| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Carcinogenesis |
Institut Gustave Roussy [A. Sp., S. B.], Department of Pathology [A. Sp.] and U. 521, INSERM [S. B.] 94800, Villejuif, France, and UPR 2169, Centre National de la Recherche Scientifique, Cancer Research Institute, B. P. 8-94801, Villejuif, France [G. G-M., A. Sa]
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONREFERENCES |
|---|
Two forms of XP have been described, depending on their clinical symptoms and biochemical defects, as "classical XP" and "XP variant" (XPV). In the classical form, seven complementation groups exist, designated A to G. The most common complementation group is XP-C found in Europe, North Africa, and the United States; whereas XP-A is essentially found in Japan. XP-C cells are exclusively deficient in the NER system that corrects the UV-induced lesions on the nontranscribed part of the genome (genome overall repair), whereas they repair normally the lesions on the transcribed strand of active genes by a mechanism known as transcriptioncoupled repair (3)
. The XPC gene codes for a protein able to recognize and bind to bulky DNA lesions. The deficiency in this protein is associated with no global genomic repair of UV-induced DNA lesions, leading to a high mutation rate in irradiated XP-C cells and, therefore, to a dramatically high level of skin cancer at ages as early as 4 years. XPV patients have a normal rate of global and gene-specific NER but are deficient in the process of postreplication repair, which enables cells to synthesize DNA despite the presence of persisting damage in template DNA (4)
. These patients exhibit an increased incidence of skin cancers, but at a later age than in XP-C (between 15 and 40 years of age). The XPV gene has been cloned recently, and it encodes a novel DNA polymerase, the polymerase 
(
), implicated in an "error-free" manner of translesion synthesis (5
, 6)
. The deficiency in this polymerase constrains the XPV cells to use, as a "rescue mechanism," a more error-prone DNA polymerase for replicating damaged templates, leading, therefore, to much higher mutation rates than in normal cells. The newly discovered mutagenic polymerases and 
are the best candidates to replace the deficient polymerase in the XPV cells (7
, 8)
.
It has been reported by Kraemer et al. (2) that the incidence of NMSCs was 4800 times greater in XP than that observed for the general population of the United States and more than 2000 times that observed for malignant melanoma. Although NMSCs have been very well studied in XP, no molecular data are available for XP melanoma, probably because of the small number of patients and the difficulty in getting enough pathological materials of this class of small, but very aggressive, tumors to carry out biochemical studies.
In this report, we studied for the first time 20 skin melanomas, isolated from four XP-C and two XPV patients, which have been characterized and confirmed by pathologists in terms of histological status. We analyzed them for their histogenetic types, their anatomical distribution, and the presence of P53 alterations (protein stabilization and gene mutations). Results show that the pathological types and body localization were different between the XP melanomas and the melanomas reported in the literature for the general population. p53 mutations were detected at a much higher frequency in XP-C melanoma (6 of 10; 60%) compared with results already published for melanoma studied in DNA repair-proficient patients (25 of 351; 7%; reviewed in Ref. 9 ), hence confirming the cellular hypermutability associated with the XP disease. Moreover, the analysis of the p53 mutations shows clearly that they were attributable to replication past unrepaired UV-induced DNA lesions for both classical XP-C patients and XPVs. The mutagenic lesions were not distributed randomly according to the two p53 DNA strands, but instead were found exclusively on the nontranscribed DNA strand in XP-C samples. This result reflects the role of transcription-coupled repair in vivo. Finally, the XP-C melanomas resemble the LMMs found essentially in the elderly.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONREFERENCES |
|---|
. All patients had multiple skin carcinomas. These human studies were carried out in agreement with the French Ethic Law at the time of the biopsies.
|
Pathology Findings.
The histogenetic type of each lesion confirmed as a primary melanoma was classified as LMM-type or other type by analogy to the international classification of malignant melanoma and the guidelines proposed by Flotte and Mihm (13)
and by McGovern et al. (14)
. Thus, LMM-type was defined in this study as melanomas arising in a context of contiguous to often-nested intraepidermal proliferation of highly pleomorphic melanocytes along the dermo-epidermis junction. Tumor thickness was measured from the stratum granulosum vertically to the depth of the tumor at its thickest part, according to Breslow (15)
. RGP and VGP patterns of invasion were evaluated as described already (16)
.
Immunohistochemistry.
Immunohistochemical examination was performed on two thin sections (for each tumor, 5 µm of formalin-fixed and paraffin-embedded archival material). The P53 protein expression was examined using the Mab DO-7 (M 700; Dako, Copenhagen, Denmark). Mab DO-7 was used at a dilution of 1:100 and incubated for 1 h at room temperature after microwave pretreatment in citrate buffer. As to amino acid sequence specificity, DO-7 is directed against a short NH2-terminal segment. Antigen localization was achieved by using the alkaline phosphatase method (Dako). Negative controls were incubated with PBS, and no positive staining was observed. Immunohistochemical staining was recorded using a semiquantitative grading, considering the proportion of tumor cells showing unequivocal positive reaction in the tumor cell nuclei. Intensity was graded as follows: 0, no staining; 1, staining in <10% of the tumor cells; 2, staining in 10–50% of the tumor cells; and 3, staining in >50% of the tumor cells.
Analysis of Mutation on the p53 Gene.
Three to five formalin-fixed, 10-µm thick, paraffin-embedded sections of primary tumors were microdissected using PixCell II laser microdissector (Arcturius, Mountain View, CA, United States) as described previously (17)
. To prevent cross-contamination during sectioning, paraffin and tissue fragments were wiped from the area with xylene between each slide. DNA from 11 paraffin sections (10–20 µm thick) and nine microdissected XP melanomas were extracted as described (18)
. DNA mutations were detected by PCR-SSCP as described already (19)
. Briefly, the coding regions of the p53 gene, covering exons 4, 5–6, 7, and 8–9 were PCR-amplified. After heat-denaturation, the PCR products were separated on Mutation Detection Enhancement (FMC-Bioproducts, Rockland, ME) gels using three different migration procedures to detect the maximum number of mutations (19)
. When shifted bands were found by SSCP, DNA was eluted, reamplified, and sequenced. The total procedure was repeated twice for each sample to exclude Taq polymerase errors. DNA from microdissected tumor samples was amplified and directly sequenced by ABIPRISM 373 automatic sequencer (Perkin-Elmer) to determine the presence of mutations. The polymorphism at codon 72 of the p53 gene was determined on tumoral and nontumoral materials.
Statistical Analyses.
Exact Ps have been calculated using the StatXact package (Statistical Software for exact nonparametric inference, 1997; Cytel Software Corporation, Cambridge, MA). All statistical tests were two-sided. A P of 
0.05 was considered statistically significant.
|
RESULTS AND DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONREFERENCES |
|---|
The XP-C patients are characterized by mutation in the XPC gene, the product of which is involved in the recognition of DNA lesions such as the UV-induced pyrimidine dimer photoproducts. The XPC protein acts only on nontranscribed genomic DNA, therefore allowing the XP-C cells to fully repair the photoproducts on the transcribed strands of active genes (20)
. Probably for that reason, the XP-C patients develop skin cancers at a very early age (in our patients, the age at onset for the first tumor ranged from 5 to 10 years), but exhibit normal neurological features. The XP-C cells were characterized by a very low level of UDS (<15%) after UV-irradiation (Fig. 1)
.
|
, but lack the DNA polymerase able to replicate across cis-syn cyclobutane thymine dimers both accurately and efficiently (6
, 21) . The absence of this enzyme increases the time of bypass at the DNA lesion, and XPV cells are therefore significantly slower than normal cells in synthesizing DNA from a UV-damaged template. As a rescue mechanism, the XPV cells are probably using another DNA polymerase able to carry out an error-prone translesion synthesis. This will lead to a higher frequency of UV-induced mutations than in normal cells (22)
. This higher level of mutations should increase the risk of cancer, which is obvious in the clinical data of the XPV patients where the appearance of skin cancers is found later than with classical XP but still much earlier than in the general population (ages at onset of first melanoma in our two XPV patients were 17 and 36 years).
We have analyzed on tumoral and nontumoral biopsies the p53 gene polymorphism at codon 72 to determine whether it could have an effect on the type of tumor and/or mutations in these patients. We found the Arg/Pro genotype in five patients and Arg/Arg in one patient, but we cannot obtain information on a putative role of this polymorphism because of the small number of patients accessible to us (Table 1)
. However, an interesting loss of heterozygosity was found to be associated with one mutation on the Arg allele in one XPV melanoma (see below).
Among the 20 melanomas analyzed, 15 melanomas were of the LMM type and 5 melanomas were of the SSM type. LMM-type melanomas were characterized by the presence of atypical dendritic melanocytes disposed along the dermo-epidermis junction. These melanocytes were highly pleomorphic with hyperchromatic nuclei and were often contiguous (Fig. 2a)
. Confluent nesting and involvement of hair follicles were present. When it occurred, invasion into the dermis was represented by atypical spindle melanocytes associated with a multinodular lymphocytic infiltrate. Neither elastolysis nor other sun-related damages that are always observed in classical LMMs in the elderly (Fig. 2b)
were present in these patients. An excess of LMM-type melanomas was also reported in XP-A patients from Japan (23)
, but their frequencies were not available in the large XP report by Kraemer et al. (2)
because in this study the types of melanomas were rarely reported. Five melanomas were characterized by a prominent intraepidermal spreading of pagetoid melanocytes (Fig. 2c)
. These SSMs had similar aspects to the most common type of melanomas in the general population. Sixteen of 20 melanomas were thinner than 1 mm, including two MIS. Four melanomas were between 1.1 and 6.6 mm in thickness. All of the tumors, including those isolated from the same XP patient, were considered as independent events because they were located at different body sites and occurred mostly at different periods of time. It is important to recall here that most XP patients develop hundreds of skin tumors, and therefore it is not abnormal to obtain several biopsies from the same patient.
|
We have analyzed an average of three to five thin sections of 10 melanomas isolated from four different XP-C patients and 10 melanomas from two XP variants (Table 1)
. SSCP of p53 sequences with subsequent DNA sequencing of abnormal migrating bands or direct sequencing of microdissected tumors revealed seven mutations; six of them were localized in the coding sequences, whereas one was in intron 4 of p53. Six mutations were found in the 10 XP-C tumors (60%), whereas only one mutation was found in 10 tumors from XPVs (10%).
p53 Mutations in Melanoma from XP-C Patients.
Among the six mutations found in XP-C patients, three mutations corresponded to tandem transitions CC to TT and two mutations are C to T transitions, both events being characteristic of the UV signature. All CC to TT tandem mutations were localized at methylated CpG sites, which are known to be methylated in the p53 gene, confirming the idea that this type of mutation is linked to the deamination of 5 methyl-cytosine (25)
. We have compared the distribution of mutations in the 8 mutations found in XP-C melanomas (six mutations in Table 1
and two mutations in Ref. 19
) and the 51 mutations reported in the database for non-XP melanomas (Ref. 26
and reviewed in Ref. 9
). The frequency of C to T transition is statistically not different between DNA-repair-proficient patients (55%) and XP-C patients (25%), whereas the frequency of tandem mutations CC to TT is statistically different between XP (67.5%) and non-XP (4%) tumors (P < 0.001). The distributions of C to T, CC to TT, and other mutations are statistically different (P < 0.001) between XP-C and non-XP melanomas. This result correlates well with previous data on p53, INK4a-ARF, and PTCH genes in NMSC from XP or normal individuals where XP mutations are characterized by about 50–60% tandem mutations, whereas non-XP tumors exhibit a frequency of 
10–15% for this type of mutation (18
, 26, 27, 28)
.
One hundred percent of the p53 mutations in XP melanomas from this study are located on pyrimidine-pyrimidine sequences that represent hot spots of DNA lesions after sun exposure. These lesions and these mutations are characteristic of the UV-B part of the solar spectrum (29 , 30) . In the XP-C group, 100% of these bipyrimidine sequences (6 of 6) were located on the nontranscribed strand of the p53 gene although the same number of bipyrimidine sites is found on the two strands of the p53 gene (170 dipyrimidine sites were found on each strand of the p53 gene between exons 4–9; P = 0.05). This result demonstrates the normal efficiency of XP-C patient cells to fully repair DNA lesions on the transcribed strand of active genes and the complete deficiency to do so on the nontranscribed strand. This is in agreement with the existence of preferential repair in vivo in the XP-C patients as also found with the p53, p16, and PTCH genes in NMSCs of XP (25) .
The repair of UV-induced DNA lesions on the transcribed strand of active genes in XP-C cells allows a better survival of the damaged cells because of a lower apoptosis rate than in other XP groups. This allows a rapid recovery of the cell cycle blockage after irradiation. However, the absence of DNA repair in the genomic DNA of XP-C patients gives rise to a high level of mutations opposite unrepaired photoproducts by translesion replication and therefore to an early onset of skin cancers. Indeed, a high level of mutations (about 60–70%) on key genes has been found in XP-C NMSCs (25)
. This is the same for XP-C melanomas, inasmuch as we found mutation in 60% of these tumors (Table 1)
. This number is much greater than in melanomas from DNA-repair-proficient patients, where a frequency of 10% can be calculated when all classes of melanomas are taken into account (9
, 31
, 32)
.
p53 Mutations in Melanoma from XPV Patients.
The p53 gene of XPV melanomas shows a low mutation frequency, because no mutation among six tumors from one patient and only one mutation among four tumors from the other XPV patient were recorded (Table 1)
. This frequency of 10% corresponds roughly to that observed in the general population. Interestingly, the mutation found in this XPV tumor is also located opposite a bipyrimidine sequence (that is, a hotspot of UV-induced DNA lesions), but this type of mutation (an A:T to G:C transition) does not correspond to a classical UV-induced one. Among 107 published p53 mutations in all skin cancers from XP patients, only three other A:T to G:C transitions were found corresponding, indeed, to a basal cell carcinoma on the cheek of another XPV at codon 135 (18)
and to two other patients whose clinical characteristics suggest to us that they should be of the variant form (28
, 33)
.
Interestingly, it is known that the mutation spectra are different between UV-irradiated classical XP and XPV cultured cells (34)
. Almost 50% of UV-induced mutations in cultured XPV cells involved A-T bp (whereas it is only 12% in classical XP cells). Among these mutations at A:Ts, almost half are A:T to G:C transitions (34)
in cells irradiated in early S phase, as found in our mutated XPV melanoma (Table 1)
. This high mutation rate on thymine photoproducts could be attributable to a less-frequent incorporation of dAMP opposite UV-photoproducts (as found in normal or classical XP cells) in XPV cells. Because of the deficiency in the translesion polymerase in XPV (5
, 21)
, it is hypothesized that another translesion polymerase will take over, but with less efficiency and more mutagenic potency, during the replication of damaged templates. The error-prone polymerases or could be the candidates (7
, 8)
. The mutagenic characteristics of the polymerase , known to produce errors in vitro by misinserting G rather than the Watson-Crick complementary base A at template T, would explain the results concerning the T to C mutations in XPV skin cancers (8)
. This effect would result in a significantly higher frequency of mutations and cancer. Reports with XPV cell extracts replicating a shuttle vector in vitro also showed a different mutation spectra as compared with wild type or classical XP cells. The accuracy of bypass replication and the type of induced mutations were different between the lagging and the leading strands, indicating that the mechanistics of bypass were different between the two strands (35)
. As we found here with p53 mutations in XPV tumors (Ref. 18
and this work), in the SV40 in vitro assay, more mutations were found to be located at Ts than at Cs. For the mutation isolated in our XPV melanoma, the bipyrimidine sequence (T–C) was located on the transcribed strand, confirming then the absence of strand selectivity for repair in this class of patients, which is in agreement with the fact that in XPV patients, a normal NER efficiency is acting on both DNA strands. In this XPV melanoma, the mutation was associated with a loss of heterozygosity of the allele Pro 72 of the p53 gene (the study of the p53 gene in normal cells of this patient showed us that it was heterozygote Arg/Pro), indicating the presence of a state of high genetic instability in this tumor. Such genomic instability has been reported in an XPV fibroblast line in the absence of an active p53 pathway (36)
. Interestingly, a preferential interaction between the wild-type P73 protein and the mutated Arg72 isoforms of P53 has been shown recently in NMSC cells. The loss of the wild-type Pro72 isoform associated with some mutations on the Arg72 allele should induce a loss of P73 DNA binding capacity and, therefore, its ability to exert a tumor-suppressor activity (37)
. These molecular modifications are exactly those found in one of the XPV melanomas studied here.
Relationship between Tumor Types and p53 Mutations.
The growth phase concept in melanoma argues that the ability to metastasize is recognizable histologically with features of a dermal nodule of invasive melanocytes (VGP; Fig. 2a
; Refs. 38
and 39
). On the contrary, as for MIS, invasive melanomas with a RGP pattern (Fig. 2a)
would be devoid of metastatic potential. Therefore, we have evaluated whether p53 mutations in XP melanomas are associated with the presence of a VGP. Although the total number of mutated melanomas is small, the mutation frequency is lower (P < 0.03) for MIS and invasive LMMs with RGP (3 of 11; 27% of mutated tumors) than for invasive LMMs with VGP (4 of 4; 100% of mutated tumors).
Immunohistochemistry of Melanoma.
Immunohistochemistry using anti-P53 antibody was performed on two thin sections for the 18 tumors indicated in Table 1
. Nuclear positivity was observed in five cases, associated in three cases with mutation of the p53 gene and in two cases classified as wild type. For three mutated melanomas, the immunohistochemistry did not reveal any P53 stabilization. Discordance between immunohistochemical results for P53 and the research for mutations was already reported (9
, 32)
. For example, a percentage of >70% of P53-positive melanomas have been found (reviewed in Ref. 32
), whereas the mutation frequency is roughly 10 times lower. Although this discrepancy could be attributable to the fact that immunohistochemistry and molecular analysis were performed on different thin sections of the tumors, these findings may indicate that p53 mutations and increased P53 protein expression are not closely related in melanomas and may represent alternative mechanisms of P53 alteration (31)
. No cytoplasmic positivity was observed in this study.
|
CONCLUSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONREFERENCES |
|---|
(5
, 21)
, which is able to replicate in vitro the thymine-thymine dimers induced by UV-light on DNA. This enzyme shows extremely limited processivity on undamaged DNA and has been shown to have a much lower fidelity than any other template-dependent DNA polymerase studied (41)
. The fundamental role of this enzyme is attested by the fact that its defect gives rise to the XPV syndrome, indicating an anticarcinogenic activity in vivo. The absence of this enzyme induces the blockage of replication forks at DNA lesions for a longer time than usual (this is the basis of the cellular characterization of the XPV cells). At stalled DNA polymerase, a "more error-prone" polymerase, such as the polymerases or , will eventually bypass the lesion, giving rise to a much higher translesion mutation rate, with the particular spectrum found in XPV cells and tumors.
Although melanomas were classified as tumors relatively independent of the mutated P53 pathway, our results indicate this is not true for those isolated from XP-C patients. This could be explained by the higher mutation rates in XP cells and XP skin tumors as measured on various protooncogenes or tumor suppressor genes because of the absence of an efficient repair process and/or by the different histogenetic types of melanomas found in XP versus DNA repair-proficient individuals. Although the number of mutations found in melanomas of XP patients is limited, the distributions of C to T, CC to TT, and other mutations are statistically different (P < 0.003) between melanomas from all XP groups and non-XP melanomas (Fig. 3)
. Moreover, it seems that the distribution of these along the p53 gene is different from the distribution of mutations found in melanomas of non-XP patients. In fact, the hotspot found in non-XP melanomas at the level of the codon 213 is not found in XP melanomas. The corresponding sequence is 5'-TCGA-3', indicating clearly that in DNA repairproficient patients, mutations can be produced at the TC bipyrimidine dimer but not in XP cells, where, in most cases, mutations arise at the CC dimer. Hotspots at 247–248 are found within the two distributions, but in DNA repair-proficient patients, the mutations are always C to T, whereas they are always CC to TT for XP tumors (Ref. 19
and Fig. 4
). This result has been already found with other XP-C patients and other analyzed genes and represents a clear example of the very high incidence of tandem mutations typical of classical XP cells (25)
.
|
|
The higher proportion of XP melanomas and LMMs in the elderly in anatomical sites receiving the greatest sun exposure (head and neck localization) compared with that of melanomas in young DNA repair-proficient individuals indicates a greater influence of UV irradiation for LMM induction. The high mutation rate in XP-C patients, attributable to global NER deficiency, is probably responsible for both the high frequency of induction of skin tumors in XP patients as well as their early appearance in childhood. The important role of unrepaired DNA photolesions in this early incidence of melanoma in XP-C patients is attested to by the mutation spectra, which shows mutations characteristic of UV exposure; whereas this is less obvious for non-XP melanomas, which are less dependent upon sun exposure (Fig. 3)
. The similarity between XP LMMs and LMMs in the elderly allows us to hypothesize that, in the elderly, the appearance of LMM is also attributable to the accumulation of UV-induced DNA lesions with time of sun exposure, leading ultimately to replication of some of the unrepaired lesions, then to mutations, and finally to this type of tumor. Interestingly, several reports showed decreasing DNA repair activity with increasing age of the individuals. These similar characteristics suggest a common pathway at the origin of the LMM in XP and in elderly people. This may represent an interesting comparison, as XP is a rare disease with well-characterized molecular events, whereas LMM is becoming a public health problem but has still poorly characterized molecular events.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 Giuseppina Giglia-Mari had a fellowship from the Ligue contre le Cancer du Val de Marne (Créteil, France). 
2 To whom requests for reprints should be addressed, at UPR 2169, Laboratory of Genetic Instability and Cancer, Institut de Recherches sur le Cancer, 7, rue Guy Moquet, B. P. 8-94801, Villejuif Cedex, France. Phone: 33-1-49-58-34-20; Fax: 33-1-49-58-34-11; E-mail: sarasin{at}infobiogen.fr
3 The abbreviations used are XP, xeroderma pigmentosum; NER, nucleotide excision repair; NMSC, non-melanoma skin cancer; LMM, lentigo maligna melanoma; UDS, unscheduled DNA; RGP, radial growth phase; VGP, vertical growth phase; SSCP, single-strand confirmational polymorphism; SSM, superficial spreading melanoma; MIS, melanoma(s) in situ.
Received 10/24/00. Accepted 1/12/01.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONREFERENCES |
|---|
. Proc. Natl. Acad. Sci. USA, 95: 6876-6880, 1998.
. Nature (Lond.), 404: 1011-1013, 2000.[Medline]
This article has been cited by other articles:
|
|
 |
|
  L. A. Fecher, S. D. Cummings, M. J. Keefe, and R. M. Alani Toward a Molecular Classification of Melanoma J. Clin. Oncol., April 20, 2007; 25(12): 1606 - 1620. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. N. Markovic, L. A. Erickson, R. D. Rao, R. R. McWilliams, L. A. Kottschade, E. T. Creagan, R. H. Weenig, J. L. Hand, M. R. Pittelkow, B. A. Pockaj, et al. Malignant Melanoma in the 21st Century, Part 1: Epidemiology, Risk Factors, Screening, Prevention, and Diagnosis Mayo Clin. Proc., March 1, 2007; 82(3): 364 - 380. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Mitchell Revisiting the photochemistry of solar UVA in human skin PNAS, September 12, 2006; 103(37): 13567 - 13568. [Full Text] [PDF]
|
|
|
|
 |
|
 S. Arad, N. Konnikov, D. A. Goukassian, and B. A. Gilchrest T-oligos augment UV-induced protective responses in human skin FASEB J, September 1, 2006; 20(11): 1895 - 1897. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. C. Bastian, A. B. Olshen, P. E. LeBoit, and D. Pinkel Classifying Melanocytic Tumors Based on DNA Copy Number Changes Am. J. Pathol., November 1, 2003; 163(5): 1765 - 1770. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y.-H. You, D.-H. Lee, J.-H. Yoon, S. Nakajima, A. Yasui, and G. P. Pfeifer Cyclobutane Pyrimidine Dimers Are Responsible for the Vast Majority of Mutations Induced by UVB Irradiation in Mammalian Cells J. Biol. Chem., November 21, 2001; 276(48): 44688 - 44694. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
