Abstract for the 2017 HEI Annual Conference
The role of air pollution on RNA oxidative stress, characterization of stress-response enzymes, and applications toward RNA-based biosensors
Lydia M. Contreras
University of Texas–Austin, USA
Background Given the importance of various RNAs in regulating cellular (and tissue) function, we are investigating potential chemical changes that are induced in mRNAs and regulatory miRNAs upon exposure to urban air mixtures. Current approaches for researching mechanistic consequences of air pollution induced stress are generally focused on measuring inflammation via expressed protein biomarkers (i.e.interleukin-8). We have previously shown that levels of 8-oxoguanosine (8OG) may present a more immediate and consistent measure of cellular stress in air pollution exposure models. However, a mechanistic understanding of the role for RNA oxidation in acute air pollution stress responses has not been investigated. The three main objectives of this work are therefore to: (1) Establish robust modified RNA immunoprecipitation sequencing (modRIPseq) protocols for identifying functionally relevant RNA oxidations in air pollution stress models, (2) Identify protein scaffolds for high affinity binding of chemically damaged RNAs, and (3) Engineer novel protein biosensors for improved efficiency in binding various RNA oxidation products to expand repertoire of modifications for modRIPseq.
Methods We exposed BEAS-2B lung cell cultures (N=3) to high levels of simulated air pollution mixtures (4ppm ozone, acrolein, methacrolein) for 90 minutes. Total RNAs for both pollution-exposed and clean-air-control samples were extracted and examined for 8OG enrichment. Through an established collaboration, these samples will also be analyzed using mass spectrometry to characterize and quantify 8OG and other oxidation products. To identify and characterize protein scaffolds for our biosensor work, we conducted electrophoretic mobility shift assays (EMSAs) of candidate proteins and the most prevailing oxidation lesions in nucleobases (8-oxoG, 8-oxoA, 5-OHC, and 5-OHU), as well as one of the most common methylation products (5mC), with n = 3. 25-mers with random sequences were synthesized each containing 6 modifications equally distributed through the oligonucleotide. Lastly, for the engineering of new biosensors, we have been developing protocols for use with two potential selection/library generation techniques- ribosome display and phage display. We have built constructs for both methods. We have performed Western blotting, PAGE-SDS and RNA electrophoresis analyses to validate successful expression of the proteins of interest.
Results Initial analyses using a genomic feature counting approach and DESeq2 highlighted differential oxidation of nuclear-localized mRNAs, long noncoding RNAs, and transcripts for several members of splicing machinery, leading to a secondary bottom up transcriptome assembly approach to investigate potential differential splicing of some genes. Network analysis also showed significant enrichment of transcripts involved in inflammatory response pathways and signaling pathways, among others. We have also shown one protein to discriminate RNAs containing either 8-oxoG, 8-oxoA or 5mC from unmodified RNA. However, it preferentially binds to 8-oxoG, with the apparent constant of dissociation (KD) decreasing by a 2-fold factor relative to the unmodified RNA. We have characterized the contribution of all the domains of this protein to the recognition (binding and specificity) of oxidized RNAs. Lastly, we have established and validated system expression to make variant libraries of a bacterial protein for screening against multiple RNA modifications in an effort to select for variants with higher affinity toward modified RNA nucleosides.