SEARCH

SEARCH BY CITATION

Keywords:

  • H3 phosphorylation;
  • Ras-MAPK pathway;
  • spectral karyotyping;
  • nonclonal chromosome aberrations;
  • pancreatic cancer

Abstract

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Activating mutations in K-Ras occur in most pancreatic cancers. We investigated whether genetic changes (K-Ras mutations) in human pancreatic cancer cell lines altered genomic instability and epigenetic events responding to Ras-mitogen activated protein kinase (MAPK) signaling by characterizing 3 human pancreatic cancer cells lines with and without activating mutations in K-Ras. Activation of the Ras-MAPK pathway results in the stimulation of the histone H3 kinase, mitogen and stress activated kinase (MSK) 1, and increased phosphorylation of histone H3 at serine 10 (H3 S10ph). MSK1 and H3 S10ph have roles in neoplastic transformation. We demonstrate that the presence of a K-Ras mutation did not correlate with elevated chromosomal aberrations or increased genomic instability. Although the levels of the epidermal growth factor receptors and MSK were similar, the Ras-MAPK pathway was differentially induced by phorbol esters (12-O tetradecanoylphorbol-13-acetate) or epidermal growth factor, with the response of this signaling pathway being cell-type specific. This response corresponded downstream at the level of chromatin where stimuli-induced elevation of H3 S10ph typically paralleled the increase in phospho-extracellular signal regulated kinase 1/2. Our results present evidence that nonclonal chromosomal aberrations and epigenetic programming responding to stimulation of the Ras-MAPK pathway may be better markers for cancer progression than the upstream mutated oncogenes. © 2008 Wiley-Liss, Inc.

Ras is one of the most commonly mutated oncogenes found in cancer. Activating mutations in K-Ras are most often in codon 12 but also in codons 13 or 61. Expression of the mutated activated K-Ras affects a myriad of downstream targets and lead to the deregulation of cellular signaling pathways as well as transformation, tumorigenesis and metastasis.1, 2 In pancreatic cancer, about 95% of all diagnosed adenocarcinomas have an acquired mutation in the K-Ras gene which leads to a constitutively activated oncoprotein.3, 4 Further, the combined activation of proto-oncogene K-Ras with inactivating mutations of tumor suppressor genes such as p53 and p16 have been linked to progression of pancreatic cancer from preinvasive lesions to metastatic disease.5 Thus, acquired somatic mutations of the K-Ras gene are associated with the early stages of pancreatic oncogenesis. Despite increased understanding of the biology and molecular pathophysiology of pancreatic cancer, the rate of survival among diagnosed patients remain poor and patients continue to be diagnosed at a late incurable state which poses a great urgency in improving strategies for early disease detection and treatment.

Ras elicits a wide array of physiological responses by triggering a protein kinase cascade known as the Ras-mitogen activated protein kinase (MAPK) signaling pathway. Stimulation of this pathway by extracellular stimuli such as growth factors, hormones and pharmacological agents such as 12-O tetradecanoylphorbol-13-acetate (TPA) activates a series of kinases that invoke epigenetic mechanisms that result in the modification of chromatin proteins concomitant with the expression of genes coding for immediate-early transcription factors and signal transduction proteins.6–11 TPA stimulation of mouse fibroblasts results in the mitogen and stress activated kinase (MSK) 1/2-mediated phosphorylation of histone H3 at serine 10 or 28, events that have been named the “nucleosomal response.”9, 12–14

Although it is well documented that K-Ras potentially plays a critical role in initiating pancreatic cancer tumorigenesis and that certain effectors such as B-Raf and HER2 which converge into the Ras-MAPK pathway are deregulated in these cells, there are limited studies that address the downstream events in cells possessing an activated K-Ras protein.5, 15 Our previous findings demonstrate that steady state levels of phospho-extracellular signal regulated kinase (ERK) 1/2 and H3 S10ph and H3 S28ph as well as MSK1 kinase activity are elevated in c-Ha-Ras-transformed mouse fibroblasts.8, 13, 16, 17 Thus, the deregulation of upstream effectors by genetic events (mutations) in these cells directly influences the downstream responses in the pathway, including reversible epigenetic events such as histone modifications. Understanding the deregulated downstream targets may facilitate identification of other players in the pathway which may be useful for prognosis, therapeutic targeting and treatment of the cancer. Whether these observations are universal and translate in other cell backgrounds particularly in pancreatic cancer cell lines containing activating K-Ras gene mutations have not been demonstrated.

Moreover, recent studies demonstrate that H3 S10ph and MSK1 are critical during neoplastic transformation18, 19 and that deregulation of H3 mitotic kinase is linked to genomic instability and oncogenesis.15, 20 Hence, there is evidence to warrant examination of downstream players of the pathway as potential targets for therapy. Nonetheless, H3 phosphorylation as a downstream event in the activation of the Ras-MAPK pathway in human pancreatic cancer cells has not been reported.

In this study, we examined the karyotypic characteristics of 3 human pancreatic cancer cells lines and observed that the presence of a K-Ras mutation did not correlate with elevated chromosomal aberrations or increased genomic instability. Further, we investigated whether human pancreatic cancer cell lines containing inherent K-Ras mutations have an altered signaling pathway in response to different stimuli when compared to cells with a wild-type K-Ras by evaluating the levels of EGFRs, phospho-ERK1/2, MSK1 and H3 S10ph. We show that stimuli such as epidermal growth factor (EGF) and TPA differentially induce the Ras-MAPK pathway in pancreatic cancer cells, with the response being cell-type specific. This response corresponds downstream at the level of chromatin where stimuli-induced elevation of phospho-ERK1/2 typically parallels an increase in H3 S10ph levels.

Material and methods

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Reagents

TPA, EGF and anti-β-actin mouse monoclonal antibodies were purchased from Sigma Chemical Co (St. Louis, MO). Anti-phospho-p44/p42 MAPK, anti-p44/p42 rabbit polyclonal and anti-histone H3 mouse monoclonal antibodies were purchased from Cell Signaling Technologies (Beverley, MA). Anti-H3 S10ph (sc-8656), anti-EGFR (sc-03) and anti-HER2/neu receptor (sc-284) rabbit polyclonal antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Anti-MSK1 sheep polyclonal antibodies were purchased from Upstate Biotechnology (Lake Placid, NY). Anti-K-Ras (ab55391) mouse monoclonal antibodies were purchased from Abcam (Cambridge, MA).

Cell lines, cell culture and treatments

The human pancreatic cancer cell lines BxPC-3, Hs766T and Panc-1 were obtained from the American Type Culture Collection (Rockville, MD) and cultured as recommended by the supplier. Both Hs766T and Panc-1 cell lines were maintained in Dulbecco's Modified Eagle Medium whereas the BxPC-3 cell line was maintained in RPMI 1640 supplemented with 10 mM HEPES and 1.0mM sodium pyruvate. All 3 cell lines were grown in their respective media supplemented with 10% (v/v) fetal bovine serum, streptomycin/penicillin antibiotics at 37°C and 5% CO2. For stimulation studies, cells were cultured until 70% confluency was reached and serum starved (SS) in 0.1% (v/v) fetal bovine serum for 48 hr to drive the majority of the cell population into G0/G1. Cells were either untreated (SS), treated with 100 nM TPA or with 50 ng/ml EGF for the indicated lengths of time in the figures. For spectral karyotyping (SKY), cells were split 24 hr before chromosome preparation as previously described.21 Cell cycle distribution was monitored by flow cytometry. The distribution of the SS cells in G0/G1,S,G2/M phases of the cell cycle were typically as follows: BxPC-3—68,12,20; Hs766T—72,13,15; Panc-1—85,5,10.

Preparation of total cell extracts and histones

Cell extracts were isolated as described previously.10 Acid extraction of histones was done as described previously.8 Protein concentrations were determined using the Bio-Rad Protein Assay as per manufacturer's instructions (Hercules, CA).

Electrophoresis and immunoblotting

Proteins were resolved by SDS (10% and 15%)-PAGE and visualized either by Coomassie Blue staining or by transfer to nitrocellulose membrane and immunochemical staining with various antibodies as per manufacturers' instructions. Enhanced chemiluminescence kits were purchased from Perkin Elmer or from Amersham Biosciences (Piscataway, NJ) for quantitative analysis using the Storm® phosphorimager. To determine the fold induction in response to stimuli, relative protein band intensities were normalized to intensities for the protein loading control such as ERK, β-actin or total H3 and each stimulated time point was normalized against the unstimulated sample.

SKY and analysis

Chromosome spreads for each pancreatic cell line were prepared using a fixation protocol from suspension as previously described.21 Briefly, adherent cells grown in log phase were trypsinized, collected and equilibrated in hypotonic solution containing KCl. To carry out fixation, cells were suspended in 3 changes of 3:1 methanol-acetic acid before dropping cell suspension onto slide.

SKY was carried out on metaphase spreads using the (Applied Spectral Imaging, Vista, CA) kit for human according to protocols recommended by the supplier including RNase A and pepsin treatment, denaturation, hybridization and detection. Image acquisition and analyses were performed using the Spectra Cube™ on a Carl Zeiss Axioplan 2 microscope with a 63× oil objective and the SKYVIEW 2.0 software.22, 23 Twenty metaphase spreads were examined per cell line in 3 independent SKY experiments.

Results

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Nonclonal chromosome aberrations reflect the genomic instability of pancreatic cancer cells

Although greater than 95% of pancreatic cancers exhibit genetic mutations in K-Ras, the development of pancreatic preinvasive lesions to aggressive carcinomas requires multiple hits in various genes for disease progression.5, 24 As a consequence, genome aberrations rather than gene mutations provide an assessment of the cancer chromosome dynamics in pancreatic cancer.24–28 We evaluated the chromosomal organization of the 3 pancreatic cancer cells using SKY. The technique allows for comparative analysis of numerical and structural chromosomal aberrations29 of individual cells in a population and enables us to assess if there is a correlative relationship between an activated Ras-MAPK pathway signaling and the chromosomal organization of pancreatic cancer cells. The pancreatic cell lines used in this study were BxPC-3, which has a wild-type Ras, and cell lines with activating mutations in K-Ras (Hs766T, codon 61 mutation; Panc-1, codon 12 mutation). Aneuploidy and structural aberrations (clonal chromosomal aberrations, CCA) specific to and characteristic of each cell line were evident (Fig. 1a–1c and Table I). For example, rearrangements such as t(X;3) (p = 0.0014) and t(3;6;16;6) (p = 1 × 10−6) characteristic for BxPC-3 were manifested in the respective karyotype and were statistically significant. These pancreatic cancer cell lines displayed specific translocated chromosomes, illustrating that all cell lines were drastically different from each other judging by their genome context (Table I). Further, the observation that all pancreatic cancer cell lines displayed different karyotypes suggested that the altered karyotypes were stochastically achieved. Regardless, we found no striking chromosomal aberrations in the 3 pancreatic cancer cell lines correlated with an activated K-Ras. Emerging studies reveal that unique aberrations known as nonclonal chromosomal aberrations (NCCAs) more accurately exhibit the heterogeneity and instability observed in cancer evolution.25–27, 30 Some cell lines are more stable than others as illustrated by the different NCCA frequencies detected (Table I). With an NCCA frequency of 60%, Panc-1 cancer cells presented with the highest NCCA frequency (Table I) and thus more genomically unstable when compared to BxPC-3 (20%) and Hs766T (15%).

thumbnail image

Figure 1. Structural and numerical chromosome aberrations exhibited by pancreatic cancer cell lines. Representative spectral karyotyping (SKY) images of (a) BxPC-3, (b) Hs766T and (c) Panc-1 illustrating their characteristic as well as unique chromosomal aberrations. Twenty mitotic figures were analyzed for each cell line to determine frequency of nonclonal chromosomal aberrations that correlated to relative genomic instability of the cell line (d). The majority of structural NCCAs represent random chromosomal rearrangements.

Download figure to PowerPoint

Table I. Nonclonal Chromosome Aberrations Reflect the Genomic Instability of Pancreatic Cell Lines
Cell lineMajor aberrationsFrequencies
BxPC-3t(X;3)35%
t(3;6;16;6)35%
t(X;11)10%
t(3;21)10%
t(10;11)10%
NCCAs20%
Hs766TT(8;5)40%
T(13;11)15%
NCCAs15%
Panc-1t(3;10)30%
t(13;17;18)20%
t(13;17)20%
t(7;19)15%
t(15;19)10%
NCCAs60%

TPA and EGF differentially stimulate phospho-ERK1/2 levels in pancreatic cancer cells

The intensity and duration of the phospho-ERK1/2 induction response varies from transient to sustained depending on the cell context examined and stimuli used.31 To ascertain whether the presence of activating K-Ras mutations confer differences in downstream effector signaling, we compared the extent and strength of phospho-ERK1/2 induction in 3 pancreatic cancer cell lines in response to EGF and TPA. EGF and TPA increased phospho-ERK1/2 levels in all 3 cell lines to different extents, with EGF having lesser effects in Hs766T and Panc-1 cells that have mutated K-Ras (Fig. 2). BxPC-3 cells, which have a wild-type K-Ras, showed a TPA-induced phospho-ERK1/2 response that was somewhat lower than that with EGF (1.5 ± 0.5 and 2.3 ± 0.8 at 60 min, respectively; n = 3; average ± SEM) (Fig. 2a, 2d and 2e). However, Hs766T, which has a mutated K-Ras, displayed a more robust and sustained phospho-ERK1/2 response to TPA when compared to EGF treatment (3.4 ± 0.3 and 1.1 ± 0.1 at 60 min, respectively; n = 3) (Fig. 2b,2d and 2e). EGF-induced phospho-ERK1/2 levels in these cells exhibited a weak and transient increase (Fig. 2b and 2e). The TPA-induced increase in phospho-ERK1/2 levels in Panc-1 cells, which also have a mutated K-Ras, was typically sustained for longer times than that observed in EGF-stimulated cells (3.6 ± 0.4 and 2.1 ± 0.4 at 60 min, respectively, n = 3) (Fig. 2c2e).

thumbnail image

Figure 2. TPA and EGF stimulate a differential induction of phospho-ERK levels in pancreatic cancer cell lines. (a) BxPC-3, (b) Hs766T and (c) Panc-1 pancreatic cancer cells were cultured until 70% confluency and serum-starved for 48 hr. Cells were either untreated (SS) or treated for 15 to 90 min with 100 nM TPA or with 50 ng/ml EGF. Total cell extracts (25 μg) were then prepared and resolved on a SDS-10%-polyacrylamide gel, transferred to a nitrocellulose membrane, and stained immunochemically with antibodies directed against ERK and phospho-ERK as indicated (d and e). The fold induction of phospho-ERK in response to stimuli (TPA or EGF) was determined as described under “Material and methods” section.

Download figure to PowerPoint

Levels of EGFRs and MSK1 are unaltered upon induction of Ras-MAPK signaling in pancreatic cancer cells

Phorbol esters (TPA) stimulate the Ras-MAPK pathway through cytoplasmic or membrane-associated targets such as protein kinase C, diacylglycerol or Ras proteins whereas EGF elicits its action primarily by binding its respective receptors to activate the same pathway.1, 32 Although each stimulus induces the pathway through different upstream effectors, both stimuli converge downstream through activated phospho-ERKs. To determine whether the difference in phospho-ERK1/2 response to EGF in Hs766T and Panc-1 was a consequence of upstream receptors, we evaluated the levels of EGFR and HER2/neu in these cells following stimulation. Neither stimulus produced any variation in EGFR or HER2/neu protein levels (Fig. 3). The K-Ras levels were similar in the 3 cell lines (Fig. 4d). Thus, the differential response of the 3 pancreatic cancer cell lines to EGF stimulation was not due to EGFR, HER2 or K-Ras levels. As the bulk of studies that examine the impact of the Ras-MAPK pathway on cancer development and progression primarily focus on membrane receptors, upstream and cytoplasmic effectors, it is of interest to examine the influence of downstream targets particularly protein modifiers at the level of chromatin.4, 33–35 MSK1 is activated in response to various stimuli and occurs downstream of the Ras-MAPK pathway. Despite activation of phospho-ERK1/2 upon EGF and/or TPA treatment, evaluation of MSK1 protein revealed similar levels before and after induction in all cell lines (Fig. 4).

thumbnail image

Figure 3. Differential response to EGF stimulation is not due to EGFR and HER2 levels. (a) BxPC-3, (b) Hs766T and (c) Panc-1 pancreatic cancer cells were cultured, serum-starved and were either untreated (SS) or treated for 15 to 90 min with 100 nM TPA or with 50 ng/ml EGF. Total cell extracts (25 μg) were then prepared and resolved on a SDS-10%-polyacrylamide gel, transferred to a nitrocellulose membrane, and stained immunochemically with antibodies directed against EGFR, HER2 and β-actin as indicated.

Download figure to PowerPoint

thumbnail image

Figure 4. MSK1 protein level is unaltered upon induction of Ras-MAPK signaling in pancreatic cancer cell lines. (a) BxPC-3, (b) Hs766T and (c) Panc-1 pancreatic cancer cells were cultured, serum-starved and were either untreated (SS) or treated for 15 to 90 min with 100 nM TPA or with 50 ng/ml EGF. Total cell extracts (25 μg) were then prepared and resolved on a SDS-10%-polyacrylamide gel, transferred to a nitrocellulose membrane, and stained immunochemically with antibodies directed against MSK1 and β-actin as indicated. (d) Immunoblots of the resolved total cell extracts (25 μg) from serum-starved cells were stained immunochemically with antibodies directed against K-Ras and MSK1 as indicated.

Download figure to PowerPoint

Induction of H3 S10ph levels parallel phospho-ERK1/2 stimulation in BxPC-3 and Hs766T but differs in Panc-1 cells

Recent studies provide evidence that H3 S10ph is critical during neoplastic transformation.18 To evaluate whether the induction of upstream phospho-ERK1/2 upon stimulation of the pathway can convey a nucleosomal response, we examined the levels of H3 S10ph following stimulation of the Ras-MAPK pathway in the pancreatic cell lines. TPA-induced activation of H3 S10ph in BxPC-3 cells was typically lower than that with EGF (1.4 ± 0.2 and 2.6 ± 0.2 for 60 min, respectively; n = 3; average ± SEM) (Fig. 5a, 5d and 5e). TPA-induced stimulation of H3 S10ph was more robust in Hs766T than in the other cell lines (Fig. 5b and 5d). For both cell lines with K-Ras mutations, TPA-induced levels of H3 S10ph were greater than those in EGF-stimulated cells (for Hs766T cells: 4.1 ± 0.3 and 0.9 ± 0.2 for 60 min, respectively; for Panc-1 cells: 2.4 ± 0.7 and 1.3 ± 0.5 for 60 min, respectively; n = 3) (Fig. 5d and 5e). However, among the cell lines analyzed, we observed the induction of H3 phosphorylation in Panc-1 cells to be the most variable from one cell preparation to the next, for example at 90 min of TPA stimulation H3 S10ph levels were 2.0 ± 0.2, 4.1 ± 0.4 and 4.3 ± 2.5 (n = 3) for BxPC-3, Hs766T and Panc-1 cells, respectively.

thumbnail image

Figure 5. Induction of H3 S10ph levels parallel phospho-ERK1/2 stimulation in BxPC-3 and Hs766T but differs in Panc-1 cells. Serum-depleted (a) BxPC-3, (b) Hs766T and (c) Panc-1 pancreatic cancer cells were treated with 100 nM TPA or 50 ng/ml EGF for 15 to 90 min. Acid-soluble nuclear histones (5 μg) were resolved on a SDS-15%-polyacrylamide gel, transferred to a nitrocellulose membrane, and stained immunochemically with anti-H3 pS10 and anti-total H3 (d and e). The fold induction H3 S10ph in response to stimuli (TPA or EGF) was determined as described under “Material and methods” section.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

In this study, we demonstrate that the Ras-MAPK pathway can be activated by TPA or EGF in human pancreatic cancer cells and lead not only to elevated levels of phospho-ERK1/2 but also to increased downstream levels of H3 S10ph. However, the strength or duration of signaling upon induction did not correlate with the activating K-Ras mutation in these cells. Both Hs766T and Panc-1 cells, which contain activating mutations of the K-Ras gene, did not display a deregulated Ras-MAPK pathway signaling as measured by irregular augmentation in levels of phospho-ERK1/2, MSK1 or H3 S10ph in comparison to wild-type K-Ras-expressing BxPC-3. Thus, it appears that the constitutive activation of K-Ras does not necessarily convey anomalous downstream signaling events in pancreatic cancer cells as we have repeatedly observed in immortalized mouse fibroblasts.8, 13, 16, 17 It is important to note though that while both the cell lines used in this study have activating mutations in the K-Ras, these cells have different intrinsic genetic backgrounds and properties owing to their distinct source of tumor origin. Importantly, various karyotypes represent different genome context defined systems.28 The same gene mutation likely has distinctive functions in different systems. As such, other factors apart from K-Ras status largely contribute to the lack of correlation between Ras and signaling events downstream and will require further study to elucidate. For example, the activity of phosphatases that counteract kinases can significantly impact signaling pathways and cellular responses.36, 37 Although we did not detect any change in EGFRs or MSK1 levels, it is possible that there are changes in their activity. Consequently, it is prudent to refrain from drawing correlations and relationships when evaluating other cell backgrounds with Ras mutations such as breast cancer (10%) and colorectal cancer (50%).3, 4

Notwithstanding the K-Ras status of the pancreatic cancer cell lines, we observed similarities in the responses to induction between levels of phospho-ERK1/2 and downstream levels of H3 S10ph. In BxPC-3 cells, both EGF and TPA increased phospho-ERK1/2 levels which coincided with the enhanced levels of H3 S10ph particularly for EGF stimulation. A more robust and sustained response of phospho-ERK1/2 and H3 S10ph in Hs766T cells was detected for TPA. As such, the pattern of stimuli-induced increase in these proteins appears to be distinct depending on the stimuli and cell context suggesting that each stimulus can dictate diverse responses in different cells as it is in the case of PC12 cells where depending on the stimuli (EGF versus NGF) used to induce a transient or sustained level of phospho-ERK1/2, the cell response can switch from proliferation to differentiation.31 Regardless, the mirrored responses of phospho-ERK1/2 and H3 S10ph irrespective of stimuli indicate that the signaling pathway remains intact downstream despite the lack of correlation with upstream K-Ras. This study suggests that downstream players such phospho-ERK1/2 and H3 S10ph serve as better markers to gauge the extent and magnitude of cellular stimulation and insults.

In the case of the Ras-MAPK signaling in Panc-1 cells, both EGF and TPA increase phospho-ERK1/2 levels with a strong preference for TPA as a stimuli. However, a much random pattern of induction upon TPA or EGF treatment for H3 S10ph was observed downstream which corresponds to the elevated frequency of NCCA in these cells (60%). In contrast, the lowest NCCA frequency of the 3 cell lines analyzed belongs to Hs766T which also exhibits the most consistent and marked H3 S10ph induction in response to stimuli. Emerging studies reveal that unique aberrations like NCCA more accurately exhibit the heterogeneity and instability observed in cancer evolution.25–27, 30 We show that the NCCA frequency of the cells not only reflects the relative genomic instability of the cell line but also correlates with the predictability of induction pattern for H3 S10ph. Since SKY evaluates the genomic content and chromosome aberrations in a specific cell in the population, it is possible that the variations observed in the H3 S10ph response reflects the dynamic alterations at the gene loci responding to upstream Ras-MAPK signaling in the cell population.

At present, there is no curative treatment for resectable and metastatic pancreatic ductal adenocarcinoma. Current chemotherapy of gemcitabine combined with EGFR tyrosine kinase inhibitor, Tarceva, provides a 4-month survival advantage at best.15 Further, there is a lack of specific tumor markers for disease staging and early diagnosis. Hence, targeting the signaling pathway and the downstream players that can affect invasiveness and oncogenic aggressivity remains an exciting avenue to pursue therapeutic intervention.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Ms. Cheryl Peltier is gratefully acknowledged for her assistance in maintenance and propagation of various cell lines. This research was supported by a grant from Canada Research Chair to J.R.D. and CIHR Frederick Banting and Charles Best Canada Graduate Scholarship to P.S.E. Imaging was performed at the Genomic Centre for Cancer Research and Diagnosis.

References

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References