Specific alterations of U1-C protein or U1 small nuclear RNA can eliminate the requirement of Prp28p, an essential DEAD box splicing factor.

While some members of the ubiquitous DExD/H box family of proteins have RNA helicase activity in vitro, their roles in vivo remain virtually unknown. Here, we show that the function of an otherwise essential DEAD box protein, Prp28p, can be bypassed by mutations that alter either the protein U1-C or the U1 small nuclear RNA. Further analysis suggests that the conserved L13 residue in the U1-C protein makes specific contact to stabilize the U1 snRNA/5' splice site duplex in the prespliceosome, and that Prp28p functions to counteract the stabilizing effect of the U1-C protein, thereby promoting the dissociation of the U1 small nuclear ribonucleoprotein particle from the 5' splice site. Thus, in addition to unwinding RNA, the DExD/H box proteins may affect RNA-RNA rearrangements by antagonizing specific RNA-stabilizing proteins.

An RNA switch at the 5' splice site requires ATP and the DEAD box protein Prp28p.

Pre-mRNA splicing requires dramatic RNA rearrangements hypothesized to be catalyzed by ATP-dependent RNA unwindases of the DExD/H box family. In a rearrangement critical for the fidelity of 5' splice site recognition, a base-pairing interaction between the 5' splice site and U1 snRNA must be switched for a mutually exclusive interaction between the 5' splice site and U6 snRNA. By lengthening the U1:5' splice site duplex, we impeded this switch in a temperature-dependent manner and prevented formation of the spliceosome's catalytic core. Using genetics, we identified the DExD/H box protein Prp28p as a potential mediator of the switch. In vitro, the switch requires both Prp28p and ATP. We propose that Prp28p directs isomerization of RNA at the 5' splice site and promotes fidelity in splicing.

Formation of a native-like subdomain in a partially folded intermediate of bovine pancreatic trypsin inhibitor.

In the folding of bovine pancreatic trypsin inhibitor (BPTI), the single-disulfide intermediate [30-51] plays a key role. We have investigated a recombinant analog of [30-51] using a 2-dimensional nuclear magnetic resonance (2D-NMR). This recombinant analog, named [30-51]Ala, contains a disulfide bond between Cys-30 and Cys-51, but contains alanine in place of the other cysteines in BPTI to prevent the formation of other intermediates. By 2D-NMR, [30-51]Ala consists of 2 regions-one folded and one predominantly unfolded. The folded region resembles a previously characterized peptide model of [30-51], named P alpha P beta, that contains a native-like subdomain with tertiary packing. The unfolded region includes the first 14 N-terminal residues of [30-51] and is as unfolded as an isolated peptide containing these residues. Using protein dissection, we demonstrate that the folded and unfolded regions of [30-51]Ala are structurally independent. The partially folded structure of [30-51]Ala explains many of the properties of authentic [30-51] in the folding pathway of BPTI. Moreover, direct structural characterization of [30-51]Ala has revealed that a crucial step in the folding pathway of BPTI coincides with the formation of a native-like subdomain, supporting models for protein folding that emphasize the formation of cooperatively folded subdomains.

Complete folding of bovine pancreatic trypsin inhibitor with only a single disulfide bond.

In the oxidative folding of bovine pancreatic trypsin inhibitor (BPTI) at neutral pH, only two one-disulfide intermediates accumulate to a significant extent, namely [5-55] and [30-51]. In this paper we describe a recombinant model of [5-55], designated [5-55]Ala, which was made by replacing the cysteine residues not involved in the disulfide bond with alanine. As judged by two-dimensional NMR, [5-55]Ala folds into essentially the same conformation as native BPTI. Moreover, like native BPTI, [5-55]Ala inhibits trypsin stoichiometrically. Thus, the disulfide-bonded intermediate [5-55] corresponds not to a partially folded protein folding intermediate but rather to an essentially completely folded protein. This conclusion provides an explanation for many of the thermodynamic and kinetic properties of [5-55] in the folding pathway of BPTI.

Role of a subdomain in the folding of bovine pancreatic trypsin inhibitor.

The disulphide-bonded intermediates that accumulate in the oxidative folding of bovine pancreatic trypsin inhibitor (BPTI) were characterized some time ago. Structural characterization of these intermediates would provide an explanation of the kinetically preferred pathways of folding for BPTI. When folding occurs under strongly oxidizing conditions, more than half the molecules become trapped in an intermediate, designated N*, which is similar to the native protein but lacks the 30-51 disulphide bond. We have tested the hypothesis that the precursor to N* is the one-disulphide intermediate [5-55], which contains the most stable disulphide in BPTI, and present evidence here that this is the case. A peptide model of [5-55], corresponding to a subdomain of BPTI, seems to fold into a native-like conformation, explaining why [5-55] does not lead to native protein and why it folds rapidly to N*. A native-like subdomain structure in a peptide model of [30-51], the other crucial one-disulphide intermediate, may explain the route by which [30-51] folds to native protein. Thus, much of the folding pathway of BPTI can be explained by the formation of a native-like subdomain in these two early intermediates. This suggests that a large part of the protein folding problem can be reduced to identifying and understanding subdomains of native proteins.