Kapitel 4. Discussion

The results presented in this work indicate the presence of two transcripts of the Dnmt1 gene whose expression varies substantially amongst the organs analyzed. Thus, high levels of expression of the ubiquitous transcript (5.4 kb) were detected in mouse brain and heart (Fig. 3.3). These results are in agreement with previous reports from Northern blot analyses in human tissues which showed that adult brain and heart express relatively large amounts of Dnmt1 mRNA . Furthermore, a new slower migrating transcript (6.2 kb) was identified in testis and skeletal muscle (Fig. 3.3). A similar longer transcript had been previously reported in spermatogenic cells (pachytene spermatocytes) of the testis . The size difference between the ubiquitous and the new transcript shown by Northern blots suggests the presence of an alternative transcriptional start site or an alternative splice isoform in testis and skeletal muscle. Indeed, by RT-PCR and RACE a new Dnmt1 cDNA was isolated which differs in the first exon. Instead of the recently described exon 1 an alternative 800 bp long exon located downstream in the gene is present in this isoform indicating a new transcriptional start site. While this work was in progress, a similar cDNA was independently isolated from testis .

The expression pattern of the new Dnmt1 isoform has been analyzed by in situ hybridization and found to be in the more specialized testis cells the haploid spermatids and at low level in skeletal muscle. The presence of a different Dnmt1 transcript in spermatids suggests that the function and regulation of Dnmt1 might be different in non-replicating (spermatids) and in replicating spermatogonial cells. Interestingly, the same isoform is also expressed in terminally differentiated post-mitotic myotubes (Fig. 3.9). During gametogenesis the overall DNA methylation level increases and is higher in the male than in the female gametes (Fig. 1.5) . The presence of a new Dnmt1 isoform in differentiating spermatids is suggestive of a role in these methylation pattern changes.

The expression of the new Dnmt1 isoform was also investigated by in situ hybridization in embryo sections and the new transcript was localized in cartilages of the ear and hip. Interestingly, different patterns of methylation have been described for specific collagen genes at different stages of differentiation and dedifferentiation of chondrocytes . However, it remains to be tested whether the alternative Dnmt1 isoform is involved in these methylation pattern changes.

Dramatic changes in genomic methylation occur during development in oogenesis and spermatogenesis (Fig. 1.5). Therefore, it was important to investigate Dnmt1 isoform expression and localization at the protein level in both organs, testis and ovaries. In testis sections, by immunofluorescence studies, spermatogonia as well as spermatids showed the presence of Dnmt1 protein (Fig. 3.7). However, in the case of spermatids the fluorescence background is high making the discrimination of signals less clear. Other studies showed also high levels of Dnmt1 protein in spermatogonia, preleptotene and leptotene spermatocytes but in contrast to this work no protein was detected in spermatids and pachytene spermatocytes where the new transcript is made.

Ovaries from adult female mice were also analyzed for the presence of the Dnmt1 protein as shown in Fig. 3.6 where the Dnmt1 protein was localized mainly in the cytoplasm of mature oocytes and in the nucleus of lutein cells. Cytoplasmic retention of Dnmt1 protein in oocytes and preimplantation embryos has been previously reported and shown to be due to the presence of a retention signal in the N-terminal regulatory domain of the protein . This active retention of Dnmt1 in the cytoplasm correlates well with the overall decrease in the genomic methylation level and thus is a likely mechanism to regulate DNA methylation by separating Dnmt1 from chromosomal DNA in the nucleus. Demethylation might occur by preventing maintenance methylation of newly synthesized DNA strands or by preventing remethylation of actively demethylated sites.


Changes in DNA methylation occur not only during gametogenesis but also during differentiation. However, conflicting data have been published concerning the exact role of DNA methylation during differentiation. Thus on one hand, artificially induced demethylation seems to stimulate myogenic differentiation , which is in accordance with the observation that the somatic Dnmt1 isoform is downregulated during myogenesis . On the other hand, forced overexpression of a truncated Dnmt1 protein, was shown to cause de novo methylation in the MyoD gene, upregulate its expression and stimulate myogenesis .

In this work, a new alternative transcript was identified not just in testis but also in skeletal muscle (Fig. 3.3) and shown to be specifically upregulated during myogenesis, while the ubiquitous transcript is downregulated (Fig. 3.9). These results could reconcile the above contradiction, in the sense that the major Dnmt1 isoform responsible for maintenance of DNA methylation patterns is indeed downregulated but an alternative isoform with potentially different properties is upregulated.

The sequence analysis of this alternative Dnmt1 isoform from skeletal muscle (Fig. 3.8) showed that it is actually identical to the one present in testis, which was published as untranslatable due to the presence of short upstream ORFs . These authors reported the inhibitory effect of upstream AUG codons based on the scanning model for translation initiation, in which the 43S pre-initiation complex enters at the 5‘-cap structure of the mRNA and migrates in the 3‘-direction to the first initiation codon that is present in an appropriate context . Under some circumstances, however, an upstream AUG codon will reduce but not abolish initiation from downstream start codons. This happens when the first AUG codon is followed shortly by a terminator codon, creating a small open reading frame (ORF) at the 5‘ end of the mRNA. One possible explanation is that, after an 80S ribosome translates the 5‘-mini ORF, the 40S ribosomal subunit remains bound to the mRNA, resumes scanning, and may reinitiate at another downstream AUG codon downstream . The translational status of the alternative Dnmt1 isoform was also analyzed experimentally . These authors found that only ~15% of this Dnmt1 transcript was associated with polysomes and that a large portion of this transcript exist as free mRNPs, suggesting that this alternative transcript is translationally regulated. However, if inhibition of translation takes place in this alternative isoform it should also take place in the somatic and in particular in the oocyte isoform which contains several upstream AUG codons. Instead the Dnmt1 protein is present at very high levels in the oocyte (Fig. 3.6). Furthermore, the suppressive influence of upstream initiation codons might be regulated. The cells have other mechanisms to counteract suppression by an upstream ORF. Thus, an upstream ORF can be bypassed either through `leaky scanning´ or through the use of an internal ribosome entry site (IRES) .

When analysing these results together with the ones already published by other groups, it is clear that each experimental approach supports a different conclusion implicating technical factors. Thus, sedimentation profiles provide an accurate measure of translational activity but the most serious limitation is that the purified spermatogenic cells used for sucrose gradients contain a broad range of stages that could obscure translational differences in individual stages. This limitation could be overcome with the transillumination-assisted microdissection technique which permits the isolation of stages of spermatogenesis from freshly dissected seminiferous tubules . On the other hand, determining the stage of initial appearance of proteins and their mRNAs is subject to uncertainties caused by differences in the sensitivity of detection of mRNAs and proteins.

It is interesting to notice that two independent promoters are used in different tissues (oocytes versus testis and skeletal muscle) to generate two different transcripts, which however, lead to the expression of the same shorter Dnmt1 isoform (Fig. 3.10). The expression of an alternative Dnmt1 isoform during oogenesis, myogenesis and


spermatogenesis however does not prove an active role in this process, especially since both isoforms have very similar biochemical properties in vitro . The shorter isoform seems to be able to substitute for the longer isoform during differentiation and to restore DNA methylation patterns in Dnmt1 null ES cells . Thus a specific role of the shorter isoform would require tissue-specific, interacting factors that could differentially interact with the Dnmt1 isoform and thus generate specificity. In this context is interesting to notice that the additional 118 amino acids of the longer isoform are highly conserved from sea urchin to human . Although the function of this amino acid stretch is not known, the presence of a putative leucine zipper suggests a role in protein-protein interactions which might be conserved throughout evolution.

First hints for a role of the short Dnmt1 isoform during development were obtained by serendipity. Since the complete somatic Dnmt1 cDNA became known and available only in 1996 a partial cDNA was used by other investigators for overexpression in myoblasts giving rise to a truncated Dnmt1 protein starting at ATG4, which is identical to the shorter isoform described here. The forced expression of this Dnmt1 isoform in myoblasts was shown to correlate with a high level of de novo methylation activity and to induce myogenic differentiation . These experiments clearly show that the shorter Dnmt1 isoform from skeletal muscle cells can indeed induce myogenic differentiation.

In the present work the tissue specific expression of Dnmt1 isoforms was shown which might play a role in the regulation of DNA methylation. However, further studies are needed involving selective Dnmt1 gene knock outs and the examination of enzymatic properties of both Dnmt1 isoforms in vivo. The elucidation of the Dnmt1 gene structure and the mapping of alternative transcriptional start sites carried out in this work make it now possible to study the transcriptional regulation and to characterize the tissue specific Dnmt1promoters.

© Die inhaltliche Zusammenstellung und Aufmachung dieser Publikation sowie die elektronische Verarbeitung sind urheberrechtlich geschützt. Jede Verwertung, die nicht ausdrücklich vom Urheberrechtsgesetz zugelassen ist, bedarf der vorherigen Zustimmung. Das gilt insbesondere für die Vervielfältigung, die Bearbeitung und Einspeicherung und Verarbeitung in elektronische Systeme.

DiML DTD Version 2.0
Zertifizierter Dokumentenserver
der Humboldt-Universität zu Berlin
HTML - Version erstellt am:
Tue Sep 17 12:11:03 2002