Abstract

   Acanthamoeba castellanii is a widespread unicellular eukaryote found in
   diverse environments, including tap water, soil, and swimming pools. It
   is responsible for severe infections, such as Acanthamoeba keratitis
   and granulomatous amebic encephalitis, particularly in individuals with
   immunocompromisation. The ability of protozoans to form dormant and
   persistent cysts complicates treatment, as current therapies are
   ineffective against cyst stages and suffer from poor specificity and
   side effects. Nitroxoline, a quinoline derivative with well-established
   antibacterial, antifungal, and antiviral properties, is a promising
   therapeutic candidate. This study aimed to elucidate cellular
   signalling events that counteract the effects of nitroxoline. In this
   study, nitroxoline significantly reduced the viability of A.
   castellanii trophozoites in a dose- and time-dependent manner, inducing
   morphological changes and apoptosis. Transcriptomic analysis revealed
   substantial alterations in gene expression, including enrichment of
   metabolic pathways, DNA damage responses, and iron ion binding.
   Nitroxoline treatment upregulated genes involved in DNA repair and
   oxidative stress response while regulating genes in the methionine and
   cysteine cycles. It also decreased the mitochondrial membrane
   potential, H₂S production, and total iron amount in A. castellanii.
   Bioinformatic analyses and molecular docking studies suggest direct
   interactions between nitroxoline and several A. castellanii proteins.
   Our research provides a comprehensive molecular map of the response of
   A. castellanii to nitroxoline, revealing significant changes in gene
   expression related to the stress response and metabolic pathways. These
   findings underscore the potential of nitroxoline as a potent
   anti-Acanthamoeba agent, offering new insights into its mechanism of
   action and paving the way for effective combinational therapeutic
   strategies.

   Keywords: Nitroxoline, Acanthamoeba castellanii, Transcriptomics,
   Molecular docking

Graphical abstract

   [33]Image 1
   [34]Open in a new tab

Highlights

     * •
       Nitroxoline triggers stress response pathways of Acanthamoeba.
     * •
       Apoptosis of Acanthamoeba was caused by nitroxoline.
     * •
       Nitroxoline induced DNA damage and decreased intracellular iron
       level.
     * •
       Nitroxoline could be added as a therapeutic candidate for the
       treatment of Acanthamoeba diseases.

1. Introduction

   Acanthamoeba castellanii is a cosmopolitan unicellular eukaryotic
   organism commonly found in tap water, swimming pools, soil, cooling
   towers, and drinking water networks. Free-living Amoeba including
   Acanthamoeba species, are opportunistic protist with a widespread
   distribution worldwide ([35]Visvesvara et al., 2007; [36]Siddiqui and
   Khan, 2012; [37]Zhang and Cheng, 2021). In humans, various Acanthamoeba
   species can cause a sight-threatening disease named, Acanthamoeba
   keratitis (AK), which is the more common disease seen by Acanthamoeba,
   found more commonly in contact lens misuse by wearers
   ([38]Cabrera-Aguas et al., 2022; [39]Niederkorn, 2021). In
   immunocompromised patients, Acanthamoeba can cause chronic infectious
   ulcers, cutaneous lesions, and furthermore, through systemic
   dissemination, make its way to the brain causing a normally fatal
   disease granulomatous amebic encephalitis (GAE)([40]Kalra et al., 2020;
   [41]Wang et al., 2021; [42]Damhorst et al., 2022). Acanthamoeba can
   cause AK in immunocompetent individuals, and its distribution is
   limited to the eye, where it can cause keratitis and blindness, if not
   diagnosed early, treated appropriately and aggressively. A study
   estimated that the global incidence of AK is 2.9 cases per million
   people with an increasing trend in incidence observed in some regions
   (Y. [43]Zhang et al., 2023).

   Acanthamoeba also has the innate ability to survive under harsh
   environmental conditions and sub-lethal drug exposure without genetic
   resistance by transiently stopping their growth, slowing their
   metabolism, and eventually shifting to dormant and persistent cysts. A
   conceptually compelling but sparsely investigated approach for
   identifying useful potentiators involves disrupting the general defence
   systems that protect pathogens from diverse antibiotics. Currently, no
   effective anti-Acanthamoeba therapies are available. Therefore,
   innovative anti-Acanthamoeba drugs are urgently required. Most current
   anti-Acanthamoeba drugs target the trophozoite-stage parasites.
   Although cyst-stage parasites do not cause clinical symptoms, drugs
   that inhibit these two stages are essential for preventing epidemics
   and protecting vulnerable populations, owing to the increase in
   anti-Acanthamoeba clinical treatment resistance ([44]Elder et al.,
   1994; [45]Lim et al., 2008; [46]Huang et al., 2021). Chemotherapy for
   Acanthamoeba infections remains challenging due to incomplete knowledge
   of its complicated pathophysiology. In cases of infection, the
   treatment regimen is often ineffective owing to delayed diagnosis, poor
   specificity, and side effects. However, this route is lengthy, with
   many hurdles and failures, including target specificity, poor
   pharmacodynamics, pharmacokinetics, and toxicity.

   Considering the limited number of anti-Acanthamoeba agents, screening
   of potential anti-Acanthamoeba agents from natural products or approved
   drugs is warranted. Nitroxoline, a quinoline derivative
   (8-hydroxy-5-nitroquinoline), is a historically used oral antibacterial
   drug ([47]O'Grady and Smith, 1966) that is structurally distinct from
   other antibiotics. In addition to its antitumor effects, it is active
   against numerous human pathogens, including multidrug-resistant
   bacteria, fungi, and viruses ([48]Fuchs et al., 2021, [49]2022a,
   [50]2022b; [51]Joaquim et al., 2021). Its antibacterial properties are
   based on the chelation of divalent cations, influencing cell membranes,
   and inhibiting intracellular enzymes such as RNA polymerase
   ([52]Hoffmann et al., 2023). From a mechanistic standpoint, nitroxoline
   partially interferes with the phosphatidylinositol
   3-kinase/AKT/mammalian target of rapamycin and rapidly accelerated
   fibrosarcoma/mitogen-activated protein kinase/extracellular
   signal-regulated kinase host cell signalling pathways ([53]Bojkova et
   al., 2023).

   Recent studies have also indicated that nitroxoline is a promising drug
   against free-living pathogenic amoebas, including Naegleria fowleri,
   Balamuthia mandrillaris, and Acanthamoeba species ([54]Haston and Cope,
   2023; [55]Laurie et al., 2018; [56]Rodríguez-Expósito et al., 2023).
   Although transcriptomic and proteomic analyses have been performed on
   these free-living amoebae, the signalling pathways involved in their
   stress responses remain limited. Furthermore, our knowledge of the
   transcriptomic landscape during nitroxoline treatment remains limited.
   This study aimed to elucidate the cellular signalling events that
   counteract the effects of nitroxoline.

2. Materials and methods

2.1. Amoeba cultivation and cysticidal activity assay

   Acanthamoeba castellanii (American Type Culture Collection (ATCC)
   30011) was obtained from the ATCC. Following established protocols,
   trophozoites were axenically cultured in a peptone–yeast–glucose
   medium. This medium comprised 10 g of protease peptone, 10 g of yeast
   extract, 10 g of glucose, 5 g of NaCl, 0.95 g of L-cysteine, 3.58 g of
   Na[2]HPO[4]ꞏ12H[2]O, and 0.68 g of KH[2]PO[4] per litre of deionised
   distilled water. The culture was incubated at 26 °C, and trophozoites,
   harvested during the late log phase after 48 h of subculture, were used
   for subsequent analyses. Encystment was induced through the following
   procedure: 5 × 10^6 trophozoites were washed twice with an encystment
   medium containing 95 mM NaCl, 5 mM KCl, 8 mM MgSO[4], 0.4 mM CaCl[2],
   1 mM NaHCO[3], 20 mM Tris-HCl, pH 9.0 and incubated in 10 mL of the
   encystment medium in the 75 cm^2 culture flask at 26 °C ([57]Hirukawa
   et al., 1998). After the encystment process was completed, the cysts
   were collected and centrifuged at 1000 g for 30 min. Subsequently, 0.5%
   SDS was used to treat the cysts for 10 min to remove immature cysts.
   Then, mature cysts were washed three times with PBS and centrifuged at
   1000 g for 15 min. Cysts (5 × 10^5 cysts/mL) were seeded in a 12-well
   plate and respectively incubated with PYG medium containing 0.1% DMSO,
   20 μM nitroxoline or 40 μM nitroxoline for 72 h. After 3 days of
   treatment, the medium was replaced with fresh PYG medium, and the
   culture was continued for recover observation for another 9 days.

2.2. Cell viability assay and IC[50] determination

   Obtained from TargetMol (Wellesley Hills, MA, USA), nitroxoline was
   dissolved in 100% dimethyl sulfoxide (DMSO; Sigma-Aldrich, St. Louis,
   MO, USA) and prepared at 100 mM concentrations. A. castellanii
   trophozoites were seeded at the log growth phase (1 × 10^4 cells per
   well) in a 96-well white microplate (Eppendorf, Hamburg, Germany) and
   incubated with 100 μL of medium containing varied nitroxoline
   concentrations (2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 40, 60, 80, and
   100 μM). And the plates were incubated at 26 °C for 24 h, 48 h, and
   72 h and observed under an inverted microscope (Olympus). Cell
   viability assay results, derived from ≥3 wells per condition and
   compared with 0.1% DMSO, were evaluated using CellTiter-Glo (Promega,
   Madison, WI, USA) following the manufacturer's guidelines.
   Subsequently, 100 μL of CellTiter-Glo reagent was introduced per well,
   mixed for 2 min on an orbital shaker, and incubated at room temperature
   for 10 min. Luminescence was recorded on a modular multimode microplate
   reader (BioTek Synergy H1), and the IC[50] values of the compound and
   growth curves were generated using GraphPad Prism 8. Each experiment
   was conducted thrice.

2.3. Flow cytometric analysis

   Trophozoite apoptosis and reactive oxygen species (ROS) were assessed
   using a PE Annexin V apoptosis detection kit I (BD Biosciences, San
   Jose, CA, USA) and a ROS assay kit (Dojindo Molecular Technologies,
   Inc., Kumamoto, Japan), respectively, following the manufacturer's
   instructions. A. castellanii trophozoites (2 × 10^6 cells/flask) were
   cultured with 20 μM and 40 μM nitroxoline for 24 h in 12.5 cm^2 cell
   culture flasks (Biofil). Trophozoites were then detached from the
   flasks, collected in transparent centrifuge tubes (15 mL), and
   centrifuged at 800×g for 5 min. These trophozoites
   (1 × 10^6 cells/tube) underwent two cold Hanks' balanced salt solution
   (HBSS) washes and were respectively resuspended in 100 μL of
   1 × binding buffer containing 5 μL of PE-Annexin V and 500 μL of ROS
   working solution containing 0.5 μL of highly sensitive DCFH-DA dye.
   Subsequently, the trophozoites for apoptosis detection were incubated
   for 15 min in darkness at room temperature, and 400 μL of 1 × binding
   buffer was added after incubating. The trophozoites for ROS detection
   were incubated for 30 min in darkness at room temperature, washed twice
   with HBSS, and resuspended with 500 μL of HBSS. Fluorescence-activated
   trophozoite sorting was performed using a FACSAria instrument (BD
   Biosciences, San Jose, CA, USA) with a 488-nm argon excitation laser.
   Analysis gates were defined using untreated amoebae, and FlowJo 10.8.1
   software (FlowJo LLC, Ashland, OR, USA) was used for data analysis.

2.4. RNA-sequencing (RNA-seq)

   Total RNA was isolated from each sample using an RNA mini kit (Qiagen,
   Hilden, Germany). RNA quality was examined using gel electrophoresis
   and Qubit (Thermo, Waltham, MA, USA). For RNA sequencing, the RNA
   samples from three biological replicates at 24 h were separated into
   three independent pools, each containing two or three distinct samples
   in equal amounts. The raw data were subsequently analysed as previously
   described ([58]Chen et al., 2024).

2.5. Real-time quantitative polymerase chain reaction (RT-qPCR)

   Total RNA was isolated using the RNeasy Plus Mini Kit (74134; QIAGEN,
   Hilden, Germany), and cDNA was synthesised using the PrimeScript™ II
   1st Strand cDNA Synthesis Kit (6210A; Takara Bio, China). The reactions
   (20 μL per well) were performed in a 96-well plate using 2 μL of SYBR
   Premix Ex Taq (Takara, Dalian, Liaoning, China) with 10 μM primers
   targeting A. castellanii GNMT, CBS, GSR, PARP, MRE11, RAD50, RAD51,
   DNA2, FEN1, APE1, and DMT1. Primers for these genes, sourced from
   published sequences ([59]Table S1), were used for RT-qPCR on an ABI
   7500 RT-PCR system (Applied Biosystems, Foster City, CA, USA). The
   2^-^△△Ct method was used to calculate the relative expression of each
   primer between the control and treatment groups, with values normalised
   to the 18S reference housekeeping gene ([60]Deng et al., 2015).

2.6. DNA damage determination

   Genomic DNA was extracted using the DNeasy Blood & Tissue Kit
   (Catalogue No. 69506; QIAGEN), following the manufacturer's protocol.
   DNA concentrations were quantified using a BioPhotometer® D30
   (Eppendorf). Trophozoite DNA damage was assessed using the DNA Damage
   Quantification Kit AP Site Counting (Dojindo, catalogue DK02). DNA was
   dissolved in TE buffer (10 mM Tris [pH 7.5] and 1 mM EDTA) at
   100 μg/mL, and 10 μL of purified genomic DNA was mixed with 10 μL of
   aldehyde reactive probe (ARP) solution in a MaxyClear microcentrifuge
   tube (1.5 mL; Corning, NY, USA) and incubated for 1 h at 37 °C. Each
   ARP-labelled DNA was purified with a filtration tube, and DNA
   concentration was measured using the BioPhotometer® D30. Purified
   ARP-derived DNA samples were diluted to 2.25 μg/mL in TE buffer, and
   90 μL of the ARP-labelled DNA solution was diluted with 310 μL of TE
   buffer. Thereafter, 60 μL of the diluted solution and 100 μL of DNA
   binding solution were added to the provided DNA high-binding plate
   well, incubated overnight at room temperature, and microwells were
   washed five times in 250 μL of 1 × wash buffer. Subsequently, 150 μL of
   diluted horseradish peroxidase-streptavidin-enzyme conjugate (1:4000
   dilution in 1 × wash buffer) was added to each well, incubated for 1 h
   at 37 °C, and microwells were washed five times in 250 μL of 1 × wash
   buffer. After adding 100 μL of substrate solution, microwells were
   incubated at 37 °C for 1 h. Absorbance at an optical density (OD) of
   650 nm was immediately measured using a modular multimode microplate
   reader (BioTek Synergy H1).

2.7. Mitochondrial membrane potential (MMP) assay

   MMP was evaluated using the MT-1 MitoMP detection kit (MT13, Dojin
   Kagaku) following the manufacturer's protocol. Treated trophozoites
   (1 × 10^6) were incubated with MT-1 working solution (v:v, 1:1000) for
   30 min at 26 °C. After treatment with the MT-1 working solution, the
   trophozoites were washed twice with HBSS and incubated in an imaging
   buffer (Kit-provided). Live trophozoite suspensions were placed on
   glass slides using a Cytospin 4 cytocentrifuge (Thermo Fisher
   Scientific) and mounted on coverslips. The fluorescence images were
   obtained using a Leica TCS SP8 microscope, and the fluorescence
   (λEx = 488 nm and λEm = 580 nm) was measured using a modular multimode
   microplate reader (BioTek Synergy H1). Images were quantified using
   ImageJ software (National Institutes of Health, Bethesda, MD, USA).

2.8. Measurement of H[2]S production in trophozoites

   Following a 24-h treatment, trophozoites (1 × 10^6) were washed twice
   with 1 × phosphate-buffered saline (PBS) and incubated for 1 h at 26 °C
   in 500 μL of HBSS containing 100 μM of 7-azido-4-methylcoumarin (AzMC)
   fluorescent probe (802409, Sigma), which selectively reacts with H[2]S
   to form a fluorescent compound. Fluorescent AzMC signal acquisition
   (λEx = 365 nm and λEm = 450 nm) was performed on a modular multimode
   microplate reader (BioTek Synergy H1), and the fluorescence images were
   obtained using a Leica TCS SP8 microscope.

2.9. Intracellular iron assay

   The amount of iron (Fe^2+ and Fe^3+) was monitored using an iron assay
   kit (Dojindo Laboratories Co., Ltd., Kumamoto, Japan). Following a 24-h
   treatment, trophozoites (2 × 10^7) were washed twice with 1 × PBS, and
   1 mL of assay buffer was added to the kit. The suspension was
   ultrasonicated for 20 min and centrifuged at 16,000×g for 15 min. After
   following the manufacturer's instructions, the standard and samples
   (n = 3 for each group) were incubated at 37 °C for 1 h. Absorbance at
   OD 593 nm was measured using a modular multimode microplate reader
   (BioTek Synergy H1).

2.10. Identifying nitroxoline target homologues in A. castellanii proteome

   In the Research Collaboratory for Structural Bioinformatics (RCSB)
   Protein Data Bank (PDB), two entries with nitroxoline as the structural
   ligand were identified: 3AI8 (Gene: CTSB, UniProt ID: [61]P07858) and
   5Y1Y (Gene: BRD4, UniProt ID: [62]O60885) ([63]Burley et al., 2023;
   [64]UniProt Consortium, 2023). To identify proteins homologous to these
   two known nitroxoline targets in A. castellanii, the complete proteome
   dataset for A. castellanii (Proteome ID: UP000011083) was retrieved
   from the UniProt proteome database
   ([65]https://www.uniprot.org/proteomes/UP000011083). Sequence alignment
   of CTSB and BRD4 against the A. castellanii proteome was performed
   using the Needleman–Wunsch algorithm with a BLOSUM62 scoring matrix
   ([66]Needleman and Wunsch, 1970). Proteins with a sequence similarity
   ≥30% were considered potential nitroxoline targets within the A.
   castellanii proteome, with the highest similarity observed at 38%
   between proteins [67]P07858 and L8H210.

2.11. Homology modelling and molecular docking

   Homology modelling and molecular docking were conducted using
   Schrödinger Suite 2022–3 (New York, NY, USA). Proteins and ligands were
   prepared using the Protein Preparation Wizard and LigPrep modules in
   the Schrödinger Suite with default settings. A receptor grid was
   generated using the Receptor Grid Generation module in Schrödinger
   Suite, and the ligand diameter midpoint boxes were defined as
   10 × 10 × 10 Å^3. Molecular docking was performed using the Ligand
   Docking module in the Schrödinger Suite with SP precision, a flexible
   ligand sampling strategy, and a van der Waals radii scaling factor of
   0.5 for the receptor and ligand ([68]Friesner et al., 2006).
   Subsequently, docking poses underwent Prime MM-GBSA computation, where
   residues within 6 Å of the docked ligands were relaxed using the
   minimise sampling method. Finally, the binding free energies (MM-GBSA
   dG Bind) and chemotype diversity were calculated by visually inspecting
   the optimised docking pose and considering the SP docking scores.

2.12. Monodansylcadaverine (MDC) autophagy staining assay in trophozoites

   An MDC staining assay kit (Beyotime, C3018S, Nanjing, China) was used
   to detect autophagic vesicles in autophagic cells. All vacuoles in the
   trophozoites were visualised using MDC staining. Following the
   manufacturer's instructions, trophozoites (1 × 10^5) after a 12-h
   treatment were washed once with 1 × PBS and incubated for 30 min at
   26 °C in 1 mL of assay buffer containing 1 μL of MDC, the tracer of
   acidic autophagic vesicles. After three washes with assay buffer,
   fluorescent MDC signal acquisition (λEx = 335 nm and λEm = 512 nm) was
   performed on a modular multimode microplate reader (BioTek Synergy H1).

2.13. Statistical analysis

   All statistical analyses were performed using the GraphPad Prism 8
   software (GraphPad Software, Version 8.0; San Diego, CA, USA).
   Comparisons between the control and treatment groups were performed
   using a one-way analysis of variance. Data are presented as
   means ± standard deviation (SD), derived from ≥3 independent
   experiments for each sample. Statistical significance was set at
   p < 0.05.

3. Results

3.1. Cell viability inhibition and morphology changes in A. castellanii after
treatment with nitroxoline

   To assess the in vitro efficacy of nitroxoline, we used the
   CellTiter-Glo® Luminescent cell viability assay. The determined IC[50]
   value is presented in [69]Fig. 1A, indicating that nitroxoline
   effectively inhibited A. castellanii trophozoites at 24 h with an
   IC[50] ≤ 16.08±0.93 μM. [70]Fig. 1B depicts the time- and
   dose-dependent decreases in A. castellanii trophozoite viability in
   response to nitroxoline. Treatment with 20, 40, 60, and 80 μM of
   nitroxoline separately led to significant reductions in trophozoite
   viability to 32.14% ± 0.97%, 20.87% ± 0.22%, 17.29% ± 0.70%, and
   9.85% ± 0.93% after 24 h (p < 0.05; [71]Fig. 1B), respectively,
   compared with the 0.1% DMSO-treated group. Specifically, significant
   declines were observed at 20, 40, 60, and 80 μM of nitroxoline after
   48 h (6.35% ± 0.48%, 2.42% ± 0.27%, 1.51% ± 0.10%, and 1.15% ± 0.04%)
   and 72 h (1.77% ± 0.26%, 0.55% ± 0.21%, 0.41% ± 0.02%, and
   0.37% ± 0.06%) (p < 0.05; [72]Fig. 1B). Data are presented as mean ± SD
   from triplicate experiments.

Fig. 1.

   [73]Fig. 1
   [74]Open in a new tab

   Suppressing A. castellanii trophozoites viability by nitroxoline. (A)
   IC[50] values of nitroxoline-treated A. castellanii trophozoites
   determined by 24 h CellTiter-Glo assay. (B) Viability of trophozoites
   following 20–80 μM nitroxoline treatments for 24, 48, and 72 h
   determined using the CellTiter-Glo assay. Data are presented as
   means ± SDs from three experiments. Blank: without any treatments.
   DMSO-treated trophozoites (0.1%) were used as the negative control. (C)
   Effects on A. castellanii trophozoites incubated with negative control
   (0.1%DMSO) and different concentration of nitroxoline for 24 h
   monitored under an inverted microscope (magnification, × 40). Scale
   bar: 20 μm. The arrows indicate the solidified trophozoites treated
   with nitroxoline. (D) Trophozoites were treated with the indicated
   compound concentrations for 24 h. Apoptosis rate was measured using PE
   Annexin-V flow cytometry. (E) Mean fluorescence intensity of PE
   Annexin-V in each group. Data are presented as means ± SDs of three
   experiments. (F) Reactive oxygen species level was measured using
   DCFH-DA flow cytometry. (G) Mean flunorescence intensity of DCFH-DA in
   each group. Data are represented as means ± SDs of three separate
   experiments. SDs, Standard deviations.

   Observation using an inverted microscope showed that the shape of the
   trophozoites was severely disrupted, and the boundary was blurred with
   increasing drug concentrations after 24 h ([75]Fig. 1C). Trophozoites
   progressively became rounded and elliptical with spinous protrusions,
   and intracellular vacuoles disappeared and clustered together after 48
   and 72 h of treatment ([76]Fig. S1A). To investigate the effect of
   nitroxoline on the cyst stage, cysts were treated with 20 μM and 40 μM
   nitroxoline for 3 days and continued the observation for another 9
   days. Trophozoites emerged in the 0.1% DMSO-treated groups on the third
   day, and appeared in the 20 μM nitroxoline-treated groups 3 days after
   the replacement of fresh medium, while no trophozoites were observed in
   the 40 μM nitroxoline-treated groups during the 12-day observation
   period ([77]Fig. S1B).

   To investigate whether nitroxoline-induced inhibition of trophozoite
   viability was linked to apoptosis, we evaluated phosphatidylserine
   levels on apoptotic trophozoite membranes using Annexin V apoptosis
   detection. Flow cytometry analysis indicated that, compared with the
   control group, nitroxoline treatment significantly promoted cell
   apoptosis ([78]Fig. 1D to E). Excessive accumulation of ROS can break
   cellular homeostasis, leading to oxidative stress ([79]Liu et al.,
   2023). ROS determined using DCFH-DA increased in the
   nitroxoline-treated groups compared with the normal-cultured group
   ([80]Fig. 1F to G). Nitroxoline treatment significantly decreased the
   MMP, indicating mitochondrial dysfunction ([81]Fig. S2).

   Collectively, these in vitro experiments demonstrate the inhibitory
   effect of nitroxoline on A. castellanii trophozoite growth.

3.2. General property and GO term enrichment analysis of transcriptome
sequencing

   Three parallel samples, separately from normal-cultured, DMSO-treated,
   20 μM nitroxoline-treated, and 40 μM nitroxoline-treated groups after
   24 h, respectively, were successfully subjected to RNA-seq ([82]Fig.
   S3A). A principal component analysis plot further revealed distinctions
   among the four trophozoite groups ([83]Fig. S3B). The expression
   profiles of the normal culture and DMSO-treated groups showed minimal
   differences. The gene expression patterns from the 20 μM
   nitroxoline-treated and 40 μM nitroxoline-treated groups were discrete
   compared with the other groups ([84]Fig. S3C), suggesting high
   variability of the expression profiles in nitroxoline-treated groups
   and indicating that many of the key variations in the genes in the
   expression profiles probably determined the drug efficacy to
   trophozoites. Volcano plots were also shown in which the blue and
   purple regions represent upregulated and downregulated genes with
   significant changes in differential abundance (log2FC < −1 or
   log2FC > 1; p < 0.05 based on Student's t-test with false discovery
   rate correction) ([85]Figs. S3D–E).

   To categorise the DEGs, GO term enrichment analysis was performed,
   which included three fundamental groups: biological processes (BP),
   cellular components (CC), and molecular functions (MF). BP analysis in
   the 20 μM nitroxoline-treated group indicated that regulated DEGs were
   predominantly involved in the cell wall macromolecule catabolic process
   ([86]Fig. 2A). Meanwhile, the significantly regulated DEGs in the CC
   category ([87]Fig. 2A) were associated with components such as the
   intrinsic component of the membrane and integral component of the
   membrane. MF analysis ([88]Fig. 2A) of 20 μM nitroxoline-treated DEGs
   revealed enrichment in oxidoreductase activity, FMN binding, iron ion
   binding, triglyceride lipase activity, and hydroxyacyl glutathione
   hydrolase activity. Nevertheless, in the 40 μM nitroxoline-treated
   group, BP analysis ([89]Fig. 2B) revealed significant alterations in a
   multitude of genes related to cellular response to DNA damage stimulus,
   DNA repair, and double-strand break repair via homologous
   recombination. In the CC category ([90]Fig. 2B), components such as
   spindles, condensed chromosomes, outer membranes, mitochondrial
   envelope, and smc5-smc6 complex were distinctly clustered. In the MF
   analysis, DEGs were enriched in ATP hydrolysis activity, ATP-dependent
   activity, FMN binding, 3–5 DNA helicase activity, and DNA binding.

Fig. 2.

   [91]Fig. 2
   [92]Open in a new tab

   GO analysis to characterise functional pathways in A. castellanii
   trophozoites affected by nitroxoline. (A, B) Bar plot of top GO terms
   for significantly changed genes in 20 (A) and 40 μM nitroxoline-treated
   trophozoites (B). Red: biological process (BP). Yellow: cellular
   component (CC). Blue: molecular function (MF).

3.3. GO interaction network and KEGG pathway analysis

   As depicted in the GO enrichment network, iron ion binding (GO:0005506)
   and oxidoreductase activity (GO:0016491) were enriched in the 20 μM
   nitroxoline-treated group ([93]Fig. 3A), and cellular response to DNA
   damage stimulus (GO:0006974) was the top cluster of enrichment in the
   40 μM nitroxoline-treated group ([94]Fig. 3B). Pathway enrichment
   analysis based on KEGG terms revealed significant enrichment of
   regulated genes in hypertrophic cardiomyopathy; metabolic pathways;
   pyruvate metabolism; and fatty acid biosynthesis in the 20 μM
   nitroxoline-treated group ([95]Fig. 3C). Nonetheless, the genes in the
   40 μM nitroxoline-treated group were significantly enriched in
   homologous recombination, Fanconi anaemia pathway, base excision
   repair, biosynthesis of secondary metabolites, and non-homologous
   end-joining ([96]Fig. 3D). Pathway enrichment analysis based on the GO
   interaction network and KEGG terms revealed significant enrichment of
   altered genes in various metabolic and DNA damage response (DDR)
   pathways. This transcriptomic response discrepancy to nitroxoline
   suggests that trophozoites adopt diverse regulatory mechanisms to
   counter external stress.

Fig. 3.

   [97]Fig. 3
   [98]Open in a new tab

   Network diagrams and KEGG analysis in A. castellanii trophozoites
   affected by nitroxoline. (A, B) Network diagrams depict the top 20 Gene
   Ontology terms interactions, grouped by pathways. Line thickness
   corresponds to the number of shared genes between two pathways. The
   ratio of overlapping genes to unique genes is ≥ 20%. The colour and
   size of the circles indicate the value of significantly regulated genes
   among each cluster. (C, D) Bar plot of top enriched KEGG terms for
   significantly changed genes in 20 and 40 μM nitroxoline-treated
   trophozoites. Statistical significance and gene percentage (number of
   genes in transcriptomic data versus all genes annotated with the KEGG
   term) are shown. KEGG, Kyoto Encyclopedia of Genes and Genomes.

3.4. Nitroxoline alters the mRNA expression of trophozoites

   The volcano plot showed that several DNA damage repair genes were
   upregulated after 24 h of treatment with nitroxoline ([99]Fig. 4A to
   B), including TOP2, PARG, BRCA2, FEN1, and RAD51. In addition, several
   other DNA damage repair genes with significant upregulation following a
   40 μM nitroxoline treatment of trophozoites were also observed
   ([100]Fig. 4B), including ATR, ATM, RAD50, RAD51 homolog, RAD54, RAD52,
   APE1, HDAC1, PARP149970 (UniProt IDs: [101]L8HC40), PARP256450 (UniProt
   IDs: [102]L8GH34), PARP114300 (UniProt IDs: [103]L8H6I0), and
   PARP338640 (UniProt IDs: [104]L8HHX0). Upon further analysis of
   down-regulated genes following 40 μM nitroxoline treatment, the
   following genes were altered in trophozoites: SAHH, GNMT, CBS, Cys1a,
   and MAT. Genes (UniProt IDs: [105]L8GGB2, [106]L8GKJ9) belonging to the
   cytochrome p450 superfamily were upregulated in 20 μM and 40 μM
   nitroxoline treatments.

Fig. 4.

   [107]Fig. 4
   [108]Open in a new tab

   Transcript profiles in trophozoites treated with nitroxoline. (A, B)
   Enhanced volcano plot of the DEGs in trophozoites under 20 and 40 μM
   nitroxoline treatments, respectively. The metabolic pathways, DNA
   damage repair pathway and mitochondrial function related genes
   significant (|Fold change| ≥1.5, Pvalue ≤0.05) are annotated. (C) Heat
   map of DEGs in nitroxoline-treated groups compared with DMSO-treated
   and normal-cultured groups (Blank). Increased and decreased abundances,
   relative to the control, are shown in red and blue, respectively. (D)
   Relative mRNA expression of SAHH, GNMT, CBS, GSR, TRX, ATM, ATR, RAD51,
   FEN1, DMT1, ATG8, ATG9, ATG16, and ATG27 under 20 and 40 μM nitroxoline
   treatments for 24 h in A. castellanii trophozoites. Gene expression was
   normalised to 18S expression levels. Results represent means ± standard
   deviations of three independent experiments. (E) Representative H[2]S
   live-cell fluorescence images after treatment of nitroxoline.
   Trophozoites were pre-treated and stained with 100 μM AzMC for 1 h.
   Scale bars, 20 μm. (F) Endogenous levels of H[2]S were detected using
   AzMC fluorescence assays (λEx = 365 nm, λEm = 450 nm). (G) DNA damage
   was assessed using avidin-biotin assays (OD = 650 nm) (H, I) Iron
   amount (Fe^2+ and Fe^3+) was monitored using absorbance assays
   (OD = 593 nm). DEGs, Differentially expressed genes; OD, Optical
   density.

   Visualising the selected DEGs within these pathways using a heatmap
   ([109]Fig. 4C) revealed distinct gene expression patterns. Nitroxoline
   can positively or negatively affect metabolic enzymes. Treatment with
   40 μM nitroxoline impacted methionine and cysteine cycles by decreasing
   the expression of GNMT, SAHH, CBS, and Cys1a ([110]Fig. 4C) and
   increasing the expression of BHMT ([111]Figs. S4A–B) in A. castellanii.
   In addition to the rescue of cell proliferation upon cysteine
   deprivation, GNMT expression normalised cellular oxidative stress, and
   the transsulfuration pathway is required to sustain glutathione levels
   and redox balance in vivo ([112]Zhu et al., 2019). Therefore, the loss
   of GNMT and CBS may disturb cellular oxidative homeostasis.
   Nevertheless, nitroxoline significantly activated various kinases, such
   as ATM and ATR, which act as the primary transducers in the DDR
   signalling cascade ([113]Fig. 4C). Many of the downstream genes
   involved in DNA damage, such as PARP, PARG, MRE11, RAD50, RAD51, and
   APE1, exhibited high expression levels after 40 μM nitroxoline
   treatment ([114]Fig. 4C). As the major contributors to detecting DNA
   damage and maintaining genomic stability, HDAC1 showed increased
   expression in the 40 μM nitroxoline-treated group. Consequently,
   several physiologically important antioxidant enzymes of the
   peroxiredoxin, thioredoxin, and glutathione systems, such as PRX, PRX2,
   TRX, TrxR, GSR, and HYR1, were significantly upregulated in the 20 and
   40 μM nitroxoline-treated groups ([115]Fig. 4C). ATG8, ATG9, ATG16L2
   and ATG27, the autophagy-related genes, represented high mRNA
   expression levels after 40 μM nitroxoline treatment ([116]Fig. 4C).

   To further verify the consistency of the transcriptome results, RT-PCR
   was performed. The expression of the vital metabolic pathway genes
   SAHH, GNMT, and CBS exhibited significant downregulation after 20 and
   40 μM nitroxoline treatment. In contrast, a significant increase in the
   redox system genes GSR and TRX and the DNA damage repair-related genes
   ATM, ATR, RAD51, and FEN1 ([117]Fig. 4D) were revealed in the 20 and
   40 μM nitroxoline-treated groups. The expression of the other DNA
   repair relative genes APE1, DNA2, PARP, PARG, MRE11, and RAD50
   ([118]Figs. S4C–H) increased only after 40 μM nitroxoline treatment,
   consistent with the transcriptomics. DMT1, involved in iron
   acquisition, was also upregulated in the 20 and 40 μM
   nitroxoline-treated groups ([119]Fig. 4D). The autophagy-related genes
   ATG8, ATG9, and ATG16 were significantly upregulated after 24 h of
   exposure to 20 or 40 μM nitroxoline ([120]Fig. 4D), while ATG27
   remained relatively changed only after 40 μM nitroxoline treatment.
   Compared with the control group, only trophozoites treated with 40 μM
   nitroxoline showed more accumulation of autophagic vacuoles ([121]Fig.
   S4I), consistent with the expression of autophagy-related genes.
   Moreover, the number of apurinic/apyrimidinic sites in DNA increased in
   the nitroxoline-treated groups, indicating probable nitroxoline-induced
   DNA damage ([122]Fig. 4G).

3.5. Nitroxoline reduces the H[2]S production of trophozoites

   8-Hydroxyquinoline derivatives can decrease the activity of CBS
   ([123]Conan et al., 2022). CBS encodes a pyridoxal
   5′-phosphate-dependent enzyme that catalyses the condensation of
   homocysteine and serine to form cystathionine, which is a branch of the
   transsulfuration pathway. The transsulfuration pathway is the primary
   route for the biosynthesis of the major cellular antioxidants cysteine
   and glutathione, and cystathionine and its downstream partner CBS
   produce the gaseous transmitter H[2]S, another ROS scavenger, as a
   byproduct of their enzymatic activity ([124]Garcia et al., 2023). Since
   CBS is also involved in H[2]S production, we measured its levels using
   an AzMC fluorogenic probe ([125]Hu et al., 2019). To investigate the
   effects of nitroxoline on trophozoites, H[2]S fluorescence imaging was
   performed using an AzMC probe ([126]Fig. 4E). A 24-h treatment with
   20 μM or 40 μM of nitroxoline significantly decreased H[2]S production
   levels by 7.90% ± 2.82% and 44.54% ± 0.82%, respectively, compared with
   the control group ([127]Fig. 4F). Notably, nitroxoline reduced the
   level of endogenous H[2]S in a dose-dependent manner. Nitroxoline
   reduced H[2]S production in cellular lysates, revealing a possible
   effect on regulating CBS expression levels or protein stability, which
   is similar to the above results showing that the mRNA expression level
   of CBS was significantly decreased after nitroxoline treatment for 24 h
   ([128]Fig. 4D).

3.6. Nitroxoline impacts iron uptake transcripts in trophozoites

   A previous study postulated that nitroxoline induces rapid iron
   starvation in the biofilms of several important pathogens ([129]Liu et
   al., 2022). We were curious whether nitroxoline could affect iron
   uptake in A. castellanii trophozoites. Full transcriptomic analysis of
   nitroxoline-treated trophozoites revealed that the iron ion-binding
   pathways ([130]Fig. 3A) were significantly affected. Nitroxoline
   treatment increased the transcription of genes involved in iron
   acquisition, including divalent metal transporter 1 (DMT1) ([131]Fig.
   4C to D). Bioinformatic analysis showed that A. castellanii genomes
   encode one gene product with homology to Plasmodium falciparum PfDMT1,
   Arabidopsis thaliana NRAMP2, and Homo sapiens DMT1 (NRAMP2). The DMT1
   homolog was ACA1_225890 in A. castellanii (AcDMT1), which shared 58.31%
   amino acid sequence identity with A. thaliana NRAMP2, 56.67% with H.
   sapiens DMT1, and 28.38% with P. falciparum PfDMT1 ([132]Table S2).
   DMT1 is essential for maintaining iron homeostasis and enables Fe^2+
   and Mn^2+ ion entry into the mitochondria. Therefore, DMT1 promotes
   mitochondrial heme synthesis, iron-sulphur cluster biogenesis, and
   antioxidant defence ([133]Wolff et al., 2014). The intracellular iron
   levels in nitroxoline-treated trophozoites were also confirmed
   ([134]Fig. 4H to I). As depicted above, nitroxoline, especially at
   40 μM, significantly decreased intracellular Fe^2+ and total Fe amounts
   in trophozoites compared with the control group. These results
   establish a structural basis for the pharmacological intervention of
   nitroxoline on trophozoites.

3.7. Searching potential targets of nitroxoline in A. castellanii proteome

   By searching ligand ’Nitroxoline’ in the RCSB PDB database, two entries
   (PDB codes: [135]3AI8 and [136]5Y1Y) were found with nitroxoline as a
   structural ligand. The first entry was human Cathepsin B (gene: CTSB,
   UniProt ID: [137]P07858), which belongs to a family of lysosomal
   cysteine proteases known as cysteine cathepsins and plays an important
   role in intracellular proteolysis. The latter belongs to the human BET
   family of BRD4 (UniProt ID: [138]O60885), a chromatin reader protein
   that recognises and binds to acetylated histones and plays a key role
   in transmitting epigenetic memory across cell divisions and
   transcription regulation. To identify possible A. castellanii proteins
   that are homologous to the known targets of nitroxoline (Cathepsin B
   and BRD4), we first downloaded the complete proteomic data of A.
   castellanii from the UniProt Proteomes database with 14,939 entries.
   The Needleman–Wunsch algorithm was used for sequence alignment using
   the BLOSUM62 scoring matrix. Cathepsin B and BRD4 sequences were
   globally aligned with all sequences in the A. castellanii proteome,
   followed by calculating sequence similarity. Unlike the high seqnece
   similarity that exceeding 80–90% when comparing human proteins with
   those from closely related mammals, the overall sequence similarity
   between human and several A. castellanii proteins is markedly lower,
   with the highest similarity observed at 38% between proteins
   [139]P07858 and L8H210. A sequence similarity cutoff (≥30%) was applied
   to isolate 639 homology proteins in A. castellanii ([140]Table S3) that
   possibley as potential targets. In addition, by comparing the GO
   annotation data of 639 homologous proteins with the GO term enrichment
   analysis outcomes illustrated in [141]Fig. 2A–C, 69 A. castellanii
   proteins overlapped to potentially bind nitroxoline and associate with
   the DEGs caused by nitroxoline treatment. Homology modelling and
   molecular docking were performed for these 69 proteins using the
   Schrödinger Suite 2022–3, allowing for conformational flexibility of
   the targets and nitroxoline. As a positive control, we redocked the
   complex structure of cathepsin B with nitroxoline (PDB ID: [142]3AI8),
   obtaining a docking scoring of −4.30 and binding free energy (MMGBSA dG
   Bind) of −18.14 ([143]Fig. 5B, [144]Table 1). After analysing the
   docking scores and binding poses, 18 proteins were identified as
   potential A. castellanii targets of nitroxoline ([145]Fig. 5,
   [146]Table 1).

Fig. 5.

   [147]Fig. 5
   [148]Open in a new tab

   Protein similarity network and docking poses of nitroxoline in the
   potential A. castellanii targets. (A) Interaction network of
   nitroxoline with potential A. castellanii targets. (B–H) The surface
   representation and two-dimensional diagrams illustrate ligand–protein
   interactions within the predicted nitroxoline-binding pockets of
   potential targets, including positive control (B), DNA repair protein
   RAD51 (C), cytochrome p450 (D), cystathionine beta-synthase (E),
   cysteine synthase (F), adenosylhomocysteinase (G), and
   betainehomocysteine methyltransferase (H) The colour coding transitions
   from dodger blue, representing the most hydrophilic regions, to orange
   to red, indicating the most hydrophobic regions, within the surface
   representation. Key residues implicated in recognising nitroxoline are
   depicted as sticks. Polar interactions are denoted by black dashed
   lines. The amino acids of potential target residues are highlighted to
   indicate positive charge (blue), negative charge (red), polar (cyan),
   and hydrophobicity (green).

Table 1.

   List of the potential nitroxoline targets in A. castellanii that also
   associated with the Gene Ontology analysed functional pathways in A.
   castellanii trophozoites affected by nitroxoline.
   Group Targets Docking Score MMGBSA dG Bind (kcal/mol) Protein Names
   Proteins associated with DNA repair L8H3I3 −6.28 −50.86 DNA repair
   protein RAD51 homolog
   L8HJY0 −6.52 −25.29 DNA repair protein
   L8GFZ7 −4.82 −12.16 Tyrosyl-tRNA synthetase
   L8GVR2 −7.16 −1.33 Flap endonuclease
   L8GHK7 −5.12 0.28 Rad51l1 protein
   Cytochrome p450 family L8H8M1 −5.21 −39.31 Cytochrome P450
   monooxygenase, putative
   L8GGB2 −8.47 −20.10 Cytochrome p450 superfamily protein
   L8GTE4 −4.66 −6.48 Cytochrome P450, putative
   L8H0Q4 −6.48 8.02 Cytochrome p450 superfamily protein
   Cysteine synthase related proteins L8GF87 −8.08 −32.41 Cystathionine
   beta-synthase
   L8HEA9 −7.04 −25.69 Cysteine synthase
   Others L8GRC9 −6.25 −26.34 DNA (cytosine-5)-methyltransferas
   L8GJL0 −6.20 −16.38 NADH cytochrome b reductase, putative
   L8H2J2 −5.42 −15.02 Betainehomocysteine methyltransferase
   L8H6B5 −7.08 −10.49 Adenosylhomocysteinas
   L8H1B7 −5.48 −2.82 Methionine aminopeptidas
   L8H7N2 −5.74 −2.79 Uncharacterized protein
   L8HK34 −5.14 4.25 histone deacetylas
   Control [149]P07858 (3AI8) −4.30 −18.14 Cathepsin B
   [150]Open in a new tab

   The 18 potential targets were categorised into four groups. The first
   group included five proteins related to DNA repair that are crucial for
   protecting the genome from damage and mutations, thereby ensuring cell
   survival. The DNA repair protein RAD51 (UniProt ID: [151]L8H3I3), which
   contains the Rad51 domain, had a good docking score of −6.28 and the
   best binding free energy (MMGBSA dG Bind) of −50.86 kcal/mol,
   significantly better than that of the control. Nitroxoline was
   stabilised through multiple stacking interactions with F48 and Y218 and
   polar interactions with K41 and K243 ([152]Fig. 5C). The second group
   comprised four proteins belonging to the cytochrome p450 family, which
   play essential roles in detoxifying and eliminating various compounds,
   the synthesis and breakdown of hormones, cholesterol synthesis, and
   vitamin D metabolism. L8GGB2, which contains the P450 domain, had the
   best docking score of −8.47 and a binding free energy of
   −20.10 kcal/mol. Nitroxoline formed salt bridges with K75 and R417 and
   engaged in extensive hydrophobic interactions with the surrounding
   residues ([153]Fig. 5D). The third group comprised two proteins
   (UniProt IDs: [154]L8GF87 and [155]L8HEA9) associated with cysteine
   synthesis. By sharing similar binding pockets, they primarily interact
   with nitroxoline through polar interactions (via K8/K92, T146/T229, and
   S233/S322) and display high docking scores and binding free energies
   (−8.08 and −7.04; MMGBSA dG Bind: −32.41 and −25.69 kcal/mol,
   respectively) ([156]Fig. 5E to F). Additionally, seven discrete
   proteins were classified as others ([157]Fig. 5A–[158]Table 1). Among
   them, adenosylhomocysteinase (UniProt ID: [159]L8H6B5) and
   betainehomocysteine methyltransferase (UniProt ID: [160]L8H2J2) had
   relatively better docking scores with nitroxoline (−7.08 and −5.42),
   being stabilised through multiple stacking and polar interactions
   ([161]Fig. 5G to H). Collectively, these bioinformatic analyses and
   molecular docking studies support the possibility of direct
   interactions between nitroxoline and several A. castellanii proteins.

4. Discussion

   We elucidated the effects of nitroxoline on A. castellanii trophozoites
   in biological activities and molecular mechanisms, as illustrated in
   the schematic. Our assays against A. castellanii showed a IC50 value of
   16.08 ± 0.93 μM on acanthamoeba. slightly higher than that on Naegleria
   fawlgri trophozoites, about 1.63 ± 0.37 μM in a previous study. And we
   also investigated the effects of nitroxoline on A. castellanii cyst
   stage. Methods for evaluating RNA expression levels, such as
   microarrays and RNA-seq, have become valuable tools for assessing the
   dynamic properties of biological systems in a fast, broad, and reliable
   manner. In addition to technological aspects, developing bioinformatic
   analytical methods and tools has enormously aided our capacity to
   extract knowledge from omics technologies.

   Reactive oxygen and nitrogen species are unavoidable byproducts of
   metabolic and energy transfer processes in oxidative life. These
   molecules play pivotal roles in regulating cellular processes. However,
   excess amounts of ROS or reactive nitrogen species cause oxidative and
   nitrosative stress, which may damage macromolecules. Free radicals
   generated in cells under oxidative stress can directly attack cellular
   membranes. Nitroxoline increased intracellular ROS levels in a
   dose-dependent manner. Consistent with this idea, nitroxoline treatment
   significantly increased GPX and GSR expression, indicating strong
   intracellular oxidation modulation. Additionally, following the
   momentum of “big data” research, Trx, TrxR, and MsrA were strongly
   upregulated after exposure to nitroxoline. Furthermore, A. castellanii
   TrxR-S reduces Trx-1, which reduces Prx-1 and MsrA, suggesting that the
   synergistic interactions of these antioxidant enzymes are integral to
   the response to oxidative stress ([162]Leitsch et al., 2021). In a more
   recent study, mitochondrial fragmentation was accompanied by
   perturbations in energy production and elevated cellular ROS production
   ([163]Aoyagi et al., 2022). In our study, nitroxoline significantly
   induced mitochondrial depolarisation in a dose-dependent manner.

   Additional evidence for the role of nitroxoline in metabolic
   reprogramming comes from proteomic data in E. coli, highlighting the
   significant impact of this drug on metabolism ([164]Deschner et al.,
   2024). When it comes to how nitroxoline impacted metabolic pathways,
   such as the methionine and cysteine cycles, the important clues come
   from the volcano plot of DEGs, which show the upregulated expression of
   BHMT and MS and downregulated expression of GNMT, SAHH, CBS, and Cys1a.
   The action of either MS or BHMT occurs during the resynthesis of
   methionine using homocysteine (HCY). GNMT plays an important role in
   methionine clearance ([165]Johnson and Cuellar, 2023). The result of
   these changes is a potential increase in methionine synthesis.
   Methionine is a major target of ROS, and its metabolism signals a state
   of nutrient deprivation ([166]Ouyang et al., 2020; [167]Parkhitko et
   al., 2019). A myraid of studies have shown that excess methionine also
   induces the transcription of ribosomal and rRNA genes and overall
   translation capacity ([168]Laxman et al., 2013). Cysteine can be
   converted from o-succinyl-L-serine by Cys1a and is also a product of
   the transsulfuration pathway and precursor for the synthesis of
   tripeptide γ-glutamyl cysteinyl glycine (glutathione) ([169]Lampe et
   al., 2023; [170]Salazar et al., 2022). Cysteine, homocysteine,
   glutathione, and cysteinyl glycine were the most abundant sulfhydryl
   groups. All these molecules are important in radical interactions as
   they are particularly susceptible to oxidation by ROS
   ([171]Chondrogianni et al., 2014). Subsequently, we found intriguing
   alternatives that nitroxoline could impact the expression of CBS, which
   is the first rate-limiting step in the transsulfuration pathway
   ([172]Zhu et al., 2019). Additionally, we traced the effect of the
   reduction in CBS expression levels to its byproduct, H[2]S, and its
   role as a gaseous transmitter or ROS scavenger within the cell
   ([173]Parkhitko et al., 2019). Following this hypothesis, H[2]S
   fluorescence imaging and qualitative analysis depicted nitroxoline
   substantially decreased the amount of H[2]S in A. castellanii
   trophozoites. H[2]S protects the cells from oxidative stress and
   modulates neuronal transmission, smooth muscle relaxation, insulin
   release, and inflammatory responses. In mammalian cells, endogenous
   H[2]S is produced by three enzymes: CSE, CBS, and 3-MST ([174]Parkhitko
   et al., 2019). However, CSE and 3-MST are not annotated in the A.
   castellanii database. Since all these metabolic pathways and genomic
   integrity safeguards play crucial roles in maintaining cellular
   homeostasis and are critical for growth, they are acutely sensed and
   trigger signalling responses. Autophagy is required for maintaining
   cellular homeostasis. Autophagy occurs when trophozoites face drug
   stress and subsequently, experience a degree of amino acid starvation.
   Therefore, whether nitroxoline could contribute to autophagy is
   conceivable. Genetic analyses have identified that several
   autophagy-related genes (such as ATG8, ATG9, ATG16 and ATG27) that
   execute and regulate autophagy were significantly upregulated. ATGs are
   organised into functional complexes that participate in each step of
   macroautophagy. Upregulating autophagy genes have also been observed in
   flies lacking GNMT ([175]Tain et al., 2020). An imbalance in HCY,
   glutathione, and cysteine metabolism can modulate autophagy
   ([176]Desideri et al., 2012; [177]Paul et al., 2018; [178]Wang et al.,
   2019). In our study, nitroxoline significantly decreased the expression
   of GNMT and CBS, indicating a potential relationship between autophagy
   and metabolic pathways. Cellular oxidative stress is also key to
   autophagic response modulation ([179]Ouyang et al., 2020). Eventually,
   after prolonged exposure to nitroxoline, the trophozoites transitioned
   from autophagy to apoptosis.

   Nitroxoline can cause DNA damage in trophozoites at high
   concentrations, which is consistent with the upregulated mRNA
   expression of an arsenal of enzymes capable of sensing replication
   stress and transducing information to influence cellular responses and
   neutralise the destructive effects of aberrant DNA structures.
   Transcriptomic and qPCR analyses showed the upregulation of ATM, ATR,
   RAD51, RAD50, MRE11, and FEN1. The DNA damage repair is a complex
   signal transduction pathway (X. [180]Zhang et al., 2023). The first
   step in DSBs (DNA double-strand breaks) is recruiting repair factors at
   DNA break sites, which are initially recognised by the MRE11-RAD50-NBS1
   complex and promote ATM activation and DNA preparation for HR
   ([181]McCarthy-Leo et al., 2022; [182]Menolfi and Zha, 2020). RAD51, a
   key protein that corrects errors in replication forks, is a canonical
   marker of DNA damage ([183]Ciccia and Elledge, 2010; [184]Paulet et
   al., 2022). Volcano analysis also revealed several significant PARPs,
   including UniProt IDs: [185]L8HC40, [186]L8GH34, [187]L8H610, and
   [188]L8HHX0. PARP is a cellular stress sensor that is activated by
   oxidative, metabolic, and genotoxic stresses, such as single-strand
   break repair and double-strand DNA breaks, and in response, directs
   cells to specific fates according to the type and strength of the
   stress stimulus.

   Homology modelling and molecular docking were performed to identify the
   potential targets of nitroxoline. Four main groups were categorised:
   five proteins related to DNA repair, four belonging to the cytochrome
   p450 family, two associated genes with cysteine synthesis, and seven
   classified as others. The DNA repair protein RAD51 (UniProt IDs:
   [189]L8H3I3), which contains the Rad51 domain, had a good docking score
   of −6.28 and the best binding free energy (MMGBSA dG Bind) of
   −50.86 kcal/mol. Cys1a (UniProt IDs: [190]L8HEA9) and CBS (UniProt IDs:
   [191]L8GF87) shared similar binding pockets, primarily interacted with
   nitroxoline through polar interactions (via K8/K92, T146/T229, and
   S233/S322), and displayed high docking scores and binding free
   energies. SAHH (UniProt ID: [192]L8H6B5) and BHMT (UniProt ID:
   [193]L8H2J2) had relatively better docking scores with nitroxoline
   (−7.08 and −5.42), being stabilised through multiple stacking and polar
   interactions. Collectively, these transcriptomic analyses and molecular
   docking studies supported the possibility of direct interactions
   between nitroxoline and several A. castellanii proteins.

   Previous studies suggested that nitroxoline exerts its antibacterial
   effects by chelating metal ions such as Fe^2+, Mn^2+, Mg^2+, and Ca^2+
   ([194]He et al., 2024; [195]Pelletier et al., 1995). Iron and copper
   are transition metal ions that can cleave hydroperoxides to form
   radicals that initiate chain reactions ([196]Chondrogianni et al.,
   2014). Our findings suggest that the intracellular iron levels (Fe^2+
   and Fe^3+) of nitroxoline-treated trophozoites were significantly
   decreased, and the transcription of DMT1 (UniProt IDs: [197]L8H7N2),
   which is involved in iron acquisition, was increased. The A.
   castellanii genome encodes a gene with homology to Plasmodium
   falciparum PfDMT1, Arabidopsis thaliana NRAMP2, and Homo sapiens DMT1
   (NRAMP2). The DMT1 homolog is AcDMT1, which shares 58.31% amino acid
   sequence identity with A. thaliana NRAMP2, 56.67% with H. sapiens DMT1,
   and 28.38% with P. falciparum PfDMT1. The observed increase in the
   expression of intracellular ferrous iron transporters suggests a
   potential increase in intracellular recycling or the liberation of
   stored iron, aiding in sustaining growth within these comparatively
   brief timeframes ([198]Lampe et al., 2023). Furthermore, the regulatory
   mechanisms underlying the changes in iron metabolism remain unclear.
   Further studies are underway to test these hypotheses.

   Putting these nascent studies together, an early mechanistic picture
   began to coalesce that nitroxoline could impact the viability of A.
   castellanii trophozoites in a dose- and time-dependent manner, with
   increased metabolic process disorders, intracellular ROS production,
   DNA damage, iron starvation, decreased endogenous H[2]S, decreased
   mitochondrial outer membrane permeabilisation, and the possibility of
   direct interactions between nitroxoline and several A. castellanii
   proteins.

CRediT authorship contribution statement

   Lijun Chen: Writing – original draft, Software, Methodology,
   Investigation. Wei Han: Software, Methodology. Wenwen Jing: Writing –
   review & editing, Methodology. Meng Feng: Writing – review & editing,
   Methodology. Qingtong Zhou: Writing – review & editing, Software,
   Methodology, Funding acquisition, Conceptualization. Xunjia Cheng:
   Writing – review & editing, Funding acquisition, Conceptualization.

Data availability

   The original contributions of this study are included in the
   article/supplementary material section.

Ethics approval

   Not applicable.

Funding

   This work was supported by the National Natural Science Foundation of
   China 81572020 and 82372278 (XC), STI2030-Major Project 2021ZD0203400
   (QZ), and Hainan Provincial Major Science and Technology Project
   ZDKJ2021028 (QZ).

Declaration of competing interest

   The authors declared no conflict of interest.

Acknowledgments