127
Immunology Today, vol. 5, No. 5, 1984
More pitfalls in the use of monoclonal antibodies SIR,
The article by Haaijman et al. (Immunol. Today 1984, 5, 56-58) on the 'assay specificity' invariably exhibited by monoclonal antibodies (Mabs) is timely. Other undesirable characteristics may be cause for greater concern-particularly demonstrations that specificity and/or the proportion of molecules expressing a given epitope may vary with the assay and protocol employed. More specifically, we have observed the following behaviour patterns with Mabs to h u m a n IgG. 1. Within a panel of Mabs reactive with Fcy epitopes expressed on all molecules of secreted IgG, individual Mabs may detect varying proportions of Ig-positive B lymphocytes or plasma cells when applied in immunocytochemistry. Thus Mabs adopted for such studies must be stringently evaluated before being used for quantitative enumeration of IgGpositive cells. 2. In direct haemagglutination, anantibody was shown to react with IgG of each subclass. However, if IgG3-sensitized cells were used in an haemagglutination inhibition assay, the M a b exhibited specificity for IgG3 subclass proteins only. Presumably the M a b has a higher affinity for the epitope presented by IgG3, probably in aggregated form, when displayed on passively sensitized erythrocytes than for monomer IgG1, IgG2 or IgG4 in solution. 3. Specificity must be established using a very extensive antigen panel. A Mab having non-IgG3 subclass specificity when tested with proteins of Caucasian origin was later shown to be reactive with a major allotypic variant commonly expressed in Mongoloid races. Similarly, an antibody initially considered as a candidate reference reagent for IgG was shown to have specificity for an epitope not expressed on IgG3m(g) molecules. It may be that immunoglobulins, as antigens, are peculiarly heterogeneous and vulnerable to changes in epitope display on denaturation and/or conformational change; however, no doubt similar experiences in other systems will accumulate. Comprehensive evaluation of Mabs to human immunoglobulins alone represents a major task for an individual laboratory and there is an obvious need and benefit to be gained from an international collaborative study. Such a study is in progress for evaluation of Mabs with specificity for h u m a n IgG and each of the four IgG subclasses 1. This study, organized under the auspices
of IUIS and W H O , wilt be completed in 1984 and will allow Mabs to be selected and adopted as reference reagents for application within a given technique using a defined assay protocol. Similar studies of Mabs with specificity for other Ig molecules could be considered and will be an item of discussion within the programme of the colloquium 'Monoclonal Antibodies to H u m a n Immunoglobulins - Specificity and Applications' to be held on 5 September as part of the
6th European Immunology Meeting in Interlaken.
The avidin-biofin system
ciated with biotin and its surfaces may always have biotin associated with it. Thus, if avidin (or anti-biotin antibody) is added to such a system, high levels of non-specific binding will result. We have observed this phenomenon (unpublished results). To obviate this problem we routinely incubate our cell monolayers in serum-free Hanks balanced salt solution for 2 h at 37°C prior to use. This allows for the internalization of the biotinassociated proteins. The problem becomes more acute in vivo as there is no way to eliminate the endogenous biotin short of rendering the animal vitamindeficient. We urge other investigators to use the avidin-biotin system where applicable but they should be aware of these limitations.
SIR,
As investigators who routinely use the avidin-biotin system for ultrastructural localization of antigens, we were keenly interested in the recent review by Meir Wllchek and Edward A. Bayer (ImmunoL Today, 1984, 5, 39-43). We share the authors' enthusiasm for the system. However, for the individual just beginning to use it and for the editor reviewing papers on the technique, we feel two points need to be stressed. First, as pointed out by Wilchek and Bayer, avidin is a basic protein with a pI of 10.5. Thus, at normal physiological pH's, avidin carries a net positive charge and binds electrostatically to negatively charged cell surface proteins. This nonspecific binding is significant (Ref. 1; and our unpublished observations). It can be overcome in two ways: extensive succinoylation of avidin 1.2; or by use of streptavidin, a biotin-binding protein with a binding affinity similar to that of avidin but with a pI of 6.5. Finn et al. 3 showed that streptavidin is as useful as succinoylated avidin. Our second comment concerns the ubiquitousness of biotin. Being a vitamin required by all living cells it is in all tissues and body fluids. In human serum, its concentration is 12.3 ng ml - 1 (3.0 x 1013 molecules ml-1). Similar levels are found in other mammals 4. The carriers for biotin are unknown but are presumed to be proteins in the serum. Because of this, a cell is probably continuously internalizing proteins asso-
Helper T cells and isotype expression
SIR, J. Th~ze, in his concise summary of evidence bearing upon the existence of isotype-switching T cells ( ImmunoI. Today 1984, 5, 61) presented, broadly: (a) indirect evidence that both 'immunoglobulin-dependent' T cells and T helper (TH) cells were required for the in-vivo
R. JEFFERIS
N. R. LING Department of Immunology, The Medical School, University of Birmingham, Birmingham, B15 2 TJ UK. Reference
1 Jefferis, R. and Skvaril, F. (1982) Immunol. Today3, 256
RANDAL E. MORRIS Department of Anatomy and Cell Biology GATHARINE B. SAELINGER Department ofMicrobiologyandMolecularGenetics, University of Cincinnati College of Medicine, Cincinnati, OH 45267-0524, USA References
1 Finn, G. M., Titus, G., Montibeller,J. A. and Hofmann, K. (1980)J. Biol. Chem. 255 242-246 2 Morris, R. E. and Saelinger, C. B. (1984) J. Histochem. Cytochem.32, 124-128 3 Finn, F. M., lwata, N., Titus, G. and Hofmann, K. (1981)Hoppe-Seyler'sZ Physiol. Chem. 362, $679-684 4 Kutsky, R. J. (1981) Handbook of Vitamins, Minerals, and Hormones, 2nd edn, Van Nostrand Reinhold Co., New York generation of IgG responses; (b) evidence that cloned T H were unable to cooperate in the production of IgG responses in vivo with unprimed B cells; and (c) (at variance with a and b) evidence that, together with primed B cells, and to a lesser extent with unprimed B cells, cloned T H were indeed competent to help the in-vitro production of manifold isotypes. May I add a further catego W to his ex-
Immunology Today, vol. 5, No. 5, 1984
More pitfalls in the use of monoclonal antibodies SIR,
The article by Haaijman et al. (Immunol. Today 1984, 5, 56-58) on the 'assay specificity' invariably exhibited by monoclonal antibodies (Mabs) is timely. Other undesirable characteristics may be cause for greater concern-particularly demonstrations that specificity and/or the proportion of molecules expressing a given epitope may vary with the assay and protocol employed. More specifically, we have observed the following behaviour patterns with Mabs to h u m a n IgG. 1. Within a panel of Mabs reactive with Fcy epitopes expressed on all molecules of secreted IgG, individual Mabs may detect varying proportions of Ig-positive B lymphocytes or plasma cells when applied in immunocytochemistry. Thus Mabs adopted for such studies must be stringently evaluated before being used for quantitative enumeration of IgGpositive cells. 2. In direct haemagglutination, anantibody was shown to react with IgG of each subclass. However, if IgG3-sensitized cells were used in an haemagglutination inhibition assay, the M a b exhibited specificity for IgG3 subclass proteins only. Presumably the M a b has a higher affinity for the epitope presented by IgG3, probably in aggregated form, when displayed on passively sensitized erythrocytes than for monomer IgG1, IgG2 or IgG4 in solution. 3. Specificity must be established using a very extensive antigen panel. A Mab having non-IgG3 subclass specificity when tested with proteins of Caucasian origin was later shown to be reactive with a major allotypic variant commonly expressed in Mongoloid races. Similarly, an antibody initially considered as a candidate reference reagent for IgG was shown to have specificity for an epitope not expressed on IgG3m(g) molecules. It may be that immunoglobulins, as antigens, are peculiarly heterogeneous and vulnerable to changes in epitope display on denaturation and/or conformational change; however, no doubt similar experiences in other systems will accumulate. Comprehensive evaluation of Mabs to human immunoglobulins alone represents a major task for an individual laboratory and there is an obvious need and benefit to be gained from an international collaborative study. Such a study is in progress for evaluation of Mabs with specificity for h u m a n IgG and each of the four IgG subclasses 1. This study, organized under the auspices
of IUIS and W H O , wilt be completed in 1984 and will allow Mabs to be selected and adopted as reference reagents for application within a given technique using a defined assay protocol. Similar studies of Mabs with specificity for other Ig molecules could be considered and will be an item of discussion within the programme of the colloquium 'Monoclonal Antibodies to H u m a n Immunoglobulins - Specificity and Applications' to be held on 5 September as part of the
6th European Immunology Meeting in Interlaken.
The avidin-biofin system
ciated with biotin and its surfaces may always have biotin associated with it. Thus, if avidin (or anti-biotin antibody) is added to such a system, high levels of non-specific binding will result. We have observed this phenomenon (unpublished results). To obviate this problem we routinely incubate our cell monolayers in serum-free Hanks balanced salt solution for 2 h at 37°C prior to use. This allows for the internalization of the biotinassociated proteins. The problem becomes more acute in vivo as there is no way to eliminate the endogenous biotin short of rendering the animal vitamindeficient. We urge other investigators to use the avidin-biotin system where applicable but they should be aware of these limitations.
SIR,
As investigators who routinely use the avidin-biotin system for ultrastructural localization of antigens, we were keenly interested in the recent review by Meir Wllchek and Edward A. Bayer (ImmunoL Today, 1984, 5, 39-43). We share the authors' enthusiasm for the system. However, for the individual just beginning to use it and for the editor reviewing papers on the technique, we feel two points need to be stressed. First, as pointed out by Wilchek and Bayer, avidin is a basic protein with a pI of 10.5. Thus, at normal physiological pH's, avidin carries a net positive charge and binds electrostatically to negatively charged cell surface proteins. This nonspecific binding is significant (Ref. 1; and our unpublished observations). It can be overcome in two ways: extensive succinoylation of avidin 1.2; or by use of streptavidin, a biotin-binding protein with a binding affinity similar to that of avidin but with a pI of 6.5. Finn et al. 3 showed that streptavidin is as useful as succinoylated avidin. Our second comment concerns the ubiquitousness of biotin. Being a vitamin required by all living cells it is in all tissues and body fluids. In human serum, its concentration is 12.3 ng ml - 1 (3.0 x 1013 molecules ml-1). Similar levels are found in other mammals 4. The carriers for biotin are unknown but are presumed to be proteins in the serum. Because of this, a cell is probably continuously internalizing proteins asso-
Helper T cells and isotype expression
SIR, J. Th~ze, in his concise summary of evidence bearing upon the existence of isotype-switching T cells ( ImmunoI. Today 1984, 5, 61) presented, broadly: (a) indirect evidence that both 'immunoglobulin-dependent' T cells and T helper (TH) cells were required for the in-vivo
R. JEFFERIS
N. R. LING Department of Immunology, The Medical School, University of Birmingham, Birmingham, B15 2 TJ UK. Reference
1 Jefferis, R. and Skvaril, F. (1982) Immunol. Today3, 256
RANDAL E. MORRIS Department of Anatomy and Cell Biology GATHARINE B. SAELINGER Department ofMicrobiologyandMolecularGenetics, University of Cincinnati College of Medicine, Cincinnati, OH 45267-0524, USA References
1 Finn, G. M., Titus, G., Montibeller,J. A. and Hofmann, K. (1980)J. Biol. Chem. 255 242-246 2 Morris, R. E. and Saelinger, C. B. (1984) J. Histochem. Cytochem.32, 124-128 3 Finn, F. M., lwata, N., Titus, G. and Hofmann, K. (1981)Hoppe-Seyler'sZ Physiol. Chem. 362, $679-684 4 Kutsky, R. J. (1981) Handbook of Vitamins, Minerals, and Hormones, 2nd edn, Van Nostrand Reinhold Co., New York generation of IgG responses; (b) evidence that cloned T H were unable to cooperate in the production of IgG responses in vivo with unprimed B cells; and (c) (at variance with a and b) evidence that, together with primed B cells, and to a lesser extent with unprimed B cells, cloned T H were indeed competent to help the in-vitro production of manifold isotypes. May I add a further catego W to his ex-
