An acidic sequence of a putative yeast Golgi membrane protein binds COPII and facilitates ER export

Abstract

We previously identified Sys1p as a high copy number suppressor of Ypt6 GTPase-deficient yeast mutants that are defective in endosome-to-Golgi transport. Here, we show that Sys1p is an integral membrane protein that resides on a post-endoplasmic reticulum (ER) organelle(s). Affinity studies with detergent- solubilized yeast proteins showed that the C-terminal 53 amino acid tail of Sys1p binds effectively to the cytoplasmic Sec23p-Sec24p COPII subcomplex. This binding required a di-acidic Asp-Leu-Glu (DXE) motif, previously shown to mediate efficient ER export of the vesicular stomatitis virus glycoprotein in mammalian cells. In Sys1p, a Glu-Leu-Glu (EXE) sequence could not substitute for the (DXE) motif. Mutations of the (DXE) sequence resulted in ER retention of approximately 30% of the protein at steady state, whereas addition of the Sys1p tail to an ER-resident membrane protein led to an intracellular redistribution of the chimeric protein. Our study demonstrates for the first time that, in yeast, a di-acidic sequence motif can act as a sorting signal for cargo selection during the formation of transport vesicles at the ER by direct binding to COPII component(s).

Fig. 1. Sys1p is a Golgi/endosome membrane protein. (A) A cleared yeast cell lysate (strain SEY6210) was divided into four aliquots that were treated for 15 min on ice with either lysis buffer (untreated), detergent, high salt or urea as indicated. After consecutive centrifugation at 10 000 and 100 000 g, the pellets (P10, P100) and the supernatants (S100) were subjected to immunoblot analysis with anti-Kar2p (ER marker), anti-Emp47p (Golgi marker) or anti-Sys1p antibodies. (B) The cleared lysate was centrifuged on sucrose gradients, and fractions (1–12, from top to bottom) were collected and subjected to immunoblot analysis with antibodies to the proteins shown to the left.

Fig. 2. Sys1p does not cycle through the ER. Cleared lysates of sec23 (A) or sec12 (B) mutant cells, grown at either permissive (25°C) or non-permissive temperature (35°C), were subjected to differential centrifugation (A) or sucrose gradient centrifugation (B) and subsequently to immunoblot analysis as described in the legend to Figure 1.

Fig. 3. Delineation of ypt6 mutant suppressor activity and membrane topology of Sys1p. (A) Schematic representation of C-terminally truncated Sys1 mutant proteins tested for suppressor activity. The four putative transmembrane domains (TM1–4) and the first and last amino acids of the various Sys1 proteins are indicated. (B) The sequence of the C-terminal tail of Sys1p is shown with arrows pointing to the end points of truncations. (C) The C-terminal region of Sys1p faces the cytosol. A 500 g supernatant of lysed cells was treated with proteinase K, and TCA-precipitated proteins were subjected to immunoblot analysis with antibodies to the luminal ER protein Kar2p, the luminal part of the Golgi protein Emp47p, the cytoplasmic region of the v-SNARE Sec22p and the C-terminus of Sys1p.

Fig. 4. The C-terminal tail of Sys1p (amino acids 151–203) interacts with the Sec23p–Sec24p complex. GST and a GST–Sys1 tail fusion protein, coupled to agarose beads, served as affinity matrix with detergent-solubilized total yeast protein. Beads were washed with buffer, and bound proteins eluted in 1 ml fractions of SDS buffer (1) or 0.5 ml fractions of 1 M NaCl (2) for SDS–PAGE and silver staining (20 µl of each fraction). Bound proteins usually appeared in two fractions, one of which is shown on the blot. Asterisks mark two unspecifically bound proteins of ∼40 and 50 kDa.

Fig. 5. Comparison of C-terminal sequences of yeast Sys1, Bap2 and Ptr2 and of mammalian ERGIC-53 and viral VSV-G proteins. (A) The (FF) and the di-acidic motifs effective in efficient ER export in mammalian cells are highlighted. In Sys1p, the (DLE) but not the (FF) motif mediates binding to COPII components. (B) The location of (DXE) motifs in Sys1p and the plasma membrane-localized transporters Bap2p and Ptr2p are shown.

Fig. 6. A di-acidic motif of the Sys1p tail is responsible for Sec23p–Sec24p binding. (A) The hydrophilic tail of Sys1p (WT) and several deletion or substitution mutants (C1–C7) were fused to GST and probed for Sec23p–Sec24p interaction. (B) COPII subcomplex binding was investigated by an affinity approach as described in the legend to Figure 4 and is shown for some tail fusions. Proteins bound to beads were eluted with 0.5 ml of 1 M NaCl, 20 µl fractions were subjected to SDS–PAGE and proteins were identified by silver staining. (C) Sec23p and Sec24p were verified with affinity-purified antibodies. (D) Sar1p, identified by immunoblot analysis, does not interact with Sec23p–Sec24p-binding Sys1p tail fusions.

Fig. 7. Deletion or mutation of the di-acidic motif from the Sys1p C-terminus moderately affects ER export. Cleared lysates from (A) wild-type yeast (strain CV1), (B) yeast cells expressing only a C-terminally truncated or (C) the indicated mutant version of Sys1p (strains CV2 and CV3) were subjected to sucrose gradient centrifugation, followed by immunoblot analysis of the proteins in gradient fractions 1–12 using anti-HA antibodies. Immunoblots were examined quantitatively using a LumiImager. Note a shift of some of the truncated Sys1(1–186) protein to the position of the ER marker Kar2p.

Fig. 8. Fusion of the Sys1p tail to the ER-resident protein Wbp1p leads to redistribution of the fusion protein in a (DXE) motif-dependent fashion. Cleared lysates of the protease-deficient strain cl3-ABYS-86, untransformed (A) or transformed with a recombinant plasmid expressing the indicated Wbp1-Sys1p tail fusion proteins (B and C), were subjected to sucrose gradient centrifugation. Proteins of different fractions were probed with antibodies specific for the ER marker Kar2p, the Golgi protein Emp47p and Wbp1p. Note the shift of the Wbp1p–Sys1 fusion protein with regard to its (DXE)-mutated version and to Wbp1 wild-type protein.