Phospholipids alter activity and stability of mitochondrial membrane-bound ubiquitin ligase MARCH5

Generation of MARCH5

All cDNAs were purchased from GenScript, and codon-optimised for overexpression in yeast (Saccharomyces cerevisiae). For large-scale overexpression in yeast, the cDNA encoding human MARCH5 was inserted into the pEMBLyex4 plasmid (Cesareni & Murray, 1987) by homologous recombination in S. cerevisiae strain PAP1500 (α ura3-52 trp1::GAL10-GAL4 lys2-801 leu2_1 his3_200 pep4::HIS3 prb1_1.6R can1 GAL) (Pedersen et al, 1996) to encode MARCH5 followed by a TEV-cleavable yEGFP and a His₁₀-tag. DNA-sequencing verified the correct assembly of the expression construct. For homologous recombination, cDNAs for MARCH5 and yEGFP were amplified by PCR and co-transformed into S. cerevisiae PAP1500 with PstI, HindIII, and BamHI digested pEMBLyex4 vector following the protocol of Gietz and Schiestl (Gietz & Schiestl, 2007). The following primers were used:

March5_fwd: 5′-ACACAAATACACACACTAAATTACCGGATCAATTCTAAGATAATTATGCCAGATCAAGCATTGCAACAA-3′5′-ACACAAATACACACACTAAATTACCGGATCAATTCTAAGATAATTATGCCAGATCAAGCATTGCAACAA-3′; March5_rev:5′-AAATTGACTTTGAAAATACAAATTTTCGGCTTCTTCTTGTTCTGGGTAGTT-3′5′-AAATTGACTTTGAAAATACAAATTTTCGGCTTCTTCTTGTTCTGGGTAGTT-3′; GFPfwTEV:5′-GAAAATTTGTATTTTCAAAGTCAATTTATGTCTAAAGGTGAAGAATTATTCACT-3′.

GFPrecrv:5′-CTTCAATGCTATCATTTCCTTTGATATTGGATCATTCAATGGTGATGGTGATGGTGATGGTGATGGTGTTTGTACAATTCATCCATACCA-3′.

Expression and membrane extraction of MARCH5

Large-scale production of MARCH5 was carried out in S. cerevisiae. First, a single colony of transfected cells was selectively propagated at 30°C until saturation in 5 ml of minimal medium (0.5% ammonium sulfate, 2% glucose, salts [64 mM KH₂PO₄, 7 mM K₂HPO₄, 20 mM MgSO₄, 17 mM NaCl, 70 μM CaCl₂, 8 μM H₃BO₃, 0.6 μM KI, 2 μM MnSO₄, 7 μM ZnSO₄, 2 μM CuSO₄, 1 μM FeCl₃, and 1 μM Na₂MoO₄], vitamins [20 μg/l biotin, 2 mg/l pantothenic acid, 2 μg/l folic acid, 10 mg/l myo-inositol, 400 μg/l niacin, 200 μg/l 4 aminobenzoic acid, 400 μg/l pyridoxine, 200 μg/l riboflavin, and 400 μg/l thiamin], and 100 μg/l ampicillin) supplemented with L-leucine and L-lysine (both at 30 mg/l) followed by propagation into minimal medium supplemented only with L-lysine. 2 ml of this culture was then used to inoculate 250 ml minimal medium supplemented with L-lysine. For expression, 2.5 liters YPG medium (1% yeast extract, 2% peptone, and 3% glycerol) supplemented with 0.5% glucose, 70 μM CaCl₂, vitamins and ampicillin was first inoculated to OD₆₀₀ = 0.05 in a 5-L-baffle flask and cultivated for 24 h overnight at 30°C. At OD₆₀₀ = 2.0–4.0 the culture was then cooled down to 15°C and subsequently induced at OD₆₀₀ = 2.5–4.0 with 250 ml induction medium (YPG medium supplemented with 20% galactose). Cells were harvested 48 h after induction by centrifugation (4,000g, 15 min, 4°C). The resulting cell pellet (ca. 80 g cell wet weight from a 9-L culture) was resuspended in Milli-Q water, aliquot in 50 ml Falcons and centrifuged at 4,000g for 10 min. The resulting cell pellet was stored at −80°C before membrane extraction. All following steps were performed a 4°C. Typically, 30 g cells were thawed and resuspended in resuspension buffer (50 mM Tris–HCl pH 7.7, 400 mM NaCl, 10% [vol/vol] glycerol, 1 mM EDTA, and 1 mM EGTA, freshly added 20 mM DTT, 5 mM PMSF, leupeptin [1 μg/ml], pepstatin A [1 μg/ml], and chymostatin [1 μg/ml] [L,P,C-protease inhibitor mix]) 1.2 times the volume of the cell pellet weight. Cells were lysed by bead beating. Therefore, resuspended cells were distributed equally into six 50 ml Falcon tubes and then saturated with glass beads (ø = 500–750 μm, Acros Organics). Every tube was vortexed 5 × 1 min with at least 1 min incubation on ice between each cycle. Beads were then washed with 100 ml lysis buffer, and the pooled cell lysate was centrifuged twice (4,000g, 10 min, 4°C). The resulting supernatant was centrifuged (215,000g, 60 min, 4°C). The membrane pellet were homogenized in a ratio of 1 g per 10 ml solubilisation buffer (50 mM Tris–HCl pH 7.7, 400 mM NaCl, 10% [vol/vol] glycerol, 1 mM TCEP, 0.1 mM EDTA, 0.5 mM PMSF, and L,P,C protease inhibitor mix) using a 30-ml tissue grinder (Wheaton). The homogenized crude membranes were flash-frozen and stored at −80°C.

Purification of MARCH5

All following steps were performed a 4°C. Crude membranes were thawed, supplemented with solubilisation buffer and finalised with 10% (wt/vol) n-Dodecyl β-D-maltoside (β-DDM; Anatrace) to a detergent concentration of 1% (wt/vol), so the final volume yielded 50 ml per g crude membranes. MARCH5 was solubilised for 1 h at 4°C, whereas stirring followed by centrifugation (22,000g, 40 min, 4°C). The supernatant was batch-incubated for 1 h with 1 ml of Nickel High-Performance resin (GE Healthcare) per 50 ml supernatant equilibrated in wash buffer 1 (50 mM Tris–HCl, pH 7.7, 100 mM NaCl, 10 mM imidazole, 5% (vol/vol) glycerol, 0.1 mM TCEP, and 0.01% (wt/vol) 2,2-didecylpropane-1,3-bis-β-D-maltopyranoside (LMNG, Anatrace)) in the presence of 10 mM imidazole. The resin supernatant mix was transferred to an empty PD-10 column resulting in a CV of 1 ml. Subsequently, the resin was washed with 20 column volumes (CV) of Wash buffer 1 followed by 20 CV of Wash buffer 2 (50 mM Tris–HCl pH 7.7, 400 mM NaCl, 90 mM imidazole, 5% [vol/vol] glycerol, 0.1 mM TCEP, and 0.01% [wt/vol] LMNG). Detergent was exchanged on the Nickel column to LMNG. The protein elution was carried out with three CV Elution buffer (50 mM Tris–HCl, pH 7.7, 100 mM NaCl, 500 mM imidazole, 5% [vol/vol] glycerol, 0.5 mM TCEP, and 0.001% [wt/vol] LMNG). GFP-tag was cleaved of MARCH5 using TEV-protease (homemade, 1:2-M ratio), while dialysing against two times 70 volumes of SEC buffer (50 mM Tris–HCl, pH 7.7, 100 mM NaCl, 5% [vol/vol] glycerol, and 0.1 mM TCEP, 0.001% [wt/vol] LMNG) using a Snakeskin dialysis tubing (3.5 kD molecular weight cut off; Thermo Fisher Scientific), for MARCH5-GFP-His₁₀ tag no TEV protease was added during the buffer exchange. Buffer was exchanged after 1 h, and dialysis continued for another 18 h. The precipitation was removed by centrifugation (4,000g, 10 min, 4°C). To remove the cleaved GFP and TEV, the supernatant was incubated with 1 ml Nickel High Performance resin equilibrated in SEC buffer for 30 min. Flow-through and wash (2 CV with SEC buffer) containing cleaved MARCH5 was collected and concentrated using an Amicon Ultra 15 molecular weight cut-off 50 kD to around 1 mg/ml. After, analytical quality control by size exclusion chromatography using a Superose 6 increase 10/300 GL column (GE Healthcare) in SEC buffer, MARCH5 was flash-frozen in liquid nitrogen and stored at −80°C. For MARCH5 in DDM, the same purification procedure was used. However, the detergent was not exchanged and instead of LMNG, 0.01% (wt/vol) DDM in the buffers was used.

Lipid-binding assay

For positive controls, 1 μl of anti-His-HRP (#7074; Cell Signalling) and 1 μl of MARCH5-GFP-His₁₀ in two concentrations (0.35 and 3.5 mg/ml) were pipetted on an empty spot on the lipid membrane strip (# P-6002; Echelon Bioscience). The dry membrane was blocked with 30 ml TBS + 3% (wt/vol) BSA overnight at 4°C, followed by incubation of 1 μg/ml MARCH5-GFP-His₁₀ in TBS + 3% (wt/vol) BSA and 0.01% (wt/vol) DDM or 0.001% (wt/vol) LMNG to ensure MARCH5 will not precipitate during incubation for 4 h at 4°C with soft agitation. The membrane was washed three times with TBS for 5–10 min, subsequently incubated with His-HRP antibody (1:2,000 dilution) in TBS + 3% (wt/vol) BSA for 1 h at room temperature. The previous wash step was repeated, followed by chemiluminescent development (Pierce ECL Western Blotting Substrate, #32132; Thermo Fischer Scientific) using Vilber Smart imaging Fusion Fx spectra or AEC staining Kit (#AEC101-KT; Sigma-Aldrich). For the lipid membrane strip using specific lipids, the 1 μl of lipid (PE (egg, chicken) (#840021), CL (heart, bovine) (#840012), PA (egg, chicken) (#840101), PC (egg, chicken) (#840051), PG (egg, chicken) (#841138), PI(4)P (brain, porcine) (#840045)) was spotted on a PVDF membrane and dried for 1 h. 1 μl of MARCH5 was spotted, serving as a positive control. The assay was performed as described above using 1 μg/ml MARCH5 in TBS + 3% (wt/vol) BSA and 0.001% (wt/vol) LMNG. However, for detection, the membrane was incubated with anti-MARCH5 (1:500, #ab185054; Abcam) and secondary antibody from anti-rabbit HRP (1:2,000, #7074; Cell Signalling) and detected by chemiluminescence using Biorad ChemiDoc Touch gel Imaging System.

Lipid preparation

PE (egg, chicken) (#840021), CL (heart, bovine) (#840012), PA (egg, chicken) (#840101), PC (egg, chicken) (#840051), and PG (egg, chicken) (#841138) were purchased as stock solutions of 10 mg/ml. PI(4)P (brain, porcine) (#840045) was purchased as a 1 mg/ml stock solution. Notably, each lipid class included in this study consisted of a mixture of lipids varying in their chain length and saturation (e.g., CL mix of 18:1 and 18:2 fatty acid chains). The desired amounts of lipids dissolved in chloroform were dried under an argon stream. The lipid films were resuspended in 0.2% (wt/vol) LMNG to a stock solution of 5 mM for a 1:10 M ratio of protein to lipid. For the titration curves of PA, PC, and CL, a separate lipid stock for each molar ratio (1:1, 1:5, 1:10, and 1:50) was prepared to a concentration that 5 μl of resuspended lipids to 35 μl MARCH5 sample resolve in the wanted molar ratio of protein to lipid (final volume of 40 μl). In the case of a used concentration of 32 μM MARCH5, a lipid stock solution of 0.19 mM for a 1:1 M ratio, 0.97 mM for a 1:5 M ratio, 1.94 mM for a 1:10 M ratio, and 9.68 mM for a 1:50 M ratio, was prepared. This allowed a constant detergent concentration in all molar ratios. To the control MARCH5 identical amount of detergent was added as for the lipid samples.

In vitro ubiquitination assay for gel-based analysis

MARCH5 was incubated with lipids at a 1:10 M ratio on ice for 60 min (see lipid preparation section) before being added to the master solution. Therefore, 1.9 μl of 5 mM corresponding lipid stocks were incubated with MARCH5 (32 μM) to a final volume of 30 μl. MARCH5 in the absence of lipids (+ control) included the same amount of LMNG as in the samples with lipid, so detergent concentration remained constant throughout the different MARCH5 lipid samples and negative control containing no E3 ligase. A master solution resulting in final concentration of 40 mM Tris–HCl, pH 7.5, 25 mM NaCl, 10 mM MgCl₂, 0.6 mM DTT, 2 mM ATP, 50 nM human E1 (#E-304; Boston Biochem), 1 μM E2 (UbcH5b, #E2-622; Boston Biochem), and 23 μM Ub (#U6253; Sigma-Aldrich) was mixed and aliquoted. 2.5 μM of MARCH5 incubated with lipids was added to the respective master solution. The reaction volume was 60 μl per reaction mix. The final detergent and lipid concentration in the master solution was 1 CMC (critical micelle concentration) (0.001%), 25 μM, respectively. For the lipid titration ubiquitination assay, it was the same experimental setup, except the reaction volume was 80 μl per reaction mix and the preparation of the MARCH5 lipid samples. 5 μl of desired stock concentration of lipids (see lipid preparation) to achieve the wanted molar ratio of protein to lipid (1:1, 1:5, 1:10, and 1:50) was added to MARCH5 (32 μM), having a final volume of 40 μl and incubated 1 h on ice prior assay. The final detergent concentration was two CMC (0.0019%), and lipid concentration in the ubiquitination assay varied between 2.5 and 125 μM. Reactions were performed at 30°C for 0–60 min or 0–120 min, and for each time point, 14 μl reaction mix was taken out and stopped by the addition of 4 μl LSB. Samples were subjected to SDS–PAGE (8 μL sample loaded on Mini protean TGX gels 4–20%, #4561096; Bio-Rad), and proteins were visualised by Coomassie blue staining (InstantBlue Coomassie protein stain, #ab119211; Abcam) or immunoblotting (2.5 μl sample loaded on gel) using Bio-Rad ChemiDoc Touch gel Imaging System, anti-ubiquitin (1:2,000, #NB300-130; Novusbio) and anti-MARCH5 (1:500, #ab185054; Abcam) and secondary antibody from anti-rabbit HRP (1:2,000, #7074; Cell Signalling), anti-mouse HRP (1:4,000, #NA931; GE Healthcare). For relative quantification of the intensity of MARCH5 and mono-ubiquitin bands, the Image laboratory software (Bio-Rad) was used, and band intensity at time point 0 of the respective samples was defined as 100%. To note, because a substrate of MARCH5 was not available for the assay, we used the ability of many Ub ligases to facilitate ubiquitination even in the absence of their substrates, forming either free ubiquitin chains, auto-ubiquitination or the E1 or E2 enzymes (Lorick et al, 1999).

In vitro de-ubiquitination assay for gel-based analysis

Deubiquitinase USP2 (#E-504; Boston Biochem) was diluted to 2.5 μM in DUB buffer (150 mM NaCl, 50 mM Tris–HCl, pH 7.5, 20 mM DTT, and 1 mM EDTA). After the performance of the in vitro ubiquitination assay for gel-based analysis (see above), 500 nM of the diluted USP2 was added to the ubiquitination mix. The reaction mix was incubated for 1 h at 30°C, before terminating the reaction with LSB. SDS–PAGE analysis was performed as previously described in the in vitro ubiquitination assay for gel-based analysis.

Ubi-real in vitro ubiquitination assay for fluorescent polarisation measurements

MARCH5-lipid samples were prepared as described above in vitro ubiquitination assay for gel-based analysis. A master solution resulting in final concentration of 40 mM Tris–HCl, pH 7.5, 25 mM NaCl, 10 mM MgCl₂, 0.6 mM DTT, 50 nM human E1 (#E-304; Boston Biochem), 1 μM E2 (UbcH5b, #E2-622; Boston Biochem), 23 μM Ub (#U6253; Sigma-Aldrich), and 100 nM F-Ub (#U-580; Boston Biochem) was mixed and aliquoted. 2.5 μM of MARCH5 incubated with lipids was added to the respective master solution or a master solution of MARCH5 without lipids as a + control. Concentrations were calculated so that desired concentrations would be achieved in a final volume of 20 μl. 18.4 μl of the master solution containing MARCH5 incubated with specific lipids were added to a Greiner 384-well, F-bottom, Hibase plate sample wells. The plate was sealed (Clear sealing tape; Molecular Dimensions). After baseline measurement, 1.6 μl ATP were added to each well to a final concentration of 4 μM, except for the–control, which included MARCH5 in the absence of lipids. The addition of ATP interrupted the measurement for around 2 min. The final detergent and lipid concentration in the master solution was 1 CMC (0.001%) 25 μM, respectively. The FP was monitored on a plate reader (Victor Nivo; Perkin Elmer) using a suitable setting for the fluorescein fluorophore with an excitation wavelength of 480, 500 nm dichroic mirror and emission wavelength of 530 nm. FP experiments were 2 h, and FP values were read every minute with a measuring time of 100 ms. In between measurements, the plate was shaken. The baseline was measured for 10 cycles (10 min). Each sample was prepared in four replicates, and the FP signal was averaged at each time point. FP signal was calculated by the Victor Nivo Software, and averages and standard deviations were calculated using GraphPad Prism. To note, fluorescent-labelled ubiquitin has a fluorophore permanently attached to its N-terminus that precludes methionine 1 (Met1)-linked ubiquitin chain formation. However, this ubiquitin linkage type is specific for the ubiquitin chain assembly complex LUBAC, which is the only known mammalian ubiquitin ligase that catalyses Met1-linked ubiquitin linkages (Hrdinka & Gyrd-Hansen, 2017).

Chemical crosslinking in vitro

Prior experiment, MARCH5 was buffer exchanged using Zeba spin column (#89882; Thermo Fisher Scientific) to 50 mM Hepes, pH 7.7, 100 mM NaCl, 5% (vol/vol) glycerol, and 0.1 mM TCEP, 0.001% (wt/vol) LMNG. Bs3 crosslinking reagent (#21586; Thermo Fisher Scientific) was prepared in a 50 mM stock solution by dissolving 10 mg Bs3 in 350 μl of 25 mM potassium phosphate, pH 7.7. MARCH5 was incubated with CL for 1 h before the experiments in a 1:10 M ratio protein to lipid as described in the lipid preparation section. To 35 μl of 19.4 μM MARCH5, 0.35 μl Bs3 was added from the stock solution to a final concentration of 0.5 μM. At each time point (0, 15, 30, and 60 min), 8 μl was taken out of the MARCH5 Bs3 mix, and the reaction was quenched by adding 1 μl 3 M Tris–HCl, pH 7.7 followed by the addition of 4 μl LSB. Crosslinking was analysed by SDS–PAGE.

Nano differential scanning fluorometry (nanoDSF)

Lipids were prepared as stated in the lipid preparation section. MARCH5 (final concentration of 0.5 mg/ml [16 μM]) was mixed with indicated lipids (stock concentration of 5 mM) in a 1:10 M ratio in a sample volume of 40 μl. The final detergent concentration was 5 CMC (0.005%). For the lipid titration experiments, 5 μl of desired stock concentration of lipids (see lipid preparation) to achieve the wanted molar ratio of protein to lipid (1:1, 1:5, 1:10, and 1:50) was added to MARCH5 (32 μM), having a final volume of 40 μl. After MARCH5 was diluted to a concentration of 0.5 mg/ml (16 μM) in buffer (50 mM Tris–HCl, pH 7.7; 100 mM NaCl; 5% glycerol; 0.1 mM TCEP; 0.001% [wt/vol] LMNG), the final detergent concentration was 12.5 CMC (0.0125%). Samples were incubated for 1 h at 4°C. Real-time simultaneous monitoring of the intrinsic tryptophan fluorescence at 330 and 350 nm was performed using a Prometheus Panta (Nanotemper) measuring a temperature range of 15°C–95°C with a temperature gradient of 1°C per minute. A minimum of triplicates was used for all experiments. Protein melting temperatures were determined as maxima of the first derivatives of the 350/330 nm ratio, analysed by Prometheus software.