Mitochondrial amidoxime‐reducing component 1 p.Ala165Thr increases protein degradation mediated by the proteasome

Metabolic dysfunction‐associated steatotic liver disease (MASLD) is a global health concern with no effective and specific drug treatment available. The rs2642438 minor allele in mitochondrial amidoxime‐reducing component 1 (MARC1) results in an aminoacidic substitution (p.Ala165Thr) and associates with protection against MASLD. However, the mechanisms behind this protective effect are unknown. In this study, we examined the consequences of this aminoacidic substitution on protein stability and subcellular localization.


| INTRODUC TI ON
3][4] MASLD encompasses a wide spectrum of complications and can progress to steatohepatitis, end stage liver disease and hepatocellular carcinoma. 5,6Although there are strategies to effectively reduce liver fat and comorbidities, approved treatments are still lacking. 7,8In this context, the identification of novel therapies is of utmost importance.
3][14] A genome-wide association study (GWAS) identified a variant (rs2642438) in the mitochondrial amidoxime-reducing component 1 (MARC1, also known as MTARC1) gene.The rs2642438 minor allele associates with lower liver triglycerides content and transaminases level and confers protection against cirrhosis and hepatocellular carcinoma. 15,16Mass spectrometric analysis revealed that carriers of the minor allele have higher concentrations of hepatic polyunsaturated phosphatidylcholines compared to non-carriers. 17Interestingly, the protective effect conferred by the mitochondrial amidoxime-reducing component 1 (MARC1) minor allele is also present against alcoholic liver disease. 18RC1 encodes for a molybdenum containing enzyme localized to the outer mitochondrial membrane, that has reductive activity and converts nitrate to nitric oxide. 19,20However, the physiological role of MARC1 in hepatic triglycerides homeostasis remains to be elucidated.
The MARC1 rs2642438 variant is a single nucleotide change resulting in an alanine (A) to threonine (T) substitution at position 165 of the protein.However, the consequences of this aminoacidic change on protein function, stability and subcellular localization are not known.
Therefore, in the present study we attempt to elucidate whether the MARC1 A165T aminoacidic change results in either a loss or gain of function.Our data reveal that overexpression of the MARC1 165T mutant protein results in lower protein levels compared to the overexpression of the A165 wild-type protein.The lower 165T protein level is the result of reduced protein stability due to accelerated proteasomal degradation, while there is no difference in subcellular localization.In conclusion, our data show that the A165T results in a loss of MARC1 function.

| Animals
All experimental procedures were approved by the Gothenburg ethics review committee on animal experiments.Female C57Bl/6N mice were purchased at 7-8 weeks of age from Charles River (Germany) and Results: MARC1 165T overexpression resulted in lower protein levels than A165 both in vivo and in vitro.Similarly, any mutant at position 165 showed lower protein levels compared to the wild-type protein.We showed that the 165T mutant protein is polyubiquitinated and its degradation is accelerated through lysine-48 ubiquitin-mediated proteasomal degradation.We also showed that the 165T substitution does not affect the MARC1 subcellular localization.

Conclusions:
This study shows that alanine at position 165 in MARC1 is crucial for protein stability, and the threonine substitution at this position leads to a hypomorphic protein variant due to lower protein levels.Our result supports the notion that lowering hepatic MARC1 protein level may be a successful therapeutic strategy for treating MASLD.

Key points
We investigated a genetic variation (rs2642438) that protects against liver diseases.This variation causes a single amino acid change in mitochondrial amidoximereducing component 1 (MARC1) and results in its lower protein levels, making it less effective.This may be a successful therapeutic strategy to maintain liver health by lowering liver MARC1 level.

| Adeno-associated virus vectors and synthesis
Vectors were purified by ion-exchange chromatography using HiTrap SP HP column (1 mL, Cytiva, 29051324).Briefly, 293T/17 cells at 72 h post transfection as described above were pelleted and sonicated with the pulse 1 second on and off at 35% amplitude, followed by Benzonase-based treatment (Merck).The treated cells were centrifuged to remove precipitated contaminants, and the supernatant was purified for AAV according to the manufacturer's instructions.The eluted AAV was concentrated using PES protein concentrators (100 KDa MWCO, Millipore) using formulation buffer containing glycerol and Plutonic® F-68 in Dulbecco's Phosphate Buffered Saline (D-PBS, Thermo Fisher).

| Hepatic overexpression of human mitochondrial amidoxime-reducing component 1 in mice
The mice (n = 4/dose group) were given a tail vein injection (100 μL) of either PBS, AAV-eGFP-Luciferase (control), AAV-hMARC1 A165 or AAV-hMARC1 165T at three dose levels: 1.0e10, 3.0e10 and 1.0e11 GC/mouse.Two weeks post dose, animals were anaesthetised with isoflurane and an orbital blood sample was taken with MultivetteR 600 K3E (Sarstedt) for plasma analysis of ALT, AST, total cholesterol and triglycerides.Liver tissue samples from the left lateral lobe were dissected and snap frozen in liquid nitrogen and stored at −80°C pending MARC1 mRNA and protein analyses.

| Hepatic mitochondrial amidoxime-reducing component 1 mRNA quantification
Liver tissue pieces were placed in a plastic tube containing a stainless-steel bead (Qiagen, 69989) and 900 μL of RLT buffer (Qiagen, 79216) and lysed for 2 × 2.5 minutes at 25 Hz using a tissue-lyser.RNA was subsequently purified using RNeasy Mini Kits (Qiagen, 74106).Isolated RNAs were reverse transcribed using the High-Capacity cDNA Synthesis kit (Applied Biosystems, 4368814), and gene expression was measured using real-time PCR with TaqMan Fast Advanced Master mix (Thermo Fisher, 4444556) on a Quantstudio 7 Flex Real-Time PCR machine (Applied Biosystems).

| Mitochondrial amidoxime-reducing component 1 overexpressing plasmids generation and site directed (in situ) mutagenesis
Human wild-type MARC1 A165 cDNA was synthesized and cloned in a pcDNA 3.1(+) vector with a V5-6XHis epitope tag at the Cterminus by GeneArt Gene Synthesis (Thermo Fisher Scientific).The rs2642438 substitution was generated by site-directed mutagenesis introducing a single base-pair change in the wild-type MARC1 gene sequence of the pcDNA3.1(+)plasmid (refer to Supplementary Information).

| Transient transfection
For transient overexpression, HuH-7 or HepG2 cells (cell culture details in suppl.Information) at 70% confluency were transfected using pcDNA3.1(+)with MARC1 wild-type, or mutant forms to the final concentration of 2.5 μg/mL and Lipofectamine 3000 transfection reagent (L3000-075; ThermoFisher Scientific), following the manufacturer's protocol.An empty vector was always used as the negative control.

| Stable cell line generation
HepG2 cell lines stably overexpressing A165 or 165T MARC1-V5 were generated using the pcDNA3.1(+)expression vector containing MARC1 A165 or MARC1 165T plasmid with a 6XHis-V5 epitope tag at the C-terminus of the target protein.An empty vector was used as a negative control.HepG2 cells were transfected using Lipofectamine 3000 reagent.Forty-eight hours post transfection, cells were diluted to 1:20 and seeded in a T150 flask, maintained in MEM media supplemented with 10% FBS, sodium pyruvate, L-glutamine, non-essential amino acids plus 500 μg/mL geneticin-G418 (stable cell media) to select resistant cells.At around 40% confluency, cells were collected and seeded in serial dilutions (1:4 to 1:8000) in a 96-well plate with stable cell media.Monoclonal colonies were gradually expanded.V5 expression levels were verified at each step by immunoblotting.
MARC1 mRNA was checked with real-time qPCR.The sequence was confirmed by Sanger sequencing.

| Immunoprecipitation
Cells stably overexpressing MARC1 A165 or 165T were plated in 6well plates and treated with MG132 (15 μM) for either 8 or 0 hours when they reached approximately 80% confluency.Following treatment, cells were chilled on ice, washed with PBS and lysed with RIPA lysis buffer (ab206996, Abcam) containing protease inhibitor.The lysate was incubated at 4°C for 30 minutes on a rotating shaker and then centrifuged at 10 000 g for 10 min at 4°C.The resulting supernatant was subjected to overnight incubation with anti-V5 magnetic beads (SAE0203, Sigma-Aldrich) on a rotating shaker at 4°C.The following day, the beads were washed three times with PBS, eluted using 2X-SDS-Laemmli buffer and saved for subsequent immunoblotting.

| Gene expression by quantitative polymerase chain reaction
For the in vitro studies, RNA from cells was extracted with RNeasy Plus mini kit (Qiagen) and retrotranscribed using highcapacity cDNA reverse transcription kit (4368813, Thermo Fisher Scientific) as per the manufacturer's protocol.Gene expression was assessed using three technical replicates by real-time qPCR using TaqMan probes [human MARC1 (Hs00224227_m1), human UBC (Hs05002522_g1), human UBE2E1 (Hs00979829), human UBE3C (Hs00904531_m1) and human ACTB (Hs01060665_g1)].Data were analysed using the 2 −ΔΔCt method and normalized to the house keeping gene ACTB.

| Subcellular fractionation
All the subcellular fractionation was performed using HepG2 cells stably overexpressing the MARC1-V5 A165 or 165T.

| Total membrane fractionation
The total membrane fractionation was performed as previously described. 26Briefly, three confluent T150 flasks were collected in PBS, centrifuged and then lysed using Buffer A (50 mM Tris-HCl, 250 mM sucrose, with protease inhibitor at pH 7.0) using 27-gauge needle for 15-20 times.After a brief centrifugation at 1000 g for 1 min, supernatant was collected and ultracentrifuged at 100 000 g for 45 min.The pellet, containing membrane proteins, was then resuspended in 10 mM Tris-HCl (pH 7.4) and divided into equal volume for solubilization.All solvents for solubilization were added to the sample (1:1) and incubated for 1 h at 4°C and then ultracentrifuged at 100 000 g for 1 h at 4°C.The resulting pellets were resuspended in PBS and saved for immunoblotting.

| Mitochondria and mitochondria-associated membrane fractionation
Mitochondria-associated membrane (MAM) fractionation were performed by adapting the protocol already available. 27Briefly, cells from 15 confluent T150 flasks of each A165 and 165T were collected using trypsinization and pooled.They were lysed in 8 mL sucrose homogenization media (.25 M sucrose, 10 mM HEPES at pH 7.6 with protease inhibitor) using a 22G needle 15 times.Next, the lysate was centrifuged at 600 g for 5 min, and the supernatant was collected and further centrifuged at 10 300 g for 10 min at 4°C.The resulting pellet was resuspended in 300 μL, loaded in percoll gradient (10 mL 30% percoll v/v in 225 mM mannitol, 25 mM HEPES and .1 mM EDTA) and ultracentrifuged at 95 000 g for 65 min at 4°C.The upper white band containing the MAM fraction, and several thin lower white bands containing the mitochondrial fraction were collected, washed and further ultracentrifuged to obtain pure MAM and mitochondria fractions.

| Endoplasmic reticulum fraction
The ER was isolated with a discontinuous sucrose gradient as already described. 11,14,27The obtained ER fraction then underwent immunoblotting using anti-V5 to detect recombinant MARC1 and anti-calnexin as a marker for ER.

| Oil-Red-O staining
Oil-Red-O (ORO) staining was performed to evaluate intracellular neutral lipid content after MARC1 silencing.The total area for ORO was determined as previously described. 28HuH-7 cells were seeded in a coverslip at a density of 3 × 10 4 cells/well in 24-well plates and treated after 24 h with scramble siRNA or MARC1 siRNA at 20 nM concentration.Forty-eight hours post-treatment, the cells were washed with PBS, with calcium and magnesium, and fixed with 2% paraformaldehyde (PFA) for 5 minutes.Intracellular lipids were stained with ORO solution (00625, Sigma-Aldrich).Cell nuclei were stained with DAPI (D9542-5MG; Sigma-Aldrich).Slides were mounted with Dako Fluorescence-Mounting Medium (S3023, Agilent Technologies).Images were acquired using an Axio-KS-400 Imaging System and AxioVision 4.8 software (Zeiss) at 40× magnification.
The ORO-stained area and the number of DAPI-stained nuclei were quantified using ImageJ (v.1.52h, NIH).The obtained ORO-stained area was normalized to the number of DAPI-stained nuclei; data shown as ORO Area μm 2 /Nuclei.

| Statistics
All statistical analyses were performed using GraphPad Prism software.For in vivo studies, p values were calculated by oneway ANOVA with Tukey's multiple comparisons test and data are presented as mean ± standard error of the mean (SEM).

| Mitochondrial amidoxime-reducing component 1 165 threonine overexpression results in lower protein levels than 165 alanine in vivo and in vitro
To understand the effect of the alanine to threonine substitution at position 165 of MARC1, we overexpressed human MARC1 A165 or MARC1 165T in mice liver using recombinant adeno-associated viruses (AAVs) at three different dose levels (1e10, 3e10 and 1e11 GC/ mouse), and compared the effects to PBS and AAV control groups (Figure 1).Hepatic MARC1 mRNA expression levels increased dosedependently following rAAV-hMARC1 A165 and 165T administration and at the highest dose level, overexpression of 165T tended to result in higher MARC1 mRNA levels compared to A165 (Figure 1A).However, when assessing hepatic protein levels of MARC1 using targeted mass-spectrometry, overexpression of MARC1 165T resulted in lower protein levels compared to MARC1 A165 at the middle and high AAV dose (Figure 1B).Mouse Marc1 protein levels, mouse Marc1 mRNA levels, body weight, spleen weight and plasma levels of ALT, AST, cholesterol and triglycerides were not affected by rAAV-MARC1 A165 and 165T administration (Figure 1C and Figure S1).In summary, the hMARC1 165T genetic variant is associated with reduced hepatic MARC1 protein levels in vivo in mice.To confirm the in vivo results, we acutely overexpressed MARC1 A165 wild-type or 165T mutant proteins tagged with a V5 at the C terminus in two different hepatocyte cell lines, namely HepG2 and HuH-7.To ensure specificity, we consistently assessed the levels of recombinant MARC1 protein by using a V5 antibody.
Consistent with the in vivo data, overexpression of MARC1 165T-V5 resulted in lower levels of protein as compared to the A165-V5 wild-type in both cell lines (Figure S2A,B,D,E), despite higher amount of mRNA level of the 165T variant (Figure S2G,H).
Following this, we acutely overexpressed MARC1-V5 A165 and 165T in MARC1 knockout HepG2 cells that are homozygous for the wild-type MARC1, that is, A165.MARC1 RNA levels were similar in these knock out cells as compared to wild-type cells indicating that endogenous MARC1 does not affect the acute MARC1-V5 overexpression (Figure S2J).

| Mitochondrial amidoxime-reducing component 1 p.Ala165 is required for protein stability
To understand whether the alanine at position 165 is required for protein stability, we substituted the MARC1 A165 amino acid with several alternative amino acids or with a deletion.Specifically, we deleted the alanine in position 165 of the protein, or changed it into: glycine, the smallest amino acid, proline an amino acid hindering alpha helix, serine the closest amino acid to threonine, valine, the closest amino acid to alanine, tryptophan the largest amino acid, tyrosine an amino acid with a phenyl group and overexpressed these mutant proteins in HepG2 and HuH-7 cells.Overexpression of these mutants showed that, irrespective of the aminoacidic change or deletion, all the mutations resulted in lower levels of MARC1 (Figure 2A,B) despite similar mRNA levels (Figure S2K,L).These data suggest that Alanine 165 is required for maintaining protein levels.

| Mitochondrial amidoxime-reducing component 1 p.165Thr intracellular degradation is accelerated through the proteasome
To understand the mechanism underlying lower protein levels observed with the 165T protein, we generated HepG2 cells stably overexpressing the MARC1 A165-V5 or 165T-V5 tagged proteins.
Stable overexpressing cells significantly recapitulated the phenotype of the acute overexpression (Figure S2C,F,I).Next, we incubated these cells with cycloheximide, an inhibitor of protein synthesis, for 8 h and measured the protein levels.We detected reduced MARC1 165T-V5 mutant protein levels after incubation with cycloheximide (200 μg/mL) and no differences in the A165-V5 wild-type protein levels (Figure 3A and Figure S3A).These data suggest that MARC1 165T-V5 lower levels are due to higher protein degradation.To formally test this hypothesis, we incubated MARC1 A165-V5 and 165T-V5 stable cells with MG132, a proteasome inhibitor, for 8 hours and measured the protein levels.After incubation with 15 μM MG132, we detected a five-fold increase in the levels of the 165T-V5 while the levels of A165-V5 were only doubled at best (Figure 3B and Figure S3B).These data suggest that MARC1 165T-V5 lower levels are due to higher proteasomal degradation.To test whether this was due to ubiquitin-mediated proteasomal degradation, we downregulated ubiquitin C (UBC), ubiquitin conjugating enzyme E2-E1 (UBE2E1) and ubiquitin protein ligase E3C (UBE3C), the key genes encoding proteins initiating ubiquitin-mediated proteasomal degradation.Downregulation of these genes resulted in rescuing the phenotype, namely restoring the 165T mutant protein levels to the same level of A165 wild-type protein (Figure 3C).
To further confirm these data, we examined polyubiquitination of MARC1 A165-V5 and 165T-V5 by immunoprecipitation with V5 antibody, after incubation with MG132.We observed that 165T exhibited higher levels of polyubiquitination after incubation with the proteasome inhibitor than the A165 (Figure 3D).
Proteins to be redirected to the proteasome for degradation are linked to the lysine-48 residue of ubiquitin.To understand whether the lysine-48 of ubiquitin is bound to MARC1, we incubated cells with MG132 for 8 hours, immunoprecipitated with a V5 antibody and then measured levels of this protein by immunoblotting with an antibody specific for ubiquitin lysine-48.We found that MARC1 165T exhibited higher levels of lysine-48 ubiquitination compared to A165 (Figure 3E).These data demonstrate that 165T is more degraded by the proteasome due to higher lysine-48 ubiquitination.Finally, immunoblotting with an antibody specific for lysine-63, which targets proteins for translocation, and regulates protein interactions, 29 revealed no differences between A165 and 165T (Figure 3F).

| Lysosomal mediated and endoplasmic reticulum degradation pathways are not involved in mitochondrial amidoxime-reducing component 1 p.165Thr degradation
In addition to the ubiquitin-proteasome system, there are other pathways for intracellular protein degradation. 30,31Among these, we investigated the endoplasmic reticulum-associated protein degradation (ERAD) and the lysosomal-mediated protein degradation pathway.
To explore potential differences between the A165 and 165T proteins in the ERAD pathway, we incubated A165-V5 and 165T-V5 cells with dibenzylquinazoline-2,4-diamine (DBeQ), for 8 hours and measured MARC1 protein levels.DBeQ (10 μM) is an ERAD pathway inhibitor that selectively reduces p97 ATPase activity.After inhibition of the ERAD pathway, there was no difference in the degradation pattern between the A165 and 165T proteins (Figure 4A and

F I G U R E 3 MARC1 165T shows increased degradation by the ubiquitin-proteasome pathway compared to A165. MARC1-V5 A165 and 165T protein level in HepG2 stable cell lines at baseline and 8 h after incubation with (A) cycloheximide (CHX, inhibitor of protein synthesis) or (B) MG132 (inhibitor of proteasome
).When at 70% confluency, cells were incubated with 200 μg/mL cycloheximide or with 15 μM of MG132 for 8 h.Then, both cells at baseline and cells 8 h after incubation were lysed; an equal volume of total proteins was loaded into 12% SDS-PAGE and immunoblotted against V5 and Calnexin (CNX) antibodies.The intensity of the western blotting bands was measured by Image Lab Software (Bio-Rad), normalized to calnexin and expressed as relative density.Protein levels were presented as relative to baseline values.MARC1-V5 A165 and 165T protein level in HepG2 stable cell lines transfected with scramble-siRNA (siSCR), siRNA for ubiquitin C (siUBC), or a combination of 3 siRNAs, that is, UBC, UBE2E1 and UBE3C (siUBC + E2 + E3), key-role players in the ubiquitination process (C).Cells stably transfected with the empty vector were used as control.Forty-eight hours after transfection, cells were harvested and processed for both qPCR and immunoblotting on 12% SDS-PAGE.To understand the ubiquitination of MARC1, stable cells (A165, 165T or negative control empty vector) were immunoprecipitated using V5-magnetic beads using RIPA buffer to only get ubiquitin from MARC1 and not from any other interacting partners (D).MG132 was added in samples to reduce proteasomal degradation of MARC1.Samples were eluted using 2X SDS-Laemmli buffer, loaded on 8% SDS-PAGE and ran 2.5 h to include highly polyubiquitinated MARC1.Then, they were immunoblotted against ubiquitin antibody.For the quantification of ubiquitin; the whole lane was quantified and presented as relative to MARC1 overexpressing cells.The smear in EV is the non-specific background.(E and F) Immunoblot of K48-linked ubiquitin; or K63-linked ubiquitin after immunoprecipitation.Samples were loaded on 8% SDS-PAGE.Data presented as mean and SD (n = 3).p values calculated by unpaired t-test.*p < .05.
Next, to inhibit the lysosomal-mediated protein degradation pathway, we incubated cells with (CQ, 25 μM) for 8 hours.CQ inhibits autophagy by impairing autophagosome fusion with lysosomes.
After inhibition of the lysosome degradation pathway, there was no difference in the degradation between the A165 and 165T proteins (Figure 4B and Figure S3D).
We also investigated if the MARC1 wild-type and mutant protein would be degraded via the unfolded protein response (UPR).We incubated A165-V5 and 165T-V5 cells for 4 hours with the inhibitors of the UPR pathways (20 μM Ceapin-A7 or .3μM GSK2606414 or .3μM Toyocamycin, along with a vehicle control, DMSO).Ceapin-A7 is a selective blocker of ATF6, GSK2606414 is a cell-permeable protein kinase Rlike endoplasmic reticulum (ER) kinase (PERK) inhibitor, and Toyocamycin is an adenosine analogue acting as an XBP1 inhibitor that affects IRE1 auto-phosphorylation.After inhibition of the UPR pathway, no difference in the protein levels was observed between A165 and 165T (Figure 4C).These findings suggest that the 165T substitution does not affect protein degradation by influencing ERAD, lysosome-mediated protein degradation and the UPR pathway.

| Mitochondrial amidoxime-reducing component 1 is localized in endoplasmic reticulum, mitochondria and mitochondria-associated membrane
To assess whether the protein-membrane interaction was affected by the MARC1 A165T substitution, we incubated the total membrane fraction from HepG2 cells stably overexpressing the 165T-V5 or A165-V5 protein with different solubilizing buffers.We found that MARC1 A165 and 165T were dissociated from the membrane fraction by the detergents, Triton X-100 and SDS, and not by incubation with 1 M NaCl or .2M Na 2 CO 3 .In line with previous studies, 20,32 these data show that MARC1 is attached to endomembranes and are thus consistent with the notion of MARC1 being an integral membrane protein.Moreover, there were no differences in regard to membrane dissociation between the wild-type and mutant proteins (Figure 5A).
To confirm the intracellular localization of MARC1 and to determine whether the amino acid change induces protein mislocalization, we purified the endoplasmic reticulum, mitochondria and MAM fractions from cells stably overexpressing the A165-V5 or 165T protein.We found that both MARC1 A165 and 165T co-localized with mitochondria, MAM fraction and ER fraction with no changes detected between A165 or 165T (Figure 5B,C).These data show that MARC1 is attached to endomembranes and that the 165T variant does not result in membrane displacement.MARC1 downregulation was >90%, resulting in ~50% reduction of intracellular lipids (Figure 6).

| DISCUSS ION
The main finding of this work is that the alanine to threonine substitution at position 165 of MARC1 results in a hypomorphic variant of the protein.This is mediated by proteasomal degradation of the MARC1 protein due to higher lysine-48 ubiquitination in 165T.
Moreover, we found that the aminoacidic change does not affect the protein localization in endomembranes.
Using a liver specific and stable overexpression system in mice, overexpression of human MARC1 165T mutant resulted in lower hepatic protein levels as compared to the A165 wild-type despite similar mRNA levels, indicating a post translational mechanism.This overexpression did not change the endogenous murine Marc1 protein levels and did not affect mouse body weight, liver transaminases or plasma lipid levels.
We recapitulated the in vivo data in two human hepatoma cell lines and generated a series of MARC1 variants at position 165: (a) testing whether the size and nature of the amino acid would be necessary for the deleterious effect by substituting A165 with threonine, serine, tryptophan and tyrosine (b) the predicted crystal structure of MARC1 indicates the 165-alanine residue is located within an alpha helix domain 20 and thus we created in situ mutants that potentially disrupt alpha helix, that is, proline, known to break alpha helix, glycine, the smallest amino acid not favouring alpha helix formation and (c) we hypothesized that by using valine the most similar amino acid to alanine, or by ablating the amino acid at position 165, we would rescue the phenotype by restoring protein levels.Notably, overexpression of all the mutants irrespective of the change resulted in similarly low amount of MARC1 proteins indicating that 165 Alanine is required at that position for protein stability.
Given that the protein level was lower in 165T compared to A165, despite higher mRNA expression, we hypothesized that the aminoacidic change would result in lower protein stability.Indeed, when cells were incubated with cycloheximide, an inhibitor of the F I G U R E 5 MARC1 is localized in endomembranes, and there is no difference between the 165T and A165 in terms of localization.(A) Total membrane fractionation and solubilization: 3 confluent T150 flasks for each of A165 and 165T MARC1-V5 overexpressing stable cells were processed for total membrane fractionation and followed by solubilization in different conditions.In the crude total membrane fraction, pellet (P) contains all the membrane proteins and supernatant (S) contains cytosolic proteins.The crude protein is then incubated with high salt concentration (1 M NaCl) to separate peripheral membrane proteins from integral proteins; .2M Na 2 CO 3 to break proteinprotein interactions without disrupting membranes; mild and denaturing detergents respectively (1% Triton-X, 2% triton-X and 2% SDS).After incubating with solvents, pellet (P) contains the proteins still in the membrane and supernatant (S) fractions contain the displaced proteins from membrane.Calnexin (CNX) is an integral membrane protein and blotted as an internal control.(B) Mitochondrial and MAM fractionation: 15 confluent T150 flasks of each A165 and 165T were used for MAM and mitochondrial fractionation.Cyto, cytosolic fraction; MAM p , pure mitochondria-associated membrane fraction; Mito cr , crude mitochondria; Mito p , pure mitochondria; PNS, post nuclear supernatant.Sample ran in 12% SDS-PAGE, blotted against a mitochondria antibody (60 kDa pyruvate dehydrogenase multienzyme complex (PDC-E2) subunit as a marker for Mitochondria; FACL4 is a marker for mitochondria-associated membrane (MAM).(C) ER fractionation: 2 T150 confluent flasks were used to isolate ER fraction.Samples were run in 12% SDS-PAGE.Calnexin (CNX) was used as the ER marker.Cyto, cytosolic fractions; ER, endoplasmic reticulum fractions; PNS, post nuclear supernatant.cellular protein synthesis, we observed a higher decay of the 165T protein over time as compared to the A165, indicating accelerated degradation.Protein degradation in cells is in large mediated via four different pathways, namely ubiquitin-mediated proteasomal degradation (UPS), endoplasmic reticulum-associated degradation (ERAD), lysosomal degradation and unfolded protein response (UPR). 31In the present study, we show that MARC1 A165 protein is degraded through the proteasomal pathway, with MARC1 165T mutant protein being polyubiquitinated to a larger extent and thus degraded faster.We also show that the higher proteasomal degradation of 165T is mediated by lysine-48 ubiquitination.
Overall, our results support the notion that the MARC1 alanine to threonine substitution at position 165 of the protein results in a hypomorphic variant due to enhanced protein degradation (Figure 7).
Consistent with this model, carriers of the minor allele of a nonsense mutation in MARC1 (p.LOF 200Ter) have lower transaminase levels and reduced risk of cirrhosis. 15RC1 is a protein located in endomembranes and in mitochondria and we tested whether the aminoacidic change may result in a displacement of MARC1 from these membranes.To test this hypothesis, we examined total membrane fractionations from cells stably overexpressing either the wild-type or mutant protein and solubilized the protein using different salts and detergents.
We observed that the MARC1 A165 wild-type and 165T mutant were dissociated upon addition of the detergents Triton X-100 or Sodium dodecyl sulphate, indicating that this is an integral membrane protein.However, we did not find any differences between A165 wild-type and 165T mutant with regard to strength of the protein interaction with membranes and its subcellular localization.
Given our findings that the alanine to threonine substitution at 165 results in a hypomorphic variant of the MARC1 protein, that is, reduced protein levels, and since this variant is associated with Limitation of our study is that we did not examine whether the aminoacidic substitution affects MARC1 enzymatic activity.Future studies with purified enzymatic protein are warranted to answer this question.However, previous studies indicate that MARC1 A165 and 165T may have similar MOCO binding and enzymatic activity 20,[32][33][34] Another limitation is that we did not elucidate the physiological role of MARC1 in hepatocyte triglyceride homeostasis and if MARC1 inactivation may affect triglyceride synthesis, beta-oxidation or very low-density lipoprotein secretion.
In conclusion, the alanine at position 165 of MARC1 is required for protein stability and its Threonine substitution results in a hypomorphic variant.Collectively, the data support the notion that downregulation of MARC1 will be an effective therapeutic strategy to reduce MASLD and ALD.However, future studies will be needed to translate these results in humans and to understand whether this strategy may be effective only in carriers of the MARC1 A165 wildtype or also 165T mutant protein carriers.

ACK N OWLED G EM ENTS
We would like to thank the staff at the AstraZeneca laboratory animal science department for help with the experiments.

Academy
Fellows from the Knut and Alice Wallenberg Foundation, Grant/ Award Number: KAW 2017.0203; the Novonordisk Distinguished Investigator Grant-Endocrinology and Metabolism, Grant/Award Number: NNF23OC0082114; the Novonordisk Project grants in Endocrinology and Metabolism, Grant/Award Number: NNF20OC0063883; the Swedish Cancerfonden, Grant/Award Number: 222270Pj Handling Editor: Luca Valenti several mutants at position 165 by in situ mutagenesis and then examined protein levels.We also generated HepG2 cells stably overexpressing MARC1 A165 or 165T to test the effect of this substitution on MARC1 subcellular localization.

For
in vitro studies, p values were calculated by unpaired Student's t-test (all plots in Figures 2-4, Figure S2D-H) or by Mann-Whitney nonparametric test (Figure 6) and data are presented as mean values ± standard deviation (SD) of at least three experiments unless specified otherwise.
To test whether hepatic downregulation of MARC1 affects the intracellular fat content, we silenced MARC1 in HuH-7 cells cultured in 2D (HuH-7 cells are homozygous for the MARC1 wild-type variant, A165), followed by Oil-Red-O staining.The silencing efficiency of F I G U R E 4 No difference in endoplasmic reticulum (ER) degradation pathways and in lysosomal degradation between MARC1 A165-V5 and 165T-V5.MARC1-V5 A165 and 165T protein level in HepG2 stable cell lines at baseline and 8 h after incubation with (A) ER associated degradation pathway (p97/VCP) inhibitor, DBeQ (10 μM) or (B) Chloroquine (CQ), lysosomal pathway inhibitor.Upon 70% confluency, cells were incubated with 10 μM DBeQ or with 25 μM CQ for 8 h.Then, both cells at baseline and cells 8 h after incubation were lysed; an equal volume of total proteins was loaded into 12% SDS-PAGE and immunoblotted against V5 and Calnexin (CNX) antibodies.The intensity of the western blotting bands was measured by Image Lab Software (Bio-Rad), normalized to calnexin and expressed as relative density.Protein levels are presented relative to baseline values.(C) Upon 70% confluency, MARC1-V5 A165 or 165T overexpressing stable cell lines at baseline and 4 h after incubation with either DMSO (vehicle control), ATF6 inhibitor (iATF6, 20 μM Ceapin-A7), PERK inhibitor (iPERK, .3μM GSK2606414) or IRE1 inhibitor (iIRE1, .3μM Toyocamycin) were lysed.An equal volume of total protein was loaded into 12% SDS-PAGE and immunoblotted against V5 and Calnexin (CNX) antibodies.The intensity of the western blotting bands was measured by Image Lab Software (Bio-Rad), normalized to calnexin and expressed as relative density.Data presented as mean and SD (n = 3).p values calculated by unpaired t-test.
a protection from MASLD, we assessed intracellular fat content in the HuH-7 cell line upon silencing of endogenous MARC1 mRNA expression.Consistently, the intracellular lipid content was reduced after downregulation of MARC1, indicating that the alanine substitution to a threonine at position 165 contributes to protection against F I G U R E 6 MARC1 downregulation decreases intracellular fat content in human hepatoma cells.HuH-7 cells were seeded in 24-well plates (3 × 10 4 cells/well) with DMEM low glucose 10% FBS and treated after 24 h with scramble siRNA or MARC1 siRNA 20 nM.Cells were stained with DAPI and Oil-Red-O (ORO) and visualized under Axio KS 400 Imaging System (Zeiss) at 40× magnification, acquired using AxioVision v4.8 software (scale bar 20 μm).ORO (red channel) was captured in Texasred.Each experiment had on average 12 images, taken in the same acquisition settings.Data shown as average of 6 independent experiments (±SD), with the total ORO area divided by the number of nuclei in the field.The p value was calculated by Mann-Whitney nonparametric test.**p < .01.MARC1 downregulation efficiency in 6 respective experiments as shown in qPCR results.liverdiseases by reducing MARC1 protein levels.However, we did not measure intracellular lipid levels after downregulating MARC1 in human primary hepatocytes.

F I G U R E 7
Amino acid substitution at A165T in MARC1 reduces protein stability by inducing ubiquitin-proteasome pathway.Figure was created with BioRe nder.com.