Journal
of Hospital
Infection
Monoclonal
(1991)
18 (Supplement
antibodies-their
A), 443450
diagnostic
potential
L. R. Haaheim
Department of Microbiology and Immunology, The Gade Institute, Armauer Hansens Bldg, University of Bergen, N-5021 Bergen, Norway
Summary:
Considering the long and excellent performance of ‘classical’ immune sera in a range of diagnostic applications, the advent of ‘tailor-made’ highly specific monoclonal antibodies has given even higher hopes of diagnostic accuracy for the future. Understanding what monoclonal antibodies are and how they compare to polyclonal sources of antibodies is absolutely essential in order to appreciate their potential. This paper will briefly focus on the production, charactterization and use of monoclonal antibodies and look critically at the future prospects. Specificity and cross-reactivity are key terms which are explained with examples in this paper. It can be anticipated that diagnostic kits in the future will be composite assays using highly specific monoclonals (or oligoclonal cocktails) in conjunction Iwith high-avidity polyclonal sera, thus taking advantage of the best features of both systems. Keywords:
Diagnostics;
microbiology;
monoclonal
antibodies.
Introduction
Hybridoma technology’ which allows the production of ‘tailor-made’ monoclonal antibodies (Mabs) against virtually any antigen, has given medical science an invaluable tool. The availability of practically unlimited amounts of highly specific reagents has led to improved diagnostic performance in microbiology. A profound misconception is that hybridoma antibodies per se are monospecific, i.e. will react exclusively with one specific antigenic determinant only. Unforeseen cross-reactions occur regularly and are to be expected, as our understanding of molecular immunology is still fragmented.
How
to make
monoclonal
antibodies
There are many reviews addressing the technical aspects of generating hybridomas.24 The following text will focus on murine h/labs, because they are the most frequently used for diagnostic purposes. A simplified flow-chart for the generation of Mabs is shown in Figure 1. Bulb/c mice are of choice and individual B-cells are vaccinated with the antigen 0195%6701/91/06A443+08
lO3.00/0
0 1991 The Hospital
443
Infection
Soaety
L. R. Haaheim
Hybridoma
flow
chart
,“,9;“,‘n l:\ HATMedium
PEG as fusing
b
+ 14 days Antibody
Al
.;:‘!,i
agent screening
c\:. Cloning
+ 14 days-
Expansion
In viva-ascitic
New
of cells;
harvest
In vitro-cell
store
1. Flow
screening
antibody
fliud Purify
Figure
antibody
diagram
supernatants
antibodies cells
of the hybridoma
technique.
immortalized by fusing them with malignant B-cells from the same type of inbred animal; in this case myeloma cells from BaZb/c mice. A variety of such myeloma cells are available commercially and those that do not secrete any murine antibodies or immunoglobulin chains themselves, are most useful. Immunization methods vary in relation to dose, route and adjuvant. Traditional schedules of intradermal, subcutaneous, intravenous or intraperitoneal routes with or without adjuvant intrasplenic immunizations may be used with nanogram quantities of antigen bound to nitrocellulose paper5 and free intrasplenic deposition of antigen.‘j
Msonoclonal
antibodies
445
The two main differences between immunization protocols for generating Mabs and polyclonal immune serum are the timing of the booster and the purity of the immunogen used. In the former case one sacrifices the animal within 2-3 days after a booster, i.e. before most B-cells have matured to antibody-secreting plasma cells. This is critical, as it has been found that the mature plasma cells are poor fusion partners. The generation of polyclonal hyperimmune serum requires a ‘pure’ immunogen, otherwise an antibody response to vaccine impurities will reduce the quality of the serum. In the case of Mabs, however, one selects only those antibodies with the properties required, disregarding other irrelevant hybridomas. The most popular way of fusing immune B-lymphocytes and myeloma cells is by use of polyethylene glycol (PEG), but viruses (e.g. Sendai) and have been successfully employed. physical means (e.g. electrofusing) Fusion alone is insufficient a,s unfused myeloma cells will quickly outgrow the delicate hybridomas. Therefore, the myelomas have been made deficient in one of the salvage DNA pathway synthesis enzymes hypoxanthine phosphoribosyltransferase (HGPRT) or thymidine kinase (TK). The addition of aminopterin, a folic acid antagonist inhibits de nouo DNA synthesis so the myeloma cells will be killed. The hybridomas will continue to grow in this :HAT medium (hypoxanthine, aminopterin, thymidin) as they have acquired DNA salvage enzymes from the B-cell fusion partner. Thus, after IO-14 days in HAT medium only hybridoma cells are left which are more robust and can be maintained in medium without aminopterin (HT m’edium). The next step is to screen the many hundreds of wells from the master fusion plates in order to identify those that secrete the antibodies required. Automated analysis is preferable; RIA and ELISA techniques are most useful. ‘Classical’ antibody assays may also be considered (e.g. agglutination, neutralization, fluorescence). Flow cytometric analyses using fluorescein-labelled conjugatles can be used for screening large numbers of hybridoma supernatants for antibodies against certain types/subsets of cells which is the most rapid tech,nique. It is important to establish a good screening assay before starting the actual fusion procedure so tlhat the hybridoma required can be identified precisely. Murine spleen cells harvested for fusion can successfully be frozen in aliquots.3 They can at a later stage be thawed and fused when the screening assay has been improved or if one is looking for Mabs against antigens/determinants that were not planned at the outset. Once the selected hybridoma cells have been cloned and recloned (i.e. isolation of single-cell cultures by seeding at low density) the hybridoma cells can be expanded into larger flasks or injected into mice for ascitic fluid production. Generally, the yield in cell culture is about 10-100 pg ml-‘, whereas ascitic fluid can contain about l-l 5 mg ml-‘.2 Purification of Mabs from cell culture supernatants is usually straightforward, whereas Mabs in
446
L. R. Haaheim
ascitic fluid will be mixed with the antibodies from the ascites-producing animal itself, thus obscuring monoclonality and making purification protocols more elaborate. The last important point to be made is the question of record-keeping. The numerous master fusion plates, each giving rise to various cloned and subcloned cultures, some being at different stages of handling, all bearing cryptic acronyms and identification letters and numbers, require meticulous records. There are examples of local tragedies due to lost tracks of events. It is good practice to concentrate on just a few master wells at first, storing the rest frozen in their original plates for later handling and characterization. Properties
and use of monoclonal
antibodies
A monoclonal antibody forms part of the ‘natural’ immune response of the animal being used. Not surprisingly, therefore, the properties of each and all of the various Mabs generated against a particular antigen will vary with respect to specificity, cross-reactions with other related (or supposedly unrelated) antigens, isotype, affinity and biological properties. What makes the hybridoma technique so useful is that after having generated a particularly good hybridoma cell line it can be used to produce unlimited quantities of identical antibodies. Cells can be stored in liquid nitrogen for an almost unlimited time (or even in a deep-freeze at - 70°C or lower for a couple of years), only later to be revitalized. In contrast, a particularly valuable hyperimmine rabbit serum cannot, strictly speaking, be replaced by immunizing new rabbits. Figure 2 depicts a scenario where the immunogen has three different epitopes exposed on its surface. One of the Mabs generated (anti-C) only recognizes the immunogen and not the two other test antigens. This antibody thus appears to be ‘specific’. Needless to say, by using a larger repertoire of test antigens one could possibly identify other antigens with similar or identical epitopes; this ‘specific’ monoclonal antibody then transgresses into ‘cross-reactivity’. Another Mab (anti-B) may react weakly with antigen 2 and appear to be cross-reactive, as does the Mab against epitope A. Other test antigens could well have given quite opposite results. Needless to say, the very definition of specificity and cross-reactions is dependent on the yardstick used for comparisons, and is therefore not a well-defined concept. Some of these points can be illustrated by data shown in Table I. The post-infection ferret sera distinguish poorly between related A/H3N2 influenza strains A/Norway/25,32,33/80 by haemagglutination-inhibition tests and cannot precisely identify their relationship with the prototype strains A/Bangkok/l and 2/79. In contrast, the three hybridoma antibodies used, clearly identify two of the three Norwegian isolates as being similar to A/Bgk/1/79, whereas the last strain appears to resemble A/Bgk/2/79. This focuses upon the point of useful diagnostic antibodies. If in this case
Monoclonal
Specificity Epitopes lmmunogen
447
antibodies
I
and cross-reoctmns Reactivity AB 3f
C 3+
3-k
Y T
Antigen
Antigen
I
2
Figure 2. Pattern of specificity and cross-reactions of monoclonal antibodies the immunogen and two related antigens, each bearing three epitopes.
the purpose had been to identify influenza strains as say HlNl) and related to recent H3N2 prototype ferret sera would be preferable. In contrast, Mabs alone could easily miss variants. On the other hand,
in reactions
with
being H3N2 (and not, strains, post-infection MC7 or MC225 used for investigating finer
Table I. Use of murine monoclonal awdpolyclonal antibodies in strain characterization of H3N2 injZuenza virus. Haemagglutination-inhibition tests using post-infection ferret sera and murine monoclonal antibodies Virus
strains
Antibodies
A/Texas/l 177 A/Bangkok/l/79 A/Bangkok/2/79
Fl 2560 640 1280
A/Norway/25/80 A/Norway/32/80 A/Norway/33/80
1280 1280 2560
5120 1280
MC7 3200 400
of Hospital
Infection
Monoclonal
(1991)
18 (Supplement
antibodies-their
A), 443450
diagnostic
potential
L. R. Haaheim
Department of Microbiology and Immunology, The Gade Institute, Armauer Hansens Bldg, University of Bergen, N-5021 Bergen, Norway
Summary:
Considering the long and excellent performance of ‘classical’ immune sera in a range of diagnostic applications, the advent of ‘tailor-made’ highly specific monoclonal antibodies has given even higher hopes of diagnostic accuracy for the future. Understanding what monoclonal antibodies are and how they compare to polyclonal sources of antibodies is absolutely essential in order to appreciate their potential. This paper will briefly focus on the production, charactterization and use of monoclonal antibodies and look critically at the future prospects. Specificity and cross-reactivity are key terms which are explained with examples in this paper. It can be anticipated that diagnostic kits in the future will be composite assays using highly specific monoclonals (or oligoclonal cocktails) in conjunction Iwith high-avidity polyclonal sera, thus taking advantage of the best features of both systems. Keywords:
Diagnostics;
microbiology;
monoclonal
antibodies.
Introduction
Hybridoma technology’ which allows the production of ‘tailor-made’ monoclonal antibodies (Mabs) against virtually any antigen, has given medical science an invaluable tool. The availability of practically unlimited amounts of highly specific reagents has led to improved diagnostic performance in microbiology. A profound misconception is that hybridoma antibodies per se are monospecific, i.e. will react exclusively with one specific antigenic determinant only. Unforeseen cross-reactions occur regularly and are to be expected, as our understanding of molecular immunology is still fragmented.
How
to make
monoclonal
antibodies
There are many reviews addressing the technical aspects of generating hybridomas.24 The following text will focus on murine h/labs, because they are the most frequently used for diagnostic purposes. A simplified flow-chart for the generation of Mabs is shown in Figure 1. Bulb/c mice are of choice and individual B-cells are vaccinated with the antigen 0195%6701/91/06A443+08
lO3.00/0
0 1991 The Hospital
443
Infection
Soaety
L. R. Haaheim
Hybridoma
flow
chart
,“,9;“,‘n l:\ HATMedium
PEG as fusing
b
+ 14 days Antibody
Al
.;:‘!,i
agent screening
c\:. Cloning
+ 14 days-
Expansion
In viva-ascitic
New
of cells;
harvest
In vitro-cell
store
1. Flow
screening
antibody
fliud Purify
Figure
antibody
diagram
supernatants
antibodies cells
of the hybridoma
technique.
immortalized by fusing them with malignant B-cells from the same type of inbred animal; in this case myeloma cells from BaZb/c mice. A variety of such myeloma cells are available commercially and those that do not secrete any murine antibodies or immunoglobulin chains themselves, are most useful. Immunization methods vary in relation to dose, route and adjuvant. Traditional schedules of intradermal, subcutaneous, intravenous or intraperitoneal routes with or without adjuvant intrasplenic immunizations may be used with nanogram quantities of antigen bound to nitrocellulose paper5 and free intrasplenic deposition of antigen.‘j
Msonoclonal
antibodies
445
The two main differences between immunization protocols for generating Mabs and polyclonal immune serum are the timing of the booster and the purity of the immunogen used. In the former case one sacrifices the animal within 2-3 days after a booster, i.e. before most B-cells have matured to antibody-secreting plasma cells. This is critical, as it has been found that the mature plasma cells are poor fusion partners. The generation of polyclonal hyperimmune serum requires a ‘pure’ immunogen, otherwise an antibody response to vaccine impurities will reduce the quality of the serum. In the case of Mabs, however, one selects only those antibodies with the properties required, disregarding other irrelevant hybridomas. The most popular way of fusing immune B-lymphocytes and myeloma cells is by use of polyethylene glycol (PEG), but viruses (e.g. Sendai) and have been successfully employed. physical means (e.g. electrofusing) Fusion alone is insufficient a,s unfused myeloma cells will quickly outgrow the delicate hybridomas. Therefore, the myelomas have been made deficient in one of the salvage DNA pathway synthesis enzymes hypoxanthine phosphoribosyltransferase (HGPRT) or thymidine kinase (TK). The addition of aminopterin, a folic acid antagonist inhibits de nouo DNA synthesis so the myeloma cells will be killed. The hybridomas will continue to grow in this :HAT medium (hypoxanthine, aminopterin, thymidin) as they have acquired DNA salvage enzymes from the B-cell fusion partner. Thus, after IO-14 days in HAT medium only hybridoma cells are left which are more robust and can be maintained in medium without aminopterin (HT m’edium). The next step is to screen the many hundreds of wells from the master fusion plates in order to identify those that secrete the antibodies required. Automated analysis is preferable; RIA and ELISA techniques are most useful. ‘Classical’ antibody assays may also be considered (e.g. agglutination, neutralization, fluorescence). Flow cytometric analyses using fluorescein-labelled conjugatles can be used for screening large numbers of hybridoma supernatants for antibodies against certain types/subsets of cells which is the most rapid tech,nique. It is important to establish a good screening assay before starting the actual fusion procedure so tlhat the hybridoma required can be identified precisely. Murine spleen cells harvested for fusion can successfully be frozen in aliquots.3 They can at a later stage be thawed and fused when the screening assay has been improved or if one is looking for Mabs against antigens/determinants that were not planned at the outset. Once the selected hybridoma cells have been cloned and recloned (i.e. isolation of single-cell cultures by seeding at low density) the hybridoma cells can be expanded into larger flasks or injected into mice for ascitic fluid production. Generally, the yield in cell culture is about 10-100 pg ml-‘, whereas ascitic fluid can contain about l-l 5 mg ml-‘.2 Purification of Mabs from cell culture supernatants is usually straightforward, whereas Mabs in
446
L. R. Haaheim
ascitic fluid will be mixed with the antibodies from the ascites-producing animal itself, thus obscuring monoclonality and making purification protocols more elaborate. The last important point to be made is the question of record-keeping. The numerous master fusion plates, each giving rise to various cloned and subcloned cultures, some being at different stages of handling, all bearing cryptic acronyms and identification letters and numbers, require meticulous records. There are examples of local tragedies due to lost tracks of events. It is good practice to concentrate on just a few master wells at first, storing the rest frozen in their original plates for later handling and characterization. Properties
and use of monoclonal
antibodies
A monoclonal antibody forms part of the ‘natural’ immune response of the animal being used. Not surprisingly, therefore, the properties of each and all of the various Mabs generated against a particular antigen will vary with respect to specificity, cross-reactions with other related (or supposedly unrelated) antigens, isotype, affinity and biological properties. What makes the hybridoma technique so useful is that after having generated a particularly good hybridoma cell line it can be used to produce unlimited quantities of identical antibodies. Cells can be stored in liquid nitrogen for an almost unlimited time (or even in a deep-freeze at - 70°C or lower for a couple of years), only later to be revitalized. In contrast, a particularly valuable hyperimmine rabbit serum cannot, strictly speaking, be replaced by immunizing new rabbits. Figure 2 depicts a scenario where the immunogen has three different epitopes exposed on its surface. One of the Mabs generated (anti-C) only recognizes the immunogen and not the two other test antigens. This antibody thus appears to be ‘specific’. Needless to say, by using a larger repertoire of test antigens one could possibly identify other antigens with similar or identical epitopes; this ‘specific’ monoclonal antibody then transgresses into ‘cross-reactivity’. Another Mab (anti-B) may react weakly with antigen 2 and appear to be cross-reactive, as does the Mab against epitope A. Other test antigens could well have given quite opposite results. Needless to say, the very definition of specificity and cross-reactions is dependent on the yardstick used for comparisons, and is therefore not a well-defined concept. Some of these points can be illustrated by data shown in Table I. The post-infection ferret sera distinguish poorly between related A/H3N2 influenza strains A/Norway/25,32,33/80 by haemagglutination-inhibition tests and cannot precisely identify their relationship with the prototype strains A/Bangkok/l and 2/79. In contrast, the three hybridoma antibodies used, clearly identify two of the three Norwegian isolates as being similar to A/Bgk/1/79, whereas the last strain appears to resemble A/Bgk/2/79. This focuses upon the point of useful diagnostic antibodies. If in this case
Monoclonal
Specificity Epitopes lmmunogen
447
antibodies
I
and cross-reoctmns Reactivity AB 3f
C 3+
3-k
Y T
Antigen
Antigen
I
2
Figure 2. Pattern of specificity and cross-reactions of monoclonal antibodies the immunogen and two related antigens, each bearing three epitopes.
the purpose had been to identify influenza strains as say HlNl) and related to recent H3N2 prototype ferret sera would be preferable. In contrast, Mabs alone could easily miss variants. On the other hand,
in reactions
with
being H3N2 (and not, strains, post-infection MC7 or MC225 used for investigating finer
Table I. Use of murine monoclonal awdpolyclonal antibodies in strain characterization of H3N2 injZuenza virus. Haemagglutination-inhibition tests using post-infection ferret sera and murine monoclonal antibodies Virus
strains
Antibodies
A/Texas/l 177 A/Bangkok/l/79 A/Bangkok/2/79
Fl 2560 640 1280
A/Norway/25/80 A/Norway/32/80 A/Norway/33/80
1280 1280 2560
5120 1280
MC7 3200 400
