Immunoglobulin E antibodies from pancreatic cancer patients mediate antibody-dependent cell-mediated cytotoxicity against pancreatic cancer cells

Abstract

In addition to allergy and parasitic infections, immunoglobulin E (IgE) has been shown recently to possess anti-viral and anti-cancer effects. We investigated serum levels of IgE, its low-affinity receptor, soluble CD23 (sCD23) in patients with pancreatic cancer and the effect of IgE against pancreatic cancer cells. Twelve patients were evaluated for pancreatic cancer by imaging and confirmed by biopsy. Fifteen healthy volunteers served as controls. Serum Igs (IgG, IgM, IgA, IgE) and sCD23 levels were determined (enzyme-linked immunosorbent assay, nephelometry) and the presence of cancer-specific IgE was assessed (fluorescence microscopy, Western blot). IgE anti-cancer activity was determined by antibody-dependent cell-mediated cytotoxicity (ADCC). Serum levels of IgE and sCD23 were elevated significantly in patients with pancreatic cancer versus controls, whereas no differences were observed in other Ig isotypes (IgG, IgM, IgA). Flow cytometry and immunofluorescence microscopy demonstrated similar presence of IgG and IgE pancreatic cancer Igs. However, Western blot analysis indicated differences in IgG and IgE antigen-specific antibodies; IgE antibody recognized a 50 kD protein. ADCC studies demonstrated that serum and purified IgE-mediated cytotoxicity against pancreatic cancer cells, effects which were reversed with anti-IgE neutralizing antibody and IgE depletion (immunoaffinity); greater cytotoxicity was observed in patient serum when compared with healthy controls. These data suggest that IgE and sCD23 may serve as useful biomarkers for patients with pancreatic cancer and may be important in the immune response to this disease in that IgE-directed therapy may help to direct treatment.

Fig. 1

Serum immunoglobulin (Ig)G, IgM and IgA in pancreatic cancer. Serum obtained from pancreatic cancer patients (n = 12) and healthy controls (n = 15) were assayed for IgG (top panel), IgM (middle panel) and IgA (bottom panel) by nephelometry (see Materials and methods). Data are expressed as mean g/l (± standard error). Significance between groups was determined by Student's t-test (P < 0·05).

Fig. 2

Serum immunoglobulin (Ig)E and sCD23 in pancreatic cancer. Serum obtained from pancreatic cancer patients (n = 12) and healthy controls (n = 15) were assayed for IgE (top panel) and soluble CD23 (bottom panel) by enzyme-linked immunosorbent assay. Data are expressed as mean (± standard error) (IgE-IU/ml; sCD23-U/ml). Significance between groups was determined by Student's t-test (P < 0·05).

Fig. 3

Presence of pancreatic cancer specific immunoglobulins (Ig) by immunofluorescence. Human pancreatic adenocarcinoma cell lines (HPAC) were cultured in the presence of serum from pancreatic cancer patients and healthy controls. IgE and IgG in serum were detected with fluorescein isothiocyanate-conjugated anti-isotype Ig. Data represent mean ± standard error of 10 sera per group. Significance between groups was determined by Student's t-test (P = not significant). Similar fluorescence patterns were detected when sera were cultured with other pancreatic cancer cell lines (MiaPaCa-2, pancreatic adenocarcinoma-1).

Fig. 4

Antigen-specific antibodies. (a) Human pancreatic adenocarcinoma cell lines (HPAC) lysates were blotted with serum obtained from pancreatic cancer patients (lanes 1 and 2) or healthy controls (lanes 3 and 4) and probed for immunoglobulin (Ig)E and IgG isotypes. (b) Lysates were obtained from MiaPaCa-2 pancreatic cancer cell line and blotted with serum from pancreatic cancer patients (lanes 1–3) or healthy control (lane 4). As shown in both figures, sera from patients with pancreatic cancer patients contained antibodies (IgG and IgE) which were specific for protein of ∼50 kD.

Fig. 5

Immunoglobulin (Ig)E-mediated antibody-dependent cell-mediated cytotoxicity: effect of neutralization. Human pancreatic adenocarcinoma cell lines (HPAC) were cultured in the presence of serum from pancreatic cancer patients and controls (n = 4/group) and cytotoxicity was determined after 18–20 h. In some experiments, anti-human IgE or IgG (10%) was added to serum before culture with target (HPAC) and effector (peripheral blood mononuclear) cells. Data were performed in triplicate with an effector : target ratio of 5 : 1 (using 5 × 10³ target cells) and 1% patient serum. Data are expressed as mean percentage of cytotoxicity as determined by lactate dehydrogenase release ± standard error of anti-IgE or IgG experiments compared with whole serum (cytotoxicity for whole serum alone averaged 17%, range 14–20%). Significance between groups was determined by Student's t-test, *P < 0·05.

Fig. 6

Immunoglobulin (Ig)E-mediated antibody-dependent cell-mediated cytotoxicity: effect of IgE depletion and purified IgE. Human pancreatic adenocarcinoma cell lines (HPAC) were cultured in the presence of either whole, IgE-depleted (immunoaffinity) serum or purified IgE obtained from (a) pancreatic cancer patients and (b) controls (n = 4/group) and cytotoxicity was determined after 18–20 h. Data were performed in triplicate with an effector : target ratio of 5 : 1 (using 5 × 10³ target cells). Purified IgE contained 20–50% of original whole serum IgE content. Data are expressed as mean percentage of cytotoxicity as determined by lactate dehydrogenase release ± standard error of anti-IgE or IgG experiments compared with whole serum (cytotoxicity for whole serum alone averaged 17%, range 14–20%). Significance between groups was determined by Student's t-test, *P < 0·05. W: whole serum; D: IgE-depleted serum; P: purified IgE.