Cloning of the complete gene for carcinoembryonic antigen: analysis of its promoter indicates a region conveying cell type-specific expression.

Carcinoembryonic antigen (CEA) is a widely used tumor marker, especially in the surveillance of colonic cancer patients. Although CEA is also present in some normal tissues, it is apparently expressed at higher levels in tumorous tissues than in corresponding normal tissues. As a first step toward analyzing the regulation of expression of CEA at the transcriptional level, we have isolated and characterized a cosmid clone (cosCEA1), which contains the entire coding region of the CEA gene. A close correlation exists between the exon and deduced immunoglobulin-like domain borders. We have determined a cluster of transcriptional starts for CEA and the closely related nonspecific cross-reacting antigen (NCA) gene and have sequenced their putative promoters. Regions of sequence homology are found as far as approximately 500 nucleotides upstream from the translational starts of these genes, but farther upstream they diverge completely. In both cases we were unable to find classic TATA or CAAT boxes at their expected positions. To characterize the CEA and NCA promoters, we carried out transient transfection assays with promoter-indicator gene constructs in the CEA-producing adenocarcinoma cell line SW403, as well as in nonproducing HeLa cells. A CEA gene promoter construct, containing approximately 400 nucleotides upstream from the translational start, showed nine times higher activity in the SW403 than in the HeLa cell line. This indicates that cis-acting sequences which convey cell type-specific expression of the CEA gene are contained within this region.

Demonstration of antibodies in patients' sera, directed against nonspecific cross-reacting antigen.

Nonspecific cross-reacting antigen (NCA), which strongly cross-reacts with carcinoembryonic antigen, was demonstrated to be an autoantigen. Antibodies directed against NCA were shown in different groups of patients, but high titers (greater than 1/64) were found only in cancer patients. A correlation between tumor mass, antigen load, and antibody titer apparently existed. Sera obtained from patients preoperatively and postoperatively differed significantly (P greater than 0.01) in the sense that titers were high only in sera sampled after the patients were treated. Nevertheless, the formation of these antibodies cannot be considered a cancer-specific phenomenon because of their existence also in patients with nonmalignant diseases.

Determination of levels of carcinoembryonic antigen and nonspecific cross-reacting antigen in the sera of neonatal infants: brief communication.

Seventy-eight sera from neonatal infants, born at full term or prematurely, were studied for their carcinoembryonic antigen (CEA) and nonspecific cross-reacting antigen (NCA) concentrations, which were compared to the normal adult concentrations. The levels of CEA in the sera were significantly higher in newborns than in adults: 9.05 ng CEA/ml in newborns as compared to 2.5 ng CEA/ml in adults (P=0.001). The levels of NCA in the sera were also higher in newborns: 164 ng NCA/ml in newborns as compared to 130 ng NCA/ml in adults. This difference in NCA levels was not significant, although 80% of the newborns had increased values (greater than 130 ng/ml). Whether the infant was born at full term or prematurely and whether the infant was a boy or girl had no statistically significant influence on the concentration of the CEA and the NCA in the infant.

Immunohistology of the antigenic pattern of a continuous cell line from a human colon tumor.

The antigenic surface pattern of a continuous cell line (HT29) derived from a human primary carcinoma of the colon was studied by the immunofluorescence technique. Monovalent and polyvalent immune sera were used. The cells of this long-term culture kept the ability to synthesize the three principal colon tumor antigens: carcinoembryonic and nonspecific cross-reacting antigens, and the membrane-associated tissular autoantigen. On the HT29 cells, which still carry the original blood group of the tumor donor, no receptors for human Ig's were detected.

Diagnostic significance of tumor markers.

Progress in the immunological detection of malignant tumors has been delayed by two major obstacles: (a) the lack of true cancer specificity of the known tumor-marker substances; and (b) the lack of a close relationship between the concentration and/or first appearance of the marker antigens and the malignant disease, especially at its early stage. An additional, important, and complicating factor implicating the specificity of the reactions of the test substances is the existence of immunological cross-reactions. Until recently, only two approaches have been pursued to resolve this difficulty of cross-reactivity, both concerned with the antigen. Now there is a third way, namely, the production of monoclonal antibodies. The poor correlation between the stage of the neoplastic disease and the level of circulating tumor markers, e.g., carcinoembryonic antigen, may be explained on a histological basis, or by inherent differences in the carcinoembryonic antigen molecule of different tumor sources. Few studies are available dealing with these problems.

Chromosomal localization of the carcinoembryonic antigen gene family and differential expression in various tumors.

Carcinoembryonic antigen (CEA) is a glycoprotein which is important as a tumor marker for a number of human cancers. It is a member of a gene family comprising about 10 closely related genes. In order to characterize mRNAs transcribed from individual genes we have identified by DNA and RNA hybridization experiments, gene-specific sequences from the 3' noncoding regions of CEA, and of nonspecific cross-reacting antigen (NCA) mRNAs, which have been recently cloned. With these probes, CEA mRNAs with lengths of 3.5 and 3.0 kilobases and an NCA mRNA species of 2.5 kilobases were identified in various human tumors. A 2.2-kilobase mRNA species, however, could only be detected in leukocytes of patients with chronic myeloid leukemia by hybridization with a probe from the immunoglobulin-like repeat domain of CEA. This region is known to be very similar among the various members of the CEA gene family, and indeed the probe hybridizes with all four mRNA species. In situ hybridization with a cross-hybridizing probe from the NCA gene localized the members of the CEA gene family to the short and to the long arm of chromosome 19. In addition, a CEA cDNA probe was found to hybridize to the long arm of chromosome 19 only.

Specificity and affinity of monoclonal antibodies against carcinoembryonic antigen.

The binding specificities of 52 well-characterized monoclonal antibodies (Mabs) against carcinoembryonic antigen (CEA) from 12 different research groups were studied by immunohistochemistry and immuno flow cytometry. In addition, the binding constant for the interaction between Mab and CEA was determined by a solution-phase assay. Cryostat sections of colon carcinoma and normal colon, stomach, liver, pancreas, and spleen were studied by immunohistochemistry. Peripheral blood granulocytes, monocytes, and lymphocytes were assayed by immuno flow cytometry. The Mabs used here have previously been classified into five essentially nonoverlapping epitope groups (GOLD 1-5) (Cancer Res., 49: 4852-4858, 1989). Most Mabs cross-reacted with different normal tissues, ranging from highly cross-reactive Mabs (positive reaction with 8 of 9 discriminating tissues) to relatively specific Mabs (positive reaction with 1 of 9 discriminating tissues). Five Mabs (10%) were specific, reacting only with colon carcinoma, normal colon mucosa, and normal gastric foveola. There was a correlation between epitope group and binding specificity. Mabs with a high degree of CEA specificity almost exclusively belonged to epitope groups 1, 2, and 3, while highly cross-reactive Mabs belonged to epitope groups 4 and 5. There was no correlation between antibody specificity and affinity for CEA. Specific Mabs with high as well as low affinity were found.

Mice transgenic for the human carcinoembryonic antigen gene maintain its spatiotemporal expression pattern.

The tumor marker carcinoembryonic antigen (CEA) is predominantly expressed in epithelial cells along the gastrointestinal tract and in a variety of adenocarcinomas. As a basis for investigating its in vivo regulation and for establishing an animal model for tumor immunotherapy, transgenic mice were generated with a 33-kilobase cosmid clone insert containing the complete human CEA gene and flanking sequences. CEA was found in the tongue, esophagus, stomach, small intestine, cecum, colon, and trachea and at low levels in the lung, testis, and uterus of adult mice of independent transgenic strains. CEA was first detected at day 10.5 of embryonic development (embryonic day 10.5) in primary trophoblast giant cells and was found in the developing gut, urethra, trachea, lung, and nucleus pulposus of the vertebral column from embryonic day 14.5 onwards. From embryonic day 16.5 CEA was also visible in the nasal mucosa and tongue. Because this spatiotemporal expression pattern correlates well with that known for humans, it follows that the transferred genomic region contains all of the regulatory elements required for the correct expression of CEA. Furthermore, although mice apparently lack an endogenous CEA gene, the entire repertoire of transcription factors necessary for correct expression of the CEA transgene is conserved between mice and humans. After tumor induction, these immunocompetent mice will serve as a model for optimizing various forms of immunotherapy, using CEA as a target antigen.

Cloning of a carcinoembryonic antigen gene family member expressed in leukocytes of chronic myeloid leukemia patients and bone marrow.

The carcinoembryonic antigen (CEA) gene family belongs to the immunoglobulin superfamily and can be subdivided into the CEA and pregnancy-specific glycoprotein subgroups. The basic structure of the encoded proteins consists of, in addition to a leader, one IgV-like and 2, 3, or 6 IgC-like domains. These domains are followed by varying COOH-terminal regions responsible for secretion, transmembrane anchoring, or insertion into the membrane by a glycosyl phosphatidylinositol tail. Here we report on the characterization of CGM6, a new member of the CEA gene subgroup, by complementary DNA cloning. The deduced coding region comprises 349 amino acids and consists of a leader, one IgV-like, two IgC-like domains, and a hydrophobic region, which is replaced by a glycosyl phosphatidylinositol moiety in the mature protein. CGM6 transcripts were only found thus far in leukocytes of chronic myeloid leukemia patients, in normal bone marrow, and in marginal amounts in normal granulocytes. The CGM6 gene product might, therefore, represent a myeloid marker. Analyses of CGM6 protein-expressing HeLa transfectants with monoclonal antibodies strongly indicate that the CGM6 gene codes for the CEA family member NCA-95.

Expression of the CEA gene family members NCA-50/90 and NCA-160 (CD66) in childhood acute lymphoblastic leukemias (ALLs) and in cell lines of B-cell origin.

The carcinoembryonic antigen (CEA) and the classical non-specific cross-reacting antigens (NCAs) belong to the CEA gene family which is part of the immunoglobulin superfamily. In normal hematopoiesis, CEA gene family members (CGMs) have only been reported on cells of myeloid and monocytic origin. In the present study, we analyzed 62 childhood acute lymphoblastic leukemias (ALLs) and seven surface immunoglobulin positive (sig+) B-cell lines for the expression of the CEA family members CEA, NCA-50/90, NCA-95, NCA-160, CGM1 and CGM7. We demonstrated that members of the CEA family were present in 76% of childhood ALLs of B- and T-cell origin. In ALLs of B-cell origin, 82% of the samples expressed at least one CEA subgroup member: 38% NCA-50/90 (CD66c), 31% NCA-160 (CD66a), and 13% both. Six of seven B-cell lines solely expressed NCA-160. In seven ALL of T-cell origin, sole NCA-160 expression was present in 29% of the cases. CEA and CGM1 were not expressed in childhood ALLs or in the sIg+ B-cell lines. In 15 ALLs and seven B-cell lines which could be analyzed for CGM7 expression, the antigen was not detected. NCA-95 was not expressed in 91% of the B-lineage ALLs, in T-lineage ALLs and in the B-cell lines. However, five B-lineage ALLs showed conflicting data on the binding patterns of two, on leukocytes specifically NCA-95 recognizing antibodies suggesting either expression of unknown forms of NCA-95 or NCA-50/90 or of a yet unknown member of the CEA family in these ALL cells. The expression of CEA subgroup members in childhood ALL cells might have prognostic impacts, as an inverse correlation exists between NCA expression on leukemic blasts and the risk factor white blood count at diagnosis.

Analysis of the specificity of CEA reactive monoclonal antibodies. Immunological support for the domain-model of CEA.

A panel of 17 monoclonal antibodies (MAbs), which are reactive with purified carcinoembryonic antigen (CEA), was tested. The MAbs were categorized into 6 groups according to their reactivity with CEA 180, CEA 160, non-specific cross-reacting antigen (NCA) 97 and NCA 50. After chemical modification of CEA (reduction, carboxymethylation, deglycosylation, enzymatic cleavage) and binding studies, the MAbs were further divided into 8 subgroups, representing 8 different antigenic sites on CEA. All MAbs bind to deglycosylated CEA. Most of the MAbs are directed against conformational determinants, since only three of them recognize reduced and alkylated CEA. The same three MAbs are able to detect 29 kDa glycosylated fragments obtained by enzymatic cleavage of CEA. These three protease V8- and trypsin-resistant fragments, probably obtained by interdomain cleavage, show a close relationship in peptide patterns, supporting the repeating structural domain-model of CEA as deduced from the cDNA sequence of CEA.

Non-specific cross-reacting antigen: characterization of specific and cross-reacting epitopes.

A cDNA for NCA-50 was cloned into the inducible expression vector pTRB1, using the polylinker site at the C-terminus of the lac Z' gene. An NCA-specific MAb (N1), NCA and CEA cross-reactive MAbs (T84.1, 192) and polyclonal antisera (anti-NCA and anti-CEA, as well as anti-PS beta G) detected the fusion protein, with a mol. wt of 155,000, which constituted about 5% of the total bacterial protein. Deletion and mutation analysis showed that all MAbs which stained positive in western blots mapped to a small region within the last third of the N-terminal domain. Superimposition of the deduced amino acid sequence of NCA-50 on the known structure of immunoglobulins reveals that the antigenic region is located on a surface loop, which corresponds to a fourth hypervariable region on the immunoglobulin heavy chain variable regions. By oligonucleotide directed site-specific mutagenesis amino acids were deduced, which constitute part of an epitope, to which the NCA-50-specific MAb, N1, binds.

The influence of the local environment on tissue architecture of colorectal carcinoma (CRC) cell aggregates and its consequence for tumour attack by lymphocytes in vitro and in vivo.

We analysed colorectal carcinoma (CRC) specimens, tumour cell spheroids and artificial tumours (ArTs) for tissue architecture, carcinoembryonic antigen (CEA) expression and lymphocyte infiltration. Two distinct organisation forms of well-differentiated CRC cells were found in vivo and in vitro. Tumour cells having contact with the tumour stroma in primary tumours, and tumour cells growing within a stroma-like structure in vitro (ArTs) were arranged as pseudoglands. In contrast, tumour cells grown as spheroids or tumour cells having lost contact with the tumour stroma in primary tumours, and most probably in the circulation, showed an inversion of the architecture of these pseudoglands, presenting their apical cell membrane to the environment. These different tumour cell formations affect lymphocytes attacking the tumour, which need contact with specific cellular membranes of polarised tumour cells, depending on the tumour architecture. Recently, we demonstrated that the CEA expression of CRC cells correlated with their resistance against LAK-cell lysis. Since CEA is mainly expressed on the apical membrane of the tumour cells, independent of the tissue architecture, the change from the pseudoglandular to the spheroid-like formation may represent an escape mechanism for malignant cells.

Isolated extracellular matrix-based three-dimensional in vitro models to study orthotopically cancer cell infiltration and invasion.

An initial event in colon cancer progression is the migration of epithelial cells through the basement membrane (BM) and the invasion of the colon submucosa, where tumour cells enter blood and lymph vessels to spread throughout the body. To interrupt this process would mean the prevention of metastasis. In order to investigate tumour cell invasion orthotopically in the human system, we established novel in vitro models which mimic normal human colon tissue (colon reproductions, CoRes) and primary colon carcinomas (artificial tumours, ArTs). These models are based on the isolated extracellular matrix (iECM) of the respective human tissues. Two isolation methods were established, the Digestion Method and the Lysis Method neither of which destroyed the characteristic architecture of the ECM found in the original tissues. BM components, i.e. laminin, fibronectin and collagen IV, were detectable in the iECM isolated with the Lysis Method but not those isolated with the Digestion Method. Scanning electron microscopic analysis of the normal colon iECM demonstrated that even if the BM was missing, the luminal surface consisted of densely packed ECM filaments which do not allow cell infiltration without degradation of the iECM. Furthermore, we demonstrated that iECM can be separately supplemented with different cell types, i.e. colorectal carcinoma cells, normal fibroblasts and immune cells at any desired concentration, combination and localisation. Therefore, these models could be used to determine the role of the BM and of the tumour cell/normal cell crosstalk in the infiltration process of human colorectal carcinoma cells.

Caring About Women and Cancer (CAWAC): a European survey of the perspectives and experiences of women with female cancers.

This paper reports on the findings of the largest ever European survey of female patients' perceptions of their cancer treatment. It has provided clarification of what women consider important in relation to their management and has identified several areas where more research is needed. It has shown that women's knowledge about cancer before diagnosis is poor and the number undergoing regular screening could be improved. Women are not being adequately prepared and educated about what to expect from treatment and steps should be taken as a matter of urgency to redress this shortcoming. It was revealed that whilst families were the primary source of support to female cancer patients, women also derive considerable support from healthcare professionals, particularly senior doctors; more attention should be paid by specialists and nurses to developing psychological skills to cope with this. In this context, further research is needed into how support groups may best meet patient needs.

Evaluation of a test system for measuring cytokine production in human whole blood cell cultures.

A simple and reproducible method is described for the measurement of mitogen-induced cytokine production in cultures of both human peripheral blood mononuclear cells (PBMC) and whole blood. In the culture supernatants the cytokines interleukin-1 alpha (IL-1 alpha), interleukin-2 (IL-2), interferon-gamma (IFN-gamma) and tumor necrosis factor alpha (TNF-alpha) were determined by a rapid and sensitive immunoassay using various monoclonal and polyclonal antibodies. Comparing the PBMC cultures with the whole blood system a good correlation was obtained if the cell number was taken into account. In the post-induction supernatants the cytokine values were found to follow typical kinetic curves. The protocol was evaluated by screening 60 cancer patients with primary disease and 60 healthy controls. A markedly reduced secretion of IFN-gamma and IL-1 alpha was found in the cancer patients compared to controls, although leukocyte and lymphocyte counts were almost identical in both groups.

Immunosuppressive activity of sera from patients with colorectal and gynecological carcinomas as evaluated by impaired IFN-gamma, IL-1 alpha and TNF-alpha production of human peripheral mononuclear cells.

The sera from patients with preoperative colorectal and gynecological carcinomas (ovarian and breast cancer) were investigated for their putative immunosuppressive activity (ISA). ISA was measured by determining the changes in the production of IFN-gamma, IL-1 alpha and TNF-alpha by human peripheral mononuclear cells (PBMC) from six normal donors. Phytohemagglutinin-stimulated PBMC were incubated with sera from patients with colorectal and gynecological carcinomas and healthy controls and in the 4-day post-inductional supernatants the cytokines were measured by an enzymoimmunological assay (ELISA). Sera from patients with carcinomas significantly decreased the IFN-gamma production as compared to the controls. In the cultures containing sera from patients with colorectal and ovarian but not breast carcinoma, significantly lower levels of IL-1 alpha and TNF-alpha were measured compared to the cultures with sera from healthy controls.

Carcinoembryonic antigen (CEA) in human colorectal cancers growing subcutaneously in nude mice.

Eleven human colorectal xenografts from 26 tumor lines established in nude mice in our laboratory were investigated for carcinoembryonic antigen (CEA) production. Serum levels of CEA in nude mice were markedly elevated in all cases but one, median values ranging from 3.7 to 42.8 ng/ml. Carcinoembryonic-antigen levels for ten nontumor-bearing nude mice ranged from 0.00 to 0.12 ng/ml (median value 0.07). A clear linear correlation could be demonstrated between log CEA serum levels and log tumor volumes in serial measurements. In one case, CEA was normal in the cancer patient's serum and gave the lowest value (3.7 ng/ml) in nude mice bearing the xenograft. By the indirect peroxidase technique, CEA was localized mainly on the apical cell membranes of the cancerous glands and in necrotic areas; only small amounts were detectable in the cytoplasm. Transplantation of human colorectal carcinomas into nude mice offers an excellent experimental in vivo system to study the mechanism of release, metabolism, and excretion of the marker.

Study of the relationship between immunohistologically demonstrated lymphocytes infiltrating human breast carcinomas and patients' survival.

Eighty-five breast carcinomas were immunostained for CD3-, CD4-, CD8-, CD16-, CD22-, CD38- and CD57-positive lymphocyte subpopulations. The results were related to follow-up data (median follow-up 46 months) of 74 patients regarding overall survival and 73 patients in respect to disease-free survival. Whereas the number of axillary lymph node metastases (P less than 0.01) and the hormone receptor status (P less than 0.01) resulted in significantly different survival curves for overall survival, not one of the lymphocyte subset infiltrats correlated significantly which overall survival. For disease-free survival, pT stage (P less than 0.01) and nodal (P less than 0.01) and hormone receptor status (P less than 0.05) proved to be prognostically important. However, disease-free survival was not influenced by the infiltration of any lymphocyte subset.

Estrogen receptor determinations in primary breast cancer. A comparison of a biochemical dextran-coated charcoal and an immunohistological technique.

The present study describes the estrogen receptor (ER) detection by an immunocytochemical assay kit (ER-ICA) on cryostat sections of 78 primary breast carcinomas. Results are compared with quantitatively measured ER levels, which were obtained by the dextran-coated charcoal (DCC) method. An excellent overall correlation between the logarithm of the ER levels, estimated by this technique, and the semiquantitative immunocytochemical evaluation was found, i.e. r = 0.73. Since the ER-ICA can be easily handled without radioactivity being involved and since it is more representative of the total tumor, we conclude (as other groups before us) that the ER-ICA is an easy-to-handle and reliable technique presenting many advantages over the DCC method.