| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Regular Articles |





¶
||**

From the Department of Pathology,*
Brigham
and Womens Hospital, Boston, Massachusetts; the Department of
Pathology,¶
Childrens Hospital, Boston,
Massachusetts; the Departments of Pediatric and Adult
Oncology,||
Dana-Farber Cancer Institute, Boston,
Massachusetts; the Division of
Hematology-Oncology,**
Childrens Hospital,
Boston, Massachusetts; the Departments of
Pathology,
and
Pediatrics,

Harvard Medical
School, Boston, Massachusetts; the Department of Clinical
Genetics,
University Hospital, Lund, Sweden;
and the Department of Pathology,
University
Hospital, Leuven, Belgium
| Abstract |
|---|
|
|
|---|
| Introduction |
|---|
|
|
|---|
| Materials and Methods |
|---|
|
|
|---|
Cytogenetic and molecular genetic data on cases 1 to 5 and 8 have
been previously published.5-7
Clinical and
histopathological data for all cases are found in Table 1
.
|
Metaphase harvest and slide preparation from tumor cell cultures
were as described previously,8
with minor modifications. G
banding was obtained by trypsin-Giemsa staining. For fluorescence
in situ hybridization, chromosome preparations were
pretreated in 2x standard saline citrate at 37°C for 1 hour,
followed by 5 minutes of digestion with pepsin using 1:25 diluted
Digest-All III (Zymed Laboratories, South San Francisco, CA).
Biotinylated and digoxigenin-labeled
satellite probes for the
centromeric regions of chromosomes 4, 8, 11, and 12 were from Oncor
(Gaithersburg, MD). The yeast artificial chromosome probes 935e9 and
946b7, flanking PLAG1, were obtained from Center dEtude
Polymorphisme Humain, Paris, France. Yeast artificial chromosome DNA
was isolated as previously described8
and labeled by
either digoxigenin or biotin by random octamer priming, using the
BioPrime DNA labeling system (Life Technologies Inc., Rockville, MD).
After labeling, DNA was purified by S-200HR spin column chromatography
(Pharmacia, Uppsala, Sweden), co-precipitated with 5x
Cot-1, and resuspended in hybridization buffer (50% formamide, 10%
dextran sulfate, 2x standard saline citrate). Probes and chromosome
preparations were co-denatured at 75°C for 2 minutes and
hybridized overnight. Biotin- and digoxigenin-labeled probes were
detected by strepatvidin-Alexa-498 (Molecular Probes, Eugene, OR) and
fluorescein anti-digoxigenin (Roche Diagnostics, Indianapolis, IN),
respectively. Chromosomes were counterstained using
diamidinophenylindole (Vector Laboratories, Burlingame, CA).
Genetic Analyses of Tissue Sections
Chromogenic in situ hybridization (CISH) was performed
on 4-µm formalin-fixed, paraffin-embedded, tissue sections on
positively charged slides (SuperFrost Plus; Menzel-Gläser,
Germany). The slides were baked overnight at 65°C and then
deparaffinized in xylene for 3 x 15 minutes, followed by washing
in 100% ethanol for 2 x 2 minutes. Air-dried slides were treated
in 100 mmol/L Tris/50 mmol/L ethylenediaminetetraacetic acid (pH 7.0)
in a microwave oven at 99°C for 15 minutes, washed briefly in
phosphate-buffered saline (PBS), and digested with Digest-All III
(Zymed Laboratories) for 1 to 7 minutes. The duration of the pepsin
digestion was optimized for each case to obtain sufficient signal
intensity without loss of tissue architecture. After washing in PBS,
the slides were immersed in 10% formalin/PBS for 1 minute and
dehydrated for 4 x 2 minutes in a 70 to 85% to 95 to 100% ethanol
series. Probes for CISH were labeled with digoxigenin and fluorescein
(New England Nuclear, Boston, MA), by random octamer priming, as
described above. Tissue sections and probes were co-denatured at 94°C
for 3 minutes and hybridized overnight at 37°C. Probe detection was
with alkaline phosphatase anti-digoxigenin (Roche Molecular
Biochemicals, Indianapolis, IN) and Fast Red substrate, and with
horseradish peroxidase anti-fluorescein and
H2O2-diaminobenzidine
(Zymed Laboratories). Tissue sections were counterstained with
hematoxylin and mounted with Glycerogel (DAKO, Carpinteria, CA). At
least 50 cells (mean, 150) were analyzed from each tumor. Tumors with
>10 cells showing one red-brown intact signal, one separate red, and
one separate brown signal at yeast artificial chromosome probe
hybridization were scored as positive for PLAG1
rearrangements. Tumors with >10 cells showing >2 signals for
centromere 8 and
2 signals with a centromeric probe for another
chromosome were scored as polysomic for chromosome 8.
| Results |
|---|
|
|
|---|
Review of the histology slides from the 16 lipoblastomas showed
that all tumors were characterized by lobules of mature and immature
fat cells including lipoblasts of the monovacuolated and
multivacuolated variety. The fibrous septae varied in thickness and
consisted of delicate and/or wiry collagen with interspersed spindle,
fibroblast-like cells, some of which had plump nuclei and abundant
cytoplasm and sometimes very small cytoplasmic vacuoles. There was
substantial difference in the histological appearance of
some tumors, prompting the classification described below. The classic
lipoblastomas (n = 11; Table 1
) usually had a
minimal myxoid componentlimited to rare microscopic foci in some
tumorsconsisting of intercellular mucin, spindle cells, and stellate
primitive mesenchymal cells, which blended with the adipocytic
component (Figure 1A)
. The myxoid
lipoblastomas (n = 2) contained abundant
interstitial mucin that comprised >50% of the specimen (Figure 1B)
.
The lipoma-like lipoblastomas (n = 2) lacked a
myxoid component and were composed predominantly of mature adipocytes
with scattered monovacuolated and multivacuolated lipoblasts (Figure 1C)
. The last tumor also lacked a myxoid component and was composed
predominantly of multivacuolated lipoblasts, some of which had central
nuclei and granular eosinophilic cytoplasm. This tumor was designated a
"lipoblastoma with hibernoma-like features" (Figure 1D)
. All tumors
were categorized using the above-mentioned descriptors (Table 1)
.
|
G-banding analyses revealed rearrangement of the 8q12 chromosome
region in 9 of 16 lipoblastomas, and a ring chromosome 8 with
unassignable cytogenetic breakpoints in the chromosome 8 long arm in
another tumor (Table 2)
. PLAG1
abnormalities were evaluated in all 16 tumors by metaphase cell
fluorescence in situ hybridization (12 cases) and/or
paraffin-section interphase cell CISH (10 cases). CISH was used for the
paraffin sections because lipid-related autofluorescence interfered
with fluorescence in situ hybridization detection methods,
particularly when attempting to evaluate the adipocytic populations in
the sections (data not shown). After optimization of the enzymatic
digestion procedure, the paraffin sections from each of the 10
lipoblastomas could be analyzed with a simultaneous appreciation of
tissue architecture, cellular morphology, and chromosomal
abnormalities. The PLAG1 in situ hybridization
analyses confirmed rearrangement of the PLAG1 region (Figure 2, a and b)
in all lipoblastomas with
cytogenetic 8q rearrangement and in one lipoblastoma that had been
diploid cytogenetically. In addition, the in situ
hybridization analyses revealed three to six copies of chromosome 8 in
three lipoblastomas, including one (case 13) that had pentasomy 8 by G
banding (Figure 3)
. Neither chromosome
banding nor molecular cytogenetic analyses showed evidence of a general
increase in DNA ploidy in the polysomy 8 lipoblastomas. Not only was
the karyotype indicative of an isolated chromosome 8 polysomy in case
13, but there were no polysomies for four comparison chromosomes (1, 4,
11, and 12, by
-satellite probes) in any of the polysomy 8 tumors
(Figure 2c)
. None of the cells in the polysomy 8 tumors had molecular
cytogenetic evidence of PLAG1 rearrangement.
|
|
|
| Discussion |
|---|
|
|
|---|
In the present study, we found PLAG1 rearrangements in 11 of
16 lipoblastomas, and we also showed that the PLAG1
nonrearranged tumors may have novel oncogenic mechanisms
involving PLAG1 copy number gains. These data
suggest an either/or relationship between PLAG1
rearrangement and PLAG1 amplification in lipoblastoma
tumorigenesis. An analogous situation has been observed in uterine
leiomyomas, where
25% (80 of 325) of tumors contain rearrangements
of the HMGIC region in 12q15, whereas another 10% (27 of
325) are trisomic for chromosome 12 (Mitelman F, Johansson B, Mertens
F: Mitelman Database of Chromosome Aberrations in Cancer.
http://cgap.nci.nih.gov/Chromosomes/Mitelman. 2001). Although
PLAG1 expression has not been compared in lipoblastomas
containing PLAG1 rearrangements
versus amplifications, we hypothesize that
PLAG1 transcriptional up-regulation will be a consistent
feature in both groups. Moreover, it is possible that the low-level
amplified PLAG1 loci are not wild type. Rather, these loci
might contain point mutations within binding sites of PLAG1
transcriptional repressors, which would thereby contribute to
PLAG1 transcriptional up-regulation. There is precedent for
promoter-region point mutations in association with nonfusion
MYC and BCL6 oncogenes in Burkitt
lymphomas22
and in follicular center cell
lymphomas,23
respectively. However, it is yet known
whether analogous genomic mechanisms contribute to transcriptional
up-regulation in nonlymphoid tumors.
Most lipoblastomas contain primitive mesenchymal cells and adipocytic cells at various stages of maturation.1 Typically, these cells are arranged in lobules showing a pattern of gradual differentiation, with the immature mesenchymal cells and multivacuolated lipoblasts at the periphery and the monovacuolated lipoblasts and adipocytes toward the center. This organization recapitulates the architecture of embryonal fat and it has been shown that some lipoblastomas eventually evolve into mature fat, presumably by terminal differentiation.3,23-26 If this is the case, one would anticipate a lipoma-like morphology in some lipoblastomas. Indeed, in the present study, two lipoma-like lipoblastomas were shown to contain PLAG1 rearrangements. However, it is unlikely that any substantial number of adult lipomas result from lipoblastoma maturation. Adult lipomas are vastly more common than lipoblastoma, and they are characterized by a cytogenetic profile very different from that in lipoblastoma. Previous cytogenetic data have shown that only 1% (2 of 216) of lipomas from patients aged 15 years or older contain 8q11-13 rearrangements (Mitelman F, Johansson B, Mertens F: Mitelman Database of Chromosome Aberrations in Cancer. http://cgap.nci.nih.gov/Chromosomes/Mitelman. 2001). The frequency of trisomy 8 in these tumors is 2% (4 of 216). This paucity of lipoma PLAG1-region rearrangements is striking, particularly becauseas discussed belowwe demonstrated PLAG1 rearrangements in the lipoma components of lipoblastomas. It follows that lipoblastomas and lipomas generally arise by distinct tumorigenic mechanisms, and possibly from different adipocytic progenitor cells.
CISH analysis of paraffin sections in six lipoblastomas revealed that the PLAG1 alterations were present in lipoblasts, mature adipocytes, stellate primitive mesenchymal cells, and fibroblast-like cells. The presence of PLAG1 rearrangements in variably differentiated mesenchymal cells suggests an origin from a transformed mesenchymal progenitor that proliferates and differentiates in a manner similar to white fat. Because lipoblastomas occur in children primarily younger than 10 years of age, we expect that such progenitors are only present, or only susceptible to PLAG1-mediated neoplastic transformation, during the first few years of life. Such a window of vulnerability in cellular development has been suggested in connection with other neoplasms, in particular hematological malignancies. However, the difficulties in determining histological origin of tumor cell populations have left this hypothesis primarily unsubstantiated in solid neoplasms.27 Lipoblastomas provide a compelling example of such a scenario, and in contrast to leukemias, in which expression of fusion transcription factor oncogenes can block cell maturation,28 lipoblastomas follow a remarkably well-ordered sequence of differentiation and remain benign. Much might be learned by characterizing the biological impact of abnormally expressed developmental factors, such as PLAG1, in benign tumors. In particular, it would be important to determine whether the benign nature of lipoblastomas is dictated by cell context (the specific progenitor cell) or whether this is a general aspect of PLAG1-mediated transformation.
| Footnotes |
|---|
Supported by the Blanceflor Boncompagni-Ludovisi, née Bildt Foundation, the John and Augusta Persson Foundation for Scientific Medical Research, the Swedish Medical Society, and a Lund Family American Cancer Society Grant.
Accepted for publication May 25, 2001.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
N. P. Agaram, M. P. Laquaglia, B. Ustun, T. Guo, G. C. Wong, N. D. Socci, R. G. Maki, R. P. DeMatteo, P. Besmer, and C. R. Antonescu Molecular Characterization of Pediatric Gastrointestinal Stromal Tumors Clin. Cancer Res., May 15, 2008; 14(10): 3204 - 3215. [Abstract] [Full Text] [PDF] |
||||
![]() |
G. Zheng and Y.-C. Yang Sumoylation and Acetylation Play Opposite Roles in the Transactivation of PLAG1 and PLAGL2 J. Biol. Chem., December 9, 2005; 280(49): 40773 - 40781. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. F. Landrette, Y.-H. Kuo, K. Hensen, S. B. van Waalwijk van Doorn-Khosrovani, P. N. Perrat, W. J. M. Van de Ven, R. Delwel, and L. H. Castilla Plag1 and Plagl2 are oncogenes that induce acute myeloid leukemia in cooperation with Cbfb-MYH11 Blood, April 1, 2005; 105(7): 2900 - 2907. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. B. Jones, J. A. Morcuende, B. R. DeYoung, G. Y. El-Khoury, J. A. Buckwalter, and F. R. Dietz Unusual Presentation of Lipoblastoma as a Skin Dimple of the Thigh. A Report of Three Cases J. Bone Joint Surg. Am., May 1, 2004; 86(5): 1040 - 1046. [Full Text] [PDF] |
||||
![]() |
C.-C. Chang and V. B. Shidham Molecular Genetics of Pediatric Soft Tissue Tumors: Clinical Application J. Mol. Diagn., August 1, 2003; 5(3): 143 - 154. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |