Concluding Remarks

The adaptation of homing endonucleases to function in group I intron splicing is reminiscent of other proteins that bind both DNA and RNA, such as the homeodomain containing Bicoid protein, T4 DNA polymerase, and p53 tumor suppressor protein (Jeffery 1999; Cassiday and Maher 2002). Many of these so-called moonlighting proteins include factors involved in DNA replication or transcription that also bind mRNAs to repress translation. The determinants for both RNA and DNA binding are not known in almost all cases; however, there is evidence that ligand-binding sites overlap in transcription factor IIIA

from Xenopus (reviewed in Cassiday and Maher 2002). Although the hypothesis that the DNA-binding sites of many of these moonlighting proteins are involved in their RNA function is very popular, our experience with homing endonuclease/maturases suggest that we can expect some surprises in this regard. Notably, it will be of great interest to see how widespread the phenomenon is of utilizing a "benign" but susceptible protein surface to develop a second function.

What remains to be learned from group I introns and their maturases? Although maturase and endonuclease function appear separate, the challenge remains to understand how members of this protein class bind with such high specificity to their cognate introns. Furthermore, the RNA-folding pathway(s) mediated by binding that leads to intron catalysis is unknown. To complicate matters, it is not clear if all maturases use the same mechanism of splicing facilitation. In particular, introns encoding maturases belong to separate structural subgroups and the maturases themselves show great divergence (Michel and Westhof 1990; Dalgaard et al. 1997). Furthermore, identical mutations in individual maturases can have opposite effects (e.g., Henke et al. 1995; Szczepanek et al. 2000). It will be essential to answer questions of maturase function by comparing and contrasting the activities of individual maturases and to identify common "maturase-motifs" if they exist. The identification of new group I intron maturases via genetic or biochemical means will be useful as the presence of the LAGLIDADG motif is not itself predictive of its RNA splicing or endonuclease function. Furthermore, it is not currently known if members of the other three major classes of intron-encoded en-donucleases have adapted to facilitate splicing (Lambowitz et al. 1999; Chevalier and Stoddard 2001).

Finally, homing endonucleases show great, natural flexibility by changing their target specificity as a means to ensure their survival. This property is being exploited with regard to engineering enzymes with novel sequence specificities to be used in gene therapy applications (Chevalier and Stoddard 2001; Gimble, this Vol.). It will be of interest to see if the same adaptability holds true for maturase RNA recognition strategies. If so, it may be possible to use maturases as scaffolds to engineer new sequence- or structure-specific RNA-binding proteins for therapeutic or gene expression studies (e.g. Campisi et al. 2001).

Acknowledgements. We thank Kristina Brady for help in manuscript preparation. Research in the authors' laboratories is supported by National Institutes of Health grant GM-62853 to M.G.C. and National Science Foundation grant MCB-013099 to R.B.W.

0 0

Post a comment