BMDMs were isolated from the tibiae and femurs of wildtype C57BL/6 mice. Cells were cultured in RPMI-1640 (GIBCO Invitrogen, Breda, the Netherlands) with 10% heat-inactivated fetal calf serum (Bodinco B.V. Alkmaar, The Netherlands), penicillin (100 U/ml), streptomycin (100 μg/ml) and L-glutamine (2 mM) (all GIBCO Invitrogen, Breda, The Netherlands), supplemented with 20% L929-conditioned medium (LCM) for 8–9 days to generate BMDMs. After attachment, macrophages were seeded at 350,000 cells per well in 24-well plates and incubated for 24 hours with oxidized low-density lipoprotein (oxLDL) (25 μg/ml; Alfa Aesar, Wardhill, MA, USA). Subsequently, macrophages were treated with 100 µM CTD-002 or GA-12 (Aten Porus Lifesciences Pvt Ltd., India) or with vehicle control [0.1% dimethyl sulphoxide (DMSO)] for 4 hours. CTD-002 is a potent small-molecule active site inhibitor of CTSD designed based on the binding interactions of pepstatin A. Docking studies showed that CTD-002 interacts with polar side chain residues of ASP33 and GLY133 of ligand binding site of the human CTSD (PDB ID: 4od9).

Ldlr^-/- mice, on a C57BL/6 background, were housed under standard conditions and given free access to food and water. Experiments were performed according to the Dutch regulation and approved by the Committee for Animal Welfare of Maastricht University (Project license number: AVD107002016743; working protocol: 2016-003-003). Female Ldlr^-/- mice (8-14 weeks of age) were fed high-fat, high-cholesterol (HFC) diet (containing 21% milk butter, 0.2% cholesterol, 46% carbohydrates and 17% casein; SAFE, Augy, France) for ten weeks and were subsequently divided into three groups (n=17-20 per group; schematic overview in Figure S1 ). To evaluate the potential side effects of intracellular CTSD inhibition, a specific small-compound inhibitor, GA-12, was designed and used in this study. HFC-fed mice were subcutaneously injected with extracellular CTSD inhibitor (CTD-002; 50 mg/kg body weight) or intracellular CTSD inhibitor, GA-12 (50 mg/kg body weight) once every two days. Vehicle treatment (5% DMSO, 40% PEG400, 10% Ethanol, 45% Saline) served as control. Ldlr^-/- mice fed a regular chow diet (9% fat, 67% carbohydrates and 24% protein) for 10 weeks were included as a control group for NASH disease phenotype (n=20). At the end of treatment, animals were euthanized by CO₂ inhalation. Collection of blood and tissue specimens, RNA isolation and cDNA synthesis were performed as described previously (7, 19).

Liver tissue was isolated and snap-frozen in liquid nitrogen and stored at -80°C. The biochemical determination of liver cholesterol and triglyceride levels were carried out as described previously (7). Briefly, 50 mg of frozen liver tissue was homogenized for 30 seconds at 5000 rpm in a closed tube with 1.0-mm glass beads and 1.0 ml SET buffer (Sucrose 250 mmol/l, EDTA 2 mmol/l and Tris 10 mmol/l). Complete cell destruction was done by 2 freeze-thaw cycles and 3 times passing through a 27-gauge syringe needle and a final freeze-thaw cycle. Protein content was measured using the BCA method Pierce, Rockford, IL). Liver cholesterol and triglyceride levels were quantified using cholesterol liquicolor kit [CHOD-PAP, Roche, Basel, Switzerland; triglycerides: GPOtrinder, TRO100, Sigma Aldrich, St. Louis, MO, USA)] following the manufacturer’s protocol on a Benchmark 550 micro-plate reader (Bio-Rad, Hercules, CA, USA).

Immunostainings were carried out on frozen liver sections (7µm) as described previously (23). Briefly, cryosections were fixated in dry acetone followed by blocking with H₂O₂ solution. Next, tissue sections were treated with avidin/biotin solution (Vector; SP2001) for 30 minutes followed by 1 hour incubation with primary antibody [infiltrated macrophages marker, Mac-1; MAB1124; clone M1/70; 1:500)], [F4/80;101201; Biolegio; (1:50)]. Sections later were incubated in secondary antibody (Rabbit anti-Rat IgG Biotin (6180-08), Southern Biotech, Birmingham, USA) which was detected using the Peroxidase Substrate kit AEC (Vector Laboratories, SK-4200, Peterborough, United Kingdom). Haematoxylin was used for nuclear counterstaining. Lipid and collagen content was assessed by using neutral lipid marker Oil Red O (ORO; O0625; Sigma-Aldrich) and Sirius red, (Direct red 80; 43664; Sigma-Aldrich) respectively according to standard protocol. Images were captured with a Nikon digital camera DMX1200 and ACT-1 v2.63 software (Nikon Instruments Europe, Amstelveen, The Netherlands).

CTSD activity was measured using the CTSD activity assay kit (MBL International, Woburn, MA) as described previously (16). Briefly, cell medium was incubated with CTSD substrate and reaction buffer for 1 hour at 37°C. Samples were measured using a fluorescence plate reader with a 328-nm excitation filter and 460-nm emission filter and CTSD activity is expressed as relative fluorescence units.

Fluorescence activated cell sorting (FACS) analysis was performed as described previously (19). Tail vein blood was collected from mice before the start of the treatment (T0) as well at the end of the treatment (T1). Using Trucount tubes (BD Biosciences, Breda, The Netherlands), staining was performed according to the manufacturer’s instructions, to detect the monocyte population (NK1.1-Ly6G-CD11b+; Ly6C). Briefly, heparinized blood samples were mixed and incubated for 10 minutes in the dark at room temperature (RT) with CD16/32 antibody (eBioscience, Halle-Zoersel, Belgium) to block Fc receptor. Samples were then gently vortexed with the appropriate antibodies (PE Mouse Anti-Mouse NK-1.1 (1:100); APC-Cy™7 Rat Anti-Mouse Ly-6G (1:100); Anti-Mouse Ly-6C-APC (1:10) (Miltenyi, Bergisch Gladbach, Germany) and incubated in the dark at RT for 20 minutes (24). All antibodies were diluted in FACS buffer (PBS, 0.1% BSA, 0.01% sodium azide). Finally, samples were mixed and incubated in the dark at RT for 15 minutes with an RBClysis solution (8.4g NH₄Cl + 0.84g NAHCO₃ in 1-liter H₂O, pH 7.2-7.4). Sample stainings were quantified using BD FACSCanto II flow cytometer (BD Biosciences).

Faecal measurement was performed on faeces collected after 24 hours from individually caged mice. Total faecal bile acid levels were specifically determined as described previously (25).

For proteomics experiments, liver homogenates were used (n=3 for each group of mice). The samples were prepared as described previously (26). Briefly, proteins were extracted, reduced with 20 mM dithiothreitol for 45 minutes and alkylated with 40 mM iodoacetamide for 45 minutes in the dark. Alkylation was terminated using 20 mM DDT which is followed by digestion with a mixture of LysC and trypsin at a ratio of 1:25 (enzyme:protein) at 37°C overnight. The digestion was stopped by formic acid. Peptide separation was performed on a Thermo Scientific (Dionex) Ultimate 3000 Rapid Separation UHPLC system equipped with a PepSep C18 analytical column (15 cm, ID 75 µm, 1,9 µm Reprosil, 120Å). Peptide samples were first desalted on an online installed C18 trapping column. After desalting peptides were separated on the analytical column with a 90-minute linear gradient from 5% to 35% Acetonitrile (ACN) with 0.1% FA at 300 nL/min flow rate. The UHPLC system was coupled to a Q Exactive HF mass spectrometer (Thermo Scientific). DDA settings were as follows. Full MS scan between 250 – 1,250 m/z at resolution of 120,000 followed by MS/MS scans of the top 15 most intense ions at a resolution of 15,000.

For protein identification and quantification, DDA spectra were analyzed with Proteome Discoverer version 2.2. The search engine Sequest was used with the Swiss-Prot mouse database (Mus musculus, TaxID 10090). The settings for the database search were similar as before (26). Proteins with a false- discovery rate ≤ 1% were considered for the analysis and were normalized to the total peptide amount. Statistical significance of changes observed in protein abundance was performed using ANOVA. The Benjamini– Hochberg method was used to correct P-values for multiple testing. Proteins were considered differently regulated if the fold change was ≥ 1.5 (log2 ≥0.58) and a p-value of ≤0.05. The differentially regulated proteins were then imported to the EnrichR tool and KEGG database (Verison 2019, mouse) was used to demonstrate the top 10 pathways of up or down-regulated proteins ranked by the combined score which is calculated by multiplying the unadjusted, rather than the adjusted, p-values with the z-scores (27).

Except for proteomics, rest of the data were analysed using GraphPad Prism software (GraphPad, San Diego, California, USA), version 6.0 for Windows. Experimental groups were compared using two-tailed student’s t-test or two-way ANOVA where appropriate or by one-way ANOVA followed by a Tukey post-hoc test. Data are expressed as mean ± SEM and were considered significant at p<0.05, in which *p<0.05; **p<0.01; or ***p<0.001 respectively.

To evaluate their potency and specificity, both the intracellular and extracellular CTSD small-compound inhibitors were tested using in vitro methods. CTSD activity was measured from the medium of bone marrow-derived macrophages (BMDMs) that were treated with either of the inhibitors. CTD-002 (i.e., the extracellular CTSD inhibitor) at a concentration of 100 µM significantly blocked CTSD activity when compared to DMSO (vehicle) as measured from the supernatant of BMDMs. In contrast, the intracellular CTSD inhibitor, GA-12, had no effect on CTSD activity in the supernatant, thus confirming the selective targeting of respective inhibitors ( Figure 1 ).

To test the effects of CTSD inhibition in NASH-associated dyslipidemia, Ldlr-/- mice on HFC diet were injected with either control or intra- or extracellular CTSD inhibitors for a period of 10 weeks. The small-compound CTSD inhibitors showed no visible signs of toxicity in mice. Food intake, body morphology and growth of all mice stayed normal throughout the 10-week experimental period. No abnormalities were detected in any of the animals at autopsy. To investigate the effects of CTSD inhibition on lipid metabolism, hepatic and plasma lipid levels were measured. After ten weeks of HFC diet, both plasma and liver cholesterol and triglycerides were significantly elevated in HFC-fed mice compared to mice on a chow diet, confirming the effect of HFC on hepatic and plasma lipid levels ( Figures S2A–D ). Further, treating HFC-fed mice with either the intracellular or the extracellular CTSD inhibitor decreased hepatic cholesterol levels compared to control mice ( Figure 2A ). However, only upon administration of the extracellular CTSD inhibitor, a reduction in hepatic triglyceride was observed ( Figure 2B ). To confirm the hepatic changes in lipid content at the histological level, we performed an oil red O staining which demonstrated elevated fat deposition in HFC mice compared to chow mice ( Figure S2C ). Moreover, in line with hepatic lipid measurements, extracellular CTSD-inhibitor treated mice showed significant reduction of intracellular and near significant reduction of sinusoidal lipid deposition compared to HFC mice ( Figures 2C, D ) suggesting a clear reduction in hepatic lipid content upon extracellular CTSD inhibition.

To further define the impact of extracellular CTSD inhibition on hepatic lipid metabolism, hepatic gene expression analysis using qPCR was performed. Compared to control and intracellular CTSD inhibitor-treated mice, inhibition of extracellular CTSD led to increased hepatic expression of cytochrome P450 7A1 (Cyp7a1), a rate-limiting enzyme in the breakdown of cholesterol to bile acids, suggesting increased cholesterol conversion only upon extracellular CTSD inhibition ( Figure 2E ). No pronounced effects in other lipid metabolism-related genes were observed ( Figure S4A ). However, plasma lipid levels remained similar among all experimental groups ( Figure 2F ). Altogether, these findings demonstrate that extracellular, rather than intracellular, CTSD inhibition decreases hepatic lipid accumulation in a mouse model for NASH.

To confirm the increased conversion of cholesterol into bile acid precursors, we opted to further investigate bile acid metabolism by assessing faecal bile acid levels. Faecal levels of primary bile acids, chenodeoxycholic acid (CDCA) and ursodeoxycholic acid (UDCA) were significantly enhanced in the extracellular inhibitor-treated group as compared to mice that received the intracellular inhibitor, indicating increased cholesterol breakdown in the former mice ( Figure 3A ). In line, secondary bile acids including α-muricholic acid and β-muricholic acid showed increased faecal levels in extracellular inhibitor-treated mice as compared to intracellular inhibitor-treated mice ( Figure 3B ). No differences in the faecal bile acid levels were observed between control and extracellular inhibitor-treated mice. Collectively, our data suggest that inhibition of extracellular CTSD activity in NASH mice results in increased disposal of cholesterol via increased conversion into bile acids leading to their subsequent excretion.

Next, we examined the role of extracellular CTSD on inflammation at both hepatic and systemic levels. Firstly, inflammatory status was significantly elevated in HFC-fed mice compared to chow mice both at systemic and hepatic levels ( Figures S2E–G ). To explore the status of systemic inflammation upon extracellular CTSD inhibition, we analysed the profile of circulating monocytes using fluorescence-activated cell sorting (FACS). However, no statistical differences were observed in the number of total and pro-inflammatory monocytes (ΔT1-T0) between three different groups of HFC mice ( Figures 4A, B ). Further, indicating the decrease of systemic inflammation upon extracellular CTSD inhibition, extracellular inhibitor-treated mice showed a significant increase in the total number of anti-inflammatory monocytes (Ly6c-) compared to the control mice and intracellular CTSD inhibitor-treated mice (near significant; p=0.07) ( Figure 4C ). However, no significant changes in T-cell population were observed upon extracellular CTSD inhibition ( Figures 4D–F ). Together, these data demonstrate minor reduction in systemic inflammation upon extracellular CTSD inhibition. Next to systemic inflammation, the effects of extracellular CTSD inhibition on hepatic inflammation were assessed by hepatic immunostainings for the inflammatory cell markers Mac-1 and F4/80. The levels of infiltrated macrophages and F4/80 positive macrophage remained unaffected upon extracellular CTSD inhibition ( Figures 4G–J ). Lastly, as expected, the extent of hepatic fibrosis, an advanced NASH feature was no different in chow and experimental groups of HFC as demonstrated by Sirius red staining ( Figure S3 ).

To obtain a comprehensive understanding of the proteomic changes upon extracellular and intracellular CTSD inhibition in the liver, label-free quantitative proteomics was performed. After data processing, a total of 1905 proteins were identified. The number of significantly differentially regulated proteins were plotted using volcano plots ( Figure S5 ). The complete list of significantly different regulated proteins (log2 ≥ 0.58) are mentioned in Tables S1–S6 .

To further understand the functions of these differentially regulated proteins, enrichment analysis was performed to identify specific pathways. The top 10 KEGG (Kyoto encyclopaedia of genes and genomes) pathways that are most significantly enriched are displayed in Table 1 . Compared to control and intracellular inhibitor-treated mice, extracellular inhibitor-treated mice showed enrichment of pathways related to lipid metabolism. Linoleic acid metabolism, steroid hormone biosynthesis, fatty acid biosynthesis and elongation were significantly enriched upon extracellular CTSD inhibition, suggesting improved lipid metabolism in these mice. Moreover, retinol metabolism, PPAR signalling, taurine/hypotaurine metabolism and caffeine metabolism were upregulated in extracellular inhibitor- treated mice ( Table 1A, B ). In addition to lipid-related pathways, ferroptosis, mineral absorption, quinone biosynthesis, ascorbate and aldarate metabolism, NAFLD and chemical carcinogenesis pathways were also enriched upon extracellular CTSD inhibition. Among the downregulated pathways in extracellular inhibitor-treated group, IL-17 pathway (proteins S100A8/A9), African trypanosomiasis, malaria (haemoglobin subunit beta 1), arginine and proline metabolism (creatinine kinase M-type) and glycosaminoglycan biosynthesis pathways (Xylosyltransferase 2) were in top 5 pathways compared to control ( Table 1C ). Interestingly, protein S100A8/A9, which is a ligand complex for Toll-like receptor 4 (TLR-4), was the most significantly downregulated protein suggesting that extracellular CTSD inhibition potentially reduces inflammation in a TLR-4 dependent manner. Moreover, pathways related to cardiac function (hypertrophic and dilated cardiomyopathy, cardiac muscle function), glycolysis, metabolic pathways (glycine, serine, threonine, thiamine, pyruvate, propanoate metabolism) and glucagon signalling were decreased in extracellular inhibitor-treated mice compared to intracellular CTSD inhibitor- treated mice ( Table 1D ).

Most importantly, intracellular inhibitor-treated mice showed downregulation of proteins involved in mitochondrial oxidative phosphorylation and electron transport chain compared to control and extracellular inhibitor-treated mice ( Tables S1 and S4 ). According to KEGG pathways, intracellular inhibitor-treated mice showed enrichment of pathways belonging to muscle contraction and glycolysis compared to control ( Table S7 ). Furthermore, proteins that are involved in hepatic stellate cell activation such as tropomyosin alpha-1 and beta chain, myosin regulatory light chain 2 and myosin 7 were upregulated. Taken together, these findings demonstrate that extracellular CTSD fraction regulates distinct pathways compared to intracellular CTSD in the context of NASH.

To better understand the link between extracellular CTSD and TLR-4 mediated inflammation, we investigated whether lipopolysaccharide (LPS), a TLR4 agonist, influences the inflammatory effects of extracellular CTSD inhibitor. OxLDL-stimulated mouse bone-marrow derived macrophages were treated with extracellular CTSD inhibitor in the presence or absence of LPS. Only in the presence of LPS, extracellular CTSD inhibitor reduced gene expression levels of inflammatory cytokines, Tnfα and Ccl2 ( Figures 5A, B ) suggesting that the anti-inflammatory effects of extracellular CTSD inhibitor are TLR-4 dependent.