| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Advances in Brief |
Innere Klinik und Poliklinik (Tumorforschung), Universitätsklinikum Essen, Westdeutsches Tumorzentrum Essen, 45122 Essen, Germany [N. S. W., G. M., J. K. S., W. B., S. S., J. S., B. O.], and Department of Microbiology and Immunology, Kimmel Cancer Institute, Jefferson Medical College, Philadelphia, Pennsylvania 19107 [Z. S., L. Y. Y. F., K. H.]
 |
ABSTRACT |
|---|
 Top ABSTRACT IntroductionMaterials and MethodsResultsDiscussionREFERENCES |
|---|
|
Introduction |
|---|
|
TopABSTRACTIntroductionMaterials and MethodsResultsDiscussionREFERENCES |
|---|
The FHIT gene spans the FRA3B fragile site and
the t(3;8)(p14.2;q24) breakpoint described in a family with hereditary
RCC (4)
. Hemizygous, interstitial, or terminal 3p
deletions involving the FHIT gene as well as down-regulated
Fhit expression have been found in 
70–85% of ccRCCs, whereas
homozygous deletions or aberrant FHIT transcripts were
rarely observed (15, 16, 17, 18, 19)
. In addition to a pro-apoptotic
activity of Fhit described recently in lung cancer cell lines in
vitro (20
, 21)
, functional evidence for
tumor-suppressive activity of the FHIT gene in
vivo has been established previously in three lung and two gastric
cancer cell lines (20
, 22)
. In cervical carcinoma cell
lines, on the other hand, restoration of FHIT expression
showed no effects on tumor growth in nude mouse experiments (23
, 24)
. In RCC, suppression of tumorigenicity by Fhit in
vivo has to date been analyzed in a single RCC cell line, RC48,
which shows a homozygous deletion involving FHIT exons 8–10
(22)
.
To further investigate the tumor-suppressive activity, and thus, the potential role of FHIT in RCC pathogenesis, we transfected FHIT cDNA expression constructs into two additional RCC cell lines, RCC-1 and SN12C (10 , 25) , which, similar to many primary ccRCCs, show low levels of endogenous Fhit expression (18) . We here report on the growth characteristics of these RCC cell lines in vitro and their tumorigenicity in vivo.
|
Materials and Methods |
|---|
|
TopABSTRACTIntroductionMaterials and MethodsResultsDiscussionREFERENCES |
|---|
Expression Constructs and Transfections.
Plasmid expression constructs containing wild-type FHIT cDNA
(pRcFHIT) and a FHIT cDNA mutated at the His96 codon
(pRcFHITH96N), significantly reducing the hydrolase activity of Fhit,
have been described (14
, 22
, 28)
. RCC-1 and SN12C cells
were stably transfected with 2 µg of pRcFHIT, pRcFHITH96N, or control
plasmid (pRcCMV), using lipofectin (Life Technologies) following the
manufacturer’s guidelines. After transfections, individual
G418-resistant colonies were selected in the presence of 500 µg/ml
active G418 (Geneticin; Life Technologies).
PCR Analyses.
RT-PCR analyses were performed to assess the endogenous FHIT
gene status within the RCC-1 and SN12C parental cell lines, using a
nested set of primers (A, 5U2 plus 3D2; B, 5U1 plus 3D1) as described
(14)
. In addition, to analyze the coding region, RT-PCR
was performed using oligonucleotide primers
5'-TGAGGACATGTCGTTCAGATTT-3' and 5'-CACTGAAAGTAGACCCGC-3', which flank
exons 5 and 9 (14)
. The presence of the exogenous
FHIT gene in transfected RCC-1 and SN12C clones was verified
by PCR analysis using the primer pair HITF and HITR as described
previously (28)
.
In Vitro Proliferation Parameters.
To examine the proliferation rates and saturation densities of cell
lines, cells were plated in 24-well cell culture plates (2
cm2/well) at a density of 4–5 x 104
cells per well. Cells from triplicate
cultures were counted once every 24 h for 8–9 days, and the
doubling time was calculated from the logarithmic part of the growth
curve. The saturation density was defined as the number of cells per
cm2 at the time of reaching confluence.
Cell Cycle Analysis.
Cells were trypsinized, fixed, washed, and incubated in DNA-staining
solution containing 50 µg/ml propidium iodide (Sigma), 5 Kunitz
units/ml bovine pancreas RNase type A (Boehringer-Mannheim, Mannheim,
Germany), and 10 µl of PCNA-FITC monoclonal mouse antibody (DAKO,
Glostrup, Denmark), and were analyzed by flow cytometry according to
the manufacturer’s recommendations (DAKO). Cell cycle analysis was
performed using a Coulter flow cytometer equipped with an argon laser
(488 nm; Coulter Electronics, Miami, FL), and data were registered and
stored in list mode. Debris was excluded by gating on a forward and
side scatter dot plot or on a DNA histogram. Fluorescence was recorded
in channel FL1 (525 nm) and channel FL2 (575 nm), using linear and
logarithmic amplification. Data were evaluated with the Multicycle
software for DNA analysis (Phoenix Flow Systems, San Diego, CA) and
with System II software (Coulter Electronics) for
sub-G0-G1 analysis.
Tumorigenicity Assay.
For the determination of tumorigenicity, 1 x 107 cells of parental RCC-1 or SN12C cells and
their derivatives were injected s.c. into the right flank of 4- to
6-week-old female NMRI-nu/nu mice (Central
animal facility, University of Essen Medical School) with groups of at
least three animals per cell line tested. Control mice were inoculated
with the parental RCC-1 and SN12C cells and with their derivatives
containing the empty vector. Animals were monitored at least weekly for
tumor formation. Mice were sacrificed when the largest tumor diameter
was 1 cm, and tumors were removed and weighed. One part of these
tumors was established in culture for further analysis, and another
part was fixed in 10% formalin for immunohistochemistry. Animal care
was provided in accordance with institutional guidelines.
Western Blot Analysis.
Preparation of cell lysates and Western blot analysis have been
described previously (22
, 28)
. Rabbit polyclonal anti-Fhit
antiserum (22)
was used at a dilution of 1:5000 for
immunoblot analysis and 1:1000 for immunohistochemical analysis.
Immunofluorescent Staining of Fhit Protein in Wild-Type and
Transfectant Cell Lines.
Cytospins of 5 x 105 RCC-1 and
SN12C parental cells and derivative FHIT transfectants (passage 5) were
fixed in 4% paraformaldehyde-PBS and permeabilized in 0.5% Triton
X-100-PBS. After washing with PBS, samples were incubated with the
rabbit polyclonal anti-Fhit antibody diluted 1:200 in PBS containing
1.5% goat serum for 1 h at 37°C. Slides were washed three times
with PBS and subsequently incubated with rhodamine-conjugated
Affinipure goat antirabbit IgG diluted 1:150 in PBS containing 1.5%
goat serum (Jackson ImmunoResearch Labs Inc., West Grove, PA) for 45
min at room temperature. Stained cells were visualized under a
fluorescence microscope (Axioplan; Zeiss, Göttingen,
Germany).
Immunohistochemical Analysis of Tumors.
Formalin-fixed tumor samples were paraffin-embedded, cut into 4-µm
sections, and mounted on Superfrost Plus glass slides (Menzel,
Braunschweig, Germany). After deparaffinization of all sections
followed by rehydratation through an ethanol series, the slides were
immersed in 0.01 M citrate buffer (pH 6.0), heated in a
microwave oven twice for 5 min for antigen recovery, and cooled for 20
min. The sections were then incubated at 37°C overnight with the
rabbit anti-Fhit primary antiserum (diluted 1:1000). Thereafter, the
sections were reacted with biotinylated goat antirabbit IgG
secondary antibody (Santa Cruz Biotechnology, Inc., Santa Cruz, TX) and
exposed to an avidin-biotin complex (Santa Cruz Biotechnology) for 45
min. Slides were washed with PBS, subsequently subjected to a 12-min
application of diaminobenzidine-peroxide solution (Santa Cruz
Biotechnology), washed again, counterstained with hematoxylin, and
coverslipped prior to evaluation.
Statistical Analyses.
Comparison of tumor-free survival of mice was performed using the
Kaplan-Meier product-limit method and the log-rank test (SPSS software,
version 7.0, SPSS Inc.). The 
2 test was
used for analyzing the rates of FHIT gene transfer into
transfected SN12C and RCC-1 cells. Cell cycle parameters were compared
using the two-sided Spearman’s test. The Wilcoxon Mann-Whitney
U test was used for comparison of tumor weights and the time
intervals between first visibility of tumors and the time of
sacrificing the animals when tumors showed diameters of 1 cm.
Two-sided P values <0.05 were considered statistically
significant.
|
Results |
|---|
|
TopABSTRACTIntroductionMaterials and MethodsResultsDiscussionREFERENCES |
|---|
600 and 450 bp) and in the RCC-1 cells (500 bp plus a
wild-type 707-bp product; Ref. 14
) when a nested set of
oligonucleotide primers was used. The wild-type-sized product (450 bp;
data not shown) was obtained using oligonucleotide primers encompassing
the FHIT coding exons 5–9, suggesting an intact
coding region of the FHIT gene in both cell lines.
Expression of the endogenous Fhit protein in the parental cell lines at
one to two passages prior to injection was analyzed by
immunofluorescence and Western blotting and showed a low degree of
endogenous Fhit expression (Figs. 1
2
I). Immunohistochemistry of tumors derived from wild-type
RCC-1 and SN12C cells (Fig. 2
II) revealed barely detectable
levels of endogenous Fhit protein.
|
|
, Fig. 2I
, and data not shown). Overall, the expression levels of
exogenous wild-type Fhit and mutant FhitH96N proteins as assessed by
Western blotting appeared to be higher in the SN12C cell lines than in
the RCC-1 cell lines (Table 1
and Fig. 1
).
|
. All cell lines transfected with the FHIT expression
constructs showed saturation densities and population doubling times
similar to those of the RCC-1 and SN12C parental cells, and did not
reveal gross alterations of cellular morphology. Cell cycle analysis in RCC-1 cell lines showed no significant differences between control cell lines consisting of either wild-type cells or cells transfected with the empty plasmid vector, and those cell lines selected for expression of exogenous Fhit. The distributions of the G0-G1, S, G2-M, and sub-G0-G1 phases (mean ± SD) were 44.9 ± 5.3%, 38.2 ± 5.7%, 17.0 ± 0.7%, and 5.0 ± 4.0% for the RCC-1 control cell lines (n = 3), and 41.4 ± 12%, 40.0 ± 12.1%, 18.6 ± 10.5, and 5.7 ± 5.4% for the FHIT-transfected RCC-1 cell lines (n = 14), respectively (P > 0.667). Likewise, no significant differences for these parameters were observed in the SN12C cells (data not shown).
Tumorigenicity of RCC-1 and SN12C Cells and Derivative Cell Lines
Expressing Exogenous Fhit.
In nude mouse assays with control cell lines, 14 of 15 mice given
injections of the parental RCC-1 cell line developed tumors within a
median of 39 days (range, 25–45 days), and mice
(n = 3) injected with RCC-1 cells harboring
the empty plasmid vector developed tumors within 33 days (Table 1)
. To
assess the tumorigenicity of RCC-1 cell lines with constitutive
expression of exogenous FHIT, five cell lines expressing the
wild-type FHIT transgene and two cell lines expressing the
mutant FHITH96N gene, respectively, were injected into nude
mice. Overall, in this series of experiments tumors developed in 22 of
26 mice, with a median latency period of 77 days (range, 41–178 days),
representing a significant delay in comparison with the controls
(median, 36 days; P < 0.00001). Four mice
remained tumor free for 171, 250, 349, and 400 days, respectively
(Table 1)
. Thus, tumorigenicity was substantially suppressed by
transfection of FHIT cDNA expression constructs.
In similar experiments using SN12C cells, control animals injected with
parental SN12C cells (n = 6) and derivative
cell lines transfected with the empty plasmid vector
(n = 3) revealed tumor growth within a median
of 18 days (range, 16–24 days; Table 1
). In contrast to the RCC-1
cells, however, the SN12C cells, which strongly expressed exogenous
Fhit (Fig. 1
) after transfection with either the wild-type or mutated
FHIT cDNA construct, showed no suppression of
tumorigenicity in 17 of 18 nude mice (94%) with latency periods
(median, 18 days; range, 8–24 days; Table 1
) for tumor development
similar to the controls (P > 0.05).
To determine the proliferation kinetics of developing tumors, time intervals between the first visibility of tumor growth and the time of reaching an approximate maximum tumor diameter of 1 cm (time of sacrificing) was compared between control mice injected with either parental cell lines or cell lines harboring the empty vector, and cell lines expressing exogenous wild-type or mutant Fhit proteins. For RCC-1, the time intervals (mean ± SD) were 17.1 ± 7.9 days for the controls (n = 17) and 12.8 ± 8.3 days for Fhit-expressing cell lines (n = 22; P = 0.096). For SN12C, the respective time intervals were 31.0 ± 13.8 days for the controls (n = 7) and 39.6 ± 13.8 days for Fhit expressing cell lines (n = 17; P = 0.337). Tumor weights (mean ± SD) in the RCC-1 controls (n = 14) and the Fhit-expressing tumors (n = 22) were 0.429 ± 0.188 g and 0.413 ± 0.314 g, respectively (P = 0.330). In the SN12C controls (n = 4) and the Fhit-expressing tumors (n = 17), tumor weights were 0.610 ± 0.151 g and 0.475 ± 0.248 g, respectively (P = 0.179). Thus, no significant differences were observed for these parameters between tumors derived from FHIT transfectants and tumors derived from control cell lines.
Analysis of Exogenous Fhit Expression in Nude Mouse Tumors.
Immunohistochemical analysis of explanted tumors from transfected SN12C
cells showed a high level of Fhit expression with strong signals in
>70% of the tumor cells (Table 1
and Fig. 2
II,
d). A different pattern of exogenous Fhit expression was
observed in the RCC-1-derived tumors. Immunohistochemistry revealed
that in 13 of 15 RCC-1-derived tumors (87%), Fhit was detected in only
1–20% of the tumor cells (Table 1)
, suggesting an outgrowth of
RCC-1-derived tumors from selected cells having down-regulated or lost
the exogenous FHIT gene (Fig. 2
II, b).
Results obtained with clone MF-D5 (Table 1)
are concordant with
previous data (22)
showing no difference in the
tumor-suppressing potential of wild-type FHIT and the mutant
FHITH96N with significantly reduced hydrolase activity.
|
Discussion |
|---|
|
TopABSTRACTIntroductionMaterials and MethodsResultsDiscussionREFERENCES |
|---|
Similar to these findings, the data reported here for the RCC cell line RCC-1, which shows a low level of intrinsic Fhit expression, provide further evidence for potential TS activity of FHIT in vivo. This activity is demonstrated by a significant delay of tumor formation in most animals injected with FHIT-transfected RCC-1 derivatives, including a small fraction of mice demonstrating a long-lasting tumor suppression. Moreover, tumors that did grow after the lag had lost Fhit expression in the vast majority of cells. Hence, taking into account the previous results obtained with cell line RC48 (22) , these data clearly establish a TS activity of FHIT in RCC cells in vivo. The restrictions to this activity, however, are presently unclear.
Immunohistochemical studies on primary ccRCC samples have suggested that Fhit inactivation may be an early event in tumors with G1 morphological grade and early clinical stage, and may be associated with tumor progression in G2/3 tumors (18 , 19) . Thus, it may be speculated that the selective TS activity of constitutive FHIT expression in the RCC-1 but not in the SN12C cells may reflect such a differential role of FHIT in the pathogenetic cascade of RCC tumorigenesis. The potential involvement of other chromosome 3p-specific TS loci within these pathways remains elusive. Both cell lines, RCC-1 and SN12C, have been shown to be susceptible to tumor suppression by the chromosome 3p loci NRC-2 and NRC-1, respectively (7, 8, 9, 10) . In the case of RCC-1, its sensitivity to phenotypic reversion by NRC-2 and FHIT, as shown here, may suggest independent TS activities of NRC-2 and FHIT. SN12C cells, on the other hand, can be reverted by reintroduction of chromosomal material containing the NRC-1 locus. Hence, if NRC-1 and Fhit were involved in a common regulatory TS pathway, the activity of Fhit would be upstream of NRC-1. A role of the VHL gene product in these systems is rather unlikely because both the RCC-1 and the SN12C cell lines contain wild-type alleles of this gene (9) . Thus, FHIT may act in RCC tumorigenesis by VHL-independent mechanisms, which have been postulated for 30–50% of sporadic ccRCCs (29) .
The functions involved in the TS activity of FHIT in vivo are not yet fully explained. Ji et al. (20) recently described an in vitro growth inhibition and/or induction of apoptosis by constitutive FHIT expression in one head and neck and in three lung cancer cell lines (including H460) lacking endogenous Fhit expression. Interestingly, inhibition of cell proliferation in these cell lines correlated with the level of exogenous Fhit expression. Likewise, Sard et al. (21) reported a significant rate of G0-G1 arrest, induction of apoptosis, and increased expression of p21waf1 in H460 lung cancer cells after transfection with FHIT. In contrast to these findings, various cancer cell lines lacking endogenous Fhit expression have been described in which no effects of FHIT on cell proliferation, cell cycle parameters, and/or induction of apoptosis in vitro have been observed irrespective of high levels of exogenous FHIT expression (22, 23, 24) . These include cell lines exhibiting reduced tumorigenicity after FHIT transfer, such as RC48 (22) , a head and neck cancer cell line (22) , and two gastric carcinoma cell lines (22) , or different cervical carcinoma cell lines that did not reveal a reduction of tumorigenicity after constitutive high-level expression of the FHIT gene (23 , 24) . Similar analyses with cells exhibiting low-level expression of endogenous Fhit, as is observed in many primary tumor samples, have been performed to date on a head and neck cancer cell line, 22B, which showed no changes of cell proliferation parameters in vitro after high-level expression of exogenous Fhit (20) . Tumorigenicity data for this cell line, however, were not reported.
In the present study, we have extended these investigations to the RCC cell lines RCC-1 and SN12C and found no detectable changes of cell cycle parameters in vitro after FHIT transfer, even in the RCC-1 cell lines with the highest level of transgene expression obtained. Nonetheless, we observed a significantly lower rate of successful FHIT gene transfer into the RCC-1 cells compared with the SN12C cells, a lower level of transgene expression, and loss of Fhit expression in most cells of tumors derived from the FHIT-transfected RCC-1 cells, suggesting a selection against FHIT expression and, hence, supporting a role of FHIT in cell proliferation or in the induction of cell death. It may be speculated that the levels of Fhit tolerated by our RCC-1 selectants were insufficient to detect changes in the in vitro growth characteristics that have been shown to be Fhit dosage dependent (20) .
In conclusion, the present study suggests that FHIT may have a role in RCC pathogenesis in vivo. In addition, the susceptibility to the potential TS activity of FHIT appears to be cell line or cell type specific. High-level expression of exogenous FHIT does not necessarily result in tumor suppression of cancer cells that either lack endogenous (e.g., cervical cancer cell lines; Refs. 23 , 24 ) or express residual low levels of endogenous Fhit (SN12C cells; this study). Furthermore, the TS activity of FHIT is not restricted to cells with a complete loss of FHIT expression by biallelic inactivation (e.g., RC48 cells; Ref. 22 ). It may also apply to cells with low-level expression of endogenous Fhit in which even a moderate elevation of Fhit expression by gene transfer may be sufficient for the induction of tumor suppression in vivo (RCC-1 cells; this study). Finally, the resistance to Fhit-induced tumor suppression by even high levels of Fhit suggest pathogenetic pathways which, if not independent of FHIT, reflect the frequent defects/losses of elements downstream of Fhit in human cancers. Identification of these pathways appears to be crucial for selecting those types of cancer that might be susceptible to FHIT gene replacement strategies.
|
Acknowledgments |
|---|
|
FOOTNOTES |
|---|
1 This work was supported by grants from
‘Deutsche Forschungsgemeinschaft,’ Bonn, Germany, and Stiftung VerUm,
Munich, Germany. 
2 To whom requests for reprints should be
addressed, at Innere Klinik und Poliklinik (Tumorforschung),
Universitätsklinikum Essen, Westdeutsches Tumorzentrum Essen,
Hufelandstrasse 55, D-45122 Essen, Germany. Phone: 49-201-723-2020
(2024); Fax: 49-201-723-2020 (5925).
3 The abbreviations used are: RCC, renal cell
carcinoma; VHL, von-Hippel-Lindau; NRC-1 and NRC-2, nonpapillary renal
carcinoma-1 and -2; TSG, tumor suppressor gene; FHIT, fragile histidine
triad; ccRCC, clear cell RCC; RT-PCR, reverse transcription-PCR; TS,
tumor suppressor.
4 H. Brauch, IKP, Stuttgart, Germany, unpublished
data.
Received 12/ 9/99. Accepted 4/17/00.
|
REFERENCES |
|---|
|
TopABSTRACTIntroductionMaterials and MethodsResultsDiscussionREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  F. Pichiorri, F. Trapasso, T. Palumbo, R. I. Aqeilan, A. Drusco, B. W. Blaser, D. Iliopoulos, M. A. Caligiuri, K. Huebner, and C. M. Croce Preclinical Assessment of FHIT Gene Replacement Therapy in Human Leukemia Using a Chimeric Adenovirus, Ad5/F35. Clin. Cancer Res., June 1, 2006; 12(11): 3494 - 3501. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Korsisaari, D. J. Rossi, K. Luukko, K. Huebner, M. Henkemeyer, and T. P. Makela The Histidine Triad Protein Hint Is Not Required for Murine Development or Cdk7 Function Mol. Cell. Biol., June 1, 2003; 23(11): 3929 - 3935. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Trapasso, A. Krakowiak, R. Cesari, J. Arkles, S. Yendamuri, H. Ishii, A. Vecchione, T. Kuroki, P. Bieganowski, H. C. Pace, et al. Designed FHIT alleles establish that Fhit-induced apoptosis in cancer cells is limited by substrate binding PNAS, February 18, 2003; 100(4): 1592 - 1597. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Roz, M. Gramegna, H. Ishii, C. M. Croce, and G. Sozzi Restoration of fragile histidine triad (FHIT) expression induces apoptosis and suppresses tumorigenicity in lung and cervical cancer cell lines PNAS, March 19, 2002; 99(6): 3615 - 3620. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Ishii, K. R. Dumon, A. Vecchione, L. Y. Y. Fong, R. Baffa, K. Huebner, and C. M. Croce Potential Cancer Therapy With the Fragile Histidine Triad Gene: Review of the Preclinical Studies JAMA, November 21, 2001; 286(19): 2441 - 2449. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Velickovic, B. Delahunt, S. Storkel, and S. K. G. Grebem VHL and FHIT Locus Loss of Heterozygosity Is Common in All Renal Cancer Morphotypes But Differs in Pattern and Prognostic Significance Cancer Res., June 1, 2001; 61(12): 4815 - 4819. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. R. Dumon, H. Ishii, A. Vecchione, F. Trapasso, G. Baldassarre, F. Chakrani, T. Druck, E. F. Rosato, N. N. Williams, R. Baffa, et al. Fragile Histidine Triad Expression Delays Tumor Development and Induces Apoptosis in Human Pancreatic Cancer Cancer Res., June 1, 2001; 61(12): 4827 - 4836. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Ishii, K. R. Dumon, A. Vecchione, F. Trapasso, K. Mimori, H. Alder, M. Mori, G. Sozzi, R. Baffa, K. Huebner, et al. Effect of Adenoviral Transduction of the Fragile Histidine Triad Gene into Esophageal Cancer Cells Cancer Res., February 1, 2001; 61(4): 1578 - 1584. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
