Artificial ER-Mitochondrion Tethering Restores Erg6 Localization and Lipid Droplet Formation in Hansenula polymorpha Δpex23 and Δpex29 Cells

Abstract

Pex23 proteins are a family of fungal endoplasmic reticulum proteins. Hansenula polymorpha contains four members, two of which, Pex24 and Pex32, function in endoplasmic reticulum-peroxisome membrane contact sites. In the absence of the other two members, Pex23 and Pex29, mitochondria are fragmented and lipid droplet numbers are reduced. We here show that in Δpex23 and Δpex29 cells an increased portion of the lipid droplet protein Erg6 (C24-methyltransferase), an enzyme involved in ergosterol biosynthesis, localizes to mitochondria. Erg6 relocalization and the reduction in lipid droplet numbers are suppressed by an artificial endoplasmic reticulum-mitochondrion tether protein. Sterol measurements showed that the presence of Erg6 at mitochondria did not cause major changes in the overall sterol composition. Our findings suggest that Pex23 and Pex29 play a role in endoplasmic reticulum-mitochondrion contact sites which prevent mitochondrial mislocalization of Erg6.

Keywords: Erg6; Pex23; Pex29; lipid droplet; membrane contact sites; mitochondria.

Figure 1.

Introduction of ERMIT restores LD abundance in Δpex23 and Δpex29. (A) Electron microscopy (EM) images of thin sections of KMnO₄-fixed glucose-grown cells of the indicated strains. ER, endoplasmic reticulum; LD, lipid droplet. Scale bar: 500 nm. Representative images are shown. (B) Quantification of LD size using EM analysis of thin sections. Data are presented as mean ± SD from two independent experiments (n = 2) of 40 random LD-containing sections. (C) Schematic representation of the ERMIT tether. (D, F) CLSM Z stack images of the indicated strains producing Erg6-mKate2 (D) or stained with Nile Red (F) with or without ERMIT. The same grey scale values were used for all strains (See Figure S2 for adapted image processing). Scale bar: 2 µm. (E,G) Normalized LD abundance based on Erg6-mKate2 (E) or Nile Red marked puncta (G) observed in Z-stack CLSM images. The abundance in WT was set to 100%. Data represent the mean from three independent experiments (n = 3) with 300 cells analyzed per experiment.

Figure 2.

Erg6 is present at LDs, ER and mitochondria in H. polymorpha Δpex23 and Δpex29 cells. (A) Single plane CLSM Airyscan images showing the localization of Erg6-GFP in glucose-grown cells of the indicated strains. GFP images are shown at two different intensity values. The minimum and maximum pixel values used for the GFP channel are indicated. Mitochondria are marked with MitoTracker Red. Orange arrows indicate LDs; yellow arrows indicate the ER. White boxes in the Overlap images indicate the regions shown in the zoom-in views. Scale bar: 2 µm. (B) Quantification of Erg6-GFP mean fluorescence intensities at mitochondria, labelled with MitoTracker Red, in the indicated strains. Data are presented as mean ± SD from three independent experiments (n = 3) with 200 cells analyzed per experiment. (C) Single-plane CLSM (Airyscan) images showing Erg6-mKate2 localization in the indicated strains with or without ERMIT. Scale bar: 2 µm.

Figure 3.

Pex23 and Pex29 accumulate at LD-MCSs, but are not essential for their formation. (A) CLSM (Airyscan, single plane) images showing the co-localization of Pex23-GFP and Pex29-GFP with Erg6-mKate2 in glucose-grown WT cells. The arrows indicate mKate spots overlapping with GFP spots. Scale bar: 2 µm. (B) Examples of EM images of thin sections of KMnO₄-fixed glucose-grown cells of the indicated strains. White lines indicate examples of distances between two membranes, as used for the quantifications in Figure 3C. ER(N), nuclear endoplasmic reticulum; ER(P), peripheral endoplasmic reticulum; LD, lipid droplet. Scale bar: 200 nm. (C) Quantification of the distances below 30 nm between LDs and other cell organelles in the indicated strains using sections of KMnO₄-fixed, glucose-grown cells. The percentages indicate the percentage of LDs in the five different categories. Mit, mitochondria; V, vacuole. Data are mean ± SD of two independent experiments (n = 2 using 40 random sections from each experiment). (D) The ergosterol content in the total cellular sterols of the indicated strains. Data are mean ± SD from three independent experiments. A student's t-test revealed no statistically significant differences.

Figure 4.

Hypothetical model of Pex23 and Pex29 protein function in LD biogenesis and Erg6 localization. (A) In WT cells, Erg6 initially sorts to the ER and localizes at LDs when these organelles are formed. Specialized ER regions involved in LD formation contain Pex23 and Pex29, but both proteins also occur at mitochondrion-ER contacts, where they either prevent transfer of Erg6 from the ER to mitochondria or facilitate the retrieval of Erg6 back to the ER from where it can be transferred to LDs. (B) In cells lacking Pex23 and Pex29 proteins, the portion of Erg6 that localizes to the mitochondria cannot be efficiently retrieved to the ER, or the trafficking of Erg6 from the ER to mitochondria is stimulated. This restricts Erg6 availability for LD formation, thereby reducing LD formation. (C) The introduction of the artificial tether ERMIT enhances the retrieval of Erg6 from the mitochondria back to the ER, supporting LD formation, or prevents transport of Erg6 from ER to mitochondria. Figure created using BioRender.