Pumilio/fem-3 mRNA-binding factor (PUF) proteins possess a recognition code for bases A, U, and G, allowing designed RNA sequence specificity of their modular Pumilio (PUM) repeats. However, recognition side chains in a PUM repeat for cytosine are unknown. Here we report identification of a cytosine-recognition code by screening random amino acid combinations at conserved RNA recognition positions using a yeast three-hybrid system. This C-recognition code is specific and modular as specificity can be transferred to different positions in the RNA recognition sequence. A crystal structure of a modified PUF domain reveals specific contacts between an arginine side chain and the cytosine base. We applied the C-recognition code to design PUF domains that recognize targets with multiple cytosines and to generate engineered splicing factors that modulate alternative splicing. Finally, we identified a divergent yeast PUF protein, Nop9p, that may recognize natural target RNAs with cytosine. This work deepens our understanding of natural PUF protein target recognition and expands the ability to engineer PUF domains to recognize any RNA sequence.The specific interaction of RNA and protein plays vital roles in RNA regulation including splicing, localization, translation, and degradation. Such recognition may be directed toward unstructured RNA requiring discrimination of RNA sequences, folded RNA motifs, or some combination of sequence and structural specificity (1). Members of the PUF 2 protein family (named after Drosophila Pumilio and Caenorhabditis elegans fem-3 mRNA-binding factor (FBF)) are sequence-specific RNA-binding proteins that regulate networks of mRNAs encoding proteins of related function (2-7). PUF proteins generally recognize the 3Ј-UTR of their target mRNAs to control the mRNA stability and translation (2-7).The RNA-binding domain of PUF proteins, known as the Pumilio homology domain (PUM-HD) or PUF domain, can bind to unstructured RNA sequences in a distinct fashion. The PUF domain of human Pumilio 1 contains eight PUM repeats, each containing three ␣-helices packed together in a curved structure (8 -10). RNA is bound as an extended strand to the concave surface of the PUF domain with the bases contacted by protein side chains. In general, each PUM repeat recognizes a single RNA base through the second helix (␣2) in an antiparallel arrangement, i.e. nucleotides 1-8 are recognized by PUF repeats 8 -1, respectively. The ␣2 helices of PUM repeats contain a 5-residue sequence, designated here as 12XX5, where the side chain at position 2 stacks with the recognized base and the side chains at positions 1 and 5 recognize the edge of the base (8, 11) (see Fig. 1A). Specific residues at these positions direct the base recognition properties of the repeat. This PUF-RNA recognition code makes it possible to modify a PUM repeat to bind a particular RNA base, producing a designed PUF domain that specifically recognizes a given 8-nucleotide RNA target. Such de novo designed RNA binders have been used to track RNA local...