A BBP-Mud2p heterodimer mediates branchpoint recognition and influences splicing substrate abundance in budding yeast

Abstract

The 3' end of mammalian introns is marked by the branchpoint binding protein, SF1, and the U2AF65-U2AF35 heterodimer bound at an adjacent sequence. Baker's yeast has equivalent proteins, branchpoint binding protein (BBP) (SF1) and Mud2p (U2AF65), but lacks an obvious U2AF35 homolog, leaving open the question of whether another protein substitutes during spliceosome assembly. Gel filtration, affinity selection and mass spectrometry were used to show that rather than a U2AF65/U2AF35-like heterodimer, Mud2p forms a complex with BBP without a third (U2AF35-like) factor. Using mutants of MUD2 and BBP, we show that the BBP-Mud2p complex bridges partner-specific Prp39p, Mer1p, Clf1p and Smy2p two-hybrid interactions. In addition to inhibiting Mud2p association, the bbpDelta56 mutation impairs splicing, enhances pre-mRNA release from the nucleus, and similar to a mud2::KAN knockout, suppresses a lethal sub2::KAN mutation. Unexpectedly, rather than exacerbating bbpDelta56, the mud2::KAN mutation partially suppresses a pre-mRNA accumulation defect observed with bbpDelta56. We propose that a BBP-Mud2p heterodimer binds as a unit to the branchpoint in vivo and serves as a target for the Sub2p-DExD/H-box ATPase and for other splicing factors during spliceosome assembly. In addition, our results suggest the possibility that the Mud2p may enhance the turnover of pre-mRNA with impaired BBP-branchpoint association.

Figure 1.

Tandem affinity purification of BBP–TAP and Mud2–TAP. (A) Autoradiogram of a 7% polyacrylamide gel of metabolically labeled proteins selected by sequential IgG agarose and calmodulin agarose chromatography from yeast that express the indicated TAP-tagged gene constructs as genomic integrants. The untagged lane shows background proteins (asterisks). Samples in lanes 1–4 were pre-treated with RNase A prior to TAP selection while samples 5–6 were not. The numbers at the left indicate the positions of unlabeled protein molecular weight markers. Bands corresponding to the untagged Mud2p and BBP proteins and the proteins with the residual CBP tags are indicated. (B) Coomassie blue stain of unlabeled Mud2-TAP and co-purifying proteins. Lane 2 shows the migration of protein molecular weight markers. (C) Yeast used in panel a treated (+) or not (−) with RNase A hybridized with probes specific for the spliceosomal snRNAs (lanes 1–4) or stained with ethidium bromide for the 25S and 18S ribosomal RNAs (lanes 5–8). (D) Single-step recovery of BBP–TAP and Mud2–TAP by TAP by calmodulin agarose chromatography. To avoid band distortion due to sample overloading, the unfractionated (total) protein lanes contain 1/3 the equivalent amount of sample. Untagged = calmodulin agarose recovered material recovered from an untagged extract.

Figure 2.

BBP–Mud2p purifies as a heterodimer. (A) HPLC resolution of molecular weight standards by sequential 2000Å and 300Å sepharose columns. (B) Western blots of the BBP–TAP and Mud2–TAP selected on calmodulin agarose and resolved by HPLC. The blots were developed with an antibody against the TAP epitope. T, total protein applied to the tandem size exclusion columns. The sample numbers correspond to the same column fractions indicated in (A).

Figure 3.

Bridged interactions of the BBP–Mud2p heterodimer. (A) Two hybrid interactions conducted in yeast wild-type for both genes (MUD2, BBP), deleted for MUD2 (mud2::KAN, BBP), or containing a non-lethal mutation within BBP (MUD2, bbpΔ56). Reporter gene activity was scored after 3 days at 30°C. (B) Direct growth assay in the absence of 2-hybrid plasmids of yeast with wild-type MUD2 and BBP alleles (WT), deleted for MUD2 (mud2::KAN, BBP), containing a non-lethal mutation within BBP (MUD2, bbpΔ56) or both mutations (mud2::KAN, bbpΔ56) at 23 and 37°C. (C) Tandem affinity selection of metabolically labeled proteins with Mud2–TAP from yeast that express the bbpΔ56 allele (lane 1) and wild-type BBP allele (lane 2). Lane 3 shows background proteins isolated from an untagged strain. The positions of unlabeled protein markers are shown on the right.

Figure 4.

Mutation of the Mud2p interaction domain of BBP suppresses sub2::KAN lethality. (A) Yeast being the wild-type or the indicated mutant alleles of the MUD2, BBP, SUB2 transformed with a URA3-linked plasmid copy of SUB2 [pSUB2(URA3)] plated at 30°C on FOA medium to select for plasmid loss. The fully wild-type strain (steak 1) and a mutant in the essential NTR2 gene (ntr2::KAN) complimented by a URA3-linked wild-type plasmid (pNTR2(URA3)) are included as positive and negative controls for growth, respectively. Streaks 6 and 7 are different meiotic isolates bearing the same bbpΔ56 mutation. (B) Re-introduction of a functional BBP allele (or an empty vector, streak 3) on a second plasmid in wild-type yeast (steak 1) or the bbpΔ56 background (streak 2). Each strain also is transformed with the pSUB2(URA3) plasmid. Growth is for 3 days at 30°C FOA medium without tryptophan.

Figure 5.

Splicing inhibition by the mud2::KAN and bbΔ56 mutations. (A) Northern analysis of RNA extracted from wild-type yeast (WT), the ts splicing mutant, prp38-1, and yeast with the bbpΔ56 and mud2::KAN single mutations and with the combined bbpΔ56 plus mud2::KAN double mutant background. RNA harvested from yeast grown continuously at 23°C (−) and after a 2 h shift to 37°C (+) was hybridized with radiolabeled probes to detect the intronless ADE3 mRNA and the RPS17A pre-mRNA and mRNA. The relative abundance of mRNA and pre-mRNA (M/P ratio) is presented below the image. (B) Northern analysis as in panel A with added samples smy2::KAN and smy2::KAN combined with bbpΔ56 or mud2::KAN. (C) Primer extension analysis of RNA isolated from the indicated yeast backgrounds that express RPS17A reporter gene constructs with the wild-type intron (WT), branchpoint mutants (HZ8, HZ3, HZ10) and 5′ splice site mutant (HZ12). (D) Primer extension of RNA from the bbpΔ56, mud2::KAN double mutant before (left) and after (right) add back of MUD2.

Figure 6.

Increased RNA export reporter activity in the mud2::KAN, bbΔ56 and mud2::KAN, smy2::KAN double mutant backgrounds. β-galactosidase activities from the spliced reporter and unspliced reporters are presented in the wild-type and mutant backgrounds. The β-galactosidase measurements were normalized by dividing each by that obtained with an intronless control reporter assayed in parallel and multiplying the resultant value by 1000. The bars represent the experimental standard deviations from four to ten replicate experiments. A two-tailed t-test of significance was used with pair wise comparisons.