| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Regular Articles |
From the Centre de Biologie Cellulaire,*
Unité Dynamique des Cellules Embryonnaires et Cancéreuses,
and Laboratoire d'Ecologie Numérique,
Université des Sciences et Technologies de Lille, Villeneuve
d'Ascq, and the Laboratoire d'Anatomie et de Cytologie
Pathologique,
Centre Oscar Lambret, and
Régulation des Processus Invasifs, de l'Angiogenèse et de
l'Apoptose§
and Immunopathologie Cellulaire
des Maladies Infectueuses,¶
Institut de
Biologie Moléculaire, Institut Pasteur, Lille, France; and
Faculté des Sciences Jean Perrin,||
Universitré d'Antois, Lens, France
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The H19 gene codes for a capped, spliced, and polyadenylated RNA. It is highly conserved in vertebrates, as homologous sequences have been detected in rabbit,10 mouse,1 chicken, monkey, and human.4,11 The protein-coding potential of H19 RNA remains uncertain, and it has been proposed that this gene may act as an RNA.11 However, introduction of deletions or point mutations into the 5'-untranslated region (5'UTR) of an ectopic H19 gene, upstream of the largest open reading frame (ORF6), enabled the production of a 26-kd protein,12 although this has not been detected in cells expressing an endogenous H19 gene.
The H19 gene is located at 11p15.5 and is imprinted with only the maternal allele being expressed.9,13 H19 maps closely to another imprinted gene, IGF-II, but in the latter case it is the paternal allele that is transcribed. It has been reported that loss of heterozygosity (LOH) of a specific parental allele could be associated with the activation of a gene in cancers,14 and LOH of 11p15 was found in a wide variety of tumors, including some Wilms' tumor15-17 and lung,18 liver,19 ovarian, and breast cancers.20,21 Loss of imprinting of IGF-II has been described in a subset of Wilms' tumors. One hallmark of Wilms' tumors is the high levels of expression of the IGF-II gene, which has generated suggestions that an overdosage of the product of this gene contributes to Wilms' tumorigenesis.22 In some Wilms' tumors (approximately one-third) the transcriptionally silent maternal IGF-II allele is activated such that IGF-II expression occurs biallelically.23,24 There is evidence (enhancer deletion) that sequences flanking the H19 gene in the mouse control the nearby IGF-II gene in cis.25,26 In the majority of Wilms' tumors the silencing of H19 has been reported.27-31 This transcriptional silencing was accompanied by DNA methylation of the maternal H19 allele and activation of the maternal IGF-II allele.27-29 Loss of imprinting of H19 and/or IGF-II has been described in various cancers, including lung carcinomas,32 rhabdomyosarcoma,33-35 hepatoblastoma,24,36 testicular germ cell tumors,37 bladder carcinomas,38 uterine cervix carcinomas,39 and esophageal cancers.40 On the contrary, in some tumors, maintenance of normal imprinting of the H19 and/or IGF-II genes were observed (colorectal,40 neuroblastoma,41 glioma,42 leiomyomata,43 and breast).44
H19 is overexpressed in a wide variety of cancers (breast,4,6 head and neck,4,39 papillary and follicular thyroid,4 uterine cervix,4,39 bladder,45,46 adrenal tumor,47 trophoblast,48 lung,4,32 and esophageal).40
To date, the actual function of the H19 gene in cancer is still a matter of debate. Hao et al49 demonstrated that introduction of an H19 cDNA construct into G401 cells or RD rhabdomyosarcoma cells (two embryonal tumor cell lines) caused morphological changes and growth retardation. These investigators also reported that one H19-transfected G401 clone no longer formed tumors when injected into nude mice and that many clones had reduced growth in soft agar. These results made the H19 gene a good candidate to be a tumor suppressor gene. This function attributed to H19 was supported by several well documented works demonstrating the silencing of the H19 gene in several Wilms' tumors.27,28 However, Reid et al50 reported that H19 expression did not correlate with tumor suppression in their G401 cells (only two of the five nontumorigenic lines expressed H19). Otherwise, Cooper et al46 demonstrated that H19 is an oncodevelopmental marker during bladder tumor progression. Ariel et al51 examined the expression of H19 in tumor arising from tissues that express this gene in fetal life, and Verkerk et al52 reported the expression pattern of H19 in testicular germ cell tumors of adolescents and adults. These studies bring evidence that H19 is not a tumor suppressor gene, and their authors proposed that its product is an oncofetal RNA. Recently, Lustig-Yariv et al53 evaluated the level of H19 expression in choriocarcinoma cell lines (JAr and JEG-3 cells) and in tumors formed by these cells after their injection into athymic nude mice; they concluded that their data assigned to the H19 gene a role in contradiction with the tumor suppressor function proposed by others. Consequently, the role of H19 is still enigmatic, and the question of the properties of the H19 product, so far an RNA, remains open.
Other studies suggested that another locus on the short arm of chromosome 11 might be involved in tumor suppression, and the likely candidate is the cyclin-dependent kinase inhibitor, the p57KIP2 gene, in band 11p15.5, which causes G1 arrest.54-58
It has been frequently demonstrated that the H19 gene is
up-regulated in vitro in differentiating cells as well as
during growth arrest.1,7,59-61
A number of growth factors,
such as insulin-like growth factor (IGF)-I and -II, epidermal growth
factor (EGF), insulin, tumor necrosis factor (TNF)-
, interferon
(IFN)-
, and transforming growth factor (TGF)-ß1, and activators or
inhibitors of protein kinase A and C modulated the H19 gene
expression level in different cell lines: vascular smooth muscle
cells,10
fetal adrenal cells,62
and cultured
adrenal cells.47
Otherwise, Leibovitch et al63
reported that the overexpression of c-mos protein in the muscle cell
line C2C12 induces a concomitant increase of H19 RNA
expression, suggesting an interrelationship between these two gene
products during muscle differentiation.
The mammary gland is a unique organ in that most of its growth, morphogenesis, and differentiation occur in the adult. During these periods, interactions between epithelial and mesenchymal cells and hormones and growth factors contribute to its development. Disorders of these interactions can result in a tumorigenic process.64,65
Observations of the H19 gene expression in normal breast and its overexpression in many tumors,4,6,44 despite the possible maintenance of genomic imprinting,44 suggest that this gene is involved in both normal organogenesis and pathological events of the mammary gland.
We indicated, in a preliminary study of the expression of H19 gene by in situ hybridization (ISH) in 13 adenocarcinomas,6 that H19 transcripts accumulate essentially within the stromal compartment of the mammary gland. The aim of the present work was 1) to extend the previous study on the expression of the H19 gene to a statistically useful number of breast cancers (102), 2) to determine the level of H19 gene expression in various subclasses of adenocarcinomas, including some which are rare, 3) to delineate the precise localization of the H19 RNA, by using antibodies raised against specific stromal or epithelial components, 4) to establish the prognostic value of the H19 RNA (localization and intensity of the H19 signal were examined, and their relationships with histological grading system and various clinicopathological information were discussed), and 5) to correlate H19 expression with molecular markers of growth activity of the tumor: steroid receptor content, Ki-67/MIB-1 antigen presence, and overexpression (abnormal) of p53 gene product, which appears to be a common event in primary mammary carcinomas.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Breast carcinoma specimens were obtained from 102 patients (100 females and 2 males) treated by mastectomy at the Center Oscar Lambret (Lille, France) in 1990 and were selected on the basis of the first and unilateral cancer. In case of fatal issue, it was necessary to be sure that the cancer was only the more probable cause of death. For each tissue sample, the following clinicopathological information was obtained: histological subclassification (invasive ductal, invasive lobular, sarcomatoid, epidermoid metaplastic, tubular, colloid mucillaneous, papillary, apocrine, intraductal, or lobular in situ), tumor (T) values (from the tumor/nodes/metastases (TNM) classification of the UICC, ranging from T0 to T4), histological grade according to Bloom and Richardson,66,67 the axillary lymph node status, and the hormone receptor status (estrogen receptor (ER) and progesterone receptor (PR)), determined in femtomoles of receptors per milligram of cytosolic proteins and considered positive above 15 fmol/mg. The presence of these receptors indicates the level of sensitivity of tumor cells to these hormones. Furthermore, the age and menopausal status of patients were known, and expectation of life was followed up until the end of 1995.
Fixation, Embedding, and Histological Staining
Immediately after resection, material was fixed with formalin (10%) for 24 hours and then dehydrated through increasing ethanol concentrations and embedded into Paraplast Plus. Five-micron sections were transferred to slides coated with 3-aminopropyl-triethoxysilane (TESPA, Aldrich) for immunohistochemical staining (IH) and to Esco Superfrost Plus (Polylabo) for in situ hybridization (ISH). For IH, tumor sections were fixed on slide by a glycerinated albumin (10%) solution.
Hemalun-phloxine-safran (HPS) staining was performed on one section of each tumor. This section indicated histological structures, and frequently this staining demonstrated heterogeneity of tumors. HPS allowed us to localize precisely the more interesting areas to be observed after various IH procedures or the ISH. As control, we analyzed normal healthy tissues from cosmetic surgery; resections originated from mature breasts of two premenopausal women.
Immunohistochemical Staining
Sections were treated with xylene to remove paraffin from tissues, which were then progressively rehydrated. Sections were preliminarily treated by a modified procedure of Balaton et al68 to restore antigen specificity before immunostaining; slides were immersed for 7.5 minutes in citrate buffer (0.01 mol/L, pH 6) heated in a pressure-cooker, and the latter was then placed for 15 minutes under cold water.
To determine precisely which cells expressed the H19 gene,
four immunostaining reactions were performed in parallel: 1) monoclonal
antibody named anti-KL1, anti-human cytokeratin specific for epithelial
cells (1:200 dilution; Immunotech, Marseille, France), 2) monoclonal
antibody anti-smooth-muscle--actin to define myoepithelial cells
(1:2000 dilution; Sigma Chemical Co., St. Louis, MO), 3)
anti-Ki-67/MIB-1 specific for a cell-cycle protein (prediluted;
Immunotech), 4) anti-p53 protein raised against the amino-terminal
amino acid sequence of both the wild-type and mutant versions of the
protein (DO-1, prediluted; Immunotech). Immunoreactions were visualized
with diaminobenzidine chromogen (Dako, Glostrup, Denmark), and sections
were post-stained with hemalun.
In Situ Hybridization
Riboprobes
pSP65 plasmids were recombined with a 1.3-kb StuI
fragment of H19 cDNA at a SmaI site. cDNA
fragments (5'
3' and 3'5') were downstream of the SP6 promoter.
Plasmids were linearized by HindIII digestion. Sense and
antisense riboprobes were synthesized in the presence of
[35S]CTP and reduced to an average 150-bp length before
use.
ISH Protocol
Basic experiments were those previously described by Quéva
et al.69
After hybridization, slides were dipped in the
NTB2 nuclear track emulsion (Kodak, Rochester, NY), heated at 45°C,
and exposed for 3 weeks. Autoradiographic revelation (D19 revelator)
and fixation (Unifix, Kodak) were performed at 12°C. A fluorescent
post-staining of the nuclei was carried on (Hoechst 33258,
Bisbenzimidine, Serva; 
= 340 nm). Coverslips were fixed by
Dako-glycergel (Sebia). Observations were performed through an
Olympus BH2 microscope.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We previously reported that in normal breast resections signal for
H19 RNA was localized within both the epithelial and
mesenchymal tissues.6
The mesenchymal compartment was
rather focally labeled. Our observations of two other normal breast
samples confirmed this initial report. However, it appears that the
H19 transcript abundance can vary from one sample to another
and even in different areas of the same section (Figure 1A)
.
|
enabled us to compare the high
difference of labeling with antisense H19 riboprobe between
normal (Figure 1A)
and tumorigenic breast tissues (Figure 1, C and F
;
Figure 2, C and F
). Based on this kind of representative observations,
each tumor was classified as overexpressing the H19 gene by
a clear high-level labeling with the probe. Determinations were
performed independently by two investigators (E. Adriaenssens and L.
Dumont). If ambiguity of classification arose, the tumor was considered
as exhibiting no more labeling than the control (normal breast). So, we
admit that the pattern of overexpression, defined in this manner, in a
given tumor, was objective. Frequently, in a given specimen we
found a heterogeneous pattern of focal H19 gene expression,
which could be explained by histological diversity and/or differences
in the grading of the carcinoma. Then, we focused our observations on
the characteristic area, which was chosen for the clinical typing of
the tumor. To delineate very precisely tumor compartments, we used two
antibodies, anti-cytokeratin named anti-KL1 and
anti-smooth-muscle--actin raised against differentiation molecules
specific for epithelial and myoepithelial cells, respectively. Only
3/102 tumors (2.9%) exhibited an exclusive but intense epithelial
labeling (Figure 1, B and C)
, but in a large majority of tumors, 94/102
(92.2%), H19 was expressed in the stromal component only
(Figure 1, D to F
and Figure 2, D to F
). H19 transcripts
were present in both compartments in 5/102 specimens (4.9%); the
epithelial signal was rather punctate (Figure 2C)
. Interestingly, an
intense signal at epithelial-stromal boundaries was located quite
frequently on epithelial or stromal cells (Figure 1, C and F
,
respectively). Whatever the classification of a tumor, only a fraction
of epithelial or stromal cells was stained. Table 1
indicates localization of
H19 RNA in ten histological subclasses of tumors.
|
|
reports information concerning
the three patterns of H19 gene expression (stromal,
epithelial, and both stromal and epithelial). Percentage of positivity,
ie, H19 RNA abundance, was indicated for the following
clinicopathological factors or parameters: T values (T0 to T4),
histological grade, estrogen and progesterone receptor status, age,
menopausal and lymph node status, and cancer evolution for 5 years
since tumorectomy. In the three specimens where H19 RNA was
exclusively observed in epithelial cells, one can notice that these
tumors were of grade III, cells were devoid of hormone receptors, and
patient death occurred within the 5 years after the surgery. This
number of cases is too small to give a statistical meaning to these
results, but it is intriguing that these 3 cases are a subset of the 21
deaths (21/102 cases) registered within this period. Indeed, we found
only 10 tumors of grade III, showing no hormone receptors, which
evolved fatally over the 5-year follow-up.
|
H19 Gene Overexpression and Various Clinicopathological Factors
Table 2
shows the abundance of the H19 RNA as a
function of various clinicopathological indications. Proportions were
compared by using a Fisher's exact probability test,70
and
the threshold P value of 0.05 was chosen. The H19
overexpression, as defined above, exhibits a very high correlation with
the T values (UICC classification; P = 1.3 x
10-5) but also with the presence of hormone receptors for
estrogen and progesterone (P = 0.0048 and
P = 0.0159, respectively). For other factors (age,
menopausal and lymph node status, histological grade, and evolution at
5 years), P values were too high to be significant; as a
consequence, the H19 gene overexpression cannot be
significantly correlated with these parameters.
H19 Gene Overexpression and Presence of a Cell Cycle Marker Protein, Ki-67/MIB-1
Proliferative-cell activity has been estimated by
immunohistochemical staining with Ki-67 antibody, and it has been shown
that MIB-1, a monoclonal antibody, can react with an epitope of the
Ki-67 protein in formalin-fixed, paraffin-embedded tissues processed by
microwave pretreatment.71
This protein is characterized by
an accumulation at the transition between G2 and M phases, and its
expression correlates with semiconservative DNA synthesis associated
with the proliferating cell nuclear antigen (PCNA) expression but not
with the DNA synthesis associated with DNA repair.72
Consequently, Ki-67 has a short half-life, and its concentration
decreases rapidly after the mitotic phase; thus, it is considered as an
accurate indicator of cell proliferation in histological
material.73
A majority of the studied tumors, 20/24
(83.3%), were mainly Ki-67/MIB-1 positive; but at the cellular level,
no co-localization could be established between the overexpression of
the H19 gene and the presence of this marker (Figure 2, B and C)
.
H19 Gene Overexpression and p53 Protein Accumulation
Ninety-five carcinomas were examined for the presence of p53
protein by using an immunohistochemical method; twenty-one cases
(23.3%) were positive. Accumulation of p53 protein was mainly inside
nuclei (Figure 2E)
, although a weak cytoplasmic signal could not be
excluded as described by Moll et al.74
The H19
and p53 labelings do not overlap (Figure 2, E and F)
. Anti-p53
immunoreaction was correlated with histological grade
(P = 0.0062) but not with a high H19
expression level (P = 0.4442) and clinical
information, including T values, T0 to T4 (P =
0.5732), estrogen (P = 0.3702) and progesterone
(P = 0.2512) receptors, age
(P = 0.4881), and menopausal
(P = 0.8383) and lymph node
(P = 0.7867) status. Correlation of p53 index
with the histological grade provided the following results: grades
I/II, P = 0.7571; grades I/III, P =
0.2001; and grades II/III, P = 0.0392.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
ISH detection of H19 RNA, associated with
immunohistochemical staining specific for epithelial or myoepithelial
cells, allowed us to precisely detect the pattern of expression of this
gene in tumors and, consequently, are complementary to previous results
obtained by Northern blot.4
H19 RNA was
preferentially located in stromal cells only (92.2% of cases),
whatever the considered histological subclasses of tumors (Table 1)
.
Adipocytes were also highly labeled. Our data establish, too, that
during breast tumorigenesis an overexpression of the H19
gene within epithelial cells was rare: in 4.9% of specimens in both
epithelial and stromal cells and in 2.9% of them in epithelial cells
only (Table 2)
.
Spanakis and Brouty-Boyé75 tested the hypothesis that predicts that the stroma also progresses along with the epithelium in a breast tumor. They asked what characteristics were likely to change in a permanent manner during tumor development, and they screened a large number of transcripts. They concluded that stromal cells from normal and pathological breast tissues present multiple irreversible differences in gene expression. During the desmoplasmic reaction, cells constructing the stroma originated mainly in fibroblasts and smooth muscle cells, the so-called myofibroblasts, which correspond to a significant percentage of cells present in breast tumors, as high as 45%.76,77 Thus, when tumorigenesis occurs, quantitative (cellular proliferation) and qualitative (disappearance of myoepithelial cells) modifications arise, and the stroma is transformed in a fibrous tissue.78 Several genes are expressed specifically in the stromal part of a breast cancer, ie, hepatocyte growth factor,79 urokinase plasminogen activator,80,81 thrombospondin-1,82 tissue factor,83 aromatase,84 and c-ets-1 transcription factor.80 Through reciprocal exchanges between epithelial and stromal cellular types, products of the latter genes interfere in the tumor growth when proliferation, angiogenesis, and invasion occur. The study of Singer et al85 is worth mentioning in this context; indeed, these authors reported a paracrine influence, mediated by soluble factors released by epithelial cells, which are able to increase expression of the IGF-II gene in stromal cells; IGF-II-expressing fibroblasts are selected specifically in the stroma of breast cancers by the malignant cells. As IGF-II and H19 genes are co-regulated by the same set of enhancers, although they are oppositely imprinted,9,13 we propose that the stromal H19 up-regulation described in the present work could be induced by paracrine factors involved in the activation of the IGF-II transcription within the same mesenchymal cells. Furthermore, the epithelium-mesenchyme interactions also play a key role in proliferation and differentiation mechanisms during normal breast development.80,86-89
Otherwise, the H19 RNA concentration observed at the
epithelium/mesenchyme boundary can reflect one issue of the
interactions or the dialogue between cancerous and stromal cells. The
conversion of fibroblast to myofibroblast is the consequence of
epithelial stimuli.90
The closer the cells are to the tumor
epithelial cells, the more they are stimulated. In another respect,
kinetics of H19 RNA paralleled the accumulation of
muscle-specific markers (-actin, MLC1/MLC3),63
and the
expression of the rabbit H19 homologue was found in
nonproliferative actin-positive cells.91
This could explain
the overexpression of the H19 gene at the epithelium/stroma
boundary in benign and malignant tumors (see also Ref. 6
). This pattern
of expression can be put in parallel with that one observed for
c-ets-1, uPA, and collagenase
expression detected in mesenchyme cells facing invasive epithelial
cells; these data suggest that epithelial cells send signals to
mesenchymal cells, which react by expressing these
genes.92,93
Owing the few number of cases (2.9%) where H19 transcripts were localized exclusively in epithelial cells, the question of the importance of this observation is posed. We can notice that this rare H19 expression pattern matches with the absence of hormone receptors and the death of patients within the 5 years after tumorectomy. These rare situations could be explained by the general deregulation of genes, which can be encountered in the advanced tumor phase of cells.
In the healthy breast, epithelial cells synthesize a basal level of H19 transcripts depending on the specimen and even the area within a given section (this study and Ref. 6 ); this indicates that in the majority of carcinomas (~92%), the tumor development is accompanied by the complete loss of H19 gene expression in cancer cells. This striking silencing of the H19 gene does not establish the final evidence of the tumor suppressor function of the gene, but our statement on H19 expression patterns fit well with this role proposed by several authors.17,27,28,30,49,94 The H19 product (a RNA) could be implicated in some differentiation (or proliferation arrest) mechanisms. Moreover, Leibovitch et al63 demonstrated that both the H19 gene and c-mos oncogene are involved in myogenic differentiation and even are necessary in the maintenance of this status. The fact that the H19 gene is expressed during embryogenesis and then turned off in almost all adult tissues, except in breast, heart, and skeletal muscles, could suggest a dual function: one in proliferative events and the other one in differentiation.
Table 2
shows that among several clinicopathological factors considered
in this study, only the T values classification (UICC) and the presence
of hormone receptors (ER and PR) gave a positive significant
correlation with H19 gene overexpression. Interestingly, the
T value is one of the three elements of the TNM classification usually
used to determine evolution and prognosis of the tumors. P
values indicate that this important clinicopathological factor is
highly correlated with H19 gene overexpression. No less
interesting is the positive significant correlation between
H19 overexpression and the presence of hormone receptors,
which can be put together with the established feature that the
estimate (in femtomoles) of these receptors indicates the level of
sensitivity of tumor cells to these hormones.95
Otherwise,
it is known that aromatase is involved in estradiol synthesis, and the
expression of the aromatase gene increases in fat tissue adjacent to
the tumor.96-98
Interestingly, H19 transcripts
were abundant in adipocytes, mostly in those located near the tumor.
The latter cells synthesize estrogens, particularly those located in
this area,99
and these hormones could account for
H19 overexpression, as it has been proposed that estrogen
could play a role in modulations of H19
expression.100
Consequently, variations of the estrogen
levels during the menstrual cycle could account for the observed
differences in H19 RNA abundance detected in various healthy
breast resections and eventually also in pathological tissues.
Now we have to discuss ISH data in parallel with information on two physiological properties of the cells overexpressing the H19 gene. Do these cells accumulate the p53 protein and are they in cycle? Relationships between p53 accumulation and pathological factors, such as the histological type and grade and the status of the ER and PR is still in dispute.101-106 The prognostic and predictive value of p53 overexpression in breast carcinomas appears weaker than hoped.107 Nevertheless, accumulation of p53 is usually associated with tumor grades and negative ER status.107 In our series of breast tumor resections, no positive correlation was provided by the comparison between p53 protein accumulation and H19 gene overexpression. Nevertheless, in another study we have demonstrated a down-regulation of the H19 promoter by the wild-type p53 protein, but not by one p53 mutant (the 143 Ala mutant).108 Discrepancies between our previous data and those reported in the present work can be explained by several outlines, not mutually exclusive. Our previous study108 was concerned with a cell line (HeLa cells) transiently or stably transfected with a p53 recombined vector, and it was focused on relationships between p53 protein and the H19 promoter and displayed the effect of an accurate p53 mutation, exhibiting a thermosensitive phenotype. On the contrary, in the present study we considered H19 gene expression in tissues originated from primary breast cancers, of which the causes are necessarily multifactorial. Furthermore, one must keep in mind that although any accumulation of p53 protein can be generally the consequence of a genetic or an epigenetic outcome, we have no indications that all of the p53 mutations induce necessarily an overexpression of the H19 gene. Otherwise, p53 protein was located exclusively in epithelial cells, and positive correlation between p53 accumulation and a mutation of the p53 gene in breast cancers varied from 62% to 92%.109 Finally, one must remember that H19 gene overexpression was anyway rare in epithelial cells.
We consider now ISH data and a feature specific of cells in cycle. A monoclonal antibody, Ki-67, has been used to demonstrate that cells are in cycle. Ki-67 identifies a nuclear nonhistone protein of 395 and 345 kd present in the nucleoli of proliferative interphase cells as well as the condensed chromatin in mitotic cells. On the contrary, cells in quiescent phase G0 lack this antigen.110,111 In this study, we used MIB-1, which is a monoclonal antibody raised against a recombinant part of the Ki-67 antigen.71 As for p53 protein detection, the MIB-1 immunostaining labeled frequently tumor cells, independently of any ISH signal specific of the H19 RNA equipment. Once more, we must remember that epithelial cells express rarely the H19 gene. Consequently, the H19 RNA seems to be not crucial in the maintenance of cells in cycle.
In conclusion, 1) H19 gene overexpression is significantly correlated to the T values (TNM classification) and the presence of hormone receptors, but with neither the p53 tumor suppressor gene product nor with a protein indicating that cells are in cycle, 2) the frequent (92.2% of adenocarcinomas) overexpression of the H19 gene in stroma could be one of the responses of mesenchymal cells to paracrine factors released by tumor epithelium (this is stressed by abundance of H19 transcripts in mesenchymal cells adjacent to epithelial tissue), 3) H19 RNA accumulates rarely in epithelial cells (7.8% of cases, but in 2.9% in malignant cells only); the general silencing of H19 in invasive cells is in agreement with considerations on which this gene has been proposed as a tumor suppressor candidate, and 4) the fold increase of a basal level of H19 gene expression in the normal breast during adulthood, as the loss of regulation inducing a frequent but complex overexpression pattern of this gene in carcinomas, seems a result of puzzling processes, reflecting the fundamental relationships between cells with different phenotype. It is unlikely that any simple mechanism will explain all of the changes of the H19 expression level that occur as the mammary gland differentiates, ages, or undergoes a neoplastic development.
|
Acknowledgements |
|---|
|
Footnotes |
|---|
Supported by grants from Association de la Recherche sur le Cancer (ARC, Villejuif), the Ligue Nationale de Lutte contre le Cancer (Paris), and the Pasteur Institute in Lille. J.J. Curgy holds grants from the Groupement des Entreprises Françaises dans la Lutte contre le Cancer (Fé-GEFLUC) and from the NORGINE PHARMA laboratories (Paris).
Accepted for publication July 24, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-fetoprotein under the control of the murine raf and Rif genes. Proc Natl Acad Sci USA 1984, 81:5523-5527
ffer HP: Maintenance of normal imprinting of H19 and Igf-2 genes in neuroblastoma. Cancer Res 1995, 55:3386-3388This article has been cited by other articles:
 |
|
 |
|
  N. Berteaux, N. Aptel, G. Cathala, C. Genton, J. Coll, A. Daccache, N. Spruyt, H. Hondermarck, T. Dugimont, J.-J. Curgy, et al. A Novel H19 Antisense RNA Overexpressed in Breast Cancer Contributes to Paternal IGF2 Expression Mol. Cell. Biol., November 15, 2008; 28(22): 6731 - 6745. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Adriaenssens, E. Vanhecke, P. Saule, A. Mougel, A. Page, R. Romon, V. Nurcombe, X. Le Bourhis, and H. Hondermarck Nerve Growth Factor Is a Potential Therapeutic Target in Breast Cancer Cancer Res., January 15, 2008; 68(2): 346 - 351. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Barsyte-Lovejoy, S. K. Lau, P. C. Boutros, F. Khosravi, I. Jurisica, I. L. Andrulis, M. S. Tsao, and L. Z. Penn The c-Myc Oncogene Directly Induces the H19 Noncoding RNA by Allele-Specific Binding to Potentiate Tumorigenesis. Cancer Res., May 15, 2006; 66(10): 5330 - 5337. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y Fellig, I Ariel, P Ohana, P Schachter, I Sinelnikov, T Birman, S Ayesh, T Schneider, N de Groot, A Czerniak, et al. H19 expression in hepatic metastases from a range of human carcinomas J. Clin. Pathol., October 1, 2005; 58(10): 1064 - 1068. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Berteaux, S. Lottin, D. Monte, S. Pinte, B. Quatannens, J. Coll, H. Hondermarck, J.-J. Curgy, T. Dugimont, and E. Adriaenssens H19 mRNA-like Noncoding RNA Promotes Breast Cancer Cell Proliferation through Positive Control by E2F1 J. Biol. Chem., August 19, 2005; 280(33): 29625 - 29636. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N Berteaux, S Lottin, E Adriaenssens, F Van Coppenolle, X Leroy, J Coll, T Dugimont, and J-J Curgy Hormonal regulation of H19 gene expression in prostate epithelial cells J. Endocrinol., October 1, 2004; 183(1): 69 - 78. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. R. Tessier, G. A. Doyle, B. A. Clark, H. C. Pitot, and J. Ross Mammary Tumor Induction in Transgenic Mice Expressing an RNA-Binding Protein Cancer Res., January 1, 2004; 64(1): 209 - 214. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Lottin, E. Adriaenssens, T. Dupressoir, N. Berteaux, C. Montpellier, J. Coll, T. Dugimont, and J. J. Curgy Overexpression of an ectopic H19 gene enhances the tumorigenic properties of breast cancer cells Carcinogenesis, November 1, 2002; 23(11): 1885 - 1895. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Elkin, I. Ariel, H.-Q. Miao, A. Nagler, M. Pines, N. de-Groot, A. Hochberg, and I. Vlodavsky Inhibition of Bladder Carcinoma Angiogenesis, Stromal Support, and Tumor Growth by Halofuginone Cancer Res., August 1, 1999; 59(16): 4111 - 4118. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
