 3. 50 0< 3.00 v 2.50 Ammonia 0 > 2.00 _ i1. 50 tD 1.00 Glutamine 50.50 0 48 96 144 1 2 TIME, HR Figure ID. Utilization of glutamine and production of ammonia by L-M cells in monolayer cultures in 2X Eagle basal medium. 137

.450 Glutamic Acid.350 LU.300 _J LiU o.250' - -Alanine 0_ C-) 1H 150 -- Proline 100 Aspartic Acid 050.025.0 0 24 48 72 96 120 144 168 192 TIME, HR Figure 2D. Production and utilization of nonessential amino acids by L-M cells in monolayer culture in 2X Eagle basal medium. 138

0.900 LUi I ) 0. 750 LI) 0 O/ a 0.600 v-,) 0.450 LU -J Glycine 0.300 0. 150 0 48 96 144 192 TIME, HR Figure 3D. Production of the nonessential amino acids glycine and serine by monolayer cultures of L-M cells in 2X Eagle basal medium. 139

TABLE IID. Comparison of the Production and Utilization of Nonessential Amino Acids by Strain L-M Cells in (2X) Eagle Basal Medium With and Without added Nonessential Amino Acids. Values Represent Maximum Amounts Produced and/or Utilized. p Moles p Moles p Moles Amino Acid Amino Acid Amino Acid Medium Amino Acid in Initial Produced/ml Utilized/ml Medium Medium Medium (2X) Eagle Glutamic acid 0 0.450 0.000 Aspartic acid 0 0.037 0.000 Alanine 0 0.270 0,240 Serine 0 0.039 0.015 Proline 0 0.160 0.150 Glycine 0 0.900 0.000 (2X) Eagle Glutamic acid 0.450 0.130 0,050 plus added Aspartic acid 0.050 0.005 0.025 nonessential Alanine 0.350 0.075 0. 325 amino acids *Serine 0.150 Proline 0.170 0 OOQ 0.070 Glycine 0.450 0 450 0.000 *Serine production and utilization in medium containing nonessential amino acids has not been determined. into the medium when they were added exogenously. In fact, with the exception of alanine and glycine, their concentration in the medium showed little variation throughout growth, suggesting that the amounts added were sufficient to maintain the necessary extra-intracellular gradient, so that excesses were not produced into the medium. Undoubtedly the cells are synthesizing nonessentials throughout growth, but are either pooling them intracellularly or synthesizing only those amounts required for cellular biosynthetic processes. Figure 4D indicates the production of protein by strain L-M cells in "(2X) Eagel" medium containing added nonessential amino acids. It is evident that the addition of nonessential amino acids did not result in a more rapid growth rate or larger growth yield. In fact, both parameters were reduced. However, this experiment will be repeated. C. GROWTH OF STRAIN L-M MOUSE CELLS IN "(2X) EAGLE" BASAL MEDIUM CONTAINING (4x) CYSTINE Cystine was originally found to be growth-promoting for strain L mouse cells by Healy, Fisher, and Parker. Morgan and Morton reported a requirement for L140

160 140 Lo |(2XE)+2XBME o I / \Amino Acids 80-, 1 a~ Added Nonessential Amino Acids 460 - 0H 20 48 96 144 192 TIME, HR Figure 4D. Protein production by L-M cells in monolayer cultures of amino acids.

cystine with heart muscle cells (5). As shown in Figure 5D, it is evident that addition of cystine at twice its normal concentration in "(2X) Eagles" medium results in a significant increase in growth rate although the cell yield is similar to that in "(2X) Eagle" medium. This indicates that the cells possess a larger requirement for (-SH-) groups than is provided by "(2X) Eagle" medium alone. Alternatively, cystine is converted to the reduced form, cysteine, which perhaps creates a more favorable redox potential for growth. It is felt that the cells possess a larger requirement for (-SH-) groups because in "(2X) Eagle" medium, the cells utilize cystine preferentially over the other (-SH-) donormethionine. Yet, when cystine reaches a certain minimum level methionine utilization is markedly increased indicating that the latter now serves as the principal (-SH-) donor. It has not yet been calculated whether doubling the cystine concentration exerts a "sparing effect" on methionine utilization. It has been postulated by a number of investigators that sulfhydryl-containing compounds may provide cells with needed (-SH-) groups for the synthesis of a protein involved in the regulation of cellular division. This postulation deserves more thorough investigation. D. SPONTANEOUS DECOMPOSITION OF GLUTAMINE IN UNINOCULATED MEDIUM INCUBATED AT 370C FOR 250 HOURS Many conflicting reports have been presented involving spontaneous decomposition of gluamine in medium incubated at various temperatures (6). The concentration of phosphate ions and the presence of serum appears to influence the spontaneous decomposition of glutamine to ammonia and pyrollidone carboxylic acid (7). It became necessary to determine the extent of spontaneous decomposition of glutamine in "(2X) Eagle" medium at 37~C. As can be seen in Figure 6D, throughout a time equal to a normal growth curve of strain L-M mouse fibroblast cells in "(2X) Eagle" medium, approximately 2C0 of the glutamine in the medium could be expected to be degraded spontaneously. 142

160 140 a- 120 CD _ 10 2XE+4X Cystine LU E 80 C / /2XE+2XBME Amino Acid LUI 60 LIU o 40 20 0 48 96 144 192 TIME, HR Figure 5D. Protein production by L-M cells in monolayer cultures in 2X Eagle basal medium with and without 4X cystine. 143

o 100 90 O 80 0 ) 70 uJ E 60 - 50 J 40. 30 20 i0 0 2 4 6 8 10 1210 0 2 4 6 8 10 12 DAYS Figure 6D. Glutamine decomposition at various time intervals in uninoculated. 2X Eagle basal medium incubated at 370C. 144

REFERENCES 1. Eagle, H., 1955. The specific amino acid requirement of a mammalian cell (Strain L) in tissue culture. J. Biol. Chem. 214:839. 2. Merchant, D. J., and K. B. Hellman. Growth of L-M strain mouse cells in a chemically defined medium. Proc. Soc. Exp. Bio. and Med. 110:194-198. 3. Eagle, H., V. I. Oyama, M. Levy, C. L. Horton, and R. Fleischman, 1956. The growth response of mammalian cells in tissue culture to L-glutamine and L-glutamic acid. J. Biol. Chem. 218:607-616. 4. Healy, G. M. D. C. Fisher, and R. C. Parker, 1954. Nutrition of animal cells in tissue culture. IX. Synthetic medium No. 703. Canad. J. Biochem. and Physiol. 32:327-337. 5. Morgan J. F. and H. J. Morton, 1955. Studies on the sulfur metabolism of tissues cultivated in vitro. I. A critical requirement for L-cystine. J. Biol. Chem. 215:539-546. 6. Tritsch, G. L. and G. E. Moore, 1962. Spontaneous decomposition of glutamine in cell culture media. Exp. Cell Res. 28:360-364. 7. Bray, H. G., S. P. James, I. M. Raffin, and W. V. Thorpe, 1949. The enzymatic hydrolysis of glutamine and its spontaneous decomposition in buffer solutions. J. Biochem. 44:625. 145

STUDIES OF BHK2, PD4 CELLS Howard Stockdale 147

GROWTH OF BHK21PD4 CELLS IN SCHEDULE 8 AND ITS MODIFICATIONS The standard medium for the growth of spinner cultures of BHK21 cells during this work was that designated "schedule 8." The formula for this is given in Table IE. A caveat is entered here that since moving to Vermont from Michigan BHK cells show increased clumping when grown in this medium. Work is currently underway to reduce the calcium level of this medium to the point where clumping is reduced without loss of cell yield. TABLE IE SCHEDULE 8 MEDIUM FOR SPINNER CULTURE Calf serum 100 ml Tryptose Phosphate 3.0 g Glucose Anhydrous 53.0 g Tris 3.63 g HEPES 4.76 g * 10 x Hanks Salts 100 ml 50 x MEM Amino Acids 30 ml 100 x Nonessential Amino Acids 30 ml 100 x BME Vitamins 30 ml 5% NaHC03 Solution 2 ml * Calcium Chloride Solution 10 ml Make to 1 liter final volume with water, and with pH adjusted to 7.5 with IN NaQH ( 5 ml). *"'10 x HANKS SALTS" NaCl 56.0 g per liter KC1 4.0 g per liter MgS04 7H20 2.0 g per liter Na2HP04 12H20 2.48 g per liter KH2PO4 1.20 g per liter * Calcium Chloride Solution 18 52 g CaC12 2H20 per liter 148

In schedule 8 medium the limiting factor is as yet unidentified. It was considered that a vitamin, a trace element, or some undefined component might limit growth. It was observed, for instance, that during growth, acidic, ninhydrin-reacting components eluting from an ion-exchange column before aspartic acid, and contributed by tryptose, showed a great deal of depletion. Solutions containing the following compounds were made. A. Cholesterol, calciferol, c-tocopherol and menadione. B. Choline, pantothenate, inositol, N-acetyl glucosamine, glutathione (red), RAD, folate, ribose, deoxyribose, ascorbate, and cyanocobalamin. C. Pyruvate, citrate, a-ketoglutarate, succinate, fumarate, malate, and oxaloacetate. Schedule 8 medium + A, B, and C; schedule 8 +50% excess (1.5 g/l) tryptose phosphate; and schedule 8 + A, B. and C + excess TP were used for the culture of BHK21,PD4 cells. No increase in growth rate or cell yield over that seen in schedule 8 medium was observed. GROWTH OF BHK21PD4 CELLS IN LOW-SERUM MEDIA It was felt that it might be advantageous to possess a type of BHK21 cell able to grow well in suspension in a low-serum medium. Three variations of schedule 8 medium, all containing 1% calf serum were made, and spinner cultures of BHK21PD4 cells set up after prior adaptation in monolayer culture. Because of such disadvantages as clumping, low cell yield, and slow growth, all three systems were abandoned. However, use of MEM medium with 1% calf serum produced cells with the following properties. After two passages in monolayer, lack of cell adhesion was seen, with cells floating free or in loose clumps. Spinner culture of BHK21PD4 in this medium yielded 2.0 x 106 cells/ml with a doubling time of 24-25 hr. Inocula may be as low as 2.5 x 105 cells/ml. No rimming or clumping was seen in spinner culture. Cells appeared to be highly glycolytic, and death was rapid after plateau phase. This latter medium was designated "MEM 1% C5" and its formula is given in Table IIE. Table IIIE gives a rough estimate of the comparative costs of producing a given number of BHK2,1D4 cells with schedule 8 and MEM 1% C5 media. 149

TABLE IIE MEM 1% CALF SERUM MEDIUM MEM powder-GIBCO Cat. No. F16 (+ NE amino acids) 10.8 g Tryptose phosphate 3.0 g NaHCO3 0.35 g Calf serum 10.0 ml Water 950.0 ml 15% (v/v) 4% 15 c.p.5. Methocell. Incorporated for spinner medium. TABLE IIIE COST OF MEDIA FOR BHK21PD4 CELLS $ Cost per Lot at Item $ Cost per Lot at Cost per Liter Medium Bulk Rate MEM 1% CS Medium Calf serum 18.00/500 ml 36.0O MEM powder (GIBCO Cat. No. F16) 37.38/50 liter lot 64.76~ Tryptose phosphate 8.25/540 g 4.86~ Total 105. 62 2 x 106 cells ml-1 formed: Cost of 1 x 109 cells = 52.81 (without tax). Schedule 8 Medium Calf serum 18.00/500 ml 360.00O Tryptose phosphate 8,25/540 g 4,860 Tris 21.00/500 g 1.7 HEPES 35.00/100 g 168.000 Glutamic acid 7.75/1000 g 0.47~ Dextrose 2 50/100 g 0.750 BME vitamins 3 o 60/100 ml 10.80O MEM amino acids 8.20/200 ml 12.300 NE amino acids 3.90/100 ml 11.700 Total 570e590 5 x 106 cells ml1 formed: Cost of 1 x 109 cells = 114.12,. Note: No adjustment is made for labor as MEM 1% CS medium is virtually preprepared, 150

GROWTH RESPONSE TO SERUM CONCENTRATION This information is incomplete, but may be of peripheral interest, During growth of BHK21PD4 cells in schedule 8 medium at varying serum levels, it was found that an increase of 1,5% in the serum content of the medium led to an increase of 1.0 x 106 cells/ml in the final yield of cells. THE EFFECT OF VARYING GLUCOSE LEVELS ON GROWTH OF BHK21PD4 CELLS In the previous report, it was noted that BHK21 cells are highly glycolytic. When cells are grown in schedule 8 medium containing 0.34o glucose, about one-third of the glucose is converted to lactic acid, a further third is not metabolized at all, and the remainder is presumably dissimilated via glycolysis and the TCA cycle. When the dry weight of cells produced is compared with the ATP yielded by glucose catabolism, an answer of 11.8 mg dry weight per Lmole ATP is obtained. This is in fairly close agreement with the value obtained for bacterial systems, generally considered to be 10 fig dry weight per 4mole ATP. It therefore appears that in schedule 8 medium, glucose is well above growth-limiting levels. Better pH control, a prerequisite for vaccine production, might be achieved if glucose is lowered to the level where it is theoretically limiting (about P0.1%) hopefully reducing the amount of lactic acid formed. Two types of experiment were attempted, (1) A series of vessels containing medium of increasing glucose content were set up to determine optimal glucose level when given as a one-step addition. (2) Alternatively it was hoped to obtain a high cell yield with minimum lactate production by adding small aliquots of glucose sequentially as required, ie., whenever growth rate appeared to decline, 1. Schedule 8 medium was made up without added glucose dispensed into spinner flasks, glucose added to levels of 0-0,3% above the basal value. pH and cell number were followed with time. Cell yields and pH delcine are given in Figure lE. Limitation of glucose does not provide an effective method of pH control without producing a concommitant fall in cell yield, Cell yield appears directly proportional to glucose added, There is some indication of slower growth at lower glucose levels, At a given time after attainment of maximal yield, cultures grown at lower glucose levels show a lower percentage of surviving cells, pH curves for these cultures fall into two distinct classes; cultures grown on 0.067% added glucose and below showing moderately steady pH, and cultures grown above this level showing a rapid pH decline and a later rise, Probably at a critical glucose level, enough lactate is formed to exceed the buffering capacity of the medium. 151

6 A Cell Yield 0 pH Decline 0. 2 -4I / 0.4 3D- 0.6 x C-)LL( 3 0. % GLUCOSE ADDED 0.8 1.0 152

2, For the growth curve involving sequential glucose additions, glucose was initially added at a level of over 0.02% above the basal level, Further additions were made throughout the growth period. Unfortunately, further additions of glucose were unable to stimulate rapid division once the growth rate had slowed and the cells appeared to be moving towards maintenance rather than growth. The final doubling time recorded was 129 hr, After 145 hr 0.140 glucose had been added with a yield of 28o,0 x 105 cells/mi and a concomitant pH drop of 0.55 unit, This shows little advantage over the one-step system where 0.134o% added glucose yielded 30.0 x 105 cells/ml with a fall of 0.52 pH unit, o It is concluded that pH control of these cultures could best be achieved by direct addition of alkali. FREEZING OF BHK21PD4 CELLS FOR SUSPENSION CULTURE An attempt was made to freeze and recover dense suspensions of BHK21PD4 cells for spinner to spinner transfer based on the procedure described by Walker, In Walker's paper dimethyl sulfoxide (DMSO) was used routinely as a preservative, and processing of cells at 4~0C was found to be superior to that at room temperature, Paul (personal communication) states that DMSO is toxic at higher temperatures, but not at 4~C. It has recently been shown in this laboratory that BHK cells suspended in their normal growth medium with 5% DMSO at 360C show a progressive inability to divide in fresh medium, The present procedure involved harvesting at room temperature and cooling the cells, resuspended in fresh medium, in ice prior to addition of DMSO. Two different media, namely schedule 8, and MEM 1% CS were used in growth and processing of the cells, and in each case cells were frozen with and without DMSO by way of comparison. Media in which the cells were frozen were the same as that used for growth except that 15% (v/v) of 40 15 cps methocell was added to the freezing medium, 100 ml spinner cultures were set up and harvested in the late-log phase when the schedule 8 and MEM 1% CS-grown cultures had the following characteristics respectively: 5304 x 105 cells/ml, > 99% viability; 22,0 x 105 cells/ ml, 87% viability. Cells were harvested at RT and resuspended in freezing medium at an estimated 20-25 x 106 cells/ml, and cooled in ice. 5% DMSO was added to half of the samples so prepared, Samples were dispensed in 2 ml aliquots in screw cap tubes instead of sealed ampoules as it was felt that these vessels were a more likely model for the bulk handling of cells, The screw cap tubes were subjected to preliminary cooling (6-8 hr) in a styrofoam box (10 in. x 10 in, x 6 in., with 1 in, wall thickness) placed in a Revco freezer and then stored in the vapor phase of liquid nitrogen, (This is essentially the technique used in J. Pauls' laboratory.) 153

After six days' storage, samples were thawed rapidly, and inoculated into 100-ml spinner cultures containing the appropriate growth medium at 2.5 x 105 cells/ml. After retrieval the characteristics of the inocula were as follows. 1. Schedule 8 cells + 5% DMSO —90% viability > 99% viability after 40 hr 2. Schedule 8 cells - No DMSO-74% viability < 1% viability after 24 hr 30 MEM 1% CS cells + 5% DMSO —89% viability-always 90-97% viability 4. MEM 1% CS cells - No DMSO —O viability No growth was observed with inocula processed without DMSO. Cells frozen in the presence of DMSO showed apparently normal growth (see Figure 2E). SURVIVAL OF BHK21PD4 CELLS AT 40C It is generally known that when cells of E. coli are suspended in a salt solution at 4~C at about 1 x 103 cells/ml, cell death is rapid. This effect is not shown if cells are at a density of 1 x 108 cells/ml, or if they are kept as a centrifuged pad. Mammalian cells may display similar behavior, and if so, cold storage without freezing may be advantageous if one wishes to use the same batch of cells for inocula over a period of a few days. Two log-phase spinner cultures grown with MEM 1% CS and schedule 8 medium and containing 1.0 and 3.0 x 106 cells/ml, respectively, were stored at 4~C with stirring. At intervals cells were used as inocula for monolayer cultures (MEM 10% TP 5% CS) inoculated at 2.5 x 105 viable cells/ml, In all but the final sample (22 days) no difficulty was found in establishing monolayer cultures, and even in this case cells began to grow after a three-day lag. Death in schedule 8 medium resulted in much cellular debris, In MEM 1% CS medium it was noted that smaller cells tended to die first. In neither of these "open" cultures did the pH rise markedly from that obtaining prior to cooling, Stationary phase cells grown on schedule 8 medium were centrifuged, the medium decanted and the cell pad stored, Monolayers were established by transferring small portions of the pad with a pasteur pipette. In this instance death was more rapid, but a viable monolayer could be obtained after 11 days storage. Results are given in Table IVE. There is a possibility of obtaining more prolonged survival by holding cells at 1-20C above the freezing point of the suspending medium,

22.0 21. 0 d 20.0 Min DT=22.5H )N Min DT=14 / Schedule 8 Cells 19.0 / A/ MEM 1% Cells 18.0 0 20 40 60 80 100 120 1 40 TIME, HR Figure 2E. Spinner growth of BHK21PD4 cells from frozen stock. 155

TABLE IVE SURVIVAL OF BHK21PD4 CELLS AT 40C A. Spinner Cultures in Growth Medium Time Schedule 8 MEM 1% CS in % Viability Cell No. % Viability Cell No. Mono Made Days x 105/ml x 10-5/ml 0 98 28.6 88 13.12 1 97 30.0 87 12.2 2 93 28,0 89 12,2 3 83 30.1 90 12,8 4 83 28.0 77 12.4 5 86 24.0 81 9.2 6 78 24.2 75 12.1 11 29 10.4 68 11,0 12 25 10.4 67 11.0 15 25 8.8 59 9.0 18 40.7 7.0 57 8.o \ 22 "25 6.9 39 6.5 B. Centrifuged Pad Time in Days 0 1 2 4 7 11 % Viability 86 35 41 47 36 ~15 Mono Made V/ / 156

UPTAKE OF AMINO ACIDS BY BHK21PD4 CELLS DURING GROWTH IN SCHEDULE 8 MEDIUM All analyses were done with a Beckman 120C amino acid analyzer. Ultrafiltrates of calf serum and tryptose phosphate were examined as a preliminary to determine whether or not these materials incorporated unidentifiable components to such an extent as to render a rough estimate of amino acid uptake impossible~. Although components other than the 21 common amino acids were observed, their concentration was relatively low and their elution times such that they did not interfere with the normally encountered amino acids, Tryptose phosphate contained nine unidentified, highly acidic components, which were eluted before aspartic acid, Three unidentified components were seen among the neutral amino acids. Lysine, histidine, arginine, tryptophan aspartic acid, threonine, serine, glutamic acid, proline, glycine, alanine, valine, methionine, isolencine, leucine, tyrosine, and phenylalamine were observedo The presence of cystine was suggested, but not confirmed, by a shoulder on the alanine peak. Some ammonia is present. Asparagine, and a possible small amount of glutamine, were identified tentatively. Upon autoclaving, there is some degradation of the "pre-aspartic acid" component and a loss of the presumed glutamine. Quantitatively, tryptose appears to provide the largest amount of amino acids of the medium components and is notably rich in leucine and lysine. At the level used (10%), calf serum does not appear to contribute a major portion of free amino acids, at least in the initial stages of growth0 There are at least two unidentified components, eluted before aspartic acid. A small amount of hydroxyproline is present, One double peak observed could be attributed to serine, threonine, and glutamine, or alternatively to a combination of two of these alone,, The major peak in the chromatogram is evidently ammonia o When an ultrafiltrate of whole medium was analyzed, the quantitative contribution of the unidentified components appeared to be relatively minor0 There is a marked increase in ammonia levels during growth, but little loss of arginine (see Table VE)o The levels of proline and glycine increase throughout growth0 Aspartic and glutamic acids are diminished by more than 90%, serine shows a 75% loss, and all other amino acids are depleted to a lesser degree. None of the essential amino acids appears to be limiting. Ammonia production, and the great depletion of aspartic acid and glutamic acid suggest that these amino acids may be dissimilated via deamination, transamination, and the TCA cycle0 A gross nitrogen balance of free amino acid uptake reveals that a maximum of 22% of all protein synthesized could be derived from free amino acids and the remainder is presumably obtained from the proteins and peptides present in calf serum and tryptose phosphate0 The behavior of serine, a presumably nonessential amino acid, is interesting in that it may be a possible source of the free glycine released, and that n157

TABLE VE UPTAKE OF FREE AMINO ACIDS BY BHK21PD4 CELLS DURING GROWTH Culture time (hr) 00.00 20.50 48.45 67.75 92.75 116.50 Change in AA Conc Cell yield (x 10-5/ml) 00.00 6.4 18.5 33.7 47.5 46.5 (4M/ml) Amino acid (iM/ml) Lysine 1,160 0.995 1,080 1.120 1,16 1.08 -0.08 Histidine 0.415 0,340 0.369 0.400 0.405 0.375 -0.04 Ammonia 0,532 0.767 0.128 1.810 3.050 3,460 +2.928 Arginine 1.140 1.040 1,030 1,070 1.070 1,040 -0.10 Tryptophan 0.078 0O074 0,550 0.078 o.o087 0.052 -0,026 Aspartic acid 0.475 0.305 0.136 0.085 o.o48 0.036 -0,439 Threonine 0.915 0.934 o.865 1.020 0.815 0.803 -0,112 Serine 1.020 o.855 0.550 0.445 0.385 0.253 -0.769 Glutamic acid 5.350 4.970 3.070 1,520 0.069 0.070 -5.280 Proline 1.060 o.890 0.995 1.240 1.380 1,66 +0.60 Glycine 0.660 0.628 o.760 o.980 0.850 0.737 +0.077 Alanine 0.858 1,010 1.520 1,970 1.21 0.708 -0o150 Valine 1.150 1,090 1,050 1,110 1.08 0.990 0.160 Methionine 0.232 0.268 0.236 0.268 0.156 0.137 -0.095 Isolencine 0,o860 o.830 0.770 0.915 0.757 o0.680 -0.160 Lencine 1,400 1,170 1.150 1,280 1.040 0.915 -o.485 Tyrosine 0,320 0.345 0.322 0.465 0.285 0.265 -o.o55 Phenylalanine 0.505 0.572 o,540 0,.667 0.487 0.425 -0.140 158

the rate of uptake per cell varies markedly throughout growth (see Figures 5E and 4E)o Both the rate of serine removed from the medium by a fixed number of cells present, and the amount of serine used in the production of new cells, show a rapid decline during exponential growth (both parameters are expressed in Ctmoles serine used per million cells over a 10-hr period). This would suggest that either a large pool of serine accumulates during growth, or that serine requirements are population-dependent, cells at high densities being better able to produce or retain serine itself, or its possible metabolite, glycine 159

5 1. 2 4L /ells 1.0 x x~~~~~~~~~3 ~~ —0.8'0 L_ 0. 6 LU Serine 0.4 0. 2 20 40 60 80 100 120 TIME, HR Figure 3E. Supernate levels of serine and glycine during growth of BHK21PD4 cells. 160

SERINE UPTAKE IN MOLES/10-6 CELLS FORMED - N3~~~~~~~~~~~~~~~~~~~ cIj W~I (3T CD 4r P CD co cu — k ci0' ON m Ii;C 001C CD oo 2 C) n~~~~~~~~C ci~~~~~~~~~31~~~~~~~~~~~~~~~/ 0 - 0~~~~~~~~~~~~~~c. c:)~~~~~~~~~~~~C H CD C) to 00 N) 0' RATE SERINE, UPTAKE IN MOLES/1U-6 CELLS PRESENT PER 10 HR. PERIOD

DISTRIBUTION LIST No. of Agency Copies Commanding General 4 U. S. Army Medical R&D Command Attn: MEDDH-SI Washington, D. C. 20315 Commanding Officer U. S. Army Medical Unit Fort Detrick Frederick, Maryland 21701 Commanding General 5 U. S. Army Materiel Command Attn: AMCRD-BC Washington, D. C. 20315 Defense Documentation Center 20 Attn: DDCIR Cameron Station Alexandria, Virginia 22314 Commanding Officer U. SO Army Combat Development Command Medical Services Agency Brooke Army Medical Center Fort Sam Houston, Texas 78234 162

Unclassified - DOCUMINT CONTROL DATA. R D (Y.qeeurtlfrrli lltloifn nrl t.tlo, bhrody of hlrtrrt end lnhdetlng annofntlon rneti he entered when rlhe overall report Is claaIfled) e. QU9t~INATlNG ACtIVITY (CCorporal.re ")taulo) I128. REPORT SECURITY CLASSIFICATION Unclassified The Regents of The University of Michigan 2b. GoOUP Ann Arbor, Michigan 48104 3. P-PONt tITLE Large Scale Fermenter Growth of Animal Cells for Virus Vaccine Production: Control of Seed Stocks and Growth Conditions (U) 4, bOEsCRIPTltE' NO.tS (Dtpe of repot and Inctlealve dlesa) Final Report (1 January 1969 to 31 August 1969) 5. A u T fe It) r(Firs naeo, middle InIts 1, toa l name) Donald J. Merchant Iloyd L. Kempe e. REPORT DAt t7a. TOTAL NO. OF PAGES 7b. NO. OF REFS December 1969 161 145 64. ChN t-RAC t OR t AN t NO. o8. ORIGINATOR'S REPORT NUMBERtSh DADA-17-67-C-7073 b. PROJIECT No. 08621-3-F C. 9b. OTHER REPORt NO(s) (Any other numbers that may be asoSgned thie report) d. to. oISTRIrUTIOn" STATEMENT Each transmittal of this document outside the Department of Defense must have prior approval of Commanding General, U.S. Army Medical R&D Command. The findings of this report are not to be construed as an official Department of the Army position unless so designated by other authorized documents. I. SUPPLEMEWTARY NOTES 12. SPONSORING MILITARY ACTlVITY U. S. Army Medical R&D Command Office of the Surgeon General Washington, DC. 20315 13. ABSTRACT Using equipment previously constructed on the program L-M mouse cells have been grown in a 5liter fermenter in medium 199 supplemented with peptone and serum. Changes in pH, O-R potential, dissolved oxygen and cell number and viability have been continuously monitored. Cells also have been grown under conditions of controlled pH and O-R potential and the relationship of O-R potential to growth rate has been determined. Work has been completed on a computer program for analysis of chromosome changes in aneuploid cell populations. The method also is applicable to study of diploid cells. Automated methods combined with computer analyses permit a rapid and sensitive monitoring of chromosomal shifts in cell population. Using the computer program, as well as more conventional methods of chromosomal analysis, data is presented on the effects of several types of environmental selection pressures on the chromosomal complement of L-M mouse cells and BHK21PD4 cells. Detailed studies of the amino acid nutrition of L-M mouse cells in chemically defined medium is presented. Combining the simplicity of Eagle basal medium with the sensitivity and accuracy of an amino acid analyzer, it has been possible to obtain critical evidence concerning the rates of use of individual amino acids, the role of imbalance of amino acids in utilization patterns and the effects of supplementing nonessential amino acids. Finally, studies of the nutrition of BHK21PD4 cells have led to the development of a very efficient medium (schedule 8) for this cell line. The comparative costs and yield is compared with more conventional media. Improved methods of storage and of freezing seed stocks of these cells also are presented. DD, NOV..1473 Unclassified Security Ctlassification

& Ita 0 ro.....Ie I..!_I~III:11 2 i- I= S ~ ~ ~ ~ ~ I N I III-2*I I i i I~I I I l lI..I I l I' iii I II I II I LI ~ ~ I II iii iii I I I J I I I lJ II II II I I I. _.I, i I I I.. II. I I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~. 0 ii

UNIVERSITY OF MICHIGAN 3 9015 03695 43II 11 llll 3 YO5 3695 5543