Variation In Antigen Expression Research Articles

Convergent Potency of Internalized Gelonin Immunotoxins across Varied Cell Lines, Antigens, and Targeting Moieties

Gelonin-based immunotoxins vary widely in their cytotoxic potency as a function of antigen density, target cell internalization and trafficking kinetics, and conjugate properties. We have synthesized novel gelonin immunotoxins using two different binding scaffold types (single-chain antibody variable fragments and fibronectin domains) targeting two different tumor antigens (carcinoembryonic antigen and EGF receptor). Constructs were characterized using an antigen-negative cell line (HT-1080), cell lines positive for each antigen (HT-1080(CEA) for carcinoembryonic antigen and A431 for EGF receptor), and a cell line positive for both antigens (HT-29). Immunotoxins exhibited K(d) values between 8 and 15 nm and showed 20-2000-fold enhanced cytotoxicity compared with gelonin (IC(50) ∼ 0.25-30 nM versus 500 nM). Using quantitative fluorescence flow cytometry, we measured internalization of gelonin (via pinocytosis) and gelonin-based immunotoxins (via antigen-dependent, receptor-mediated endocytosis). Results were matched with cytotoxicity measurements made at equivalent concentration and exposures. Unexpectedly, when matched internalization and cytotoxicity data were combined, a conserved internalized cytotoxicity curve was generated that was common across experimental conditions. Considerable variations in antigen expression, trafficking kinetics, extracellular immunotoxin concentration, and exposure time were all found to collapse to a single potency curve on the basis of internalized immunotoxin. Fifty percent cytotoxicity occurred when ∼ 5 × 10(6) toxin molecules were internalized regardless of the mechanism of uptake. Cytotoxicity observed at a threshold internalization was consistent with the hypothesis that endosomal escape is a common, highly inefficient, rate-limiting step following internalization by any means tested. Methods designed to enhance endosomal escape might be utilized to improve the potency of gelonin-based immunotoxins.

PAX-5: A Valuable Immunohistochemical Marker in the Differential Diagnosis of Lymphoid Neoplasms

Undifferentiated tumors and hematolymphoid neoplasms can be diagnostically challenging due to potential overlap of morphologic features and variant antigen expression. PAX-5, a transcription factor expressed throughout B-cell maturation, is detected in most B-cell neoplasms including those that lack expression of mature B-cell markers, such as classical Hodgkin lymphoma (cHL), B-lymphoblastic leukemia and B-cell lymphomas following rituximab therapy. The lack of PAX-5 expression in most CD30-positive non-hematopoietic malignancies (embryonal carcinoma and seminoma) and T-cell lymphomas, such as anaplastic large cell lymphoma (ALCL), suggests that the absence of PAX-5 may be used to confirm non-B-cell lineage. The goal of this study was to retrospectively assess PAX-5 immunoreactivity in diagnostic samples of hematolymphoid and other non-hematopoietic malignancies. Diagnostic lymph node, decalcified core bone marrow biopsies and tissue sections from 111 archived paraffin-embedded tissue blocks and a tissue lymphoma microarray were immunostained using a monoclonal antibody to PAX-5. The corresponding hematoxylin and eosin stained tissue sections and additional immunostains were simultaneously evaluated. PAX-5 immunoreactivity in neoplastic cells was scored as positive or negative. This study was exempted by the Institutional Review Board for Human Research. Nuclear PAX-5 immunoreactivity was detected in 88% (36/41) of Hodgkin lymphoma, all cases of diffuse large B-cell lymphoma (n=72), small B-cell lymphomas (n=5), B-lymphoblastic leukemia/lymphoma and mixed phenotype acute leukemia with B-cell lineage (n=5). PAX-5 was not detected in ALCL (n=22), T-cell lymphoblastic leukemia/lymphoma, mixed phenotype acute leukemia with T-cell lineage (n=5), acute myeloid leukemia (n=4), carcinoid tumors with typical morphology (n=5), melanoma (n=3), and undifferentiated/metastatic tumors (n=8). Non-neoplastic bone marrow sections showed scattered nuclear staining in small B-cell lymphocytes/hematogones. The detection of PAX-5 immunoreactivity resulted in the reclassification of two cases of ALCL to cHL. Overall, our results demonstrate that including PAX-5 in a panel with other immunomarkers helps establish B-cell lineage and increases diagnostic yield.

Selective Enrichment Media Affect the Antibody-Based Detection of Stress-Exposed Listeria monocytogenes due to Differential Expression of Antibody-Reactive Antigens Identified by Protein Sequencing

Selective enrichment broths are frequently used to recover stressed Listeria cells to detectable levels, but the ability of antibodies to detect these cells from various commonly used enrichment media is unknown. In this study, a polyclonal (PAb) and monoclonal (MAb) antibody were used to examine the variation in antigen expression on healthy or stress-recovered Listeria monocytogenes cells grown in brain heart infusion broth, buffered Listeria enrichment broth (BLEB), Listeria repair broth (LRB), University of Vermont medium (UVM), and Fraser broth (FB) for immunodetection. Indirect enzyme-linked immunosorbent assay (ELISA) data showed that L. monocytogenes subjected to stresses (acid, cold, heat, and salt) and then grown in BLEB gave the highest reaction with the anti-Listeria PAb while those grown in LRB gave the highest reaction with the MAb C11E9. Cells grown in UVM and FB gave poor ELISA values with both antibodies. Western blotting with PAb revealed differential expression of surface proteins of 62, 58, 50, 43, and 30 kDa on L. monocytogenes cells, with most proteins displaying elevated expression in BLEB and LRB but reduced or no expression in UVM or FB. Similar differential expressions were noticed for C11E9. PAb-reactive proteins were identified as putative LPXTG-motif cell-wall anchor-domain protein (62 kDa; lmo0610), flavocytochrome C fumarate reductase chain A homolog protein (58 kDa; lmo0355), enolase (50 kDa; lmo2455), glyceraldehyde 3-phosphate dehydrogenase (43 kDa; lmo2459), and hypothetical phospho-sugar binding protein (30 kDa; lmo0041), respectively, and the MAb-reactive 66-kDa protein was confirmed to be N-acetylmuramidase (lmo2691). In conclusion, BLEB and LRB favorably supported increased expression of antigens and proved to be superior to UVM and FB for immunodetection of stressed L. monocytogenes cells.

Variable var transition rates underlie antigenic variation in malaria.

Plasmodium falciparum erythrocyte membrane protein 1 (PfEMP1) is expressed on the surface of infected erythrocytes where it plays a central role in both infected erythrocytes cytoadhesion and immune evasion. Switches in clonal expression of PfEMP1 result in antigenic variation that facilitates long-term chronic infection of the host. The var gene family encodes PfEMP1 variants, with transcriptional switching between different var variants providing the molecular basis for antigenic variation. Despite the importance of var transcriptional switching in the evasion of the immune response, little is known about the way in which this process is regulated. Here we report the measurement of transition on and off rates for a series of var gene variants. We find (i) that on and off rates for a given variant are dissimilar, (ii) that these rates vary dramatically among different variants, and (iii) that in isogenic clones expressing the same var gene, both on and off rates are constant and appear to be an intrinsic property of that particular gene. These data would suggest that the information that determines the probability of the activation or silencing of var genes is present in their surrounding DNA. Furthermore, some transitions appear to be disallowed depending on the recent variant antigen expression history of the parasite clone. These findings have important implications for both the underlying molecular mechanisms of antigenic variation and the processes that promote chronicity of infection in vivo.

Receptor for the globular heads of C1q (gC1q-R, p33, hyaluronan-binding protein) is preferentially expressed by adenocarcinoma cells.

Combinatorial Ig libraries with phage display allow in vitro generation of human Ig fragments without the need to maintain hybridomas in ongoing cell culture or to select circulating Ig from human serum. Identifying tumor-associated antigens on the surface of intact tumor cells, as opposed to purified proteins, presents a challenge due to the difficulty of preserving complex 3-D epitopic sites on the cell surface, the variable expression of antigens on different malignant cell types and the stereotactic interference of closely associated proteins on the intact membrane surface limiting accessibility to antigenic sites. A combinatorial Ig library of 10(10) clones was generated from the cDNA of PBMCs derived from patients with breast adenocarcinoma. Following subtractive panning, the library was enriched for Ig (Fab fragment) binding to intact adenocarcinoma cells and the resultant Fabs were screened against a cDNA expression library, itself generated from breast cancer cells. Using this approach, we isolated clones from the cDNA library expressing gC1q-R, a glycoprotein comprising the major structure of C1, the first component of the complement system. gC1q-R is a 33 kDa glycoprotein expressed not only on the cell surface but also intracellularly, with motifs that target it to mitochondria and complete homology with HABP and human HeLa cell protein p32, which is copurified with pre-mRNA SF2. Sequencing of the gene encoding tumor-associated gC1q-R did not reveal any consistent tumor-specific mutations. However, histochemical staining with anti-gC1q-R MAb demonstrated marked differential expression of gC1q-R in thyroid, colon, pancreatic, gastric, esophageal and lung adenocarcinomas compared to their nonmalignant histologic counterparts. In contrast, differential expression was not seen in endometrial, renal and prostate carcinomas. Despite high expression in breast carcinoma, gC1q-R was also expressed in nonmalignant breast tissue. Although the precise relation of gC1q-R to carcinogenesis remains unclear, our finding of tumor overexpression and the known multivalent binding of gC1q-R to not only C1q itself but also a variety of circulating plasma proteins as well as its involvement in cell-to-cell interactions suggest that gC1q-R may have a role in tumor metastases and potentially serve in molecule-specific targeting of malignant cells.

Variant antigen expression in malaria infections: posttranscriptional gene silencing, virulence and severe pathology.

Variant antigen expression in malaria infections: posttranscriptional gene silencing, virulence and severe pathology.

A perspective on clonal phenotypic (antigenic) variation in protozoan parasites.

Intra-clonal phenotypic (antigenic) variation is used by many pathogens to evade the consequences of immune-mediated killing by mammalian hosts. In this substantially theoretical article, I emphasise that antigenic variation (sensu stricto) involves no change in genotype; its importance as a mechanism for promoting pathogen transmission and its polyphyletic origin. From a functional perspective, antigenic variation is constrained by the requirement to meet five conditions. These are: capability to express several antigens against which functional immunity predominates; capability to interact with the environment; mutually exclusive expression of variable antigens in each cell within an infection; mutually exclusive expression in the within-host pathogen population and the capability for population growth within a host. Meeting these conditions leads to chronicity of infection and high rates of hierarchical and reversible switching of expression between variable antigens. The organisation of hierarchical expression is discussed in some detail.

Bivariate Flow Cytometric Analysis of Proliferation Associated Nuclear Antigens in Solid Tumor Cells

Bivariate flow cytometry was used for analysis of DNA content and proliferation associated nuclear antigens (PCNA, Ki-67, p105, p145, p120) in the T24 cell line, in 34 fresh solid tumors, and in 4 normal tissue specimens. Antigens were analyzed in both fixed and unfixed cells. In solid tumors and T24 cells, Larsen's washless method yielded the highest expression of PCNA and Ki-67, whereas p105, p145, and p120 were best demonstrated by other methods. Antigen expression was generally lower in normal tissues than in solid tumor (P < 0.05), but there was great variation in antigen expression among all tumor types. Antigen expression did not correlate with DNA ploidy or with percent S phase. The data suggest that cell fixation methods must be optimized for each antigen studied. Our results demonstrate the feasibility and methodology of dual flow cytometry on minute clinical specimens by the washless method. (The J Histotechnol 18:17, 1995)

Variation in the deposition of the antibodies at different anatomical sites in linear IgA disease of adults and chronic bullous disease of childhood.

The diagnosis of linear IgA disease of adults (LAD) and chronic bullous disease of childhood (CBDC) relies upon finding a linear band of IgA at the basement membrane zone on direct immunofluorescence. This study examines the regional variation in antigen expression in the skin of affected individuals. Direct immunofluorescence was performed on biopsies from four different sites in 17 patients with these diseases. In two patients a biopsy from the volar surface of the forearm was negative, but other sites were positive; in the remaining patients there was no variation in antibody expression with site. It is therefore recommended that, if a single diagnostic biopsy is to be taken, the volar surface of the forearm is avoided.

The relationship of variable antigen expression and population growth rates in Trypanosoma brucei.

The relationship between variable antigen type (VAT) expression and trypanosome growth rates was investigated. Growth rates in mice were compared between pairs of cloned trypanosome populations, each of which homogeneously expressed a different VAT. All three pairwise combinations of GUTats (Glasgow University Trypanozoon antigen types) 7.3, 7.4 and 7.5 were analysed twice and all three combinations of GUTats 8.2, 8.3 and 8.4 were compared once. The lines expressing different VATs were of the same passage history within each group. In a sensitive assay of relative growth, no significant differences were found in four of six experiments using GUTats 7.3, 7.4 and 7.5 or in one of three experiments using GUTats 8.2, 8.3 and 8.4. In the experiments in which differences were observed, the data were analysed further to compare the population doubling times of lines. These times differed by less than 10% in all cases. We conclude that variable antigen expression exerts a small (possibly negligible) effect on rates of trypanosome population growth.

EXPRESSION OF ABH BLOOD GROUP ANTIGENS IN HUMAN HEART TISSUE AND ITS RELEVANCE TO CARDIAC TRANSPLANTATION

Blood group antigen expression was assessed in cardiac tissue taken during heart transplantation. The samples were from ten donors and nine recipients, of whom seven were blood group A, three were group B, two were group AB, and seven were group O. Cryostat sections were studied by immunofluorescence microscopy. Blood group antigens were confined to the mesothelial cells on the surface of the epicardium and the cardiovascular endothelium where they were consistently demonstrated according to erythrocyte blood group using a rabbit anti-A serum, monoclonal antibodies against A and B antigens, and the H antigen-specific lectin Ulex europaeus. The cardiac muscle itself was negative for blood group antigen. The results indicate that there is some variation in antigen expression between individuals, which may explain why transplants have occasionally been successful in breaching the blood group ABO barrier.

Analysis of antigen expression at multiple tumor sites in epithelial ovarian cancer

The question of whether the antigenic phenotype of human epithelial ovarian cancer varies in a given patient between the primary tumor and metastatic sites or among metastatic sites themselves is an important issue in planning potential therapeutic strategies for ovarian cancer. We have obtained tumor specimens from at least two separate sites during operations on 12 patients with epithelial ovarian cancer, and we have typed these specimens with a group of 18 monoclonal antibodies that react with cell-surface glycoprotein and carbohydrate antigens, including blood group antigens. Antibodies with relative specificity for malignant cells as well as those that detect more widely distributed epithelial antigens were used. A total of 31 specimens from 12 patients with advanced adenocarcinoma (8 serous, 3 undifferentiated, 1 endometrioid) of the ovary were studied, including fresh ascites cells in two patients. Frozen sections of tumor specimens were stained with the antibodies by the indirect immunoperoxidase technique and graded semiquantitatively. Little difference was seen in antigenic expression of tumors that were obtained from various sites in the same patient for either the epithelial cell markers or blood group markers. Intratumoral antigenic heterogeneity was seen, but this was generally quite consistent within a given patient's specimens. As anticipated, variations in antigen expression were seen among specimens from different patients. The antigenic phenotype of the tumor specimens in a given patient, as determined immunohistochemically by our group of antibodies, showed only minor variation among primary and metastatic sites.

Antigenic Heterogeneity in Human Ovarian Cancer

WELCH, WILLIAM R.; NILOFF, JONATHAN M.; ANDERSON, DEBORAH; BATTAILE, ANN; EMERY, SHERRY; KNAPP, ROBERT C.; BAST, ROBERT C. JR. Author Information

Antigenic heterogeneity in human ovarian cancer

Primary and metastatic tumor tissues from 21 patients with ovarian epithelial cancer were studied with a panel of 8 monoclonal antibodies. Primary tumors reacted with 1 to 7 antibodies (mean, 3.5). Heterogeneity was observed even within histologic subtypes. Comparison of metastases (including ascites) with their respective primaries revealed differences in antigen profile in each of 10 cases studied. In one patient variable antigen expression was observed in five ascites samples collected over a 12-month period. These observations of antigenic heterogeneity and modulation with respect to site and time suggest that single monoclonal antibody immunotherapy would not be appropriate for all patients or even for a single patient over time.

Monoclonal antibodies in neuro-oncology

Many studies have suggested the possible existence of tumor-associated antigens in brain gliomas. Strong evidence for the existence of such cell determinants was provided by recent investigations using hybridoma technology. The possibility of obtaining monoclonal antibodies (MAbs) against glioma-associated antigens should help to allow their identification, purification, and characterization. Utilizing MAbs as reagents of predefined specificity, a number of central and peripheral nervous system antigens could be detected. The molecules recognized by MAbs in glioma cells can be subdivided into four categories: [1] biochemical defined proteins, [2] specificities shared by nervous system-lymphoid cells, [3] oncoembryonic-oncofetal determinants, and [4] tumor-restricted antigens. Of greater significance is the heterogeneity of antigen expression among various individual glioma cells observed in frozen sections of tumor biopsies. Using a panel of MAbs, the phenotypic heterogeneity, i.e., the variation in antigen expression can be documented within and among malignant gliomas and cell lines derived from them. In spite of this the characteristic pattern of antibody binding to brain tumors makes MAbs the potentially best reagents for immuno-histochemical application in surgical neuropathology. Moreover, immuno-cytological screening of tumor cells in the cerebrospinal fluid has also proved to be valuable. The localization of radio-labelled MAbs in experimental and human gliomas growing subcutaneously and intracranially in athymic nude mice were explored by radioscintigraphy and autoradiography. Imaging experiments with 131I-labelled MAbs recognizing epitopes on the glioma cell surface showed high levels of specific activity in xenografts. Preliminary data indicate that administration of 131I-MAbs as well as drug conjugates (daunomycin-MAbs) causes a depression of glioma cell proliferation in vitro as well as delayed tumor growth and thus prolonged survival time of tumor-bearing mice. The mechanisms of antibody delivery and transport of "immunotoxins" from the vascular compartment to intracerebral tumor tissue are presently a subject of discussion. The complexity of this area necessitates comprehensive experimental work in order to define the factors involved in the delivery of MAbs to brain to tumor tissue and thus optimize the rate of blood-to-tumor transport. Current investigations have shown that it is possible to image malignant human gliomas using radio-labelled antibodies. The next step will be to attain target immunotherapy. The use of MAbs as carrier molecules for clinical applications might soon be possible.

Histogenesis of Extramammary and Mammary Paget Cells

To define the histogenesis of the Paget cell and possibly identify differences in cells from the two sites, six vulvar and 23 mammary specimens from Paget's disease lesions were studied for immunocytochemical antigens. All vulvar and 21 (91%) mammary lesions were strongly reactive for glandular cytokeratin. All lesions showed immunopositive Paget cells with epithelial membrane antigen (EMA). An apocrine antigen, gross cystic disease fluid protein (GCDFP-15), decorated 66.5% and 56.5% of extramammary and mammary lesions, respectively. All vulvar Paget cells stained for carcinoembryonic antigen (CEA), a frequency significantly greater than the 35% in mammary lesions (p = 0.02). However, CEA is expressed by both eccrine and apocrine sweat glands and their tumors. Vulvar Paget cells were negative for lysozyme, casein, lactalbumin, and S100 protein, compared with 9%, 4%, 4%, and 26% in nipple lesions. S100 protein expression is similar to that in mammary ductal carcinoma (32%). The glandular origin of both extramammary and mammary Paget cells is indicated by the presence of glandular cytokeratin, EMA, and CEA. Approximately 60% of all cases in both sites showed evidence of apocrine derivation (GCDFP-15 positivity). Variable antigen expression suggests possible malignant transformation of pluripotent germinative cells able to differentiate in an apocrine or an eccrine direction, or in both.

Heterogeneity of antigen expression in benign and malignant breast and ovarian epithelial cells

Expression of II distinct antigen families has been studied in normal and malignant epithelial cells from breast and ovary by avidin-biotin immunoperoxidase staining of frozen tissue sections. Each of the antigens was expressed in a fraction of the normal and malignant breast tissues from different individuals. Among breast and ovarian cancers, a similar fraction of tumors expressed each of the determinants. An epitope on a 42-kDa glycoprotein was expressed significantly more often by ovarian carcinomas than by benign ovarian epithelium. Several determinants on 66-, 240-kDa or high-molecular-weight proteins or glycoproteins were expressed significantly more often on benign breast epithelium than on normal ovarian epithelium. Substantial heterogeneity in antigen expression was observed among different tumor specimens. Each of 14 epithelial ovarian cancers expressed a distinctive combination of antigens. Among 18 breast cancers, 16 distinct antigenic phenotypes were observed. Interestingly, comparable heterogeneity was observed in the expression of epitopes on benign breast and ovarian epithelium, compatible with the possibility that the monoclonal reagents recognized polymorphic determinants. Substantial variations in antigen expression were also observed among cells within each neoplasm. However, when antibodies against 5 different determinants were used in combination, a majority of cells could be stained intensely in all 13 breast carcinomas tested and more than 95% of tumor cells could be stained in 10 of the 13.

Chapter 17 Detection of small cell carcinoma in bone marrow aspirates using monoclonal antibodies and mixtures of monoclonal antibodies

Abstract Four monoclonal antibodies and 4 mixtures of these were used to stain 8 control bone marrows, 8 bone marrows known to be infiltrated with SCLC and 20 cytologically negative bone marrows from patients with SCLC. A small population of haemopoietic cells stained with each of the reagents. Clumps of stained cells were seen in all cytologically positive bone marrows. Varation of antigen expression was seen but this was largely overcome by the use of mixtures of monoclonal antibodies. Clumps of stained cells were detected in 4 of 20 (20%) cytologically negative bone marrows. Monoclonal antibodies are more sensitive than routine cytological analysis at detecting SCLC infiltrating bone marrow. This method is limited both by the presence of stained single cells in the normal haemopoietic population and by variation of antigen expression by SCLC, our monoclonal antibodies and 4 mixtures of these were used to stain 8 control bone marrows, 8 bone marrows known to be infiltrated with SCLC and 20 cytologically negative bone marrows from patients with SCLC. A small population of haemopoietic cells stained with each of the reagents. Clumps of stained cells were seen in all cytologically positive bone marrows. Variation of antigen expression was seen but this was largely overcome by the use of mixtures of monoclonal antibodies. Clumps of stained cells were detected in 4 of 20 (20%) cytologically negative bone marrows. Monoclonal antibodies are more sensitive than routine cytological analysis at detecting SCLC infiltrating bone marrow. This method is limited both by the presence of stained single cells in the normal haemopoietic population and by variation of antigen expression by SCLC.

A monoclonal antibody CIBr17 recognizes a myoepithelium-specific antigen in human mammary gland.

A murine monoclonal antibody (MAb) CIBr17, raised against the human breast carcinoma cell line PMC42, reacts specifically with myoepithelial cells in normal human breast. This IgGl antibody recognizes a approximately 110kDa glycoprotein that is expressed on the cell surface and junctional membranes of PMC42 monolayer cultures. The CIBr17 antigen is present in two major glycosylated forms with approximate pls of 5.2 and 5.5 respectively in PMC42 cells. The tissue specificity of CIBr17 was assessed on frozen sections of PLP-fixed tissues by means of a 4-layer immunoperoxidase technique. CIBr17 has reacted with a variety of epithelium-derived tissues and some smooth muscle cells. Within many epithelial tissues, CIBr17 has demonstrated specific staining of particular epithelial cell types. Within normal breast and most benign breast lesions, antibody CIBr17 stained only myoepithelial cells. No staining of luminal epithelium, basement membranes or stromal elements was observed. In sclerosing adenosis, CIBr17 stained areas of pronounced myoepithelial differentiation, while in duct epitheliosis variable staining of proliferating cells was observed. In breast carcinomas, CIBr17 demonstrated variable antigen expression. In most tumors, CIBr17 either did not stain any tumor cells or stained only a small number of tumor cells spread randomly throughout the tumor. In several ductal carcinomas, however, CIBr17 stained the majority of tumor cells.

Heterogeneous expression of cell-surface antigens in normal epithelia and their tumours, revealed by monoclonal antibodies.

Most monoclonal antibodies that have been raised to human epithelial tumours bind to only some of the cells in a tumour, showing that tumour cells are very heterogeneous in their expression of antigens. Normal epithelia show the same heterogeneity of antigen expression, as also do cell lines and clones of epithelial cells in culture. It is not related to the mitotic cell cycle. Many, probably most of the antigenic determinants to which the antibodies bind are carbohydrate structures. It is not clear whether variations in antigen expression reflect variations in the differentiated state of the cells or merely variations in the carbohydrate structures on otherwise identical cells, nor is ir clear whether antibodies could be made that bind to all tumour cells by avoiding antibodies to carbohydrate structures. The normal and apparently reversible nature of this heterogeneity of antigen expression conflicts with conventional views that heterogeneity among cells of a tumour is due to permanent genetic change. The heterogeneity within normal clones suggests that cloning is not an adequate way to study heterogeneity in tumour cells. The implications of heterogeneous expression of antigens within tumours for therapeutic and diagnostic application of antibodies are discussed.