Characterization of SPP inhibitors suppressing propagation of HCV and protozoa

Significance

Signal peptide peptidase (SPP) is an essential host factor for propagation of hepatitis C virus (HCV). Here, we show that dibenzoazepine-type γ-secretase inhibitors suppressed the maturation of all genotypes of HCV core proteins through a specific interaction with Val223 in SPP, and no drug-resistant virus emerged after several passages of HCV in the presence of the SPP inhibitors. In addition, SPP encoded by Plasmodium falciparum was functionally similar to human SPP, and treatment with the SPP inhibitors suppressed the propagation of protozoa, including P. falciparum and Toxoplasma gondii. Structural analysis in silico revealed that Phe258 of SPP participates in binding to the inhibitors. Compounds possessing a high affinity to Val223/Phe258 in SPP might be novel therapeutics for chronic hepatitis C and protozoiasis.

Abstract

Signal peptide peptidase (SPP) is an intramembrane aspartic protease involved in the maturation of the core protein of hepatitis C virus (HCV). The processing of HCV core protein by SPP has been reported to be critical for the propagation and pathogenesis of HCV. Here we examined the inhibitory activity of inhibitors for γ-secretase, another intramembrane cleaving protease, against SPP, and our findings revealed that the dibenzoazepine-type structure in the γ-secretase inhibitors is critical for the inhibition of SPP. The spatial distribution showed that the γ-secretase inhibitor compound YO-01027 with the dibenzoazepine structure exhibits potent inhibiting activity against SPP in vitro and in vivo through the interaction of Val223 in SPP. Treatment with this SPP inhibitor suppressed the maturation of core proteins of all HCV genotypes without the emergence of drug-resistant viruses, in contrast to the treatment with direct-acting antivirals. YO-01027 also efficiently inhibited the propagation of protozoa such as Plasmodium falciparum and Toxoplasma gondii. These data suggest that SPP is an ideal target for the development of therapeutics not only against chronic hepatitis C but also against protozoiasis.

Hepatitis C virus (HCV) is a major causative agent of chronic liver diseases, including steatosis, cirrhosis, and hepatocellular carcinoma (1). HCV infection is also epidemiologically correlated with extrahepatic manifestations such as type 2 diabetes, mixed cryoglobulinemia, and non-Hodgkin lymphoma (2). Combination therapy using pegylated IFN plus ribavirin achieved an ∼50% sustained virological response (SVR) in patients infected with genotype (GT) 1b with high viral loads (3). With the introduction of direct-acting antivirals (DAAs) targeting viral protease and polymerase (4), over 90% of chronic hepatitis C patients achieved SVR (5, 6). However, HCV that is resistant to DAAs has already been reported (7), and thus there is need for novel therapeutics with a low frequency of emergence of breakthrough viruses.

HCV, which belongs to the family Flaviviridae, possesses a positive-sense and single-stranded RNA genome and has seven major genotypes (GT1–GT7) (8). The viral RNA is translated into a large single polyprotein (∼3,000 amino acids) and is processed into 10 viral proteins through cleavage by host and viral-encoded proteases. The core protein of HCV is a component of viral capsids, and E1 and E2 are envelope glycoproteins. The p7 protein acts as a proton pump for the efficient release of the virus. Nonstructural (NS) 2 and 3 proteins possess protease activities. NS4 is thought to be a scaffold for the viral replication complex. NS5A interacts with various host factors and regulates viral replication. NS5B has RNA-dependent RNA polymerase activity (9, 10). HCV is trapped by glycosaminoglycans such as heparin and heparan sulfate on the cell surface and then is transferred to protein receptors. HCV enters the cells via endocytosis, replicates on the endoplasmic reticulum (ER) membrane, and buds into the ER lumen.

The core protein of HCV is a multifunctional protein present in many cellular components, such as the nucleus, ER, lipid droplets, lipid rafts, and mitochondria. The core protein participates in apoptosis, autophagy, cell cycles, and oncogenesis (11–13). Although the precise molecular mechanisms of HCV-induced pathogenesis remain unknown (14), liver-specific HCV core transgenic (CoreTg) mice have been reported to exhibit insulin resistance, steatosis, and hepatocellular carcinoma (15, 16). Sterol regulatory element binding protein 1c (SREBP-1c), which positively regulates the production of saturated and monounsaturated fatty acids and triglycerides, is enhanced in the CoreTg mouse liver (17), suggesting that the core protein participates in the development of liver diseases and extrahepatitis manifestations.

HCV core protein is cleaved off from the polyprotein at amino acid position 191/192 by the host signal peptidase, and the signal sequence in the C-terminal region is further cleaved by the intramembrane aspartic protease known as “signal peptide peptidase” (SPP) to form a mature core protein (18). We previously demonstrated that the maturation of HCV core protein by cleavage with SPP is essential for the stable expression of the core protein and the induction of steatosis in CoreTg mice, and we observed that immature core protein unprocessed by SPP was degraded by proteasomes after ubiquitination by TRC8, an E3 ubiquitin ligase (19). Those findings suggested that SPP would be an ideal target for the development of therapeutics against chronic hepatitis C.

SPP has nine transmembrane domains that are members of the family of GxGD-type intramembrane cleaving proteases, including γ-secretase (20). The SPP gene is also encoded in protozoa such as Plasmodium falciparum and was suggested to be essential for their survival (21). SPP was also reported to cleave cellular substrates such as the histocompatibility antigen HLA-E (22), the unsplicing variant of X-box binding protein 1 (XBP1μ) (23), heme oxygenase-1 (HO-1) (24), and viral proteins (25–29). Although some of the inhibitors for γ-secretase inhibit SPP (19), the molecular mechanisms underlying the ability of the compounds to inhibit γ-secretase and SPP remain unclear.

In our present investigation of the structure–activity relationship of the γ-secretase inhibitors, we determined that the dibenzoazepine-type structure of γ-secretase inhibitors plays a crucial role in their inhibition of SPP, and we identified a compound, YO-01027, that exhibits potent inhibition of SPP in vitro and in vivo. Our findings also demonstrate that human SPP (hsSPP) and P. falciparum SPP (pfSPP) are functionally similar and that YO-01027 suppresses the propagation of protozoa, including P. falciparum and Toxoplasma gondii. In addition, the treatment of CoreTg mice with YO-01027 suppressed core-induced steatosis. We also show that treatment with YO-01027 suppressed the production of infectious particles of HCV by inhibiting the maturation of HCV core proteins and that no drug-resistant virus emerged even after 15 passages in the presence of the inhibitor. In a subsequent analysis using an in silico structural-based approach, we demonstrate that Val223 of SPP interacts with the dibenzoazepine structure of YO-01027 and that Phe258 of SPP participates in binding to inhibitors for SPP and γ-secretase. These data provide clues for the development of novel SPP-targeting therapeutics for chronic hepatitis C and protozoiasis.

Materials and Methods

Cell Lines.

Mouse embryonic fibroblasts (MEFs), a human embryonic kidney cell line (HEK293T), and two human hepatocellular carcinoma cell lines (Huh7 and Huh7.5.1) were cultured in DMEM supplemented with 10% FBS, 100 U/mL penicillin, and 100 µg/mL streptomycin. GT1 (Con1 strain) (30) and GT2 (JFH1 strain) (31) replicon cells were cultured in DMEM supplemented with 10% FBS, 100 U/mL penicillin, 100 µg/mL streptomycin, and 1 mg/mL of G418.

Mice.

Mice were maintained under a 12-h light/dark cycle (lights on at 08:00 AM) at room temperature (23 °C ± 2 °C). All animal experiments conformed to the Guidelines for the Care and Use of Laboratory Animals (32) and were approved by the Institutional Committee of Laboratory Animal Experimentation of the Research Institute for Microbial Diseases, Osaka University. All efforts were made to minimize animal suffering and to reduce the number of animals used in the experiments.

Preparation of HCV and T. gondii.

HCV derived from the GT2a JFH-1 strain mutated in E2, p7, and NS2 as shown (33) was prepared by serial passages in Huh7.5.1 cells. Briefly, 1.5 × 10⁶ Huh7.5.1 cells seeded on a 10-cm dish were incubated for 1 d and were inoculated with HCV at a multiplicity of infection (MOI) of 1.0, and the culture medium was changed to fresh medium at 2 h postinfection. Culture supernatants were collected at 4 d postinfection, and infectious titers were determined. J6/JFH1 and Con1/JFH1 were prepared as previously described (34, 35).

Antibodies and Reagents.

The following antibodies were obtained from the sources indicated: anti-HCV NS5A monoclonal antibody (5A27) (36), anti-HCV core mouse monoclonal antibody (Fujirebio), anti-actin mouse monoclonal antibody (A228; Sigma), anti-HA rat monoclonal antibody (3F10; Roche), anti-GFP mouse monoclonal antibody (JL-8; Clontech), Alexa Fluor (AF) 488-conjugated anti-rabbit IgG antibodies (Life Technologies), and horseradish peroxidase-conjugated anti-FLAG mouse monoclonal antibody (M2; Sigma). YO-01027 was obtained from SYNthesis med chem, and LY-411575 was obtained from Sigma. RO-4929097, avagacestat, LY-450139, DAPT, MK-0752, telaprevir (VX-950), and daclatasvir (BMS-790052) were purchased from Selleck Chemicals. Compound E and compound 34 were purchased from Santa Cruz Biotechnology.

Plasmids.

A lentiviral vector, FUGW (#14883) was obtained from Addgene. FUGW was inserted into the puromycin N-acetyl-transferase gene and the GFP gene under the internal ribosomal entry site (IRES) sequence, and the resulting plasmids were designated “FUIPW” and “FUIGW,” respectively (19). The plasmid pCMV-VSV-G (#8454) was obtained from Addgene. The plasmids pMDLg/pRRE and pRSV-Rev were obtained from David Huang, The Walter & Eliza Hall Institute, Parkville, Australia. The cDNAs of GT1 (TN strain), GT1 (H77C strain), GT2 (J6 strain), GT2 (J8 strain), GT2 (JFH1 strain), GT3 (S52 strain), GT4 (ED43 strain), GT5 (SA13 strain), GT6 (HK6a strain), and GT7 (QC69 strain) were synthesized by Integrated DNA Technologies (IDT). These cDNAs were amplified by PCR and cloned into FUIGW together with FLAG tag sequences. The SPP gene encoded in P. falciparum was synthesized by IDT, amplified by PCR, and cloned into FUIPW together with an ER retrieval sequence fused with HA tag sequences (19). SPP mutants were generated by the overlap extension method and were cloned into FUIPW. All cloning procedures were performed using the In-Fusion HD cloning kit (Clontech), and the sequences were confirmed by using an ABI Prism 3130 genetic analyzer.

Generation of Lentiviruses.

HEK293T cells (2 × 10⁶) were seeded on a 10-cm dish and incubated at 37 °C for 1 d. The lentiviral transfer vector FUIPW (1.5 µg), 2 µg of pMDLg/pRRE, 2 µg of pRSV-Rev, and 1 µg of pCMV-VSV-G were mixed with 500 µL of Opti-MEM and 40 µL of polyethylenimine and were incubated for 15 min. The DNA complex was inoculated into HEK293T cells, and the culture medium was changed at 4 h posttransfection. The culture supernatants were collected at 3 d posttransfection. For the infection of lentivirus, 2 × 10⁵ cells (2 mL) were seeded on six-well plates and were incubated for 1 d. The virus-containing culture supernatants (2 mL) and 8 µL of Polybrene (4 mg/mL; Sigma) were inoculated into cells and centrifuged at 1,220 × g for 45 min at 32 °C. Stable cell lines were selected by puromycin at 2 d postinfection.

qRT-PCR.

qRT-PCR for HCV RNA was performed using a TaqMan RNA-to-Ct 1-Step Kit and a ViiA7 real-time PCR system (Life Technologies). The following primers were used: HCV, 5′-GAGTGTCGTGCAGCCTCCA-3′ and 5′-CACTCGCAAGCACCCTATCA-3′; GAPDH, 5′-TGTAGTTGAGGTCAATGAAGGG-3′ and 5′-ACATCGCTCAGACACCATG-3′. The following probes were used: HCV, 5′-6-FAM/CTGCGGAAC/ZEN/CGGTGAGTACAC/-3′IABkFQ; GAPDH, 5′-6-FAM/AAGGTCGGA/ZEN/GTCAACGGATTTGGTC/-3′IABkFQ. HCV RNA was determined by the ΔΔCt method using GAPDH as an internal control.

HCV Titration.

Huh7.5.1 cells were seeded on 24-well plates (3 × 10⁴ cells per well) and were incubated for 1 d. The culture supernatants serially diluted by medium were inoculated and incubated for 2 h. Then the culture supernatants were removed, 1% methylcellulose containing DMEM supplemented with 10% FBS, 100 U/mL penicillin, and 100 µg/mL streptomycin was added, and the culture was further incubated for 2 d. The supernatants were removed, washed once with PBS, and then incubated with 4% paraformaldehyde (PFA) in PBS for 2 h. The cells were washed with PBS three times and were permeabilized by incubation with 0.2% Triton X-100 containing PBS for 15 min. After washing with PBS three times, the cells were incubated for 1 h with anti-NS5A antibody (1/2,000) diluted by 2% FBS/PBS at room temperature. After washing with PBS three times, the cells were incubated for 1 h with AF488-conjugated anti-rabbit antibody (1/2,000) diluted by 2% FBS/PBS at room temperature. After the cells were washed three final times with PBS, the viral proteins expressing foci were counted under an immunofluorescent microscope (Olympus).

Immunoblotting.

Cell lysates were prepared by adding lysis buffer consisting of 20 mM Tris⋅HCl (pH 7.4), 135 mM NaCl, 1% Triton X-100, 1% glycerol, and protease inhibitor mixture tablets (Roche), incubation for 30 min at 4 °C, and centrifugation at 14,000 × g for 15 min at 4 °C. The supernatants were incubated at 95 °C for 5 min. Proteins were resolved by SDS/PAGE (Novex gels; Invitrogen) and were transferred onto nitrocellulose membranes. The membranes were then blocked with Tris-buffered saline containing 20 mM Tris⋅HCl (pH 7.4), 135 mM NaCl, 0.05% Tween 20, and 5% skim milk, incubated with the primary antibody at room temperature for 1 h, and then incubated with the HRP-conjugated secondary antibody at room temperature for 1 h. The immune complexes were visualized with SuperSignal West Femto substrate (Pierce) and were detected by an LAS-3000 image analyzer system (Fujifilm).

Determination of IC₉₉ in the Combination of SPP Inhibitors and DAAs.

Huh7 cells were seeded on a 24-well plate (3 × 10⁴ cells per well) and incubated for 24 h. The cells were then infected with HCV (JFH1 strain) at an MOI of 5 and were incubated for 2 h. The supernatants were replaced with fresh medium containing compounds. At 4 d postinfection, the infectious titers in the culture supernatants were determined by using a focus-forming assay. IC₉₉ curves were generated using GraphPad Prism 7 software (GraphPad).

Dosing.

Dosing of YO-01027 was performed as previously described (37). Briefly, YO-01027 was formulated in 1-mM solutions containing 0.5% (hydroxypropyl)methyl cellulose (Sigma) and 0.1% Tween-80 in H₂O. CoreTg mice were dosed i.p. once a day for 14 d with vehicle or with 5.0 µmol/kg YO-01027. BALB/c mice were dosed i.p. once a day with vehicle or with 5.0 µmol/kg or 10 µmol/kg YO-01027 at 24 h postinfection.

Live Imaging Analysis.

The 8-wk-old BALB/c mice were i.p. infected with 1 × 10² parasites (RH strains) expressing luciferase in 100 mL PBS, and bioluminescence was assessed on the indicated days after infection. For the detection of bioluminescence emission, mice were i.p. injected with 3 mg of d-luciferin (Promega) in 200 mL PBS, maintained for 5 min, and then anesthetized with isoflurane (Dainippon Sumitomo Pharma). Abdominal photon emission was assessed during a 3-s exposure by an in vivo imaging system (IVIS 100; Xenogen) and was analyzed as described previously (38).

Hematoxylin and Oil Red O Staining.

Formalin-fixed livers were impregnated with 30% sucrose and frozen with Tissue-Tek OCT compound (Sakura Finetek). Ten-micrometer-thick frozen sections were stained with hematoxylin (H&E). To visualize lipids, frozen sections were stained with Oil Red O. The frozen sections were washed with running tap water for 5 min, rinsed with 60% isopropanol for 1 min, and then stained with freshly prepared Oil Red O solution for 15 min. The sections were then rinsed with 60% isopropanol, followed by light staining of the nuclei with HE for 15 s. The sections were observed under a microscope (Olympus).

Homology Modeling and Docking Simulation of SPP.

Structural information of a presenilin/SPP homolog (PSH) from Methanocellus marisnigri JR1 was obtained from the Protein Data Bank (PDB)(PDB ID code: 4HYG.A; https://pdbj.org/mine/resources/4hyg) as a template to model SPP. Jr1 SPP sequences were aligned by HHPred (https://toolkit.tuebingen.mpg.de/#/tools/hhpred), and a structural model of SPP was built using the program Spanner (https://sysimm.ifrec.osaka-u.ac.jp/spanner/). Poorly aligned and flexible regions of modeled SPP were selected manually and resampled by Spanner. The 3D information of YO-01027 was obtained from ZINC (https://zinc.docking.org/substance/22056928), and docking simulation of YO-01027 with SPP was conducted using the AutoDock Vina docking program (39). Structural alignment of modeled SPP and presenilin-1 (PDB ID code: 5A63.B; https://pdbj.org/mine/summary/5a63) was conducted by the SALIGN server (https://modbase.compbio.ucsf.edu/salign/) (40). The electrostatic potential was calculated using an Adaptive Poisson–Boltzmann Solver (APBS) tool (41–43). To render the structure images (and make mutants of SPP in silico), the open-source PyMOL Molecular Graphics System 1.8.4.0 was used.

Cell Viability Assay.

Huh7 cells (2 × 10⁴ cells in 50 µL of culture medium) were added to black 96-well plates (Corning) and then serially diluted compounds (50 µL) were added to each well. Cell viability was measured by using a CellTiter-Glo Luminescent Cell Viability Assay (Promega) according to the manufacturer’s protocol.

Parasite Culture and Drug Treatment.

P. falciparum strain 3D7 was cultured in RPMI 1640 medium (Nacalai Tesque) supplemented with 25 µg/mL l-glutamine, Hepes, NaHCO₃, 50 µg/mL hypoxanthin, gentamicin, 10% human serum, and RBCs at 3% hematocrit in an atmosphere of 5% CO₂, 5% O₂, and 90% N₂ at 37 °C as previously described (44). Parasites were synchronized to ring stage by three 5% sorbitol treatments (45) and then were treated with the indicated drugs for 48 h. Parasitemia and morphological changes were observed under a microscope. The ME49 and RH strains of T. gondii were maintained in Vero cells in RPMI 1640 supplemented with 2% FBS, 100 U/mL penicillin, and 0.1 mg/mL streptomycin.

Statistical Analysis.

Statistical analyses were performed using GraphPad Prism software (GraphPad). Significant differences were determined using Student’s t test and are indicated with asterisks (*P < 0.05) and double asterisks (**P < 0.01) in each figure. Significant differences in the in vivo survival data were determined using a log-rank test and are indicated with asterisks (*P < 0.05) and double asterisks (**P < 0.01) in each figure.

Results

The Dibenzoazepine-Type Structure in γ-Secretase Inhibitors Participates in the Inhibitory Activity of SPP.

Our prior study demonstrated that the cleavage of HCV core protein by SPP is essential for maturation of the core protein and that an immature core protein generated by the inhibition of SPP activity was degraded by proteasomes after ubiquitination by TRC8 (Fig. 1A) (19). In the present study, to assess the inhibitory activity of γ-secretase inhibitors for SPP, we constructed a dicistronic lentiviral vector to express the FLAG-tagged core protein and GFP (Fig. 1B). Among the γ-secretase inhibitors we examined, YO-01027 exhibited more potent suppression of the maturation of HCV core protein than LY-411575 or RO-0492907, as we reported previously (19).

Fig. 1.

The dibenzoazepine-type structure of γ-secretase inhibitors participates in the inhibitory activity of SPP. (A) Schematic of the maturation of HCV core protein and its inhibition by SPP. HCV core protein is cleaved off from the precursor polyprotein by signal peptidase and then matures through the processing by SPP. The immature core protein that was not processed by SPP due to the inhibition of SPP was degraded by the proteasome pathway through ubiquitination by the E3 ligase TRC8. (B) Structure of a dicistronic lentiviral vector encoding the FLAG-tagged core protein under the ubiquitin promoter and GFP under the encephalomyocarditis virus (EMCV) IRES. (C) SPP inhibitory activity of the γ-secretase inhibitors. Huh7 cells expressing FLAG-tagged core protein were treated with the compounds at concentrations of 0.1, 0.3, 1.0, or 3.0 µM for 2 d, and the mature core protein was detected by immunoblotting. Data are representative of three independent experiments.

In contrast, other compounds such as compound E, compound 34, LY-450139, DAPT, avagacestat, and MK-0752 showed no effect on the processing of HCV core protein (Fig. 1C). We therefore classified these γ-secretase inhibitors into two groups based on whether they inhibited SPP activity (Fig. 2) and found that the γ-secretase inhibitors capable of inhibiting SPP activity have a conserved dibenzoazepine-type structure. YO-01027 and compound E have the same structure except for dibenzoazepine, and YO-01027 but not compound E inhibited SPP activity, suggesting that dibenzoazepine-type γ-secretase inhibitors can inhibit SPP.

Fig. 2.

Classification of γ-secretase inhibitors. (Upper) All the γ-secretase inhibitors exhibiting inhibition of SPP have a dibenzoazepine structure (green circle). (Lower) The other γ-secretase inhibitors exhibiting no inhibitory activity toward SPP lack the dibenzoazepine-type structure. Notably, the only difference between YO-01027 and compound E is the dibenzoazepine structure (purple circle).

YO-01027 Is a Potent Inhibitor of the Propagation of HCV and Protozoa.

To examine the toxicity of the γ-secretase inhibitors, we determined the viability of Huh7 cells treated with the inhibitors at a concentration of 3 µM. Although treatment with LY-411575, YO-01027, and LY-450139 exhibited no significant cell toxicity, avagacestat showed cell toxicity at 4 d posttreatment (Fig. S1A). Next, to examine the effect of YO-01027 and LY-411575 on the replication of HCV, we treated GT2 and GT1 replicon cells with YO-01027 and LY-411575 for 3 d. Treatment with YO-01027 and LY-411575 exhibited no effect on the replication of HCV RNA (Fig. S1B).

To examine the effects of the γ-secretase inhibitors on the production of infectious HCV particles, Huh7 cells infected with HCV were treated with various concentrations of the inhibitors. The IC₉₉ values were 230 nM for YO-01027 and 76.7 μΜ for LY-411575 (Fig. 3A), suggesting that YO-01027 exhibited a more potent inhibitory effect on the production of infectious particles in the culture supernatants (Fig. 3A). We did not see any effects by using LY-450139 and avagacestat (Fig. S1C). Next, we examined the effects of LY411575 and YO-01027 on the propagation of two chimeric HCVs, J6/JFH1 and Con1/JFH1, that possessed the structural proteins of GT2 and GT1, respectively. Huh7 cells infected with the chimeric viruses were incubated with LY-411575 or YO-01027. Propagation of both J6/JFH1 and Con1/JFH1 was significantly suppressed by the treatment with LY-411575 and YO-01027. In particular, treatment with YO-01027 impaired the propagation of J6/JFH1 to an undetectable level (Fig. S1D). To examine the synergistic effects of SPP inhibitors and DAAs on the propagation of HCV, we first determined the IC₉₉ values of telaprevir and daclatasvir on HCV propagation; these values were 114 nM and 5.28 nM, respectively (Fig. S1E). Next, Huh7 cells infected with HCV were treated with SPP inhibitors and DAAs, and the IC₉₉ values against each combination were determined (Fig. 3B). All IC₉₉ values were plotted on an isobologram based on the Loewe combination index (46), and the combination of SPP inhibitors and DAAs was found to exhibit synergistic effects on the propagation of HCV (Fig. 3C). Collectively, these results suggest that YO-01027 is a potent inhibitor of the production of infectious HCV particles.

Fig. 3.

YO-01027 is a potent inhibitor of the production of infectious particles of HCV. (A) Huh7 cells infected with HCV JFH1 strain at an MOI of 5 were treated with LY-411575 or YO-01027 for 4 d. Infectious titers in the culture supernatants were determined by a focus-forming assay. The P values were determined by a Student's t test. The IC99 values were calculated by using GraphPad Prism software. FFU, focus-forming units. (B) Huh7 cells infected with HCV JFH1 strains at an MOI of 5 were treated with the SPP inhibitors LY-411575 (0, 300, 1,000, or 3,000 nM) or YO-01027 (0, 30, 100, or 300 nM) and the DAAs telaprevir (TPV) or daclatasvir (DCV) for 4 d. Infectious titers in the culture supernatants were determined by a focus-forming assay. (C) Fractional IC₉₉ (FC) values were calculated as the IC₉₉ of an inhibitor combination divided by the IC₉₉ of an inhibitor alone. The effect of synergy was investigated by isobologram analysis. (D) SPPKOHuh7 cells were lentivirally transduced with hsSPP and pfSPP. SPPKOHuh7 cells stably expressing either hsSPP (Hs) or pfSPP (Pf) were selected with puromycin, and the expression of hsSPP and pfSPP was confirmed by Western blotting. (E) Huh7 and SPPKOHuh7 cells stably expressing hsSPP (Hs) and pfSPP (Pf) were infected with a lentivirus expressing FLAG-tagged HCV core protein. The mature core protein was detected by Western blotting. (F) SPPKOHuh7 cells stably expressing hsSPP (Hs) and pfSPP (Pf) were infected with HCV at an MOI of 5. Infectious titers in the culture supernatants were determined by a focus-forming assay at 4 d postinfection. The P values were determined by a Student's t test. (G) SPPKOHuh7 cells stably expressing pfSPP were infected with lentivirus vector expressing HCV core protein and were treated with YO-01027 at a concentration of 0.1, 0.3, 1.0, or 3.0 µM. The matured core protein was detected by immunoblotting. (H) Treatment with YO-01027 at a concentration of 1, 10, or 20 μM for 4 d suppressed the production of infectious particles of HCV in SPPKOHuh7 cells expressing pfSPP upon infection with HCV at an MOI of 5. The P values were determined by a Student's t test. (I) The growth rate of P. falciparum treated with YO-01027 or LY-411575 was measured. The P values were determined by a Student's t test. (J) MEFs were infected with the T. gondii strain ME-49, which possesses a luciferase gene, at an MOI of 1 and then were incubated with the inhibitors for 1 d. The luminescent signal was measured. The P values were determined by a Student's t test. (K) MEFs, SPP-KO MEFs (SPPKOMEFs), Huh7 cells, and SPPKOHuh7 cells were infected with T. gondii at an MOI of 1. The luminescent signal was measured at 1 d postinfection. The P values were determined by a Student's t test. ns, not significant. (L) BALB/c mice i.p. infected with T. gondii (RH strain: 1 × 10² parasites) were i.p. treated with 5 μmol/kg or 10 μmol/kg of YO-01027 (n = 6 mice for each treatment) or vehicle (n = 6 mice) at 24 h postinfection. Bioluminescent analysis was performed at the indicated time points to visualize parasite growth. (M) The survival of mice was monitored. The P values were determined by a log-rank text. Data are the mean ± SD of two independent experiments performed with cell lines (A, B, F, and H–K) and are representative of three (G) or two (D, E, L, and M) independent experiments. Significant differences are indicated by double asterisks (**P < 0.01) or a single asterisk (*P < 0.05).

A previous study demonstrated that the SPP gene is encoded in protozoa such as P. falciparum and is likely to be essential for their survival (21). To examine the role of P. falciparum-encoded SPP (pfSPP) on the maturation of HCV core protein, we determined the expression of HCV core protein in SPP-knockout Huh7 (SPPKOHuh7) cells by assessing the expression of pfSPP (Fig. 3D). The expression of core protein in SPPKOHuh7 was restored by the exogenous expression of either hsSPP or pfSPP (Fig. 3E), suggesting that pfSPP can compensate for the function of hsSPP in cleaving HCV core protein in SPPKOHuh7 cells.

Next, to examine the effect of pfSPP on the propagation of HCV, we determined the intracellular viral RNA and infectious titers in the supernatants of SPPKOHuh7 cells expressing either hsSPP or pfSPP upon infection with HCV. HCV propagation in SPPKOHuh7 cells was restored by the expression of either pfSPP or hsSPP (Fig. 3F and Fig. S1F). In addition, the maturation of core protein (Fig. 3G) and HCV propagation (Fig. 3H) were impaired in SPPKOHuh7 cells expressing pfSPP by treatment with YO-01027, suggesting that pfSPP was functionally similar to hsSPP.

For investigation of the effects of the γ-secretase inhibitors on the propagation of P. falciparum, synchronized P. falciparum was inoculated with hematocrit and γ-secretase inhibitors. Treatment with YO-01027 exhibited 10-fold higher inhibition of the propagation of P. falciparum compared with treatment with LY-411575 (Fig. 3I). In addition, propagation of the ME49 strain of T. gondii, which possesses a luciferase reporter gene, was also significantly more inhibited by treatment with YO-01027 than by treatment with LY-411575 (Fig. 3J). Treatment with LY-411575 or YO-01027 at concentrations up to 30 µM exhibited no significant cell toxicity in SPPKOHuh7 cells expressing pfSPP (Fig. S1G).

Next, to examine the effect of host SPP on the propagation of the ME49 strain of T. gondii, we inoculated parental and SPP-knockout MEFs and Huh7 cells with the ME49 strain. The propagation of T. gondii was similar among the cell lines (Fig. 3K), suggesting that a parasite but not the host SPP participates in the propagation of T. gondii. Finally, we examined the effect of YO-01027 on T. gondii in vivo. BALB/c mice i.p. infected with the RH strain possessing a luciferase reporter gene were treated with 5 μmol/kg or 10 μmol/kg of YO-01027 i.p. at 24 h postinfection, and the growth of parasites was determined using an IVIS imaging system. The growth of parasites was significantly suppressed by treatment with YO-01027 (Fig. 3L), and the survival times of infected mice were slightly extended by the YO-01027 treatment (Fig. 3M). Collectively, these results suggest that YO-01027 is capable of inhibiting the propagation of HCV and protozoa.

YO-01027 Ameliorates Liver Steatosis in CoreTg Mice.

To determine the effect of YO-01027 on the liver pathogenesis in CoreTg mice, we administered YO-01027 to the mice for 2 wk. Although the mRNA expression of the HCV core protein in the livers of CoreTg mice and the body weights of the mice were not changed by the treatment with YO-01027 (Fig. 4A), the expression of the core protein was significantly reduced by the treatment (Fig. 4B). Oil Red O staining of the liver sections revealed that the lipid accumulation in the livers of the CoreTg mice was also ameliorated by 2-wk treatment with YO-01027 when the mice were 24 wk old (Fig. 4C).

Fig. 4.

Treatment of YO-01027 ameliorates liver steatosis in CoreTg mice. (A) CoreTg male mice (24 wk old) were i.p. administered YO-01027 or vehicle one time/d for 14 d. (Upper) Total RNA was obtained from the livers of CoreTg mice (n = 5), and core mRNA was determined by qPCR. Data are the relative expression levels after standardization by actin expression and are shown as the means ± SE. (Lower) Body weights of mice before and after YO-01027 administration (n = 5). (B) The livers of CoreTg mice treated with YO-01027 or vehicle for 2 wk or left untreated (WT) were subjected to immunoblotting. (C) Liver sections of mice after treatment with YO-01027 were stained with Oil Red O (Upper) and H&E (Lower). (Scale bars, 100 μm.) (D) The expression of mRNA of SREBP-1c (Left), SREBP-1a (Center), and SREBP-2 (Right) in the livers was determined by qPCR (n = 5). Data are the mean ± SD from two independent measurements in five mice per compound treatment. A significant reduction (*P < 0.05) of SREBP-1c expression was observed in CoreTg mice in response to YO-01027 treatment. Images in B and C are representative of at least three independent mice per compound treatment. The P values were determined by a Student's t test.

The synthesis of liver fatty acids is tightly regulated by SREBP-1c, a family of steroid regulatory element-binding proteins. Here, the mRNA expression of SREBP-1c, but not that of SREBP-1a or SREBP-2, was reduced in the livers of CoreTg mice by the YO-01027 treatment (Fig. 4D). These results suggest that treatment with YO-01027 ameliorates liver steatosis in CoreTg mice by suppressing the expression of HCV core protein.

SPP Inhibitors Are Broad-Spectrum Antivirals Against HCV of All Genotypes with a Low Risk of the Emergence of Drug-Resistant Breakthrough Viruses.

Next, to assess the effects of YO-01027 on the maturation of HCV core proteins derived from various genotypes of HCV, we constructed expression vectors for HCV core proteins derived from GT1 to GT7. The treatment of Huh7 cells expressing these core proteins with YO-01027 efficiently suppressed the maturation of core proteins derived from all HCV genotypes (Fig. 5A). For a further examination of the emergence of drug-resistant viruses by treatment with the inhibitors, culture supernatants from Huh7.5.1 cells infected with HCV and treated with 1 μΜ of LY-411575, 0.1 μΜ of YO-01027, or 0.1 µM of telaprevir for 4 d were inoculated into naive Huh7.5.1 cells treated with same compounds (Fig. 5B). After five rounds of passages, HCV in the culture supernatants exhibited resistance to the treatment with telaprevir (Fig. 5C), but HCV was sensitive to the treatment with LY-411575 and YO-01027 even after 15 passages (Fig. 5D), suggesting that the SPP inhibitors are broad-spectrum antivirals against all genotypes of HCV with a low risk of the emergence of drug-resistant breakthrough viruses.

Fig. 5.

SPP inhibitors are broad-spectrum antivirals against HCV of all genotypes with a low risk of emergence of drug-resistant breakthrough viruses. (A) Inhibitory effects of YO-01027 against the processing of core proteins of various genotypes of HCV. FLAG-tagged core proteins of GT1 (TN strain), GT1 (H77C strain), GT2 (J6 strain), GT2 (J8 strain), GT2 (JFH1 strain), GT3 (S52 strain), GT4 (ED43 strain), GT5 (SA13 strain), GT6 (HK6a strain), and GT7 (QC69 strain) were expressed in Huh7 cells and were treated with YO-01027 at a concentration of 0.1, 0.3, 1.0, or 3.0 µM. HCV core protein was detected by immunoblotting. Images are representative of at least three independent experiments. (B) Culture supernatant of Huh7.5.1 cells infected with HCV in the presence of LY-411575 (1.0 µM), YO-01027 (0.1 µM), or telaprevir (TPV, 0.1 µM) was passaged in naive cells in the presence of these reagents. After five or 15 passages, the supernatants were collected, and the inhibitory effects of the reagents were determined. (C) After five passages in the presence or absence of telaprevir, culture supernatants were inoculated into Huh7.5.1 cells in the presence of various concentrations of telaprevir, and the infectious titers were determined 4 d postinfection. (D) After 15 passages in the presence or absence of LY-411575 or YO-01027, the culture supernatants were inoculated into Huh7.5.1 cells in the presence of various concentrations of the inhibitors, and infectious titers were determined 4 d postinfection. Data in C and D are the mean ± SD of two independent experiments. **P < 0.01. P values were determined by a Student's t test.

Val223 of SPP Interacts with the Dibenzoazepine-Type Structure of YO-01027.

We next investigated the molecular interactions between SPP and γ-secretase inhibitors. Because the crystal structure of a presenilin/SPP homolog (PSH) from M. marisnigri JR1 has been resolved (47), we constructed the 3D structure of human SPP in silico based on the structure of PSH (Fig. 6A). The enzymatic active site of SPP had been identified as Asp219 and Asp265 (20), and the structures of PSH and presenilin-1 show a cavity-like structure in an enzymatic active site and a large hole surrounded by TM2, TM3, TM5, and TM7 (47, 48). The modeled structure of SPP showed the same cavity-like structure and a large hole (Fig. 6B).

Fig. 6.

In silico profiling of the structure of SPP. (A) Homology modeling of hsSPP was deduced based on a PSH derived from M. marisnigri JR1 (PDB ID code: 4HYG). The enzymatic active site of SPP was identified in Asp219 and Asp265, which are located on TM6 and TM7, respectively. (B) The structure of SPP is shown with an electrostatic surface representation. As in previous studies of PSH and presenilin-1 (47, 48), the modeled structure of SPP exhibited a cavity-like structure and a large hole.

To examine the interaction of SPP with γ-secretase inhibitors, we performed a docking simulation by using YO-01027. This in silico simulation indicated that the dibenzoazepine structure of YO-01027 interacts with Val223, Phe192, or Ala259 of SPP (Fig. 7A). To determine the amino acid residue responsible for the interaction with YO-01027, we established SPPKOHuh7 cell lines stably expressing mutant SPP with Val223, Phe192, and Ala259 replaced by Ile, Trp, and Leu, respectively.

Fig. 7.

Val223 of SPP interacts with the dibenzoazepine-type structure of YO-01027. (A) The docking simulation of YO-01027 revealed three putative drug-binding sites in SPP (sites 1–3). (B) SPPKOHuh7 cells were lentivirally transduced with wild-type or mutant SPP in which Val223 was replaced with Ile (V223I), Phe192 was replaced with Try (F192W), or Ala259 was replaced with Leu (A259L), and SPP expression was confirmed by immunoblotting after selection with puromycin. (C) Huh7 cells and SPPKOHuh7 cells stably expressing wild-type or mutant SPP were infected with a lentivirus expressing the FLAG-tagged HCV core protein. The mature core protein was detected by immunoblotting. (D) SPPKOHuh7 cells expressing wild-type or mutant SPP were transduced with a lentiviral vector encoding the FLAG-tagged HCV core protein and were treated with YO-01027 at a concentration of 0.1, 0.3, 1.0, or 3.0 µM for 2 d. The mature core protein was detected by immunoblotting. (E) Close-up views of the interaction of SPP and YO-01027. The dibenzoazepine structure of YO-01027 interacted with SPP Val223. Asp219 of the enzymatic active site of SPP was also close to YO-01027. The data are representative of two (B and C) or three (D) independent experiments.

The results showed that the expression of SPP was comparable in the wild type and mutants (Fig. 7B) and that the SPP mutants retained the ability to process HCV core proteins (Fig. 7C). Although the mutant SPPs in which Phe192 was replaced by Trp and Ala259 was replaced by Leu were still sensitive to the YO-01027 treatment, the substitution of Ile for Val223 resulted in resistance to the YO-01027 treatment (Fig. 7D). In addition, the SPP mutant in which Val223 was replaced by Ile also showed resistance to other dibenzoazepine-type γ-secretase inhibitors, i.e., LY-411575 and RO-0492907 (Fig. S2). These results suggest that YO-01027 interacts directly with enzymatic active sites (Asp219 and Asp265) of SPP and that the dibenzoazepine structure of YO-01027 interacts with Val223 of SPP (Fig. 7E).

Phe258 in SPP Determines the Accessibility of the γ-Secretase Inhibitors.

To further elucidate the structure–activity relationship between γ-secretase inhibitors and SPP, we compared the docking simulation of YO-01027 to SPP with that to presenilin-1 (Fig. 8A). We focused on Phe258 of SPP because this amino acid is close to the dibenzoazepine structure in YO-01027. Although the 3D structure of presenilin-1 around SPP Phe258 has not yet been determined (Fig. 8B), the amino acid residue corresponding to SPP Phe258 in presenilin-1 is Gly378 (Fig. 8C). We therefore hypothesized that the benzyl group of SPP Phe258 prevents compound E from interacting with SPP (Fig. 8C), because the only difference between YO-01027 and compound E is the dibenzoazepine structure.

Fig. 8.

Phe258 in SPP determines the accessibility of the γ-secretase inhibitors. (A) Close-up views of the interaction of SPP Phe258 with YO-01027. (B) The deduced structure of SPP (blue) was compared with that of human presenilin-1 (PDB ID code: 5A63; yellow). The structure of presenilin-1 corresponding to SPP Phe258 was not resolved. (C) Amino acid comparison between SPP and presenilin-1. (Upper) The amino acid residue of presenilin-1 corresponding to SPP Phe258 is Gly. (Lower) A hypothetical model of the interaction of SPP with compound E. (Left) In wild-type SPP, Phe258 is an obstacle to the incorporation of compound E. (Right) A small side chain such as Gly258 accepts the incorporation of compound E into the P258G SPP mutant. (D) Docking simulation of compound E with the P258G SPP mutant revealed an interaction similar to that seen between YO-01027 and wild-type SPP (Fig. 6E). (E) SPPKOHuh7 cells expressing the P258G SPP mutant were transduced with a lentiviral vector encoding the FLAG-tagged HCV core protein and were treated with compound E (Upper) or YO-01027 (Lower) at a concentration of 0.1, 0.3, 1.0, or 3.0 µM for 2 d. The mature core protein was detected by immunoblotting. The data are representative of three independent experiments.

To test our hypothesis, we constructed the in silico structure of a mutant SPP in which Phe258 was replaced with Gly, and we performed a docking simulation of both the wild-type SPP and mutant SPP with compound E. Although the wild-type SPP exhibited no interaction with compound E in our docking simulation, the Phe258Gly mutant of SPP interacted with compound E much as the wild-type SPP interacted with YO-01027 (Fig. 8D).

To further confirm our in silico data in vitro, we established SPPKOHuh7 cells expressing the Phe258Gly mutant of SPP. The expression of SPP was comparable in the wild type and the Phe258Gly mutant (Fig. S3A), and the SPP mutant still retained the ability to process HCV core proteins (Fig. S3B). SPPKOHuh7 cells expressing the Phe258Gly mutant of SPP were treated with either compound E or YO-0127 upon the expression of HCV core protein. Treatment not only with YO-01027 but also with compound E resulted in significant inhibition of the core protein expression in SPPKOHuh7 cells expressing the SPP Phe258Gly mutant (Fig. 8E), whereas other compounds such as semagacestat, avagacestat, compound 34, DAPT, and MK-0752 showed no effect on the processing of HCV core protein (Fig. S3C). Collectively, these results suggest that SPP Phe258 determines the accessibility of the γ-secretase inhibitors to an active site of SPP.

Discussion

Our previous research revealed that SPP plays crucial roles in the propagation of HCV and the development of core-induced pathogenesis and that SPP is therefore a promising target for the development of novel therapeutics against chronic hepatitis C (19). In addition, SPP encoded in protozoa participates in their propagation (21), suggesting that SPP inhibitors would be applicable as not only antivirals but also as antiprotozoals. However, the molecular mechanisms underlying the suppression of SPP by the γ-secretase inhibitors have been largely unknown. Therefore in the present study we attempted to determine the interaction between the γ-secretase inhibitors and SPP.

Our findings showed that some γ-secretase inhibitors are capable of inhibiting SPP activity. Although YO-01027 shares the same structure as compound E, except for the dibenzoazepine structure, YO-01027 inhibited SPP in our experiments, but compound E did not, suggesting that the dibenzoazepine-type structure plays a crucial role in the inhibition of SPP. Our in silico docking simulation of hsSPP with YO-01027 suggested that three binding sites and the mutation of Val223 in SPP abolished the inhibitory activity of YO-01027, indicating that SPP Val223 interacts with the dibenzoazepine-type structure in YO-01027.

Our results also showed that Phe258 is a spatial obstacle to the incorporation of compound E into SPP. Because the inhibitory activity of YO-01027 was still higher than that of compound E in the Phe258Gly mutant of SPP, we speculate that the interaction between SPP and the dibenzoazepine-type structure of YO-01027 might be responsible for the potent inhibitory activity of YO-01027. Our comparison of the amino acid sequences of members of the SPP family revealed that the amino acid residue Val223 in SPP is conserved not only in presenilin-1 but also in SPP-like 2A (SPPL2A), SPPL2B, and SPPL3. Therefore, dibenzoazepine-type γ-secretase inhibitors might be able to inhibit other SPP family enzymes. In contrast, Phe258 in SPP is conserved in SPPL3 but not in presenilin-1, SPPL2A, or SPPL2B. Thus, enhancing the interaction of a compound containing SPP Val223 and Phe258 might improve SPP-specific inhibition.

We also observed that compound E and compound 34 have a benzodiazepine structure, but compound 34 could not inhibit the activity of the Phe258Gly SPP mutant (Fig. S3C), suggesting that structures other than the dibenzoazepine structure participate in determining specificity for SPP inhibition. These data could provide clues for the development of novel inhibitors that are specific for SPP.

Although treatment with DAAs targeting viral proteins including NS3, NS5A, and NS5B have achieved a 90% SVR rate (5, 6), the emergence of viruses resistant to DAAs is an important issue that must be considered (7). The inhibition of host proteins that are crucial for efficient viral propagation is a suitable method of suppressing the emergence of drug-resistant breakthrough viruses (49, 50). HCV has various genotypes and a highly diverse amino acid sequence due to the low fidelity of RNA-dependent RNA polymerase, and thus it is difficult for antivirals targeting viral proteins to inhibit all genotypes of HCV. In this study, we demonstrated that the inhibition of SPP suppressed both the maturation of HCV core proteins derived from all the HCV genotypes examined and the propagation of HCV without the emergence of drug-resistant viruses. Moreover, combined treatment with SPP inhibitors and DAAs showed synergistic effects, suggesting that combination therapy with SPP inhibitors and DAAs might be an effective treatment for patients with chronic hepatitis C, with a low frequency of emergence of drug-resistant viruses.

HCV belongs to the family Flaviviridae, which contains four genera: Flavivirus, Pestivirus, Hepacivirus, and Pegivirus (51, 52). The C-terminal signal sequences of the core proteins of Hepacivirus and Pestivirus are processed by SPP (53), and those of Flavivirus are cleaved by the viral protease NS3 (54). Therefore, cleavage of the signal sequence of the core protein is essential for the maturation of the core protein in viruses of the family Flaviviridae.

We also explored the possibility of the application of SPP inhibitors in other infectious diseases. Some protozoa encode SPP in their genomes, and SPP has been suggested to play a crucial role in the life cycle of these protozoa (21). Here we observed that SPP inhibitors suppressed the propagation of P. falciparum and T. gondii by inhibiting the SPP encoded in their genomes, not by the inhibiting SPP in the host cell genomes. LY-411575, NITD679, and NITD731 were previously shown to inhibit the activity of pfSPP (21), and all three of these compounds contain the dibenzoazepine structure, which is consistent with our data (Fig. 2).

Collectively, these results suggest that SPP is an ideal drug target for the development of novel therapeutics for chronic hepatitis C and protozoiasis. However, the development of SPP-specific inhibitors is still required, because all the compounds used in this study were developed as γ-secretase inhibitors. The phase III clinical trials of semagacestat (LY-450139) resulted in no improvement of Alzheimer’s disease and unfortunately enhanced tumorigenesis of skin and other cancers (55). In this study, we showed that LY-450139 had no inhibitory activity on SPP (Fig. S1C). Therefore, specific inhibitors for SPP which exert no inhibitory effects on γ-secretase are needed for the treatment of chronic hepatitis C. Toward this goal, further investigations will be necessary to elucidate the 3D structure of SPP and the structural interaction between SPP and YO-01027.