Expression of oncofoetal pancreatic antigens in hamster adult pancreas during experimental carcinogenesis

The mucinous domain of pancreatic carboxyl-ester lipase (CEL) contains core 1/core 2 O-glycans that can be modified by ABO blood group determinants

Carboxyl-ester lipase (CEL) is a pancreatic fat-digesting enzyme associated with human disease. Rare mutations in the CEL gene cause a syndrome of pancreatic exocrine and endocrine dysfunction denoted MODY8, whereas a recombined CEL allele increases the risk for chronic pancreatitis. Moreover, CEL has been linked to pancreatic ductal adenocarcinoma (PDAC) through a postulated oncofetal CEL variant termed feto-acinar pancreatic protein (FAPP). The monoclonal antibody mAb16D10 was previously reported to detect a glycotope in the highly O-glycosylated, mucin-like C terminus of CEL/FAPP. We here assessed the expression of human CEL in malignant pancreatic lesions and cell lines. CEL was not detectably expressed in neoplastic cells, implying that FAPP is unlikely to be a glycoisoform of CEL in pancreatic cancer. Testing of the mAb16D10 antibody in glycan microarrays then demonstrated that it recognized structures containing terminal GalNAc-α1,3(Fuc-α1,2)Gal (blood group A antigen) and also repeated protein sequences containing GalNAc residues linked to Ser/Thr (Tn antigen), findings that were supported by immunostainings of human pancreatic tissue. To examine whether the CEL glycoprotein might be modified by blood group antigens, we used high-sensitivity MALDI-TOF MS to characterize the released O-glycan pool of CEL immunoprecipitated from human pancreatic juice. We found that the O-glycome of CEL consisted mainly of core 1/core 2 structures with a composition depending on the subject's FUT2 and ABO gene polymorphisms. Thus, among digestive enzymes secreted by the pancreas, CEL is a glycoprotein with some unique characteristics, supporting the view that it could serve additional biological functions to its cholesteryl esterase activity in the duodenum.

Human fetoacinar pancreatic protein: an oncofetal glycoform of the normally secreted pancreatic bile-salt-dependent lipase

A fetoacinar pancreatic protein (FAP) associated with the ontogenesis, differentiation and oncogenic transformation of the human exocrine pancreas has been purified from pancreatic juices of patients suffering from pancreatitis or duodenal cancers invading the pancreas [Escribano and Imperial (1989) J. Biol. Chem. 264, 21865-21871]. This protein has striking similarities, i.e. M(r), amino acid composition and N-terminal sequence, to the bile-salt-dependent lipase (BSDL) of normal human pancreatic secretion. The aim of this study was to gain further insight into the nature of the two proteins. Reactivity with the mouse monoclonal antibody J28 (mAb J28), which characterizes FAP, and enzyme activity could not be dissociated during biochemical purification of BSDL. Furthermore, a polyclonal antiserum raised against purified human BSDL reacted completely with FAP in Western-blot analysis giving additional support to the idea of similar molecular structures for BSDL and FAP. However, by the same technique, mAb J28 reacted with a relatively restricted population of BSDL molecules. The classical BSDL preparation could be separated into molecules bearing the J28 epitope and those devoid of it by immunoaffinity on immobilized mAb J28. The two subpopulations had identical N-terminal sequences and some differences in their amino acid compositions. However, they had different carbohydrate compositions. J28-epitope-bearing molecules were active on BSDL substrates, although their specific activity was decreased. These results are consistent with the existence of two closely related polypeptide chains with different glycan counterparts. Therefore, if the name FAP is reserved for molecules bearing the J28 epitope, which is linked to a carbohydrate-dependent structure. FAP could represent an oncofetal-related variant of BSDL. Our result is the first demonstration of the existence of an oncofetal-type subpopulation of an otherwise normally secreted human pancreatic enzyme.

Immunosurveillance by T-lymphocytes in pretumoral stages of chemically induced pancreatic carcinogenesis

Cellular immune-response during pancreatic carcinogenesis induced by N-nitroso-bis(2-hydroxy-propyl)amine in Syrian Golden hamsters was studied using a mouse antiserum to hamster T-lymphocytes in indirect immunofluorescence. The chronology of lesions in this model is, acinar cell atypia, cystadenoma, ductal hyperplasia and adenocarcinoma. Lymphocyte infiltration began before microscopic lesions. Starting as an interstitial and interlobular migration, this earliest population was composed of various kind of mononuclear cells including T-cells. As pancreatic lesions proceeded, an abundant lymphocyte supply through newly formed capillaries (angiogenesis), gave rise to inter- and intralobular nodules composed almost exclusively of T-cells. Migrating from nodules, T-cells invaded and readily destroyed the exocrine tissue. Formation of hyperplasic ducts and of adenocarcinoma was accompanied by considerable accretion of the basal membrane (fibrosis). T-cells were located outside and around this basal membrane so that they never invaded the ductal epithelium. Our results suggest there is an effective immunosurveillance in the early stages of transformation that becomes ineffective at later stages as a consequence of T-cells' inability to pass through the basal membrane barrier surrounding the ductal epithelium in preneoplasic lesions (ductal hyperplasia) and in adenocarcinoma. Extending our observations to human pancreatic cancer could provide a new insight in cellular immunosurveillance and, as a consequence might, help cellular immunotherapeutic approaches for this almost fatal disease.

Radioimmunolocalization of the monoclonal antibody J28 in early transformation stages in N-nitrosobis(2-hydroxypropyl)amine-induced pancreatic tumors in the Syrian golden hamster

We describe the in vivo localization of radiolabeled mAb J28, a murine monoclonal antibody characterizing the oncodevelopmental human fetoacinar pancreatic (FAP) protein, at different stages of chemical induction of pancreatic tumors in the Syrian golden hamster. Before doing localization studies in this model, we looked at the cross-reactivity of mAb J28. Semiquantitative dot-blot analysis demonstrated that the antigen recognized in hamster pancreas has an oncodevelopmental expression pattern, while a molecular mass identical to that of human FAP was deduced from sodium dodecyl sulfate/polyacrylamide gel electrophoresis/nitrocellulose immunoblot. 125I-labeled mAb J28 was administered through micro-osmotic pumps to hamsters treated with N-nitrosobis(2-hydroxypropyl)amine (BHP). This was done at three intervals that roughly correspond to the latent period, pretumoral stages, and terminal cancerogenesis in two independent groups of hamsters. Both studies allowed similar results: (a) mAb J28 accumulated almost specifically in the pancreas; (b) maximal accumulation was associated with pleomorphic alterations of the acinar cell tissue at pretumoral stages; (c) no accumulation was found in the case of adenocarcinoma of the pancreas. It is concluded that FAP behaves as a marker of preneoplastic lesions, and therefore that radioimmunoimaging with mAb J28 might help with early diagnosis of pancreatic cancer.

Expression of blood group-related carbohydrate antigens in normal human pancreatic tissue

The expression of type 1, 2 and 3 chain carbohydrate structures in 15 normal pancreata was investigated by immunohistochemical methods using well-defined monoclonal antibodies. Surgical biopsies of pancreata were obtained from kidney donors while the organs were still perfused. Type 1 chain structures were abundantly expressed including monosialylated(ms)- and disialylated(ds)-Le(a) antigens, which have previously been associated with cancer. Type 2 chain structures were represented by H and Le(y) antigens and to a lesser extent by the precursor structure N-acetyllactosamine, whereas Le(x), dimeric Le(x), and ms-Le(x) were only sporadically observed, in contrast to fetal pancreatic tissue, in which Le(x) has been found to be abundantly expressed. H chain 3 antigen was found in nearly all specimens, whereas precursor structures Tn, sial-Tn and T antigens were absent. Desialylation unmasked the T antigen in all specimens. Absence/masking of Tn, T and related antigens is of special interest, since these antigens are associated with tumor development in other tissues, and may be of importance in pancreatic cancer diagnostics in the future.

Effects of insulin and somatostatin on the growth and the colony formation of two human pancreatic cancer cell lines

The effects of insulin and somatostatin on the growth and the colony formation of two human pancreatic cancer cell lines, BxPC-3 and SOJ-6, were studied. The BxPC-3 cell line (American Type Culture Collection no. CRL 1687) was derived from a moderately differentiated pancreatic adenocarcinoma. The SOJ-6 cell line is a subclone of SOJ that was initiated from ascites of a well-differentiated pancreatic adenocarcinoma. Both cell lines express fetoacinar pancreatic antigen, an antigen that might be associated with early transformation stages. However, these lines have different proliferation and tumoral powers. SOJ-6 cells showed an almost twofold higher division rate over BxPC-3 cells when cultured in RPMI-1640 medium containing 10% fetal bovine serum. The tumorigenic degree of SOJ-6 cells, as assessed by tumor growth in nude mice, was about three times greater than that of BxPC-3. The in vitro growth of BxPC-3 cells was significantly promoted by insulin, and was slightly inhibited by somatostatin, whereas the growth of SOJ-6 cells was not influenced by these hormones. Using a clonogenic assay in soft agar, the average ratio of colony numbers formed by SOJ-6 and BxPC-3 was about 10/1, indicating a good correlation between the colony formation and tumorigenic degree in vivo. In this test, the number of colonies formed by BxPC-3 cells was increased about twofold in insulin-supplemented medium. On the other hand, somatostatin inhibited the colony formation by a factor of four to six. However, no hormonal modulation of the colony formation of SOJ-6 cells was observed. Our data show that pancreatic cancer cell lines respond differently to pancreatic hormones, and suggest that this may be correlated to a tumour stage or a tumour type.

The diagnostic value of the foetoacinar pancreatic (FAP) protein in cancer of the pancreas; a comparative study with CA19/9

The serum diagnostic value of the foeto-acinar pancreatic protein (FAP protein), an oncofoetal pancreatic antigen, was tested in 201 patients. Of these, 112 suffered from malignant disease (57 patients had pancreatic carcinoma and 55, extra-pancreatic malignancies) and 89 had benign disease (49 patients with hepato-pancreato-biliary disease and 40 with other benign disease). FAP protein was measured by a competitive radioimmunoassay. In this technique, the normal cut-off level was 10% inhibition. This was deducted from values in 32 normal sera. FAP protein levels superior to 10% inhibition were found in 86% of patients with pancreatic cancer, in 31% with non-pancreatic malignancy, in 69% with benign hepato-pancreato-biliary disease and in 20% with other benign diseases. Accordingly, sensitivity of FAP protein for pancreatic carcinoma was 86% and specificity, 66%. However, high FAP protein levels (greater than 30% inhibition) were almost exclusively seen in patients with pancreatic cancer. At this cut-off level, specificity increased to 95% but sensitivity decreased to 51%. Determination of the carbohydrate antigen CA19/9 was made in parallel by a commercially available assay. At the cut-off level of 37 u ml-1, CA19/9 in our serum panel had a sensitivity of 74% for pancreatic carcinoma and a specificity of 88%. In pancreatic cancer 55 out of 57 patients had elevated levels of either FAP protein or CA19/9 (sensitivity; 96%).

Differentiation antigens in fetal human pancreas. Reexpression in cancer

An antiserum was raised against pancreatic extracts obtained from human fetuses under 5 months of gestational age. After absorption with adult tissues, this antiserum specifically recognized antigens located in the cytoplasm of fetal pancreatic acini. All of the examined pancreatic tissues, ranging from 3 to 5 months of gestational age, showed a strong positive reaction of most of the acinar cells. The number of stained acini and the staining intensity gradually decreased from 5 months onwards and by the 7th-8th month only a few cells remained positive. Adult pancreas was completely negative as were a variety of normal adult and fetal tissues. This antiserum also reacted with tumor structures in 18/18 pancreatic adenocarcinomas as well as with pancreatic acini in the vicinity of tumor. Primary carcinomas of the liver, large bowel, stomach, breast, urinary bladder, lung and other localizations did not react with this antiserum. In some cases of chronic pancreatitis (3/12) a reaction was observed in a few acinar cells. Immunoblot assay after polyacrylamide electrophoresis revealed, in both fetuses and tumors, two main antigens of approximately 60 kDa and 110 kDa relative molecular weight. Several minor components were also observed. These results suggest that our polyclonal antiserum defines a new group of oncodevelopmental antigens with high organ specificity.

Isolation and characterization of two oncofetal glycoproteins from hamster pancreas using concanavalin A and preparative electrophoresis

Pancreatic fetal acinar antigens in the Syrian golden hamster, which are associated with development of the pancreas, have been previously described. In this study, two major antigens were isolated from fetal pancreas using affinity chromatography on Con A-Sepharose and preparative electrophoresis. Homogenates from fetal and adult pancreas were analyzed for their ability to bind to concanavalin A. This lectin allowed obtention of eluted fractions accounting for 2 and 0.7%, respectively, of the protein content in crude extracts. Concanavalin A-positive fraction from fetal pancreas contained two major carbohydrate-reactive glycoproteins of relative molecular weight (Mr) 80 000 and 58 000 in SDS-polyacrylamide gel electrophoresis. Both behaved as fetal antigens in nitrocellulose blot immunoassay. Similar experiments with chemically induced tumors of the pancreas led to a concanavalin A fraction containing the 80 and 58 kDa fetal glycoproteins; but in this case, the fraction was quite heterogeneous. Our data provide new support for the existence of differentiation antigens in the acinar cells of the pancreas, and indicate that two major ones are glycoproteins. Moreover, both are expressed in pancreatic tumors.

Expression by human fetal organs of organ-specific cancer neoantigens as measured by leukocyte adherence inhibition

Human cancers express organ-specific cancer neoantigens (OSN) as determined by in vitro leukocyte responses to extracts of cancers by the tumor host. In this study, we determined whether the OSNs were normal developmental proteins that were expressed by fetal organs and re-expressed with oncogenesis. Fetal extracts, principally of lung and colon but also of liver and kidney, were tested for their ability to induce leukocyte adherence inhibition (LAI) as compared to extracts from adult tissues of the same organ. Leukocytes from lung cancer patients showed positive LAI responses to 13- and 19-week fetal lung tissue. Likewise, leukocytes from colon cancer patients showed positive LAI responses to 14- and 19-week fetal colon tissue, whereas leukocytes from control subjects did not. Neither group responded positively to 21-week fetal organs. Criss-cross experiments showed that the fetal antigen was organ specific. Multiparous pregnant women showed positive LAI responses to cancer extracts but not to extracts from normal tissues of the same organ. The pattern of the LAI response was bell-shaped. Positive LAI responses to lung and breast cancer were detected at 4 to 7 months gestation and peaked at 5 months. To the fetal colon, LAI positive responses were detected at 5 to 8 months gestation, with the peak response at 6 months. The results indicate that OSN of cancers are also expressed by fetal organs and sufficient antigen is shed by fetal organs to sensitize pregnant women. Older fetal organs (21 weeks) and adult organs do not express an immunogenic or antigenic OSN.