Numerous observations link the loss of imprinting of insulin-like growth factor 2 (IGF2) and an overdosage of this growth factor gene with cancer, in general, and with Wilms’ tumorigenesis, in particular. It is not known, however, if loss of imprinting correlates with specific stages of neoplasia or if allelic expression patterns vary within the tumor. By applying an allele-specific in situ hybridization technique to formalin-fixed thin sections, we show that the parental IGF2 alleles can be differentially expressed, not only in Wilms’ tumors, but also in nephrogenic rests (which represent premalignant lesions) of Wilms’ tumor patients. Moreover, a subpopulation of mesenchymal cells, which surrounds tumor nodules, expresses IGF2 biallelically irrespective of the imprinted state of IGF2 within the tumor. These data show that Wilms’ tumorigenesis involves epigenetic heterogeneity as visualized by variable allelic IGF2 expression patterns.

Wilms’ tumor is a malignancy of the embryonic kidney, which has proven to be a productive model for understanding the pathogenesis of human neoplasia (1, 2, 3, 4). It has been proposed that nephrogenic rests, which are small foci of abnormally persistent blastema-derived cells, are potential precursors of Wilms’ tumors (1). Histological forms of these preneoplastic lesions are associated with Wilms’ tumor-predisposing syndromes, such as BWS3(5). BWS is associated with a fetal overgrowth condition that has been suggested to be caused by the perturbed regulation of imprinted genes in the 11p15.5 region (1, 2, 3, 4). One of these candidates is IGF2, which produces a growth factor (6) that in normal circumstances is expressed primarily from the paternal allele (7, 8, 9, 10). In addition, the organ-specific expression of IGF2 during human fetal development correlates with a pattern of organ overgrowth and the predisposition of tumors in the BWS (11).

Several independent lines of research show that the epigenetically silenced maternal IGF2 allele is reactivated (LOI) in Wilms’ tumors, which presumably results in a growth factor overdosage (1, 2, 3, 4). Significantly, it has been shown that the epigenetic status of the IGF2/H19 region is perturbed in normal kidneys of BWS, as well as non-BWS cases of Wilms’ tumor (12). By using an allele-specific in situ hybridization technique (13, 14, 15), we show here that IGF2 LOI is more complex than previously realized. First, it is not stage-specific because biallelic expression of the IGF2 gene can appear in the normal context in premalignant lesions and postneoplastically. Second, the pattern of imprinting and LOI seems to be surprisingly complex because opposite alleles can be found expressed monoallelically in individual cell subpopulations within the tumor tissue and in nephrogenic rests of Wilms’ tumor patients.

Tissue Samples and Nucleic Acid Extraction.

Specimens representing Wilms’ tumors, kidneys, and nephrogenic rests were collected from patients who (except for specimens 8 and 9) had received preoperative chemotherapy. These specimens include materials from patients suffering from BWS (specimen 2; the patient later developed fatal and anaplastic Wilms’ tumors, specimen 3) and the Denys-Drash syndrome (specimen 6), and four patients had multiple tumors (specimens 1–4 and 6). Routinely processed formalin-fixed, paraffin-embedded tissues were used for ISH. DNA and RNA were extracted from snap-frozen tissues, fibroblast cultures, and from peripheral leukocytes, as has been described previously (16).

IGF2 Imprinting Status.

The tissue specimens were scored for ApaI polymorphisms within exon 9 of human IGF2, as has been described (8). Eleven of the tumor specimens were heterozygous for the ApaI polymorphic site and were examined more closely, as summarized in Table 1. The overall functional IGF2 imprinting status was determined by reverse transcription-PCR, as has been described (8). PCR-amplified DNA from each specimen, except cases 8 and 9, was sequenced over the ApaI site to confirm sequence identity (Sequenase Version 2.0, USB).

ASISH.

Serial 5-μm sections from formaldehyde-fixed and paraffin-embedded tissues or 4% paraformaldehyde-fixed freeze sections were used for ASISH, as has been described (13, 14). The slides were dipped in NTB2 (Kodak) emulsion, developed and counterstained in Mayer’s hematoxylin before mounting. ISH analysis using an IGF2 antisense RNA probe was carried out as has been described (13).

The purpose of this work was to examine the imprinting status of the IGF2 gene at the cellular level in Wilms’ tumors by applying the method of ASISH analysis (13, 14, 15). Therefore, we first identified 12 Wilms’ tumor patients who were heterozygous for the diagnostic ApaI polymorphic site at the IGF2 locus (Table 1). A conventional reverse transcription-PCR imprinting assay revealed that IGF2 was biallelically expressed in seven of these Wilms’ tumors (Table 1 and data not shown). Four of the specimens included nephrogenic rests, the precursor of Wilms’ tumor (1). The high resolution of the ASISH approach allowed us to specifically address the temporal window during which the LOI for IGF2 is manifested. Fig. 1 shows the result of such an analysis. The primary tumor of specimen 2, which was derived from a patient suffering from BWS, displayed homogeneous levels of IGF2 transcripts derived from both parental alleles. Because this pattern of allelic expression could be extended to nephrogenic rests and the patient suffered from bilateral Wilms’ tumors, IGF2 LOI is an early and perhaps constitutional event in this patient. The BWS patient developed a relapse 6 years later, which showed anaplastic histology and eventually led to the fatal progression of the disease. In contrast to the primary tumor, the allelic IGF2 expression patterns of the relapse were heterogeneous, with subpopulations of cells expressing IGF2 monoallelically (Fig. 2, A-C). This pattern persisted throughout the tumor (Fig. 2 and data not shown). We conclude that the fatal progression of Wilms’ tumorigenesis is in this patient associated with the emergence of variable allelic expression patterns of the IGF2 gene.

Similarly, IGF2 LOI is associated with tumor progression in another patient (specimen 9; Table 1). Whereas the vast majority of tumor cells expressed IGF2 monoallelically, a subpopulation of tumor cells expressed IGF2 biallelically (17, 18). It seems that both biallelic expression and loss of biallelic expression can occur as a result of tumor progression. This is not restricted to malignant transformation, however, because also nephrogenic rests could express both parental IGF2 alleles in a heterogeneous manner (Fig. 3). When comparing the allelic expression pattern in the remainder of the Wilms’ tumor specimens, it became apparent that the heterogenous allelic expression pattern of IGF2 represented a common feature. Table 1 summarizes these results, which show that five different Wilms’ tumor specimens display a heterogeneous imprinting status (i.e., the IGF2 alleles are differentially expressed in cell subpopulations within the tumor tissue).

An unexpected outcome of the ASISH analysis was the observation that IGF2 LOI could be observed in a subpopulation of mesenchymal cells in five of the Wilms’ tumors. Fig. 4 shows one such example (specimen 10); although IGF2 is monoallelically expressed in the tumor cells, both parental alleles are active in a small subset of hyperplastic mesenchymal cells. Interestingly, the cells displaying biallelic IGF2 expression specifically line the tumor nodules, which in this instance expressed IGF2 monoallelically (Fig. 4). Other examples show variable allelic expression patterns of IGF2 in fibrovascular stroma (Table 1; data not shown).

We have shown here that there is a considerable heterogeneity with respect to allelic IGF2 expression patterns in both nephrogenic rests and Wilms’ tumors, as well as in the mesenchyme of Wilms’ tumors. This epigenetic heterogeneity could be observed in the majority of specimens (five of seven Wilms’ tumors) that expressed IGF2 biallelically. In some instances, IGF2 seemed to be expressed monoallelically from opposite parental alleles within the tissue, although the small size of such cell populations occasionally precluded an unambiguous assessment of allelic expression patterns in adjacent slides. Nonetheless, these data imply that the link between the origin of Wilms’ tumor and epigenetic disturbances (17, 18, 19, 20, 21, 22, 23) is more complex than previously anticipated. Whereas the present study analyzed primary tumors, other studies have revealed that both complete hydatidiform moles (14) and a choriocarcinoma cell line display hypervariable allelic IGF2 and H19 expression patterns (24), suggesting that epigenetic heterogeneity may not be uncommon in association with human pathology.

The stochastic appearance of IGF2 LOI in both premalignant and malignant cells could reflect an epigenetic lesion that is constitutive or which involves only a subpopulation of cells. By analogy to the mosaicism of allelic IGF2 expression patterns in human androgenetic cells (14, 25), the pathological neutralization of parental marks of imprinted genes may generate epigenetic heterogeneity during Wilms’ tumorigenesis. It is potentially interesting in this context that in vitro culturing of mouse embryos leads to dynamic readjustment of the methylation pattern of an imprinting control region upstream of the H19 gene.4 These data imply that an absence of a normal, or the presence of an abnormal physiological stimulus during in vitro development randomly switches the parental identity of the region that represses maternal Igf2 expression (26) to generate a mosaic methylation pattern. Similarly, premalignant cells may experience a modified genetic and/or physiological set up that directly or indirectly generates an epigenetic heterogeneity. We have earlier postulated that such an epigenetic mosaicism facilitates selection processes that may eventually lead to an IGF2 overdosage and absence of H19 expression during Wilms’ tumorigenesis (17, 25).

Given that the perceived maintenance of biallelic expression may require a selectable function, it is a disturbing fact that Wilms’ tumors do not produce high levels of IGF II ligand (27), presumably due to translation repression of the cytoplasmic IGF2 transcripts (18). The observed epigenetic heterogeneity with respect to allelic IGF2 expression patterns, therefore, could reflect absence of a function for the IGF2 gene in the malignant phenotype of Wilms’ tumor. According to this scenario, the IGF2 overdosage and biallelic expression only provide a selectable marker during a specific window in the tumorigenic process. In the absence of a growth selection pressure, variable allelic IGF2 expression patterns may reappear. Although these scenarios are compatible with the data reported here, we cannot rule out the possibility that the biallelic IGF2 expression patterns are not necessarily causative, but simply diagnostic for other epigenetic events occurring at linked and currently unknown loci.

Our observation that the mesenchyme, which surrounds some of the tumor nodules, expresses IGF2 biallelically in some of the specimens is potentially interesting. There are three main and not mutually exclusive explanations to this result. Firstly, the mesenchymal cells and the tumor cells may belong to the same pool of epigenetically heterogeneous precursor cells (28). The increased dosage of IGF2 expression could somehow facilitate an interaction with the tumor nodules. Secondly, biallelic IGF2 expression is simply a marker for epigenetic selection mechanisms operating on other loci, as discussed above. Thirdly, the tumor nodules and the juxtaposed mesenchyme may interact to induce IGF2 LOI in the mesenchyme. Although the detailed allelic expression patterns are currently not known, we note with interest that normal tissue of colon cancer (29) and breast cancer (30) patients display IGF2 LOI.

We submit that IGF2 LOI is associated with heterogenous allelic expression patterns in Wilms’ tumor. It may now be essential to address whether or not IGF2 LOI is the result of a genome-wide, regional or locus-specific breakdown of epigenetic systems during tumorigenesis and how these relate to DNA methylation (31).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

      
1

Supported by the Swedish Cancer Research Foundation (RmC/CF; to R. O.), the Swedish Pediatric Cancer Foundation (BCF; to R. O. and F. H.), the Swedish Natural Science Research Council (NFR; to R. O.), the Wenner-Gren Foundation (to H. C.), and NIH Grant CA65145 (to A. F. P.).

            
3

The abbreviations used are: BWS, Beckwith-Wiedemann syndrome; IGF, insulin-like growth factor; LOI, loss of imprinting; ISH, in situ hybridization; ASISH, allele-specific ISH.

      
4

L. Liu, M. Pilartz, C. Kanduri, P. Wentgel, U. Eriksson, and R. Ohlsson, unpublished observations.

Fig. 1.

IGF2 allele usage in neoplastic and preneoplastic kidney tissues of a child affected by BWS (specimen 2). Allele A and allele B depict transcripts that were detected by the labeled oligoprobe, which was complementary to ApaI noncutting and cutting alleles, respectively. Magnification is 32-fold for all panels.

Fig. 1.

IGF2 allele usage in neoplastic and preneoplastic kidney tissues of a child affected by BWS (specimen 2). Allele A and allele B depict transcripts that were detected by the labeled oligoprobe, which was complementary to ApaI noncutting and cutting alleles, respectively. Magnification is 32-fold for all panels.

Close modal
Fig. 2.

IGF2 allele usage in a relapse (specimen 3; see Table 1). A, autoradiographic images of allelic expression patterns of IGF2. Right, a computer-colored overlay to illustrate the variable allelic expression patterns of IGF2 in this Wilms’ tumor. The red-lilac, blue, and yellowish colors depict monoallelic A, B, and biallelic expression, respectively (magnification, 1.3-fold). B, a region within the boxed area in A is magnified to document that the images of the heterogeneous allelic expression patterns of the IGF2 gene do not reflect absence of particular tissue cells (magnification, 25-fold). C, an enlarged portion of B, as indicated by the boxed area in the left-most image (magnification, 100-fold).

Fig. 2.

IGF2 allele usage in a relapse (specimen 3; see Table 1). A, autoradiographic images of allelic expression patterns of IGF2. Right, a computer-colored overlay to illustrate the variable allelic expression patterns of IGF2 in this Wilms’ tumor. The red-lilac, blue, and yellowish colors depict monoallelic A, B, and biallelic expression, respectively (magnification, 1.3-fold). B, a region within the boxed area in A is magnified to document that the images of the heterogeneous allelic expression patterns of the IGF2 gene do not reflect absence of particular tissue cells (magnification, 25-fold). C, an enlarged portion of B, as indicated by the boxed area in the left-most image (magnification, 100-fold).

Close modal
Fig. 3.

Mosaic IGF2 imprinting status in hyperplastic nephrogenic rests (specimen 6). A, D, and G, bright-field views of areas displaying differential imprinting status. B, E, and H, dark-field views of allele A-generated transcripts. C, F, and I, dark-field views of allele B-generated transcripts. B and C, E and F, and H and I are consecutive sections. Magnification, 33-fold.

Fig. 3.

Mosaic IGF2 imprinting status in hyperplastic nephrogenic rests (specimen 6). A, D, and G, bright-field views of areas displaying differential imprinting status. B, E, and H, dark-field views of allele A-generated transcripts. C, F, and I, dark-field views of allele B-generated transcripts. B and C, E and F, and H and I are consecutive sections. Magnification, 33-fold.

Close modal
Fig. 4.

Tumor-specific connective tissue expresses IGF2 biallelically. ASISH analysis was performed as outlined in “Materials and Methods.” A-D and E-H display two different areas within the same Wilms’ tumor (specimen 10). A, B, E, and F, bright-field views of tissue morphology. C, D, G, and H, dark-field views of allelic IGF2 expression patterns. WT, Wilms’ tumor; CT, connective tissue. ∗, false-positive dark-field signal. Magnification, 40-fold.

Fig. 4.

Tumor-specific connective tissue expresses IGF2 biallelically. ASISH analysis was performed as outlined in “Materials and Methods.” A-D and E-H display two different areas within the same Wilms’ tumor (specimen 10). A, B, E, and F, bright-field views of tissue morphology. C, D, G, and H, dark-field views of allelic IGF2 expression patterns. WT, Wilms’ tumor; CT, connective tissue. ∗, false-positive dark-field signal. Magnification, 40-fold.

Close modal
Table 1

Allelic expression patterns of IGF2 in clinical variants of Wilms’ tumora

SpecimenTissueDNARNA
Wilms’ tumor A/B −/B; A/B 
 Tumor mesenchyme  −/B; A/B 
Wilms’ tumor  A/B 
 Nephrogenic rests A/B A/B 
 Tumor mesenchyme  A/B 
3b Wilms’ tumor A/B A/B; A/−; −/B 
 Tumor mesenchyme  A/B; A/− 
Wilms’ tumor A/B −/B 
Wilms’ tumor A/B −/B 
Wilms’ tumor A/B NDc 
 Nephrogenic rests  A/B; A/−; −/B 
Wilms’ tumor A/−d A/− 
 Nephrogenic rests A/−d A/− 
Wilms’ tumor A/B A/B 
Wilms’ tumor A/B A/B; A/− 
 Tumor mesenchyme  A/−; A/B 
10 Wilms’ tumor A/B A/−; A/B 
 Tumor mesenchyme  A/−; A/B 
11 Wilms’ tumor A/B −/B 
 Nephrogenic rests  −/B 
12 Wilms’ tumor A/B −/B; A/B 
SpecimenTissueDNARNA
Wilms’ tumor A/B −/B; A/B 
 Tumor mesenchyme  −/B; A/B 
Wilms’ tumor  A/B 
 Nephrogenic rests A/B A/B 
 Tumor mesenchyme  A/B 
3b Wilms’ tumor A/B A/B; A/−; −/B 
 Tumor mesenchyme  A/B; A/− 
Wilms’ tumor A/B −/B 
Wilms’ tumor A/B −/B 
Wilms’ tumor A/B NDc 
 Nephrogenic rests  A/B; A/−; −/B 
Wilms’ tumor A/−d A/− 
 Nephrogenic rests A/−d A/− 
Wilms’ tumor A/B A/B 
Wilms’ tumor A/B A/B; A/− 
 Tumor mesenchyme  A/−; A/B 
10 Wilms’ tumor A/B A/−; A/B 
 Tumor mesenchyme  A/−; A/B 
11 Wilms’ tumor A/B −/B 
 Nephrogenic rests  −/B 
12 Wilms’ tumor A/B −/B; A/B 
a

Alleles A and B of IGF2 depict the Apa I noncutting and cutting alleles, respectively.

b

Fatal progression of Wilms’ tumor 6 years later (compare with specimen 2).

c

ND, not determined because the tissue was too necrotic.

d

Loss of heterozygosity.

We are grateful to Drs. Bruce Beckwith and Gary Franklin for valuable advice and Dr. Lena Scheibenpflug for assistance with clinical materials.

1
Bove K. E., McAdams A. J. The nephroblastomatosis complex and its relationship to Wilms’ tumor.
Perspect. Pediatr. Pathol.
,
3
:
185
-223,  
1976
.
2
Reeve A. E. Role of genomic imprinting in Wilms’ tumor and overgrowth disorders.
Med. Pediatr. Oncol.
,
27
:
470
-475,  
1996
.
3
Moulton T., Chung W-Y., Yuan L., Hensle T., Waber P., Nisen P., Tycko B. Genomic imprinting and Wilms’ tumor.
Med. Pediatr. Oncol.
,
27
:
476
-483,  
1996
.
4
Feinberg A. P. Multiple genetic abnormalities of 11p15 in Wilms’ tumor.
Med. Pediatr. Oncol.
,
27
:
484
-489,  
1996
.
5
Beckwith J. B. Precursor lesions of Wilms’ tumor: clinical and biological implications.
Med. Pediatr. Oncol.
,
21
:
158
-168,  
1993
.
6
Stewart C. E., Rotwein P. Growth, differentiation, and survival: multiple physiological functions for insulin-like growth factors.
Physiol. Rev.
,
76
:
1005
-1026,  
1996
.
7
Rainier S., Johnson L., Dobry C., Ping A., Grundy P., Feinberg A. Relaxation of imprinted genes in human cancer.
Nature (Lond.)
,
362
:
747
-749,  
1993
.
8
Ogawa O., Eccles M., Szeto J., McNoe L., Yun K., Maw M., Smith P., Reeve A. Relaxation of insulin-like growth factor II gene imprinting implicated in Wilms’ tumor.
Nature (Lond.)
,
362
:
749
-751,  
1993
.
9
Ohlsson R., Nyström A., Pfeifer-Ohlsson S., Töhönen V., Hedborg F., Schofield P., Flam F., Ekström T. J. IGF2 is parentally imprinted during human embryogenesis and in the Beckwith-Wiedemann syndrome.
Nat. Genet.
,
4
:
94
-97,  
1993
.
10
Giannoukakis N., Deal C., Paquette J., Goodyer C., Polychronakos C. Parental genomic imprinting of the human IGF2 gene.
Nat. Genet.
,
4
:
98
-101,  
1993
.
11
Hedborg F., Holmgren L., Sandstedt B., Ohlsson R. The cell type-specific IGF2 expression during early human development correlates to the pattern of overgrowth and neoplasia in the Beckwith-Wiedemann syndrome.
Am. J. Pathol.
,
145
:
802
-817,  
1994
.
12
Okamoto K., Morison I. M., Taniguchi T., Reeve A. E. Epigenetic changes at the insulin-like growth factor II/H19 locus in developing kidney is an early event in Wilms’ tumorigenesis.
Proc. Natl. Acad. Sci. USA
,
94
:
5367
-5371,  
1997
.
13
Adam G., Cui H., Miller S., Flam F., Ohlsson R. Allele-specific in situ hybridization (ASISH) analysis: a novel technique which resolves differential allelic usage of H19 within the same cell lineage during human placental development.
Development (Camb.)
,
122
:
839
-847,  
1996
.
14
Ohlsson R., Cui H., Flam F., Fisher R., Miller S., He L., Pfeifer S., Adam G. I. R. Absence of discriminative parental imprints is linked with epigenetic mosaicism of the IGF2 and H19 genes.
Dev. Genes Evol.
,
209
:
113
-119,  
1999
.
15
Svensson K., Mattsson R., James T., Wentzel P., Pilartz M., MacLaughlin J., Miller S., Olsson T., Eriksson U., Ohlsson R. The paternal allele of the H19 gene is progressively silenced during early mouse development: the acetylation status of histones may be involved in the generation of variegated expression patterns.
Development (Camb.)
,
125
:
61
-69,  
1998
.
16
Ohlsson R., Hedborg F., Holmgren L., Walsh C., Ekström T. J. Overlapping patterns of IGF2 and H19 expression during human development: biallelic IGF2 expression correlates with a lack of H19 expression.
Development (Camb.)
,
120
:
361
-368,  
1994
.
17
Cui H., Hedborg F., He L., Nordenskjöld A., Sandstedt P-O. S., Ohlsson R. Inactivation of H19, an imprinted and putative tumor repressor gene, is a preneoplastic event during Wilms’ tumorigenesis.
Cancer Res.
,
57
:
4469
-4473,  
1997
.
18
Li Y-M., Franklin G., Cui H. M., Svensson K., He X. B., Adam G., Ohlsson R., Pfeifer S. The H19 transcript is associated with polysomes and may regulate IGF2 expression in trans.
J. Biol. Chem.
,
273
:
28247
-28252,  
1998
.
19
Hao Y., Crenshaw T., Moulton T., Newcomb E., Tycko B. Tumor-suppressor activity of H19 RNA.
Nature (Lond.)
,
365
:
764
-767,  
1993
.
20
Moulton T., Crenshaw T., Hao Y., Moosikasuwan J., Lin N., Dembitzer F., Hensle T., Weiss L., McMorrow L., Loew T., Kraus W., Gerald W., Tycko B. Epigenetic lesions at the H19 locus in Wilms’ tumor patients.
Nat. Genet.
,
7
:
440
-447,  
1994
.
21
Steenman M. J., Rainier S., Dobry C. J., Grundy P., Horon I. L., Feinberg A. P. Loss of imprinting of IGF2 is linked to reduced expression and abnormal methylation of H19 in Wilms’ tumor.
Nat. Genet.
,
7
:
433
-439,  
1994
.
22
Feinberg A. J. Genomic imprinting and gene activation in cancer.
Nat. Genet.
,
4
:
110
-113,  
1993
.
23
Tycko B. Genomic imprinting and cancer Ohlsson R. eds. .
Genomic Imprinting: An Interdisciplinary Approach
,
25
:
133
-170, Springer-Verlag Berlin, Heidelberg-New York  
1999
.
24
He L., Cui H., Walsh C., Mattsson R., Lin W., Anneren G., Pfeifer-Ohlsson S., Ohlsson R. Hypervariable allelic expression patterns of the imprinted IGF2 gene in tumor cells.
Oncogene
,
16
:
113
-119,  
1998
.
25
Ohlsson R., Tycko B., Sapienza C. Monoallelic expression: “there can only be one.”.
Trends Genet.
,
14
:
435
-438,  
1998
.
26
Thorvaldsen J. L., Duran K. L., Bartolomei M. S. Deletion of the H19 differentially methylated domain results in loss of imprinted expression of H19 and Igf2.
Genes Dev.
,
12
:
3693
-3702,  
1998
.
27
Haselbacher G. K., Irminger J. C., Zapf J., Ziegler W. H., Humbel R. E. Insulin-like growth factor II in human adrenal pheochromocytomas and Wilms’ tumors: expression at the mRNA and protein level.
Proc. Natl. Acad. Sci. USA
,
84
:
1104
-1106,  
1987
.
28
Zhuang Z., Merino M. J., Vortmeyer A. O., Bryant B., Lash A. E., Wang C., Deavers M. T., Shelton W. F., Kapur S., Chandra R. S. Identical genetic changes in different histologic components of Wilms’ tumors.
J. Natl. Cancer Inst.
,
89
:
1148
-1152,  
1997
.
29
Cui H., Horon I. L., Ohlsson R., Hamilton S. R., Feinberg A. P. Loss of imprinting in normal tissue of colorectal cancer patients with microsatellite instability.
Nat. Med.
,
4
:
1276
-1280,  
1998
.
30
van Roozendaal C. E., Gillis A. J., Klijn J. G., van Ooijen B., Claassen C. J., Eggermont A. M., Henzen-Logmans S. C., Oosterhuis J. W., Foekens J. A., Looijenga L. H. Loss of imprinting of IGF2 and not H19 in breast cancer, adjacent normal tissue and derived fibroblast cultures.
FEBS Lett.
,
16
:
107
-111,  
1998
.
31
Feinberg A. P. Imprinting of a genomic domain of 11p15 and loss of imprinting in cancer: an introduction.
Cancer Res.
,
59
:
1743s
-1746s,  
1999
.