Profiling the physiological pitfalls of anti‐hepatitis C direct‐acting agents in budding yeast

Abstract

Sofosbuvir and Daclatasvir are among the direct‐acting antiviral (DAA) medications prescribed for the treatment of chronic hepatitis C (CHC) virus infection as combination therapy with other antiviral medications. DAA‐based therapy achieves high cure rates, reaching up to 97% depending on the genotype of the causative hepatitis C virus (HCV). While DAAs have been approved as an efficient and well‐tolerated therapy for CHC, emerging concerns about adverse cardiac side effects, higher risk of recurrence and occurrence of hepatocellular carcinoma (HCC) and doubts of genotoxicity have been reported. In our study, we investigated in detail physiological off‐targets of DAAs and dissected the effects of these drugs on cellular organelles using budding yeast, a unicellular eukaryotic organism. DAAs were found to disturb the architecture of the endoplasmic reticulum (ER) and the mitochondria, while showing no apparent genotoxicity or DNA damaging effect. Our study provides evidence that DAAs are not associated with genotoxicity and highlights the necessity for adjunctive antioxidant therapy to mitigate the adverse effects of DAAs on ER and mitochondria.

Introduction

Hepatitis C virus (HCV) is a liver-infecting RNA virus that belongs to the family Flaviviridae. Over 100 million people are chronically infected around the world, with the major complications of HCV infection being liver cirrhosis and emergence of hepatocellular carcinoma (Mohd Hanafiah et al., 2013; Blach et al., 2017). Egypt has the sharpest worldwide prevalence rates of HCV infection, with at least 15% of the population infected (Sievert et al., 2011). Current guidelines recommend HCV treatment with direct-acting antivirals (DAAs)-based regimen for all HCV-infected patients, irrespective of the degree of liver fibrosis (World Health Organization, 2018). Sofosbuvir (SOF) is a novel nucleotide analogue prodrug that selectively inhibits the HCV non-structural protein 5B (NS5B), an RNA-dependent RNA polymerase crucial for the replication of HCV (Fung et al., 2014). SOF is a component of the first oral, interferon (IFN)-free treatment authorized for a treatment of CHC infection (Asselah, 2014; Falade-Nwulia et al., 2017). Daclatasvir (DCV) is one of the new DAAs that interfere with other stages of the HCV life cycle (Belema and Meanwell, 2014; Lim and Gallay, 2014). It is the first selective inhibitor of HCV non-structural protein 5A (NS5A), and interference with this replication complex limits viral multiplication (Fung et al., 2014; Degasperi et al., 2015). The combination of both drugs with or without ribavirin provoked sustained virological response (SVR) in patients in different stages of advanced liver diseases (Sulkowski et al., 2014; Wyles et al., 2015; Moshyk et al., 2016; Boglione et al., 2017; Pol et al., 2017).

Although DAAs initially showed significant promise when introduced to the market, subsequent studies noted a few limitations and deleterious side effects of these treatments. Therefore, a detailed investigations to explore interventions of DAAs with molecular pathways, cellular physiology and subcellular organelles are required (Maciocia et al., 2016; Abouelkheir Abdalla et al., 2017; Kim et al., 2017; Andrade et al., 2018; Hirose et al., 2019; Kogiso et al., 2019). Researchers were particularly concerned over the possible increase in incidence and relapse of HCC recorded in HCV patients after DAA therapy (Giovannini et al., 2020; Lithy et al., 2020; Muzica et al., 2020; Pop et al., 2020).

The budding yeast Saccharomyces cerevisiae is routinely used as a model organism to resolve biological pathways and evolutionary processes conserved between yeast and higher eukaryotes, and more recently as a platform to investigate newly synthesized therapeutics (Parsons et al., 2006). The simple culture conditions and easy genetic manipulation of the yeast genome facilitate its extensive usage in various lines of biological research. Additionally, S. cerevisiae was the first unicellular eukaryote with a complete sequenced genome. The yeast genome has been completely sequenced since 1996 and comprises 6.275 genes organized on 16 chromosomes; the function of more than 92% of yeast genes has been identified (Botstein et al., 1997). Comparative genomic analysis revealed that nearly 40% of yeast genes share conserved primary structure with at least one structural or functional homolog of human protein-coding genes (Parsons et al., 2003). Moreover, 30% of human genes with identified function in genetic and pathological disorders have yeast orthologs (Foury, 1997). Due to this remarkable homology in the genome and the extreme conservation of basic cellular pathways, studies in yeast have facilitated the analysis of major cellular processes, such as mRNA translation and decay (Sonenberg et al., 2000; Coller and Parker, 2004), DNA repair mechanisms (Tsukuda et al., 2005) and the cell cycle regulation (Hartwell and Kastan, 1994). In this study, we used budding yeast for a detailed inspection of the adverse effects of DAAs on cell physiology, genome stability, and on subcellular organelles’ architecture and integrity. Our chemical-genetic and organelle biology approach of our study revealed apparent effects on the dynamics of ER and mitochondria, while excluding any DNA damaging or genotoxic effects of DAAs.

Results Susceptibility of budding yeast to SOF and DCV

We initially determined the minimum inhibitory concentration (MIC) of SOF (Fig. 1A) and DCV (Fig. 1B) in budding yeast cells. SOF and DCV were dissolved in DMSO, and twofold serial dilutions were prepared and added to the liquid culture of budding yeast adjusted to the same optical density. The effect of both drugs in different concentrations on growth and survival was monitored by reading the optical density after 16 h incubation. We found that the concentration of 64 µM DCV was sufficient to inhibit the yeast growth compared with mock-treated control cells (Fig. 1C), whereas concentrations higher than 100 µM of SOF were required to induce growth inhibition (Fig. 1D). These results indicate that DCV is less tolerated by budding yeast than SOF. In the same context, isogenic sets of polyploid yeast exhibited similar growth pattern to haploid strain when plated on plates containing inhibitory concentrations of each drug that were deduced from the growth curve of haploid yeast (Fig. 1E).

Fig. 1. Minimal inhibitory concentration (MIC) determination of Sofosbuvir and Daclatasvir. Chemical structures of (A) Sofosbuvir (SOF) and (B) Daclatasvir (DCV). (C) Dose–response curve of twofold serial dilutions of Daclatasvir (mean ± SD; n = 3 independent biological replicates). (D) Dose-response curve of twofold serial dilutions of Sofosbuvir (mean ± SD; n = 3 independent biological replicates). (E) Representative drop dilution of exponentially growing isogenic series of yeast strain of different ploidy on YPD plates supplemented with 128 µM Sofosbuvir (SOF) or 64 µM Daclatasvir (DCV). Plates were imaged after 48 h of incubation at 30 °C.

Exposure of budding yeast to DAAs perturbs cell volume

Perturbation of budding volume and morphology of budding yeast reflects several cellular stress signals and serves as an discernible phenotype of physiological aberrations, such as unfolded protein response (UPR), stress signals and cell cycle deregulations including cell cycle arrest and ageing (Scrimale et al., 2009; Moreno et al., 2019). Pathways that disturb cell volume are tightly linked to cell cycle progression, efficient protein folding machinery and DNA quality control pathways. Previously, fundamental regulators and genetic networks that control cell size in yeast and their homologs in mammalian cells were identified using systematic screens of yeast mutants that shift the size profile to smaller, or larger volume distributions (Jorgensen et al., 2002; Zhang et al., 2002; Ohya et al., 2005; Dungrawala et al., 2012; Soifer and Barkai, 2014; Yahya et al., 2014). To explore the effect of SOF or DCV on budding yeast volume, we grew budding yeast in presence of sublethal doses of each drug and compared them with identical cultures grown with drugs known to induce ER stress (Dithiothreitol, DTT) (Hamdan et al., 2017), UPR (Tunicamycin; TN) (Zhao et al., 2019), and genotoxic stress (Hydroxyurea, HU) (Soriano-Carot et al., 2012). As a control, we grew budding yeast with a vehicle of the tested drugs. Additionally, a yeast mutant (Δcln3) (Yahya et al., 2014) with a large cell volume due to a deletion of the G1 cyclin and another mutant (Δsfp1) with a small budding volume due to a deletion of SFP1, the master transcription factor of ribosomal biogenesis, were analysed (Fig. 2A) (Sudbery, 2002; Cipollina et al., 2008; Albert et al., 2019). Budding yeast treated with DAAs showed significantly enlarged budding volume compared with control cells (CT (control): 48.87 ± 13.28; SOF:71.3 ± 28.37 (P < 0.0001); DCV: 78.01 ± 34.17 (P < 0.0001)), although not as large as in cells treated with DTT (116.0 ± 34.83 (P < 0.0001)), TN (112.2 ± 34.80 (P < 0.0001)), HU (90.30 ± 31.13 (P < 0.0001)) or Δcln3 (100.7 ± 44.91 (P < 0.0001)), while the control Δsfp1 showed the expected small volume (32.69 ± 12.36 (P < 0.0001)) (Fig. 2B). The prominent effects of DAAs on cell volume denote possible effects on cellular organelles and cell physiology.

Fig. 2. Sofosbuvir and Daclatasvir enlarge budding volume.A. Representative bright-field images of haploid yeast cell grown for 4–6 h in YPD media (control; CT), or media supplemented with 64 µM Sofosbuvir (SOF), 32 µM Daclatasvir (DCV), or other drugs that affect cell volume (5 mM Dithiothreitol; DTT, 5 µM Tunicamycin; TN and 200 mM Hydroxyurea; HU). ∆cln3 mutant was used as a control for large cell volume and ∆sfp1 for small cell volume. Scale bar: 5 μm.B. Quantification of individual volumes at budding of (A) yeast cells (box plots represent 25th percentile, median, 75th percentile, the whiskers extend to the minimum and maximum values; n = 400 cells; 3 independent biological replicates). The statistical analysis was done using one-way ANOVA and Tukey–Kramer post hoc test (***P [less than] 0.001).

DAAs disrupt the integrity and the morphology of ER

To further dissect the effects on cellular organelles and other physiological aspects that are affected by tested drugs, we investigated the effect of DAAs on ER architecture. As a positive control, we used tunicamycin that is known to disturb ER integrity (Schuck et al., 2009; Austriaco, 2012) in a background expressing Sec63-GFP, a subunit of the Sec62/Sec63 complex involved in post-translational translocation of proteins to ER and commonly employed to evaluate the ER integrity (Schuck et al., 2009; Chaillot et al., 2015). Sec63 is an essential ER translocation protein required for protein import to ER that localizes to both cortical ER (cER) and nuclear ER (nER). ER organization as judged microscopically by Sec63-GFP distribution and compartmentalization was assessed in cells challenged with DAAs and in control cells. The control cells exhibited a clear and well-characterized ER organization with the nER around the nuclear compartment, the cER adjacent to plasma membrane at the cell periphery and few cytoplasmic ER tubes (Fig. 3A). However, in cells treated with tunicamycin, the ER became fragmented, and the GFP signal was diffused in the cytoplasm. The nER integrity was moderately or entirely disrupted in some cells (Fig. 3A). Cells treated with DAAs contained large cytoplasmic foci, possibly representing collapsed ER (Fig. 3A). Taken together, these findings indicate that DAAs disrupt ER organization. DAAs-treated cells also showed enhanced upregulation (relative to untreated cells) of Hsp26 (CT: 0.56 ± 0.15; SOF:1.77 ± 0.15 (P < 0.0001); DCV: 1.50 ± 0.1 (P < 0.0001)), a small cytosolic heat shock protein (sHSP) with a molecular chaperone function (Petko and Lindquist, 1986; Carmelo and Sá-Correia, 1997), albeit to a lower level than in cells treated with TN (3.03 ± 0.20 (P < 0.0001)) (Fig. 3B and C). The altered ER morphology and upregulated markers of UPR response indicate that DAAs induce ER stress and accumulation of unfolded protein.

Fig. 3. Effect of Sofosbuvir and Daclatasvir on ER architecture.A. Representative fluorescence images of Sec63-GFP localization in control cells (CT), cells treated with 64 µM Sofosbuvir (SOF), 32 µM Daclatasvir (DCV) or 5 µM Tunicamycin (TN) as a positive control for ER stressing agent for 4–6 h. Cortical (cER) and nuclear ERs (nER) are labelled. Scale bar, 5 µm.B. Expression analysis of heat shock protein Hsp26. Cells expressing Hsp26-6HA were exponentially grown for 4–6 h in YPD media supplemented with DMSO, 64 µM Sofosbuvir (SOF), 32 µM Daclatasvir (DCV) or 5 µM Tunicamycin (TN). Levels of Hsp26-6HA were analysed by Western blot, and Rps23 was used as a loading control.C. Relative abundance quantification of Hsp26-6HA in (B). Mean ± SD (3 independent biological replicates). The statistical analysis was done using one-way ANOVA and Tukey–Kramer post hoc test (***P [less than] 0.001).

Yeasts treated with DAAs show no apparent genotoxicity

The major concerns about any new drug arise from possible mutagenic or genotoxic effects. To investigate possible genotoxic effects of DAAs, we monitored the phosphorylation status of Rad53, an essential protein kinase required for DNA damage signalling (Cordón-Preciado et al., 2006). Upon exposure to DNA damaging agent, Rad53p amplifies initial signals through modifying the phosphorylation state of downstream protein targets that recognize DNA damage and block replication (Allen et al., 1994; Weinert et al., 1994). Yeast cultures arrested in G1 by alpha factor were released into media with the tested drugs or a vehicle; as a positive control for DNA damage, we treated yeast culture with HU. Cells treated with HU showed a dramatic increase of Rad53 protein levels (2.00 ± 0.26 (P < 0.0001)) and noticeable shift to high phosphorylated forms, while yeasts cells treated with DAAs (SOF: 0.45 ± 0.12 (P = n.s.); DCV: 0.44 ± 0.14 (P = n.s.)) or the control (CT: 0.42 ± 0.22) show no change in the level or in the phosphorylation status of Rad53 (Fig. 4A and B); in the same experiment we monitored the budding index as an indicator of the cell cycle progression, cells treated with DAAs were able to pass G1, enter S phase and normally behave as control cells while HU-treated cells clearly arrested in S phase (Fig. 4C). Drugs that interfere with microtubule dynamics and tubulin polymerization, for example benomyl, are highly genotoxic and lead to aneuploidy due to incorrect segregation of sister chromatids (Georgieva et al., 1990; Mailhes and Aardema, 1992). A sensitive assay to record correct segregation was used, where segregation of chromosome 4 labelled by tetO/TetR-GFP system near centromeres was monitored in cells released from mitotic arrest into media with or without the tested drugs (He et al., 2000). Again, no significant changes were observed in cells treated with DAAs (SOF: 93.0 ± 2.64 (P = n.s.) and DCV: 92.0 ± 4.00 (P = n.s.)) compared with control cells (CT: 92.33 ± 3.05), while the benomyl treated cells showed clear signs of chromosomes mis-segregation (Benomyl: 55.0 ± 7.55 (P < 0.0001)) (Fig. 4D and E). Together, these results suggest no genotoxic effects of the DAAs.

Fig. 4. Sofosbuvir and Daclatasvir cause no genotoxic stress.A. Cells expressing Rad53-6HA were exponentially grown in YPD media, arrested in G1 phase with α-factor and then released into YPD media supplemented with DMSO, 64 µM Sofosbuvir (SOF), 32 µM Daclatasvir (DCV) or 200 mM Hydroxyurea (HU) as a positive control for genotoxic stress. Samples were collected at time 0 and after 2 h post-treatments. Basal levels and phosphorylation of Rad53-6HA were analysed by Western blot. Ponceau staining was used as a loading control.B. Relative abundance quantification of Rad53-6HA (relative to time 0 before treatment) in (A). Mean ± SD (3 independent biological replicates). The statistical analysis was performed using one-way ANOVA and Tukey–Kramer post hoc test (***P [less than] 0.001).C. Budding index for the respective cells in (A), samples were collected at the indicated time points, budding index was calculated after microscopic examination (at least 200 cells per count were scored). Mean ± SD is shown from 3 independent biological replicates.D. Chromosome segregation (centromeres of chromosome 4 labelled by CEN4-tetO/TetR-GFP; in green) in cells grown in media supplemented with DMSO, 64 µM Sofosbuvir (SOF) 32 µM Daclatasvir (DCV) or benomyl (10 μg ml−1). Spc29-RFP stains the spindle pole bodies; in red. Scale bar, 5 µm.E. Percentage (mean ± SD) of correct chromosome segregation of cells in (D); (correct segregation was identified as a distribution of the 2 chromatids between mother and daughter cells, while existence of two dots in one cell was scored as a mis-segregation event, at least n = 300 cells for each experiment were counted; 3 independent biological replicates). The statistical analysis was done using one-way ANOVA and Tukey–Kramer post hoc test (***P [less than] 0.001).

DAAs reduce the integrity and quantity of mitochondrial DNA

To evaluate possible effects of DAAs on mitochondrial DNA, we stained DNA of live cells growing in the presence of DAAs with DAPI (Fig. 5A). Treated cells showed a decrease in the mitochondrial DNA. We quantified the mtDNA content relative to genomic DNA using qPCR with specific primers that target COX3, a gene encoded in the mitochondrial DNA and the housekeeping ACT1 in the genomic DNA. As a control, we used a yeast rho⁰ strain lacking the mtDNA. The distribution of mtDNA was significantly altered after the treatment, as DAAs-treated cells showed a significant decrease of mtDNA compared with control cells (CT: 1.00 ± 0.0; SOF: 0.71 ± 0.01 (P < 0.0001); DCV: 0.74 ± 0.04 (P < 0.0001)) (Fig. 5B). Additionally, we evaluated mitochondrial Cox2 protein levels by Western blot and observed a significant decrease (SOF: 0.65 ± 0.01 (P < 0.0001); DCV: 0.76 ± 0.06 (P < 0.0001)) compared with control cells (CT: 1.54 ± 0.19, Fig. 5C and D)). As a negative control, we treated cells with chloramphenicol (CHL) that inhibits mitochondrial translation and thus reduced Cox2 protein levels (0.44 ± 0.09 (P < 0.0001)). These results show that DAAs reduced the levels and the integrity of mtDNA.

Fig. 5. Sofosbuvir and Daclatasvir affect integrity and copy number of mitochondrial DNA.A. Mitochondrial nucleoid signals in control cells, 64 µM Sofosbuvir (SOF), 32 µM Daclatasvir (DCV) and cells depleted of mitochondrial DNA (rho0) as a negative control. Cells were grown to mid-logarithmic growth phase (6–8 h) in YPG media and stained with DAPI to visualize DNA. Scale bar, 5 µm.B. Quantification of mitochondrial DNA copy number by qPCR for the respective cells in (A). Total DNA was extracted from respective cells, and qPCR was performed using primers for COX3 (mitochondrial DNA locus) and ACT1 (housekeeping chromosomal DNA locus). Mean ± SD (n = 3 independent biological replicates). The statistical analysis was done using one-way ANOVA and Tukey–Kramer post hoc test (***P [less than] 0.001).C. Cox2 protein levels analysed by Western blot in control cells, 64 µM Sofosbuvir (SOF), 32 µM Daclatasvir (DCV) or with 50 µg ml−1 Chloramphenicol (CHL) as a negative control.D. Relative abundance quantification of Cox2 in (C). Mean ± SD (3 independent biological replicates). The statistical analysis was done using one-way ANOVA and Tukey–Kramer post hoc test (***P [less than] 0.001).

DAAs cause distinct mitochondrial fragmentation and impair mitochondrial respiration

Mitochondria are cellular organelles with a main function to generate energy. In addition to this role, mitochondria are involved in maintenance of calcium homeostasis, metabolism of amino acid and nucleotides, and biosynthesis of iron–sulphur clusters, haeme and ubiquinone (Young, 2017). Yeast strains that were cultured in growth media supplemented with DAAs showed enhanced fragmentation of mitochondria similar to the pattern exhibited by yeast strain depleted of mitochondrial DNA, whereas control cultures exhibited the typical network morphology of mitochondria, when mitochondrial integrity was examined by microscopy (Fig. 6A). To test a direct effect of altered mitochondrial integrity and loss of mtDNA, we tested oxygen concentration is well-established indicator for mitochondrial function. Indeed, yeasts treated with DAAs showed significantly reduced oxygen concentration and lower basal respiration (Fig. 6B).

Fig. 6. Sofosbuvir and Daclatasvir alter mitochondria integrity and impair mitochondria respiration.A. Representative fluorescence microscopy images of control cells, 64 µM Sofosbuvir (SOF) or 32 µM Daclatasvir (DCV) and respiratory deficient cells (rho0) as a negative control. Cells harbouring pYX232-mtGFP (a plasmid for expression of mitochondria-targeted GFP in yeast) were grown to mid-logarithmic growth phase (6–8 h) in YPG, and then subjected to fluorescent microscopy to visualize mitochondria. Scale bar, 5 µm.B. Oxygen concentration of respective cells in (A). Values are mean oxygen concentration ±SD (n = 3 independent biological replicates). The statistical analysis was performed using Kruskal–Wallis test followed by Dunn post hoc test (***P [less than] 0.001).

Discussion

DAAs display high performance and satisfactory safety profile in patients since their introduction into drug markets for the control of Hepatitis C. DAAs have dramatically changed the HCV treatment landscape, with elevated sustained virologic response (SVR), potential benefits in CHC treatment and improved compliance (Asselah, 2014; Virlogeux et al., 2016; Shiha et al., 2018; Heffernan et al., 2019). DAAs-based therapy of HCV infections has achieved high virologic cure rates (>90%) superior to the interferon-based therapy [interferon (INF) plus ribavirin (RBV)], with a low SVR not exceeding 50% (Seo et al., 2017). Common side effects of DAAs are fatigue, nausea, dizziness and anaemia. However, with the wide application of DAAs combination regimens, recent clinical studies indicated unwanted side effects and development of certain types of cancer, which raises concerns about possible effects of DAAs on genome stability and other aspects of cellular physiology.

DAAs display no signs of genotoxicity

Yeast-based screens have been well established for various chemical biology studies to monitor genotoxicity and cytotoxicity, identify DNA damaging agents, determine the underlying mechanisms and configure side effects of the newly synthesized drugs on functional organelles and metabolism (Kock et al., 2009; dos Santos, 2012; van Bui et al., 2015; Zhou et al., 2016; Fasullo et al., 2017; Delerue et al., 2019). Polyploid yeast shows prominent sensitivity compared with haploid or diploid yeasts when challenged with drugs and chemicals that induce genotoxicity or interfere with chromosome maintenance or organization, including DNA damaging and microtubule depolymerizing drugs (Storchová et al., 2006; Yahya et al., 2021). In case of DAAs, both haploid and polyploid yeasts exhibited similar sensitivity profile at the same inhibitory concentration of DAAs. Moreover, cells treated or untreated with DAAs displayed similar budding profile and segregated their duplicated chromosomes with the same efficiency. Additionally, DAAs-treated yeast did not accumulate high phosphorylated forms of Rad53, a well-known marker of genotoxic stress response. All these results suggest that DAAs treatment has negligible genotoxic effects.

DAAs disturb the architecture of ER and mitochondria

ER has been frequently reported as a nonspecific target of several antiviral agents (Ganta and Chaubey, 2019). Therefore, we tested the global effect of DAAs on organelles integrity. Indeed, yeast cells treated with DAAs showed discernible signature of ER disturbance in the form of fragmented or collapsed ER, accompanied with upregulation of cytosolic heat shock protein Hsp26, possibly in response to UPR provoked by ER stress.

Previous studies showed that the antiviral nucleoside analogues, such as zidovudine (AZT), can generate mitochondrial toxicity upon long-term treatment (Sun et al., 2014; Jin et al., 2017). Our study found that mitochondria were strongly affected in response to DAAs. Treated cells suffered from alterations of mitochondrial dynamics, accumulated fractured mitochondria and showed distortion of the network pattern of healthy mitochondria. Moreover, DAAs-treated cells displayed significant loss of mtDNA. The exact role of DAAs in disturbing the mitochondrial physiology and organization has not been investigated in detail, although mitochondrial toxicity triggered by antiviral drugs and DAAs has been previously recognized (Lewis and Dalakas, 1995; Kakuda, 2000; Kim et al., 2017). The undesirable effects of DAAs on mitochondrial morphology and the level of mtDNA likely caused the reduced oxygen consumption and lower basal respiration rate. Our observation may provide a mechanistic explanation for the clinical side effects that include fatigue and dizziness observed in most patients. In the same context, drugs that interfere with mitochondrial homeostasis were found to promote alterations of cardiac physiology (Varga and Pacher, 2018). Together with our findings, this may explain the reported cardiac toxicity associated with DAAs (Ucciferri et al., 2017; El Missiri et al., 2020).

DAAs and hepatitis B reactivation

Due to pharmacological eradication of HCV, DAAs may possibly reactivate hepatitis B virus (HBV) in formerly infected patients. Thus, the European Medicines Agency (EMA) and the United States Food and Drug Administration (FDA) have recommended checking all HCV patients for HBV coinfection before starting DAA-based therapy for HCV to reduce the risk of HBV reactivation (Cheung et al., 2016; Chen et al., 2017; Holmes et al., 2017; Loggi et al., 2017; Blackard and Sherman, 2018; El Kassas et al., 2018; Jiang et al., 2018; Wang et al., 2019; Miyasaka et al., 2020; Palacios-Baena et al., 2020). Our findings may present another indirect link between induction of ER stress and mitochondrial disturbance due to DAAs treatment and HBV reactivation, although this assumption will require further investigation. It was reported that HBV can induce ER stress and activate UPR. The HBV-induced ER stress and UPR signalling increase the folding capacity of the infected cells by upregulating chaperons as a response to the accumulated unfolded proteins. HBV can employ this folding protein surplus for processing viral proteins, genome replication or virion assembly (Lazar et al., 2014). Additionally, the DAAs-induced ER stress may instigate stimuli sufficient for reactivating inactive HBV.

Experimental procedures Strains, plasmids and growth conditions

Yeast strains and bacterial plasmids used in this study are listed in Table S1. Yeast strains were based on the W303 or S288C genetic backgrounds of a mating type (Winston et al., 1995). Yeast-diluted precultures were grown exponentially for 4–5 generations in YP (1% yeast extract, 2% peptone), or in synthetic complete media (SC; 1.34% yeast nitrogen base, 0.04% complete synthetic mix) supplemented with 2% glucose (YPD or SDC), or 2% galactose (YPG or SGC), aerated by shaking at 30 °C. Yeast cells were arrested in late G1 by treating exponential cultures with 5 μg ml⁻¹ α-factor (Core Facility, Max Planck Institute of Biochemistry, Martinsried, Germany) for 105 min at 30 °C, or arrested in mitosis by adding 30 µg ml⁻¹ nocodazole for 3 h at 30°C (Santa Cruz Biotechnology; Dallas, TX, USA).

General genetic methods

Standard protocols were used for DNA manipulations and transformation of yeast cells. Single null mutants and epitope fusion to target genes were constructed using the homologous recombination approach after PCR amplification of the antibiotic resistance kanMX4 cassettes with short homologous sequences flanking the gene of interest and selection for geneticin resistance (Wach et al., 1994; Janke et al., 2004). Gene deletions were confirmed by PCR analysis. For adding the C-terminus epitope tags, Phusion DNA Polymerase (Thermo Fisher Scientific; Waltham, MA, USA) was used.

Determination of growth sensitivities

Sensitivity assays to DAAs were carried out by spotting 1:10 serial dilutions of exponential cultures onto YPD plates supplemented with the tested drugs at 30 °C for 48 h (Pijuan et al., 2015).

Minimal inhibitory concentration (MIC) determination

A log phase culture in complete synthetic media was diluted to 0.02 OD₆₀₀ as determined in a standard spectrophotometer, in a sterile 96-well plate (Thermo Fisher Scientific) with rounded bottom. Tested drugs were diluted in the same medium (2-fold serial dilution) starting from a concentration of 512 μM. 100 μl of the diluted culture was added to 100 μl of diluted chemicals using a multichannel pipette and mixed three times (3 wells were mixed with a blank as a control and one well was free from cells and drug), then covered and incubated at 30 °C for 16 to 20 h with gentle shaking. Plates were then removed, and OD measurements were monitored using GloMax^® Discover Microplate Reader (Promega; Madison, WI, USA).

Oxygen concentration analysis

Cells were grown exponentially in YPG medium. 125 µM DCV or 500 µM SOF was added, and the cultures were further grown for 6–8 h. 1 OD₆₀₀ cells were collected by centrifugation (5000 g, 5 min, room temperature) and resuspended in water. An Oxygraph (Hansatech Instruments; Pentney, UK) was calibrated with oxygen-saturated and oxygen-depleted water, using sodium dithionite for oxygen removal. Oxygen levels of 900 μl of saturated growth medium were measured until a plateau was reached, 100 μl of cell suspension was added and oxygen levels were measured continuously until they reached 0 μmol ml^–1.

Western blot analysis

Total cell lysate was prepared from cell pellets with 5 OD₆₀₀, where 15 µl of 5 M urea was added to cell pellet and boiled for 2 min immediately. An equal volume of glass beads was added, then cell suspensions were vortexed for 8 min at room temperature, and 50 µl of 2% SDS and 0.125 M Tris–HCl pH 6.8 were then added, vortexed for 2 additional min, boiled for 2 min and centrifuged. The protein concentration was determined by the Micro DC protein assay (Bio-Rad, Richmond, CA, USA). Total protein was separated in polyacrylamide gel and blotted to PVDF membranes (GE Healthcare, Chicago, IL, USA). The membranes were blocked with 5% skim-milk and washed in TBS-0.1% Tween-20 buffer (25 mM Tris, 50 mM NaCl, 2.5 mM KCl and 0.1% Tween-20). Afterwards, the membranes were incubated with the specific primary antibodies overnight at 4 °C. After three washes, the membranes were incubated with the secondary antibody coupled to the horseradish peroxidase for 2 h at room temperature (Table S2). Membranes were incubated with Amersham ECL Prime Western Blotting Detection Reagent and visualized by iBright™ CL1000 Imaging System (Thermo Fisher Scientific). The intensity of individual bands was measured using ImageJ software (NIH, Bethesda, MD, USA; http://rsb.info.nih.gov/ij).

DNA isolation and mitochondrial quantitative PCR analysis

Total DNA was extracted using YeaStar™ Genomic DNA Kit (Zymo Research; Irvine, CA, USA) according to the manufacturer's instructions from mid-logarithmic growth cultures. Total DNA was subjected to qPCR using the iQ-Supermix (Bio-Rad). qPCR reactions were performed in 96-well plates with 20 μl of volume using Maxima™ SYBR Green qPCR Master Mix (Thermo Fisher Scientific) according to the manufacturer's instructions. For mtDNA and nuclear DNA, loci within the COX3 and ACT1 were amplified respectively (see Table S2 for primer sequences). mtDNA levels were calculated relative to nuclear DNA by the 2^−∆∆CT method (Livak and Schmittgen, 2001).

Immunofluorescence

Differential interference contrast (DIC) and fluorescence microscopy images were captured from living cells with Axioscope 2 microscope (Zeiss; Jena, Germany). Images were acquired with a Cool-SNAP-HQ 12-bit monochrome digital camera (Roper Scientific; Trenton, NJ, USA). Yeast cultures were grown exponentially in liquid growth medium (YPD or YPG) supplemented with the indicated drugs. Cell samples were collected and concentrated 10-fold by centrifugation (12 000 g, during 1 min at room temperature); then, cell suspensions were incubated with 1 μg ml⁻¹ 4′,6-diamidino-2-phenylindole (DAPI) for 5 min to stain the DNA. More than 200 cells from at least three independent experiments were analysed in each measurement.

Miscellaneous

Cell volume was determined from bright-field images and measured by BudJ plugin (Ferrezuelo et al., 2012) using ImageJ; budding index was calculated according to (Amponsah et al., 2021). All chemical compounds used in this work are listed in Table S3.

Statistical analyses

Data are presented as mean ± standard deviation (SD) and box plots showing the median, box edges [25th–75th percentiles], and the whiskers prolonged to the minimum and maximum values. Statistical significance was assessed by application of the Kolmogorov–Smirnov normality test followed by one-way ANOVA and Tukey–Kramer post hoc test (parametric test), or Kruskal–Wallis test followed by Dunn post hoc test (non-parametric test). P-values are indicated by asterisks *P < 0.05; **P < 0.01; ***P < 0.001. All statistical tests were conducted using GraphPad Prism software v.8.0.2 (GraphPad; La Jolla, CA, USA).

Acknowledgements

G.Y. is grateful to the Alexander-von-Humboldt Foundation (AVH) for granting the Return Fellowship EGY 1194703 GF-P. R.A. and M.M.A. would like to extend their sincere appreciation to the Researchers Supporting Project number (RSP-2020/96), King Saud University, Riyadh, Saudi Arabia.

Conflict of interest

The authors declare that they have no conflict of interest.

Funding Information

G. Y. was funded by Return Fellowship EGY 1194703 GF-P awarded by the Alexander-von- Humboldt Foundation. J. P. was supported by ‘Beques de Recerca IRSJD-Carmen de Torres 2020’. The funders had no role in study design, analysis or decision to publish.