RNA motif discovery by SHAPE and mutational profiling (SHAPE-MaP). Human immunodeficiency virus 1 strain:NL4-3

Many central biological processes are RNA structure-mediated, but the higher-order structures of most RNAs are unknown, making it difficult to understand how RNA structure governs biological function. Here we measure sites of 2'-hydroxyl acylation by SHAPE reagents in a single direct step by massively parallel sequencing. We exploit the ability of reverse transcriptase to misread a SHAPE-modified nucleotide and to incorporate a nucleotide non-complementary to the original sequence in the newly synthesized cDNA. SHAPE mutational profiling (SHAPE-MaP) makes possible de novo identification of functional motifs based on the extent and entropy of the underlying RNA structure. Structure modeling based on fully automated analysis recovered greater than 90% of accepted base pairs in complex RNAs of known structure and was used to define a second-generation model for the structure of an HIV-1 RNA genome. The HIV-1 model contains all known large-scale structured motifs, identified previously unknown elements, and allowed us to define four pseudoknotted structures, all validated by viral replicative fitness or RNA structure perturbation. RNA elements with functional roles in viral replication are overwhelmingly located in genome regions with high levels of well-determined structure (p = 0.002). SHAPE-MaP is independent of sequencing platform and library construction scheme, can interrogate both local and large-scale RNA features, yields accurate and high-resolution secondary structure models, detects scarce and low abundance RNAs, allows the structural consequences of sequence polymorphisms to be disentangled in single experiments, and will ultimately democratize RNA structure analysis.