Biodefense function of omental milky spots through cell adhesion molecules and leukocyte proliferation

To evaluate the immunological functions of the greater omentum in the peritoneal cavity, the localization of cell adhesion molecules (CAMs) on mesothelial cells and leukocytes in the omental milky spots were studied in normal and lipopolysaccharide (LPS)-stimulated mice by means of immunoelectron microscopy. The milky spots featured numerous leukocytes among the dome-shaped mesothelial cells, even in the normal stable state. Leukocyte integrins LFA-1, Mac-1, and VLA-4 were preferentially localized to microvilli and ruffles of macrophages and lymphocytes. The mesothelial cells of the milky spots showed higher ICAM-1 levels than did those of other omental regions, and fibronectin was detected in the stomata. The number of leukocytes markedly increased following an increase in proliferating cell nuclear antigen (PCNA)-positive cells in the milky spots after LPS stimulation. The mesothelial cells contained VCAM-1 newly restricted to the microvilli and increasing amounts of ICAM-1. These results show that the omental milky spots are active sites for leukocyte migration and peritoneal leukocyte supply because of the presence of adhesion molecules and active cell proliferation. Proliferative active leukocytes and those that have migrated from vessels pass through the stomata via an interaction of VLA-4 and fibronectin, adhere to the microvilli of the activated mesothelial cell surface as the result of an interaction between ICAM-1/VCAM-1 and integrins, and exude into the peritoneal cavity. Much of the exudation and adhesion of leukocytes seen in the milky spots of LPS-stimulated mice may be attributable to an increase in cell proliferation and in the amounts of ICAM-1 and VCAM-1.     

Defective B1 cell homing to the peritoneal cavity and preferential recruitment of B1 cells in the target organs in a murine model for systemic lupus erythematosus

We previously reported that B lymphocyte chemoattractant (BLC; CXCL13) was highly and ectopically expressed in aged (NZB x NZW)F1 (BWF1) mice developing lupus nephritis, and that B1 cells were preferentially chemoattracted toward BLC. We demonstrate in this study that B1 cells fail to home to the peritoneal cavity in aged BWF1 mice developing lupus nephritis, and that they are preferentially recruited to the target organs including the kidney, lung, and thymus when injected i.v. In contrast, B1 cells homed to the peritoneal cavity in aged BALB/c mice as effectively as in young mice. Accumulation of B1 cells to the omentum milky spots was also impaired in aged BWF1 mice compared with young mice. CD11bhighF4/80high cells with macrophage morphology were confirmed to be a major cell source for BLC in the peritoneal cavity both in young and aged BWF1 mice. However, the number of BLC-producing peritoneal macrophages was markedly decreased in aged BWF1 mice. These results suggest that the decreased number of BLC-producing peritoneal macrophages together with ectopic high expression of BLC in aged BWF1 mice result in abnormal B1 cell trafficking during the development of murine lupus.     

A rapid and simple method for visualizing milky spots in large fixed tissue samples of the human greater omentum

ABSTRACT

Milky spots are unique lymphoid structures in the greater omentum that participate in both immune homeostasis of the peritoneal cavity and formation of omental metastases. We developed a rapid and simple staining method to enable macro- or stereomicroscopic identification of these miniscule structures in large samples of fixed human greater omentum. By immersing approximately 6 × 4 cm samples of omental tissue in hematoxylin, these samples could be evaluated quickly for the presence of milky spots. We used an alum hematoxylin variant containing 1 g hematoxylin, 50 g aluminium ammonium sulfate, 0.2 g sodium iodide, 1 g citric acid and 50 g chloral hydrate. This staining method enabled us to determine the number, location, dimensions and topographical relation of milky spots to other structures. Our method also facilitates isolation of milky spots for further investigation. Hematoxylin imparts a blue color to the milky spots, which remains in place during further processing for paraffin embedding. This enabled easy recognition of milky spots during transfer through various solutions and permitted selection of relevant paraffin slides prior to additional staining.     

Insufficient ability of omental milky spots to prevent peritoneal tumor outgrowth supports omentectomy in minimal residual disease

Background: The greater omentum is frequently involved in the course of gastrointestinal and ovarian tumors. Therefore, common practice in surgical treatment for especially gastric and ovarian cancer includes removal of the greater omentum. Paradoxically, many immune cells, such as macrophages that accumulate in so-called milky spots, reside within the omentum and are cytotoxic against tumor cells ex vivo. Consequently, omental macrophages might play an important role in killing tumor cells, and may hereby prevent development into local peritoneal recurrences. In the present study, we therefore evaluated the role of the omentum and the clinical relevance of omentectomy in minimal residual disease (MRD). Methods: Tumor cell dissemination patterns on the omentum in a rat model were examined using DiI-labelled CC531s tumor cells. Additionally, intra peritoneal (i.p.) tumor load was investigated in rats that underwent omentectomy or sham laparotomy followed by i.p. injection of CC531s cells on day 21, which represented MRD. Results: At 4 h post injection, tumor cells predominantly adhered on milky spots. Number of cells thereafter declined rapidly suggesting initial tumor killing functions in these specific immune aggregates. Despite initial reduction observed in milky spots, numbers of tumor cells however increased at fatty tissue stripes that border the omentum. This indicated proliferation at these locations, which corresponded to macroscopic observations of the omenta on day 21 after tumor cell injection. Omentectomy resulted in reduced intra-abdominal tumor load, which was completely attributable to the absence of the omentum, as tumor development did not differ on other sites. Even in the MRD group microscopic clusters of tumor cells located in the omentum eventually developed into macroscopic nodules.Conclusion: Since the ability of omental milky spots is, even in MRD, insufficient to prevent intra abdominal tumor outgrowth, omentectomy, which reduces tumor load, is recommended in surgical treatment of intra abdominal tumors that are prone to disseminate intraperitoneally.     

Cellular subsets of the milky spots in the human greater omentum

Summary The cellular composition of the human milky spots was investigated on surgically removed specimens of the greater omentum of three 8-month-old infants operated on for neuroblastoma. Monoclonal antibodies and immunohistochemical methods for recognition of macrophages, B-lymphocytes and T-lymphocytes and toluidine-blue staining for mast cells were used. The mean number of cells in one milky spot amounted to 570±33. This cell population was composed of 47.5% macrophages, 29.1% B-lymphocytes, 11.7% T-lymphocytes and 6.1% mast cells. Since inflammation was absent in the material investigated, the numerical data found in the present paper could be regarded as representative cell levels of normal milky spots.     

Differential molecular and anatomical basis for B cell migration into the peritoneal cavity and omental milky spots

The constitutive migration of B cells from the circulation into the peritoneal cavity and back is essential for peritoneal B cell homeostasis and function. However, the molecular machinery and the anatomical basis for these migratory processes have hardly been investigated. In this study, we analyze the role of integrins as well as the role of the omentum for B2 cell migration into and out of the peritoneal cavity of mice. We demonstrate that alpha(4)beta(7) integrin-mucosal addressin cell adhesion molecule 1 interaction enables B2 cell migration from the circulation into omental milky spots but not into the peritoneum. In contrast, alpha(4)beta(1) integrin mediates direct entry of B2 cells into the peritoneal cavity as well as their retention at that site, limiting B2 cell egress via the draining parathymic lymph nodes. Surgical removal of the omentum results in a 40% reduced immigration of B2 cells from the circulation into the peritoneum but does not impair B cell exit from this compartment. In conclusion, these data reveal the existence of alternative routes for B2 cell entry into the peritoneal cavity and identify integrins as key factors for peritoneal B2 cell homeostasis, mediating B2 cell migration into and out of the peritoneal cavity as well as their retention at this site.     

The involvement of omentum and its milky spots in the dynamics of peritoneal macrophages

The investigation has been carried out on stimulated and unstimulated peritoneal cavities of rats. China ink and Corynebacterium parvum were injected i.p. both as peritoneal stimuli and markers. Omenta were picked up at time intervals beginning with 10 min and up to seven days after the i.p. injection. The light and electronmicroscopic investigation showed after 10-30 minutes labeled macrophages stuck as monolayers on some peritoneal areas corresponding to the milky spots which developed in size and number. Days after the i.p. injection the labeled macrophages were found deeper in the milky spots. After the fourth day they appeared in the regional lymph nodes. The milky spots contained also large lymphocytes and plasma cells. The results suggest that milky spots are not only places of resident macrophages development and release in the peritoneal cavity but also their exit pathways. Therefore the omentum leads the traffic of peritoneal macrophages. The developed milky spots play also the role of lymphoid structures providing grounds for macrophage-lymphocyte contacts.     

In Vivo and Ex Vivo Approaches to Study Ovarian Cancer Metastatic Colonization of Milky Spot Structures in Peritoneal Adipose

High-grade serous ovarian cancer (HGSC), the cause of widespread peritoneal metastases, continues to have an extremely poor prognosis; fewer than 30% of women are alive 5 years after diagnosis. The omentum is a preferred site of HGSC metastasis formation. Despite the clinical importance of this microenvironment, the contribution of omental adipose tissue to ovarian cancer progression remains understudied. Omental adipose is unusual in that it contains structures known as milky spots, which are comprised of B, T, and NK cells, macrophages, and progenitor cells surrounding dense nests of vasculature. Milky spots play a key role in the physiologic functions of the omentum, which are required for peritoneal homeostasis. We have shown that milky spots also promote ovarian cancer metastatic colonization of peritoneal adipose, a key step in the development of peritoneal metastases. Here we describe the approaches we developed to evaluate and quantify milky spots in peritoneal adipose and study their functional contribution to ovarian cancer cell metastatic colonization of omental tissues both in vivo and ex vivo. These approaches are generalizable to additional mouse models and cell lines, thus enabling the study of ovarian cancer metastasis formation from initial localization of cells to milky spot structures to the development of widespread peritoneal metastases.     

Mesothelial stomata overlying omental milky spots: Scanning electron microscopic study

Summary The presence in the rat omentum of intercellular pores (the classical stomata of von Recklinghausen) between the mesothelial cells overlying aggregates of lymphoreticular cells (the classical milky spots of Ranvier) and the apparent migration of lymphocytes through these stomata were recorded for the first time by scanning electron microscopy. Previous studies on passage of cells across the peritoneum and omentum used experimentally administered cells, while in the present study no cells were administered to the rats and their own lymphocytes were observed in situ. The possible role of lymphocytes in the peritoneal cavity is also briefly discussed.     

Characterization of Omental Immune Aggregates during Establishment of a Latent Gammaherpesvirus Infection

Herpesviruses are characterized by their ability to establish lifelong latent infection. The gammaherpesvirus subfamily is distinguished by lymphotropism, establishing and maintaining latent infection predominantly in B lymphocytes. Consequently, gammaherpesvirus pathogenesis is closely linked to normal B cell physiology. Murine gammaherpesvirus 68 (MHV68) pathogenesis in laboratory mice has been extensively studied as a model system to gain insights into the nature of gammaherpesvirus infection in B cells and their associated lymphoid compartments. In addition to B cells, MHV68 infection of macrophages contributes significantly to the frequency of viral genome-positive cells in the peritoneal cavity throughout latency. The omentum, a sheet of richly-vascularized adipose tissue, resides in the peritoneal cavity and contains clusters of immune cell aggregates termed milky spots. Although the value of the omentum in surgical wound-healing has long been appreciated, the unique properties of this tissue and its contribution to both innate and adaptive immunity have only recently been recognized. To determine whether the omentum plays a role in gammaherpesvirus pathogenesis we examined this site during early MHV68 infection and long-term latency. Following intraperitoneal infection, immune aggregates within the omentum expanded in size and number and contained virus-infected cells. Notably, a germinal-center B cell population appeared in the omentum of infected animals with earlier kinetics and greater magnitude than that observed in the spleen. Furthermore, the omentum harbored a stable frequency of viral genome-positive cells through early and into long-term latency, while removal of the omentum prior to infection resulted in a slight decrease in the establishment of splenic latency following intraperitoneal infection. These data provide the first evidence that the omentum is a site of chronic MHV68 infection that may contribute to the maintenance of chronic infection.     

Activated omentum becomes rich in factors that promote healing and tissue regeneration

In order to study the mechanism by which an omental pedicle promotes healing when applied to an injured site, we injected a foreign body into the abdominal cavity to activate the omentum. One week after the injection, we isolated the omentum and measured blood vessel density, blood content, growth and angiogenesis factors (VEGF and others), chemotactic factors (SDF-1α), and progenitor cells (CXCR-4, WT-1). We found that the native omentum, which consisted mostly of adipose tissue, expanded the mass of its non-adipose part (milky spots) 15– to 20-fold. VEGF and other growth factors increased by two– to four-fold, blood vessel density by three-fold, and blood content by two-fold. The activated omentum also showed increases in SDF-1α, CXCR-4, and WT-1 cells (factors and cells positively associated with tissue regeneration). Thus, we propose that an omentum activated by a foreign body (or by injury) greatly expands its milky-spot tissue and becomes rich in growth factors and progenitor cells that facilitate the healing and regeneration of injured tissue. This work was partly supported by a grant (no. 2000–241 to A.K.S.) from the Juvenile Diabetes Foundation International.     

Milky spots in the human greater omentum. Macroscopic and histological identification

Macroscopic and histological investigations were made from surgical specimens demonstrating milky spots in the human greater omentum from subjects of various ages. The milky spots in the human greater omentum appear as tiny, cotton-wool-like masses to the naked eye. Histologically, the milky spots consisted mainly of many macrophages with diffuse cytoplasmic esterase reaction products and esterase-negative B lymphocytes surrounding the vascular networks. Macrophages phagocytosed many carbon particles which were introduced as a carbon suspension during the operation. The vascular networks were blood capillary convolutions with a glomerular shape. Silver impregnation showed the delicate networks of reticular fibers which constitute the framework of the organ. The number of milky spots was highest in infancy and gradually decreased with age.     

Effect of intraperitoneal administration of granulocyte/macrophage-colony-stimulating factor in rats on omental milky-spot composition and tumoricidal activity in vivo and in vitro

 Milky spots in the greater omentum are small accumulations of leucocytes that consist mainly of macrophages and have recently shown to be a selective dissemination site of intraperitoneal (i. p.) inoculated tumour cells. However, milky-spot macrophages show tumoricidal activity and may, therefore, be an excellent source of effector cells suited for local immunotherapy. In the present study we first examined whether granulocyte/macrophage- colony-stimulating factor (GM-CSF) treatment of isolated milky-spot macrophages affects the cytotoxicity against syngeneic colon carcinoma cells (CC531) in vitro. Secondly, we studied the influence of intraperitoneal GM-CSF administration on the number and antitumour activity of milky-spot and peritoneal macrophages. All studies were performed in Wag/Rij rats in which a syngeneic colon carcinoma cell line (CC531) is available. The results of the in vitro study showed that GM-CSF treatment of the omental macrophages led to an increased cytotoxicity against the tumour cell line. Intraperitoneal administration of 1000 U GM-CSF daily for 7 consecutive days demonstrated both an enhanced antitumour activity of the milky-spot macrophages and an increase in the milky-spot macrophage population. An increase in the proliferative capacity, according to bromodeoxyuridine incorporation, was shown in the milky-spot macrophages. Taking into account both the enhanced macrophage number and their enhanced activity upon i. p. GM-CSF treatment, the milky-spot macrophages may provide a rationale for local intraperitoneal immunotherapy in the prevention of intra-abdominal tumour growth. Received: 11 April 1996 / Accepted: 21 May 1996     

The Effects of a Tsetse DNA Virus Infection on the Functions of the Male Accessory Reproductive Gland in the Host Fly Glossina morsitans centralis (Diptera; Glossinidae)

Freshly deposited third instar Glossina morsitans centralis larvae were infected with the tsetse DNA virus by microinjection, and at emergence adult males were separated from the females and fed on rabbit blood every second day for 8 days. A control group treated with sterile saline were handled similarly. They were dissected, and comparative observations made on the appearance and size of the accessory reproductive glands (ARG) in infected and control males. Regularly fed 8-day-old males from infected and control groups were mated to 2-day-old normal females obtained from the insectay. After separation from copula, the females were dissected and the uteri examined for the presence and quality of the spermatophore. The spermathecae were also examined for insemination. ARG tissues from the control and virus infected regularly fed 8-day-old male flies were fixed and processed for electron microscopic studies. The ARGs from control flies were found to be milky in appearance, whereas those from virus-infected flies were transparent in most parts. The ARGs from virus-infected males were significantly smaller in diameters (F = 42.26, p < 0.0001) and shorter (F = 200.4, p < 0.0001) than those of the controls. Most of the virus-infected males failed to form a complete spermatophore, whereas almost all the controls formed complete spermatophore as observed in the uteri of the female mates (Χ² = 111.661, p < 0.0001). The infected males that formed partial spermatophores and those that did not form any at all failed to inseminate their female mates. Histological studies of the ARGs revealed some lesions in the epithelial cells characterized by degeneration of cytoplasmic organelles and detachment of the muscle layer from the basal plasma membrane. However, no virus particles were observed in the affected cells. Received: 18 November 1998 / Accepted: 10 December 1998     

Identification of the proteins related to cytochrome P450 induced by fenvalerate in a <Emphasis Type="Italic">Trichoplusia ni</Emphasis> cell line

In order to reveal the metabolic reaction to the presence of fenvalerate mediated by P450 in insects, we used the trypan blue exclusion technique and 3-(4,5-dimethylthiazol)-2,5-diphenyltrazolium bromide (MTT) reduction assay to assess the vitality of Trichoplusia ni (Tn) cells treated with fenvalerate, and observed dose- and time-dependent changes in total cellular P450s. In addition, two-dimensional electrophoresis (2-DE) and matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF-MS) were used to identify the proteins involved in the fenvalerate reaction process. Finally, the cDNA of P450 fragments was cloned and real-time RT-PCR was performed. Our data showed that at the 0–15 μmol/L challenge concentration of fenvalerate, at which the vitality of Tn cells was not affected (p > 0.05), there was a tendency toward a dose- and time-response of total cellular P450s, which peaked at the 9 h (p < 0.05) and 12 h (p < 0.01) time points following 12.5 μmol/L stimulation with fenvalerate. The 2-DE assay detected more than 1300 protein spots in each two-dimensional gel, of which 33 spots displayed significant differences. Among the changed spots, three isoforms of P450 were identified. One of the three P450 cDNA fragments (CYP4L4) was cloned and sequenced, and its expression in treated Tn cells increased significantly (p < 0.01). It was found that fenvalerate induced the expression of P450s in insect cells. This suggests that fenvalerate could be metabolized by CYP4L4 through a hydroxylation reaction in insect cells.     

Biodegradation of biphenyl by the ascomycetous yeast Debaryomyces vanrijiae

Cells of the yeast strain Debaryomyces vanrijiae SBUG 770, grown with glucose, converted biphenyl to 4-hydroxybiphenyl as the major metabolite. In addition, 2-hydroxybiphenyl was formed in minor amounts. No further degradation of these substances was detected. However, these monohydroxylated derivatives were oxidised by alkane-grown cells in the presence of the co-metabolic substrate, tetradecane. Under these conditions 2-hydroxybiphenyl was oxidised to 2,5-dihydroxybiphenyl, and 4-hydroxybiphenyl was rapidly metabolised by formation of two major metabolites. One was identified as 3,4-dihydroxybiphenyl. Characterisation of the second product as 4-phenylmuconolactone points to a further metabolism of the initially formed dihydroxylated biphenyl via ortho-ring fission. Received: 8 May 1998 / Received revision: 26 June 1998 / Accepted: 2 July 1998     

Morphological features of milky spots of the rat omentum in inflammation

The omental milky sports of the rat have been examined with the light and electron microscope after intraperitoneal stimulation by particulate coal and zymosan. No positive correlation between cell level proliferation in milky spots and alteration of their size has been found. The increase in size of milky spots is due to the inflow of cells from the blood and bone marrow. The development in milky spots of a great number of macrophages "tubercles" and multinucleated giant cells in the experimental conditions may respond to inflammation. The administration of zymosan result in the influx of lymphocytes forming lymphatic follicle-like structure. Despite the absence of germinative centres, the appearance of a great number of lymphoblasts and plasma cells in the milky spots provides the evidence of the active antibody production aimed at immunological protection of abdominal cavity.     

Humanisation and characterisation of PR1A3, a monoclonal antibody specific for cell-bound carcinoembryonic antigen

Carcinoembryonic antigen (CEA) is highly expressed by most tumours of gastrointestinal origin, but its use as a target for tumour therapy is complicated by the high levels of soluble CEA that are found circulating in the blood of cancer patients. A monoclonal antibody PR1A3 has been prepared, which binds preferentially to cell-surface rather than soluble CEA, this cell selectivity should make PR1A3 an ideal candidate for antibody-targeted tumour therapy. PR1A3 has been humanised and shown to retain its cell-surface specificity and affinity. Stable expression of the humanised antibody from chinese hamster ovary (CHO) cells has been achieved after transfection and amplification. Since PR1A3 binds preferentially to cell-associated CEA, a cell-free enzyme-linked immunosorbent assay (ELISA) has been developed to allow characterisation and routine assay of the antibody. This assay was developed using a recombinant chimeric protein constructed by cloning the domain of CEA that is bound by PR1A3 (the B3 domain) into a hybrid gene containing the Fc portion of IgG and three domains of biliary glycoprotein. Stable expression of this hybrid protein has been achieved from CHO cells. In ELISA both humanised and murine PR1A3 bound strongly to this antigen but only at a minimal level to soluble CEA. Two binding sites for the antibody were found on the gastric carcinoma cell line MKN45, one of higher affinity (1 nM) and the other at lower affinity (60 nM). Similar affinities were found for both murine and humanised antibodies. The data presented make it unlikely that the differential binding to cell-surface as distinct from soluble CEA can be accounted for by low affinity of PR1A3 for CEA, and provides further support for the hypothesis that some conformational change takes place on CEA release from cells and that it is this change that blocks PR1A3 binding to its epitope. Received: 5 October 1998 / Accepted: 19 November 1998     

Modeling the Early Steps of Ovarian Cancer Dissemination in an Organotypic Culture of the Human Peritoneal Cavity

The pattern of ovarian cancer metastasis is markedly different from that of most other epithelial tumors, because it rarely spreads hematogenously. Instead, ovarian cancer cells exfoliated from the primary tumor are carried by peritoneal fluid to metastatic sites within the peritoneal cavity. These sites, most notably the abdominal peritoneum and omentum, are organs covered by a mesothelium-lined surface. To investigate the processes of ovarian cancer dissemination, we assembled a complex three-dimensional culture system that reconstructs the lining of the peritoneal cavity in vitro. Primary human fibroblasts and mesothelial cells were isolated from human omentum. The fibroblasts were then mixed with extracellular matrix and covered with a layer of the primary human mesothelial cells to mimic the peritoneal and omental surfaces encountered by metastasizing ovarian cancer cells. The resulting organotypic model is, as shown, used to examine the early steps of ovarian cancer dissemination, including cancer cell adhesion, invasion, and proliferation. This model has been used in a number of studies to investigate the role of the microenvironment (cellular and acellular) in early ovarian cancer dissemination. It has also been successfully adapted to high throughput screening and used to identify and test inhibitors of ovarian cancer metastasis.     

Pharmacological administration of granulocyte/macrophage-colony-stimulating factor is of significant importance for the induction of a strong humoral and cellular response in patients immunized with recombinant carcinoembryonic antigen

Eighteen colorectal carcinoma patients without macroscopic disease after surgery were immunized using recombinant (r) human (h) carcinoembryonic antigen (CEA) with (n = 9) or without (n = 9) the addition of soluble granulocyte/macrophage-colony-stimulating factor (GM-CSF). The dose of rhCEA per immunization was 100 μg (n = 6), 316 μg (n = 6) or 1000 μg (n = 6). rhCEA was given s.c. on day 1 and 80 μg/day of GM-CSF s.c. on days 1–4. The schedule was repeated six times during a period of 9 months. All patients in the GM-CSF group developed a strong rhCEA-dose-dependent IgG antibody response while only one-third of the non-GM-CSF patients mounted a weak antibody response. All patients (9/9) in the GM-CSF group developed a strong rhCEA-specific proliferative T cell response as well as type I T cells (interferon γ secretion). In 45% of the patients also a weak type II T cell response (interleukin-4 secretion) was evoked. Both MHC-class-I- and -II restricted rhCEA-specific T cells were noted. A specific cellular response (proliferation and/or cytokine secretion) against native hCEA could be found in 8/9 patients in the GM-CSF group, although at a significantly lower level than against rhCEA. In the non-GM-CSF group a weak rhCEA-specific T cell response was induced. Three patients had a proliferative response, 4 patients type I T cells and 6 patients type II T cells. No signs of autoimmune reactions were noted. Local pharmacological administration of GM-CSF seemed to be a prerequisite for the induction of a strong immunity against baculovirus-produced hCEA protein. However, the cellular response against native CEA was of a significantly lower magnitude. Received: 13 November 1997 / Accepted: 21 May 1998