The Wilms' tumor suppressor gene (WT1) was previously identified as being imprinted, with frequent maternal expression in human placentae and fetal brains. We examined the allele-specific expression of WT1 in cultured human fibroblasts from 15 individuals. Seven of 15 fibroblast lines were heterozygous for polymorphic alleles, and the expression patterns were variable, i.e., equal, unequal or monoallelic paternal expression in three, two and two cases, respectively. Exclusive paternal expression of WT1 was also shown in non-cultured peripheral lymphocytes from the latter two individuals. The allele-specific expression profiles of other imprinted genes, IGF2 and H19, on human chromosome 11 were constant and consistent with those in other tissues. Our unexpected observations of paternal or biallelic expression of WT1 in fibroblasts and lymphocytes, together with the previous findings of maternal or biallelic expression in placentae and brains, suggest that the allele-specific regulatory system of WT1 is unique and may be controlled by a putative tissue- and individual-specific modifier.
Genomic imprinting is defined as a gamete-of-origin dependent modification causing differential expression of the two alleles of a gene (1 ). In mice and humans, 17 genes have been shown to undergo imprinting, five of which were expressed from a maternal and 12 from a paternal allele (2 ). Most of these represent either tissue- developmental stage- or species-specific expression profiles, implying the presence of a cellular modifier(s) which regulates the appropriate imprinted expression (3 ). Since aberrant imprinting may cause embryonic abnormalities, genetic diseases and tumors (2 ), information on tissue-specific expression or cellular mosaicism in expression profiles of imprinted genes is important to elucidate the biological significance and molecular basis of imprinting.
The WT1 gene, located on chromosome 11p13, has been isolated and shown to be frequently inactivated in Wilms' tumor, a childhood kidney cancer (4 ,5 ). The gene encodes a zinc finger DNA-binding protein and is presumed to function as a transcriptional repressor (6 ). WT1 is highly expressed in the developing kidney and urogenital tract, but also in most Wilms' tumors and hematopoietic tumors, suggesting that the gene plays a role in urogenital development and hematopoietic differentiation (7 ,8 ). Each parental allele of WT1 was previously demonstrated to be equivalently expressed in normal fetal kidneys and Wilms' tumors (9 ). However, WT1 has been identified as being imprinted, with maternal expression in about half of pre-term placental villus and fetal brain tissue (10 ). Further extensive studies showed that maternal monoallelic expression was observed in 39% of the samples, while the expression of other samples was biallelic (11 ). In the present study, we examined allele-specific expressions of WT1 as well as IGF2 and H19 on human chromosome 11 in fibroblasts and lymphocytes. Unlike findings in placentae and fetal brains, we unexpectedly observed that WT1 was expressed from the paternal allele, at least in some individuals.
Using a dinucleotide repeat (DR) polymorphism in the 3' untranslated region (UTR) of the WT1 gene, we assessed the allelic expression of WT1 in cultured human fibroblasts from 15 normal individuals. WT1 was expressed exclusively from the paternal allele in two of seven informative lines with the polymorphism (cases 1 and 2, Fig. 1 A), while expression of WT1 was detected from both paternal and maternal alleles in the remaining five lines (cases 3-7, Table 1 ). To exclude the tube-to-tube variability in band patterns that has frequently been observed in the microsatellite PCR assay, we also performed PCR amplification with primers flanking a WT1-HinfI restriction fragment length polymorphism (RFLP) (10 ). The result was consistent with those using the microsatellite PCR assay (case 1, Fig. 1 B). We also examined the allelic expression of WT1 using the DR polymorphism in fresh lymphocytes from two individuals who represented paternal expression in fibroblasts (cases 1 and 2). As in the fibroblasts, exclusive paternal expression of WT1 was shown in the lymphocytes (Fig. 1 C).
Table 1
We have demonstrated that cultured human fibroblasts and lymphocytes showed paternal or biallelic expression of WT1 in some cases, in contrast with previous findings of maternal or biallelic expression of WT1 in human placental villus and fetal brain tissue (10 ,11 ), suggesting a unique allele-specific expression profile of WT1. WT1 was also expressed exclusively from the paternal allele in non-cultured peripheral lymphocytes derived from two individuals with paternal expression in their cultured fibroblasts. Unlike WT1, expression profiles of IGF2 and H19 were constant and consistent with what has been observed in other tissues. Thus, it is feasible that paternal expression of WT1, at least in some cases, reflected the in vivo status rather than a cell culture artifact. These observations support an idea that inter-individual variation results in variable allele-specific expression patterns of WT1.
Hence, WT1 showed either paternal, biallelic or maternal expression in some cell types from different individuals, although the functional importance of WT1 for these tissues showing polymorphic expression remains to be clarified. The observed patterns of paternal and maternal expression in different tissues possibly represent extreme skewing of an allelic exclusion mechanism of WT1 expression. Thus, kidney apparently shows equal allelic expression of WT1, but individual cells could only express one allele, with different cells showing a paternal- or maternal-only pattern. An analogous situation has been observed in the random X inactivation in female somatic cells or allelic inactivation of olfactory receptor genes (14 ,15 ), and skewed X inactivation has been reported in humans and mice (16 ).
Tissue-, developmental stage- or individual-specific biallelic expression of imprinted genes has been generally considered to result from a relaxation or loss of imprinting (17 ,18 ). This is suggested by evidence from biallelic expression of IGF2 and H19 in human neoplasms (12 ,13 ) and a subgroup of Beckwith-Wiedemann syndrome (19 ) characterized by a high incidence of embryonal tumors. The term `relaxation of imprinting,' or `loss of imprinting,' implies that a negative signal on the repressed allele is extinguished or not recognized, allowing biallelic expression (17 ,18 ). In the case of WT1, however, a feasible explanation for the unexpected expression patterns is that WT1 is subjected to imprinting during gametogenesis, and, later, a tissue- and individual-specific modifier(s) determines the allelic expression pattern. Although this explanation of allele-specific expression of WT1 may or may not be valid for all imprinted genes, it is worthwhile to mention that two interesting exceptions for IGF2 and H19 with expression from only the otherwise silent allele have been previously described (20 ,21 ). The existence of genetic modifiers of functional imprinting is also suggested by previous observations with polymorphic IGF2 imprinting in humans, as well as transgene imprinting in mice (22 ,23 ). Since variable allele-specific expression profiles of imprinted genes have been implicated in human genetic disorders and susceptibility to tumors (24 ), or even if biological meanings may not be uncovered, it is valuable to examine inter- and intra-individual variabilities in the allelic expression of WT1 for better understanding of regulatory mechanisms involved in genomic imprinting.
Fifteen skin biopsies from different individuals were obtained with standard techniques. Fibroblast cultures were maintained at 37°C in Dulbecco's modified Eagle's medium (DMEM, Sigma) supplemented with 10% fetal calf serum and antibiotics (penicillin-streptomycin, Gibco BRL) and expanded until sufficient cells were present for nucleic acid extraction. Fresh lymphocytes were also obtained from the two individuals with monoallelic WT1 expression in fibroblasts.
PCR analysis of genomic DNA was performed as previously reported (12 ), with minor modifications. Prior to reverse transcriptase (RT), the total RNA samples were digested with RNase-free DNase I to prevent DNA contamination. To investigate the possibility of DNA contamination in the RNA sample, total RNA was incubated with (RT+) or without (RT-) M-MLV RT (Gibco BRL), including an oligo (dT)15 primer. The same primers and PCR conditions were used for PCR analysis of cDNA as those used for genotyping. No amplification was detected in the absence of RT, excluding DNA contamination.
To detect the WT1-DR polymorphism in the 3' UTR (25 ), primer 400 (5'-AATGAGACTTACTGGGTGAGG-3') was end-labeled with [[gamma]-32P]dATP (Amersham) using T4 polynucleotide kinase, followed by PCR amplification using primers 400 and 401 (5'-TTACACAGTAATTTCAAGCAACGG-3') with 35 cycles, at 94°C for 60 s, 60°C for 30 s and 72°C for 60 s. PCR products were resolved on electrophoresis in 8% polyacrylamide/8 M urea sequencing gels and exposed to an Imaging Screen Cassette-BI (Bio-Rad). The area under each allele peak was integrated and compared with other alleles in the lane, using the GS-525 Molecular Imager system (Bio-Rad). Primers WTHfa (5'-AATCAGAGAGCAAGGCATCG-3') and WTHfb (5'-GTGCAAGGAGGTATGTACATC-3') were used for the WT1-HinfI RFLP analysis. A total of 35 PCR cycles, at 94°C for 60 s, 58°C for 30 s and 72°C for 60 s, were performed. To assess allele-specific expression of IGF2 and H19, a transcribed ApaI RFLP in the 3' UTR of IGF2 and an RsaI RFLP in exon 5 of H19 were analyzed. The primers and PCR conditions used were: for detection of the IGF2-ApaI RFLP, I1 (5'-CTTGGACTTTGAGTCAAATTGG-3') and I3 (5'-CCTCCTTTGGTCTTACTGGG-3'), 35 cycles at 94°C for 60 s, 55°C for 30 s and 72°C for 60 s; for the detection of H19-RsaI RFLP, H1 (5'-TACAACCACTGCACTACCTG-3') and H3 (5'-TGGAATGCTTGAAGGCTGCT-3'), 35 cycles at 94°C for 1 min, 62°C for 3 min and 72°C for 5 min in the presence of 15% glycerol. PCR products were ethanol-precipitated, digested with the enzyme, electrophoresed on a 5% polyacrylamide gel and ethidium-bromide stained.
We would like to thank Dr Masao Yamada (National Children's Medical Center) for providing primers 400 and 401, Dr Takashi Takahashi (Aichi Cancer Center Research Institute) for valuable suggestions. The present study was supported by the Japan Science and Technology Corporation (JST) and the Mitsubishi Foundation, Japan.
Human Molecular Genetics
Pages
Introduction
Results
Discussion
Materials And Methods
Cell lines
PCR and RT-PCR
Acknowledgements
References
Case
Sample/passage
Sourcea
Sex
Age (years)
WT1 expression
1
Fibroblast/p4
FS
M
24
Paternalb
Lymphocytec
PB
Paternal
2
Fibroblast/p4
FS
M
23
Paternal
Lymphocytec
PB
Paternal
3
Fibroblast/p10
FS
F
Unequald
4
Fibroblast/p3
VS
M
51
Unequald
(kidney
Equal)
5
Fibroblast/p3
VS
M
71
Equal
(kidney
Equal)
6
Fibroblast/p19
FB
M
Fetus
Equal
(kidney
Equal)
7
Fibroblast/p10
UC
F
0
Equal
(placenta
Equal)
REFERENCES
This page is maintained by OUP admin. Last updated Sat Oct 18 13:29:50 BST 1997. Part of the OUP Journals World Wide Web service.
Copyright
Oxford University Press, 1997