Cancer Research Cancer Epigenetics Sign up for Cancer Research eTOC's
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Schneider, F.
Right arrow Articles by Schlag, P. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Schneider, F.
Right arrow Articles by Schlag, P. M.
[Cancer Research 61, 4605-4611, June 1, 2001]
© 2001 American Association for Cancer Research

Tumor Biology

Overexpression of Sialyltransferase CMP-Sialic Acid:Galß1,3GalNAc-R {alpha}6-Sialyltransferase Is Related to Poor Patient Survival inHuman Colorectal Carcinomas1

Frank Schneider, Wolfgang Kemmner2, Wolfgang Haensch, Gudrun Franke, Stephan Gretschel, Uwe Karsten and Peter Michael Schlag

Robert-Rössle-Klinik at the Max Delbrück Center for Molecular Medicine, Department for Surgery and Surgical Oncology, Charité, Campus Buch [F. S., W. K., W. H., G. F., S. G., P. M. S.], and Max Delbrück Center for Molecular Medicine [U. K.], D-13122 Berlin, Germany


   ABSTRACT
Top
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Thomsen-Friedenreich (TF)-related blood group antigens, such as TF, Tn, and their sialylated variants, belong to a family of tumor-associated carbohydrates. The aim of the present study was to examine tumor-associated alterations of glycosyltransferases involved in the biosynthesis of the TF glycotope in colorectal carcinomas. To this end, glycosyltransferase expression was examined in 40 cases of colorectal carcinoma specimens classified according to the WHO/Union International Contre Cancer guidelines and in "normal" mucosa of the same patients. Occurrence of TF glycotope was examined by immunohistochemistry with the monoclonal antibody A78-G/A7. Expression of sialyltransferases CMP-sialic acid:Galß1,3GalNAc-R {alpha}3-sialyltransferase I and II (ST3Gal-I and ST3Gal-II) and CMP-sialic acid:Galß1,3GalNAc-R {alpha}6-sialyltransferase (ST6GalNAc-II) and of core 2 ß1,6-N-acetylglucosaminyltransferase was determined by reverse transcription-PCR in the same cryostat sections used for immunohistochemistry. Additionally, {alpha}2,3-sialyltransferase enzyme activity was studied in each of these tissues. The TF glycotope was detected in 7% of the normal mucosa, but in 57% of the carcinoma samples. Expression of {alpha}2,3-sialyltransferases ST3Gal-I, ST3Gal-II, and enzyme activity of {alpha}2,3-sialyltransferase was significantly increased (P < 0.001) in carcinoma specimens compared with normal mucosa. ST3Gal-I mRNA expression was significantly increased (P = 0.05) in cases showing invasion of lymph vessels. Expression of ST6GalNAc-II was significantly increased (P = 0.04) in cases with metastases to lymph nodes along the vascular trunk. Moreover, ST6GalNAc-II expression provides an prognostic factor for patient survival (log rank, P = 0.02). In an attempt to study the functional relevance of the glycosyltransferases for TF biosynthesis, SW480 colorectal cells were transfected with each of the enzymes, and cell surface expression of the TF glycotope was examined by flow cytometry. The presence of TF was not altered by transfection of the cells with either sialyltransferase ST3Gal-I or ST3Gal-II. However, successful transfection with core 2 ß1,6-N-acetylglucosaminyltransferase led to reduced expression of TF. In contrast, increased cell surface expression of TF was found after ST6GalNAc-II transfection. Thus, expression of TF on the cell surface of SW480 colorectal carcinoma cells depends on the ratio of core 2 ß1,6-N-acetylglucosaminyltransferase and ST6GalNAc-II. Earlier immunohistological studies demonstrated that TF is a prognostic factor for patient survival. Our results suggest that sialyltransferase ST6GalNAc-II is of crucial relevance for the prognostic significance of TF.


   INTRODUCTION
Top
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
TF3 -related blood group antigens such as TF, Tn, and their sialylated variants, are oncofetal carbohydrate structures. Expression of these antigens on glycoconjugates of the cell surface has been correlated with tumor prognosis and metastasis (1, 2, 3, 4, 5) . The biosynthesis of the TF glycotope (Galß1,3-N-acetyl-D-galactosamine) depends on the activity of a number of glycosyltransferases, starting with peptidyl-GalNAc-transferase to form the Tn structure (6) . In the second step, there are three different possible reaction paths: formation of sialosyl-Tn (Fig. 1Citation , Sia6-Tn) through the activity of sialyltransferase ST6GalNAc-I; formation of core 3 structures through a core 3 ß1,3GlcNAc transferase; or synthesis of the core 1 structure TF (Fig. 1Citation , middle) through the action of a ß1,3-galactosyltransferase. However, in most "normal" mucosa, TF is not expressed but processed by C2GNT to a core 2 (Fig. 1)Citation structure, which then is subject to further chain extension and/or {alpha}2,3-sialylation. Direct sialylation of TF by {alpha}2,3-sialyltransferases (Fig. 1Citation , Sia3Core-1), which catalyzes the transfer of sialic acid to Galß1,3-N-acetyl-D-galactosamine, interferes with further chain extension. In addition, {alpha}2,6-sialylation of TF (Fig. 1Citation , Sia6Core-1) or of 3-sialylated core 1 to disialylated core 1 (Fig. 1Citation , DiSiaCore-1) by ST6GalNAc-II stops further processing because ST6GalNAc-II competes with C2GNT for the C-6 of GalNAc. However, once core 2 is formed, further conversions to complex O-glycans can be catalyzed (Fig. 1Citation , Sia3Core-2). Increased levels of the {alpha}2,3-sialyltransferases and decreased levels of C2GNT have been found in breast cancer cells (7 , 8) . Similarly, Yang et al. (9) in an analysis of homogenates prepared from colorectal carcinomas, found increased enzyme activity of {alpha}2,3-sialyltransferases and decreased C2GNT activity compared with normal mucosa.



View larger version (19K):
[in this window]
[in a new window]
[Download PPT slide]
  		Fig. 1. Biosynthetic pathways of O-glycan structures studied in this report. For the synthesis of the core 1 structure (TF), a ß1,3-galactosyltransferase using the acceptor Tn is responsible. However, in most normal mucosa, TF is not expressed but processed by core 2 ß1,6-GlcNAc transferase C2GNT (1) to a core 2 structure, which then becomes subject to further chain elongation and/or {alpha}2,3-sialylation (2) to 3-sialylated core 2 (Sia3Core-2). Direct sialylation of TF by {alpha}2,3-sialyltransferases (2), ST3Gal-I, or ST3Gal-II leads to 3-sialylated core 1 (Sia3Core-1). Another possibility is {alpha}2,6-sialylation (3) of TF to 6-sialylated core 1 (Sia6Core-1) or of 3-sialylated core 1 to disialylated core 1 (DiSiaCore-1) by sialyltransferase ST6GalNAc-II (3). This stops further processing because ST6GalNAc-II competes with C2GNT (1) for the C-6 of GalNAc. NeuAc, N-acetylneuraminic acid.

 
In the present study, the expression of glycosyltransferases involved in biosynthesis of the TF glycotope was analyzed in surgical specimens from colorectal carcinomas. In detail, we determined GM1-dependent {alpha}2,3-sialyltransferase enzyme activity (10) and mRNA expression of O-glycan-specific {alpha}2,3-sialyltransferases ST3Gal-I (11) and ST3Gal-II (12) , of core 2 ß1,6-GlcNAc transferase (13) , and of ST6GalNAc-II (14) . The occurrence of TF and the extent of sialylation of TF was examined by immunohistochemical detection. Experimental results were compared with histopathological data and the clinical follow-up. In an attempt to study the functional relevance of glycosyltransferases for TF biosynthesis, cell surface expression of TF was examined after transfection of a colorectal cell line with each of the enzymes under study.


   MATERIALS AND METHODS
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Materials.
Oligonucleotides were synthesized using an Applied Biosystems 380B automated DNA synthesizer and used after high-performance liquid chromatographic purification (BioTez, Berlin, Germany). Guanidine thiocyanate-phenol-chloroform, ethidium bromide, DMEM, fetal bovine serum, PBS, glutamine, and G418 were bought from Gibco (Life Technologies, Inc., Eggenstein, Germany). Vectors pcDNA4/TO, pcDNA3.1, and pcDNA3.1/V5/His-TOPO; zeocin; and Eukaryotic TOPO TA Cloning Kit were from Invitrogen (Groningen, the Netherlands). HBSS and trypsin/EDTA were purchased from Biochrom (Berlin, Germany). Cytidine-5-monophospho-N-acetylneuraminic acid, RNase-free DNase, DNA standards, and random hexamers were obtained from Boehringer-Mannheim (Germany). Moloney murine leukemia virus reverse transcriptase was from Promega (Mannheim, Germany). Agarose and sodium cacodylate were from Roth (Karlsruhe, Germany). AmpliTaq polymerase was from Perkin-Elmer (Weiterstadt, Germany). Deoxynucleotide triphosphates and Sephadex G50 were from Amersham-Pharmacia (Freiburg, Germany). FITC-conjugated goat antimouse IgG was from Dianova (Hamburg, Germany). The lectins jacalin and PNA were from Vector. SuperFect was from Qiagen (Hilden, Germany). Vibrio cholerae neuraminidase was from DadeBehring (Marburg, Germany). TF-PAA polyacrylamide conjugate was obtained from Syntesome (Munich, Germany). All other reagents were obtained from Sigma.

Patients and Tissue Samples.
A total of 40 patients suffering from colorectal carcinoma were examined (19 women, 21 men). Their age range was 40–93 years, with a median age of 69.5 years. The range of follow-up was 1–40 months, with a median follow-up time of 24 months. In 25 patients, the carcinoma was located in the colon; in 15, it was located in the rectum. Eighteen carcinomas had formed distant metastases at the time of surgery. Surgical specimens were classified and characterized according to the WHO/Union International Contre Cancer guidelines. Nonmalignant mucosa was scraped off, and carcinoma tissue was carefully excised from tumor specimens. Tissue was snap-frozen in liquid nitrogen. In each case, tumor tissue and normal mucosa from the same patient were examined. All data, including gender, age, stage of disease, and pathological factors, were prospectively recorded.

Statistical Analysis.
The Wilcoxon signed-rank test was used for statistical analysis of differences between mRNA expression of a given enzyme in carcinoma and normal tissue of the same patient (matched pairs). Kruskal-Wallis and Mann-Whitney tests were used for the evaluation of possible statistical correlations between enzyme expression and histopathological parameters. Survival analysis was performed for the overall survival time. Survival curves were obtained according to Kaplan-Meier, and the significance of their differences was assessed by the log-rank test. Hazard ratios, which indicate the relative risk of death of one group compared with the other, were calculated through Cox regression. To analyze the association of several predictor variables with survival, we performed a multivariate Cox regression analysis with a standard model of approved prognostic markers such as tumor staging and presence of distant metastases.

Cell Culture.
Human colon cancer cell lines SW480 and LS174T were cultured in DMEM supplemented with 10% fetal bovine serum, 2 mM glutamine, and antibiotics (100 units/ml penicillin and 100 mg/ml streptomycin). All cells were grown at 37°C in a humidified atmosphere of 6% CO2.

Immunohistochemistry.
Immunohistochemistry was performed using an indirect immunofluorescence technique. Serial cryostat sections were used to achieve comparable molecular and immunofluorescence probes. A mAb against TF antigen, A78-G/A7 (15) , was used to describe the topographical antigen distribution. The antibody was used as supernatant at a dilution of 1:10. As a secondary antibody, an AMCA-conjugated goat antimouse immunoglobulin antibody (DAKO, Hamburg, Germany) at a dilution of 1:5 was used. AMCA fluorochrome was activated by short-wavelength light at 340–350 nm. Emission was in the blue region at 440–460 nm. Negative controls were prepared by substitution of the primary antibody by Tris-buffered saline.

Nuclear DNA was counterstained with the fluorescent vital dye propidium iodide (Becton Dickinson, Heidelberg, Germany) at a dilution of 2 µl/100 ml. Propidium iodide was detected in the orange range of the spectrum with a 562–588 nm band pass filter.

To determine the extent of sialylation of the TF antigen, neuraminidase predigestion of tissue sections was carried out according to a modified histochemical standard procedure. Briefly, slides were incubated with sialidase from Vibrio cholerae diluted 1:10 in 0.1 M sodium acetate buffer (pH 5.5) containing 9 mM CaCl2 and 154 mM NaCl for 60 min at 37°C. Staining was visualized by means of an inverse epifluorescence microscope Axiovert 100 M, (Zeiss, Jena, Germany), with the excitation filter adjusted to AMCA excitation. All slides were examined at a magnification of x630. Setup was configured as follows: Gamma, 1.0; enhancement, red 0.6, green 0.8, blue 0.8; exposure time, 5.04 s; gain, 1 dB. Original interpretation of the immunofluorescent slides was made by direct visual inspection. Scoring was performed as follows: 0, no staining; 1, positive staining of less than one-third of the epithelial cells in a section; 2, positive staining of more one-third and less than two-thirds of the epithelial cells; 3, positive staining of more than two-thirds of the cells.

RT-PCR for Glycosyltransferases.
Serial cryostat sections were examined by a pathologist. Only sections containing predominantly carcinoma tissue were analyzed. Carcinoma tissue from adjacent sections was powdered in liquid N2 by a tissue homogenizer (Braun, Germany). RNA was extracted using guanidine thiocyanate-phenol-chloroform (16) and treated with RNase-free DNase. RNA yield was determined spectrophotometrically. Five micrograms of total cellular RNA were reverse transcribed by Moloney murine leukemia virus reverse transcriptase with random hexamers. Expression of the ß-actin housekeeping gene was used to determine cDNA yield and integrity and to check for possible contamination with genomic DNA. To this end, an intron-spanning ß-actin-specific primer set was designed: ß-actin forward primer, 5'-GGC ATC GTG ATG GAC TCC G-3'; ß-actin reverse primer, 5'-GCT GGA AGG TGG ACA GCG A-3' (amplicon length, 622 bp).

Enzyme expression in a given tissue was determined by RT-PCR simultaneously for all glycosyltransferases under study. C2GNT was amplified according to Shimodaira et al. (17) using forward primer 5'-GTG CTC AGA ATG GGG CAG GAT GTC ACC TGG-3' and reverse primer 5'-TCA CTA CTA GGA TTC TCC CCA GCA AGC TCC-3' (amplicon length, 381). ST3Gal-I (11) was amplified using forward primer 5'-ATG AGG TGG ACT TGT ACG GC-3' and reverse primer 5'-AAC GGC TCC AGC AAG ATG-3' (amplicon length, 253). ST3Gal-II (18) was amplified using forward primer 5'-CCC TGC TCT TCA CCT ACT CG-3' and reverse primer 5'-GCA TCA TCC ACC ACC TCT-3' (amplicon length, 282). ST6GalNAc-II was amplified using forward primer 5'-CTG CCA GTA AAT TCA AGC TGC-3' and reverse primer 5'-TTG CTT GTG ATG AAT CCA TAG G-3' (amplicon length, 184).4 ß-Actin was coamplified within the same tube with 0.2 µl of ß-actin primer.

For all sample sequences, cycling conditions were 1 min at 94°C, 1 min at 59°C, and 2 min at 72°C. The PCR reaction mixture contained 1.5 µl of PCR buffer [100 mM Tris-HCl (pH 8.3), 500 mM KCl], MgCl2 (see below), 5 µl of sample cDNA, 1.5 µl of 2 mM deoxynucleotide triphosphates, 0.2 µl of AmpliTaq polymerase, 0.2 µl of ß-actin primer (original concentration, 0.25 µM), enzyme-specific primers, and H2O to a volume of 15 µl. The volume of enzyme-specific primers (original concentration, 50 µM) was 0.8 µl in reactions to detect ST3Gal-I and C2GNT and 0.7 µl in reactions to detect ST3Gal-II and ST6GalNAc-II. The volume of MgCl2 (original concentration, 25 mM) was 1 µl in reactions to detect ST3Gal-I and 0.8 µl in reactions to detect ST3Gal-II, ST6GalNAc-II, and C2GNT. Reaction products obtained after 26–28 cycles were electrophoresed in 3% agarose gels containing ethidium bromide. All reaction products were sequenced and compared with published data. Fluorescence was measured by a Fluor-Imager SI (Molecular Dynamics, Krefeld, Germany). For semiquantitative analysis of the RT-PCR data, the fluorescence of each sample was compared with the fluorescence of ß-actin coamplified within the same tube. The density of each enzyme band was compared with the appropriate ß-actin band by calculating the ratio [(fluorescence units of the enzyme band/fluorescence units of ß-actin) x 100]. The detection limit was 1 ng of double-stranded DNA. The assay was linear up to 25 ng. RNA expression of each sample was determined at least in five independent experiments.

For comparison, enzyme expression was also examined by competitive RT-PCR using mimic standards. Internal standards were identical to the sample with regard to the primer annealing sequences, but differed in sequence length. The volume of the standards in the PCR reaction was 0.2 µl (original concentration, 1 pM). The volume of enzyme-specific primers (original concentration, 50 µM) was 0.8 µl in reactions to detect ST3Gal-I and 0.7 µl in reactions to detect ST3Gal-II. The volume of MgCl2 (original concentration, 25 mM) was 0.8 µl in reactions to detect ST3Gal-I and 0.7 µl in reactions to detect ST3Gal-II. Cycling parameters were as described above.

Fluorometric Sialyltransferase Assay.
Tissue was minced with scissors. After the minced tissue was washed with ice-cold PBS (pH 7.3) at 4°C, cells were pelleted and incubated on a rocking plate for 30 min at 4°C in 500 µl of 0.25 M sodium cacodylate buffer (pH 6.5) containing 0.4% Triton X-100. Supernatants were centrifuged at 20,000 x g at 4°C for 20 min. The protein content in diluted supernatants was determined by the BIO-RAD (Munich) protein assay using BSA as standard. Enzyme activity was measured fluorometrically according to Gross and Brossmer (19) with modifications. Acceptors used were the ganglioside GM1 (20 mg/ml), TF-PAA (1 mg/ml), or asialo-fetuin (20 mg/ml).

Briefly, 5 µl of enzyme extract containing 5 µg of protein was incubated with 8 µl of 0.25 M sodium cacodylate buffer (pH 6.5), 2.6 µl of acceptor GM1 or asialo-fetuin or 17.2 µl of TF-PAA, 2 µl of BSA (20 mg/ml), 2.8 µl of 10 µM FITC-labeled cytidine 5'-monophospho-N-acetylneuraminic acid, and 14.6 µl of H2O (for TF-PAA, only 5 µl of H2O were added) for 60 min at 37°C. The reaction was stopped by the addition of 5.4 µl of cytidine 5'-triphosphate.

Twenty-five µl of the assay mixture were applied to a liquid chromatography system. Fluorescence was detected at an excitation wavelength of 490 nm and an emission wavelength of 520 nm. Fluorescently labeled glycoprotein was separated from donor substrate on a Sephadex G50 column (12.5 x 0.4 cm) eluted with 0.1 M Tris-HCl (pH 8.5) at a flow rate of 0.5 ml/min. Endogenous activity without acceptor GM1 was subtracted. One unit of activity is defined as the amount of enzyme catalyzing the formation of 1 µmol of product/min.

Transfection of Cell Lines.
Human cDNA encoding the Galß1,4GlcNAc {alpha}2,3-sialyltransferase gene, ST3Gal-I, was obtained from Dr. J. C. Lefebvre (Université De Nice, Nice, France; Ref. 12 ), ST3Gal-II was obtained from Dr. Eric Sjoberg (Cytel Corporation, San Diego, CA), ST6GalNAc-II from Dr. Georgia Sotiropoulou (University of Patras, Greece), and C2GNT from Dr. M. Fukuda (Burnham Institute, La Jolla, CA; Ref. 17 ). In each case, enzyme cDNA was cloned into the vector pcDNA3.1. Human colon carcinoma cells SW480 (American Type Culture Collection) were transfected with this vector using the SuperFect reagent. Cells were grown to 40% confluence. Five µg of DNA were dissolved in 150 ml of DMEM and mixed with 30 µl of SuperFect. After a 10-min incubation at room temperature, this solution was added to the cells. Three h later, the medium was removed, and the cells were washed three times with PBS and incubated with fresh growth medium. After a 48-h incubation period, growth medium was replaced again with medium containing 700 µg/ml G418 or 750 µg/ml zeocin. Resistant clones were maintained continuously in growth medium containing 400 µg/ml G418 or 250 µg/ml zeocin. Expression vector pcDNA3.1, containing the full-length cDNA of ST3Gal-II, was transfected into SW480 cells. The coding regions of ST3Gal-I and ST6GalNAc-II were cloned into the expression vector pcDNA3.1/V5/His-TOPO. Both vectors were transfected into SW480 cells. The same vector containing ST6GalNAc-II was also transfected into LS174T cells. The coding region of C2GNT was inserted into the expression vector pcDNA4/TOPO and transfected into SW480 cells. SW480 and LS174T cells were also mock-transfected with plasmids pcDNA3.1 and pcDNA4/TOPO without an insert. The vector pcDNA3.1/V5/His-TOPO contains a sequence that codes for a series of six histidine residues that function as a metal-binding domain in the translated protein. Recombinant protein was purified by immobilized metal-affinity chromatography (Invitrogen). Cells were lysed by two freeze-thaw cycles. Protein was washed and eluted under native conditions using imidazole elution buffers in increasing concentrations. The elution was monitored by the absorbance of the fractions at 280 nm. The fraction that contained the peak absorbance was concentrated by Mr 10,000 cutoff low-protein binding centrifugal units.

Flow Cytometry.
For flow cytometric analysis, cells were detached with trypsin-EDTA (0.05%–0.02%) in PBS. Cells were washed twice with PBS and allowed to recover for 1 h at 37°C. Cells were then incubated for 30 min at 37°C in 0.1 M imidazole buffer (pH 6.8) containing 0.1% CaCl and 0.4% NaCl with or without V. cholerae neuraminidase. Cells were resuspended in staining buffer (HBSS containing 4% BSA and 30 mM NaN3). For examination of the cell surface expression of TF, digoxigenin-labeled lectins PNA and jacalin were used with the TF-specific mAb A78-G/A7 (Dr. Karstens). PNA was used in a 1:600 dilution for neuraminidase-treated cells and in a 1:200 dilution for the untreated cells. Jacalin was diluted 1:170, and the cell culture supernatant containing antibody A78-G/A7 was used undiluted. Cells were incubated with lectins or antibody for 1 h at 4°C and washed twice with staining buffer. Binding of the antibody A78-G/A7 was detected by the secondary FITC-goat antimouse IgG. Flow cytometric analyses were performed on a FACScan flow cytometer using Cell Quest software (Becton Dickinson).


   RESULTS
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Immunohistochemistry.
Occurrence of the TF glycotope in colorectal carcinomas was determined by immunohistochemistry using the O-glycan-specific mAb A78-G/A7 (20) . This antibody reacts with Galß1,3GalNAc{alpha}/ß residues irrespective of the carrier molecule, and it does not cross-react with related ß-galactosides. Results from indirect immunofluorescence detection of TF epitopes on colorectal mucosa and corresponding carcinomas were analyzed according to the reaction sites of normal cells and tissues within the tumor epithelium. TF reactivity was lower in normal mucosa (positive staining in 7% of cases) than in tumor tissue (positive staining in 57% of cases). The number of cases with >33% TF-positive cells in a section increased after sialidase treatment. This increase was moderate in the normal mucosa (from 0 to 7% positive cases) but strong in carcinomas (from 20 to 66% positive cases). Cases that became positive after neuraminidase treatment are referred to as expressing sialosyl-TF. To quantify sialosyl-TF, the difference in TF-specific reactivity before and after sialidase treatment was calculated. Removal of sialic acid did not change TF-specific reactivity in 35% (14 of 40) of the carcinomas, but this procedure led to a moderate increase of the histological score in 43% (17 of 40) of the cases and to a high increase in 23% (9 of 40) of the cases.

Semiquantitative RT-PCR.
Enzyme expression was always compared with that of ß-actin coamplified within the same tube. The deviation between five independent measurements was on average 10.1% for ST3Gal-I, 10.4% for ST3Gal-II, 14.0% for ST6GalNAc-II, and 11.3% for C2GNT. The ratio between ß-actin and sample fluorescence was cycle independent between cycles 24 and 28 (21 , 22) . Similar results were obtained by use of quantitative measurements using mimic standards (data not shown).

Glycosyltransferase Expression and Enzyme Activity in Human Colorectal Mucosa Specimens and Carcinomas.
Enzyme expression was determined by RT-PCR simultaneously for all glycosyltransferases under study (Table 1)Citation . As an example, RT-PCR of one of the samples is depicted in Fig. 2Citation . For all cases, the median of the ratio of expression in tumor versus normal mucosa was 2.29 for ST3Gal-II, 1.93 for ST3Gal-I, 1.47 for ST6GalNAc-II, and 1.31 for C2GNT. According to the Wilcoxon signed-rank test, the difference between expression of ST3Gal-I and ST3Gal-II in carcinoma and normal tissue was highly significant (P < 0.001) for both enzymes. Sialyltransferase activity toward GM1 was significantly increased in carcinomas (P < 0.001). This enzyme activity correlated with mRNA expression of ST3Gal-I (Spearman’s correlation coefficient, P = 0.013). Expression of C2GNT was also observed in biopsy samples of seven tumor-free healthy patients who underwent a coloscopy. No significant differences in ST6GalNAc-II or C2GNT expression were detected between normal mucosa and carcinoma tissue. Increased expression of ST6GalNAc-II (P = 0.06) but decreased expression of C2GNT (Table 2)Citation were found in tissues displaying pronounced occurrence of TF.


View this table:
[in this window]
[in a new window]

  		Table 1 Glycosyltransferase mRNA expression and GM1-dependent {alpha}2,3-sialyltransferase enzyme activity in surgical specimens of human colorectal tissue

 


View larger version (85K):
[in this window]
[in a new window]
[Download PPT slide]
  		Fig. 2. Expression of glycosyltransferases in colorectal carcinoma tissue. Fluorescence of electrophoretically resolved, ethidium bromide-stained PCR products of ß-actin and glycosyltransferase target sequences. Glycosyltransferases were amplified according to "Materials and Methods." The top band shows the reaction product of ß-actin amplification. The tissue:amplicon pairs are as follows: Lane 1, carcinoma case 35:ST3Gal-I + ß-actin; Lane 2, normal mucosa case 35:ST3Gal-I + ß-actin; Lane 3, carcinoma case 35:ST3Gal-II + ß-actin; Lane 4, normal mucosa case 35:ST3Gal-II + ß-actin; Lane 5, carcinoma case 35:C2GNT + ß-actin; Lane 6, normal mucosa case 35:C2GNT + ß-actin; Lane 7, carcinoma case 35:ST6GNT-II + ß-actin; Lane 8, normal mucosa case 35:ST6GNT-II + ß-actin; Lane std, DNA standards VII (0.019–1.11 kbp).

 

View this table:
[in this window]
[in a new window]

  		Table 2 Glycosyltransferase mRNA expression and GM1-dependent {alpha}2,3-sialyltransferase enzyme activity in comparison to immunohistochemical detection of the TF/glycotope

 
Correlation of Glycosyltransferase Expression with Histopathological Parameters.
Enzyme expression of the carcinoma specimens was correlated with histopathological parameters, such as invasiveness (pT), presence of lymph node metastases (pN) or of distant metastases (M), invasion to lymph (L) or blood vessels (V), using the Mann-Whitney and Kruskal-Wallis tests. A significant increase (P = 0.05) of ST3Gal-I expression was detected in carcinomas with invasion to lymph vessels (L). ST3Gal-I expression was also increased in cases with lymph node metastases (P = 0.08). The formation of lymph node metastases along the vascular trunk (pN3) correlated significantly (P = 0.04) with ST6GalNAc-II mRNA expression. In addition, carcinomas that had invaded lymph (L) or blood vessels (V) showed increased ST6GalNAc-II expression. No significant correlations were found with the formation of distant metastases, although {alpha}2,3-sialyltransferase enzyme activity toward GM1 and ST6GalNAc-II mRNA expression was increased and C2GNT mRNA expression was decreased in metastasizing carcinomas (Table 1)Citation .

Association between Glycosyltransferase Expression and Survival of Patients.
To determine the prognostic value of glycosyltransferase expression for the overall survival of patients, three groups were formed containing equal numbers of cases (n = 13). ST6GalNAc-II expression was found to be highly associated (P = 0.02) with overall patient survival (Fig. 3)Citation . Accordingly, in univariate Cox regression analysis of survival, only ST6GalNAc-II expression was significantly increased (P = 0.001). In multivariate Cox regression analysis expression of ST6GalNAc-II (P = 0.018) was associated with survival. Because invasiveness (pT), and metastasis to regional lymph nodes (pN), are factors with established relevance for patient survival, in the multivariate analysis these variables were compared with glycosyltransferase expression levels. Association was highly significant only for pN (P = 0.027).



View larger version (18K):
[in this window]
[in a new window]
[Download PPT slide]
  		Fig. 3. Overall survival analysis of the patients (n = 40) with high, moderate, or low ST6GalNAc-II expression. Kaplan-Meier survival curves were obtained, and the significance of the differences was assessed by the log-rank test. Three groups were formed: ST6 low, expression of ST6GalNAc-II was <15% of that of ß-actin (n = 13); ST6 moderate, expression of ST6GalNAc-II was 15–50% of that of ß-actin (n = 13); and ST6 high, expression of ST6GalNAc-II was >50% of that of ß-actin (n = 14). ST6GalNAc-II expression was found to be highly associated (log-rank, P = 0.02) with patient survival. X axis, survival time after surgery; Y axis, surviving fraction.

 
Transfection of Colorectal Carcinoma Cells with Sialyltransferases ST3Gal-I, ST3Gal-II, ST6GalNAc-II, and C2GNT.
Colorectal carcinoma cells SW480 were transfected with vector pcDNA3.1 or pcDNA4/TOPO containing cDNA encoding for ST3Gal-I, ST3Gal-II, ST6GalNAc-II, or C2GNT. The enzyme mRNA expression profile of successfully transfected cells was examined by duplex RT-PCR (see above), and the results were compared with those for mock transfectants. Transfection of SW480 cells with ST3Gal-II cDNA resulted in an 2.3-fold overexpression. No such permanent increase was detected in the case of ST3Gal-I. Transfection of SW480 cells with C2GNT cDNA resulted in a 7-fold increase in C2GNT expression. RT-PCR analysis of SW480 cells transfected with ST6GalNAc-II cDNA failed to show a significant increase. In contrast, LS174T cells transfected with ST6GalNAc-II showed strong mRNA expression of this enzyme. No expression of ST6GalNAc-II was detectable in LS174T mock transfectants.

To assess ST6GalNAc-II enzyme activity resulting from transfection, recombinant enzyme protein with a polyhistidine tag was purified from transfected cells. Sialyltransferase activity was then determined using a fluorometric sialyltransferase assay with acceptors asialo-fetuin or TF-PAA. Recombinant ST6GalNAc-II showed high activity with asialo-fetuin but not with TF-PAA (Table 3)Citation . On the other hand, recombinant ST3Gal-I enzyme was able to sialylate TF-PAA. Accordingly, Samyn et al. (14) have shown that the ST6GalNAc-II enzyme demands an acceptor with a protein backbone and does not accept the TF disaccharide alone.


View this table:
[in this window]
[in a new window]

  		Table 3 Enzyme activity of recombinant glycosyltransferases

 
For analysis of the cell surface expression of carbohydrate groups, fluorescence-activated cell sorting using the lectin PNA, which is specific for Galß1,3GalNAc residues on glycoproteins or glycolipids, or the mAb A78-G/A7, which is specific for TF groups on O-glycans was carried out (Table 4)Citation . Sialylation of TF was not influenced by manipulation of ST3Gal-I or ST3Gal-II expression, but sialylation of glycoconjugates other than TF was altered. SW480 as well as LS174T cells showed increased binding of the TF antibody after transfection with ST6GalNAc-II. Thus, increased expression of TF in the ST6GalNAc-II transfectants could be demonstrated. In contrast, cell surface expression of TF was reduced after C2GNT transfection.


View this table:
[in this window]
[in a new window]

  		Table 4 Cell surface expression of transfectants as measured by fluorescence-activated cell sorting

 

   DISCUSSION
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
A number of studies have demonstrated a correlation of cell surface expression of TF with malignancy; however, the molecular background remains to be understood. Recently, in an immunohistochemical study of 264 cases of colorectal carcinomas, Baldus et al. (5) demonstrated that TF immunoreactivity as detected by mAb A78-G/A7 is an independent prognostic factor indicating a lower survival probability. In the present study we tried to combine immunohistochemistry using mAb A78-G/A7 with analysis of the glycosyltransferases involved in biosynthesis of the TF glycotope in surgical specimens from colorectal carcinomas. In an attempt to understand the functional relevance of the glycosyltransferases, we also studied TF expression in colorectal cell lines that had been transfected with the pertinent enzymes.

Immunohistochemistry demonstrated, in agreement with earlier studies, that TF was present almost exclusively in carcinoma tissues and that TF reactivity increases drastically after sialidase treatment (2 , 20) . Although part of the increased accessibility of the TF structure might be attributable to desialylation of neighboring oligosaccharides, these results indicate that TF is highly sialylated in tumor tissue. Accordingly, mRNA expression of sialyltransferases ST3Gal-I and ST3Gal-II and {alpha}2,3-sialyltransferase activity toward GM1 were significantly increased in tumor tissue. GM1-dependent {alpha}2,3-sialyltransferase enzyme activity correlated mainly with ST3Gal-I mRNA expression. A similar increase of {alpha}2,3-sialyltransferase activity in colorectal carcinomas has been demonstrated by Yang et al. (9) , but a direct comparison between TF immunoreactivity and glycosyltransferase activity in the same sections has yet to be performed. Sialylation of TF (core 1) turns off the branching reaction catalyzed by C2GNT that leads to core 2. The reverse, however, is also possible: if core 2 is formed first, it could be easily {alpha}2,3-sialylated, i.e., C2GNT has to act before {alpha}2,3-sialylation (23) . Thus, an enhanced {alpha}2,3-sialyltransferase activity could lead to an increased formation of sialosyl-TF (Fig. 1Citation , Sia3Core-1), but also of Sia3-Core 2 (Fig. 1Citation , Sia3Core-2). Our data suggest that {alpha}2,3-sialylation facilitates metastasis through the lymphoid pathway because ST3Gal-I mRNA expression was increased in cases with metastases to the lymph nodes and in cases with invasion to lymph vessels.

Another sialyltransferase acting on TF is ST6GalNAc-II, which forms an {alpha}2,6-linkage (14) . This type of sialylation of TF turns off the branching reaction catalyzed by C2GNT that leads to core 2. However, because C2GNT and ST6GalNAc-II compete for the same C-6 atom of GalNAc, the reverse is not possible: {alpha}2,6-sialylation excludes further branching and vice versa. Overexpression of ST6GalNAc-II was particularly pronounced in carcinomas with involved lymph nodes along larger blood vessels. Moreover, this study is the first to demonstrate the prognostic value of enhanced expression of ST6GalNAc-II. The mortality of patients with a strong ST6GalNAc-II mRNA expression was higher, resulting in shorter overall survival (Fig. 3)Citation . In multivariate Cox regression, ST6GalNAc-II expression performed well when compared with established prognostic markers such as presence of lymph node metastasis (pN) or invasiveness (pT).

In an attempt to further analyze the functional relevance of the glycosyltransferases to the presence of the TF glycotope, colorectal cell lines were transfected with ST3Gal-I, ST3Gal-II, ST6GalNAc-II, and core-2 ß1,6GlcNAc transferase. RT-PCR demonstrated a severalfold increase in glycosyltransferase expression in the transfected cells. Recombinant ST6GalNAc-II enzyme was purified from transfected cells. ST6GalNAc-II enzyme activity was much higher in the transfectants than in mock transfectants. The presence of the TF glycotope on the surface of successfully transfected cells was examined by flow cytometry. Expression of TF was not altered by transfection of cells with either sialyltransferase ST3Gal-I or ST3Gal-II. However, transfection with C2GNT led to drastically reduced expression of TF in comparison with that of mock-transfected cells. This seems to be attributable to the processing of the core 1 structure TF to core 2 through C2GNT-dependent ß1,6-N-acetylglucosaminylation. On the other hand, transfection of cells with ST6GalNAc-II resulted in increased expression of TF on the cell surface (Table 4)Citation . Thus, {alpha}2,6 sialylation blocks further processing to the core 2 structure because of the competition of ST6GalNAc-II with C2GNT for the same C-6 atom of GalNAc. In contrast to other findings, {alpha}2,6-sialylation attributable to ST6GalNAc-II but not {alpha}2,3-sialylation seems to be the major obstacle for the processing of TF to core 2 structures. From the in vitro results, a correlation of TF expression with that of ST6GalNAc-II is predicted.

Two mechanisms regarding the TF glycotope are involved in tumorigenesis and metastasis of colorectal carcinomas. An early event in tumorigenesis is up-regulation of {alpha}2,3-sialylation of Galß1,3-N-acetyl-D-galactosamine residues. Overexpression of ST3Gal-I may contribute to metastasis because it is correlated with increased metastasis to regional lymph nodes and enhanced invasion to lymph vessels. Another mechanism is up-regulation of ST6GalNAc-II, which blocks the normal processing to the core 2 structure and which is significantly associated with a shorter overall survival of the patients.


   ACKNOWLEDGMENTS
 
We thank S. Grigull and I. Wendler for excellent technical assistance. We also thank Dr. J. C. Lefebvre (Université De Nice, Nice, France) for providing us with ST3Gal-I cDNA, Drs. J. Paulson and E. Sjoberg (Cytel Corp., San Diego, CA) for ST3Gal-II cDNA, Dr. G. Sotiropoulou (University of Patras, Greece) for ST6GalNAc-II cDNA, and Dr. M. Fukuda (The Burnham Institute, La Jolla, CA) for the generous gift of C2GNT cDNA.


   FOOTNOTES
 
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 18 U.S.C. Section 1734 solely to indicate this fact.

1 Supported by the Wilhelm-Sander-Stiftung (Project 98.060.1). Back

2 To whom requests for reprints should be addressed, at Robert-Rössle-Klinik at the Max Delbrück Center for Molecular Medicine, Department for Surgery and Surgical Oncology, Lindenberger Weg 80, D-13122 Berlin, Germany. Phone: 49-30-9406-2506; Fax: 49-30-9406-2846; E-mail: wkemmner{at}mdc-berlin.de Back

3 The abbreviations used are: TF, Thomsen-Friedenreich glycotope; GalNAc, N-acetyl-D-galactosamine; GlcNAc, N-acetyl-D-glucosamine; ST3Gal-I/II, CMP-sialic acid:Galß1,3GalNAc-R {alpha}3-sialyltransferase I/II; ST6GalNAc-II, CMP-sialic acid: Galß1,3GalNAc-R {alpha}6-sialyltransferase; C2GNT, core 2 ß1,6-N-acetylglucosaminyltransferase; PNA, peanut agglutinin; mAb, monoclonal antibody; TF-PAA, Galß1,3 N-acetyl-D-galactosamine conjugated to poly[N-(2-hydroxyethyl)acrylamide]; AMCA, 7-amino-4-methylcoumarin-3-acetic acid; RT-PCR, reverse transcription-PCR. Back

4 Sotiropoulou, G., Anisowicz, A., and Sager, R. Isolation and cloning from human mammary epithelial cells of a complete cDNA sequence homologous to other known sialyltransferases (embl:HS14550). Genebank entry, 1999. Back

Received 11/13/00. Accepted 4/ 2/01.


   REFERENCES
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Springer G. F. T and Tn, general carcinoma autoantigens. Science (Wash. DC), 224: 1198-1206, 1984.[Abstract/Free Full Text]
  2. Cao Y., Karsten U., Liebrich W., Haensch W., Springer G. F., Schlag P. M. Expression of Thomsen-Friedenreich-related antigens in primary and metastatic colorectal carcinomas. A reevaluation. Cancer (Phila.), 76: 1700-1708, 1995.[Medline]
  3. Fukuda M. Possible roles of tumor-associated carbohydrate antigens. Cancer Res., 56: 2237-2244, 1996.[Abstract/Free Full Text]
  4. Victorzon M., Nordling S., Nilsson O., Roberts P. J., Haglund C. Sialyl Tn antigen is an independent predictor of outcome in patients with gastric cancer. Int. J. Cancer, 65: 295-300, 1996.[Medline]
  5. Baldus S. E., Zirbes T. K., Hanisch F. G., Kunze D., Shafizadeh S. T., Nolden S., Monig S. P., Schneider P. M., Karsten U., Thiele J., Holscher A. H., Dienes H. P. Thomsen-Friedenreich antigen presents as a prognostic factor in colorectal carcinoma: a clinicopathologic study of 264 patients. Cancer (Phila.), 88: 1536-1543, 2000.[Medline]
  6. Brockhausen I. Biosynthesis and functions of O-glycans and regulation of mucin antigen expression in cancer. Biochem. Soc. Trans., 25: 871-874, 1997.[Medline]
  7. Brockhausen I., Yang J. M., Burchell J., Whitehouse C., Taylor Papadimitriou J. Mechanisms underlying aberrant glycosylation of MUC1 mucin in breast cancer cells. Eur. J. Biochem., 233: 607-617, 1995.[Medline]
  8. Lloyd K. O., Burchell J., Kudryashov V., Yin B. T., Taylor P. J. Comparison of O-linked carbohydrate chains in MUC-1 mucin from normal breast epithelial cell lines and breast carcinoma cell lines. Demonstration of simpler and fewer glycan chains in tumor cells. J. Biol. Chem., 271: 33325-33334, 1996.[Abstract/Free Full Text]
  9. Yang J. M., Byrd J. C., Siddiki B. B., Chung Y. S., Okuno M., Sowa M., Kim Y. S., Matta K. L., Brockhausen I. Alterations of O-glycan biosynthesis in human colon cancer tissues. Glycobiology, 4: 873-884, 1994.[Abstract/Free Full Text]
  10. Kemmner W., Kruck D., Schlag P. Different sialyltransferase activities in human colorectal carcinoma cells from surgical specimens detected by specific glycoprotein and glycolipid acceptors. Clin. Exp. Metastasis, 12: 245-254, 1994.[Medline]
  11. Kitagawa H., Paulson J. C. Differential expression of five sialyltransferase genes in human tissues. J. Biol. Chem., 269: 17872-17878, 1994.[Abstract/Free Full Text]
  12. Giordanengo V., Bannwarth S., Laffont C., Van Miegem V., Harduin Lepers A., Delannoy P., Lefebvre J. C. Cloning and expression of cDNA for a human Gal(ß1-3)GalNAc {alpha}2,3-sialyltransferase from the CEM T-cell line. Eur. J. Biochem., 247: 558-566, 1997.[Medline]
  13. Bierhuizen M. F., Fukuda M. Expression cloning of a cDNA encoding UDP-GlcNAc: Gal ß 1-3-GalNAc-R (GlcNAc to GalNAc) ß 1-6GlcNAc transferase by gene transfer into CHO cells expressing polyoma large tumor antigen. Proc. Natl. Acad. Sci. USA, 89: 9326-9330, 1992.[Abstract/Free Full Text]
  14. Samyn P. B., Krzewinski R. M., Steelant W. F., Delannoy P., Harduin L. A. Molecular cloning and functional expression of human ST6GalNAc II. Molecular expression in various human cultured cells. Biochim. Biophys. Acta, 1474: 201-211, 2000.[Medline]
  15. Karsten U., Butschak G., Cao Y., Goletz S., Hanisch F. G. A new monoclonal antibody (A78-G/A7) to the Thomsen-Friedenreich pan-tumor antigen. Hybridoma, 14: 37-44, 1995.[Medline]
  16. Chomczynski P., Sacchi N. Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal. Biochem., 162: 156-159, 1987.[Medline]
  17. Shimodaira K., Nakayama J., Nakamura N., Hasebe O., Katsuyama T., Fukuda M. Carcinoma-associated expression of core 2ß-1,6-N-acetylglucosaminyltransferase gene in human colorectal cancer: role of O-glycans in tumor progression. Cancer Res., 57: 5201-5206, 1997.[Abstract/Free Full Text]
  18. Kim Y. J., Kim K. S., Kim S. H., Kim C. H., Ko J. H., Choe I. S., Tsuji S., Lee Y. C. Molecular cloning and expression of human Gal ß 1,3GalNAc {alpha}2,3-sialyltransferase (hST3Gal II). Biochem. Biophys. Res. Commun., 228: 324-327, 1996.[Medline]
  19. Gross H. J., Brossmer R. Characterization of human plasma sialyltransferase using a novel fluorometric assay. Clin. Chim. Acta, 197: 237-247, 1991.[Medline]
  20. Cao Y., Stosiek P., Springer G. F., Karsten U. Thomsen-Friedenreich-related carbohydrate antigens in normal adult human tissues: a systematic and comparative study. Histochem. Cell Biol., 106: 197-207, 1996.[Medline]
  21. Petretti T., Schulze B., Schlag P. M., Kemmner W. Altered mRNA expression of glycosyltransferases in human gastric carcinomas. Biochim. Biophys. Acta, 1428: 209-218, 1999.[Medline]
  22. Petretti T., Kemmner W., Schulze B., Schlag P. M. Altered mRNA expression of glycosyltransferases in human colorectal carcinomas and liver metastases. Gut, 46: 359-366, 2000.[Abstract/Free Full Text]
  23. Kuhns W., Rutz V., Paulsen H., Matta K. L., Baker M. A., Barner M., Granovsky M., Brockhausen I. Processing O-glycan core 1, Galß1-3GalNAc{alpha}-R. Specificities of core 2, UDP-GlcNAc: Gal ß1-3GalNAc-R(GlcNAc to GalNAc) ß6-N-acetylglucosaminyltransferase and CMP-sialic acid: Gal ß1-3GalNAc-R {alpha}3-sialyltransferase. Glycoconj. J., 10: 381-394, 1993.[Medline]



This article has been cited by other articles:


Home page
 GlycobiologyHome page
S. Julien, E. Adriaenssens, K. Ottenberg, A. Furlan, G. Courtand, A.-S. Vercoutter-Edouart, F.-G. Hanisch, P. Delannoy, and X. Le Bourhis
ST6GalNAc I expression in MDA-MB-231 breast cancer cells greatly modifies their O-glycosylation pattern and enhances their tumourigenicity
Glycobiology, January 1, 2006; 16(1): 54 - 64.
[Abstract] [Full Text] [PDF]

Home page
 Cancer Res.Home page
Y. Kakiuchi, S. Tsuji, M. Tsujii, H. Murata, N. Kawai, M. Yasumaru, A. Kimura, M. Komori, T. Irie, E. Miyoshi, et al.
Cyclooxygenase-2 Activity Altered the Cell-Surface Carbohydrate Antigens on Colon Cancer Cells and Enhanced Liver Metastasis
Cancer Res., March 1, 2002; 62(5): 1567 - 1572.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Schneider, F.
Right arrow Articles by Schlag, P. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Schneider, F.
Right arrow Articles by Schlag, P. M.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online