Distinct requirements for intra-ER sorting and budding of peroxisomal membrane proteins from the ER

Abstract

During de novo peroxisome biogenesis, importomer complex proteins sort via two preperoxisomal vesicles (ppVs). However, the sorting mechanisms segregating peroxisomal membrane proteins to the preperoxisomal endoplasmic reticulum (pER) and into ppVs are unknown. We report novel roles for Pex3 and Pex19 in intra-endoplasmic reticulum (ER) sorting and budding of the RING-domain peroxins (Pex2, Pex10, and Pex12). Pex19 bridged the interaction at the ER between Pex3 and RING-domain proteins, resulting in a ternary complex that was critical for the intra-ER sorting and subsequent budding of the RING-domain peroxins. Although the docking subcomplex proteins (Pex13, Pex14, and Pex17) also required Pex19 for budding from the ER, they sorted to the pER independently of Pex3 and Pex19 and were spatially segregated from the RING-domain proteins. We also discovered a unique role for Pex3 in sorting Pex10 and Pex12, but with the docking subcomplex. Our study describes an intra-ER sorting process that regulates segregation, packaging, and budding of peroxisomal importomer subcomplexes, thereby preventing their premature assembly at the ER.

Figure 1.

Pex3 is required for budding of the RING-domain protein Pex2 from the ER. (A) Cells transformed with plasmids expressing specified HA-tagged proteins were assayed for growth on methanol medium. Cells were grown overnight in YPD, and ∼0.1 OD₆₀₀/ml was further inoculated into methanol medium. Cell growth was measured after 48 h. All the HA-tagged proteins complemented their respective deletion mutants. The assay was repeated twice with similar results. (B) Genetic backgrounds of the strains used in the budding assay in C. (C) ER-budding assays were performed as described in Materials and methods. Peroxisomal proteins in the budded ppVs were detected by immunoblotting. S1, supernatant cytosolic fraction; NTP, nucleoside triphosphates; TBPS, buffer control. 30 µl of 80-µl reaction was analyzed. Input is ∼10% of PYCs used in each reaction. The experiment was repeated three times with similar results.

Figure 2.

Pex2-GFP and Pex17-GFP localization in WT and mutant cells. Fluorescence microscopy analysis of methanol-grown cells (3 h) expressing Pex2-GFP or Pex17-GFP. Cells were grown in YPD and switched during exponential phase to methanol medium. DIC, differential interference contrast. Bar, 2 µm. (A) Colocalization of Pex2-GFP expressed from the inducible alcohol oxidase (AOX) promoter with the ER marker Sec61-mCherry in WT and mutant cells. (B) Colocalization of Pex17-GFP expressed from the inducible AOX promoter with the ER marker Sec61-mCherry in WT and mutant cells. (C) Localization of Pex2-GFP and Pex17-GFP with Pex3-RFP after 6 h in methanol medium. Each localization experiment was repeated more than five times with similar results.

Figure 3.

Pex2-GFP and Pex17-GFP are trafficked through the ER. (A) Colocalization of Pex17-GFP expressed from its endogenous promoter with the ER marker Sec61-mCherry in WT and mutant cells. (B) Colocalization of Pex2-GFP expressed from the AOX promoter with the ER marker Sec61-mCherry in WT cells at specified time points.

Figure 4.

Subcellular fractionation of Pex2-GFP and Pex17-GFP. (A) TCA precipitates of intact cells expressing Pex2-GFP or Pex17-GFP from the specified promoter were obtained as described in Materials and methods. Western blotting was performed with anti-GFP monoclonal antibody. GFP* is the cleaved fragment of Pex2-GFP and is ∼10 kD larger than the GFP fragment cleaved from Pex17-GFP. Notably, Pex17-GFP expressed from the endogenous promoter showed no significant cleavage. (B) Cells expressing Pex2-GFP or Pex17-GFP from the inducible AOX promoter were grown in methanol medium (8 h) and fractionated to obtain PNS, 20S, 20P, 200S, and 200P fractions as described in Materials and methods. The pellet fractions were resuspended in the initial volume, and equal volumes of each fraction was analyzed by SDS-PAGE and immunoblotting. (C) Cells expressing Pex17-GFP from endogenous promoter were fractioned and analyzed as described in B. All the subcellular fractionation experiments were repeated three times with similar results.

Figure 5.

Localization of Pex2 and Pex17 in pex1Δ and pex6Δ cells. (A) Pex17-mCherry and Pex2-GFP localization in pex1Δ and pex6Δ cells. Bar, 2 µm. (B) Pex2-GFP and Pex17-GFP localization relative to Sec61-mCherry in pex1Δ cells. (C) Model summarizing the localization and sorting requirements for PMPs. Step 1 shows the RING-domain protein Pex2 dispersed at the cell periphery, whereas the docking complex proteins are sorted to the ER in cells lacking Pex3 or Pex19 (Fig. 2, A–C). Similar localization patterns were observed in pex3Δ cells, as well as in pex3Δ pex19Δ double mutants. However, when both Pex3 and Pex19 were reintroduced as seen in pex1Δ or pex6Δ cells, both Pex2 and Pex17 localized to the pER and subsequently budded out of the ER, forming ppVs as in A and B. Each localization experiment was repeated more than three times with similar results.

Figure 6.

Pex3 is required for the interaction between Pex19 and RING-domain proteins. (A) Co-IP was performed by immunoprecipitating Pex19-FLAG using M2-agarose beads as described in Materials and methods. Immunoblotting was performed with the specified antibodies. The cells were harvested after 6 h in methanol medium. Co-IP was repeated three times with similar results. Ctrl, control; HRP, horseradish peroxidase. (B) Model summarizing the co-IP data. Pex19 presumably docks with Pex3 at the ER, enabling it to interact stably with the RING-domain proteins. This ternary complex initiates intra-ER sorting of the RING-domain proteins to the pER. This is evident from the localization of fluorescence-tagged Pex2 in pex1Δ cells, where in the presence of both Pex3 and Pex19, Pex2 sorts to a punctate pER structure (Fig. 5, A and B). Failure to sort to the pER in the absence of Pex3 could be the reason that the RING-domain proteins do not bud in pex3Δ cells. However, the docking complex proteins do sort and bud because they interact with Pex19 in the absence of Pex3.

Figure 7.

Pex19 is required for the interaction between Pex3 and RING-domain, as well as docking complex, proteins. (A) Co-IP was performed by immunoprecipitating Pex3-3HA using rat monoclonal antibodies conjugated with Sepharose beads as described in Materials and methods. The cells were harvested after 6 h in methanol medium. Coimmunoprecipitated proteins were detected by immunoblotting. Co-IP was repeated three times with similar results. (B) The updated model from Fig. 5 B depicts Pex19-dependent association of Pex3 with the docking complex proteins at the pER. This is supported by the co-IP observations in Fig. 6 A and the localization of Pex17 to the pER in pex1Δ cells (Fig. 5 B). The interaction of Pex3 with the docking complex proteins in pex1Δ cells in A suggests that Pex3 also sorts with the docking complex proteins because Pex3 is required to sort Pex10 and Pex12 into the docking complex ppVs (Fig. 8 B).

Figure 8.

Immunoisolation and protein composition of preperoxisomal vesicles. Immunoisolation of preperoxisomal vesicles was performed as described in Materials and methods. Pex2-3HA and Pex17-3HA were used as anchors for isolation of vesicles from 20S. After antibodies coupled to Sepharose beads were used to capture the ppVs, beads were washed three times, and associated membranes were eluted in SDS sample buffer. The immunoisolate was analyzed with specified antibodies. The experiment was performed with cells harvested after 6 h in methanol medium. (A) Immunoisolation was performed with 20S fractions from WT, pex1Δ, and pex6Δ cells expressing either Pex17-3HA and Pex2-GFP or Pex2-3HA and Pex17-GFP expressed from the constitutive GAP promoter. In pex3Δ cells, Pex2-GFP was not detectable in the 20S fraction, probably because it is retained in the ER (Fig. 1 C) and thus was not included in the experiment. pex1Δ cells expressing P_AOX-Pex1 were switched from YPD to methanol medium (+Methanol) for 6 h or were continued in YPD (−Methanol) before immunoisolation of vesicles. (B) Immunoisolation was performed with 20S fractions from WT, pex1Δ, and pex6Δ cells expressing either Pex17-3HA or Pex2-3HA. The expression of Pex2-3HA was considerably lower than that of Pex17-3HA in pex1Δ, pex6Δ, and pex3Δ cells, and the blot (right) was exposed for a longer duration (∼10×). The experiment was repeated more than three times with similar results.

Figure 9.

Localization of Pex12-GFP in WT and mutant cells. Pex12-GFP and mCherry-Sec61 localization in WT, pex19Δ, pex3Δ, pex1Δ, and pex6Δ cells. Bar, 2 µm. Each localization experiment was repeated three times with similar results.

Figure 10.

Pex3 and Pex19 are required for the sorting of Pex12 into vesicles containing the docking subcomplex proteins. (A) Interaction of Pex17-3HA with Pex3, RING-domain, and docking complex proteins in WT, pex19Δ, pex3Δ, and pex1Δ cells. The co-IP experiment was repeated twice with similar results. Ctrl, control. (B) Model for the role of Pex19 and Pex3 in the sorting and budding of RING-domain and docking complex proteins from the ER. Both Pex3 and Pex19 must interact with RING-domain proteins at the cortical ER to form stable ternary complexes (Figs. 6 A and 7 A), which are then sorted to the pER domains from which the distinct ppVs bud (Fig. 1 C). In the absence of either Pex3 or Pex19, the RING-domain proteins are dispersed over the cortical ER and fail to localize to the punctate pER (Figs. 2 A and 9). Because Pex3 is a membrane protein and Pex19 is a cytosolic protein that docks with Pex3, and because Pex3 has its own intra-ER sorting signal (Fakieh et al., 2013), Pex3 must sort the RING proteins to the pER. In contrast to the RING-domain proteins, neither Pex3 nor Pex19 is required for the intra-ER sorting of a docking protein, such as Pex17, to the pER (Fig. 2 B). Furthermore, although Pex3 is not necessary for the budding of ppV-D vesicles (Fig. 1 C), it is found associated with the vesicles immunoisolated with Pex17-HA (Fig. 8 B) because these vesicles contain RING-domain proteins, Pex10 and Pex12 (Fig. 8 B), whose intra-ER sorting is dependent on Pex3 and Pex19 (Fig. 9). New functions revealed for Pex3 by this study are (1) its role in de novo peroxisome biogenesis and (2) its role in the intra-ER sorting and subsequent budding of the RING-domain proteins from the pER, but not for either process for the docking complex proteins. For Pex19, we present new information that it docks with Pex3 at the ER membrane, where it both bridges and stabilizes the interactions with the RING-domain proteins, to allow intra-ER sorting of these proteins to the pER. Subsequently, as shown previously, Pex19 is also required for budding of both ppVs from the ER (Lam et al., 2010; Agrawal et al., 2011).