Rearrangements at the 11p15 Locus and Overexpression of Insulin-Like Growth Factor-II Gene in Sporadic Adrenocortical Tumors*
CHRISTINE GICQUEL, XAVIER BERTAGNA, HÉLÈNE SCHNEID, MARIE FRANCILLARD-LEBLOND, JEAN-PIERRE LUTON, FRANÇOIS GIRARD, AND YVES LE BOUC
Laboratoire d’Explorations Fonctionnelles Endocriniennes, Hôpital Trousseau (C.G., H.S., F.G., Y.L.B.), 75012 Paris; and Clinique des Maladies Endocriniennes et Métaboliques, Hôpital Cochin (X.B., M.F .- L., J .- P.L.), 75014 Paris, France
ABSTRACT
Little is known about the pathophysiology of sporadic adrenocortical tumors in adults. Because loss of heterozygosity at the 11p15 locus has been described in childhood tumors, particularly, in adrenocortical tumors, associated with the Beckwith-Wiedemann syndrome and be- cause insulin-like growth factor-II (IGF-II) is a crucial regulator of fetal adrenal growth, we looked for structural analysis at the 11p15 locus and IGF-II gene expression in 23 sporadic adrenocortical adult tumors: 6 carcinomas (5 with Cushing’s syndrome and 1 nonsecreting) and 17 benign adenomas (13 with Cushing’s syndrome, 1 pure androgen secreting, and 3 nonsecreting).
Twenty-one patients were informative at the 11p15 locus, and six (four carcinomas and two adenomas) of them (28.5%) exhibited 11p15 structural abnormalities in tumor DNA (five, an uniparental disomy
and one, a mosaicism). In a single case that could be further studied, a paternal isodisomy was observed.
Very high IGF-II mRNA contents were detected in seven tumors (30%; 5 of the 6 carcinomas and 2 of the 17 adenomas). They were particularly found in tumors with uniparental disomy at the 11p15 locus. Overall, a strong correlation existed between IGF-II mRNA contents and DNA demethylation at the IGF-II locus.
These data show that genetic alterations involving the 11p15 locus were highly frequent in malignant tumors, but found only in rare adenomas. These results in combination with evidence for overexpres- sion of IGF-II from the 11p15.5 locus suggest that abnormalities in structure and/or expression of the IGF-II gene play a role as a late event of a multistep process of tumorigenesis. (J Clin Endocrinol Metab 78: 1444-1453, 1994)
I INSULIN -like growth factors (IGF-I and IGF-II) are poly- peptides involved in metabolism, growth, and cell differ- entiation. They are synthesized in a variety of tissues and have endocrine and auto/paracrine mechanisms of action depending on the original tissue (1, 2). Both peptides are normally produced in adrenocortical cells (3-5). IGF-I through its action on steroidogenesis enzymes maintains the differentiated functions of the cell (4, 6). The precise role of IGF-II in mature adrenocortical gland is less well known.
Adrenocortical tumors in human (7-13) have a low inci- dence and distribute almost evenly between the benign ad- enomas and the malignant carcinomas. Because there is no absolute clinical, biological, anatomical, or even histological criteria, in many cases the benign or malignant nature of a purely localized tumor cannot be strictly asserted.
The pathological mechanisms responsible for the growth and/or steroid overproduction of adrenocortical tumors re- main mysterious, except in some peculiar conditions such as the Beckwith-Wiedemann syndrome (BWS) (14), the McCune-Albright syndrome (15), and the Li-Fraumeni syn-
Address requests for reprints to: Dr. Christine Gicquel, Laboratoire d’Explorations Fonctionnelles Endocriniennes, Hôpital Trousseau, 26 Avenue Arnold Netter, 75012 Paris, France.
* This work was supported in part by Assistance Publique Hopitaux de Paris (Contrat de Recherche Clinique 913104), in part by the Univer- sity Paris VI, Faculté Saint-Antoine, and in part by INSERM (Contrat de Recherche Externe 920709 and Réseau de Recherche Clinique “COMETE*).
drome (16), where different molecular defects have been identified.
IGF-II overexpression is a common feature of many tumors (17-20), including adrenocortical tumors (21). Specific ab- normalities of the 11p15 (where the IGF-II gene maps) have been described in embryonic tumors (sporadic, familial, or associated with BWS) (20, 22-26). Moreover, the IGF-II gene is submitted to parental imprinting (27-29), and it has been suggested that abnormalities of genomic imprinting are in- volved in BWS (24, 25, 28-30). In adrenocortical tumors, allelic losses at the 11p15 locus have been found in adenomas (31) and carcinomas (32) associated with BWS as well as in familial carcinomas (33).
In contrast with these embryonic tumors, sporadic adult adrenocortical tumors have not been so well studied. We examined the 11p15 region and IGF-II expression in a series of such tumors classified as benign or malignant according to classical criteria. We showed that IGF-II overexpression was a common feature (30%), often associated with unipar- ental disomy at the 11p15 locus. These genetic events were frequent in malignant tumors, but found only in rare ade- nomas, which exhibited the largest volume and nuclear pleomorphism.
Subjects and Methods
Patients
Twenty-three patients, 16-64 yr old, were included in this study. The hormonal status was evaluated as previously described (12). The diag-
nosis of Cushing’s syndrome was made from both clinical features and biological evidence of glucocorticoid oversecretion (increased urinary free cortisol and abnormal response to the high dose dexamethasone test).
Staging of the tumor as a localized, regional, or metastatic disease was based on clinical data, radiological studies, and CT scanning and was corroborated by the findings at surgery and on pathological exam- ination (12). Biological and histological data are summarized in Table 1. The numbers for the patients were assigned chronologically.
Six patients were diagnosed as having adrenocortical carcinomas, five of them with Cushing’s syndrome. Two patients had metastases, one presented a local recurrence, one had kidney invasion, and two had localized disease but histological data suggestive of malignancy. All of them had large tumors, with weights ranging from 135-2310 g.
Seventeen patients were diagnosed as having adrenocortical adeno- mas, 14 were secreting [13 of them with Cushing’s syndrome and 1 (patient 15) with clinical and biochemical features of androgen overse- cretion]. In patient 20, who had only high blood pressure, the diagnosis of Cushing’s syndrome was assumed on the basis of an abnormal response to the high dose dexamethasone test with undetectable ACTH and corticotropic insufficiency after unilateral adrenalectomy. Tumor weights ranged from 9-34 g. Seven tumors (patients 6, 14, 15, 16, 18, 21, and 25) showed nuclear pleomorphism, and therefore, histological data were not completely unequivocal.
Tissue fragments, obtained after surgery, were frozen in liquid nitro- gen and stored at -80 C until DNA/RNA extraction. Control constitutive DNA was available for all patients from peripheral blood leukocytes or normal peritumoral adrenocortical tissue in two patients with adenomas (patients 10 and 11; Table 1).
Material and Methods
Isolation of DNA and RNA from blood and tissues
Leukocyte DNA was prepared as previously described (34). Tumor DNA was extracted simultaneously with RNA by the guanidium thio-
cyanate cesium method (20, 35). DNA was dialyzed overnight against 0.001 mol/L EDTA and 0.01 mol/L Tris-HC1, pH 7.5, and further treated as leukocyte DNA.
Spectrometry at 260 and 280 nm was used for DNA and RNA quantification. Nucleic acid integrity was checked by ethidium bromide staining.
Probes
The different human probes used in this study were the 4-kilobase (kb) HindIII fragment of the H-ras-1 probe (pRSVEyras) (36); the 2.7-kb BgIII-XhoI fragment of the insulin probe (37); the 892- to 2231-basepair (bp) fragment of the IGF-II exon 9 cDNA, which is part of the coding sequence for exon 9 localized after the first site of polyadenylation; the 663-bp EcoRI fragment of the IGF-II cDNA (38); the 5.5-kb EcoRI fragment from the D11S12 probe (pADJ762) (39); the 827-bp calcitonin cDNA probe (40); the 0.8-kb HincII fragment from the catalase probe (pCATint1) (41); the 667-bp EcoRI-BamHI fragment of the IGF-I cDNA (42); and the 1.1-kb glyceraldehyde 3-phosphate deshydrogenase (G3PDH) probe (Clontech, Palo Alto, CA), which was used for verifying loading on dot blots. DNA probes were labeled with [a-32P]deoxy-ATP as previously described (43).
Southern blot analysis
Ten micrograms of constitutive (leukocyte or normal adrenal) and tumoral DNA were digested with restriction endonucleases according to the manufacturer’s instructions. Enzymes were chosen on the basis of the polymorphic patterns yielded: TaqI for the H-ras-1 gene (36), Sacl and TaqI for the insulin gene (44), Avall and Sacl for the IGF-II gene (20), Taql and Mspl for the D11S12 probe (39), Taql for the calcitonin gene (45), and TaqI for the catalase gene (41).
The restriction fragment length polymorphisms detected with the different probes used are shown in Table 2. Electrophoresis and transfer of DNA fragments were performed as previously described (34).
| Patient no.ª | Age | Clinical data | Biological data | Histological data | |||||
|---|---|---|---|---|---|---|---|---|---|
| Clinical presentation | 1st symptom to diagnosis (months) | Tumor stage at diagnosis | UFC (nmol/day)& | ACTH (pmol/L)€ | Tumor wt (g) | Hemorrhage/ necrosis/ invasion | Nuclear pleomorphism | ||
| Carcinomas (n = 6) | |||||||||
| 1 | 20 | Cushing's S | 12 | Metastases | ND | <1 | 135 | Yes | No |
| 2 | 49 | Cushing's S | 6 | Regional | 257 | <1 | 1550 | Yes | Yes |
| 3 | 42 | Cushing's S | 24 | Metastases | 480 | ND | 1230 | Yes | Yes |
| 4 | 22 | Flank pain | 2 | Localized | 138 | 1.8 | 160 | Yes | Yes |
| 5 | 16 | Cushing's S | 36 | Recurrence | 7395 | <1 | 2010 | Yes | Yes |
| 24 | 34 | Cushing's S | 8 | Localized | 1140 | <1 | 2310 | Yes | Yes |
| Adenomas (n = 17) | |||||||||
| 6 | 29 | Cushing's S | 30 | Localized | 1245 | <1 | 17 | No | Yes |
| 7 | 62 | Cushing's S | 42 | Localized | 295 | <1 | 9 | No | No |
| 8 | 57 | Cushing's S | 48 | Localized | 2420 | <1 | 17 | No | No |
| 9 | 26 | Cushing's S | 24 | Localized | 745 | <1 | 10 | No | No |
| 10 | 39 | Cushing's S | 48 | Localized | 1357 | <1 | 15 | No | No |
| 11 | 36 | Cushing's S | 12 | Localized | 361 | <1 | 18 | No | No |
| 13 | 27 | Cushing's S | 36 | Localized | 268 | <1 | 12 | No | No |
| 14 | 53 | Cushing's S | 48 | Localized | 436 | <1 | 30 | No | Yes |
| 15 | 24 | Virilization | 84 | Localized | 108 | 8.8 | 34 | No | Yes |
| 16 | 35 | Cushing's S | 9 | Localized | 803 | 1 | 12 | No | Yes |
| 17 | 40 | Cushing's S | 66 | Localized | 411 | <1 | 15 | No | No |
| 18 | 39 | Cushing's S | 60 | Localized | ND | <1 | 12 | No | Yes |
| 19 | 48 | Cushing's S | 30 | Localized | 709 | <1 | 10 | No | No |
| 20 | 64 | HBP | 36 | Localized | 127 | <1 | 20 | No | No |
| 21 | 55 | Cushing's S | 24 | Localized | 728 | <1 | 20 | No | Yes |
| 23 | 52 | Incidentaloma | 18 | Localized | 70 | ND | 15 | No | No |
| 25 | 54 | Flank pain | 2 | Localized | 111 | 1.1 | 30 | Yes | Yes |
UFC, Urinary free cortisol; HBP, high blood pressure; ND, not done.
” The numbers for the patients were assigned chronologically.
Normal, 55-250.
· Normal, 8 ± 3.9.
| Probe | Localization | Restriction endonuclease | Fragment lengths (kb) | ||
|---|---|---|---|---|---|
| H-ras-1 | 11p15.5 | TaqI | 3.5/2.5ª | 2.3 | |
| IGF-II | 11p15.5 | AvaII | 1.6 | 1.1/0.9ª | 0.4 |
| SacI | 10/7.6ª | 2.6 | |||
| Insulin | 11p15.5 | SacI | 7.5/6ª | ||
| TaqI | 5.6/4.5ª | ||||
| D11S12 | 11p15.5 | TaqI | 8.3/3.2ª | 3.9 | |
| MspI | 2.1/1.7ª | 1.3 | 1.1 | ||
| 1.5/1.3ª | 1.3 | 1.1 | |||
| Calcitonin | 11p15.1 | TaqI | 8/6.5ª | 3 | 2.3 |
| Catalase | 11p13 | TaqI | 3.5/2.5ª | 1 | |
ª Restriction fragment length polymorphism.
Dot and Northern blot analyses
For dot blot analysis, RNA samples were dotted onto a Hybond-N membrane (Amersham International, Aylesbury, United Kingdom) using a Hybri-dot apparatus (BRL, Gaithersburg, MD). For Northern blot analysis, denatured total RNA was loaded onto 1.2% agarose-formal- dehyde gels, then transferred to Hybond-N membranes, as previously described (20). Normal control adrenal RNA was obtained from glands surgically removed during large nephrectomy for kidney cancer. Placen- tal RNA, human colon adenocarcinoma cell line (SW613) RNA, and human hepatoma cell line (HepG2) RNA provided samples with high IGF-II mRNA expression.
Quantification
Hybridization signals were quantified by densitometric analysis using a GS300 scanning densitometer and a GS370 one-dimensional electro- phoresis data system (Hoefer Scientific Instruments, San Francisco, CA).
IGF-I and IGF-II protein assay
Serum IGF-I and IGF-II levels were determined after separation by acidic gel filtration, using methods previously described (46).
Serum peptide levels were measured in nine patients with variable IGF-II mRNA expression in tumors: four with high IGF-II expression (patients 2, 4, 5, and 15) and five with low or normal IGF-II expression (patients 7, 10, 11, 13, and 16).
Results
Analyses at the 11p15 locus
Leukocyte and tumor DNA were analyzed in all 23 pa- tients. Six patients had malignant tumors (patients 1-5 and 24), and 17 had benign tumors (patients 6-11, 13-21, 23, and 25). Analyses at the 11p15 locus were performed using five DNA markers: H-ras-1, IGF-II, insulin, D11S12, and calcitonin. Twenty-one patients (6 of 6 carcinomas and 15 of 17 adenomas) were informative for 1 or more of these 11p15 markers (Table 3). Of the 21 informative cases, 6 exhibited loss of heterozygosity (LOH) in their tumors (patients 3, 4, 5, 24, 10, and 14; Table 3), whereas all had normal hetero- zygous profiles in their leukocyte DNA (Fig. 1).
For patients 3, 4, 5, 24, and 14, the tumor DNA profile showed highly disproportionate allele intensity, with an al- most complete loss of one allele and a double dose of the remaining allele, showing a uniparental disomy (Fig. 1). These abnormal profiles were found at the insulin locus for patient 3; at the ras, insulin, IGF-II, and calcitonin loci for patient 4; at the H-ras-1, D11S12, and calcitonin loci for
| Patient no. | H-ras-1, 11p15.5 | Insulin, 11p15.5 | IGF-II, 11p15.5 | D11S12, 11p15.5 | Calcitonin, 11p15.1 | Catalase, 11p1.3 | Structural rearrangement |
|---|---|---|---|---|---|---|---|
| Carcinomas | |||||||
| 1 | BB | AB | AB | BB | AB | AB | No |
| 2 | AB | BB | AA | BB | BB | AB | No |
| 3 | BB | B | AA | BB | BB | AA | UPD |
| 4 | A | A | A | BB | B | A | UPD |
| 5 | B | BB | AA | B | B | AA | UPD |
| 24 | AA | AA | A | BB | BB | AA | UPD |
| Adenomas | |||||||
| 6 | BB | BB | AA | AB | AB | AA | No |
| 7 | AB | AB | AA | BB | BB | AA | No |
| 8 | BB | BB | AA | AB | BB | AA | No |
| 9 | BB | BB | AA | BB | AB | AA | No |
| 10 | Ab | aC | Ab | BB | BB | AB | Mosaicism |
| 11 | BB | BB | AB | BB | BB | AA | No |
| 13 | BB | AB | AA | BB | AB | AA | No |
| 14 | A | BB | B | B | BB | B | UPD |
| 15 | AB | AA | AA | AB | BB | AA | No |
| 16 | AB | BB | AB | AB | AB | AA | No |
| 17 | AA | AA | AA | BB | BB | AA | NI |
| 18 | BB | BB | AA | BB | BB | AA | NI |
| 19 | BB | BB | AB | BB | BB | AA | No |
| 20 | BB | AB | AA | BB | BB | AB | No |
| 21 | AA | BB | AB | BB | BB | AB | No |
| 23 | AB | AA | AA | BB | AB | AB | No |
| 25 | AB | AB | AA | AB | AA | AA | No |
C, A third polymorphic allele; A or B, uniparental disomy, A or B being the duplicate remaining allele; Ab or aC, mosaicism with disproportionate profile; the less abundant allele is printed in lower case; UPD, uniparental disomy; NI, noninformative. Abnormal profiles are emphasized by boxes.
A
Avall / IGF II
B Taql / Insulin
pat nº4
pat nº10
pat nº14
pat nº1 pat nº3 pat nº4
L
T
L
NA Tco Tcl
L
T
kb
LTLTLT
kb
1.6
5.6
1.1
0.9
4.5
0.5
C Taql / H-ras 1 pat nº2 pat nº4 pat nº5
0.4
LTLTLT
kb
3.5
2.5
patient 5; at the IGF-II locus for patient 24; and at the H- ras-1, IGF-II, and D11S12 loci for patient 14 (Table 3). Densitometric quantification confirmed the duplication of the remaining allele in all tumors (Fig. 2; the 0.4-kb fragment was used as an internal control for the amount of DNA in the AvaII-IGF-II system). A weak hybridization signal was observed at the position of the disappearing allele, which varied between 6-18% of the overall signal of the two polymorphic bands and most likely corresponded to a slight contamination with normal stromal cell DNA in the tumor.
Parental analysis could be performed for patient 4; using the AvaII-IGF-II system, the maternal allele (0.9 kb) had been lost in the tumor, whereas two copies of the paternal allele (1.1 kb) were present (Fig. 3A). The calcitonin locus study indicated that the father was heterozygous and showed that two copies of the same paternal chromosome 11 were re- tained in the tumor, revealing a paternal isodisomy in the tumor (Fig. 3B).
For patient 10, normal heterozygous profile was found in leukocyte DNA as in normal adrenal DNA when tumoral DNA showed slightly disproportionate allele intensities; the 1.1-kb allele was more intense than the 0.9-kb allele (ratio, 61%/39%; Figs. 1 and 2). The same profile was found in two different parts of the same tumor selected for the predomi- nance of compact cells (T co) or clear cells (T cl). Results obtained at the IGF-II locus using the enzyme SacI and also at the H-ras1 and insulin loci provided the same patterns (data not shown).
For the 15 other informative patients at 11p15 locus, no LOH was found, including patients 1 and 2 who had carci- nomas (Fig. 1, B and C).
Overall, LOH was found in 4 of the 6 informative carci- nomas (patients 3, 4, 5, and 24) and 2 of the 15 informative adenomas (patients 10 and 14; Table 3).
Analyses at the 11p13 locus (Table 3)
The same DNAs were studied at the catalase locus (11p13). Among the six patients with uniparental disomy or mosai- cism at 11p15 loci, three (patients 4, 10, and 14) were informative at the catalase locus. Patients 4 and 14 exhibited the same uniparental disomy as described for the 11p15 loci. For patient 10, the mosaicism was not found at the 11p13 locus. For the five other informative patients (patients 1, 2, 20, 21, and 23), no LOH was detected.
Figure 4 summarizes the distribution of allelic losses along the short arm of chromosome 11. This figure emphasizes that when patients were informative, the IGF-II/insulin locus was always involved, and that allelic losses could extend to the 11p13 region (patients 4 and 14).
IGF-II mRNA expression in tumors
IGF-II mRNA expression was measured in all tumors by dot blot assay. The results were compared with those found in normal adult adrenocortical glands (Table 4) and in tissues and cells known to express high levels of IGF-II mRNA (human placenta and the SW613 and HepG2 cell lines). The cDNA probe G3PDH was used as control for RNA loading.
Very high levels of IGF-II mRNA expression (>100 times those in normal human adult adrenocortical tissues) were found in five of the six carcinomas (Table 4). Four of these five carcinomas were those with uniparental disomy at the 11p15.5 locus (patients 3, 4, 5, and 24), and the fifth was from a patient who was only informative for the H-ras-1 locus and who did not exhibit LOH at this locus (patient 2). Part of these results are shown in Fig. 5.
Increased levels of IGF-II mRNA were also found in 2 of the 17 adenomas (Table 4 and Fig. 5): in patient 14, who showed uniparental disomy, and in patient 15, who did not exhibit LOH at the 11p15 locus.
patient 3
leukocyte
1.1
patient 4
4.5
leukocyte
tumor
1.6
tumor
5.6
0.9
0.5
0.4
patient 10
0.9
patient 14
leukocyte
leukocyte
tumor
tumor
1.1
1.6
0.9
1.1
0.4
0.4
1.6
0.5
0.5
IGF-II mRNA expression was analyzed by Northern blot for tumoral and control adrenocortical RNA. The sizes of the messengers (4.8 kb transcribed from promoter P4, and 6 and 2.2 kb transcribed from promoter P3) were the same in tumoral and normal adrenals whatever the level of expres- sion (Fig. 6). The relative distribution of the mRNA species indicated a preferential use of promoter P3 (6- and particu- larly 2.2-kb mRNA species) compared to promoter P4 in the tumor tissues (Fig. 6B). The high hybridization signal around 2 kb is specific, corresponding to the 2.2-kb IGF-II mRNA. When Northern blots were hybridized with the 892- to 2231- bp fragment of IGF-II exon 9 cDNA (which specifically detects the 4.8- and 6-kb IGF-II mRNAs), the 2.2-kb signal decreased dramatically (Fig. 6B).
Methylation of the IGF-II gene
The restriction enzyme AvalI is highly sensitive to DNA methylation. The restriction site GC(A/T)CC at the 3’-end of exon 7 of the IGF-II gene is cut by AvalI only when it is demethylated, generating an additional 0.5-kb fragment with
a concomitant decrease in the 1.6-kb fragment (20). The methylation of this restriction site has been shown to be a tissue-specific phenomenon.
A DNA demethylation index at the IGF-II locus was calculated. The ratio of the 0.5-kb/0.5 + 1.6-kb AvalI frag- ments is an index of IGF-II DNA demethylation (20, 30). IGF-II demethylation was not detected in leukocyte DNA. In tumors, DNA demethylation ranged from 11-100%, and in normal adrenal tissues, DNA demethylation was 10 ± 4.5% (mean + SD; Table 4). A significant correlation was found between IGF-II mRNA expression and the IGF-II demethyl- ation index (r = 0.68; P < 0.001; Table 4). The highest indexes of IGF-II DNA demethylation were found in the five patients with uniparental disomy, with demethylation, re- spectively, of 100%, 84%, 100%, 69%, and 64% for patients 3, 4, 5, 24, and 14 (Table 4).
A typical demethylation pattern (using the IGF-II-Avall system) is shown in Figs. 1A and 2 for patients 4 and 14, with an important decrease in the amount of the 1.6-kb and the appearance of the 0.5-kb fragment. This demethylation
A
Ava II / IGF II
B
Taq I / calcitonin
kb
LLTL
LLTL
kb
1.6
8
6.5
1.1 0.9
0.5
0.4
process was gene specific and was not found for other genes, such as the IGF-I gene located at the 12q22-24.1 locus (data not shown).
Serum IGF-I and IGF-II protein levels
Systemic levels of IGF-I ranged from 108-352 ng/ml (mean ± SD, 300 ± 56 ng/ml). They were in the normal range, except in one subject (patient 5; 108 ng/ml). IGF-II levels ranged from 610-1478 ng/ml (mean ± SD, 1280 ± 265
ng/ml) and were in the normal range, except in one subject (patient 7; 610 ng/ml).
Discussion
Adrenocortical tumors in human adults are poorly char- acterized on biochemical, anatomical, and histological grounds, which establish their secretory activity and, with some uncertainty, their benign or malignant nature. Because IGF-II is normally produced in human adrenals, and abnor- malities have been described in congenital adrenocortical tumors at the 11p15 locus, we studied this chromosomal region and IGF-II gene expression in sporadic adult human adrenocortical tumors.
Our data show that genetic losses involving 11p segments are important determinants in the tumorigenesis of sporadic adrenocortical tumors in human adults. In 6 of 21 informative tumors, somatic DNA rearrangement was found on the short arm of chromosome 11 at the 11p15 locus. For 5 of these 6 tumors (patients 3, 4, 5, 14, and 24), the loss of 1 allele was associated with a duplication of the remaining allele.
In a single case for which a family analysis was possible (patient 4), the remaining allele was of paternal origin. Pref- erential retention of paternal allele at the 11p15 locus has been described in other tumors, such as Wilms’ tumors and embryonic rhabdomyosarcomas (20, 25, 47). In patient 4, whose father was heterozygous at the calcitonin locus, we showed that tumor DNA retained two copies of the same paternally derived chromosome 11, demonstrating in this case a paternal isodisomy. Patient 10 exhibited only slightly disproportionate intensities of the polymorphic bands. The fact that three different genes (H-ras-1, IGF-II, and insulin) had the same unbalanced ratios of their two alleles (39%/ 61%), whereas these ratios were 49%/51% in leukocyte and normal adrenal DNA, favored a mosaicism cell distribution; one clonal, with allelic loss at the 11p15.5 locus, and the other without it.
15.5
15.4
15.3
15.2
15.1
.13
H ras 1
Insulin
IGF II
D11S12
calcitonin
catalase
patients
3
??????????
????????? 777777777 777777777777777777
4
77777777}
5
7777777777777777?
?????????
14
777777777
77777777
24
77777777 ?? ?????????
????????? ??????????????????
10
777777777 77777777
| Patient no. | 11p structural rearrangement | IGF-II mRNA content" | IGF-II DNA demethylation indexb |
|---|---|---|---|
| Carcinomas | |||
| 1 | No | 0,4 | 19 |
| 2 | No | >100 | 50 |
| 3 | UPD | >100 | 100 |
| 4 | UPD | >100 | 84 |
| 5 | UPD | >100 | 100 |
| 24 | UPD | >100 | 69 |
| Adenomas | |||
| 6 | No | 0.5 | 18 |
| 7 | No | 1 | 15 |
| 8 | No | 2.1 | 25 |
| 9 | No | 3.2 | 13 |
| 10 | Mosaicism | 0.5 | 25 |
| 11 | No | 0.4 | 19 |
| 13 | No | 2.1 | 16 |
| 14 | UPD | >100 | 64 |
| 15 | No | >100 | 36 |
| 16 | No | 0.35 | 15 |
| 17 | NI | 0.9 | 11 |
| 18 | NI | 0.4 | 33 |
| 19 | No | 0.6 | 13 |
| 20 | No | 0.2 | 13 |
| 21 | No | 0.4 | 14 |
| 23 | No | 1.2 | 17 |
| 25 | No | 0.4 | 16 |
UPD, Uniparental disomy; NI, noninformative.
” IGF-II mRNA content is expressed on the basis of densitometric analysis relative to normal adult adrenocortical tissue taken as refer- ence (=1) and normalized with respective G3PDH values.
· Expressed as a ratio, as defined in Results; demethylation of normal adrenocortical tissue (mean + SD) = 10 ± 4.5% (n = 4).
These genetic losses are probably not random chromo- somal losses associated with the malignant genotype in gen- eral, because they were frequent. The structural re- arrangement of the IGF-II gene had a functional correlate. IGF-II gene overexpression was frequently found (30% of all the tumors and 83% of the carcinomas) and was constant in tumors with uniparental disomy (>100 times the level in normal adrenal).
The same IGF-II mRNAs species (4.8 kb and 6 and 2.2 kb, respectively, transcribed from promoters P4 and P3) (48) were found in tumors as in normal adrenal tissue, but with preferential use of promoter P3 (6 and particularly 2.2 kb) in tumoral tissues. mRNA degradation could not be completely excluded, although it is not supported by the large decrease in the 2.2-kb signal when Northern blots were hybridized with the 892- to 2231-bp fragment of IGF-II exon 9 cDNA.
High IGF-II overexpression might be caused by epigenetic modification of the IGF-II gene, such as tissue-specific meth- ylation. The strong correlation between IGF-II gene expres- sion and the demethylation level of this gene in the tumors supports this mechanism, which has been showed for other neoplasms (49). Demethylation was not found for genes other than IGF-II. This is different from other tumors, in which extensive demethylation of DNA correlates with tu- mor progression (50, 51).
The high prevalence of the IGF-II gene rearrangement and
the strong correlation between uniparental disomy and IGF- II overexpression strongly suggest that the 11p15.5 region is a major determinant for adrenocortical tumorigenesis. It has been suggested that IGF-II gene itself was the gene candidate (WT2) for BWS (52). Alternatively, because in BWS the absence of a maternal allele confers a higher risk of tumors than the sole duplication of the paternal allele (14, 24), because LOH at the 11 p15 locus is described in many tumors (53), and because transfer of 11p15.5-14.1 region suppresses tumorigenicity of nephroblastoma or rhabdomyosarcoma cell lines (54, 55), this region might contain a tumor suppressor gene. The H19 gene, which is highly expressed during de- velopment and also maps to the 11p15.5 region, has recip- rocal imprinting to the IGF-II gene, as its maternally derived allele is active (56) and could be this tumor-suppressor gene (57).
Some of our adrenocortical tumors (patients 2 and 15) had high IGF-II mRNA contents without LOH. IGF-II gene over- expression could be explained by relaxation of the IGF-II gene with biallelic expression of both maternal and paternal alleles. Loss of imprinting was found in 65-69% of Wilms” tumors not undergoing LOH (28, 29).
Whatever the mechanisms that lead to IGF-II gene over- expression (uniparental disomy, abnormal methylation pat- tern, or loss of imprinting), the causal relation between the high content of IGF-II mRNA and the growth of adrenocor- tical tumors remains to be established. Further studies should analyze the content of IGF-II peptides and the presence of IGF-II receptors, which should ultimately determine the bi- ological consequence of the genetic abnormalities that have been described. It is also likely that other molecular events are involved in the pathophysiology of adrenocortical tu- mors. In one carcinoma (patient 1), we found no evidence of abnormal IGF-II gene organization or expression. In a prior study, Yano et al. (58) systematically looked for allelic losses in six adrenocortical carcinomas and showed a constant LOH at the 17p locus, whereas LOH at 11p and 13q loci were found in only half of the malignant tumors. No gene abnor- mality was found in benign tumors or hyperplastic glands. Together with our data, these results suggest that 11p re- arrangement is a late event in adrenocortical tumorigenesis. These data are very close to those described for some em- bryonal tumors and demonstrate that congenital tumorigenic mechanisms could be common in sporadic adult tumors.
Finally, and importantly, the high prevalence of IGF-II gene overexpression in malignant adrenocortical tumors compared with that in their benign counterparts is striking. It indicates first that classical criteria to distinguish adreno- cortical tumors are evidently insufficient in many cases, and secondly that the systematic determination of IGF-II gene structure and expression in these tumors should become an essential prognostic factor.
Acknowledgments
We are greatly indebted to Dr. Jullienne for the gift of the calcitonin probe, to Dr. Holthuizen for the gift of the insulin probe and the 892- to 2231-bp fragment of IGF-II exon 9 cDNA, to Dr. Bos for the gift of
IGF II
G3PDH
IGF II
G3PDH
patients
patients
nº 13
nº 6
nº 1
nº 7
nº 9
nº 24
nº 15
nº 19
nº 25
nº 20
normal adrenal
nº 17
normal adrenal
placenta
placenta
Hep G2 cell line
Hep G2 cell line
ug RNA
8
4
2
1
0.5
4
4
NaOH
ug RNA
8
4
2
1
0.5
4
4
NaOH
A
normal adrenal
SW613 cell line
Hep G2 cell line
B
patients
placenta
Hep G2 cell line
patient 2
7
9
10
19
21
a
b
c
kb
-6 —
+4.8-
-2.2-+
15
5
5
5
2.5
ug RNA
1
5
3 h
48 h
14 h
exposure
1 h
3 h
the H-ras-1 probe, to Dr. Munnich for the gift of the D1IS12 probe, and to Dr. Mannens for the gift of catalase probe.
References
1. Daughaday WH, Rotwein P. 1989 Insulin-like growth factors I and II. Peptide, messenger ribonucleic acid and genes structures, serum and tissue concentrations. Endocr Rev. 10:68-91.
2. Humbel RE. 1990 Insulin-like growth factors I and II, a review. Eur J Biochem. 190:445-462.
3. Voutilainen R, Miller WL. 1987 Coordinate tropic hormone regu- lation of mRNAs for insulin-like growth factor II and the cholesterol side-chain-cleavage enzyme, P450scc, in human steroidogenic tis- sues. Proc Natl Acad Sci USA. 84:1590-1594.
4. Penhoat A, Chatelain PG, Jaillard C, Saez JM. 1988 Charateriza- tion of insulin-like growth factor I and insulin receptors on cultured bovine adrenal fasciculata cells. Role of these peptides on adrenal cell function. Endocrinology. 122:2518-2526.
5. Kamio T, Shigematsu K, Kawai K, Tsuchiyama H. 1991 Immu- noreactivity and receptor expression of insulinlike growth factor I and insulin in human adrenal tumors. An immunohistochemical study of 94 cases. Am J Pathol. 138:83-91.
6. Pham-Huu-Trung MT, Villette JM, Bogyo A, Duclos JM, Fiet J, Binoux M. 1991 Effects of insulin-like growth factor I (IGF-I) on enzymatic activity in human adrenocortical cells. Interactions with ACTH. J Steroid Biochem. 39:903-909.
7. Hutter AM, Kayhoe DE. 1966 Adrenal cortical carcinoma. Clinical features of 138 patients. Am J Med. 41:572-580.
8. King DR, Lack EE. 1979 Adrenal cortical carcinoma. A clinical and pathologic study of 49 cases. Cancer. 44:239-244.
9. Bertagna C, Orth DN. 1981 Clinical and laboratory findings and results of therapy in 58 patients with adrenocortical tumors admitted to a single medical center (1951 to 1978). Am J Med. 71:855-875.
10. Nader S, Hickey RC, Sellin RV, Samaan NA. 1983 Adrenal cortical carcinoma. A study of 77 cases. Cancer. 52:707-711.
11. Freeman DA. 1986 Steroid hormone-producing tumors in man. Endocr Rev. 7:204-220.
12. Luton JP, Cerdas S, Billaud L, et al. 1990 Clinical features of adrenocortical carcinoma, prognostic factors and the effect of mito- tane therapy. N Engl J Med. 322:1195-1201.
13. Ross NS, Aron DC. 1990 Hormonal evaluation of the patient with an incidentally discovered adrenal mass. N Engl J Med. 323:1401- 1405.
14. Junien C. 1992 Beckwith-Wiedemann syndrome, tumourigenesis and imprinting. Curr Opin Genet Dev. 2:431-438.
15. Weinstein LS, Shenker A, Gejman PV, Merino MJ, Friedman E, Spiegel AM. 1991 Activating mutations of the stimulatory G protein in the McCune-Albright syndrome. N Engl J Med. 325:1688-1695.
16. Malkin D, Li FP, Strong LC, et al. 1990 Germ line p 53 mutations in a familial syndrome of breast cancer, sarcomas, and other neo- plasms. Science. 250:1233-1238.
17. Scott J, Cowell J, Robertson ME, et al. 1985 Insulin-like growth factor-II gene expression in Wilms’ tumour and embryonic tissues. Nature. 317:260-262.
18. Reeve AE, Eccles MR, Wilkins RJ, Bell GI, Millow LJ. 1985 Expression of insulin-like growth factor-II transcripts in Wilms’ tumour. Nature. 317:258-260.
19. Gloudemans T, Prinsen I, Van Unnik JAM, Lips CJM, Den Otter W, Sussenbach JS. 1990 Insulin-like growth factor gene expression in human smooth muscle tumors. Cancer Res. 50:6689-6695.
20. Schneid H, Seurin D, Noguiez P, Le Bouc Y. 1992 Abnormalities of insulin-like growth factor (IGF-I and IGF-II) genes in human tumor tissue. Growth Regul. 2:45-54.
21. Ilvesmäki V, Kahri AI, Miettinen PJ, Voutilainen R. 1993 Insulin- like growth factors (IGFs) and their receptors in adrenal tumors: high IGF-II expression in functional adrenocortical carcinomas. J Clin Endocrinol Metab. 77:852-858.
22. Koufos A, Grundy P, Morgan K, et al. 1989 Familial Wiedemann- Beckwith syndrome and a second Wilms’ tumor locus both map to 11 p15.5. Am J Hum Genet. 44:711-719.
23. Ping AP, Reeve AE, Law DJ, Young MR, Boehnke M, Feinberg
AP. 1989 Genetic linkage of Beckwith-Wiedmann syndrome to 11 p15. Am J Hum Genet. 44:720-723.
24. Henry I, Puech A, Riesewijk A, et al. 1993 Somatic mosaicism for partial paternal isodisomy in Wiedemann-Beckwith syndrome: a post-fertilization event. Eur J Hum Genet. 1:19-29.
25. Grundy P, Telzerow P, Paterson MC, et al. 1991 Chromosome 11 uniparental isodisomy predisposing to embryonal neoplasms. Lan- cet. 338:1079-1080.
26. Schneid H, Vazquez MP, Seurin D, Le Bouc Y. 1991 Loss of heterozygosity in non-tumoral tissue in two children with Beckwith- Wiedemann syndrome. Growth Regul. 1:168-170.
27. De Chiara TM, Robertson EJ, Efstratiadis A. 1991 Parental im- printing of the mouse insulin-like growth factor II gene. Cell. 64:849-859.
28. Rainier S, Johnson LA, Dobry CJ, Ping AJ, Grundy PE, Feinberg AP. 1993 Relaxation of imprinted genes in human cancer. Nature. 362:747-749.
29. Ogawa O, Eccles MR, Szeto J, et al. 1993 Relaxation of insulin-like growth factor II gene imprinting implicated in Wilms’ tumour. Nature. 362:749-751.
30. Schneid H, Seurin D, Vasquez MP, Gourmelen M, Cabrol S, Lebouc Y. 1993 Parental allele specific methylation of the human insulin-like growth factor II gene and Beckwith-Wiedemann syn- drome. J Med Genet. 30:353-362.
31. Hayward NK, Little MH, Mortimer RH, Clouston WM, Smith PJ. 1988 Generation of homozygosity at the c-Ha-ras-1 locus on chro- mosome 11p in an adrenal adenoma from an adult with Wiede- mann-Beckwith syndrome. Cancer Genet Cytogenet. 30:127-132.
32. Henry I, Jeanpierre M, Couillin P, et al. 1989 Molecular definition of the 11p15.5 region involved in Beckwith-Wiedemann syndrome and probably in predisposition to adrenocortical carcinoma. Hum Genet. 81:273-277.
33. Henry I, Grandjouan S, Couillin P, et al. 1989 Tumor-specific loss of 11p15.5 alleles in del11p13 Wilms’ tumor and in familial adre- nocortical carcinoma. Proc Natl Acad Sci USA. 86:3247-3251.
34. Gicquel C, Le Bouc Y, Luton JP, Girard F, Bertagna X. 1992 Monoclonality of corticotroph macroadenomas in Cushing’s disease. J Clin Endocrinol Metab. 75:472-475.
35. Chirgwin JM, Przybyla AE, MacDonald RJ, Rutter WJ. 1979 Isolation of biologically active ribonucleic acid from sources enriched in ribonuclease. Biochemistry. 18:5294-5299.
36. Pierotti MA, Radice P, Biunno I, Borrello MG, Cattadori MR, Della Porta G. 1986 Detection of two TaqI polymorphisms in the VTR region of the human HRAS1 oncogene. Cytogenet Cell Genet. 43:174-180.
37. Bell GI, Pictet RL, Rutter WJ, Cordell B, Tischer E, Goodman HM. 1980 Sequence of the human insulin gene. Nature. 284:26-32.
38. Le Bouc Y, Noguiez P, Sondermeijer P, Dreyer D, Girard F, Binoux M. 1987 A new 5’-non-coding region for human placental insulin-like growth factor II mRNA expression. FEBS Lett. 222:181- 185.
39. Barker D, Holm T, White R. 1984 A locus on chromosome 11p with multiple restriction site polymorphisms. Am J Hum Genet. 36:1159-1171
40. Le Moullec JM, Jullienne A, Chenais J, et al. 1984 The complete sequence of human preprocalcitonin. FEBS Lett. 167:93-97.
41. Quan F, Korneluk RG, Tropak MB, Gravel RA. 1986 Isolation and characterization of the human catalase gene. Nucleic Acids Res. 14:5321-5335.
42. Le Bouc Y, Dreyer D, Jaeger F, Binoux M, Sondermeijer P. 1986 Complete characterization of the human IGF-I nucleotide sequence isolated from a newly constructed adult liver cDNA library. FEBS Lett. 196:108-112.
43. Feinberg AP, Vogelstein B. 1983 A technique for radiolabeling DNA restriction endonuclease fragments to high specific activity. Anal Biochem. 132:6-13.
44. Bell GI, Horita S, Karam JH. 1984 A polymorphic locus near the human insulin gene is associated with insulin-dependent diabetes mellitus. Diabetes. 33:176-183.
45. Hoppener JW, Steenbergh PH, Zandberg J, et al. 1984 Localization of the polymorphic human calcitonin gene on chromosome 11. Hum Genet. 66:309-312.
46. Binoux M, Seurin D, Lassarre C, Gourmelen M. 1984 Preferential measurement of insulin-like growth factor (IGF) I-related peptides in serum with the aid of IGF-binding proteins (IGF BPs) produced by rat liver in culture. Estimation of serum IGF BP levels. J Clin Endocrinol Metab. 59:453-462.
47. Scrable H, Cavenee W, Ghavimi F, Lovell M, Morgan K, Sapienza C. 1989 A model for embryonal rhabdomyosarcoma tumorigenesis that involves genome imprinting. Proc Natl Acad Sci USA. 86:7480- 7484.
48. Holthuizen E, Le Roith D, Lund PK, et al. 1991 Revised nomen- clature for the insulin-like growth factor genes and transcripts. In: Spencer EM, ed. Modern concepts of insulin-like growth factors. New-York: Elsevier; 733-736.
49. Gloudemans T, Pospiech I, Van der Ven LTM, et al. 1992 Expres- sion and CpG methylation of the insulin-like growth factor II gene in human smooth muscle tumors. Cancer Res. 52:6516-6521.
50. Gama-Sosa MA, Slagel VA, Trewyn RW, et al. 1983 The 5- methylcytosine content of DNA from human tumors. Nucleic Acids Res. 11:6883-6894.
51. Feinberg AP, Vogelstein B. 1983 Hypomethylation distinguishes
genes of some human cancers from their normal counterparts. Nature. 301:89-92.
52. Little M, Van Heyningen V, Hastie N. 1991 Dads and disomy and disease. Nature. 351:609-610.
53. Seizinger BR, Klinger HP, Junien C, et al. 1991 Report of the committee on chromosome and gene loss in human neoplasia. Cytogenet Cell Genet. 58:1080-1096.
54. Dowdy SF, Fasching CL, Araujo D, et al. 1991 Suppression of tumorigenicity in Wilms’ tumor by the p15.5-14 region of chromo- some 11. Science. 254:293-295.
55. Koi M, Johnson LA, Kalikin LM, Little PFR, Nakamura Y, Fein- berg AP. 1993 Tumor cell growth arrest caused by subchromosomal transferable DNA fragments from chromosome 11. Science. 260:361-364.
56. Ferguson-Smith AC, Sasaki H, Cattanach BM, Surani MA. 1993 Parental-origin-specific epigenetic modification of the mouse H19 gene. Nature. 362:751-755.
57. Hao Y, Crenshaw T, Moulton T, Newcomb E, Tycko B. 1993 Tumour-suppressor activity of H19 RNA. Nature. 365:764-767.
58. Yano T, Linehan M, Anglard P, et al. 1989 Genetic changes in human adrenocortical carcinomas. J Natl Cancer Inst. 81:518-523.