Development and engineering of RNA aptamers for riboswitching and biosensing

RNAs can form highly complex three-dimensional structures that allow them to bind target molecules with high affinity and specificity. These so-called RNA aptamers are not only found in nature but can also be selected in vitro using a method called SELEX. In theory, this method allows the selection of aptamers for any target, such as small molecules, proteins and even cells. For this selection process, a target is immobilized and incubated with a randomized pool of RNA sequences. During a washing step, only those sequences remain that bound to the target, while non-binding ones are removed. The bound RNA sequences are then recovered, amplified, and used for additional cycles. This process is repeated until the original RNA pool is enriched with binding sequences. The presented work employs a relatively novel variation of the SELEX protocol called RNA Capture-SELEX. In this method, the randomized pool of RNA sequences is immobilized and not the target of the selection. This is achieved via a capture oligonucleotide that binds a constant sequence located within the randomized part of each sequence in the pool. This offers two key advantages over the traditional protocol. Firstly, the target of the selection remains in solution and is not chemically immobilized. Especially the structure of smaller molecules can be drastically changed through such a chemical immobilization, which can lead to less successful and unspecific enrichments. Secondly, to be released from the capture oligonucleotide, the aptamer must undergo a structural change upon target binding. This means that RNA Capture-SELEX innately selects RNA aptamers that change stability and structure depending on the presence or absence of a ligand. This structural change can be exploited for the design and selection of synthetic RNA constructs such as riboswitches, ribozymes, and biosensors. The first part of this work consists of the characterization and engineering of a new, tobramycin binding riboswitch. It is one of the best performing constructs in its class of translation initiation controlling riboswitches and even rivals the tetracycline riboswitch. The riboswitch sequence that was initially selected contained two independent tobramycin binding sites with Kds of 1.1 nM and 2.4 μM and it is capable of differentiating tobramycin from the closely related substances kanamycin A and B. With mutational analysis, ITC and 1H NMR spectroscopy the secondary structure of the riboswitch could be established, and it was found that the two binding sites are located in different parts of the structure. The two binding sites were turned into separate constructs and separately analyzed. The high affinity binding site with a Kd of 1.1 nM was solely responsible for the observed regulation and engineered into a minimal riboswitch sequence with a size of only 33 nt and good performance. The second part of this work comprised of the selection and characterization of a caffeine binding RNA aptamer and its use in a riboswitch, aptazyme and biosensor. After in vitro selection and in vivo yeast screening, the riboswitch was systematically shortened. In a doped pool screening approach, the riboswitch was partially randomized, but it was found that the originally selected sequence was already close to optimal. The caffeine binding aptamer was then used for the engineering of a hammerhead ribozyme based aptazyme. The aptazyme achieved a 2x gene regulation when placed in a 3’ UTR in yeast. The aptamer was then grafted onto an iSpinach fluorogenic aptamer and thereby turned into a fluorescence-based biosensor that could detect caffeine above a concentration of 100 µM. The third part of this work focused on a proof-of-concept for a modular, aptamer-based lateral flow assay. For this, a levofloxacin binding aptamer, that was also selected using RNA Capture-SELEX, was used as a bioreceptor. The biosensor made use of the principle that aptamers selected with RNA Capture-SELEX displace a DNA capture oligonucleotide when binding a ligand. The lateral flow assay specifically detected levofloxacin and could distinguish it from the closely related compound ciprofloxacin. With a limit of visual detection of 100 µM sample concentration, the assay is not sensitive enough for a direct use in environmental testing. However, the presented assay establishes a proof-of-concept for the development of biosensors based on RNA Capture-SELEX selected aptamers.