All H. polymorpha strains used in this study are listed in online resource Table S1. Yeast cells were cultivated at 37 °C in mineral medium (van Dijken et al. 1976) supplemented with 0.5% glucose (Sigma Aldrich, #14431-43-7), 0.5% methanol (Sigma-Aldrich, #67-56-1) or a mixture of 0.3% methanol and 0.1% glycerol (Sigma-Aldrich, #G7893). Leucine (Sigma-Aldrich #L8912) was added to a final concentration of 60 μg/ml, when necessary.

For imaging experiments, cells were grown on glucose to an optical density (OD660) of 1.0. For induction of the TEV protease Nia, cells were grown to the exponential growth phase on glucose (OD660 1.8–2.0) and diluted to an OD660 of 0.1 in mineral medium (0.3% methanol and 0.1% glycerol). The OD660 was measured using a spectrophotometer (StaRRcol, #SC-60-S).

Selection of positive transformants was performed on YPD plates (1% yeast extract (Gibco, #2441197), 1% peptone (Gibco, #2024469), and 1% dextrose (glucose)) containing 100 μg/ml zeocin (InvivoGen #ant-zn-05), 100 μg/ml nourseothricin (Werner Bioagents/Jena Bioscience, #AB-102XL), and/or 300 μg/ml hygromycin (InvivoGen #ant-hg-5). Escherichia coli DH5α (Hanahan 1983) was used for cloning purposes, and selection was performed using LB plates (1% Bacto tryptone (Gibco, #211705), 0.5% yeast extract (Gibco, #212750), and 0.5% NaCl (Sigma-Aldrich, #S9888)) containing 100 μg/ml ampicillin (Sigma-Aldrich, #A9518-5G).

Plasmids used are listed in online resource Table S2. Primers used are listed in online resource Table S3.

For this study, the unfolded protein region of S. cerevisiae nucleoporin (Nsp1, #YJL041W, UniProt #P14907) was chosen. Three variants of the ScNsp1 genomic sequence, encoding the unfolded protein region of Nsp1, were created: 903 bp (half-length Nsp1, 301 aa), 1803 bp (full-length Nsp1, 601 aa), and 3600 bp (two full-length Nsp1 in sequence fused together, 1200 aa).

Plasmid pHIPN18 GFP-Nsp1301-ePTS1 was constructed using plasmid pHIPN18 mGFP. A Nsp1301 fragment consisting of 903 bp was produced by polymerase chain reaction (PCR) amplification using primers Nsp1-F and Nsp1-R300. These primers contain an NdeI restriction site, ePTS1, stop codon, and a XbaI restriction site. The corresponding PCR product and pHIPN18 mGFP were digested with NdeI (Thermo Scientific, #FD0583) and XbaI (Thermo Scientific, #FD0684) and ligated together. Plasmid pHIPN18 GFP-Nsp1601-ePTS1 was constructed with a similar procedure but using the Nsp1-R600 reverse primer instead, producing a 1809-bp fragment.

The GFP-TEV-Nsp11200-ePTS1 plasmid was constructed by PCR amplification of two fragments and the pHIPN18 mGFP plasmid as a backbone. Firstly, a fragment containing an NdeI restriction site, a nine-nucleotide linker region, a BamHI (Thermo Scientific, #FD0054) restriction site, TEV protease cleavage sequence, the 1809 bp of NSP1601, and a SalI restriction site (Thermo Scientific, #FD0644) was produced using primers BamHI-TEV-NSP1-F and TEV-NSP1-SalI-R. Secondly, a fragment containing an NdeI restriction site, a nine-nucleotide linker region, a SalI restriction site, the 1809 bp of NSP1601, the 1ePTS1 sequence, a stop codon, and a XbaI restriction site (NSP1-600-12SKL) was constructed using primers Sali-Nsp1-F and Nsp1-R600. These two fragments were combined using the SalI restriction site and then ligated into NdeI and XbaI restriction enzyme-digested pHIPN18 mGFP backbone.

The 2xHA-Nsp1601-ePTS1 plasmid was constructed by DNA synthesis (GenScript) of the cloning plasmid pUC57 2HA-NSP1601-AarI. This plasmid contains a HindIII restriction site, a start codon, two copies of the HA tag, the first 178 nucleotides of Nsp1 and the AarI restriction site and was digested with HindIII (Thermo Scientific, #FD0504) and AarI (Thermo Scientific, #ER1581 and New England Biolabs, #R0745S). This fragment was ligated into HinIII + AarI-digested GFP-TEV-Nsp11200-ePTS1 to produce the 2xHA-Nsp1601-ePTS1 plasmid.

To construct plasmid 2xHA-TEV-Nsp11200-ePTS1, a 938-bp fragment was produced by PCR amplification with primers 345.FW.2xHA and 346.RV.2xHA, using plasmid GFP-TEV-Nsp11200-ePTS1 as a template. The 938-bp fragment was isolated and digested with restriction enzymes NotI (Thermo Scientific, #FD0593) and BamHI and ligated into the plasmid GFP-TEV-Nsp11200-ePTS1 digested with the corresponding enzymes to produce the 2xHA-TEV-Nsp11200-ePTS1 plasmid.

Genes were expressed under control of alcohol oxidase 1 (AOX1), alcohol dehydrogenase 1 (ADH1), or endogenous promoter. For expression under control of the AOX1 and ADH1 promoter, the full-length gene was cloned. All plasmids were linearized and integrated into the genome as described before (Faber et al. 1994; Saraya et al. 2012). DNA restriction enzymes were used as recommended by the suppliers (Thermo Scientific or New England Biolabs). PCR for cloning was carried out with Phusion High-Fidelity DNA Polymerase (Thermo Scientific, #F530S). Colony PCR was carried out using DreamTaq DNA Polymerase (Thermo Scientific, #EP0701). For all DNA ligation, the Rapid DNA Ligation kit was used (Thermo Scientific, #K1423). For DNA sequence analysis, the Clone Manager 5 program (Scientific and Educational Software, Durham, NC) was used. For fluorescence microscopy (FM) localization studies, Pex3-mKate2 was used as peroxisomal marker.

For trichloroacetic acid (TCA, Sigma-Aldrich, #T6399) protein precipitation (McCammon et al. 1994) of 3 OD660 units yeast culture, 10 mM ZnCl2 (Sigma- Aldrich, #208086) was added to the 12.5% TCA solution and to the 80% acetone (Supelco, #90872) wash solution to inhibit the activity of the yeast strains expressing the TEV protease NIa (Dougherty et al. 1989; Joseph and Savithri 2000). Samples were prepared in Laemmli buffer (Bio-Rad, 4xLaemmli #1,610747). Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was performed using 7.5% discontinuous gels as described previously (Baerends et al. 2000) or using 4–15% gradient gels (Bio-Rad, 4–15% Mini-PROTEAN® TGX™ Precast Protein Gels, #456108). Equal amounts of protein were loaded per lane.

Semi-dry transfer (Thermo Scientific, Pierce G2 Fast Blotter) of proteins from the SDS gel to a nitrocellulose membrane (Amersham Protran 0.2 NC, #10600001) was performed as described previously (Kyhse-Andersen 1984) with the 60 min standard transfer time program. Blots were decorated with mouse monoclonal antiserum against GFP (Santa Cruz Biotechnology, Inc. #sc-9996), mouse monoclonal antiserum against anti-HA tag (Roche, #11583816001), or rabbit polyclonal antiserum against elongation factor 1 alpha (EF1α), which served as a loading control (Kiel et al. 2007). Secondary goat anti-rabbit immunoglobulin G (IgG) (Thermo Scientific, #31460) or goat anti-mouse IgG antibodies (Thermo Scientific, #31430) conjugated to horseradish peroxidase were used for detection.

For detection, chemiluminescent solution (Amersham ECL Prime, #RPN2232) was applied to the membranes according to the manufacturer’s guidelines (Bio-Rad, Chemidoc Imaging System). Blots were imaged using the Bio-Rad ChemiDoc Imaging System, quantified using Bio-Rad ImageLab software (version 6.1), and edited using Adobe Illustrator (version 28.7.1). The density of each band measured was standardized by dividing the density of the corresponding loading control band.

A Zeiss Axioscope A1 fluorescence microscope with a 63× 1.40 NA objective (Carl Zeiss), CoolSNAP HQ2 digital camera (Photometrics CoolSNAP HQ2, Birmingham, UK), and MICRO-MANAGER software (version 1.4) was used for capturing images. The GFP signal was visualized with a 470/40 nm bandpass excitation filter, a 495 nm dichromatic mirror, and a 525/50 nm bandpass emission filter. The mKate2 fluorescence was visualized with a 587/25-nm bandpass excitation filter, a 605 nm dichromatic mirror, and a 647/70 nm bandpass emission filter. ImageJ (version 1.53u) and Adobe Illustrator (version 28.7.1) were used for image analysis and editing. All FM images within each figure were processed similar unless otherwise stated.

Airyscan images were captured with a confocal laser scanning microscope (Carl Zeiss, LSM800) equipped with a 32-channel gallium arsenide phosphide photomultiplier tube (GaAsP-PMT), Zen 2009 software (Carl Zeiss), and a 63× 1.40 NA objective (Carl Zeiss) for oil immersion. The GFP signal was visualized by excitation with a 488 nm laser, and mKate2 was visualized with a 561 nm laser. Cells were imaged at room temperature. Fiji (version 1.53t) and Adobe Illustrator (version 28.7.1) were used for image analysis and editing. All Airyscan images within each figure were processed similarly.

For Airyscan imaging, cells were washed with 1× phosphate-buffered saline (PBS) (10 mM Na2HPO4 (Millipore, # 567550), 0.137 M NaCl (Sigma-Aldrich, #S9888), 0.27 mM KCl (Sigma-Aldrich, #P9541), and 0.18 mM KH2PO4 (Sigma-Aldrich, # P5655), then fixed in 1% formaldehyde (FA) (Sigma-Aldrich, #F8775) solution for 15 min on ice and washed in 1 × PBS prior to imaging. For the 1% (FA) solution, we prepared 16% (w/v) FA at 65 °C on a heating block and dilute 16× in 1 × PBS to make a 1% (w/v) FA solution (62.5 μL/1 mL PBS) (Krikken et al. 2020). Cells were imaged at room temperature.

Immuno-EM (iEM) was performed as described previously (Thomas et al. 2018). For iEM, polyclonal antiserum raised against alcohol oxidase (αAox) (van der Klei et al. 1988) and catalase (Cta) (Keizer et al. 1992; Waterham et al. 1993) were used as previously described (Salomons et al. 2000). Labeling of HA was performed using monoclonal antibodies (Sigma-Aldrich H9658; 1:100 dilution), followed by goat anti-mouse IgG antibodies conjugated to 6-nm gold particles (Aurion, the Netherlands). For 12 peroxisomes, the distance of each gold particle to the peroxisomal membrane was measured using the line selection tool in Fiji (version 1.53t).