Frequently during RNA processing, sections of RNA, called introns, must be spliced out and the flanking sequences rejoined. A type of intron known as Group I is remarkable in that the intron RNA is actually the catalyst for the excision reaction (Cech et al., 1981). RNA catalysts are known as ribozymes.

My laboratory studies group I introns. Some group I introns can perform this reaction in vitro and are said to be self-splicing. Others require the assistance of proteins; the few identified thus far stabilize or help fold the intron RNA (Weeks and Cech, 1996; Caprara et al., 2001). Mutational analysis has shown that the excision of some group I introns depends upon a maturase protein encoded by the intron itself (Lazowska et al., 1980). The first biochemical assay for maturase activity was developed in our laboratory: a maturase from Aspergillus nidulans directly and significantly facilitates excision of an intron with limited self-splicing activity (Ho et al, 1997). The intron (AnCOB) is located in the apocytochrome b gene. A high resolution secondary structure may be seen as a pdf file.

Interestingly the maturase protein can also be classified as belonging to a group of proteins known as DNA homing endonucleases (Dujon, 1989). These are generally encoded in introns (or inteins) of about 1,000 base pairs. The proteins recognize (or home in on) a target sequence of 14 - 40 base pairs that corresponds to an intronless version of the DNA sequence flanking their own intron. They cleave this site and initiate the process of inserting a copy of their intron sequence into the cut site. By convention the AnCOB intron-encoded homing endonuclease is named I-AniI. A large number of DNA endonucleases similar to I-AniI have been identified but the majority do not appear to have maturase activity. The defining characteristics of a maturase are not evident from sequence analysis. It is commonly believed that the RNA splicing activity evolved from a protein which originally functioned solely as a DNA endonuclease (Belfort, 1990; Lambowitz and Perlman, 1990). However it is unlikely that this new function arose simply by fusion of an additional gene sequence. The raises the general question "how does a protein acquire a new function without increasing in size and how easily can this can be detected?". This is particularly relevant to annotating genomes. Attribution of function to a gene on the basis of similarity of either amino acid sequence or tertiary structure to a previously characterized protein may constitute only partial characterization of the sequence since an unknown number of proteins may "moonlight" and perform a second, unrelated and unforeseen task (Jeffery, 1999).

We are curious to learn more about how such proteins develop an additional function. Intriguingly a number of well-known transcription factors not only bind DNA but also have RNA-binding properties (Cassiday and Maher, 2002). It is clearly of interest to understand the origins and the structures of RNA and DNA binding sites on the same protein. Biochemical and genetic data from our lab suggest that the two sites for binding RNA and DNA on I-AniI are distinct although there may be some overlap (Geese et al., 2003; Bolduc et al., 2003).

Dr. Barry Stoddard (Fred Hutchinson Cancer Research Center) along with Dr Mark Caprara (Center for RNA Molecular Biology, Case Western Reserve University Medical School) have collaborated with us to solve the molecular structure of I-AniI complexed with its DNA substrate at a resolution of 2.6 (Bolduc et al., 2003). The right figure marks regions of clustered positively charged amino acids conserved between three maturases. These are potential regions on the protein that might be utilized to bind RNA and one has been identified as specifically required for RNA splicing but not DNA cleavage (Bolduc et al., 2003). The overall three dimensional structures of I-AniI and I-CreI, a prototype homing endonuclease (Heath et al., 1997), are surprisingly similar given the dissimilarity of their amino acid sequences. This now makes identifying the RNA binding determinants of I-AniI all the more intriguing since I-CreI does not have maturase activity. We are now particularly interested in determining the molecular structure of the maturase complexed to its group I intron RNA.