(610a) Biomedical Potential of Viral and Endogenous Double-Stranded RNA Detection

Detection of viruses, especially the ones related to infectious diseases, has attracted much attention. In recent years, we have seen an increasing number of viral infection outbreaks, such as Middle East respiratory syndrome coronavirus and Zika virus, and they are spreading on a global scale at an increasingly alarming rate. Nucleic acid based methods are the most commonly employed schemes for virus detection. This includes the use of polymerase chain reaction (PCR), which is routinely performed to amplify and detect viral nucleic acids. While nucleic acid-based detection is highly specific, it has the disadvantage that some prior knowledge of the virus is needed in order to prepare the relevant detection methods. Such approach can be limiting when faced with novel or mutated viruses of unknown genomic sequences. Considering that in many situations it is necessary to determine the infection status of a person without having much information regarding the nature of the virus involved, it would be highly desirable to have a primary screening tool capable of detecting many different viruses without knowing their nucleotide sequences. To design these universal virus detectors, it is important to identify biomarkers that are commonly present across many viruses. Recently, the use of long dsRNAs as a universal virus marker has gained popularity.

Long dsRNAs are generated as byproducts of viral replication and are produced abundantly by DNA viruses and positive-strand RNA viruses, and also by a limited number of negative-strand RNA viruses. Moreover, certain viruses such as reovirus, rotavirus, and cypovirus are known to contain dsRNA genome. As these long dsRNAs are a common signature of viruses, they are recognized by mammalian innate immune system and trigger a cascade of immune response reactions. Particularly, dsRNA sensors of the innate immune response system recognize the length and the terminal structure of the dsRNAs rather than their specific sequences, thus these immune response proteins are capable of recognizing a wide range of viruses that generate long dsRNAs.

In this study, we utilize a combination of dsRNA recognizing J2 antibody and reactive poly(pentafluorophenyl acrylate) (PPFPA) to construct a virus detection platform with superior sensitivity. PPFPA attached silicon substrates are first prepared, and the reactive PFP units are utilized to immobilize J2 antibodies that recognize dsRNAs longer than 40 base pairs (bp). We show that these surfaces can successfully distinguish dsRNAs from single-stranded RNAs (ssRNAs), as well as synthetic long dsRNAs (76 bp) from short ones (19 bp), demonstrating their potential in specifically detecting only the long dsRNAs generated by viruses. Furthermore, to enhance dsRNA detection sensitivity, we devise a two-step detection process analogous to sandwich enzyme-linked immunosorbent assay (ELISA) where following dsRNA capture, the bound dsRNAs are then visualized using fluorophore-tagged antibodies that also recognize RNAs’ double-stranded secondary structure. By utilizing the developed platform, long dsRNAs can be detected and visualized from total RNA mixture as well as from total cell lysates, which contain a mixture of various abundant contaminants such as DNAs, proteins, and other cellular ss- and dsRNAs. Lastly, we demonstrate the application of the developed platform in detecting long dsRNAs generated by hepatitis C virus (HCV) and hepatitis A virus (HAV) without utilizing information on viral RNA sequences.

Although dsRNAs are strongly associated with immune response to viral infection, increasing evidences suggest that human cells naturally express endogenous dsRNAs that can regulate antiviral machineries in various cellular contexts such as during the cell cycle and response to stressors. Moreover, accumulation of endogenously encoded dsRNAs is related to the onset of autoimmune and age-related macular degeneration. Lastly, dsRNAs also play a key role during cellular response to chemotherapy where treatment of the DNA-demethylating agent leads to cell death by inducing the transcription of endogenous dsRNA genes, which subsequently activate antiviral machineries. Considering the biomedical potential of dsRNAs, we further develop a small organic molecule to quantify the collective expression level of all types of dsRNAs expressed in human cells.

We investigate the potential of using photochromic spiropyrans as a tool to detect and profile the overall expression of dsRNAs. Our goal is three folds: 1) To establish spiropyran as a molecule that can interact with dsRNAs, 2) To characterize the interactions between dsRNA and spiropyran, and 3) To apply spiropyran to detect changes in levels of endogenous dsRNAs in response to various stressors. We find that the open form of spiropyran, merocyanine (MC), can moderately interact with dsRNAs, which results in protonation of MC to MCH+ and alters the characteristic UV-Vis absorbance spectrum of the compound. By quantifying this change, we can access the amount of dsRNA present in the solution. The effect is greater for the GC pair than that of the AU pair and mixture of the two types of nucleotide pairs also show strong characteristic decrease in the light absorbance. Moreover, using RNases with different substrate specificities, we can enrich and detect long dsRNAs by removing most of the ssRNAs and short hairpin RNAs from the cell extract. Furthermore, we apply our approach to demonstrate the potential usage of MC and elevation of dsRNA expression as a predictive marker for cellular response to the DNA-demethylating agent 5-aza-2'-deoxycytidine (decitabine), which is commonly used to treat colorectal cancer and pre-leukemic disorder myelodysplasia (MDS).

Collectively, our work presents the first example of polymer grafted surfaces as well as small organic molecule to capture and detect long dsRNAs. These findings establish potential of dsRNA detection in biomedical applications ranging from viral infection to cancer