Natural IgM Antibodies and Immunosurveillance Mechanisms against ...

[CANCER RESEARCH 63, 7995– 8005, November 15, 2003]

Natural IgM Antibodies and Immunosurveillance Mechanisms against Epithelial Cancer Cells in Humans Stephanie Bra¨ndlein, Tina Pohle, Nele Ruoff, Ewa Wozniak, Hans-Konrad Mu¨ller-Hermelink, and H. Peter Vollmers Institute of Pathology, University of Wu¨rzburg, Wu¨rzburg, Germany

ABSTRACT Malignancy is like a chronic disease, and the immune system is permanently involved in recognizing and eliminating the transformed cells. The human hybridoma technology offers the unique opportunity to study the mechanisms, structures, and targets involved in recognition and elimination of aberrant cells. Thousands of tumor-reactive human monoclonal antibodies were isolated by this technique from cancer patients and from healthy donors, and all of these antibodies were IgM antibodies; no IgG and IgA antibodies were found. Fourteen of these antibodies were selected for DNA sequence analysis, characterization of their binding patterns, and determination of their origin and genetics. All of the IgM antibodies studied expressed only few or no mutations at all (germ-line coded), bound to carbohydrates on modified tumor-specific receptors and induced apoptosis. The degree of cross-reactivity to other tumors correlated reciprocally with the number of mutations in coding regions. By using an anti-idiotypic antibody we were able to show that the IgM-producing cells were of CD5ⴙ B-cell origin. The data presented here indicate that the innate immunity and natural IgM antibodies play an important role in immunosurveillance mechanisms against epithelial tumors in humans.

INTRODUCTION During its lifetime each multicellular organism is permanently exposed to transformed cells, which arise spontaneously or by inducing factors. Even if not every transformed cell has the ability and potency for malignant behavior, the important question is not why malignant cells arise, but instead, why malignancy occurs so infrequently (1). An early recognition and a rapid elimination of transformed cells by immune mechanisms is essential for keeping an organism free of tumor cells. The innate or natural immunity, a part of the immune system, which has for a long time been defined by immunologists as unspecific background immunity, seems to be of greater importance than has been thought (2). The importance of natural killer cells and natural antibodies for the defense against bacterial infections has been shown in detail, but only very little data on immunity against transformed cells, supporting a specific role of the innate immunity in tumor defense of humans, has thus far been published (3–5). One reason for that lack of data is the fact that only xenoimmunizations were performed to study cellular and humoral immunity against tumor cells, and to define tumor-specific targets. The results, however, obtained after ⬃30 years of xenoimmunizations, were disappointing. No truly tumor-specific immune cells or antibodies and no truly tumor-specific receptors, e.g., receptors expressed exclusively on malignant cells, were described (6 – 8). We believe that tumor-specific “nonself” structures can only be detected in syngeneic systems. The conventional human hybridoma technology (Trioma technique) offers the unique possibility to study the humoral immunity against malignant cells and to characterize new tumor-related membrane receptors. Fully human monoclonal antibodies for diagnosis and therapy can be generated by this technique, and

the molecular mechanisms of tumor defense can be investigated (3–5, 9 –15). We described recently a series of natural human monoclonal antibodies, which were isolated from different cancer patients by using the human hybridoma technique. These antibodies were germ-line coded, tumor-specific, functionally active, and bound to carbohydrate structures (4, 5, 9, 13–15). By using this approach, we were also able to characterize two new receptors (13, 14). The receptor identified by the human monoclonal antibody SC-1, which was isolated from a patient with a signet ring cell carcinoma of the stomach, is a specific modified isoform of decay-accelerating factor-B, expressed exclusively on the membrane of stomach carcinoma cells (9, 13, 15). The antibody SC-1 reacts with almost all diffuse-type and with ⬃20% of intestinal-type adenocarcinomas, and induces apoptosis of stomach cancer cells (11–13, 15–19) both in vitro and in experimental in vivo systems (9, 11, 16). Clinical studies showed that SC-1 induced regression and apoptosis of primary stomach cancers without any toxic cross-reactivity with normal tissue (17). The fully human germ-line coded monoclonal IgM antibody PAM-1 was isolated from a patient with a stomach carcinoma, too (20). PAM-1 binds to a membrane receptor present on almost all of the epithelial cancers of every type and origin. When evaluated on nonmalignant tissue, the only PAM-1-specific reactivity was an intracellular binding to proteins in the Golgi apparatus of the kidney (14). The PAM-1 receptor was purified from tumor cell membrane extracts and was found to be a Mr 130,000 integral membrane glycoprotein (14), homologous to cysteine-rich fibroblast growth factor receptor 1, which has thus far only been detected and described in Golgi of embryonic chicken cells and in Chinese hamster ovary cells (21). This post-transcriptionally modified receptor is expressed on almost all epithelial cancers of every type and origin, but not on healthy tissue. It is also present on precursor lesions found in: Helicobacter pylori-induced gastritis, intestinal metaplasia, and dysplasia of the stomach; ulcerative colitis-related dysplasia and adenomas of the colon; Barrett metaplasia and dysplasia of the esophagus; squamous cell metaplasia and dysplasia of the lung; and cervical intraepithelial neoplasia I-III (5, 14). In this study we investigated the origin, reactivity patterns, and genetics of tumor-specific human monoclonal IgM antibodies isolated from cancer patients and healthy donors. We give evidence that innate immunity is not only responsible for the recognition and elimination of bacterial structures but also for the removal of transformed epithelial cells. MATERIALS AND METHODS

Somatic Hybridization. Hybridomas were prepared and cultured as described previously (9). Briefly, lymph nodes or spleens obtained from lung-, pancreas-, or colon-carcinoma patients, or healthy donors during surgery were prepared by mechanical disruption of the tissue. The resulting single cell suspension was fused at a 2:1 ratio with the heteromyeloma HAB-1 with Received 6/19/03; revised 8/1/03; accepted 9/4/03. PEG1500, followed by subsequent distribution on 24-well plates. Hybridomas The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with were cultured in RPMI 1640 containing 10% FCS and Hypoxanthin Amino18 U.S.C. Section 1734 solely to indicate this fact. pterin Thymidin (HAT) supplement (PAA, Pasching, Austria). After 4 – 6 weeks Request for reprints: H. Peter Vollmers, Institut fu¨r Pathologie, Universita¨t Wu¨rthe supernatants were screened for antibody production in an ELISA. The zburg, Josef-Schneider Str. 2, D-97080 Wu¨rzburg, Germany. Phone: 49-931-20147898; average growth rate was ⬃80%, ⬃50% of the grown clones produced immuFax: 49-931-20147798; E-mail: [email protected]. 7995

NATURAL IgM ANTIBODIES AND IMMUNOSURVEILLANCE

noglobulins. Positive clones were tested immunohistochemically on autologous tumor tissue sections, and clones, which showed a positive reaction, were subsequently recloned. cDNA Synthesis and Reverse Transcription-PCR. Total RNA was prepared as described previously (22). cDNA synthesis from total RNA obtained from hybridoma cell lines was performed with 5 ␮g total RNA using Life Technologies, Inc. (Eggenstein, Germany) Moloney murine leukemia virus reverse transcriptase according to the manufacturer’s instructions. The amplification of VH and VL genes was carried out in a 25 ␮l volume with 1.75 mM MgCl2, 0.4 pM primer, 200 ␮M of each dNTP, and 1 unit Taq polymerase (MBI Fermentas, St. Leon-Rot, Germany). The PCR products were amplified using the following cycle profiles: 95°C for 2 min, followed by 35 cycles of 94°C for 30 s, 65°C for 30 s (for VH3 and VH4 primers), 60°C for VH1, VH2, VH5, VH6, and 52°C for VL primers, respectively, with a final extension at 72°C for 4 min. Sequence Analysis. PCR products were purified by gel electrophoresis through 2% agarose (Roth, Karlsruhe, Germany) and extraction from gel using a Jetsorb gel extraction kit (Genomed, Bad Oeynhausen, Germany). Cloning of PCR fragments was performed with pCR-Script Amp SK(⫹) cloning kit (Stratagene, Heidelberg, Germany). Ten positive clones were sequenced using the DyeDeoxy Termination Cycle Sequencing kit (Applied BioSystems Inc., Weiterstadt, Germany) and analyzed with an automated DNA sequencer ABIPrism373. Both strands were sequenced using T3 and T7 primers. The sequences were analyzed using DNASIS for Windows software and IMGT/VQUEST databases. Glycosidase Assays. Detached and washed cells were resuspended in RPMI 1640 containing 10% FCS and incubated for 1 h on ice, then counted, and cytospins were prepared. After air-drying, cytospin preparations were acetone-fixed (10 min), washed, and incubated with 20 ␮U/ml O-glycosidase or 5 mU/ml N-glycosidase (Boehringer, Mannheim, Germany) for 4 h at 37°C. Then, slides were washed and stained immunohistochemically. Immunohistochemical Staining of Acetone-Fixed Cells. The cytospins were blocked for 30 min with PBS/BSA (0,1%). After washing with Tris/NaCl three times, the sections were incubated with the different human antibodies or mouse anticytokeratin 8 antibody diluted 1:50 with BSA/PBS (Dako, Hamburg, Germany) for 30 min as a positive control. Then they were washed three times with Tris/NaCl, followed by incubation with secondary antibodies (peroxidase-labeled rabbit antihuman or rabbit antimouse conjugate 1:50) for 30 min. After washing three times with Tris/NaCl and incubation in PBS for 10 min, staining was performed with diaminobenzidine (0.05%)-hydrogen peroxide (0.02%) for 10 min at room temperature. The reaction was stopped under running tap water, and the sections were counterstained with hematoxylin. After mounting with glycerol-gelatin, the sections were analyzed using light microscopy Immunohistochemical Staining of Paraffin Sections. Paraffin-embedded human tissues were sectioned (2 ␮m), deparaffinized, and heated in citric acid (pH 5.5) in a pressure cooker for 5 min. The sections were blocked with BSA (5 mg/ml) diluted in PBS for 30 min at room temperature. Treated sections were then incubated either with the different IgM antibodies (10 ␮g/ml) or with positive control antibodies (anticytokeratin 8 antibody or anticytokeratin 5/6 antibody; Dako, Hamburg, Germany; diluted 1:20 with BSA/PBS) for 2.5 h at 37°C in a humidified incubator. The sections were then washed three times with Tris/NaCl, followed by incubation with peroxidase-labeled rabbit antihuman IgM antibody diluted 1:50 in PBS containing 30% rabbit serum (for human IgM antibodies) or rabbit antimouse conjugate (Dako) diluted 1:50 in PBS containing 30% human AB plasma (for positive control antibodies) at room temperature for 1 h. After washing three times with Tris/NaCl, the tissue sections were incubated in PBS for 10 min before staining with diaminobenzidine (0.05%)-hydrogen peroxide (0.02%) for another 10 min at room temperature. The reaction was stopped under running tap water, and the sections were counterstained with hematoxylin. After mounting with glycerol-gelatin, the sections were analyzed using light microscopy. Immunohistochemical Staining of Cryo-Sections. The frozen human tissues were sectioned (4 ␮m), fixed in acetone, air-dried, and washed with Tris/NaCl. Then the cryo-sections were blocked with milk powder (3%) diluted in PBS for 30 min. After washing with Tris/NaCl three times, the sections were incubated with the different human antibodies, unrelated human monoclonal IgM as a negative control (Chrompure IgM, Dianova, in the same concentration) or mouse anticytokeratin 8 antibody diluted 1:50 with BSA/

PBS (Dako) for 30 min as a positive control. Then they were washed three times with Tris/NaCl, followed by incubation with secondary antibodies (peroxidase-labeled rabbit antihuman or rabbit antimouse conjugate 1:50) for 30 min. After washing three times with Tris/NaCl and incubation in PBS for 10 min, staining was performed with diaminobenzidine (0.05%)-hydrogen peroxide (0.02%) for 10 min at room temperature. The reaction was stopped under running tap water, and the sections were counterstained with hematoxylin. After mounting with glycerol-gelatin, the sections were analyzed using light microscopy MTT1 Proliferation Assay. The MTT assay using the human stomach adenocarcinoma cell line 23132/87 was performed as described (20). This carcinoma cell line was derived from a freshly prepared primary culture of a gastric tumor patient (23, 24). In general, for the tests early passages (⬍10) were used to avoid cell culture artifacts. Briefly, trypsinized cells were diluted to 1 ⫻ 106 cells/ml in complete growth medium, and 50 ␮l of cell suspension was added to each well of a 96-well plate. Fifty ␮l of the different antibodies and controls diluted with complete growth medium was then added to the wells, and the plates were incubated for 48 h at 37°C in a humidified incubator. To demonstrate normal growth, the cells were supplemented with complete growth medium (control 1). Unrelated human IgMs in the same concentration (Chrompure IgM; Dianova) served as the negative control (control 2). For measurement, 50 ␮l of MTT solution (5 mg/ml) were added to each well, and the plates were incubated for 30 min at 37°C. After incubation, the plates were centrifuged at 800 ⫻ g for 5 min followed by the removal of the MTT solution. The stained cell pellet was dissolved in 150 ␮l dimethylsulphoxide, and absorption was measured at wavelengths of 540 nm and 690 nm. Apoptosis Assay. The extent of antibody-induced apoptosis on human stomach carcinoma cells (23132/87) was analyzed by the Cell Death Detection ELISAPLUS kit (Roche, Mannheim, Germany). For this assay 1 ⫻ 104 tumor cells were plated on 96-well plates and incubated in the presence of the different IgM antibodies for 48 h at 37°C and 7% CO2 in a humidified incubator. To demonstrate normal growth, the cells were supplemented with complete growth medium (control 1). Unrelated human IgMs in the same concentration (Chrompure IgM; Dianova) served as the negative control (control 2). After incubation the cells were centrifuged for 10 min at 200 ⫻ g, and the supernatants were removed followed by an incubation with lysis buffer for 30 min at room temperature. After centrifugation the supernatants were transferred into a streptavidin-coated microtiter plate, and immunoreagent was added (mixture of 10% antihistone-biotin, 10% anti-DNA peroxidase (POD), and 80% incubation buffer) and incubated for 2 h at room temperature on a microtiter plate shaker at 250 rpm. After incubation, unbound components were removed by washing with incubation buffer. POD was determined photometrically with an ABTS substrate (1 ABTS tablet in 5-ml substrate buffer). The antibody-induced apoptosis was measured at 405 nm against ABTS solution as a blank (reference wavelength ⬃490 nm). Anti-Idiotypic Antibody. BALB/c mice were immunized three times with 50 ␮g purified SC-1 antibody, and after 14 days spleen cells were immortalized by fusion to the mouse myeloma cells (Ag-8). Resulting monoclonal antibodies were tested in an ELISA for binding on a panel of human IgM antibodies, including SC-1. The murine monoclonal IgG antibody produced by clone 6/22 was the only one, which was found to be specific for SC-1. Immunofluorescence Staining. For the double staining assay, cryosections of human lymphatic tissue were incubated with the anti-idiotypic antibody 6/22, a pan-B antibody and an anti-IgM antibody for 30 min at room temperature, and subsequently incubated with a rabbit-antimouse secondary antibody coupled with FITC or tetramethylrhodamine isothiocyanate for additional 60 min at room temperature. After mounting with glycerol-gelatin, the sections were analyzed using a fluorescence microscope.

RESULTS IgM Antibodies from Tumor Patients. Lymphocytes from spleen, lymph nodes, and the tumor itself obtained from ⬎60 cancer patients with different carcinomas served as sources for ⬃260 fusion experiments. No preselection regarding age and gender of patients, 1 The abbreviation used is: MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide.

7996


Table 1 Hybridization experiments

Origin of lymphocytes

Tumor Stomach (Diffuse type carcinoma)

Stomach (Intestinal type carcinoma)

Number of patients

Tumor staging/grading

Spleen

2

Tumor

5

Spleen

3

Tumor

5

T1N0, T3N2 G2, G3 T1N1-T4N1 G2, G3 T1N0–T2N2 G2–G3 T1N0–T3N1 G2–G3 T2N0 G3 T2N0 G2 T2N1 G2, G3 T1N0–T3N1 G2, G3 T1N1 G2 T1N0–T3N2 G2, G3 T1N0–T4N0 G2, G3 T2N0–T4N0 G2, G3 T1N0–T3N1 G2, G3

Esophagus (Squamous cell carcinoma)

Lymph node Spleen Tumor Tumor

1 1 7 20

Pancreas (Adenocarcinoma) Lung (Adenocarcinoma)

Spleen Lymph node

1 6

Lung (Squamous cell carcinoma)

Lymph node

7

Prostate (Adenocarcinoma)

Lymph node

8

Breast (Adenocarcinoma)

Lymph node

4

Colon (Adenocarcinoma)

tumor staging, and grading, was done. Comparing the lymphocyte sources, the best fusion efficacy was obtained with lymphocytes from spleens (Table 1). Altogether ⬎18,000 growing triomas were investigated in an ELISA test for antibody production, and about half of them produced human immunoglobulins. Triomas from stomach, pancreas, lung, and breast cancer patients produced an excess of antibodies of IgM isotype, whereas triomas from colon and prostate cancer patients were more balanced with regard to the amount of IgG and IgA immunoglobulin isotypes. The ELISA-positive clones were additionally tested for tumorspecific reactivity by immunohistochemistry on a panel of cancerous and noncancerous tissues. The results showed that all of the tumorspecific antibodies were of IgM isotype (for details see Table 1). Specificity of IgM. The reaction patterns of the human monoclonal antibodies were investigated in more detail by immunohistochemistry on a broad panel of cryo- and paraffin-embedded primary carcinomas and normal tissues. The tissues were not preselected regarding age and gender of the patients. Three to 30 different cases of each tumor type were tested. The specimens were taken from different tumor stages and grades to analyze reaction patterns of the antibodies for a spectrum as broad as possible of different epithelial malignancies. The tested antibodies showed a heterogeneous reactivity pattern with tumor tissues. The antibodies PAM-1, LM-1, PM-2, SAM-3, SAM-4, SAM-6, PM-1, and CM-1 reacted with most of the carcinomas tested in this study, whereas the reactivity of the antibodies SC-1, CM-2, and SAM-2 was restricted to specific carcinomas (for details

ELISA positive immunoglobins

Tumor-specific immunoglobins

Number of fusions (⫻100)

Growing clones

IgM

IgG

IgA

IgM

IgG

IgA

120

9412

3278

578

644

131

0

0

5

247

86

49

24

3

0

0

45

3607

1191

275

239

42

0

0

5

149

63

17

22

2

0

0

3 15 7 20

93 615 273 993

13 227 48 163

11 166 39 59

8 152 44 76

1 42 7 19

0 0 0 0

0 0 0 0

7 16

546 567

106 78

21 55

20 37

27 11

0 0

0 0

22

863

156

62

33

20

0

0

14

560

90

84

91

23

0

0

6

85

15

6

4

3

0

0

see Table 2). Fig. 1 shows some examples of tumor-specific staining patterns of several antibodies. All of the tested antibodies showed a specific intensive staining of tumor cells, whereas the surrounding tissue was not stained. On normal tissues none of the antibodies exhibited any binding activity (data not shown). Targets of IgM. In addition to the specificity analysis on tissues we performed some studies on the antibody reactivity. Previously published results indicated that the tumor-specific IgM antibodies SC-1 and PAM-1 bind to carbohydrate residues on the corresponding receptors, which are post-transcriptionally modified (13, 14). On the basis of these data the binding characteristics of selected IgM antibodies were analyzed. Cytospin preparations of tumor cells were incubated with O- and N-glycosidase, and after that tested by immunohistochemistry for binding of the IgM antibodies. It was found that four of the five antibodies tested bound to a carbohydrate residue on specific receptors. For example, as shown in Fig. 2, exemplary treatment of lung carcinoma cells with N-glycosidase led to a dramatic decrease in LM-1 antibody staining (Fig. 2D), whereas incubation with dephosphorylation buffer or digestion with O-glycosidase (data not shown) had no effect on the binding of the antibody LM-1. Genetics of IgM. To investigate the genetic origin of these human monoclonal IgM antibodies, the variable gene regions of the immunoglobulin heavy (VH) and light chain (VL) genes were amplified, cloned, and sequenced. The sequences were compared with germ-line sequences of the corresponding germ-line genes to detect somatic mutations. The results are presented in Table 3, A and B. The degree

Table 2 Reactivity pattern of human monoclonal IgM antibodies on tumor tissues Tissue

Carcinoma type

PAM-1 ⫹/⫺

LM-1 ⫹/⫺

PM-2 ⫹/⫺

SAM-3 ⫹/⫺

SAM-4 ⫹/⫺

SAM-6 ⫹/⫺

PM-1 ⫹/⫺

CM-1 ⫹/⫺

SAM-1 ⫹/⫺

SAM-5 ⫹/⫺

CM-2 ⫹/⫺

SC-1 ⫹/⫺

SAM-2 ⫹/⫺

Esophagus

Squamous cell Adeno (Barrett) Adeno (diffuse) Adeno (intestinal) Adeno Adeno (ductal) Adeno Squamous cell Invasive (ductal) Invasive (lobular) Adeno Adeno Adeno

5/0 11/2 5/0 4/1 13/0 8/0 8/3 8/1 5/0 5/0 7/1 8/1 9/0

3/0 4/0 4/0 2/1 3/0 6/0 5/1 6/0 4/0 4/0 3/0 3/0 7/0

3/0 4/0 4/0 3/0 3/0 6/0 3/0 3/0 4/0 4/0 3/0 3/0 6/1

3/0 4/0 4/0 3/0 3/0 5/1 5/1 6/0 4/0 3/1 3/0 3/0 7/0

3/0 4/0 4/0 2/1 2/1 6/0 4/0 3/0 4/0 4/0 3/0 3/0 7/0

3/0 4/0 4/0 3/0 3/0 6/0 3/0 3/1 4/0 4/0 3/0 4/0 5/2

3/0 4/0 4/0 3/0 2/1 5/1 4/1 3/2 3/1 3/1 2/1 2/2 3/4

3/0 4/0 2/2 3/0 25/2 3/5 3/1 3/1 4/0 3/2 2/1 3/0 4/3

5/0 4/0 2/2 2/1 3/0 4/2 3/1 2/2 3/1 2/2 0/3 2/1 5/2

3/0 2/2 3/1 3/0 0/3 3/3 2/2 3/2 3/1 3/1 0/3 3/0 6/1

0/3 0/4 0/4 0/3 11/16 0/6 0/5 0/5 0/4 0/4 0/3 0/3 0/7

0/3 0/4 15/15 4/13 0/3 0/6 0/4 0/3 0/4 0/4 0/3 0/3 0/4

5/0 0/4 0/4 0/3 0/3 0/6 0/5 0/5 0/4 0/4 0/3 0/3 0/7

Stomach Colon Pancreas Lung Breast Ovary Uterus Prostate

7997


Fig. 1. Immunohistochemical staining of antibodies SAM-1 to SAM-6 on different carcinomas. Tumor tissue sections were stained with positive control (anticytokeratin 5/6 for esophageal squamous cell carcinoma, anticytokeratin 8 for invasive lobular adenocarcinoma of the breast, adenocarcinoma of the lung, adenocarcinoma of the uterus, and gastric adenocarcinoma), unrelated human IgM antibody as a negative control and the different IgM antibodies: A, antibody SAM-1 on adenocarcinoma of the lung; B, antibody SAM-2 on gastric adenocarcinoma; C, antibody SAM-3 on esophageal squamous cell carcinoma; D, antibody SAM-4 on gastric adenocarcinoma; E, antibody SAM-5 on adenocarcinoma of the uterus; F, antibody SAM-6 on invasive lobular adenocarcinoma of the breast. Original magnification, ⫻100.

of identity of the nucleotide sequences of the VH segment to those of the closest reported germ-line VH genes ranged from 92.3 to 100%, as summarized in Table 3A. The antibodies were expressed by different VH genes of the VH3, and VH5 gene family. Interestingly an accumulation of several families was observed. Most of the antibodies were expressed by VH genes of DP49 and DP47. The close homology of the VH regions to the germ-line genes and the low R:S ratio, which is an indicator for affinity maturation of antibodies, indicated that none of the antibodies studied was affinity maturated by somatic mutation due to antigen contact. The degree of identity of the nucleotide sequences of the VL segment to their most homologous VL germ-line genes ranged from 94.8 to 100% (Table 3B). All but one of the antibodies used ␭-light chain genes. The R:S ratio was again low. The antibodies were expressed by different VL genes of the VL1, VL2, VL3, and VL5 gene family. A preference of the VL3 gene family was observed, but other gene families were used, too.

Functional Activity of IgM. To examine the functional activity of the antibodies isolated from cancer patients in vitro, we used the colorimetric mitochondrial hydroxylase assay (MTT), which is a standard assay for proliferation (13). In this assay primary cultures of human carcinoma cells were incubated with the IgM antibodies SAM-4, SAM-5, and SAM-6. After 48 h all of the tested antibodies inhibited cell proliferation, whereas in the controls with human unrelated IgM no inhibition of cell proliferation was observable (Fig. 3A). For additional investigation of the inhibitory effect of the antibodies on tumor cell growth, the Cell Death Detection ELISAPLUS kit was used. This apoptosis assay enables a clear distinction between necrosis and specific antibody-induced apoptosis. The experiments clearly showed that each antibody induced apoptosis in carcinoma cells after 48 h of incubation (Fig. 3B). IgM Antibodies from Healthy Donors. To prove that natural IgM antibodies with tumor specificity exist in humans, spleen cells from

7998


from healthy donors, the VH and VL genes were amplified, cloned, and sequenced. The sequences were compared with germ-line sequences in the IMGT/V-QUEST database to identify the most homologous germ-line genes and to detect somatic mutations. The results are presented in Table 4. The degree of identity of the nucleotide

Fig. 2. N-Glycosidase treatment of lung carcinoma cells. Cytospin preparations of the lung adenocarcinoma cell line Colo-699 were treated with N-Glycosidase and stained with anticytokeratin 8 and antibody LM-1. A, anticytokeratin 8 before treatment; B, antibody LM-1 before treatment; C, anticytokeratin 8 after treatment; and D, antibody LM-1 after treatment with N-glycosidase.

healthy donors were immortalized with the heteromyeloma HAB-1, and the resulting antibodies were tested for antitumor activity. From a small series of clones two IgM antibodies were selected, which showed a positive reaction with different tumor tissues but not with normal nontransformed cells. Figs. 4 and 5 show an example of the reactivity patterns of the antibodies NORM-1 and NORM-2 on several tumor tissues in comparison with staining data on normal tissues of the same organs. The antibodies isolated from healthy donors showed a specific staining of tumor cells, whereas the surrounding tissue and normal tissue was not stained. Genetics of NORM-1 and NORM-2. To investigate the genetic origin of the human monoclonal IgM antibodies, which were isolated

Fig. 3. Functional analysis of antibodies SAM-4, SAM-5, and SAM-6. The consequences of antibody treatment on the proliferation of gastric carcinoma cell line 23132/87 were measured by (A) MTT proliferation assay and (B) Cell Death Detection ELISAPLUS. A, inhibition of cell proliferation with antibodies SAM-4, SAM-5, and SAM-6 measured after 48 h. B, antibody induced apoptosis detected by Cell Death Detection ELISAPLUS apoptosis assay after 48 h. Control 1, to demonstrate normal conditions, cells were supplemented with complete growth medium but without any antibody. Control 2, unrelated human IgM was added in a similar concentration.

Table 3 Sequence-analysis of human monoclonal IgM antibodies isolated from cancer patients A. Heavy chain Antibody SC-1 PAM-1 CM-1 SAM-2 SAM-5 SAM-3 SAM-6 SAM-4 PM-2 PM-1 SAM-1 CM-2

VH-family

Germ-line gene

Homology [%]

R:S FR

R:S CDR

VH3 VH3 VH3 VH3 VH3 VH3 VH3 VH3 VH3 VH3 VH3 VH5

DP-49, (IgHV3-30/3-30.5*01) DP-49, (IgHV3-30/3-30.5*01) DP-49, (IgHV3-30/3-30.5*01) DP-49, (IgHV3-30/30.5*01) DP-49, (IgHV3-30/30.5*01) DP-49, (IgHV3-30/30.5*01) (IgHV3-30.3*01) DP-47 (IgHV3-23) DP-47, (IgHV3-23*01) DP-47, (IgHV3-23*01) DP-42 (IgHV3-53) DP-73, (IgHV5-51*01)

97 100 100 92,3 99,3 100 100 100 100 100 99,6 97,2

4/2 0/0 0/0 2/2 0/0 0/0 0/0 0/0 0/0 0/0 0/0 2/1

2/0 0/0 0/0 5/7 0/2 0/0 0/0 0/0 0/0 0/0 0/1 3/2

Homology [%]

R:S FR

R:S CDR

96,8 100 98,6 99,6 98,7 99,0 99,3 98,6 94,8 97,2 98,2 —

3/2 0/0 1/1 1/0 3/0 2/1 1/1 3/0 6/3 0/2 3/2 —

3/0 0/0 2/0 0/0 0/1 0/0 0/0 0/1 4/2 4/2 0/0 —

B. Light chain Antibody SC-1 SAM-1 SAM-2 SAM-6 PAM-1 PM-1 CM-1 SAM-4 SAM-5 CM-2 PM-2 SAM-3

VL-family

␭ ␭ ␭ ␭ ␭ ␭ ␭ ␭ ␭ ␭ ␭ ␬

Germ-line gene DPL-23, (IgLV3-1*01) DPL-23, (IgLV3-1*01) DPL-23, (IgLV3-1*01) DPL-23, (IGLV3-1*01) LV318, (IgLV3-21*03) (IgLV3-10*01) (IgLV3-25*03) DPL-8, (IgLV1-40*01) DPL-5, (IgLV1-51*01) DPL-11, (IgLV2-14*01) (IgLV5-45*01) —

7999


proliferation was measured by photospectrometry. Fig. 6A shows that the antibodies isolated from healthy donors inhibited tumor cell proliferation after 48 h, whereas the controls with unrelated human IgM did not show any growth inhibition. To additionally investigate the inhibitory effect of the antibodies NORM-1 and NORM-2 on tumor cell growth, the Cell Death Detection ELISAPLUS kit was used. The results showed that each antibody induced apoptosis in carcinoma cells after 48 h of incubation (Fig. 6B). Origin of Natural IgM Antibodies. Natural IgM antibodies are produced by CD5⫹ B cells (B1 cells), which belong to the innate immunity and first defense effectors against pathogenic organisms. These B1 cells represent a unique set of lymphocytes, which differ phylogenetically, phenotypically, and functionally from the conventional B2-lymphocytes. We used the murine anti-SC-1 idiotypic antibody 6/22 to identify the B-cell clone carrying the SC-1 immunoglobulin receptor by im-

Fig. 4. Immunohistochemical staining of antibody NORM-1 on different carcinomas and normal tissues. Tumor tissue sections were stained with positive control (anticytokeratin 7 for pancreatic adenocarcinoma, and anticytokeratin 8 for adenocarcinoma of stomach and colon), unrelated human IgM antibody as a negative control and the IgM antibody NORM-1. A, adenocarcinoma of the colon; B, diffuse type stomach carcinoma; C, adenocarcinoma of the pancreas. Normal tissue sections were stained with H&E, unrelated human IgM antibody as a negative control and the IgM antibody NORM-1: D, normal colon tissue; E, normal gastric tissue; F, normal pancreas tissue. Original magnification, ⫻100.

sequences of the VH segment to those of the closest reported germline VH genes was 100%. VH genes of the VH3 gene family expressed the antibodies. The close homology of the VH regions to the germ-line genes and the low R:S ratio again indicated that none of these antibodies were affinity maturated by somatic mutation due to antigen contact. The degree of identity of the nucleotide sequences of the VL segment to their most homologous VL germ-line genes ranged from 99.3 to 99.6%, with both antibodies using ␭-light chain genes. The R:S ratio was low again, and the mutations were restricted to the framework region. Functional Activity of NORM-1 and NORM-2. To examine whether the antibodies isolated from healthy donors showed a functional activity in vitro, the MTT assay was used again. The tumor cells were incubated with antibody NORM-1 and NORM-2 for 48 h. Thereafter the amount of antibody-induced inhibition of tumor cell

Fig. 5. Immunohistochemical staining of antibody NORM-2 on different carcinomas and normal tissues. Tumor tissue sections were stained with positive control (anticytokeratin 8), unrelated human IgM antibody as a negative control, and the IgM antibody NORM-2: A, diffuse type stomach carcinoma; B, adenocarcinoma of the lung; C, adenocarcinoma of the colon. Normal tissue sections were stained with H&E, unrelated human IgM antibody as a negative control, and the IgM antibody NORM-2: D, normal gastric tissue; E, normal lung tissue; F, normal colon tissue. Original magnification, ⫻100.

8000


Table 4 Sequence-analysis of human monoclonal IgM antibodies isolated from healthy donors A. Heavy chain Antibody

VH-family

Germ-line gene

Homology [%]

R:S FR

R:S CDR

NORM-1 NORM-2

VH3 VH3

DP-47, (IgHV3-23*01) DP-77, (IgHV3-21*01)

100 100

0/0 0/0

0/0 0/0

Homology [%]

R:S FR

R:S CDR

99,3 99,6

1/1 1/0

0/0 0/0

B. Light chain Antibody

VL-family

NORM-1 NORM-2

␭ ␭

Germ-line gene IgLV3-10*01 IgLV1-40*01

munofluorescence double-staining experiments. As illustrated in Fig. 7A, the SC-1-producing cell was CD5 positive as well as IgM positive. Similar stainings were done on different lymphatic material of other patients and healthy donors. Fig. 7B clearly shows that this CD5/SC-1-positive cell was also present in lymphoid tissue of other cancer patients and in addition in healthy persons. These data demonstrate that the SC-1 antibody is produced by a B1/CD5⫹ cell and is part of the natural immunity. Reactivity Pattern of Natural IgM Antibodies. The reactivity pattern of the antibodies was investigated in detail by immunohistochemistry on a variety of paraffin- and cryo-embedded carcinomas and normal tissues. On normal tissues no specific bindings could be observed. The antibodies PAM-1, LM-1, PM-2, SAM-3, SAM-4, and SAM-6, for example, reacted with a lot of the tested carcinomas, whereas the reaction of antibody CM-2 was restricted to colon carcinomas, and the antibodies SC-1 and SAM-2 reacted only with stomach carcinomas (for details see Table 2).

Fig. 6. Functional analysis of antibodies NORM-1 and NORM-2 in vitro. The consequences of antibody treatment on the proliferation of gastric carcinoma cell line 23132/87 were measured by (A) MTT proliferation assay and (B) Cell Death Detection ELISAPLUS: A, inhibition of cell proliferation with antibodies NORM-1 and NORM-2 measured after 48 h. B, antibody induced apoptosis detected by Cell Death Detection ELISAPLUS apoptosis assay after 48 h. Control 1, to demonstrate normal conditions, cells were supplemented with complete growth medium but without any antibody. Control 2, unrelated human IgM was added in a similar concentration.

The reactivity of all of the sequenced human monoclonal tumorspecific IgM antibodies was compared with the mutation rate. Interestingly, there was a striking correlation between the number of mutations and the reactivity pattern. Germ-line antibodies without mutations (e.g., PAM-1, PM-2, SAM-3, and SAM-4) always reacted with a broad spectrum of different tumors within their VH family, whereas the spectrum of reactivity decreased with mutational events (e.g., SC-1, CM-2, SAM-1, SAM-5, and SAM-2). Antibody PAM-1 (100% homology to germ-line gene) for example bound to almost all of the tumors tested, whereas antibody SC-1 (97% homology to germ-line gene) bound only to stomach cancer cells. This indicates that few mutational events determine the reactivity pattern of the antibodies. These observations are illustrated in Fig. 8. A remarkable reciprocal correlation was detected with regard to antibody reactivity and the amount of mutational events. DISCUSSION The conventional human hybridoma technology is an optimal approach to study the humoral immunity of humans against transformed cells. Immortalizing human lymphocytes by somatic hybridization generates human monoclonal antibodies, which will not only provide new therapeutic human antibodies in the fight against cancer, but will, even more importantly, also define important new targets on cancer tissues and give new information about the humoral immunity of humans against cancer. Our investigation and analysis of several thousands of tumor-reactive human monoclonal antibodies from cancer patients and healthy donors showed that all of the tumor-specific antibodies had several features in common. All of the tumor-specific human monoclonal antibodies isolated from cancer patients, regardless of the tumor-type or the lymphocyte source, were of IgM isotype. No tumor-specific IgGs or IgAs and no affinity-maturated tumor-specific antibodies were isolated. All of the IgM antibodies were germ-line coded and belonged to distinct gene families. In addition, most of them bound to carbohydrates and were functionally active in vitro by inducing apoptosis. Furthermore, our data showed that tumor-specific antibodies were also present in healthy donors, and these antibodies were solely of one isotype, namely IgM, too. The reactivity pattern of all of these IgMs correlated reciprocally with their mutational status, i.e., with increasing mutation frequency the grade of reactivity to other tumors decreased. The origin of one of these IgM antibodies, SC-1, was determined by using an anti-idiotypic antibody, and we demonstrated that the human antibody SC-1 was produced by CD5⫹ lymphocytes and that these SC-1 producing cells were also present in lymphatic tissue of healthy donors. Malignancy is like a chronic disease, and evolution had to develop a defense system that enabled organisms to detect and to remove transformed cells early and sufficiently. Components of the innate and the adapted immunity are responsible for the elimination of “nonself” cells. In invertebrates immunity is restricted to the innate system, and

8001


Fig. 7. Determination of SC-1/CD5⫹ origin by immunofluorescence double staining on spleen cells by using an anti-SC-1 idiotypic antibody (A) and determination of SC-1-producing cells on different lymphatic tissues from healthy donors by using immunoperoxidase staining with anti SC-1 idiotypic antibody (B). A, to investigate the origin of antibody SC-1 producing cell, immunofluorescence double staining was performed with an anti-SC-1 idiotypic antibody (clone 6/22), a FITC labeled antihuman IgM antibody, and tetramethylrhodamine isothiocyanate-labeled antihuman IgM and anti-CD5⫹ antibodies. The staining was analyzed using a fluorescence microscope. The same cells, which are positive for SC-1, are positive for human IgM and for CD5 as well. B, to investigate the existence of SC-1-producing CD5⫹ cells in nonautologous lymphatic tissue, immunoperoxidase staining was performed with antiSC-1 idiotypic antibody on autologous spleen, on a spleen from a SC-1 negative patient, and on a spleen from a healthy donor. Negative control, unrelated human IgM antibody.

vertebrates are equipped in addition with the adaptive, trainable immune system (25–28). The innate response is invariable and works by using a transmitted germ-line coded pool of specific receptors. The adapted response is derived from the first response, and its diversity is based on mutational events. The innate immunity (also referred to as “unspecific immunity”) consists of natural killer cells, dendritic and mast cells, macrophages, and natural IgM antibody producing B cells (29 – 40). This system can distinguish between self and nonself, and is responsible for the first specific immune response directed against bacteria, viruses, and malignant cells (40 – 42). This response is T-cell independent, which means that antigen presentation by T cells is not required (43). The cells involved in innate immunity do not recognize specific single structures but instead specific patterns, which are conservative and expressed independently from mutational events (44). This recognition system guarantees that the immune response need not follow all of the mutational changes, but can focus on the detection of structures, which are most likely involved in primary cell stability and cell preservation mechanisms. The immune-competent cells fulfill that task by using specific sets of germ-line coded receptors, which belong to distinct protein families (44, 45). The acquired genetic

variability of the innate immunoglobulin receptors is achieved by combinatorial association of germ-line immunoglobulin genes. Additional deletions, additions, and mistakes in recombination events guarantee a genetic and receptor variability, which is sufficient to cover a broad spectrum of antigens on pathogenic organisms and gives a sufficient protection without additional mutational adaptation (46 –52). Thus, the innate immune recognition is entirely different from the diverse recognition of the adaptive immunity. The most potent molecules of the immune system are pentameric IgM antibodies. Their unique structure is the prerequisite for the specific recognition and function of the innate immunity. The antibodies described here were selected in vitro for their antiproliferative activity. Most of the investigated human IgM antibodies isolated from cancer patients and from healthy donors induced apoptosis by binding to tumor-specific receptors (see also Refs. 4, 15). This effect was demonstrated on different established tumor cell lines (4) and on primary tumor cells isolated from cancer patients (9, 11, 23, 24). In addition, the human monoclonal IgM antibody SC-1 showed its apoptotic effect not only in vitro, but also in vivo, in animals as well as in patients when used in a clinical study of stomach cancer (13, 18). In contrast, former published IgM antibodies, which also bind specif-

8002


Fig. 8. Reciprocal correlation of antibody specificity versus genetic variation. Overall mutations in variable heavy and light immunoglobulin chain regions are compared with the amount of reactivity of the several antibodies on different tumor tissues. Antibodies without any mutations show a very broad reactivity on the tested carcinomas, whereas the reactivity of antibodies with mutations is restricted to several tumor types.

ically to the cell surface of epithelial cancer cells, did not induce apoptosis (14, 53). Therefore, it is unlikely that the apoptotic activity was the result of in vitro selection processes. It seems likely that this apoptotic activity is specific for most of these natural IgM antibodies and that the type of receptor and the specific epitope are important for inducing the apoptotic activity, not simply the membrane binding. However, IgM antibodies are excellent complement activators as well, and it is possible that they remove aberrant cells also via complement activation or antibody-dependent cellular cytotoxicity. Because we did not select the antibodies for complement activation, we cannot exclude that both the isolated apoptosis inducing human antibodies and the antibodies that do not induce apoptosis in vitro show cell lysis mechanisms like complement activation in vivo. Abnormal and modified glycolipid and glycoprotein synthesis by tumor cells is a common feature of malignant cells, and very often results in the expression of these modified structures on the surface (54, 55). It is also known that these structures elicit a strong immune response (56). It is likely that most of the human monoclonal IgM antibodies described in this study bound most likely to epitopes on carbohydrate structures of specific receptors, because the digestion with endo-glycosidases led to a significant decrease of antibody binding to the tumor cell. A tumor-specific expression and antibody recognition of carbohydrate epitopes was already described for the human IgM antibodies SC-1 and PAM-1 (13, 14). These two natural immunoglobulins bind to carbohydrate epitopes present on a tumorspecific variant of CD55/decay-accelerating factor (SC-1) and on a new post-transcriptionally modified isoform of cysteine-rich fibroblast growth factor receptor 1 (PAM-1), respectively. Removal of these epitopes by glycosidase treatment diminished the binding of the antibodies. The exact carbohydrate structures and modifications, detected by the human antibodies, as well as the mechanisms leading to cell death are the subject of additional investigations. Natural IgM antibodies are produced by CD5⫹ B cells, which belong to the innate immunity and first defense effectors against pathogenic organisms. These B1 cells represent a unique set of lymphocytes, which differ phylogenetically, phenotypically, and functionally from the conventional B2 lymphocytes (38, 57– 66). B1 cells arise in the organism before B2 cells, they represent the major pool of neonatal B cells, and they decrease with increasing amounts of B2 cells (63, 67– 69). A typical marker for B1 cells is the expression of the des pan T glycoprotein CD5, together with membrane-bound IgM, IgD, B220, CD23, and CD43 (70). In normal individuals these B1 (CD5⫹) cells are responsible for the production of natural IgM and IgA antibodies (38, 71–75).

We verified the hypothesis of the CD5⫹ origin of natural antibodies by using an anti-idiotypic antibody against the SC-1 IgM antibody. This anti-SC-1 idiotypic antibody was made in mice and is a specific SC-1 antibody. We used this murine antibody to identify the SC-1 immunoglobulin receptor-positive B-cell clone with immunofluorescence double-staining experiments, and we successfully demonstrated that the SC-1-producing cell is CD5 positive. In addition, similar immunoperoxidase stainings were done on lymphoid material of other patients and healthy donors, and the results showed that this CD5⫹/ SC-1-positive cell was also present in lymphoid tissue of healthy persons. From those results we concluded that the SC-1 antibody is made by a CD5⫹ B cell (B1 cell) and that it is part of the natural immunity. A good evidence for the theory of tumor-specific IgM antibodies being inherited natural antibodies is the existence of these antibodies in healthy persons. To prove that natural IgM antibodies with tumor specificity exist as an immunological reactivity platform in humans, spleen cells from “healthy” individuals were immortalized and the resulting antibodies tested for antitumor activity. Among a small series of selected clones two IgM antibodies could be identified, which showed a reactivity against different tumor tissues but not against nontransformed cells. Moreover, these antibodies showed similar genetic origins as the tumor-specific antibodies isolated from cancer patients. The sequence analysis showed that the antibodies isolated from healthy donors were members of the family of naturally occurring, nonaffinity maturated antibodies, too. The in vitro functional analysis of the tested human monoclonal IgM antibodies isolated from healthy donors demonstrated that these antibodies, like the antibodies found in cancer patients, not only reduced cell proliferation, but also induced apoptosis in cancer cells. IgM antibodies with tumor reactivity were detected in sera of several healthy donors. But it was neither proven whether these antibodies were really natural antibodies nor was it excluded that these antibodies were autoantibodies (40, 45, 76 – 81). Without specificity and genetic analysis it is not possible to classify these observations. The data presented here indicate that the same group of naturally occurring IgM antibodies found both in healthy individuals and in cancer patients are responsible for the defense against malignant cells. The reactivity of all of the sequenced human monoclonal tumorspecific IgM antibodies in this study was compared with the mutation rate. Interestingly, there was a striking correlation between the number of mutations and the reactivity pattern. Germ-line coded antibodies with no mutations (PAM-1, PM-2, SAM-3, and SAM-4) always reacted with a broad spectrum of different tumors within their VHfamily, whereas the spectrum of reactivity decreased with mutational events (SC-1, CM-2, and SAM-5). Antibody PAM-1 (100% homology to germ-line gene) for example bound to nearly all of the tumors tested, whereas antibody SC-1 (97% homology to germ-line gene) bound to stomach cancer cells only. This indicates that few mutational events might determine the reactivity pattern, and the most important question is whether these mutations are the result of inheritance or of learning, or whether the natural immunity is trainable, like the adaptive. In this article we give evidence that origin, reactivity patterns, and genetics of these truly tumor-specific antibodies are close to what we know about the defense against bacterial structures. Therefore, these data might indicate that the innate immunity is also responsible for immunosurveillance mechanisms against malignant epithelial cells.

ACKNOWLEDGMENTS

8003

We thank Annette Vollmers for improving the manuscript.


REFERENCES 1. Alberts, B., Bray, D., Lewis, J., Raff, M., Roberts, K., and Watson, J. D. Cancer. In: M. Robertson, R. Adams, S. M. Cobert, D. Goertzen, (eds.), Molecular Biology of the Cell, Ed. 3 ed., pp. 1255–1291. New York: Garland Publishing, Inc., 1994. 2. Casali, P., and Notkins, A. L. CD5⫹ B lymphocytes, polyreactive antibodies and the human B-cell repertoire. Immunol. Today, 10: 364 –368, 1989. 3. Vollmers, H. P., and Bra¨ ndlein, S. Nature⬘s best weapons to fight cancer: reloading human monoclonal IgM antibodies. Hum. Antibodies, 11: 131–142, 2002. 4. Bra¨ ndlein, S., Lorenz, J., Ruoff, N., Hensel, F., Beyer, I., Mu¨ ller, J., Neukam, K., Eck, M., Illert, B., Mu¨ ller-Hermelink, H. K., and Vollmers, H. P. Human monoclonal IgM antibodies with apoptotic activity isolated from cancer patients. Hum. Antibodies, 11: 107–119, 2002. 5. Bra¨ ndlein, S., Beyer, I., Eck, M., Bernhardt, W., Hensel, F., Mu¨ ller-Hermelink, H. K., and Vollmers, H. P. CFR-1, a new marker for precancerous epithelial lesions defined by the human monoclonal antibody PAM-1. Cancer Res. 63, 2052–2061, 200. 6. Canney, P. A., Moore, M., Wilkinson, P. M., and James, R. D. Ovarian cancer antigen CA125: a prospective clinical assessment of its role as a tumour marker. Br. J. Cancer, 50: 765–769, 1984. 7. Mayer, A., Chester, K. A., Flynn, A. A., and Begent, R. H. Taking engineered anti-CEA antibodies to the clinic. J. Immunol. Methods, 231: 261–273, 1999. 8. de Bono, J. S., and Rowinsky, E. K. The ErbB receptor family: a therapeutic target for cancer. Trends Mol. Med., 8: 19 –26, 2002. 9. Vollmers, H. P., O’Connor, R., Mu¨ ller, J., Kirchner, T., and Mu¨ ller-Hermelink, H. K. SC-1, a functional human monoclonal antibody against autologous stomach carcinoma cells. Cancer Res., 49: 2471–2476, 1989. 10. Pfaff, M., O’Connor, R., Vollmers, H. P., and Muller-Hermelink, H. K. Human monoclonal antibody against a tissue polypeptide antigen-related protein from a patient with a signet-ring cell carcinoma of the stomach. Cancer Res., 50: 5192–5198, 1990. 11. Vollmers, H. P., Da¨ mmrich, J., Ribbert, H., Wozniak, E., and Mu¨ ller-Hermelink, H. K. Apoptosis of stomach carcinoma cells induced by a human monoclonal antibody. Cancer (Phila.), 76: 550 –558, 1995. 12. Vollmers, H. P., Hensel, F., Hermann, R., Dammrich, J., Wozniak, E., Gessner, P., Herrmann, B., Zimmermann, U., and Muller-Hermelink, H. K. Tumor-specific apoptosis by the human monoclonal antibody SC-1: A new therapeutical approach for stomach cancer. Oncology Reports, 5: 35– 40, 1998. 13. Hensel, F., Hermann, R., Schubert, C., Abe, N., Schmidt, K., Franke, A., Shevchenko, A., Mann, M., Muller-Hermelink, H. K., and Vollmers, H. P. Characterisation of glycosylphosphatidylinositol-linked molecule CD55/Decay-accelerating factor as the receptor for antibody SC-1-induced apoptosis. Cancer Res., 59: 5299 –5306, 1999. 14. Hensel, F., Brandlein, S., Eck, M., Schmidt, K., Krenn, V., Kloetzer, A., Bachi, A., Mann, M., Muller-Hermelink, H. K., and Vollmers, H. P. A. novel proliferationassociated variant of CFR-1 defined by a human monoclonal antibody. Lab. Investig., 81: 1097–1108, 2001. 15. Hensel, F., Hermann, R., Brandlein, S., Krenn, V., Schmausser, B., Geis, S., MullerHermelink, H. K., and Vollmers, H. P. Regulation of the new coexpressed CD55 (decay-accelerating factor) receptor on stomach carcinoma cells involved in antibody SC-1-induced apoptosis. Lab. Investig., 81: 1553–1563, 2001. 16. Vollmers, H. P., Hensel, F., Hermann, R., Da¨ mmrich, J., Zimmermann, U., and Mu¨ ller-Hermelink, H. K. Specific apoptotic activity of the human monoclonal antibody SC-1 on the human stomach cancer cells in vitro. Immunobiology, 179: 394, 1997. 17. Vollmers, H. P., Zimmermann, U., Krenn, V., Timmermann, W., Illert, B., Hensel, F., Hermann, R., Thiede, A., Wilhelm, M., Ruckle-Lanz, H., Reindl, L., and MullerHermelink, H. K. Adjuvant therapy for gastric adenocarcinoma with the apoptosisinducing human monoclonal antibody SC-1: first clinical and histopathological results. Oncol. Rep., 5: 549 –552, 1998. 18. Hermann, R. Hensel, F., Muller, E. C., Keppler, M., Souto-Carneiro, M., Brandlein, S., Muller-Hermelink, H. K., and Vollmers, H. P. Deactivation of regulatory proteins hnRNP A1 and A2 during SC-1 induced apoptosis. Hum. Antibodies, 10: 83–90, 2001. 19. Vollmers, H. P., Da¨ mmrich, J., Hensel, F. Ribbert, H., Meyer-Bahlburg, A., UfkenGaul, T., Korff, M., and Mu¨ ller-Hermelink, H. K. Differential expression of apoptosis receptors on diffuse and intestinal type stomach carcinoma. Cancer (Phila.), 79: 433– 440, 1997. 20. Vollmers, H. P., Dammrich, J., Ribbert, H., Grassel, S., Debus, S., Heesemann, J., and Muller-Hermelink, H. K. Human monoclonal antibodies from stomach carcinoma patients react with Helicobacter pylori and stimulate stomach cancer cells in vitro. Cancer (Phila.), 74: 1525–1532, 1994. 21. Burrus, L. W., Zuber, M. E., Lueddecke, B. A., and Olwin, B. B. Identification of a cysteine-rich receptor for fibroblast growth factors. Mol. Cell. Biol., 12: 5600 –5609, 1992. 22. Krenn, V., Ko¨ nig, A., Hensel, F., Berek, C., Souto-Carneiro, M. M., Haedicke, W., Wang, Y., Vollmers, H. P., and Mu¨ ller-Hermelink, H. K. Molecular analysis of rheumatoid factor (RF)-negative B cell hybridomas from rheumatoid synovial tissue: evidence for an antigen-induced stimulation with selection of high mutated IgVH and low mutated IgVL/lambda genes. Clin. Exp. Immunol., 115: 168 –175, 1999. 23. Vollmers, H. P., Stulle, K., Da¨ mmrich, J., Pfaff, M., Papadopoulos, T., Betz, C., Saal, K., and Mu¨ ller-Hermelink, H. K. Characterization of four new gastric cancer cell lines. Virchows Arch. B Cell Pathol. Incl. Mol. Pathol., 63: 335–343, 1993. 24. Saal, K., Vollmers, H. P., Mu¨ ller, J., Ko¨ hler, J., Ho¨ hn, H., and Mu¨ ller-Hermelink, H. K. Cytogenetic differences between intestinal and diffuse types of human gastric carcinoma. Virchows Arch. B. Cell Pathol. Incl. Mol. Pathol., 64: 145–150, 1993. 25. Moretta, A., Bottino, C., Mingari, M. C., Biassoni, R., and Moretta, L. What is a natural killer cell? Nat. Immunol., 3: 6 – 8, 2000.

26. Amakata, Y., Fujiyama, Y., Andoh, A., Hodohara, K., and Bamba, T. Mechanism of NK cell activation induced by coculture with dendritic cells derived from peripheral blood monocytes. Clin. Exp. Immunol., 124: 214 –222, 2001. 27. Soloski, M. J. Recognition of tumor cells by the innate immune system. Curr. Opin. Immunol., 13: 154 –162, 2001. 28. Berczi, I., Bertok, L., and Chow, D. A. Natural immunity and neuroimmune host defense. Ann. N. Y. Acad. Sci., 917: 248 –257, 2000. 29. Bohn, J. Are natural antibodies involved in tumour defense? Immunol. Lett., 69: 317–320, 1999. 30. Boes, M. Role of natural and immune IgM antibodies in immune responses. Mol. Immunol., 37: 1141–1149, 2000. 31. Dunn, G. P., Bruce, A. T., Ikeda, H., Old, L. J., and Schreiber, R. D. Cancer immunoediting: from immunosurveillance to tumor escape. Nat. Immunol., 3: 991– 998, 2002. 32. Colucci, F., Di Santo, J. P., and Leibson, P. J. Natural killer cell activation in mice and men: different triggers for similar weapons? Nat. Immunol., 3: 807– 813, 2002. 33. Kronenberg, M., and Gapin, L. The unconventional lifestyle of NKT cells. Nat. Rev. Immunol., 2: 557–568, 2002. 34. Kuppen, P. J., Gorter, A., Hagenaars, M., Jonges, L. E., Giezeman-Smits, K. M., Nagelkerke, J. F., Fleuren, G., and van de Velde, C. J. Role of NK cells in adoptive immunotherapy of metastatic colorectal cancer in a syngeneic rat model. Immunol. Rev., 184: 236 –243, 2001. 35. Parkin, J., and Cohen, B. An overview of the immune system. Lancet, 357: 1777– 1789, 2001. 36. Zitvogel, L. Dendritic and natural killer cells cooperate in the control/switch of innate immunity. J. Exp. Med., 195: 9 –14, 2002. 37. MacPherson, G., Kushnir, N., and Wykes, M. Dendritic cells, B cells and the regulation of antibody synthesis. Immunol. Rev., 172: 325–334, 1999. 38. Martin, F., and Kearney, J. F. B1 cells: similarities and differences with other B cell subsets. Curr. Opin. Immunol., 13: 195–201, 2001. 39. Ochsenbein, A. F., and Zinkernagel, R. M. Natural antibodies and complement link innate and acquired immunity. Immunol. Today, 21: 624 – 630, 2000. 40. Moretta, L., Biassoni, R., Bottino, C., Mingari, M. C., and Moretta, A. Natural killer cells: a mystery no more. Scand. J. Immunol., 55: 229 –232, 2002. 41. Greenberg, S., and Grinstein, S. Phagocytosis and innate immunity. Curr. Opin. Immunol., 14: 136 –145, 2000. 42. Mekori, Y. A., and Metcalfe, D. D. Mast cells in innate immunity. Immunol. Rev., 173: 131–140, 2000. 43. Janeway, C. A., Jr. How the immune system works to protect the host from infection: a personal view. Proc. Natl. Acad. Sci. USA, 98: 7461–7468, 2001. 44. Janeway, C. A., Jr. Approaching the asymptote? Evolution and revolution in immunology. Cold Spring Harb. Symp. Quant. Biol., 54: 1–13, 1989. 45. Medzhitov, R. Toll-like receptors and innate immunity. Nature Rev. Immunol., 1: 135–145, 2001. 46. Cedar, H., and Bergman, Y. Developmental regulation of immune system gene rearrangement. Curr. Opin. Immunol., 11: 64 – 69, 1999. 47. Lewis, S. M. The mechanism of V(D)J joining: lessons from molecular, immunological, and comparative analyses. Adv. Immunol., 56: 27–150, 1994. 48. Gellert, M. Recent advances in understanding V(D)J recombination. Adv. Immunol., 64: 39 – 64, 1997. 49. Schatz, D. G., Oettinger, M. A., and Schlissel, M. S. V(D)J recombination: molecular biology and regulation. Annu. Rev. Immunol., 10: 359 –383, 1992. 50. Constantinescu, A., and Schlissel, M. S. Changes in locus-specific V(D)J recombinase activity induced by immunoglobulin gene products during B cell development. J. Exp. Med. 185, 609 – 620, 1997. 51. Papavasiliou, F., Jankovic, M., Gong, S., and Nussenzweig, M. C. Control of immunoglobulin gene rearrangements in developing B cells. Curr. Opin. Immunol., 9: 233–238, 1997. 52. Rothenberg, E. V. Stepwise specification of lymphocyte developmental lineages. Curr. Opin. Genet. Dev., 10: 370 –379, 2000. 53. Hensel, F., Knorr, C., Hermann, R., Krenn, V., Muller-Hermelink, H. K., and Vollmers, H. P. Mitogenic autoantibodies in Helicobacter pylori-associated stomach cancerogenesis. Int. J. Cancer, 81: 229 –235, 1999. 54. Karsten, U., Diotel, C., Klich, G., Paulsen, H., Goletz, S., Mu¨ ller, S., and Hanisch, F. G. Enhanced binding of antibodies to the DTR motif of MUC1 tandem repeat peptide is mediated by site-specific glycosylation. Cancer Res., 58: 2541–2549, 1998. 55. Egea, G., Franci, C., Gambus, G., Lesuffleur, T., Zweibaum, A., and Real, F. X. cis-Golgi resident proteins and O-glycans are abnormally compartmentalized in the RER of colon cancer cells. J. Cell Sci., 105: 819 – 830, 1993. 56. Apostolopoulos, V., Loveland, B. E., Pietersz, G. A., and McKenzie, I. F. CTL in mice immunized with human mucin 1 are MHC-restricted. J. Immunol., 155: 5089 – 5094, 1995. 57. Hayakawa, K., Hardy, R. R., Herzenberg, L. A., and Herzenberg, L. A. Progenitors for Ly-1 B cells are distinct from progenitors for other B cells. J. Exp. Med., 161: 1554 –1568, 1985. 58. Kipps, T. J. The CD5 B cell. Adv. Immunol., 47: 117–185, 1989. 59. Forster, I., Gu, H., Muller, W., Schmitt, M., Tarlinton, D., and Rajewsky, K. C. D.5 B cells in the mouse. Curr. Top. Microbiol. Immunol., 173: 247–251, 1991. 60. Hardy, R. R. Variable gene usage, physiology and development of Ly-1⫹ (CD5⫹) B cells. Curr. Opin. Immunol., 4: 181–185, 1992. 61. Haughton, G., Arnold, L. W., Whitmore, A. C., and Clarke, S. H. B-1 cells are made, not born. Immunol. Today, 14: 84 – 87, 1993. 62. Herzenberg, L. A. B-1 cells: the lineage question revisited. Immunol. Rev., 175: 9 –22, 2000. 63. Rothstein, T. L. Cutting edge commentary: two B-1 or not to be one. J. Immunol., 168: 4257– 4261, 2002.

8004


64. Hayakawa, K., and Hardy, R. R. Development and function of B-1 cells. Curr. Opin. Immunol., 12: 346 –353, 2000. 65. Wortis, H. H., and Berland, R. Cutting edge commentary: origins of B-1 cells. J. Immunol., 166: 2163–2166, 2001. 66. Su, I., and Tarakhovsky, A. B-1 cells: orthodox or conformist? Curr. Opin. Immunol., 12: 191–194, 2000. 67. Hayakawa, K., Hardy, R. R., Parks, D. R., and Herzenberg, L. A. The “Ly-1 B” cell subpopulation in normal immunodefective, and autoimmune mice. J. Exp. Med., 157: 202–218, 1983. 68. Lalor, P. A., Stall, A. M., Adams, S., and Herzenberg, L. A. Permanent alteration of the murine Ly-1 B repertoire due to selective depletion of Ly-1 B cells in neonatal animals. Eur. J. Immunol., 19: 501–506, 1989. 69. Hamilton, A. M., Lehuen, A., and Kearney, J. F. Immunofluorescence analysis of B-1 cell ontogeny in the mouse. Int. Immunol., 6: 355–361, 1994. 70. Moris, D. L., and Rothstein, T. L. CD5⫹ B (B-1) cells and immunity. E. C. Snow, Ed. Handbook of B and T Lymphocytes, p. 421. San Diego: Academic, 1994. 71. Lam, K. P., and Rajewsky, K. B cell antigen receptor specificity and surface density together determine B-1 versus B-2 cell development. J. Exp. Med., 190: 471– 477, 1999. 72. Sidman, C. L., Shultz, L. D., Hardy, R. R., Hayakawa, K., and Herzenberg, L. A. Production of immunoglobulin isotypes by Ly-1⫹ B cells in viable motheaten and normal mice. Science (Wash. DC), 232: 1423–1425, 1986. 73. Forster, I., and Rajewsky, K. Expansion and functional activity of Ly-1⫹ B cells upon transfer of peritoneal cells into allotype-congenic, newborn mice. Eur. J. Immunol., 17: 521–528, 1987.

74. Kroese, F. G., Ammerlaan, W. A., and Kantor, A. B. Evidence that intestinal IgA plasma cells in ␮, ␬ transgenic mice are derived from B-1 (Ly-1 B) cells. Int. Immunol., 5: 1317–1327, 1993. 75. Ishida, H., Hastings, R., Kearney, J., and Howard, M. Continuous anti-interleukin 10 antibody administration depletes mice of Ly-1 B cells but not conventional B cells. J. Exp. Med., 175: 1213–1220, 1992. 76. Luhm, J., Brand, J. M., Koritke, P., Hoppner, M., Kirchner, H., and Frohn, C. Large-scale generation of natural killer lymphocytes for clinical application. J. Hematother. Stem. Cell Res., 11: 651– 657, 2002. 77. Lanier, L. L., and Phillips, J. H. Natural killer cells. Curr. Opin. Immunol., 4: 38 – 42, 1992. 78. Miller, J. S. Biology of natural killer cells in cancer and infection. Cancer Investig., 20: 405– 419, 2002. 79. Bohn, J. Josimovic-Alasevic, O., Settmacher, U., Kiessig, S. T., Lukowsky, A., Volk, H. D., Diamantstein, T., Von Baehr, R., and Jahn, S. Human monoclonal IgM antibodies from foetal B-cell hybridomas directed against a surface antigen on human tumour cells. Res. Immunol., 143: 879 – 891, 1992. 80. Ollert, M. W., David, K., Schmitt, C., Hauenschild, A., Bredehorst, R., Erttmann, R., and Vogel, C. W. Normal human serum contains a natural IgM antibody cytotoxic for human neuroblastoma cells. Proc. Natl. Acad. Sci. USA, 93: 4498 –503, 1996. 81. David, K., Ollert, M. W., Vollmert, C., Heiligtag, S., Eickhoff, B., Erttmann, R., Bredehorst, R., and Vogel, C. W. Human natural immunoglobulin M antibodies induce apoptosis of human neuroblastoma cells by binding to a Mr 260, 000 antigen. Cancer Res., 59: 3768 –3775, 1999.

8005