The Inverse Effect Hypothesis

It is broadly accepted that the MSL complex functions to specifically upreg-ulate the expression of genes from the single male X chromosome. However, there is another school of thought that believes this is an oversimplified model of dosage compensation. There are several reasons for this. For example, dosage compensation occurs not only in males (1X:2A), but also in metamales (1X:3A), metafemales (3X:2A) and triploid intersexes (2X:3A; Birchler 1992; Birchler et al. 2003). The dosage compensation mechanism described above cannot account for these scenarios. This and other inconsistencies have led the Birchler group to use the inverse effect hypothesis to explain dosage compensation.

It was observed that when large sections of chromosomes were deleted (segmental aneuploidy) the most common effect observed was a genome-wide upregulation of genes (Birchler and Schwartz 1979; Sabl and Birchler 1993). This phenomenon is called inverse dosage effect, where the deleted region often contains negative regulators and loss thereof results in the upregulation of many unlinked genes. The larger the deletion is, the more global the inverse effect. Importantly, this also means that the deleted segment (or parts thereof) may not be downregulated, as the global upregulation feeds back on its non-deleted homologue and returns its expression to normal levels; however, much of the rest of the genome suffers an approximate twofold increase in gene expression (Birchler et al. 2001).

The similarity to Drosophila, where the presence of a single X chromosome in males essentially generates an aneuploid state, is obvious. Such an aneu-ploid state could result in an inverse dosage effect on the autosomes, whereas the single X chromosome is automatically compensated by the upregulation feedback. Such a large-scale twofold upregulation from the autosomes would be lethal for a cell. Birchler and colleagues have proposed that the function of the DCC is to sequester the MOF histone acetyltransferase to the X chromosome (Birchler et al. 2001,2003; Hiebert and Birchler 1994). This sequestration would remove H4K16Ac and its transcriptional activation effects from the autosomes to the X chromosome. This loss of H4K16Ac would cause a drop in gene expression to near that in females and so rescue the autosomes from an otherwise hyperactive state.

Some experimental support for this model comes from work by Bhadra et al. In some msl mutant males, binding of the MSL complex to the X chromosome is disrupted and the complex becomes associated with all the chromosomes. This results in an increase in the level of acetylation on the autosomes and a corresponding increase in the level of gene expression (Bhadra et al. 1999, 2000). Moreover, gene expression experiments in mle or mof mutants, which lack an MSL complex, revealed that transgenes on the X chromosome remained dosage compensated; however, many autosomal transgenes are upregulated (Birchler 1996). Finally, ectopic expression of MSL-2 in females induces MSL complex formation, which, according to the previous model, should cause an increase in expression of genes on the X chromosome. This, however, is not so. Bhadra and coworkers did not detect any increased expression from X-chromosomal transgenes tested (Bhadra et al. 1999, 2000).

In this model, most of the X chromosome is automatically compensated, but having attracted excessive amounts of H4K16Ac, a mechanism is needed to counteract its activating effects. Similar to results obtained by Corona et al., Birchler's group has shown that X-linked genes show increased expression in ISWI mutant individuals (Corona et al. 2002; Pal Bhadra et al. 2005). This has led them to propose that it is the repressive actions of ISWI that limit the hyperactivating effects ofhyperacetylation of the X chromosome (Pal Bhadra et al. 2005).

It is not completely clear whether there is one true dosage compensation system. There is experimental evidence to support both possibilities. Further characterisation of the members of the DCC and identification of interacting proteins will help to elucidate this important mechanism of large-scale gene regulation.

Acknowledgements We would like to thank members of the lab for critical reading of the manuscript. Mikko Taipale for use of his template in Fig. 1. S.R. was supported by an EMBO long term fellowship and is currently supported by an HFSP fellowship. Research in our lab is partially funded by DFG; SFB "Transregio 5" and EU NoE "Epigenome".

0 0

Post a comment