Molecular Cancer Research Holland AM No Date
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 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 Thompson, E. J.
Right arrow Articles by Gendler, S. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Thompson, E. J.
Right arrow Articles by Gendler, S. J.
Molecular Cancer Research 4:489-497 (2006)
© 2006 American Association for Cancer Research

Signaling and Regulation

Tyrosines in the MUC1 Cytoplasmic Tail Modulate Transcription via the Extracellular Signal-Regulated Kinase 1/2 and Nuclear Factor-{kappa}B Pathways

Eric J. Thompson1, Kandavel Shanmugam1, Christine L. Hattrup2, Kari L. Kotlarczyk1, Albert Gutierrez2, Judy M. Bradley1, Pinku Mukherjee1 and Sandra J. Gendler1,2

1 Cancer Center Scottsdale and 2 Mayo Clinic College of Medicine, Mayo Clinic Arizona, Scottsdale, Arizona

Requests for reprints: Sandra J. Gendler, Mayo Clinic Arizona, Johnson Research Building, 13400 East Shea Boulevard, Scottsdale, AZ 85259. Phone: 480-301-7062; Fax: 480-301-7017. E-mail: gendler.sandra{at}mayo.edu


   Abstract
Top
 Notes
 Abstract
Introduction
 Results
 Discussion
 Materials and Methods
 Acknowledgements
 References
 
Much of the ability of the MUC1 oncoprotein to foster tumorigenesis and tumor progression likely originates from the interaction of its cytoplasmic tail with proteins involved in oncogenic signaling. Many of these interactions are regulated by phosphorylation, as the cytoplasmic tail contains seven highly conserved tyrosines and several serine/threonine phosphorylation sites. We have developed a cell line–based model system to study the effects of tyrosine phosphorylation on MUC1 signaling, with particular emphasis on its effects on gene transcription. COS-7 cells, which lack endogenous MUC1, were stably infected with wild-type MUC1 or a MUC1 construct lacking all seven tyrosines (MUC1 Y0) and analyzed for effects on transcription mediated by the extracellular signal-regulated kinase 1/2 (ERK1/2) and nuclear factor-{kappa}B (NF-{kappa}B) pathways. COS.MUC1 Y0 cells showed heightened active ERK1/2 with increased activator protein-1 (AP-1) and signal transducer and activator of transcription 3 (STAT3) transcriptional activity; there was also a simultaneous decrease in NF-{kappa}B transcriptional activity and nuclear localization. These changes altered the phenotype of COS.MUC1 Y0 cells, as this line displayed increased invasion and enhanced [3H]thymidine incorporation. Analysis of the three lines also showed significant differences in their cell cycle profile and bromodeoxyuridine incorporation when the cells were serum starved. These data support the growing evidence that MUC1 is involved in transcriptional regulation and link MUC1 for the first time to the NF-{kappa}B pathway. (Mol Cancer Res 2006;4(7):489–97)


   Introduction
Top
 Notes
 Abstract
 Introduction
Results
 Discussion
 Materials and Methods
 Acknowledgements
 References
 
MUC1 is a tumor antigen and oncoprotein that is overexpressed in most tumors, including breast, pancreatic, ovarian, and colon cancers (1). MUC1 is a heterodimer consisting of a large, glycosylated extracellular domain and a smaller subunit consisting of a short extracellular stem, the transmembrane sequence, and a 72–amino acid cytoplasmic tail (collectively called MUC1-CT). Although the reasons for its oncogenic activity remain unclear, much recent work has focused on the ability of this protein to interact with signaling molecules, primarily through the MUC1-CT. The MUC1-CT has multiple phosphorylation sites, including seven tyrosine residues, and associates with many kinases, including glycogen synthase kinase 3ß (2), protein kinase C{delta} (3), c-Src (4, 5), and all four members of the ErbB family (6, 7). MUC1 interactions with other proteins can be regulated by phosphorylation: for example, MUC1-ß-catenin binding is increased by c-Src, ErbB1, or protein kinase C{delta} phosphorylation but decreased by glycogen synthase kinase 3ß phosphorylation (4). Increased MUC1-ß-catenin interaction after MUC1-CT phosphorylation led to decreased E-cadherin-ß-catenin association and stimulated anchorage-independent growth, showing a functional role for MUC1-CT phosphorylation. An additional argument for the importance of the MUC1-CT tyrosine residues is that six of the seven are 100% conserved in mammals (1, 8), likely reflecting an essential role for tyrosine phosphorylation in regulating MUC1-associated signaling. The overexpression, oncogenic activity, and phosphorylation-dependent interactions of MUC1 make the study of the role of MUC1 tyrosine phosphorylation important.

Among the most interesting recent findings regarding MUC1 oncogenic activity are its presence in the nucleus and its ability to interact with and regulate transcription factors. MUC1 can interact with p53-responsive promoter elements and can regulate transcription through binding to the p53 regulatory domain (9). In addition, the MUC1-CT can be found in the nucleus in association with ß-catenin (10) and can alter transcription of a ß-catenin-driven reporter construct in a manner depending on the presence of Tyr46 in the MUC1-CT (11). These studies support a role for MUC1 that was previously considered unlikely, if not impossible: involvement of a transmembrane glycoprotein—long thought to be involved only in steric modulation of adhesion—in transcriptional regulation in the nucleus.

These studies emphasize the importance of examining MUC1 in a new light: searching for novel signaling pathways that could be affected by MUC1 and seeking to clarify the role of MUC1 in pathways to which it has already been linked. One such signaling network is the extracellular signal-regulated kinase 1/2 (ERK1/2) pathway. Epidermal growth factor treatment of mouse mammary glands resulted in tyrosine phosphorylation of the MUC1-CT, which correlated with activation of ERK1/2 as seen by dual phosphorylation (dpERK1/2; ref. 6). Cell culture studies show similar results, as both expression of a CD8/MUC1 chimera in MUC1-non-expressing cells (12) and binding of Pseudomonas aeruginosa to MUC1 on airway cells (13) increased dpERK1/2 levels. Interestingly, when the seven MUC1-CT tyrosine residues were mutated in this CD8/MUC1 chimera, endogenous dpERK1/2 levels decreased and were not able to be stimulated by CD8 antibody treatment of the mutant chimera. Increased dpERK1/2 has been linked to enhanced proliferation and alterations in cell motility and invasive capacity in part because it can alter gene expression via the signal transducer and activator of transcription 3 (STAT3), the activator protein-1 (AP-1) complex, and other transcription factors (14). STAT3 activity is increased in a wide variety of tumor types, including breast cancer, and is capable of stimulating proliferation while inhibiting apoptosis (15, 16). AP-1 is a heterodimeric transcription factor composed of c-Fos, c-Jun, or other related proteins. Although the specific targets of AP-1 depend on the composition of the heterodimer, in general, AP-1 transcription is thought to signal for increased invasive capacity, proliferation, and cellular survival (17).

The involvement of MUC1 in prosurvival signaling is not surprising, considering that MUC1 has long been thought to play a role in protecting epithelial layers and enhancing epithelial cell survival (1). Mice lacking MUC1 show little or no phenotype when housed in a pathogen-free environment but develop chronic reproductive tract infections when housed in nonsterile conditions (18). Exposure of cultured cells to inflammatory cytokines can up-regulate MUC1 expression, indicating that increased MUC1 levels may be an important part of cellular stress responses (19). MUC1 has also been linked to decreased apoptosis in response to oxidative stress and genotoxic agents (ref. 20; e.g., by suppressing p53-responsive gene targets). An additional mechanism by which MUC1 could affect cellular stress response is through modulation of the nuclear factor-{kappa}B (NF-{kappa}B) pathway. NF-{kappa}B is a heterodimeric transcription factor composed of a variety of subunits from the Rel family of proteins (21). It is heavily involved in regulating response to cellular stress typically through increased transcription of proinflammatory and antiapoptotic targets. Classic regulation of NF-{kappa}B occurs largely through controlling its cellular localization: when inactive, NF-{kappa}B is sequestered in the cytoplasm by I{kappa}B (inhibitors of NF-{kappa}B) proteins. Phosphorylation of I{kappa}B by the I{kappa}B kinase complex causes degradation of I{kappa}B, thus releasing NF-{kappa}B to enter the nucleus and affect transcription of a variety of gene targets. Given the importance of phosphorylation in this network, it is likely that other proteins involved in affecting NF-{kappa}B activity will also show phosphorylation-dependent regulation.

This report describes a model for studying the importance of MUC1-CT tyrosine-based signaling using a non-tumor-derived cell line to clarify the regulation of MUC1 in cells lacking tumor-associated genetic and epigenetic changes. Expression of wild-type MUC1 (MUC1 WT) and a mutant lacking the seven cytoplasmic tail tyrosine residues (MUC1 Y0) in COS-7 cells showed striking differences in transcriptional regulation leading to alterations in invasion and [3H]thymidine incorporation. Our results confirm that MUC1 can modulate the activity of ERK1/2 and its downstream targets, AP-1 and STAT3. Interestingly, the MUC1 Y0 mutant is far more potent than MUC1 WT in activating the ERK1/2 pathway, suggesting the presence of previously undetected, tyrosine-dependent interactions between MUC1 and negative regulators of this pathway. In addition, we describe for the first time the ability of MUC1 to alter NF-{kappa}B-responsive transcription, adding yet another important pathway to the growing list of signaling networks affected by the MUC1 oncoprotein.


   Results
Top
 Notes
 Abstract
 Introduction
 Results
Discussion
 Materials and Methods
 Acknowledgements
 References
 
Recent studies have shown that MUC1 is capable of regulating transcription, which may in part explain the effect of MUC1 on events involved in oncogenesis. To study the importance of MUC1-CT tyrosine phosphorylation in this setting, COS-7 cells were infected with constructs expressing MUC1 WT or the MUC1 Y0 mutant where all seven tyrosine residues in the MUC1-CT were changed to phenylalanine (Fig. 1A ). Empty vector infection (COS.Neo) was used as a control. The COS-7 line was chosen because it does not express endogenous MUC1; it is important to note that although this line was immortalized with SV40 T antigen it was derived from normal kidney cells and therefore lacks many of the genetic and epigenetic alterations that occur in tumor-derived cells. Expression of MUC1 was confirmed by Western blots using an antibody directed against the MUC1-CT (Fig. 1B), whereas proper surface localization of both MUC1 constructs was seen by flow cytometry on nonpermeabilized cells (Fig. 1C). Note that the cytoplasmic tail of MUC1 Y0 shows a shift in electrophoretic mobility relative to MUC1 WT; this likely reflects a decrease in phosphorylation due to mutation of the tyrosine residues. Several bands appear in the blot for the MUC1-CT, which are thought to represent differences in glycosylation or other post-translational modifications of this subunit.


Figure 1
View larger version (46K):
[in this window]
[in a new window]
  		FIGURE 1. MUC1 WT and MUC1 Y0 can be expressed in COS-7 cells. A. Sequence of the MUC1 WT cytoplasmic tail and the MUC1 Y0 mutant, which has the seven tyrosine residues changed to phenylalanine (highlighted). B. COS-7 cells stably expressing Neo control, MUC1 WT, or MUC1 Y0 constructs. Whole-cell lysate was blotted for expression of the MUC1 cytoplasmic tail using the CT2 monoclonal antibody. C. Cell surface expression of MUC1 was confirmed by fluorescence-activated cell sorting. Nonpermeabilized cells were stained with HMPV-FITC antibody, which is directed against the MUC1 extracellular domain: COS.Neo, COS.MUC1 WT (unshaded), and COS.MUC1 Y0 (shaded). D. Analysis of tyrosine phosphorylation in immunoprecipitations with a MUC1 antibody (CT2) or hamster preimmune serum. Immunoprecipitations were blotted for phosphotyrosine (top) or MUC1-CT (bottom). Representative of three independent experiments.

 
Given that the MUC1-CT can be tyrosine phosphorylated in cells expressing endogenous MUC1, we analyzed whether this is also true of the transfected protein. Immunoprecipitations for MUC1 from all three cell lines (Fig. 1D) were blotted for phosphotyrosine and the MUC1-CT. Only COS.MUC1 WT cells exhibit a phosphotyrosine band that overlaps with the MUC1-CT; as expected, COS.MUC1 Y0 cells show no phosphotyrosine, whereas COS.Neo cells display neither MUC1 nor phosphotyrosine at this size. The faint phosphotyrosine band seen in all lanes is most likely immunoglobulin light chain, which runs slightly above the MUC1-CT. Immunoprecipitations with hamster preimmune serum were used as controls; a faint band at the approximate size of the MUC1-CT appears in the preimmune serum precipitation from COS.MUC1 WT, likely representing a nonspecific association that is much weaker than the specific precipitation of MUC1 from these cells.

Much work done in tumor-derived cell lines and in mouse tumor models points to a link between MUC1 and the ERK1/2 kinases. We therefore examined the consequences of MUC1 WT and MUC1 Y0 expression on these proteins and their downstream effectors. Western blots for dpERK1/2 (Fig. 2A ) show a mild increase in ERK1/2 activation in COS.MUC1 WT cells relative to COS.Neo as expected based on previous reports (6, 22). Surprisingly, COS.MUC1 Y0 cells display a substantial increase in dpERK1/2 levels compared with either of the other two cell lines. ERK 2 (p42) seems to be activated more strongly than ERK1 (p44) by expression of either MUC1 WT or MUC1 Y0. ERK1/2 total protein levels are consistent across all three cell lines. This increase in dpERK1/2 is serum dependent but not responsive to epidermal growth factor treatment, as these cell lines show identical dpERK1/2 levels with this treatment (data not shown). To confirm that the increase in dpERK1/2 corresponds to enhanced transcriptional activity of ERK1/2 effectors, COS.Neo, COS.MUC1 WT, and COS.MUC1 Y0 cells were transiently transfected with a reporter construct containing the luciferase gene driven by six tandem AP-1 consensus DNA-binding sites. Although there is little difference in AP-1 activity between COS.Neo and COS.MUC1 WT, the COS.MUC1 Y0 line have a 4- to 5-fold increase in AP-1 activity (Fig. 2B), corroborating the striking increase in dpERK1/2 in these cells. A second factor, which can mediate transcription downstream of ERK1/2, is STAT3. Transcription of a STAT3-responsive reporter was also increased in COS.MUC1 Y0 cells compared with COS.MUC1 WT and COS.Neo (Fig. 2C).


Figure 2
View larger version (16K):
[in this window]
[in a new window]
  		FIGURE 2. ERK1/2 phosphorylation and activity of AP-1 and STAT3 are increased in COS.MUC1 Y0. A. Whole-cell lysates were blotted for dpERK1/2 and total ERK1/2 (arrows). The same membrane was used for both blots. B. AP-1 transcriptional activity was measured in COS.Neo, COS.MUC1 WT, and COS.MUC1 Y0 cells by transfection of a luciferase reporter driven by AP-1 consensus sites. Relative luciferase units for COS.Neo were set to 1; data for the other lines are fold change relative to COS.Neo. C. Transcriptional activity of STAT3 was measured as described in B. *, P < 0.001, compared with COS.Neo.

 
Given the important role of the ERK1/2 pathway in cellular proliferation and survival, we examined the effect of MUC1 WT and MUC1 Y0 expression on other proteins involved in these events. No differences were seen in levels of p53, p21, or p27 in the COS-derived cell lines (data not shown), likely excluding MUC1 regulation of p53 as an important pathway in these cells. In contrast, a luciferase reporter responsive to NF-{kappa}B indicated that MUC1 Y0 expression may alter basal NF-{kappa}B transcription activity. COS.MUC1 WT cells show no significant change in NF-{kappa}B activity compared with COS.Neo (Fig. 3A ). Activity of NF-{kappa}B in COS.MUC1 Y0 cells, however, is at approximately one fifth the level seen in the other two cell lines, indicating that mutation of the MUC1-CT tyrosine residues has a negative influence on NF-{kappa}B-mediated transcription. Reverse transcription-PCR array analysis of a NF-{kappa}B target gene, interleukin-8, showed greatly reduced transcription in COS.MUC1 Y0 cells compared with the other lines (data not shown). To corroborate these data, we analyzed nuclear localization of the p65 subunit of NF-{kappa}B under steady-state conditions. As expected, basal levels of nuclear p65 are relatively low in all three cell lines, but there is significantly less nuclear p65 in the COS.MUC1 Y0 cells compared with COS.Neo or COS.MUC1 WT (Fig. 3B and C).


Figure 3
View larger version (44K):
[in this window]
[in a new window]
  		FIGURE 3. NF-{kappa}B activity and nuclear localization are decreased in COS.MUC1 Y0. A. Transcription of a luciferase reporter driven by NF-{kappa}B consensus sites was assessed in COS.Neo, COS.MUC1 WT, and COS.MUC1 Y0 cells. Relative luciferase units for COS.Neo were set to 1; data for the other lines are fold change relative to COS.Neo. B. Immunocytochemical staining of the p65 subunit of NF-{kappa}B. Areas of colocalization of PI and anti-p65-FITC were colored white for clarity (Overlay). C. Quantitation of nuclear p65 shown as box plots. Staining was quantitated using the FITC fluorescence intensity of 150 PI-positive nuclei. Center line, mean relative fluorescence units (RFU) from 150 nuclei; top and bottom boundaries, range of data from the 25th to 75th percentile relative fluorescence unit scores. *, P < 0.001, compared with COS.Neo.

 
With the striking differences seen in AP-1 and NF-{kappa}B transcriptional activity, we wanted to examine cellular events tied to these pathways. Given that AP-1 transcription is capable of increasing the invasive potential of cells and that a previous report showed that MUC1-derived peptides could stimulate invasion (23), this phenomenon was examined using a standard Transwell assay. Serum-starved cells were plated on top of Matrigel, with medium containing 10% serum in the lower chamber as an attractant. Not surprisingly based on the transcriptional profile, there was little change seen in COS.MUC1 WT cells compared with COS.Neo (Fig. 4 ). However, a striking increase in invasion was seen in COS.MUC1 Y0 cells relative to the other lines, with ~80% of COS.MUC1 Y0 cells successfully invading into the matrix.


Figure 4
View larger version (16K):
[in this window]
[in a new window]
  		FIGURE 4. COS.MUC1 Y0 cells display increased invasive potential. Serum-starved cells were plated in Transwell chambers coated with Matrigel, with serum-containing medium in the lower chamber as an attractant. Invasion was determined as the percentage of cells that successfully entered the Matrigel. Columns, average of three independent experiments. *, P < 0.001, compared with COS.Neo.

 
We next studied the effect of MUC1 WT and MUC1 Y0 expression on proliferation and survival. [3H]Thymidine incorporation assays revealed a 10- to 20-fold increase in nucleotide incorporation in the COS.MUC1 Y0 line compared with COS.Neo and COS.MUC1 WT (Fig. 5A ). This result would seem to reflect a substantial gain in proliferation in response to MUC1 Y0 expression. However, growth curves obtained for all three lines showed only slightly higher cell counts in the COS.MUC1 Y0 line, with no difference seen until the seventh day in culture (Fig. 5B). The disparity between these results is not due to increased cell death, as baseline apoptosis was not higher in the COS.MUC1 Y0 cells compared with the other two lines (data not shown). Cell cycle analysis indicated a decrease in the percentage of COS.MUC1 Y0 cells in G0-G1 phase, with a simultaneous increase in the percentage of cells in G2-M (Fig. 5C). No significant differences in cell cycle profile exist between the COS.Neo and COS.MUC1 WT lines.


Figure 5
View larger version (38K):
[in this window]
[in a new window]
  		FIGURE 5. [3H]Thymidine incorporation and cell cycle profile are altered in COS.MUC1 Y0. A. Cells plated at two different densities (5,000 or 25,000 per well) were incubated with [3H]thymidine and harvested 24 hours later. Uptake of radiolabel is shown in counts/min (cpm). B. COS.Neo, COS.MUC1 WT, and COS.MUC1 Y0 cells were plated in triplicate and counted at 24-hour intervals for 7 days. C. Cells were stained with PI and analyzed by fluorescence-activated cell sorting for cell cycle distribution. Results are percentage of cells in G0-G1, S, or G2-M phases. *, P < 0.01, compared with COS.Neo.

 
To examine these results, we studied bromodeoxyuridine (BrdUrd) incorporation in these cells either in the presence of serum or after 24 hours of serum starvation. BrdUrd is incorporated into DNA both during DNA synthesis and in other processes, such as DNA repair. COS.Neo, COS.MUC1 WT, and COS.MUC1 Y0 show little difference in BrdUrd incorporation when maintained in normal growth medium (data not shown). However, after 24 hours of serum starvation, there is an ~2-fold difference in BrdUrd incorporation in the COS.MUC1 Y0 cells compared with the other cell lines (Fig. 6A ). Quantification of BrdUrd-positive nuclei confirms statistical significance of these results (Fig. 6B).


Figure 6
View larger version (21K):
[in this window]
[in a new window]
  		FIGURE 6. COS.MUC1 Y0 show increased BrdUrd incorporation. A. Immunocytochemical staining of BrdUrd incorporation. Serum-starved cells were treated with BrdUrd 1 hour before fixation. Areas of colocalization of PI (red) and anti-BrdUrd (green) are shown in yellow. Confocal settings were determined using the COS.Neo line and maintained for all three cell lines to ensure comparability. B. Quantitation of BrdUrd incorporation shown as box plots. Staining was quantitated using the FITC fluorescence intensity of 150 PI-positive nuclei. Center line, mean relative fluorescence units from 150 nuclei; top and bottom boundaries, range of data from the 25th to 75th percentile relative fluorescence unit scores. *, P < 0.001, compared with COS.Neo.

 

   Discussion
Top
 Notes
 Abstract
 Introduction
 Results
 Discussion
Materials and Methods
 Acknowledgements
 References
 
Although it has long been known that MUC1 is highly overexpressed in a wide range of tumors, it is only very recently that some of the effects of MUC1 up-regulation have come to be understood. MUC1 levels and MUC1-CT tyrosine phosphorylation have been linked to ERK1/2 activity in mice (6), whereas phosphorylation of the MUC1-CT by Src, ErbB1, or protein kinase C{delta} regulates its association with proteins, such as ß-catenin and glycogen synthase kinase 3ß, in cell lines (3, 4, 7). Protein kinase C{delta}–mediated phosphorylation of MUC1 led to anchorage-independent cell growth, which likely resulted from decreased association of ß-catenin with E-cadherin on MUC1-CT phosphorylation. Despite these studies, however, it remains unclear what the functional consequences of MUC1-CT tyrosine phosphorylation are. Similarly, it is unknown whether all seven MUC1-CT tyrosine residues are phosphorylated or which of the residues is most important for MUC1-related signaling. These studies were designed to examine both transcriptional activity and cell behavior in response to exogenous expression of MUC1 WT, which shows constitutive tyrosine phosphorylation in this system compared with the tyrosine-lacking mutant MUC1 Y0.

Interestingly, the COS-7-derived cell lines seem to show differential regulation of MUC1 compared with many other models. Given previous reports, we expected that MUC1 WT would result in increased ERK1/2 activity leading to increased proliferation and invasion. We also expected that mutation of the MUC1-CT tyrosine residues would diminish or abrogate these effects. There are several possible reasons to explain why our data disagree with expectations. First, it is important to note that prior studies (e.g., refs. 6, 12) were done largely in cancer-derived or embryonic cell lines or in mouse tumor models. COS-7 cells are SV40-immortalized derivatives of normal adult kidney fibroblasts and would therefore lack many of the genetic and epigenetic alterations that accompany oncogenesis. COS-7 cells are not normal, but the results seen in this line may more accurately reflect the role of MUC1 in non-tumor-derived systems, although the possibility exists that SV40 T antigen may affect MUC1-related signaling. Second, COS-7 cells are derived from male green monkey kidney, so species- or cell type–specific regulation could be involved. Finally, our studies use the entire MUC1 molecule, in contrast to other reports examining chimeric MUC1 proteins. One such study in COS-7 cells noted that anti-CD8 stimulation of the CD8/MUC1 chimera activated ERK2 through the Ras pathway (22), supporting our finding that MUC1 is linked to this pathway in these cells. Although such chimeras are useful for studying alterations in the MUC1-CT after extracellular stimulation, it is not clear whether this accurately represents the behavior of intact MUC1, making study of the entire molecule very important.

Many of the interactions attributed to MUC1 involve phosphorylation of the MUC1-CT. The fact that mutation of the tyrosine residues in the MUC1 Y0 construct does not cause the expected decrease in survival or invasion likely reflects a unique advantage of the COS-7 model system that has been lost or masked in tumor-based models. Specifically, the MUC1-CT tyrosine residues may in fact be involved in one or more previously uncharacterized interactions with inhibitory factors that cannot associate with the MUC1 Y0 mutant. Examples of such factors could include phosphatases or proteins capable of mislocalizing MUC1 or its signaling partners. In normal or close-to-normal cells, these inhibitory factors could down-regulate the potentially oncogenic signals associated with MUC1 expression, which would explain the failure of the COS.MUC1 WT cell line to show altered transcription via the ERK1/2 or NF-{kappa}B pathways. Lacking tyrosine residues, MUC1 Y0 would not face inhibition and could therefore stimulate survival and invasion using pathways that do not require MUC1 tyrosine phosphorylation. Alternatively, MUC1 Y0 may bind and sequester inhibitory factors that do not associate with MUC1 WT, thereby preventing these factors from dampening proliferative or invasive cellular signaling. As tyrosine-dependent inhibition of MUC1 signaling has not been described in other models, the factor(s) responsible may be decreased in tumor-based systems or may be prevented from interacting with the MUC1-CT due to altered signaling or competition with other MUC1-binding proteins.

An alternative possibility is that the MUC1-CT tyrosine residues may be required for proper "coupling" between the extracellular and CT domains. A previous report (24) found that deletion of either the MUC1 tandem repeats or cytoplasmic sequence resulted in increased invasion and metastasis in vivo, suggesting that proper control of MUC1 function may depend on coordinate regulation of both subunits of the protein. Mutation of the MUC1-CT tyrosines may disrupt the link between the intracellular and extracellular portions of MUC1, resulting in misregulation of its adhesive properties. In agreement with this idea, expression of COS.MUC1 Y0 resulted in enhanced invasion into Matrigel. Intriguingly, COS.MUC1 Y0 cells also show a dramatic enhancement of ERK1/2 phosphorylation and activity of the STAT3 and AP-1 transcription factors, which could contribute to the increased invasion seen in this cell line. ERK1/2 activity can alter several pathways capable of driving tumor progression, including transcription of invasion-stimulating molecules, such as matrix metalloproteinases (14).

MUC1 expression has been linked to increased ERK activity before, although the mechanism by which this occurs remains largely unexplored. P. aeruginosa binding to MUC1 in airway cells (13) and antibody stimulation of a CD8-MUC1-CT chimera in embryonic cells (12) activated ERK1/2 downstream of MUC1. Similarly, we have shown that overexpression of human MUC1 in the mouse mammary gland results in increased dpERK1/2 compared with wild-type or MUC1-null glands (6). Note that these studies saw the most striking increase in ERK1/2 activation after stimulation, whereas this report shows alteration only in baseline dpERK1/2. The identical responses of the COS-7-derived cell lines after epidermal growth factor stimulation support the idea that these cells show regulation of MUC1 that was not characterized in previous studies, as several groups have shown that MUC1 tyrosine phosphorylation is important in this pathway (6, 7). The ability of MUC1 Y0, but not MUC1 WT, to stimulate ERK1/2 signaling in COS-7 cells could reflect a tyrosine-dependent interaction of MUC1 with proteins involved in inhibiting ERK1/2 activity, such as the dual-specificity phosphatases capable of inactivating dpERK1/2.

The increased ERK1/2 activity in COS.MUC1 Y0 cells would suggest that the sharp increase in [3H]thymidine incorporation in this line should correspond to enhanced proliferation. It is intriguing that the growth curves do not reflect a significant increase in proliferation in COS.MUC1 Y0 cells compared with the other two lines. Routine passaging in cell culture also indicates similar growth rates among the three cell lines. The [3H]thymidine must therefore be involved in nonproliferative pathways, such as DNA repair or synthesis of DNA that is not associated with cell division (e.g., endoreduplication, telomere synthesis, or incorporation into the mitochondrial genome). Some light is shed on this riddle by the results of the BrdUrd incorporation studies. Given that COS-7 cells cycle quite rapidly, the majority of BrdUrd incorporation in the presence of serum is likely due to DNA synthesis. It is therefore not surprising that the three lines showed little difference in BrdUrd incorporation in normal growth medium. However, as the rate of proliferation slows after serum removal, other causes for BrdUrd incorporation are more likely to be visible. The increased incorporation of both [3H]thymidine and BrdUrd in COS.MUC1 Y0 may therefore reflect higher levels of DNA repair in these cells. This might explain the change in the cell cycle profile of the COS.MUC1 Y0 line: the apparent accumulation of cells in G2-M could reflect a repair-related checkpoint that prevents these cells from completing mitosis, which would therefore clarify the failure of this line to proliferate more rapidly than the other two. If so, this would be the first study to suggest a connection between MUC1 and genomic repair mechanisms. Studies are ongoing to confirm that the altered nucleotide incorporation does indeed reflect an involvement in DNA repair.

This is also the first report to correlate MUC1 expression with regulation of NF-{kappa}B activity. Although preliminary, it seems that MUC1 can alter NF-{kappa}B-responsive reporter activity, p65 nuclear localization, and transcription of an endogenous target gene, interleukin-8. It is not yet clear what role MUC1 plays in this pathway, but there are at least two mechanisms by which it could be proposed to affect NF-{kappa}B activity and localization. First, MUC1 could modulate association of proteins within this pathway either by directly influencing NF-{kappa}B itself or its regulatory proteins. MUC1 could interact with members of the I{kappa}B kinase or I{kappa}B families to alter their regulation of NF-{kappa}B localization, for example. Second, MUC1 could directly affect NF-{kappa}B localization by changing its rate of transport through the nuclear pore complex. Although the MUC1-CT lacks a nuclear localization signal, binding to NF-{kappa}B could allow it to alter the accessibility of the NF-{kappa}B nuclear localization signal. If the MUC1-CT is not tyrosine phosphorylated, as is the case with MUC1 Y0, it could mask the NF-{kappa}B nuclear localization signal directly or cause a conformational change in NF-{kappa}B that hides the nuclear localization signal. Phosphorylation of the MUC1-CT could then release this masking effect or cause a different conformational change that reveals the NF-{kappa}B nuclear localization signal, thus facilitating transport into the nucleus. Studies are ongoing to clarify the role of MUC1 in NF-{kappa}B signaling.

The finding that mutation of the MUC1-CT tyrosine residues affects basal NF-{kappa}B activity indicates for the first time that MUC1 can influence this important cellular pathway. In addition to this novel result, we confirm previous reports showing that MUC1 regulates the ERK1/2 pathway and present evidence that MUC1 can alter AP-1-mediated transcription. These data suggest that MUC1 tyrosine phosphorylation may be involved in previously uncharacterized negative regulatory interactions that dampen ERK1/2 activity. Finally, our results also hint at a role for MUC1 in regulating DNA repair, although confirmation of this phenomenon is still in progress. The possible significance of the effect of MUC1 on the ERK1/2 and NF-{kappa}B pathways is apparent, as these networks are tied to a multitude of cellular events important for tumorigenesis and metastasis. Further study of the influence of tyrosine phosphorylation on the MUC1 oncoprotein will likely clarify its important role in tumor formation and progression.


   Materials and Methods
Top
 Notes
 Abstract
 Introduction
 Results
 Discussion
 Materials and Methods
Acknowledgements
 References
 
Cloning of MUC1 WT and MUC1 Y0 Vectors
MUC1 Y0 was created using the QuickChange mutagenesis kit (Stratagene, La Jolla, CA). Briefly, primers based on the MUC1 sequence were designed containing single-base alterations resulting in mutation of the cytoplasmic tail tyrosine residues to phenylalanine. Successful mutation was confirmed with DNA sequencing. MUC1 Y0 and MUC1 WT were cloned into the pLNCX.1 vector for retroviral infection.

Cell Culture and Retroviral Infection
COS-7 cells (American Type Culture Collection, Manassas, VA) were maintained in DMEM (Invitrogen, Carlsbad, CA) supplemented with 10% FCS, 1% glutamax (Invitrogen), and 1% penicillin/streptomycin. Cell counts were done every 24 hours after plating, with three wells of each line counted in triplicate (i.e., a total of nine counts per line each day). For retroviral infection, GP2-293 packaging cells (stably expressing the gag and pol proteins) were cotransfected with the appropriate MUC1 construct and a vector expressing the VSV-G envelope protein. After 48 hours, medium was removed from the transfected packaging cells and cleared of debris by centrifugation at 3,000 rpm. Virus was pelleted from the cleared medium; this was resuspended in growth medium containing 8 µg/mL polybrene (hexadimethrine bromide) and incubated overnight with COS-7 cells that had been pretreated for 2 to 3 hours with polybrene. COS-7 cells were selected with 0.5 mg/mL G418, beginning 48 hours postinfection, and cells were maintained as polyclonal populations. Two independent infections of all three constructs were done, with similar results; expression of the infected constructs was stable throughout the span of experiments. For phosphotyrosine analysis, cells were treated for 30 minutes before lysis with 200 nmol/L sodium orthovanadate (25).

Western Blots, Immunoprecipitations, and Antibodies
Briefly, cells were lysed in HEPES buffer (20 mmol/L HEPES, 150 mmol/L NaCl, 1% Triton X-100, 2 mmol/L EDTA) containing protease (Complete inhibitor cocktail; Roche, Indianapolis, IN) and phosphatase inhibitors (10 mmol/L sodium fluoride, 2 mmol/L sodium vanadate, 50 µmol/L ammonium molybdate). Protein concentration was determined by BCA assay (Pierce, Rockford, IL) and equal quantities of lysate were loaded on SDS-PAGE gels. For immunoprecipitation, 1 mg protein was incubated with an antibody specific to MUC1 or preimmune serum in TNEN [50 mmol/L Tris, 150 mmol/L NaCl, 1 mmol/L EDTA, 0.5% NP40 (pH 7.4)]. Antibody complexes were captured with protein G agarose beads (Pierce) and eluted in 2x reducing sample buffer for loading onto gels. The MUC1 cytoplasmic tail antibody, CT2, was made in-house by the Mayo Clinic Arizona Immunology Core (6). Antibodies to phosphotyrosine (BD Biosciences, San Jose, CA), ERK1/2 (Cell Signaling, Danvers, MA), NF-{kappa}B (Santa Cruz Biotechnology, Santa Cruz, CA), and dpERK1/2 (Sigma, St. Louis, MO) were used according to manufacturer's recommendations. Fluorescence-activated cell sorting analysis of MUC1 was done with a FITC-conjugated antibody (HMPV) that recognizes an epitope in the extracellular domain of MUC1 (PharMingen, San Jose, CA).

Flow Cytometry
Staining for MUC1 surface localization was done on cells that were trypsinized but not permeabilized. For cell cycle analysis, cells were serum starved 24 hours, fixed in –20°C ethanol, treated with RNase A to remove RNA, and stained with propidium iodide (PI; 1 mg/mL in 0.1% sodium citrate). Flow cytometry for all assays was done on a FACScan flow cytometer. Cell cycle data were analyzed using the ModFit software program.

Immunocytochemistry
Cells were plated on chamber slides and grown to the desired confluence for BrdUrd incorporation or p65 NF-{kappa}B staining. For BrdUrd incorporation, cells were either maintained in normal medium or serum starved for 24 hours. After 1-hour incubation with BrdUrd (50 µmol/L), cells were washed, fixed in –20°C ethanol, permeabilized with 0.5% Tween 20, and stained with undiluted anti-BrdUrd antibody (a kind gift of Dr. R. Fonseca, Mayo Clinic Arizona) and anti-mouse Alexa 488 secondary (Invitrogen). For p65 staining, cells maintained in complete growth medium were fixed, permeabilized in –20°C methanol, and stained with anti-p65 primary (Santa Cruz Biotechnology) and anti-mouse Alexa 488 secondary antibodies. Quantitation for both stains was done by determining the FITC fluorescence intensity in 150 PI-positive nuclei (50 nuclei in each of three x200 fields).

Invasion Assays
Transwell chambers (BD Biosciences) were coated with Matrigel (Fisher, Houston, TX). Cells were plated in serum-free medium over the Matrigel and cultured for 24 hours. Medium containing 10% serum was used in the bottom well as an attractant. After incubation, noninvaded cells and Matrigel were removed from half of the wells (samples, containing only invaded cells) and retained in the other half (controls, containing all cells). Membranes were stained with 0.5% crystal violet in 20% methanol, washed, and destained in 10% acetic acid. Samples and controls were loaded in quadruplicate into 96-well plates, which were read at 570 nm. Percent invasion was determined as the average sample absorbance over the average control absorbance multiplied by 100.

[3H]Thymidine Incorporation Assays
Cells were plated in quadruplicate at low density (5 x 103 or 25 x 103) in normal growth medium in 96-well plates. [3H]Thymidine was added in fresh medium (1 µCi/well) 24 hours after plating and cells were permitted to grow for another 24 hours. At this time, cells were washed to remove excess radioactivity, trypsinized, and harvested to a filter plate, which was then read on a TopCount plate reader.

Luciferase Reporter Assays
Cells were transiently transfected with constructs expressing the luciferase reporter gene under the control of consensus binding sites for the appropriate transcription factors. Cotransfection with ß-galactosidase was used to control for transfection efficiency. Cells were lysed in reporter lysis buffer (Promega, Madison, WI) 48 hours after transfection, and equal volumes of lysate were added in triplicate to 96-well OptiPlates (Packard, Meridien, CT). Luciferase substrate (Promega) was added to each well and plates were read on a luminometer. The Gal-Screen kit (Tropix, Foster City, CA) was used to determine ß-galactosidase activity, and values were normalized by dividing the average luciferase activity for each sample by its average ß-galactosidase activity. For each experiment, relative luciferase values for COS.Neo control cells were set to 1 and values for the other cell lines were calculated as fold increase or decrease compared with COS.Neo.

Statistical Analysis
Statistics were analyzed with JMP 5.1.2 software (SAS Institute, Inc., Cary, NC). Ps were generated using the two-tailed Student's t test, and significance was confirmed using the Wilcoxon rank sum and Pearson {chi}2 tests.


   Acknowledgements
Top
 Notes
 Abstract
 Introduction
 Results
 Discussion
 Materials and Methods
 Acknowledgements
References
 
We thank Joseph Loftus for the retroviral expression system, Marvin Ruona for graphics preparation, Gargi Basu, Greg Ahmann, and Carole Rohl for experimental assistance, and Irene Beauvais for administrative assistance.


   Notes
Top
 Notes
Abstract
 Introduction
 Results
 Discussion
 Materials and Methods
 Acknowledgements
 References
 
Grant support: American Cancer Society Steven S. Bielfelt postdoctoral fellowship PF-04-256-01-MGO (E.J. Thompson), Department of Defense Breast Cancer Research predoctoral award W81XWH-04-1-0300 (C.L. Hattrup), and NIH R01 grant CA64389 (S.J. Gendler).

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.

Note: E.J. Thompson and K. Shanmugam contributed equally to this work.

Current address for E.J. Thompson: AZ Biodesign Institute, Arizona State University, Tempe, AZ.

Received 2/ 8/06; revised 4/25/06; accepted 5/10/06.


   References
Top
 Notes
 Abstract
 Introduction
 Results
 Discussion
 Materials and Methods
 Acknowledgements
 References
 

  1. Gendler SJ. MUC1, the renaissance molecule. J Mammary Gland Biol Neoplasia 2001;6:339–53.[CrossRef][Medline]
  2. Li Y, Bharti A, Chen D, Gong J, Kufe D. Interaction of glycogen synthase kinase 3ß with the DF3/MUC1 carcinoma-associated antigen and ß-catenin. Mol Cell Biol 1998;18:7216–24.[Abstract/Free Full Text]
  3. Ren J, Li Y, Kufe D. Protein kinase C{delta} regulates function of the DF3/MUC1 carcinoma antigen in ß-catenin signaling. J Biol Chem 2002;277:17616–22.[Abstract/Free Full Text]
  4. Li Y, Kuwahara H, Ren J, Wen G, Kufe D. The c-Src tyrosine kinase regulates signaling of the human DF3/MUC1 carcinoma-associated antigen with GSK3ß and ß-catenin. J Biol Chem 2001;276:6061–4.[Abstract/Free Full Text]
  5. Al Masri A, Gendler SJ. Muc1 affects c-Src signaling in PyV MT-induced mammary tumorigenesis. Oncogene 2005;24:5799–808.[CrossRef][Medline]
  6. Schroeder JA, Thompson MC, Gardner MM, Gendler SJ. Transgenic MUC1 interacts with epidermal growth factor receptor and correlates with mitogen-activated protein kinase activation in the mouse mammary gland. J Biol Chem 2001;276:13057–64.[Abstract/Free Full Text]
  7. Li Y, Ren J, Yu W, et al. The epidermal growth factor receptor regulates interaction of the human DF3/MUC1 carcinoma antigen with c-Src and ß-catenin. J Biol Chem 2001;276:35239–42.[Abstract/Free Full Text]
  8. Spicer AP, Duhig T, Chilton BS, Gendler SJ. Analysis of mammalian MUC1 genes reveals potential functionally important domains. Mamm Genome 1995;6:885–8.[CrossRef][Medline]
  9. Wei X, Xu H, Kufe D. Human MUC1 oncoprotein regulates p53-responsive gene transcription in the genotoxic stress response. Cancer Cell 2005;7:167–78.[CrossRef][Medline]
  10. Wen Y, Caffrey TC, Wheelock MJ, Johnson KR, Hollingsworth MA. Nuclear association of the cytoplasmic tail of MUC1 and ß-catenin. J Biol Chem 2003;278:38029–39.[Abstract/Free Full Text]
  11. Huang L, Ren J, Chen D, Li Y, Kharbanda S, Kufe D. MUC1 cytoplasmic domain coactivates Wnt target gene transcription and confers transformation. Cancer Biol Ther 2003;2:702–6.[Medline]
  12. Wang H, Lillehoj EP, Kim KC. MUC1 tyrosine phosphorylation activates the extracellular signal-regulated kinase. Biochem Biophys Res Commun 2004;321:448–54.[CrossRef][Medline]
  13. Lillehoj EP, Kim H, Chun EY, Kim KC. Pseudomonas aeruginosa stimulates phosphorylation of the airway epithelial membrane glycoprotein Muc1 and activates MAP kinase. Am J Physiol Lung Cell Mol Physiol 2004;287:L809–15.[Abstract/Free Full Text]
  14. Viala E, Pouyssegur J. Regulation of tumor cell motility by ERK mitogen-activated protein kinases. Ann N Y Acad Sci 2004;1030:208–18.[CrossRef][Medline]
  15. Hodge DR, Hurt EM, Farrar WL. The role of IL-6 and STAT3 in inflammation and cancer. Eur J Cancer 2005;41:2502–12.[CrossRef][Medline]
  16. McCubrey JA, May WS, Duronio V, Mufson A. Serine/threonine phosphorylation in cytokine signal transduction. Leukemia 2000;14:9–21.[CrossRef][Medline]
  17. Kundu JK, Surh YJ. Molecular basis of chemoprevention by resveratrol: NF{kappa}B and AP-1 as potential targets. Mutat Res 2004;555:65–80.[Medline]
  18. DeSouza MM, Surveyor GA, Price RE, et al. MUC1/episialin: a critical barrier in the female reproductive tract. J Reprod Immunol 1999;45:127–58.[Medline]
  19. O'Connor JC, Julian J, Lim SD, Carson DD. MUC1 expression in human prostate cancer cell lines and primary tumors. Prostate Cancer Prostatic Dis 2005;8:36–44.[CrossRef][Medline]
  20. Yin L, Huang L, Kufe D. MUC1 oncoprotein activates the FOXO3a transcription factor in a survival response to oxidative stress. J Biol Chem 2004;279:45721–7.[Abstract/Free Full Text]
  21. Pande V, Ramos MJ. NF-{kappa}B in human disease: current inhibitors and prospects for de novo structure based design of inhibitors. Curr Med Chem 2005;12:357–74.[Medline]
  22. Meerzaman D, Shapiro PS, Kim KC. Involvement of the MAP kinase ERK2 in MUC1 mucin signaling. Am J Physiol Lung Cell Mol Physiol 2001;281:L86–91.[Abstract/Free Full Text]
  23. Schroeder JA, Adriance MC, Thompson MC, Camenisch TD, Gendler SJ. MUC1 alters ß-catenin-dependent tumor formation and promotes cellular invasion. Oncogene 2003;22:1324–32.[CrossRef][Medline]
  24. Kohlgraf KG, Gawron AJ, Higashi M, et al. Contribution of the MUC1 tandem repeat and cytoplasmic tail to invasive and metastatic properties of a pancreatic cancer cell line. Cancer Res 2003;63:5011–20.[Abstract/Free Full Text]
  25. Wang H, Lillehoj EP, Kim KC. Identification of four sites of stimulated tyrosine phosphorylation in the MUC1 cytoplasmic tail. Biochem Biophys Res Commun 2003;310:341–6.[CrossRef][Medline]



This article has been cited by other articles:


Home page
 J. Biol. Chem.Home page
P. K. Singh, M. E. Behrens, J. P. Eggers, R. L. Cerny, J. M. Bailey, K. Shanmugam, S. J. Gendler, E. P. Bennett, and M. A. Hollingsworth
Phosphorylation of MUC1 by Met Modulates Interaction with p53 and MMP1 Expression
J. Biol. Chem., October 3, 2008; 283(40): 26985 - 26995.
[Abstract] [Full Text] [PDF]

Home page
 J. Immunol.Home page
T. L. Tinder, D. B. Subramani, G. D. Basu, J. M. Bradley, J. Schettini, A. Million, T. Skaar, and P. Mukherjee
MUC1 Enhances Tumor Progression and Contributes Toward Immunosuppression in a Mouse Model of Spontaneous Pancreatic Adenocarcinoma
J. Immunol., September 1, 2008; 181(5): 3116 - 3125.
[Abstract] [Full Text] [PDF]

Home page
 Mol Cancer ResHome page
Q. Shen, J. J. Rahn, J. Zhang, N. Gunasekera, X. Sun, A. R.E. Shaw, M. J. Hendzel, P. Hoffman, A. Bernier, and J. C. Hugh
MUC1 Initiates Src-CrkL-Rac1/Cdc42-Mediated Actin Cytoskeletal Protrusive Motility after Ligating Intercellular Adhesion Molecule-1
Mol. Cancer Res., April 1, 2008; 6(4): 555 - 567.
[Abstract] [Full Text] [PDF]

Home page
 Hum ReprodHome page
M.-K. Shyu, M.-C. Lin, J.-C. Shih, C.-N. Lee, J. Huang, C.-H. Liao, I-F. Huang, H.-Y. Chen, M.-C. Huang, and F.-J. Hsieh
Mucin 15 is expressed in human placenta and suppresses invasion of trophoblast-like cells in vitro
Hum. Reprod., October 1, 2007; 22(10): 2723 - 2732.
[Abstract] [Full Text] [PDF]

Home page
 Cancer Res.Home page
P. K. Singh, Y. Wen, B. J. Swanson, K. Shanmugam, A. Kazlauskas, R. L. Cerny, S. J. Gendler, and M. A. Hollingsworth
Platelet-Derived Growth Factor Receptor {beta}-Mediated Phosphorylation of MUC1 Enhances Invasiveness in Pancreatic Adenocarcinoma Cells
Cancer Res., June 1, 2007; 67(11): 5201 - 5210.
[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 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 Thompson, E. J.
Right arrow Articles by Gendler, S. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Thompson, E. J.
Right arrow Articles by Gendler, S. J.

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