Prp8p dissection reveals domain structure and protein interaction sites

Abstract

We describe a novel approach to characterize the functional domains of a protein in vivo. This involves the use of a custom-built Tn5-based transposon that causes the expression of a target gene as two contiguous polypeptides. When used as a genetic screen to dissect the budding yeast PRP8 gene, this showed that Prp8 protein could be dissected into three distinct pairs of functional polypeptides. Thus, four functional domains can be defined in the 2413-residue Prp8 protein, with boundaries in the regions of amino acids 394-443, 770, and 2170-2179. The central region of the protein was resistant to dissection by this approach, suggesting that it represents one large functional unit. The dissected constructs allowed investigation of factors that associate strongly with the N- or the C-terminal Prp8 protein fragments. Thus, the U5 snRNP protein Snu114p associates with Prp8p in the region 437-770, whereas fragmenting Prp8p at residue 2173 destabilizes its association with Aar2p.

FIGURE 1.

Structure of the gene dissection transposon and dissected prp8. The transposon was custom-built from three main components: (1) PRP8 gene downstream flanking region (DFR) containing signals for transcription termination, polyadenylation, and stop codons in all three reading frames; (2) KanMX4 cassette (Wach et al. 1994) conferring Kanamycin-resistance in Escherichia coli and G418-resistance in yeast (KanMX4); (3) PRP8 gene upstream flanking region (UFR) containing signals for initiation of transcription and translation. The transposon is flanked by 19 bp “Mosaic End” (ME) sequences recognized by Tn5 transposase. An example of the structure of one of a library of transposon-dissected prp8 genes is shown (not to scale). The transposon terminates transcription and translation of the upstream prp8 gene sequences and drives expression of the prp8 gene sequences downstream, producing two contiguous protein fragments via translation of two separate mRNAs. Each fragment is tagged at its N terminus with a nuclear localization signal (NLS). The disrupted alleles were introduced into the TAP-tagged chromosomal copy of PRP8 in yeast strain RG8T by homologous recombination using PCR-amplified DNA; the outer primer set was for partition after codon 436, and the inner primer set was for partition after codons 770 and 2173 to avoid generating the 9-bp direct repeat (DR) generated during transposition.

FIGURE 2.

TAP purification of Prp8-associated proteins. (A) TAP-purified proteins were fractionated in a 4%–12% PAGE gel and analyzed by mass spectrometry. (B) Effect of salt on the association of the pairs of dissected Prp8 polypeptides. Splicing extracts were incubated with IgG-agarose in the presence of 150, 300, or 500 mM NaCl. Proteins were fractionated, blotted, and probed with anti-8.6 antibodies. (Lanes 1) Nontagged full-length Prp8p, (lanes 2) Tap-tagged full-length Prp8p, (lanes 3) TAP-tagged Prp8p dissected at residue 770, (lanes 4) TAP-tagged Prp8p dissected at residue 2173.

FIGURE 3.

Detection of proteins associated with the N- or C-terminal fragments of Prp8p. (A) Splicing extracts were first incubated with horse IgG agarose in 500 mM NaCl to pull down only the C-terminal Prp8p fragments (lanes 1–5). The supernates were then incubated with anti-8.6-adsorbed protein A Sepharose beads (lanes 6–10). After PAGE, Prp8p fragments that bound to the beads were detected by Western blotting, using anti-8.6 antibodies. (Lanes 1,6) full-length TAP-Prp8 protein; (lanes 2,7) full-length nontagged Prp8p; (lanes 3,8) Prp8T/436p; (lanes 4,9) Prp8T/770p; (lanes 5,10) Prp8T/2413p. The blot from A was stripped and reprobed with anti-Snu114 antibodies (B) or with anti-Aar2p antibodies (C). (D) Splicing extracts from strains producing nontagged full-length Prp8p (lane 1), Prp8T/ 770p (lanes 2–4), Prp8T/2173p (lanes 5–7), and full-length Prp8Tp (lanes 8–10) were incubated with IgG-agarose in the presence of 150 mM NaCl. The precipitates (Co-IP), 10% total, and 10% supernates (Sup) after the pulldown were analyzed by Western blotting with both anti-8.6 and anti-Aar2p antibodies.

FIGURE 4.

Glycerol gradient fractionation of Prp8p, Snu114p, and Aar2p. RG8T cell extract containing full-length Prp8Tp (A) and RG8T/2173p extract containing Prp8T/2173p (B) were fractionated in 10%–30% glycerol gradients. Alternate gradient fractions were analyzed by Western blotting with anti-Snu114p and anti-Aar2p antibodies. (*) Aar2p near the top of the gradient (lanes 25).

FIGURE 5.

Diagram showing the positions of viable breakpoints relative to other known features in Prp8p (not to scale). Vertical lines above the protein indicate the positions at which Prp8p can be split in two and still function in trans, in vivo. The domains identified in this work are indicated. Numbers refer to amino acid residues. PPP, proline-rich region; NLS, putative nuclear localization signal; RRM, RNA recognition motif; 3.2, a region of particularly high amino acid sequence conservation; MPN, predicted Mpr-1, Pad-1, N-terminal domain; a, b, c, d, and e, genetically defined regions (Kuhn and Brow 2000); 5′SS X-link, a five-amino-acid region identified by photocross-linking to contact the 5′ splice site in pre-mRNA (Reyes et al. 1999). For further details about features in Prp8p, see Grainger and Beggs (2005).