Julia Haas 1, Johann Roque 1, Beth Anderson 1, Ariel Thelander 1, Will Ford 2, Aparajita Chakraborty 3, Paul Robustelli 3, Blanton Tolbert 2, and Brittany Morgan 1.

University of Notre Dame, Department of Chemistry & Biochemistry 1, University of Pennsylvania, Department of Biochemistry & Biophysics 2, and Dartmouth College, Department of Biochemistry and Cell Biology 3.

Abstract:
RNA Binding Proteins (RBPs) control many cellular processes including transcription, translation, and alternative splicing. RBPs have the most disease annotated mutations of any protein class, but these proteins have historically been considered “unligandable” due to their i) highly dynamic and disordered nature and ii) high level of conservation within their RNA binding domains (RBDs). We hypothesized that differences in protein dynamics could be exploited to enable covalent ligands to selectively target conserved cysteines within RBDs. To investigate this, we screened a library of cysteine-reactive covalent ligands against homologous RBPs with differences in their protein dynamics: heterogeneous nuclear ribonucleoproteins H and F (hnRNP H and F). hnRNPs H and F are an excellent model system for RBP conservation because they have over 89% sequence similarity, three parallel quasi-RNA recognition motifs (qRRMS), and three conserved cysteines (C22, C34, & C122) within their RBDs. Despite this conservation, we observed differences in the covalent labeling of hnRNP H and F during in vitro screenings. Notably, we found one small molecule with a ten-fold difference in selectivity for hnRNP H over hnRNP F and we consistently observed greater covalent labeling of cysteines in hnRNP H. This work demonstrates that covalent ligands can be used to differentially engage conserved cysteines. Moreover, our findings suggest that the unique dynamics of hnRNP H, specifically slower conformational exchange, may contribute to greater covalent labeling. By understanding how protein dynamics contribute to covalent selectivity, we can cultivate a powerful new approach for targeting RBPs and other dynamic and disordered proteins.

Rajat Mudgal1, Dishari Thornhill1, Kai-Neng Chou1, Akira Ono1

1Dept. of Microbiology and Immunology, University of Michigan Medical School, Ann Arbor, MI

Abstract:
HIV-1 continues to cause mortality and morbidity to people living with HIV despite effective antiretrovirals. A multifunctional viral structural protein Gag drives HIV-1 particle formation at the plasma membrane. Matrix (MA) and nucleocapsid (NC) are two major Gag domains that bind RNA during HIV-1 particle assembly. RNA binding to MA highly basic region (MA-HBR) is hypothesized to prevent premature or promiscuous membrane binding of Gag. The vast majority of MA-bound RNAs in cells are tRNAs, but the details of Gag-tRNA binding remain to be determined.
In the current study, we applied modified photoactivatable ribonucleoside-enhanced cross-linking and immunoprecipitation (PAR-CLIP) approaches to an NC-deficient Gag variant (delNC) and full-length Gag. We found that most tRNAs that bind to membrane-binding-incompetent (MBI) version of delNC Gag were those present abundantly in the cytosol. However, selenocysteine tRNA (SeCTCA), a less abundant tRNA, was over-represented in the MBI delNC Gag-bound tRNAs. Importantly, the tRNAs bound less to membrane-binding-competent delNC Gag than to MBI delNC Gag, suggesting mutual exclusivity between Gag-membrane and Gag-tRNA binding. In the context of full-length Gag, SeCTCA binds only MA-HBR, while other tRNAs bind both MA-HBR and NC. Preliminary data showed that overexpression of SeCTCA in a T cell line significantly decreased HIV release, suggesting a physiological role of SeCTCA in HIV-1 life cycle. Ongoing studies are aimed at elucidating the mechanism by which SeCTCA inhibits HIV-1 particle release and identifying the interaction interphase between SeCTCA and MA-HBR, which can be exploited as a potential target of novel RNA-based antiviral therapeutic strategies.

Y. Yuan1, A. Linskens1, R. De Gouvea2, H. Liu2, Y. Xiao2, C. Hsieh1, S. Barmada1, S. Yadlapalli1,2

Cellular and Molecular Biology Training Program1 and Department of Cell and Developmental Biology2, University of Michigan Medical School, Ann Arbor, Michigan 48109, USA

Abstract:
Circadian clocks are cell-autonomous timekeepers that regulate ~24-hour rhythms in gene expression. While basic mechanisms of circadian transcription-translation feedback are well-understood, spatiotemporal dynamics of clock transcripts and their processing remains unexplored.

Here, we demonstrate that a specific intron (P) of timeless (tim), a Drosophila clock gene, is post-transcriptionally spliced at nuclear speckles, which serves as a novel rate-limiting mechanism in circadian clocks by introducing a delay between transcription of tim mRNA and accumulation of its translatable form in the cytoplasm. Upon transcription, tim mRNAs localize to nuclear speckles, where ‘P’ intron is spliced before nuclear export. Removing this intron causes cytoplasmic accumulation of tim mRNAs, faster accumulation of TIM protein and shortening of the clock to ~22 hours. Inclusion of this ‘P’ intron into reporter minigene transcripts is sufficient for their nuclear localization in both Drosophila neurons and human U2OS cells, and, notably, addition of a 6-bp canonical branchpoint sequence to the intron changes its splicing from post-transcriptional to co-transcriptional. Knockdown of spliceosomal factors, including Prp3 and Prp19, specifically impairs post-transcriptional splicing of the ‘P’ intron, resulting in elevated nuclear retention of tim mRNAs, decreased TIM protein abundance, and disrupted clock. Finally, we found that post-transcriptional splicing at nuclear speckles is a conserved mechanism in other transcripts, including an ALS risk factor unc-13.

Together, our results reveal that post-transcriptional splicing of the timeless ‘P’ intron at nuclear speckles is a rate-limiting step in circadian clocks. Moreover, spatiotemporal dynamics of intron splicing may play underappreciated roles in gene regulation at large.