ILAR Journal Online, Volume 31(1) 1989: Perspectives on Animal Use

Online Issues

<< All Back-issues

<< This Issue's Table of Contents

ILAR Journal V31(1) 1989
Perspectives on Animal Use

Refinement of Monoclonal Antibody Production and Animal Well-being
Michael W. McGuill and Andrew N. Rowan

Mr. McGuill is a second-year veterinary student at Tufts University School of Veterinary Medicine. Dr. Rowan is associate professor in the Department of Environmental Studies and director of the Tufts Center for Animals and Public Policy at the Tufts University School of Veterinary Medicine. He is a member of the ILAR News Editorial Panel and the ILAR Committee on Pain and Distress in Laboratory Animals. Address reprint requests to Dr. Rowan, Tufts School of Veterinary Medicine, 203 Harrison Avenue, Boston, MA 02111.

Introduction

In 1975, two Cambridge University researchers published a report describing how they successfully fused myeloma and antibody-producing cells to produce a set of single cells that could be grown in large quantities (hybridoma) (Kohler and Milstein, 1975). Each cloned cell secreted only a single antibody type with a single binding specificity and affinity (monoclonal). Standard immunization protocols using a single antigen result in the secretion of a range of antibodies with varying specificities and affinities for the antigen. Monoclonal antibodies (MABs) have had an enormous impact on research and disease diagnosis, and Kohler and Milstein were deservedly awarded a Nobel Prize (in 1984) for their technical discovery.

MABs are used in an increasingly wide variety of experimental and clinical applications. One common way to produce MABs involves injecting hybridoma cells into the peritoneal cavities of laboratory animals, particularly mice. Antibody in the resulting ascitic fluid of hybridoma-inoculated mice is of much higher titer than in the serum of animals traditionally immunized, which makes mouse ascites production the preferred procedure when it is feasible. Yelton et al. (1981) made the point that a single hybridoma-inoculated mouse can produce as much antibody as can be obtained from multiple bleedings of a well-immunized goat.

MABs can be produced in cell culture with high levels of purity, but titers of MABs in the ascitic fluid of hybridoma-inoculated mice have been reported to be at least 1,000 times greater than in the spent medium of cell culture (Galfre and Milstein, 1981). More recently, Epstein and Epstein (1986) reported that ascites fluid MAB titers average 1-10 mg/ml, and spent fluid titers from cell culture average 0.01-0.1 mg/ml. Advances in cell culture technology have made mass production of MABs competitive, as the recent growth in the antibody-production sectors of the biotechnology industry indicates. For example, using an automated cytostat for MAB production, Fazekas de St. Groth (1986) reported an average yield of 10-20 mg MAB per day per 500 ml cytostat. Under the conditions described by Brodeur et al. (1984), an average mouse produces 6.5 ml of ascites fluid over three days. If MAB titers are between 1 and 10 mg/ml, as reported by Epstein and Epstein (1986), a single mouse would produce between 2 and 20 mg MAB per day. However, a mouse will produce ascites for only a few days, while hybridomas in culture often secrete antibody for much longer. Roder et al. (1986) reported high yields in culture beyond 18 months.

For production of human MABs, in vitro techniques are preferred. Truitt et al. (1984) reported that, even though mouse ascites titers of human MABs are more than I00 times greater than cell culture fluid titers, the expense and effort involved in adapting and growing human/human hybridomas in athymic mice make that approach less practical. Epstein and Epstein (1986) discussed further advantages of in vitro and in vivo production of MABs.

It should not be assumed that in vitro procedures are inherently more humane. For instance, in vitro preparations sometimes use (fetal) calf serum, which may be obtained from animals under dubious circumstances. The "feeder" cells reported to promote selection and cloning of hybridomas (Cahoon et al., 1984) are obtained from the mouse peritoneal cavity. A recent paper (Lane et al., 1988) described a procedure to increase by 16-fold the number of feeder cells, primarily macrophages, that can be collected from the peritoneal cavities of mice by preinjecting them with the inflammatory agents pristane and incomplete Freund's adjuvant (IFA), which are known to cause distress in large doses.

In most of the literature on in vivo production of MABs, concerns for maximizing antibody production seem to take precedence over concerns for the welfare of the experimental animals. Given the growing awareness of animal welfare issues within public and scientific communities, it is surprising that these new techniques, which have had a major impact on the number of animals used in research, have been developed and described with so little overt attention given to the alleviation of animal pain, distress, and suffering.

Stresses and Potential Stresses of Ascites and MAB Production

A number of experimental manipulations required to maximize antibody production do cause animal stress or suffering. Brodeur et al. (1984) reported that survival times of mice reflect the hybridoma cell concentrations injected into them intraperitoneally (IP). For instance, mice injected with 3.5 × 10⁷cells survive 8.5 days on average, while mice injected with 3.2 × 10⁶ cells (the number that yields optimum ascites production) survive 12.7 days on average. Also, pristane, which causes distress in large doses, is injected into the peritoneal cavity to "prime" the cavity and increase ascites and antibody yield.

A recent paper by Amyx (1987) addressed pretreatment with pristane. According to Amyx, pristane is thought to induce granulomatous reactions and interfere with peritoneal fluid drainage. Large volumes of pristane injected IP into mice are associated with weight loss, a hunched appearance, and lack of activity. Amyx suggests lowering the dose of pristane to minimize these effects. In spoken comments at a "Colloquium on Recognition and Alleviation of Animal Pain and Distress" (American Veterinary Medical Association, Chicago, May 1987), Amyx reported that the usual dose of 0.5 ml pristane produces strong distress symptoms, while a smaller dose of 0.2 ml produces milder distress symptoms.

Brodeur et al. (1984), in an examination of the impact of different experimental manipulations on ascites fluid volume and antibody titer, found that 0.5 ml of pristane resulted in the formation of a greater volume of high-titer ascitic fluid than did 2.0, 1.0, and 0.2 ml of pristane, but there was considerable variation from one batch to the next, with average MAB titer levels in the fluid ranging from log 3 to log 5. Survival times of mice injected with 1.0, 0.5, and 0.2 ml pristane were similar (11.8, 11.4, and 11.2 days, respectively). The difference between the high-titer fluid volumes produced by 0.5 ml (24 ml) and 0.2 ml (18 ml) of pristane pretreatment, therefore, may not be that significant.

Some researchers (Colwell et al., 1986; Kwan et al., 1980) reported success using 0.2 ml pristane for priming. Recently, Hoogenraad and Wraight (1986) reported no significant differences in ascites production following preinjection with 0.1, 0.2, and 0.5 ml of pristane. Mice preinjected with 0.1 ml of pristane yielded an average of 9.65 ml of ascitic fluid, and mice preinjected with 0.5 ml pristane yielded an average of 9.72 ml of ascitic fluid. Therefore, use of 0.5 ml pristane may not justify the additional distress associated with the greater volume of pristane. However, 0.1 ml pristane was the lowest volume that resulted in all the mice developing ascites tumors. If 0.05 ml was used for priming, then only two-thirds of the mice developed ascites tumors.

Because pristane is a suspected carcinogen, alternatives to pristane have been evaluated, especially for production of antibodies for use in humans. Gillette (1987) reported that of seven agents studied as priming agents for MAB production, only IFA produced results comparable or superior to pristane. Other agents evaluated included proteose-peptone, thioglycollate, corn oil, mineral oil, and complete Freund's adjuvant. Mueller et al. (1986) reported time savings when using IFA priming instead of pristane priming. In their study, mice primed with IFA could be injected with hybridoma cells as soon as one day after priming and then tapped (fluid removed) for high titer ascitic fluid within 14 days of priming. Mice primed with pristane are usually not tapped for two to three weeks after priming.

In neither of these studies was the stress caused by the priming agents evaluated. In Gillette's study, mice primed with IFA survived for more taps once ascitic fluid was produced than did mice primed with pristane, but the implications of this result in terms of mouse suffering are unclear. IP injection of high doses of complete Freund's adjuvant are associated with signs of distress similar to those caused by high doses of pristane, but lower pristane doses are associated with mild, transient clinical signs (Amyx, 1987). The same may be true of IFA doses. Further studies are needed to find what, if any, doses of IFA cause minimum stress to mice while still producing high titer ascitic fluid.

Without evidence to the contrary, we should assume that ascites, by itself, causes discomfort and distress in mice because it is known to cause discomfort in humans. In humans, mild ascites is generally painless (Burnside and McGlynn, 1987), although patients with massive tense ascites are frequently unable to ambulate and experience abdominal discomfort, indigestion, and heartburn (Mauch and Ultmann, 1985; Pockros and Reynolds, 1986). Elevation of the diaphragm due to ascites is associated with dyspnea, orthopnea, or tachypnea (Mauch and Ultmann, 1985). In mice used for MAB production, ascites fluid and the tumors causing the ascites fluid occupy peritoneal space and result in distension of the abdomen. One may remove the ascites fluid, but the cell mass continues to increase and, presumably, will cause distress and suffering. Brodeur et al. (1984) reported that draining ascitic fluid as soon as it accumulates is necessary to reduce mortality.

A recent paper by Brodeur and Tsang (1986) reported that first-generation crossbred mice (BALB/c x SW) yielded four times as much ascitic fluid as was obtained from BALB/c mice. The male mice, having an average size of 35 g, produced about 28 ml of high-titer ascitic fluid over 8.9 days, or about 3.2 ml per day. Using these crossbreeds could result in the use of fewer mice, which answers the need to reduce animal use as much as possible. However, the paper does not address the possible additional stress to each individual animal that results from such a rapid build-up of ascites mass or, perhaps more importantly, from removal of such large volumes of ascites fluid during repeated samplings.

In human clinical medicine, paracentesis of ascitic fluid is discouraged because of fear of hypovolemia (Kao et al., 1985; Lamont et al., 1983). For patients with massive ascites, Lamont et al. (1983) recommended gradual paracentesis of up to 3 liters over 4 hours. Boyer (1986) recommended that ascitic patients should not lose more than 0.3-0.5 kg fluid per day. Pockros and Reynolds (1986) reported that mobilizing large amounts of ascitic fluid (more than 1 liter per day) in cirrhotic patients by rapid diuresis results in hypovolemia (a 24-percent drop in plasma volume) and renal insufficiency. In cirrhotic patients with ascites and edema, these negative effects are not seen following rapid diuretic mobilization of ascites or large-volume paracentesis, at least initially, because the edematous fluid is mobilized preferentially without causing blood volume contraction (Kao et al., 1985; Pockros and Reynolds, 1986; Quintero et al., 1985). However, hypovolemia and renal insufficiency appear after edema is depleted (Pockros and Reynolds, 1986).

The literature does not address possible effects of paracentesis of ascitic fluid from the mouse, but with samplings averaging 3 ml--a volume that is well over the total blood volume of a mouse--the possibility of physiologic stress due to hypovolemia must be considered. If investigators wait up to three days between tappings, sample volumes may be significantly greater, and the potential for hypovolemia may likewise be increased. (Galfre and Milstein [1981] recommended sampling every one to three days; Epstein and Epstein [1986] recommended sampling at intervals of two to three days.)

Golba et al. (1974) reported that frequent IP injections in rats are stressful regardless of the kind of material injected. Thus, IP sampling of ascitic fluid in mice may, by itself, be an additional stressor to consider in designing experimental protocols for harvesting MABs.

Recommendations

Galfre and Milstein (1981) suggested that, "in the long run, even large-scale preparations of pure monoclonal antibody may use spent medium from cultured cells as a more humane and better-controlled source. But this will depend on the technological development of large-scale cell culture methods." This is the optimum approach, and apparently such technology is now competitive. But in vivo methods are still used because the costs of establishing in vitro culture facilities at individual institutions are high or because particular hybridomas cannot be grown in culture without difficulty.

The technology to replace in vivo hybridoma techniques with in vitro cultures is being developed and refined. In the short term, a similar effort should be applied to reducing the stress and suffering of animals that are used for in vivo methods. For instance, lower doses of pristane (0.1-0.2 ml) should be used instead of the more conventional 0.5-ml dose. Also, animals should be humanely killed when they show signs of the moribund condition to minimize suffering associated with the final stages of dying. Many institutions now permit only two taps, the final tap being part of a terminal procedure.

Furthermore, research is necessary to provide a better understanding of the stresses endured by animals under protocols designed to maximize production of monoclonal antibodies. The effects of large intraperitoneal tumor masses, of frequent and large-volume paracentesis, and of pristane priming must be studied to provide a scientific basis for developing procedures that minimize animal distress and suffering.

References

Amyx, H. L. 1987. Control of animal pain and distress in antibody production and infectious disease studies. J. Am. Vet. Med. Assoc. 191( 10): 1287-1289.

Boyer, T. D. 1986. Removal of ascites: What's the rush? Gastroenterology 90:2022-2023.

Brodeur, B. R., and P. S. Tsang. 1986. High yield monoclonal antibody production in ascites. J. lmmunol. Methods 86:239-241.

Brodeur, B. R., P. Tsang, and Y. Larose. 1984. Parameters affecting ascites tumor formation in mice and monoclonal antibody production. J. lmmunol. Methods 71:265-272.

Burnside, J. W., and T. J. McGlynn. 1987. P. 226 in Physical Diagnosis. Baltimore: Williams & Wilkins.

Cahoon, B. E., R. J. Sugasawara, R. S. Cubicciotti, and A. E. Kara. 1984. Influence of macrophage-conditioned media on cloning efficiency, antibody synthesis, and growth rate of hybridomas. Hybridoma 3(I ):75.

Colwell, D. E., S. M. Michalek, and J. R. McGhee. 1986. Method for generating a high frequency of hybridomas producing monoclonal IgA antibodies. Methods Enzymol. 121:42-51.

Epstein, N., and M. Epstein. 1986. The hybridoma technology: 1. Production of monoclonal antibodies. Adv. Biotechnol. Processes 6:179-218.

Fazekas de St. Groth. S. 1986. Automated production of monoclonal antibodies in a cytostat. Methods Enzymol. 121:360-375.

Galfre, G., and C. Milstein. 1981. Preparation of monoclonal antibodies: Strategies and procedures. Methods Enzymol. 73:1-46.

Gillette, R. W. 1987. Alternatives to pristane priming for ascitic fluid and monoclonal antibody production. J. Immunol. Methods 99:21-23.

Golba, S., M. Golba, and T. Wilczok. 1974. The effect of trauma, in the form of intraperitoneal injections or puncture of the orbital venous plexus, on peripheral WBC count in rats. Acta Physiol. Poi. 25(4):339-345.

Hoogenraad, N. J., and C. J. Wraight. 1986. The effect of pristane on ascites tumor formation and monoclonal antibody production. Methods Enzymol. 121:375-385.

Kao, H. W., N. E. Rakov, E. Savage, and T. B. Reynolds. 1985. The effect of large volume paracentesis on plasma volume--a cause of hypovolemia? Hepatology 5(3):403-407.

Kohler, G., and C. Milstein. 1975. Continuous cultures of fused cells secreting antibody of predefined specificity. Nature 256:495-497.

Kwan, S. P., D. E. Yelton, and M. D. Scharff. 1980. Production of monoclonal antibody. Genet. Eng. 2:31-46.

Lamont, J. T., R. S. Koff, and K. J. Isselbacher. 1983. Cirrhosis. !n Harrison's Principles of Internal Medicine, 10th ed., R. G. Petersdorf, R. A. Adams, E. Braunwald, K. J. Isselbacher, J. B. Martin, and J. D. Wilson, eds. New York: McGraw-Hill.

Lane, R. D., W. Renno, V. Nepomuceno, C. Schafer, and R. L. Mellgren. 1988. The influence of stimulated peritoneal feeder cells and mitogens upon antibody secreting hybridomas. Hybridoma 7(3):289-299.

Mauch, P. M., and J. E. Ultmann. 1985. Treatment of malignant ascites. Pp. 2150-2153 in Cancer: Principles and Practice of Oncology, vol. 2, 2d ed., V. T. DeVita, S. Hellman, and S. A. Rosenberg, eds. Philadelphia: Lippincott.

Mueller, U. W., C. S. Hawes, and W. R. Jones. 1986. Monoclonal antibody production by hybridoma growth in Freund's adjuvant primed mice. J. lmmunol. Methods 87:193-196.

Pockros, P. J., and T. B. Reynolds. 1986. Rapid diuresis in patients with ascites from chronic liver disease: The importance of peripheral edema. Gastroenterology 90:1827-1833.

Quintero, E., P. Gines, V. Arroyo, A. Rimola, F. Bory, R. Planes, J. Viver, J. Cabiera, and J. Rodes. 1985. Paracentesis versus diuretics in the treatment of cirrhotics with tense ascites. Lancet i:611-612.

Roder, J. C., S. P. C. Cole, and D. Kozbor. 1986. The EBV hybridoma technique. Methods Enzymol. 121:140-167.

Truitt, K. E., J. W. Larrick, A. A. Raubitschek, D. W. Kuck, and S. W. Jacobson. 1984. Production of human monoclonal antibody in mouse ascites. Hybridoma 3(2):195-199.

Yelton, D. E., S. B. Roberts, and M. D. Scharff. 1981. Hybridomas and monoclonal antibodies. Lab. Manage. 19(I): 19-24.

Commentary: Kevin Kenny

Mr. Kenny is a graduate student in immunology at the New York State College of Veterinary Medicine, Ithaca, New York.

The paper by McGuill and Rowan deals with refinement of techniques involved in monoclonal antibody (MAB) production. Three techniques comprise the focus of the article: in vivo production of MAB in ascitic fluid, the use of peritoneal exudate cells (PEC) as feeder layers, and the use of fetal bovine serum (FBS) in hybridoma media. Over the past few years, the Mastiffs Research Laboratory at Cornell University has been involved in the production of MAB specific for a number of bacterial-derived (staphylococcal cellular and extracellular antigens) and animal-derived (bovine immunoglobulins) antigens. During this time, the opportunity has arisen to refine certain techniques used in MAB production. The following comments represent observations made in this laboratory along with those of other scientists regarding the various techniques described by McGuill and Rowan.

Good yields of MABs have been achieved by culturing cells in vitro in both dialysis tubing and in bioreactors. Sjogren-Jansson and Jeansson (1985) originally described the dialysis tubing method, and it has recently been modified by Jwo and LoVerde (1988). Cells are placed in dialysis tubing, which is then sealed and placed in a tissue culture flask. Media is added, and the cells are incubated for 10 days, being agitated by shaking twice a day. Media in the flask is changed every second day. A yield of 8-11 mg (MAB) per tube can be attained for MAB of either lgM or IgG subclasses. It has also been possible to culture hybridoma cells in a bioreactor for prolonged periods with yields of antibody ranging from 100-500 mg per week (Evans and Miller, 1988).

The use of alternatives to PEC as feeder layers for hybridomas has also been examined. Parent myeloma cells seeded at a concentration of 107 cells per 96-well plate (105 cells per well) in hypoxanthine-aminopterin-thymidine (HAT) media support fusions and cloning adequately. Spleen cells and PEC from mice destined for euthanasia have been harvested, suspended in 90 percent serum and I0 percent dimethyl sulfoxide (DMSO), and frozen in liquid nitrogen. After thawing and viability counting, cells are seeded at 2 × 105 cells per well and 7 × 103 cells per well for spleen cells and PEC, respectively. It is imperative that mice used as cell donors not be infected with pathogenic viruses. Recently, interleukin-6 derived from a rodent bladder carcinoma cell line and used as a 5 percent v/v media supplement has been shown to adequately replace feeder cells (Dr. Kenneth Caroll, National Institute of Higher Education, Dublin, Ireland, personal communication, 1988).

Considerations of cost and animal welfare encouraged this laboratory to examine alternatives to FBS. We have found that gamma globulin-free horse serum supports the growth of bovine hybridomas and MAB production in a manner similar to FBS. Media composed of 65 ml Dulbecco's minimal essential media (D-MEM), 25 ml of HL-1 (Ventrex Labs, Portland, Maine), 10 ml of Optimem (Gibco Labs, Gaithersburg, Maryland), plus 13 percent FBS was found to be superior to D-MEM plus 20 percent FBS for conducting fusions.

I feel certain that there are people reading this journal who have pioneered methods to improve MAB production, reduce animal use, and improve animal welfare. It is of great importance that such information be circulated through the columns of this journal.

References

Evans, T. L., and R. A. Miller. 1988. Large scale production of murine monoclonal antibodies using hollow fiber bioreactives. Biotechniques 6:762-767.

Jwo, J., and P. T. LoVerde. 1988. Large scale production of monoclonal antibodies. Biotechniques 6:734-738.

Sjogren-Jansson, E., and S. Jeansson. 1985. Large scale production of monoclonal antibodies in dialysis tubing. J. Immunol. Methods 84:359-364.