Direct sorting of the yeast uracil permease to the endosomal system is controlled by uracil binding and Rsp5p-dependent ubiquitylation

Abstract

The yeast uracil permease, Fur4p, is downregulated by uracil, which is toxic to cells with high permease activity. Uracil promotes cell surface Rsp5p-dependent ubiquitylation of the permease, signaling its endocytosis and further vacuolar degradation. We show here that uracil also triggers the direct routing of its cognate permease from the Golgi apparatus to the endosomal system for degradation, without passage via the plasma membrane. This early sorting was not observed for a variant permease with a much lower affinity for uracil, suggesting that uracil binding is the signal for the diverted pathway. The FUI1-encoded uridine permease is similarly sorted for early vacuolar degradation in cells exposed to a toxic level of uridine uptake. Membrane proteins destined for vacuolar degradation require sorting at the endosome level to the intraluminal vesicles of the multivesicular bodies. In cells with low levels of Rsp5p, Fur4p can be still diverted from the Golgi apparatus but does not reach the vacuolar lumen, being instead missorted to the vacuolar membrane. Correct luminal delivery is restored by the biosynthetic addition of a single ubiquitin, suggesting that the ubiquitylation of Fur4p serves as a specific signal for sorting to the luminal vesicles of the multivesicular bodies. A fused ubiquitin is also able to sort some Fur4p from the Golgi to the degradative pathway in the absence of added uracil but the low efficiency of this sorting indicates that ubiquitin does not itself act as a dominant signal for Golgi-to-endosome trafficking. Our results are consistent with a model in which the binding of intracellular uracil to the permease signals its sorting from the Golgi apparatus and subsequent ubiquitylation ensures its delivery to the vacuolar lumen.

Figure 1.

Uracil triggered delivery of its specific permease to the vacuole via the VPS pathway. (A) Wild-type and end3Δ cells transformed with pFL38GalFUR4-GFP were cultured on galactose to midexponential growth phase. Glucose and uracil were then added and cells were cultured for an additional 2 h. Cells withdrawn from cultures before (0) and after (2 h) the addition of glucose and uracil were examined for fluorescence and with Nomarski optics. (B) End3Δ cells transformed with pFL38 GalFUR4-GFP and pep12Δ cells transformed with either pFL38 GalFUR4-GFP or pFL38 GalFUR4^KR-GFP were cultured with lactate and then with added galactose, in the absence or presence of uracil. The cellular distribution of newly synthesized permease was assessed, over 2 h for end3Δ cells and 3 h for pep12Δ cells to take into account their very slow growth, by fluorescence microscopy, after staining the vacuoles with CMAC as described in MATERIALS AND METHODS. (C) Extracts from end3Δ cells that had produced Fur4-GFP for 2 h in the absence or presence of uracil were fractionated on equilibrium sucrose density gradients. Sucrose gradient concentration increases from left to right in the figure. Aliquots of the various fractions were analyzed by western immunoblotting for Fur4-GFP/GFP and Pma1p. Molecular weight markers are indicated in kDa.

Figure 3.

The delivery of uracil permease to the vacuolar lumen was compromised in cells with impaired ubiquitylation. (A) Wild-type cells producing either the SA or the KR variant of Fur4-GFP under control of the Gal10 promoter were used. Cells were cultured with galactose to midexponential growth phase. Glucose and uracil were then added and the cells cultured for an additional 2 h. Cells withdrawn from cultures before (0) and after (2 h) the addition of glucose and uracil were examined for fluorescence. (B) GFP-tagged versions of SA or KR variant of Fur4p were produced for 2 h in the presence of uracil, in cells first grown on lactate as a carbon source, with galactose then added to start permease synthesis. Wild-type, npi1/rsp5, npi2/doa4, and tul1Δ mutants were used. Permease distribution (GFP) and vacuolar staining (CMAC) were assessed.

Figure 2.

Uridine promoted the direct routing of newly synthesized uridine permease to the vacuole. (A) WT and end3Δ cells producing Fui1-GFP under control of the Gal10 promoter, were cultured on galactose to midexponential growth phase. Glucose (2%) and uridine were then added and cells cultured for an additional 2 h. Cells withdrawn from cultures before (0) and after (2 h) the addition of glucose and uridine were examined for fluorescence and with Nomarski optics. (B) Fui1-GFP was produced for 2 h in end3Δ cells in the absence or presence of uridine (urd). Permease distribution (GFP) and vacuolar staining (CMAC) were assessed.

Figure 4.

Uracil permease is ubiquitylated during trafficking from the Golgi apparatus to the vacuole, in a Rsp5p-dependent manner. The DOA4 gene had been deleted in the cells used here, which harbored either the wild-type or a mutant version (npi1) of the RSP5 gene. The cells were cotransformed with plasmid Yep96, encoding normal ubiquitin under the control of the Cup1 promoter (Ellison and Hochstrasser, 1991), and either pFL38GalFUR4^SA-GFP or pFL38GalFUR4^KR-GFP. Cells were first cultured with lactate as the carbon source, and CuSO4 (100 μM) and galactose were then added and the cells cultured for 2 h in the presence of uracil. Protein extracts were prepared and fractionated on equilibrium sucrose density gradients. (A) Aliquots of the various fractions from RSP5 cells producing the SA variant of Fur4-GFP were analyzed by Western immunoblotting for Fur4-GFP, GFP, and Pma1p. Sucrose gradient concentration increases from left to right in the figure. (B) The eight central fractions of the gradient shown in A were pooled and analyzed in parallel to equivalent fractions from the same cells that have produced HA-tagged ubiquitin (from the Yep112 plasmid; Ellison and Hochstrasser, 1991) instead of normal ubiquitin. A small line indicates the increase in the molecular weight of a ubiquitin conjugate. (C) As for A except that fractions of gradients from extracts of npi1 mutant cells producing Fur4^KR-GFP were analyzed by Western immunoblotting for Fur4-GFP, GFP, Pma1p, and Vat2p, a marker of the vacuolar membrane. (D) Pools of the eight central fractions of gradients from RSP5 and npi1 cells producing the SA or KR variant of Fur4-GFP were analyzed by immunoblotting with anti-GFP antibodies.

Figure 5.

The binding of uracil is the main signal for the efficient sorting of Fur4p to the endosomal system. (A) End3Δ cells transformed with pFL38GalUb-FUR4-GFP or pFL38GalUb-FUR4^KR-GFP were cultured with lactate, and the medium was then supplemented with galactose to induce the synthesis of GFP-tagged permease. After 2 h in the absence or presence of uracil, permease fluorescence (GFP) and vacuole staining (CMAC) were assessed. (B) WT cells cotransformed with Yep46FUR4 and either pFL38GalFUR4^SA,272-GFP or pFL38GalFUR4^SA-GFP were subjected to galactose induction in the presence of uracil and examined as in A.

Figure 6.

Polyubiquitylation is required for efficient trafficking along the endocytic pathway and subsequent vacuolar degradation. (A) WT cells producing either Fur4^KR-GFP or Ub-Fur4^KR-GFP under control of the Gal10 promoter were supplemented with glucose and the cells cultured for an additional 2 h in the absence or presence of uracil. Cells withdrawn from cultures before (0) and after the addition of glucose + uracil (2 h) were labeled with CMAC and examined to determine permease distribution (GFP) and the pattern of vacuolar staining (CMAC). Total protein extracts from cells withdrawn before (0) and after the addition of glucose for 2 h in the absence (-) or presence (+) of uracil were analyzed by Western blotting with anti-GFP antibodies. Potential SDS-resistant dimers of permease are indicated by an asterisk. (B) npi1/rsp5 cells producing either Ub-Fur4-GFP or Ub-Fur4^KR-GFP were cultured with galactose to midexponential growth phase, then allowed to grow for 2 h in the presence of glucose and uracil, and examined as in A for GFP and CMAC fluorescence. Membrane-enriched fractions from these cells were prepared before the addition of glucose and analyzed by immunoblotting with anti-GFP antibodies. A small vertical line indicates potential ubiquitin conjugates.

Figure 7.

Presumed sorting steps for newly synthesized uracil permease. At the Golgi level, free permease (oval) is targeted to the cell surface (1), whereas uracil-liganded permease (rectangle) is sorted to the VPS pathway (2). An induced change in the conformation of Fur4p may modify its association with lipid rafts, accounting for its sorting to either plasma membrane or endosomes. On delivery to the endosome, uracil-liganded Fur4p may be efficiently ubiquitylated (•) and sorted to inward budding vesicles of MVB (3), resulting in its degradation in the vacuolar lumen or it may be maintained at the surface of MVB due to defective ubiquitylation, resulting in its becoming a permanent resident of the vacuolar membrane (4). It may be retrieved by the recycling pathway to the plasma membrane (5) if uracil is removed or in cases of underubiquitylation (○). If the sorting decision governed by uracil takes place only at the endosome level, the various possible routes from that location remain valid.