Lipase maturation factor 1 affects redox homeostasis in the endoplasmic reticulum

The construct for LMF1 with a C‐terminal polyhistidine tag was previously described (Babilonia‐Rosa & Neher, 2014). Site‐directed mutagenesis of LMF1‐His was employed to obtain C‐to‐A mutations. Plasmids pSWAN‐GFP37TAG and pSWAN‐pBpaRS allowed for expression of suppressor tRNA and tRNA synthetase for incorporation of pBPA for photocrosslinking and were provided by P. Schultz (Liu et al, 2007). AvrII and BsrGI sites were used to remove GFP from the pSWAN‐GFP37TAG and insert LMF1 with an 8X C‐terminal polyhistidine tag. The amber codon was incorporated in the pSWAN‐LMF1 in residue F262 using site‐directed mutagenesis to generate pSWAN‐Lmf1‐F262Amber. For LPL‐V5 and LPL‐His, human LPL was excised from pCMV‐SPORT6‐LPL (Open Biosystems), a C‐terminal V5 or His tag was added, and it was inserted into a pcDNA5⁄FRT vector (Thermo Fisher Scientific) using HindIII/BcII sites. To make the plasmid for expression of Txnip, the cDNA sequence of human Txnip was synthesized as a GBlock by IDT. It was digested using HindIII and XhoI and inserted in pCNDA5/FRT. Plasmids for expressing mCherry‐Sec61β (#49155), YPet‐tagged fibronectin (#65421), thioredoxin (#21614), and pSpCas9n‐2A‐Puro (#48141) were obtained from Addgene. Thioredoxin with a C35S mutation was generated using Quickchange. For CRISPR/Cas, plasmids pSN348 and pSN349 were generated from pGL3‐U6‐sgRNA‐PGK‐ (Addgene plasmid #51133) by cutting with Bsa1 and inserting annealed oligo pairs 1 (ccgggaagactgggtactcggatc and aaacgatccgagtacccagtcttc) and 2 (ccggccgcgggcgccgccattgtt and aaacaacaatggcggcgcccgcgg). The ERroGFP‐S4 plasmid was a gift of Yasuyoshi Sakai group and Jun Hoseki at Kyoto University (Hoseki et al, 2016). The LDLR‐GFP plasmid was a gift from Elizabeth Tarling at UCLA. A plasmid for expression of human ERdj5 with a V5 tag was obtained from the UNC Chapel Hill Lenti‐cDNA Core facility (ORFeome clone # BC117299, RefSeq Gene NM_018981.2).

Stable HEK293 Flp‐In™ (Thermo Scientific) cells lines were generated with LPL‐V5, LPL‐His, or LMF1‐His as per manufacturer's direction. Specifically, Fugene 6 (Promega) was used as a transfection reagent and cells were selected with 200 μg/ml of hygromycin. HEK293 cells were transiently transfected with Fugene 6 and 1 μg of DNA for LMF1‐His and the respective lipase construct for the DSP crosslinking experiment. cld/cld and cld/wt MEFs (Boedeker et al, 2001) were obtained from the Doolittle laboratory. Transfection of cld/cld cells on six‐well plates was performed with 2 μg of DNA and X‐tremeGENE HP per manufacturer's instructions. All cell lines were maintained in Dulbecco's modified Eagle's medium with 10% fetal bovine serum, 1% penicillin/streptomycin, and 2 mM l‐glutamine (complete media).

CRISPR‐Cas‐mediated LMF1 deletions were generated essentially as described (Shen et al, 2014). Briefly, HEK 293 Flp‐In™ (Thermo Scientific) cells were transfected with 0.66 μg each of pSpCas9n‐2A‐Puro, pSN348, and pSN349 using Fugene 6 (Promega). Cells were split into plates coated with fibronectin and selected with 2 μg/ml of puromycin until untransfected cells died. Single colonies were selected, split into 96‐well plates, and tested for their ability to secrete LPL. Genomic DNA was harvested from cells not secreting LPL, and the first exon of LMF1 was amplified by PCR, cloned, and sequenced.

siRNA transfections were performed with 20 nM of RNA oligonucleotide and 7 μl of Lipofectamine RNAiMax (Life Technologies) per six‐well plate, and samples were collected 48 h post‐transfection. Validated Silencer Select siRNAs were obtained from Life Technologies: negative control No. 1 (4390843), ERp44 (s22965), PDI (s439), ERp72 (s18446), UGGT1 (132932), UGGT2 (112074), and ERdj5 (132773). Knockdown of protein expression was quantitated by Western blot using a ChemiDoc MP Imaging System from Bio‐Rad and Image Lab Software. The percent of remaining protein expression was calculated with the following formula: ((Candidate/GAPDH)/(NC/GAPDH))*100. Experiments were done in triplicate.

For pull‐down assays of LMF1‐His from HEK Flp‐In stable cell lines, two T‐75 flasks per condition were seeded with 7.5 × 10⁶ cells. LMF1‐His expression was induced with 2 μg/ml tetracycline the next morning. The next day, cells were trypsinized, washed with 1× PBS, crosslinked with 2 mM DSP (ProteoChem) with agitation for 30 min at room temperature, and quenched with 150 mM Tris pH 8. The cells were lysed by douncing in buffer 1 (250 mM NaCl, 10 mM Tris pH 8, and 1 mM PMSF). Following centrifugation for 20 min at 15,600 × g, pellets were resuspended by douncing with 1.4 ml of buffer 1. Lysates were incubated overnight with agitation and the addition of 20 mM fos‐choline 12 (Anatrace). After centrifugation as above, 5% glycerol, 40 mM imidazole, and Ni‐NTA agarose beads (Qiagen) were added to the supernatant. Beads were bound for 1 h at 4°C and then washed three times with 4 ml of buffer 2 (1 M NaCl, 10 mM Tris pH 8, 3 mM fos‐choline 12.5% glycerol, 40 mM imidazole, and 1 mM PMSF) in 11 ml poly‐prep chromatography columns (Bio‐Rad). LMF1 complexes were eluted with buffer 3 (250 mM NaCl, 10 mM Tris pH 8, 3 mM fos‐choline 12, 400 mM imidazole, and 1 mM PMSF). Eluate fractions were loaded on 8% SDS–PAGE, transferred to PVDF (Bio‐Rad) for 75 min at 100 V, and blocked with 5% non‐fat milk or BSA.

For photocrosslinking with pBpa, HEK293 cells were plated in six‐well plates at 10.5 × 10⁵ cells per well; four wells were utilized per sample. The next day, 1 μg of pSWAN‐pBpaRS and pSWAN‐Lmf1‐F262Amber was transfected with Fugene 6. After 24 h, the media was replaced with media containing 2 μg/ml of tetracycline, 0.5 mM pBpa (Chem‐Impex International), and 0.3% of DMSO. Control wells lacked pBpa but contained 0.3% of DMSO. Cells were grown with pBPA for 24 h. The media was changed to 1× PBS, and the cells were kept on ice and crosslinked with a UV lamp at a wavelength of 365 nm for 30 min at 2.5 cm from the lamp. LMF1 was purified from cells as described above.

HEK293∆LMF1 were transfected with LMF1 C‐to‐A constructs 24 h before media collection. Three hours before media collection, lipase secretion was promoted by the addition of 600 μl of DMEM with 1% FBS and 15 USP units/ml of heparin per 9.5 cm² well. The amount of lipase in media, soluble cell fraction and insoluble cell fraction, was measured by Western blot. Cells were trypsinized and lysed with lysis buffer (150 mM NaCl, 1.0% Triton X‐100, 50 mM Tris pH 8.0) and then separated by centrifugation. The soluble lysate fraction was removed, and the pellet fraction was solubilized with 5× SDS–PAGE sample buffer. Media, lysate, and pellet fractions were separated using 12% SDS–PAGE.

HEK293 + LMF1 cells were plated in a six‐well plate at 0.5 × 10⁶ cells/ml. Cells were mock‐transfected or transfected with 1 μg of the Txnip plasmid. LMF1‐His expression was induced overnight with 2 μg/ml tetracycline, and the next morning, cells were precipitated with ice‐cold, 8% TCA, washed with acetone, and resuspended with an automated homogenizer in 50 mM Tris pH 8, 4% SDS with or without 10 mM AMS. Prior to precipitation, cells were either untreated or treated with 1 mM H₂O₂ for 30 min or 10 mM DTT for 10 min. Samples were separated on 12% SDS–PAGE for 3 h at 180 V and then analyzed by Western blot as described below.

HEK293 + LMF1 cells or HEK293∆LMF1 were plated in a T75 flask at a density of 5 × 10⁶ cells 48 h prior to use. Each flask was transfected with 12 μg of either mCherry (control), ERdj5‐V5, or Trx‐C35S plasmid 24 h before harvesting. Expression was induced with tetracycline at the time of transfection. At harvesting, cells were washed with cold 10 mM NEM in PBS, trypsinized, and pelleted. Cells were resuspended in PBS and TCA precipitated with 8% TCA on ice for 30 min and pelleted at 13,000 × g at 4°C for 10 min. Pellets were resuspended in denaturing alkylating buffer (6 M urea, 1% SDS, 100 mM Tris pH 7.5) plus 50 mM NEM. Lysates were diluted 1:10 in Triton buffer (2% Triton‐X100, 50 mM Tris pH 7, 150 mM NaCl, 6 M urea) plus 20 mM NEM. Fos‐choline 12 was added to 3 mM, and resuspended lysates were incubated overnight at 4°C. Lysates were cleared by centrifugation at 13,000 × g for 5 min at 4°C. Cleared lysates were batch bound to 150 μl of Ni‐NTA Excel (GE Healthcare) resin for 2 h at 4°C. Beads were washed 4 × with 3 ml of Triton buffer plus 10 mM imidazole and then eluted 2 × 150 μl with Triton buffer plus 500 mM imidazole. Eluted samples were incubated with 2.5 M urea in 2× SDS loading dye with or without 50 mM DTT for 10 min at room temperature and then heated to 72°C for 4 min before resolution by SDS–PAGE. Westerns were performed as described below.

For Western blots, samples mixed with 5× SDS loading dye (0.2 M Tris pH 8.0, 1% SDS, 30% glycerol, ±40 mM DTT) and separated using SDS–PAGE. Proteins were then transferred to PVDF using a wet transfer (Bio‐Rad). Blots were blocked for 2 h. Blots were then incubated with antibodies at dilutions listed below for 1 h at room temperature. Blots were washed three times in TBST for 10 min each. Blots were incubated with anti‐rabbit or anti‐mouse HRP secondary antibody at 1:20,000 in 2.5% milk/TBST for 45 min and then washed three times in TBST. Blots were incubated with ECL substrate (ADVANSTA WesternBright, BioExpress) and proteins detected with a ChemiDoc MP Imaging System from Bio‐Rad and Image Lab Software. For Western blots of LPL and PL aggregates, gels were soaked in 10 mM DTT for 10 min prior to transfer.

LMF1‐His was detected with an anti‐His antibody (1:5,000 AbD Serotec). PDI, ERp72, and ERp44 antibodies were obtained from Cell Signaling, used at 1:1,000 in 5% BSA, and with HRP‐conjugated anti‐rabbit (Southern Biotech) at 1:5,000. UGGT1, UGGT2, ERdj5, TRX, and TXNIP antibodies were obtained Abcam; used at 1:500, 1:1,000, 1:500, 1:1,000, and 1:600, respectively, in 5% BSA with HRP‐conjugated anti‐mouse or anti‐rabbit antibodies from Southern Biotech. Mouse anti‐V5 antibody from Bio‐Rad was diluted 1:5,000 in 5% non‐fat milk in TBST. The c‐Myc antibody (MA1980, Invitrogen) was used at a dilution of 1:1,000 in 5% non‐fat milk in TBST.

Assays were based on previously described methods (Francis et al, 2002; Poet et al, 2017). To generate mRNA, the pCMV‐SPORT6‐LPL template plasmid was linearized with HindIII. RNA was transcribed using the mMessage mMachine SP6 Transcription kit (Invitrogen) per the manufacturer's directions. To generate SP cells, HEK293 or HEK293∆LMF1 was grown to 70% confluence, trypsinized, pelleted, and resuspended in KHM buffer (110 mM KOAc, 2 mM MgOAc, 20 mM HEPES pH 7.2) with Digitonin (20 μg/ml) and then incubated for 8 min on ice. Cells were pelleted at 100 × g for 4 min at 4°C and resuspended in potassium acetate/HEPES buffer (90 mM HEPES pH 7.2, 50 mM KOAc pH 7.2) and incubated for 10 min on ice. Cells were pelleted again at 100 × g for 4 min at 4°C and resuspended in micrococcal nuclease buffer (KHM supplemented with 1.5 mM CaCl₂) with micrococcal nuclease (NEB) added for 12 min at room temperature. Digestion was arrested with 4 mM EGTA, and cells were pelleted at 100 × g for 4 min at 4°C, washed with KHM buffer, and resuspended in 50 μl KHM buffer. Cells were counted and used immediately.

Metabolic labeling and pulse‐chase analysis: Translation mix was prepared using the Flexi rabbit reticulocyte lysate (Promega) with the following components added to a final volume of 25 μl: 16.5 μl lysate, KCl at 70 mM, amino acid mixture minus methionine at 200 μM, RNasin (Promega) at 40 U/μl, 750 ng RNA, 8.8 × 10⁴ SP cells, 11 μCi of EasyTagTM EXPRESS³⁵S mix (PerkinElmer), and nuclease‐free water. Mixtures were incubated at 30°C for 1 h, and then, 1 mM cycloheximide (Sigma) was added to stop translation. At the indicated time points, 15 μl of reaction mix was removed and added to either SDS loading dye and put on ice or 14 mM AMS (Invitrogen) and incubated for 1 h in the dark. 1× SDS loading dye was added to all samples and boiled for 3 min at 80°C. Samples were spun down, and lysates of equal volumes were separated by SDS–PAGE. SDS Gels were dried, exposed, and imaged on a Typhoon FLA 9500 (GE), and quantification was carried out in FiJi.

Gel slices containing purified LMF1 with pBPA‐ or DSP crosslinked‐binding partners were analyzed by the Duke proteomic facility. Control gel slices were obtained from samples prepared without pBpa (pBpa experiment) or without tetracycline for induction (DSP). Coomassie‐stained SDS–PAGE bands were subjected to standardized in‐gel trypsin digestion. Extracted peptides were lyophilized to dryness and resuspended in 12 μl of 0.2% formic acid/2% acetonitrile. Each sample was subjected to chromatographic separation on a Waters NanoAcquity UPLC equipped with a 1.7 μm BEH130 C₁₈ 75 μm I.D. X 250 mm reversed‐phase column. The mobile phase consisted of (A) 0.1% formic acid in water, and (B) 0.1% formic acid in acetonitrile. Following a 3 μl injection, peptides were trapped for 3 min on a 5 μm Symmetry C₁₈ 180 μm I.D. X 20 mm column at 5 μl/min in 99.9% A. The analytical column was then switched in‐line, and a linear elution gradient of 5% B to 40% B was performed over 30 min at 400 nl/min. The analytical column was connected to a fused silica PicoTip emitter (New Objective, Cambridge, MA) with a 10‐μm tip orifice and coupled to a Q Exactive Plus mass spectrometer (Thermo Scientific) through an electrospray interface operating in a data‐dependent mode of acquisition. The instrument was set to acquire a precursor MS scan from m/z 375–1,675 with MS/MS spectra acquired for the 10 most abundant precursor ions. For all experiments, HCD energy settings were 27v and a 120 s dynamic exclusion was employed for previously fragmented precursor ions.

Raw LC‐MS/MS data files were processed in Proteome Discoverer (Thermo Scientific) and then submitted to independent Mascot searches (Matrix Science) against a SwissProt database (Human taxonomy) containing both forward and reverse entries of each protein (20,322 forward entries). Search tolerances were 5 ppm for precursor ions and 0.02 Da for product ions using trypsin specificity with up to two missed cleavages. Carbamidomethylation (+57.0214 Da on C) was set as a fixed modification, whereas oxidation (+15.9949 Da on M) and deamidation (+0.98 Da on NQ) were considered dynamic mass modifications. All searched spectra were imported into Scaffold (v4.3, Proteome Software), and scoring thresholds were set to achieve a peptide false discovery rate of 1% using the PeptideProphet algorithm. The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD009641 (https://doi.org/10.6019/PXD009641).

To assess cell sensitivity to BSO, DTT, and tunicamycin, first 1 × 10⁵ cells/well of cld/cld or cld/wt were plated in a six‐well plate. The next day, either vehicle control, 10 mM L‐BSO, 2.5 μg/ml tunicamycin, or 5 mM DTT was added. After 24 (tunicamycin or BSO) or 12 (DTT) hours of treatment, cells were removed from the plate with trypsin, stained with trypan blue, and live cells were manually counted using a hemocytometer. For HEK293/HEK293∆LMF1 cells, cells were plated in six‐well plates coated with fibronectin (Invitrogen). Cells were treated with 2 μg/ml Tunicamycin for 20 h, with 10 mM L‐BSO for 24 h or 10 mM DTT for 12 h. Cells were trypsinized, stained, and live cells were counted on a Countess II cell counter (Invitrogen). For both cell types, survival was reported relative to vehicle control. Three wells were counted for each condition and the average survival was calculated, and three of these independent experiments were carried out per condition.

For all experiments, cells were plated in poly‐d‐lysine coated 33‐mm glass bottom Petri dishes, with 14 mm microwell, No. 1.5 cover glass (MatTek). For characterization of fibronectin in HEK293, HEK293∆LMF1, cld/wt, and cld/cld cells, 1.3 × 10⁵ cells were transfected with X‐tremeGENE (Roche) with 1.2 μg plasmid expressing fibronectin‐YPet and mCherry‐Sec61β. The transfection media was removed 6–7 h post‐transfection, and the cells were washed with PBS and then returned to culture media. Images were taken 24 h post‐transfection at 37°C in a CO₂ chamber with a Zeiss LSM 710 confocal microscope and a Plan Apo 40×/1.4 NA oil objective. For quantification, the images were first deconvolved using the AutoQuant 3D Deconvolution function. A threshold for each probe was then determined using the Threshold Plugin of Fiji. Using these thresholds, colocalization percentages were calculated using Imaris (Bitplane).

For characterization of the redox environment in cells, HEK293 or HEK293∆LMF1 cells were transfected with 1 μg ERroGFP‐S4 plasmid and in some experiments 0.4 μg LPL‐mCherry plasmid with Fugene HD (Promega) 24 h prior to imaging. Cells were imaged on a Zeiss LSM 710 laser scanning microscope as described above. All images were taken at identical gain, offset, and laser intensities between groups. To assess cellular redox state, ERroGFP was excited by 405 and 458 nm lasers and emission was collected between 504 and 533 nm in independent frames. Background intensity was subtracted automatically with Fiji, and three random ROIs were drawn on each cell body and mean intensity was measured in Fiji. The mean intensity of tracks 405 to 458 were compared to generate a ratio describing the oxidized GFP state (405 nm) to reduced GFP state (458). For each of three independent experiments, three spots from 17 cells were imaged. Representative data are shown.