Structure of the human 20S U5 snRNP.

The 20S U5 small nuclear ribonucleoprotein particle (snRNP) is a 17-subunit RNA-protein complex and a precursor of the U4/U6.U5 tri-snRNP, the major building block of the precatalytic spliceosome. CD2BP2 is a hallmark protein of the 20S U5 snRNP, absent from the mature tri-snRNP. Here we report a high-resolution cryogenic electron microscopy structure of the 20S U5 snRNP, shedding light on the mutually exclusive interfaces utilized during tri-snRNP assembly and the role of the CD2BP2 in facilitating this process.

Characterization of the SF3B1-SUGP1 interface reveals how numerous cancer mutations cause mRNA missplicing.

The spliceosomal gene SF3B1 is frequently mutated in cancer. While it is known that SF3B1 hotspot mutations lead to loss of splicing factor SUGP1 from spliceosomes, the cancer-relevant SF3B1-SUGP1 interaction has not been characterized. To address this issue, we show by structural modeling that two regions flanking the SUGP1 G-patch make numerous contacts with the region of SF3B1 harboring hotspot mutations. Experiments confirmed that all the cancer-associated mutations in these regions, as well as mutations affecting other residues in the SF3B1-SUGP1 interface, not only weaken or disrupt the interaction but also alter splicing similarly to SF3B1 cancer mutations. Finally, structural modeling of a trimeric protein complex reveals that the SF3B1-SUGP1 interaction "loops out" the G-patch for interaction with the helicase DHX15. Our study thus provides an unprecedented molecular view of a protein complex essential for accurate splicing and also reveals that numerous cancer-associated mutations disrupt the critical SF3B1-SUGP1 interaction.

Structural basis of branch site recognition by the human spliceosome.

Recognition of the intron branch site (BS) by the U2 small nuclear ribonucleoprotein (snRNP) is a critical event during spliceosome assembly. In mammals, BS sequences are poorly conserved, and unambiguous intron recognition cannot be achieved solely through a base-pairing mechanism. We isolated human 17S U2 snRNP and reconstituted in vitro its adenosine 5´-triphosphate (ATP)­dependent remodeling and binding to the pre­messenger RNA substrate. We determined a series of high-resolution (2.0 to 2.2 angstrom) structures providing snapshots of the BS selection process. The substrate-bound U2 snRNP shows that SF3B6 stabilizes the BS:U2 snRNA duplex, which could aid binding of introns with poor sequence complementarity. ATP-dependent remodeling uncoupled from substrate binding captures U2 snRNA in a conformation that competes with BS recognition, providing a selection mechanism based on branch helix stability.

Critical role of the finger loop in arrestin binding to the receptors.

We tested the interactions with four different G protein-coupled receptors (GPCRs) of arrestin-3 mutants with substitutions in the four loops, three of which contact the receptor in the structure of the arrestin-1-rhodopsin complex. Point mutations in the loop at the distal tip of the N-domain (Glu157Ala), in the C-loop (Phe255Ala), back loop (Lys313Ala), and one of the mutations in the finger loop (Gly65Pro) had mild variable effects on receptor binding. In contrast, the deletion of Gly65 at the beginning of the finger loop reduced the binding to all GPCRs tested, with the binding to dopamine D2 receptor being affected most dramatically. Thus, the presence of a glycine at the beginning of the finger loop appears to be critical for the arrestin-receptor interaction.

Structural basis of arrestin-3 activation and signaling.

A unique aspect of arrestin-3 is its ability to support both receptor-dependent and receptor-independent signaling. Here, we show that inositol hexakisphosphate (IP6) is a non-receptor activator of arrestin-3 and report the structure of IP6-activated arrestin-3 at 2.4-Å resolution. IP6-activated arrestin-3 exhibits an inter-domain twist and a displaced C-tail, hallmarks of active arrestin. IP6 binds to the arrestin phosphate sensor, and is stabilized by trimerization. Analysis of the trimerization surface, which is also the receptor-binding surface, suggests a feature called the finger loop as a key region of the activation sensor. We show that finger loop helicity and flexibility may underlie coupling to hundreds of diverse receptors and also promote arrestin-3 activation by IP6. Importantly, we show that effector-binding sites on arrestins have distinct conformations in the basal and activated states, acting as switch regions. These switch regions may work with the inter-domain twist to initiate and direct arrestin-mediated signaling.