Date of Degree


Document Type


Degree Name





Kevin Ryan

Committee Members

Prabodhika Mallikaratchy

Nancy Greenbaum

Mark Emmerson

Ian M. Willis

Subject Categories

Life Sciences




Synthetic DNA-based small RNA expression vectors offer potential advantages over chemically synthesized RNA, such as lower cost, greater chemical stability, easier synthesis and longer shelf life. We are working to develop a novel biomedical technology that employs synthetic single-stranded circular oligonucleotides (coligos) as promoter-independent small RNA expression vectors. Coligos were initially designed to code for pre-miRNA mimics and were prepared by circularizing the first-strand cDNA sequence (i.e., the template strand) of selected genomic human pre-miRNAs. However, we are exploring coligos as general expression vectors for a variety of small RNAs used in biomedical research. To efficiently use coligos as a small RNA expression platform, and produce small RNA of defined sequence, it is important to be able to accurately predict and control transcription initiation and termination. Coligo transcription initiation is promoter-independent; previous studies suggest that transcription initiation is in part driven by secondary structure features and, to a lesser extent, sequence. Transcription termination is less clear and prone to produce readthrough transcripts. Here we present the results of two in vitro selections designed to uncover secondary structure and sequence features that can lead to precise and efficient promoter-independent transcription initiation of coligo templates by RNA polymerase III. Alpha-amanitin inhibition studies and immunoprecipitation experiments have demonstrated that RNA Pol III is involved in coligo transcription. A previously identified “initiation motif” was randomized to produce combinatorial libraries that were used to simultaneously test multiple coligo structures and sequences. In the first in vitro selection, 12 nucleotides in the initiation motif were randomized to produce a coligo library that contains 412 (about 16.7 million) unique sequences, each represented about 720,000 times in a 20 pmol sample used in each experiment. In the second in vitro selection, two juxtaposed 3-nucleotide sequences adjacent to the coligo stem were randomized, in a more focused region, to produce a coligo library that contains 46 (about 4,096) unique sequences, each represented about 2.94 x 109 times in a 20 pmol sample. A multiple-round in vitro selection method was designed and optimized. After multiple rounds of selection, the final selected libraries were sequenced through next-generation sequencing (NGS). The most highly represented sequences were analyzed and the secondary structure of their corresponding coligos predicted in silico. The results of the in vitro selections confirm that transcription initiation is driven in part by the secondary structure of the initiator motif, independently of the coligo sequence. The secondary structures of the most-represented sequences share similar characteristics. Coligo-seq, a small RNA-seq method variation developed in our laboratory, was used to sequence the transcripts produced by the selected coligo library. NGS sequencing results suggest that RNA Pol III preferentially initiates transcription in a templated pyrimidine located within the initiation motif of the coligo template. Here we also present the results of our attempt to identify non-natural nucleotides that lead to site-specific coligo transcription termination. A non-natural nucleotide residue was included in the coligo template to induce transcription termination in a region already prone to provoke transcription termination by RNA Pol III. The non-natural nucleotides were either incorporated during synthesis or modified post-synthetically. Various non-natural nucleotides were tested; some have a significant effect on transcription termination, but also had effects on initiation, perhaps because they were placed structurally close to the initiation site. In this thesis, we present the progress made in our overall goal to construct a synthetic coligo platform template that produces a small RNA transcript with well-defined 5’ and 3’ ends.

Included in

Life Sciences Commons