| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Regular Article |
From the Institute of Pathology,*
and the Clinic for
Urology,
University Basel, Basel, Switzerland;
VYSIS Incorporated,
Downers Grove, Illinois;
and the Laboratory of Cancer Genetics,§
National Human Genome Research Institute, National Institutes of
Health, Bethesda, Maryland
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Recent evidence has implicated CDKN2A located at chromosome 9p21 to be frequently aberrant in the germline of members of familial melanoma kindreds, but also in bladder and other solid tumors.6 CDKN2A encodes a 156-amino acid protein that exclusively associates with CDK4 and CDK6, inhibiting their complexation with D-type cyclins and the consequent phosphorylation of the retinoblastoma protein.7 This contributes to cell-cycle arrest. Inactivation of CDKN2A has been described as a consequence of homozygous deletion, rearrangement, hypermethylation, or point mutation.8-10 The involvement of CDKN2A in human cancer is controversial. It has been shown that the rates of homozygous deletions or mutations is smaller in primary tumors than in cell lines.9,11
The role of CDKN2A in CRCC is uncertain.1,12 Homozygous deletions of CDKN2A have been detected in up to 56% of kidney cancer cell lines.13 Hypermethylation of the promoter region of CDKN2A were reported in 23% of RCC cell lines.10
To further determine the significance of CDKN2A alterations in primary CRCC we performed: 1) sequence analysis of CDKN2A in 113 CRCCs; 2) fluorescence in situ hybridization (FISH) to search for homozygous CDKN2A deletions; 3) immunohistochemical CDKN2A expression analysis using a renal tumor microarray containing 532 renal tumors; and 4) loss of heterozygosity (LOH) analysis using four polymorphic microsatellites on 9p.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
All renal tumors from the archive of the Institute for Pathology, University of Basel, were reviewed by one pathologist (HM). One hundred thirteen consecutive nephrectomy specimens from 1985 to 1994 with CRCC14 were selected for this study. Histological grading and tumor staging were done according to Thoenes and colleagues15 and International Union Against Cancer (UICC).16 There were 11 grade 1, 71 grade 2, and 31 grade 3 tumors. Forty tumors were stage pT1, eight were pT2, 64 were pT3, and one was pT4. An extensive tissue sampling to ensure perinephric fat infiltration may account for the relative high proportion of high-stage cases in this tumor set. The mean tumor diameter was 6.6 ± 3.0 cm. There was lymph node metastasis at nephrectomy in seven patients and evidence for hematogenous metastasis (pM1/cM1) in 19 patients.
Overall and tumor-specific survival data were obtained by reviewing the hospital records, by direct communication with the attending physicians, and from the Cancer Registry of Basel. Patients were evaluated from the time of biopsy diagnosis to the last known follow-up. Overall survival was available from 100 patients and tumor-specific survival was available from 83 patients. The mean follow-up time was 48.9 months (median, 48 months; range, 2 to 131 months). Thirty-five patients had a 5-year follow-up. Forty-one patients died within 5 years after surgery. There was a significant association between short tumor-specific survival and presence of metastasis (P < 0.0001), pT stage (P < 0.001), and histological grade (P < 0.05).
DNA Preparation
Formalin-fixed, paraffin-embedded tumor tissue from each of the 113 patients was selected on the basis of hematoxylin and eosin (H&E)-stained sections to ensure a minimum of 75% tumor cells in the samples. Twenty to 30 mg of normal and tumor tissue was scraped away from the paraffin blocks using a scalpel. Areas of normal tissue were defined using H&E tissue sections. Deparaffinizing of the tissues and DNA extraction were performed according to the QIAmp Tissue Kit protocol (Qiagen, Basel, Switzerland).
Microsatellite Analysis
Analysis of allelic deletions was performed using primers for
microsatellites D9S970 (9p12-9p13), D9S171
(9p13), D9S1748 (9p21), and D9S156 (9p21). Primer
sequences were obtained from the Genome Data Base. The locus of
D9S1748 lies upstream of exon 1 of
CDKN2A.17
Primers were labeled with
T4-polynucleotide kinase (Catalys, Wallisellen, Switzerland) and
32P-ATP (Amersham/Pharmacia, Zurich,
Switzerland) for microsatellite analysis. Polymerase chain
reaction (PCR) amplification was performed in a total of 15 µl
containing 50 to 100 ng DNA, 1x Taq buffer (Qiagen), 200
µmol/L dNTPs, 3 pmol of each primer, 0.4 pmol
32P-labeled primer, and 1 U Taq DNA
polymerase (Qiagen). An initial denaturation step of 95°C for 3
minutes followed by 35 cycles each of 95°C for 30 seconds, 50°C to
55°C for 30 seconds, and 72°C for 1 minute with a final extension
step of 72°C for 5 minutes comprised the PCR profile. Products were
separated by electrophoresis in denaturing 6% polyacrylamide gels
followed by autoradiography as previously described.18
For
informative cases, allelic loss was scored if the radiographic signal
of one allele was >50% reduced in the tumor DNA as compared with the
corresponding normal allele.
FISH for Detection of CDKN2A Deletions
FISH was performed as previously described.19
One
section from each tissue block was stained with H&E to ensure the
presence of at least 90% tumor cells and nuclei of formalin-fixed
tissue blocks were dissociated and dropped onto slides. Two-color FISH
was performed using a 180-kb Spectrum Orange-labeled 9p21 probe (Vysis,
Downer’s Grove, IL), spanning the minimal homozygously deleted region
that includes CDKN2A and excludes CDKN2B as
described by Cairns and colleagues,1
together with a
corresponding Spectrum Green-labeled centromeric 9 
satellite probe
(CEP 9, Vysis). Slide pretreatment, hybridization, and washing
procedures were performed as previously described.19
The
hybridization mixture contained 3 µl of each of the probes and Cot-1
DNA (1 mg/ml). Slides were counterstained with 0.2 µmol/L
4,6-diamidino-2-phenylindole. At least 100 nuclei were selected for
scoring using 4,6-diamidino-2-phenylindole staining. A loss of one
CDKN2A allele was defined as presence of less 9p21 than
centromere 9 signals in >40% of nuclei. A tumor was considered
monosomic for chromosome 9 if >50% of the nuclei showed only one
signal for (chromosome) 9p21 and CEP9. A homozygous deletion of 9p21
was considered, if >50% of the nuclei showed centromere 9 signals
without 9p21 signals.
Sequence Analysis
Exon 1 and exon 2 of CDKN2A were amplified with two
primer sets (Table 1)
, each resulting in
two overlapping fragments. PCR samples for the first PCR contained
100 ng of genomic DNA, 10 pmol of each primer, 1x Taq
buffer (Qiagen), 200 µmol/L of each nucleotide (dATP, dCTP, dGTP,
dTTP) and 1 U Taq polymerase (Qiagen) in a 20 µl total
volume. PCR conditions for the step-down PCR were: 95°C for 3
minutes; 10 cycles of 95°C for 40 seconds, 62.5°C to 64.6°C for
40 seconds, 72°C for 1 minute; 25 cycles of 95°C for 40 seconds,
55.5°C to 57°C for 30 seconds, 72°C for 1 minute; 72°C for 5
minutes. PCR conditions for exon 3 were: 95°C for 3 minutes; 35
cycles of 95°C for 40 seconds, 55°C for 30 seconds, 72°C for 1
minute; 72°C for 5 minutes.
|
IRD800-labeled primers (MWG-Biotech, Ebersberg, Germany) were used to direct cycle-sequencing of the PCR products. Cycle sequencing was done according to the protocol of the manufacturer’s instructions (MWG-Biotech). Sequence products were analyzed on a LICOR-DNA sequencer (Model 4000). Tumor DNA that showed a sequence change was subjected to a second PCR and sequence analysis. In those cases the matched normal DNA was also examined.
Immunohistochemistry for CDKN2A Expression on a Renal Tumor Tissue Microarray
To evaluate the prevalence of CDKN2A expression in renal tumors, a renal tumor microarray was used containing tumor specimens from 532 renal tumors and tissue from six normal kidneys. The construction of the tumor microarray was previously described.20,21 There were 383 conventional (clear cell), 57 papillary, 23 chromophobe, three collecting duct RCCs, and 17 oncocytomas. The presence of tumor tissue on the arrayed samples was verified on one H&E-stained section. Sections (5-µm thick) were cut for immunohistochemistry. Standard indirect immunoperoxidase procedures were used for immunohistochemistry (ABC-Elite; Vector Laboratories, Burlingame, CA) on the tumor microarray as previously described.21 A well-characterized antibody to CDKN2A7 was used for detection of CDKN2A expression (1:25, Ab-1; Oncogene Research Products, Cambridge, MA) after pronase pretreatment. Tumors were considered positive if an unequivocal nuclear positivity was seen in tumor cells. Cases of prostate and bladder carcinoma with nuclear CDKN2A expression served as external positive control.
Statistics
Contingency table analysis was used to analyze the relationship between allelic deletion, grade, stage, and presence of metastasis. Overall survival rates were plotted using the Kaplan-Meier method. Statistical differences between tumors with and without 9p deletions were determined with the log-rank test. The proportional hazard model was used to test for independent prognostic information.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Eighty-eight normal/tumor pairs could be interpreted for LOH
analysis on chromosome 9p. Twenty-one of 88 CRCCs (24%) showed allelic
deletion with at least one microsatellite. Most deletions were detected
at D9S171 on 9p13. Twelve of 59 informative tumors (20%)
displayed allelic loss in this region. In contrast, only seven of 73
informative cases (10%) showed LOH at D9S1748 (9p21), eight
of 68 (9%) demonstrated allelic deletion at D9S156 (9p21),
and five of 68 (7%) displayed LOH at D9S970 (9p12-9p13).
Allelic deletion at only one of the microsatellite loci examined were
found in six tumors (10%) at D9S171, followed by three
tumors at D9S1748 and D9S156 (4% each),
respectively, and two tumors at D9S970 (Figure 1, A and B)
.
|
. Also, the separate
analysis of the individual microsatellites did not reveal a
relationship between LOH and morphological parameters. Tumor-specific
survival data were available for 62 patients being informative for at
least one 9p microsatellite. Fifteen of these patients showed LOH.
Tumor-specific survival rates were only analyzed in stage pT3 RCCs,
because there were too few tumor-related deaths in pT1/2 RCCs. In
univariate survival analysis patient prognosis was associated with 9p
deletion in stage pT3 RCCs. Tumor-specific 5-year survival was 58% for
patients without 9p deletion (n = 30), whereas
all patients with 9p deletions (n = 7) died of
disease (P = 0.01, Figure 2
). The analysis of pT3 CRCCs without
metastasis at the time of nephrectomy also revealed a significant
association between 9p deletions and poor prognosis
(P = 0.02), although the number of tumors with
deletions was small (n = 4) within this
analysis. Cox proportional hazard analysis including the variables
histological grade, presence of metastasis, and 9p deletion showed that
LOH at 9p was not an independent predictor of poor prognosis in pT3
CRCCs (Table 3)
.
|
|
|
There was a 24-bp deletion within exon 1 in 13 of 113 tumors
(12%) (Figure 3A)
. One patient had this
deletion also in the normal DNA. This in-frame mutation results in a
truncation of 8 amino acids of CDKN2A. Because of the special feature
of the sequence in the start coding region of CDKN2A there
were 25 possible variants of deletions leading to the same sequence
alteration (Figure 3B)
. The 24-bp deletion was not associated with
tumor stage or presence of metastases. Interestingly, the deletion was
not detected in grade 1 RCCs, but in 9% of grade 2 and 22% of grade 3
tumors (Table 4)
. This trend did not
reach significance (P = 0.07). There was no
relationship between the 24-bp deletions and tumor-specific survival
(P = 0.15).
|
|
One hundred eight of 113 CRCCs (96%) showed a G to C transversion in
the noncoding sequence of exon 3 that was also found in the matched
normal DNA of each tumor. Eighty-nine of these tumors were homozygous
and 19 were heterozygous for this transversion. There were only five
tumors showing the wild-type sequence of the gene data
base.22
Interestingly, the homozygous C/C genotype was
associated with a low grade of differentiation
(P < 0.05) (Table 5)
. To further evaluate the
frequency of the C haplotype in the normal population, we
analyzed 48 blood samples of healthy donors. The analysis revealed a
C-allele frequency of 90% of the individuals
tested.
|
High-quality hybridization signals for both centromeric and gene-specific probes were obtained in 54 tumors. Four of 54 CRCCs (7%) showed physical 9p21 deletions, but none of the tumors had a homozygous CDKN2A deletion according to our definition. There were two additional tumors with chromosome 9 monosomy.
CDKN2A Expression Tested on a Renal Tumor Tissue Microarray
To determine the frequency and potential implications of CDKN2A
expression in RCC, we analyzed a cohort of 532 renal tumors on a tumor
microarray. We observed six CRCCs with nuclear CDKN2A expression. These
tumors showed neither CDKN2A mutations nor
D9S1748 deletions. Papillary (n =
57), chromophobe (n = 23), and collecting duct
carcinomas (n = 3) as well as oncocytoma
(n = 17) did not show CDKN2A expression.
Prostate and bladder carcinomas served as external positive controls.
In these controls, a strong nuclear staining was detectable.
Representative tumors are shown in Figure 4, A and B
.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Allelic imbalance has generally been considered to represent loss of genetic material, but the use of PCR-based techniques makes it difficult to differentiate allele loss from allele gain. Although it cannot be excluded that some of our tumors had gains we think that the vast majority of our tumors analyzed showed LOH. In approximately half of the cases signal reduction of one allele was near absolute in the tumors. Only four of 24 tumors (17%) with LOH displayed a 50% reduction of one allele. In addition, our FISH results coincide with previous cytogenetic data in that regional or entire gains of chromosome 9 are rare events in kidney tumors.27 Hence, signal reduction of one allele very likely reflects loss rather than gain of genetic material on chromosome 9.
CDKN2A has been implicated as a potential target for the deletion on 9p in a variety of neoplasms including RCCs, because CDKN2A was found inactivated by homozygous deletion and hypermethylation of the promoter region.9,10 In our analysis, only seven of 73 CRCCs (10%) demonstrated allelic deletion at D9S1748 locus that lies within CDKN2A. A low frequency of CDKN2A deletions was also corroborated by our FISH analysis showing CDKN2A deletion in four of 54 CRCCs (7%). The absence of homozygous deletions by FISH is consistent with previous findings of Cairns and colleagues,1 who detected inactivation of CDKN2A by homozygous deletion only in one of 42 primary RCCs. Although our FISH probe spans the minimal homozygously deleted region at CDKN2A described by Cairns and colleagues,1 we cannot exclude the presence of homozygous deletions <180 kb. However, the large proportion of homozygous CDKN2A deletions in RCC cell lines described by Kinoshita and colleagues4 may rather occur as result of a selective overgrowth of tumor cells with CDKN2A homozygous losses in tissue cultures.
The most common mechanism for CDKN2A inactivation in human cancer cell lines is through homozygous deletion, whereas intragenic mutations occur in a smaller proportion of tumors and cell lines.9,11 Cairns and colleagues11 were not able to detect point mutations of CDKN2A in 42 RCCs with chromosome 9p losses and argued that CDKN2A is not the primary target of the 9p21 loss. Importantly, we detected three different CDKN2A sequence alterations, one in exon 1, one in exon 2, and an additional alteration in the 3'-untranslated region of exon 3. Thirteen of 113 CRCCs (12%) showed a 24-bp deletion in exon 1, causing the loss of amino acid residues 4 to 11.
The same mutation was also detected at a lower rate in melanoma and
prostate cancer. Kumar and colleagues28
found a 24-bp
deletion in two of 31 (6%) sporadic melanomas and Komiya and
colleagues29
reported a 24-bp deletion in one of 51 (2%)
prostate cancers. Because of the special sequence structure of the two
repeat units located at the 5' end of exon 1, there are 25 possible
variants leading to the same sequence alteration (Figure 3B)
. The
relevance of these mutations for gene expression or protein function of
CDKN2A is unclear. This repeat region seems to be unstable because
duplications of this 24-bp repeat were found in familial
melanomas.30-32
Based on the functional studies of
CDKN2A, this mutant form was functionally normal with respect to CDK4
binding.33
As both 24-bp insertion and deletion lie
outside the ankyrin-repeat region of CDKN2A, these mutations
may not have any negative effects on the interaction between CDKN2A and
CDK4. Whether the 24-bp deletion in CRCC weakens CDKN2A
expression in tumors remains to be determined. The association of these
mutations with higher grades raises the possibility that this mutation
is associated with progression or genetic instability of RCC.
The ratio of wild-type and mutated CDKN2A alleles was approximately equal in the majority of the tumors suggesting a mechanism that leads to a gain of heterozygosity. As some tumors showed less amounts of the mutant allele the existence of a small subpopulation of heterozygous or homozygous tumor cells cannot be excluded. Therefore, the observed mutation rate in our tumors might be a slight underestimation.
The G to A transition in codon 140 of exon 2 is likely to be a genetic polymorphism rather than a somatic mutation. This polymorphism has been described in many primary tumors.34-36 The transition results in an amino acid exchange from alanine to threonine. This point mutation may not influence CDKN2A function, because the altered protein inhibits the CCND1-CDK4/6 complex in the late G1 phase of the cell cycle similarly to the wild-type protein.37
The G to C transversion in the noncoding sequence (3'-UTR) of exon 3 (nucleotide 540) represents a known polymorphism, which was present in tumor and normal tissue of 108 patients. The same transversion was found by Ueki and colleagues38 in eight of 30 primary astrocytomas and in 11 of 50 CEPH DNAs. The transversion was heterozygous in all these cases. This represents a G-allele frequency of 0.12 and a C-allele frequency of 0.88. Kumar and colleagues28 described a G-allele frequency of 0.18 and a C-allele frequency of 0.82 in sporadic melanoma. In our study, 96% of CRCCs showed a heterozygous polymorphism, the C-allele frequency was 0.88 and the G-allele frequency was 0.12. The identical G and C allele frequencies in DNA extracted from blood of 48 healthy donors argues against a role of this polymorph-ism for a CRCC disposition.
Methylation of the promoter region is another mechanism for CDKN2A inactivation in cell lines and primary tumors.10 Merlo and colleagues39 demonstrated methylation in one RCC cell line, suggesting that inactivation because of hypermethylation may play a role in RCC. However, a previous analysis by Clifford and colleagues40 detecting CDKN2A hypermethylation only in one of 44 primary RCCs (2%) suggested that this mechanism may not be relevant for CDKN2A inactivation in primary CRCC.
Mutations affecting CDKN2A, homozygous deletions, or methylation of the promoter region might result in a lack of CDKN2A expression. In this analysis, none of the normal kidney tissues and only six CRCCs in the renal tumor microarray displayed detectable levels of CDKN2A. It has been reported that CDKN2A expression is low or even undetectable in most normal human tissues that reflects the basal physiological level of CDKN2A.41,42 Therefore, the negative phenotypes of CRCCs found in this study may rather reflect the normal physiological state than a reduced expression. Frequent expression of CDKN2A can be seen in prostate,43 lung,44 ovarian,45 and breast carcinoma.46 CDKN2A expression has been associated with tumor progression and poor prognosis in ovarian47 and breast cancer46 and a higher likelihood of tumor recurrence in prostate cancer.43 Up-regulation of CDKN2A, resulting in expression of CDKN2A, may develop through different mechanisms. Cellular stress,48 hyperthermia,49 and UV irradiation50 have been reported to trigger CDKN2A expression. Some data suggest an association between CDKN2A and retinoblastoma protein (RB1) expression, in which the absence of functional RB1 limits CDKN2A activity. Indirect mechanisms for CDKN2A expression might be because of E2F1 or CCND1 expression.7
In summary, CDKN2A alterations (mutations and/or deletions)
are present in a subgroup of primary CRCCs. Importantly, LOH was more
frequent at D9S171, which was initially mapped to 9p21
(Genome Database; Gene Map ‘96). Subsequent mapping has placed this
marker to 9p13 10 cM centromeric to CDKN2A (GeneMap’99).
The significantly lower rates of allelic losses found centromeric
(D9S970) and telomeric (D9S1748 and
D9S156) to D9S171 strongly support the existence
of an unknown tumor suppressor gene proximal to the CDKN2A
locus. Such a tumor suppressor gene might also be involved in other
tumor types because preferential allelic loss at D9S171 has
been observed in lung,51
head and neck,52
brain,53
breast,54
and esophagus
carcinomas,55
as well as in melanoma,56
mesothelioma,57
neuroblastoma,58,59
pituitary
adenoma,60
and childhood acute lymphoblastic
leukemia.61
|
Acknowledgements |
|---|
|
Footnotes |
|---|
Supported by Swiss National Science Foundation (31-50752.97).
Accepted for publication October 24, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  R. Beroukhim, J.-P. Brunet, A. Di Napoli, K. D. Mertz, A. Seeley, M. M. Pires, D. Linhart, R. A. Worrell, H. Moch, M. A. Rubin, et al. Patterns of Gene Expression and Copy-Number Alterations in von-Hippel Lindau Disease-Associated and Sporadic Clear Cell Carcinoma of the Kidney Cancer Res., June 1, 2009; 69(11): 4674 - 4681. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Klatte, P. N. Rao, M. de Martino, J. LaRochelle, B. Shuch, N. Zomorodian, J. Said, F. F. Kabbinavar, A. S. Belldegrun, and A. J. Pantuck Cytogenetic Profile Predicts Prognosis of Patients With Clear Cell Renal Cell Carcinoma J. Clin. Oncol., February 10, 2009; 27(5): 746 - 753. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Cheng, G. T. MacLennan, S. Zhang, M. Wang, M. Zhou, P.-H. Tan, S. Foster, A. Lopez-Beltran, and R. Montironi Evidence for Polyclonal Origin of Multifocal Clear Cell Renal Cell Carcinoma Clin. Cancer Res., December 15, 2008; 14(24): 8087 - 8093. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R Natrajan, W Warren, B Messahel, J S Reis-Filho, M-A Brundler, J S Dome, P E Grundy, G Vujanic, K Pritchard-Jones, and C Jones Complex patterns of chromosome 9 alterations including the p16INK4a locus in Wilms tumours J. Clin. Pathol., January 1, 2008; 61(1): 95 - 102. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. D. Jones, J. N. Eble, M. Wang, G. T. MacLennan, B. Delahunt, M. Brunelli, G. Martignoni, A. Lopez-Beltran, S. M. Bonsib, T. M. Ulbright, et al. Molecular Genetic Evidence for the Independent Origin of Multifocal Papillary Tumors in Patients with Papillary Renal Cell Carcinomas Clin. Cancer Res., October 15, 2005; 11(20): 7226 - 7233. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Gunawan, A. von Heydebreck, T. Fritsch, W. Huber, R.-H. Ringert, G. Jakse, and L. Fuzesi Cytogenetic and Morphologic Typing of 58 Papillary Renal Cell Carcinomas: Evidence for a Cytogenetic Evolution of Type 2 from Type 1 Tumors Cancer Res., October 1, 2003; 63(19): 6200 - 6205. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Schraml, A. Hergovitz, F. Hatz, M. B. Amin, S. D. Lim, W. Krek, M. J. Mihatsch, and H. Moch Relevance of Nuclear and Cytoplasmic von Hippel Lindau Protein Expression for Renal Carcinoma Progression Am. J. Pathol., September 1, 2003; 163(3): 1013 - 1020. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
