Evolution of Substrates and Components of the Pro/N-Degron Pathway

Abstract

Gid4, a subunit of the ubiquitin ligase GID, is the recognition component of the Pro/N-degron pathway. Gid4 targets proteins in particular through their N-terminal (Nt) proline (Pro) residue. In Saccharomyces cerevisiae and other Saccharomyces yeasts, the gluconeogenic enzymes Fbp1, Icl1, and Mdh2 bear Nt-Pro and are conditionally destroyed by the Pro/N-degron pathway. However, in mammals and in many non-Saccharomyces yeasts, for example, in Kluyveromyces lactis, these enzymes lack Nt-Pro. We used K. lactis to explore evolution of the Pro/N-degron pathway. One question to be addressed was whether the presence of non-Pro Nt residues in K. lactis Fbp1, Icl1, and Mdh2 was accompanied, on evolutionary time scales (S. cerevisiae and K. lactis diverged ∼150 million years ago), by a changed specificity of the Gid4 N-recognin. We used yeast-based two-hybrid binding assays and protein-degradation assays to show that the non-Pro (Ala) Nt residue of K. lactis Fbp1 makes this enzyme long-lived in K. lactis. We also found that the replacement, through mutagenesis, of Nt-Ala and the next three residues of K. lactis Fbp1 with the four-residue Nt-PTLV sequence of S. cerevisiae Fbp1 sufficed to make the resulting "hybrid" Fbp1 a short-lived substrate of Gid4 in K. lactis. We consider a blend of quasi-neutral genetic drift and natural selection that can account for these and related results. To the best of our knowledge, this work is the first study of the ubiquitin system in K. lactis, including development of the first protein-degradation assay (based on the antibiotic blasticidin) suitable for use with this organism.

Figure 1.

Two-hybrid (Y2H) binding assays with S. cerevisiae and K. lactis proteins. The Pro and Ser residues (N-terminal in these test proteins) are colored red and green, respectively. In rows 1 and 2, as shown previously, S. cerevisiae Gid4 binds to S. cerevisiae P-Fbp1 but not to the otherwise identical S-Fbp1 mutant that bears Nt-Ser. In rows 3 and 4, the same result but with K. lactis Gid4 (binding to S. cerevisiae P-Fbp1 but not to S-Fbp1). In row 5, same as row 3 but in the absence of S-Fbp1, a test for the absence of the Y2H signal in the presence of K. lactis Gid4 alone. Analogous tests with S. cerevisiae Fbp1 alone were described previously. In rows 6 and 7, no detectable binding of K. lactis Gid4 to the wt K. lactis S-Mdh2 protein and its N-terminally truncated P³⁷-Mdh2 derivative (see Figure S2 and Results and Discussion).

Figure 2.

Promoter reference technique and its use for degradation assays with K. lactis and S. cerevisiae proteins. (A and B) Tetracycline (Tc)-based promoter reference technique (see Results and Discussion and refs and 40). (C) In lane 1, kilodalton markers. Tc/PRT chase was performed at 30 °C with wt K. lactis A-Fbp1_3f (Kl-A-Fbp1_3f) during transition from growth on ethanol (EtOH) to growth on glucose in wt (lanes 2–5) or gid4Δ (lanes 6–9) S. cerevisiae. The bands of Kl-A-Fbp1_3f and fDHFR_ha (a reference protein) are indicated on the right. (D) Same as panel C but with S. cerevisiae in glucose-containing medium. (E) Quantification of data in panel F. (F) In lane 1, kilodalton markers. Same as in panel D, in wt S. cerevisiae, but Tc/PRT chases with S. cerevisiae SP-Pck1_3f (Sc-SP-Pck1_3f) (lanes 2–5) and with K. lactis SP-Pck1_3f (Kl-SP-Pck1_3f) (lanes 6–9) during the transition from growth on ethanol (EtOH) to growth on glucose (see Materials and Methods). (G) In lane 1, kilodalton markers. Tc/PRT chases, in wt (lanes 2–5) and in gid2Δ (lanes 6–9) S. cerevisiae grown in glucose-containing medium, with K. lactis Gid4 N-terminally tagged with the myc₉ epitope (Kl-_myc9Gid4). Mouse DHFR (the indicated reference protein) was long-lived in S. cerevisiae irrespective of whether it contained its wt and Nt-acetylatable Met-Asp Nt sequence (as in this study) or the Met-Pro Nt sequence (the latter was converted, cotranslationally, to the non-Nt-acetylatable Nt-Pro).

Figure 3.

Replacing the Nt-AGIC sequence of K. lactis Fbp1 with the Nt-PTLV sequence of S. cerevisiae Fbp1 makes the resulting ptlv-Kl-Fbp1 short-lived in K. lactis. wt K. lactis Fbp1_3ha, bearing the AGIK Nt sequence (after the removal of Nt-Met), was expressed in wt K. lactis from the native P_FBP1 promoter and was C-terminally tagged with the triple-ha tag. This K. lactis protein is denoted as agic-Kl-Fbp1_3ha (after removal of Nt-Met). In a separate and otherwise wt K. lactis strain, the agic-Kl-Fbp1_3ha mentioned above was replaced with ptlv-Kl-Fbp1_3ha (after removal of Nt-Met), bearing the Nt-PTLV sequence of wt S. cerevisiae Fbp1 (see Results and Discussion). (A) In lane 1, kilodalton markers. In lanes 2–5, wt K. lactis cells expressing (from the native P_FBP1 promoter) the endogenous wt K. lactis agic-Kl-Fbp1_3ha (after removal of Nt-Met) were grown for 20 h in ethanol medium and thereafter shifted to glucose medium, followed by the blasticidin-based chase-degradation assay. Lanes 6–9, same as lanes 2–5 but with the otherwise wt K. lactis strain expressing ptlv-Kl-Fbp1_3ha (after removal of Nt-Met), instead of agic-Kl-Fbp1_3ha. Lanes 10–13, same as lanes 6–9, but with gid4Δ K. lactis. Lanes 10–13, same as lanes 6–9, but in the absence of added blasticidin. (B) In lane 1, kilodalton markers. Lanes 2–5, same as lanes 2–5 in panel A, but the blasticidin-based chase was carried out with wt K. lactis cells [expressing, from the native P_FBP1 promoter, the endogenous wt K. lactis agic-Kl-Fbp1_3ha (after removal of Nt-Met)] grown in glucose alone. Lanes 6–9, same as lanes 2–5 but with the otherwise wt K. lactis strain expressing ptlv-Kl-Fbp1_3ha (after removal of Nt-Met), instead of agic-Kl-Fbp1_3ha. Lanes 10–13, same as lanes 6–9, but with gid4Δ K. lactis.