(C) PLOS One [1]. This unaltered content originally appeared in journals.plosone.org.
Licensed under Creative Commons Attribution (CC BY) license.
url:https://journals.plos.org/plosone/s/licenses-and-copyright

------------



Differential contribution of two organelles of endosymbiotic origin to iron-sulfur cluster synthesis and overall fitness in Toxoplasma

['Sarah Pamukcu', 'Lphi', 'Univ Montpellier', 'Cnrs', 'Montpellier', 'Aude Cerutti', 'Yann Bordat', 'Sonia Hem', 'Bpmp', 'Inrae']

Date: 2021-12

Iron-sulfur (Fe-S) clusters are one of the most ancient and ubiquitous prosthetic groups, and they are required by a variety of proteins involved in important metabolic processes. Apicomplexan parasites have inherited different plastidic and mitochondrial Fe-S clusters biosynthesis pathways through endosymbiosis. We have investigated the relative contributions of these pathways to the fitness of Toxoplasma gondii, an apicomplexan parasite causing disease in humans, by generating specific mutants. Phenotypic analysis and quantitative proteomics allowed us to highlight notable differences in these mutants. Both Fe-S cluster synthesis pathways are necessary for optimal parasite growth in vitro, but their disruption leads to markedly different fates: impairment of the plastidic pathway leads to a loss of the organelle and to parasite death, while disruption of the mitochondrial pathway trigger differentiation into a stress resistance stage. This highlights that otherwise similar biochemical pathways hosted by different sub-cellular compartments can have very different contributions to the biology of the parasites, which is something to consider when exploring novel strategies for therapeutic intervention.

Toxoplasma gondii is a ubiquitous unicellular parasite that harbours two organelles of endosymbiotic origin: the mitochondrion, and a relict plastid named the apicoplast. Each one of these organelles contains its own machinery for synthesizing iron-sulfur clusters, which are important protein co-factors. In this study, we show that interfering with either the mitochondrial or the plastidic iron-sulfur cluster synthesizing machinery has a profound impact on parasite growth. However, while disrupting the plastidic pathway led to an irreversible loss of the organelle and subsequent death of the parasite, disrupting the mitochondrial pathway led to conversion of the parasites into a stress resistance form. We used a comparative quantitative proteomic analysis of the mutants, combined with experimental validation, to provide mechanistic clues into these different phenotypic outcomes. Although the consequences of disrupting each pathway were manifold, our data highlighted potential changes at the metabolic level. For instance, the plastidic iron-sulfur cluster synthesis pathway may be important for maintaining the lipid homeostasis of the parasites, while the mitochondrial pathway is likely crucial for maintaining their respiratory capacity. Interestingly, we have discovered that other mutants severely impacted for mitochondrial function, in particular the respiratory chain, are able to survive and initiate conversion to the stress resistance form. This illustrates a different capacity for T. gondii to adapt for survival in response to distinct metabolic dysregulations.

Like in plants and algae, apicoplast-containing Apicomplexa seem to harbour the three (ISC, SUF and CIA) Fe-S cluster synthesis pathways. Although the CIA pathway was recently shown to be important for Toxoplasma fitness [ 25 ], investigations in apicomplexan parasites have been so far mostly conducted in Plasmodium species and they essentially focused on the apicoplast-located SUF pathway [ 26 – 31 ]. The SUF pathway was shown to be essential for the viability of malaria parasites during both the erythrocytic and sexual stages of development, and has thus been recognized as a putative avenue for discovering new antiparasitic drug targets (reviewed in [ 32 ]). Contrarily to the ISC pathway, which is also present in the mammalian hosts of apicomplexan parasites, the SUF pathway may indeed yield interesting specificities that may be leveraged for therapeutic intervention. However, very little is known about Fe-S clusters synthesis in other apicomplexan parasites, including T. gondii. For instance, out of the four known metabolic pathways hosted by the apicoplast, Fe-S cluster synthesis was the only one remaining to be functionally investigated in T. gondii, while the others were all shown to be essential for the tachyzoite stage of the parasite (a fast replicating developmental stage responsible for the symptoms of the disease) [ 33 – 36 ]. Here, we present the characterization of two T. gondii mutants we generated to specifically impact the plastidic and mitochondrial SUF and ISC pathways, respectively. Our goal was to assess the relative contributions of these compartmentalized pathways to parasite development and fitness.

Fe-S clusters are simple and ubiquitous cofactors involved in a great variety of cellular processes. As their name implies, they are composed of iron and inorganic sulfur whose chemical properties confer key structural or electron transfer features to proteins in all kingdoms of life. They are important to the activities of numerous proteins that play essential roles to sustain fundamental life processes including, in addition to electron transfer and exchange, iron storage, protein folding, oxygen/nitrogen stress sensing, and gene regulation [ 17 ]. The synthesis of Fe-S clusters and their insertion into apoproteins requires complex machineries and several distinct pathways have been identified in bacteria for synthesizing these ancient cofactors [ 18 ]. They include the ISC (iron-sulfur cluster) pathway for general Fe–S cluster assembly [ 19 ], and the SUF (sulfur formation) pathway [ 20 ] that is potentially activated in oxidative stress conditions [ 21 ]. Eukaryotes have inherited machineries for synthesizing Fe-S cluster through their endosymbionts [ 22 ]. As a result, organisms with both mitochondria and plastids, like land plants, use the ISC pathway for assembling Fe-S clusters in the mitochondria and the SUF pathway for Fe-S clusters in the plastids [ 23 ]. Additional protein components that constitute a cytosolic Fe-S cluster assembly machinery (CIA) have also been identified: this pathway is important for the generation of cytosolic, but also of nuclear Fe-S proteins, and is highly dependent on the ISC mitochondrial pathway for providing a sulfur-containing precursor [ 24 ].

The phylum Apicomplexa comprises a large number of single-celled protozoan parasites responsible for serious disease in animals, including humans. For example, this phylum includes parasites of the genus Plasmodium that are responsible for the deadly malaria, and Toxoplasma gondii a ubiquitous parasite that can lead to a severe pathology in immunocompromised individuals. Apicomplexan parasites evolved from a photosynthetic ancestor and many of them still retain a plastid [ 5 , 6 ]. This plastid, named the apicoplast, originated from a secondary endosymbiotic event [ 7 , 8 ]. It has lost its photosynthetic properties as the ancestors of Apicomplexa switched to an intracellular parasitic lifestyle [ 9 ]. The apicoplast nevertheless still hosts four main metabolic pathways [ 10 , 11 ]: a so-called non-mevalonate pathway for the synthesis of isoprenoid precursors, a type II fatty acid synthesis pathway (FASII), part of the heme synthesis pathway, and a Fe-S cluster synthesis pathway. As the apicoplast is involved in these important biological processes for the parasite, and as they markedly differ from those of the host (because of their algal origin), that makes it a valuable potential drug target. Apicomplexan parasites also generally contain a single tubular mitochondrion, although its aspect may vary during parasite development [ 12 , 13 ]. The organelle is an important contributor to the parasite’s metabolic needs [ 14 ]. It classically hosts tricarboxylic acid (TCA) cycle reactions, which are the main source of electrons that feeds the mitochondrial electron transport chain (ETC) which generates a proton gradient used for ATP production. It also contains additional metabolic pathways, like a Fe-S cluster synthesis pathway and part of the heme synthesis pathway operating in collaboration with the apicoplast. The latter reflects obvious functional links between the organelles and potential metabolic interactions, which is also illustrated by their physical connection during parasite development [ 15 , 16 ].

Endosymbiotic events were crucial in the evolutionary timeline of eukaryotic cells. Mitochondria and plastids evolved from free-living prokaryotes that were taken up by early eukaryotic ancestors and transformed into permanent subcellular compartments that have become essential for harnessing energy or synthesizing essential metabolites in present-day eukaryotes [ 1 ]. As semiautonomous organelles, they contain a small genome, but during the course of evolution a considerable part of their genes have been transferred to the cell nucleus. Yet, they rely largely on nuclear factors for their maintenance and expression. Both organelles are involved in critically important biochemical processes. Mitochondria, which are found in most eukaryotic organisms, are mostly known as the powerhouses of the cell, owing to their ability to produce ATP through respiration. Importantly, they are also involved in several other metabolic pathways [ 2 ], including the synthesis of heme groups, steroids, amino acids, and iron-sulfur (Fe-S) clusters. Moreover, they have important cellular functions in regulating redox and calcium homeostasis. Similarly, plastids that are found in plants, algae and some other eukaryotic organisms, host a diverse array of pathways that contribute greatly to the cellular metabolism [ 3 ]. While often identified mainly as compartments where photosynthesis occurs, plastids host other important metabolic pathways. For example, they are involved in the assimilation of nitrogen and sulfur, as well as the synthesis of carbohydrates, amino acids, fatty acids and specific lipids, hormone precursors, and also Fe-S clusters. The best-characterized plastid is arguably the plant cell chloroplast, but not all plastids have photosynthetic function, and in land plants they are in fact a diverse group of organelles that share basal metabolic pathways, but also have specific physiological roles [ 4 ].

Results

Use of label-free quantitative proteomics to identify pathways affected by TgNFS2 or TgISU1 depletion There is a wide variety of eukaryotic cellular processes that are depending on Fe-S cluster proteins. To get an overview of the potential T. gondii Fe-S proteome, we used a computational tool able to predict metal-binding sites in protein sequences [56], and performed subsequent manual curation to refine the annotation. We identified 64 proteins encompassing various cellular functions or metabolic pathways that included, beyond the Fe-S synthesis machinery itself, several DNA and RNA polymerases, proteins involved in redox control and electron transfer, and radical S-adenosylmethionine (SAM) enzymes involved in methylation and methylthiolation (S2 Table). HyperLOPIT data or manual curation helped us assign a putative localization for these candidates. A considerable proportion (19%) of these were predicted to localize to the nucleus, where many eukaryotic Fe-S proteins are known to be involved in DNA replication and repair [57]. Yet, strikingly, most of the predicted Fe-S proteins likely localize to the endosymbiotic organelles. Several (19%) are predicted to be apicoplast-resident proteins, including radical SAM enzymes lipoate synthase (LipA) [58] and MiaB, a tRNA modification enzyme [59], as well as the IspG and IspH oxidoreductases of the non-mevalonate isoprenoid pathway [60]. Finally, for the most part (43%), candidate Fe-S proteins were predicted to be mitochondrial, with noticeably several important proteins of the respiratory chain (SDHB, the Fe-S subunit of the succinate dehydrogenase complex, Rieske protein and TgApiCox13) [61–63], but also enzymes involved in other metabolic pathways such as heme or molybdopterin synthesis. CRISPR/Cas9 fitness scores [39] confirmed many of these putative Fe-S proteins likely support essential functions for parasite growth. We sought to confirm these results experimentally. Thus, in order to uncover the pathways primarily affected by the depletion of TgISU1 and TgNFS2, and to identify potential Fe-S protein targets, we conducted global label-free quantitative proteomic analyses. Like most plastidic or mitochondrial proteins, candidate Fe-S acceptors residing in these organelles are nuclear-encoded and thus need to be imported after translation and have to be unfolded to reach the stroma of the organelle. This not only implies the addition of the Fe-S cofactor should happen locally in the organelle, but also that this may have a role in proper folding of these proteins. We thus assumed that disrupting a specific pathway may have a direct effect on the stability and expression levels of local Fe-S proteins. Cellular downstream pathways or functions may also be affected, while other pathways may be upregulated in compensation. Parasites were treated for two days with ATc (cKD TgISU1-HA) or three days (cKD TgNFS2-HA, as it takes slightly longer to be depleted, Fig 3A), prior to a global proteomic analysis comparing protein expression with the ATc-treated TATi ΔKu80 control. For each mutant, we selected candidates with a log2(fold change) ≤-0.55 or ≥0.55 (corresponding to a ~1.47-fold change in decreased or increased expression) and a p-value <0.05 (ANOVA, n = 4 biological replicates) (S3 and S4 Tables and Fig 6A and 6B). To get a more exhaustive overview of proteins whose amounts varied drastically, we completed this dataset by selecting some candidates that were consistently and specifically absent from the mutant cell lines or only expressed in these (S3 and S4 Tables). PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 6. Change in protein expression induced by TgNFS2 and TgISU1 depletion. Volcano plots showing the protein expression difference based on label-free quantitative proteomic data from TgNFS2 (A) and TgISU1 (B) mutants grown in the presence of ATc. X-axis shows log2 fold change versus the TATi ΔKu80 control grown in the same conditions, and the Y-axis shows -log10(p value) after ANOVA statistical test for n = 4 independent biological replicates. Selected variant protein categories are highlighted in color. C) Venn diagram representation of the shared and unique proteins whose expression is affected by the depletion of TgNFS2 and TgISU1. D) and E): mapping of less or more abundant candidates (circled dots) in the TgNFS2 and TgISU1 mutants, respectively, on the spatial proteome map representation of the hyperLOPIT data, highlighting clusters denoting putative subcellular localization. Full details available at: https://proteome.shinyapps.io/toxolopittzex/. https://doi.org/10.1371/journal.ppat.1010096.g006 Overall, depletion of TgISU1 led to a higher variability in protein expression and while the pattern of expression was essentially specific for the respective mutants, a number of shared variant proteins were found (Fig 6C and S5 Table). For instance, common lower expressed candidates include a SAM synthase, possibly reflecting a general perturbation of SAM biosynthesis upon loss of function of Fe-S-containing radical SAM enzymes [64]. Using dedicated expression data [65,66] available on ToxoDB.org we realized that, strikingly, many of the common variant proteins were stage-specific proteins (S5 Table). For instance, the protein whose expression went down the most is SAG-related sequence (SRS) 19F. The SRS family contains GPI-anchored surface antigens related to SAG1, the first characterized T. gondii surface antigen, and whose expression is largely stage-specific [67]. This protein, SRS19F, may be most highly expressed in stages present in the definitive host [66,68]. Conversely, SRS44, also known as CST1 and one of the earliest marker of stage conversion to bradyzoites [69], was upregulated in both mutants. Several other bradyzoite proteins whose expression increased included Ank1, a tetratricopeptide-repeat protein highly upregulated in the cyst-stages but not necessary for stage conversion [70], aspartyl protease ASP1, an α-galactosidase, as well as several dense granule proteins (GRA). Dense granules are specialized organelles that secrete GRA proteins that are known to participate in nutrient acquisition, immune evasion, and host cell-cycle manipulation. Many GRA have been characterized in the tachyzoite stage, but several are stage-specific and expressed in bradyzoites [71]. It should be noted that bradyzoite-specific proteins were generally much strongly expressed upon TgISU1 depletion than TgNFS2 depletion. Nevertheless, altogether these results show that altering either the plastidic or the mitochondrial Fe-S cluster synthesis pathway led to an initial activation of the expression of some markers of the bradyzoite stage, whose involvement in the stress-mediated response is well documented [54].

Depletion of TgNFS2 has an impact on the apicoplast, but also beyond the organelle We next focused on proteins that varied specifically upon depletion of TgNFS2 (S3 Table). Using the hyperLOPIT data available on ToxoDB.org, we assessed the putative localization of the candidates (Fig 6D and S7A Fig) and we also defined putative functional classes based on manual curation (Fig 6A and S7B Fig). Surprisingly, few apicoplast proteins were impacted. This could reflect a limited impact on apicoplast Fe-S apoproteins, but this is in contradiction with the late, yet pronounced, effect we see on the organelle in the absence of TgNFS2 (Fig 5A and 5B). There might also be a bias due to an overall low protein abundance: less than half of the apicoplast candidates of the predicted Fe-S proteome (S2 Table) were robustly detected even in the control for instance, including our target protein TgNFS2. Finally, of course it is possible that depletion of Fe-S clusters, while impacting the functionality of target proteins, did not have a considerable effect on their abundance. We sought to verify this for apicoplast stroma-localized LipA, a well-established and evolutionarily-conserved Fe-S cluster protein, which was found to be only marginally less expressed in our analysis (S3 Table). LipA is responsible for the lipoylation of a single apicoplast target protein, the E2 subunit of the pyruvate dehydrogenase (PDH) [33]. Using an anti-lipoic acid antibody on cKD TgNFS2-HA protein extracts, we could already see a marked decrease in lipoylated PDH-E2 after only one day of ATc incubation (Fig 7A). This was not due to a general demise of the apicoplast as it considerably earlier than the observed loss of the organelle (Fig 5A and 5B), and levels of the CPN60 apicoplast marker were clearly not as markedly impacted (Fig 7A). This is also unlikely to reflect a general decrease of TgPDH-E2 levels upon TgNFS2 knock-down, as our quantitative proteomics data, which was performed after 3 days of ATc incubation, show the same amount of unique peptides for TgPDH-E2 (32 ± 2.5 for the control and 32.25 ± 2.75 for the mutant, ~60% of sequence coverage). This finding suggests apicoplast Fe-S-dependent activities may be specifically affected in this mutant, which would happen before observing the general demise and loss of the organelle. Long term incubation of cKD TgNFS2-HA parasites with ATc and co-staining with apicoplast and inner membrane complex (IMC) markers, revealed general cell division defects, including organelle segregation problems and an abnormal membranous structures (Fig 7B). Overall, this suggests impacting the Fe-S cluster synthesis pathway in the apicoplast had important consequences beyond the organelle itself. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 7. TgNFS2 depletion impacts apicoplast-related pathways and has deleterious effects on parasite replication. A) A decrease in the lipoylation of the E2 subunit of pyruvate dehydrogenase (TgPDH-E2), which depends on the Fe-S-containing lipoyl synthase called LipA in the apicoplast, was observed by immunoblot using an anti-lipoic acid antibody on cell extracts from cKD TgNFS2-HA parasites kept with ATc for an increasing period of time. TgCPN60 was used as a control for apicoplast integrity. TgSAG1 was used as a loading control. Decrease of lipoylated TgPDH-E2 was quantified by band densitometry and normalized with the internal loading control. Data represented are mean ±SEM of n = 3 independent experiments. ** p ≤ 0.01, *** p ≤ 0.001 ANOVA comparison. B) cKD TgNFS2-HA parasites that were grown in the presence of ATc for 5 days were co-stained with anti-TgIMC3 (to outline parasites and internal buds) and anti-CPN60 (an apicoplast marker), which highlighted abnormal membrane structures and organelle segregation problems. Scale bar represents 5 μm. DNA was labelled with DAPI. DIC: differential interference contrast. https://doi.org/10.1371/journal.ppat.1010096.g007

Depletion of TgISU1 initiates conversion into bradyzoites Another feature highlighted by the quantitative proteomics analysis of the TgISU1 mutant is the change in the expression of stage-specific proteins (S4 Table). The expression of several bradyzoite-specific proteins including GRAs and proteins of the SRS family, was strongly increased. At the same time, some tachyzoite-specific SRS and GRA proteins were found to be less expressed. This was supporting the idea that intracellularly developing parasites lacking TgISU1 may convert into bona fide cyst-contained bradyzoites, as suggested by our initial morphological observations (Fig 5C). To verify this experimentally, we used a lectin from the plant Dolichos biflorus (DBL), which recognizes the SRS44/CST1 glycoprotein that is exported to the nascent wall of differentiating cysts [69]. We could see that during continuous growth of cKD TgISU1-HA parasites in the presence of ATc, there was an increasing number of DBL-positive structures (Fig 9A). This was quantified during the first 48 hours of intracellular development (Fig 9B) and, interestingly, was shown to mimic the differentiation induced by nitric oxide, a known factor of stage conversion [73], and a potent damaging agent of Fe-S clusters [74]. We combined RNAseq expression data for tachyzoite and bradyzoite stages [66] to establish a hierarchical clustering of the SRS proteins detected in our quantitative proteomics experiments for the two mutants (Fig 9C). Overall, this revealed an increase in the expression of bradyzoite-specific SRS in the TgISU1 mutant, although not all bradyzoite-specific SRS were strongly increased, perhaps reflecting an atypical or incomplete stage conversion. As mentioned earlier, some were also upregulated in the TgNFS2 mutant but in much lesser proportions. The strongest increase in bradyzoite-specific SRS expression upon TgNFS2 depletion was for SRS44/CST1, which happens to be the protein DBL preferentially binds to [69]. However, contrarily to the TgISU1 mutant, labelling experiments did not indicate any detectable increase in DBL recruitment in the TgNFS2 mutant (Fig 9B), confirming that impairing the plastidic Fe-S center synthesis pathway does not trigger full stage conversion. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 9. Depletion of TgISU1 triggers parasite differentiation. A) cKD TgISU1-HA parasites were grown in the presence of ATc and labelled with anti-TgIMC3 (to outline parasites) and a lectin of Dolicos biflorus (DBL) to specifically outline nascent cyst walls. Scale bar represents 10 μm. DNA was labelled with DAPI. DIC: differential interference contrast. B) Quantification of DBL-positive vacuoles after 24 hours or 48 hours of culture of 1) the cKD TgISU1-HA and cKD TgNFS2-HA mutants in the presence or absence of ATc 2) the Tati ΔKu80 cell line, as a negative control, 3) the Tati ΔKu80 cell line in the presence of 100μM nitric oxide (NO), as a positive control. Data are from n = 3 independent experiments. Values are mean ±SEM. * p ≤ 0.05, ** p ≤ 0.01, Student’s t-test C) Clustering of bradyzoite (Bz) or tachyzoite (Tz)-specific proteins of the SRS family shows specific enrichment of bradyzoite proteins upon TgISU1 depletion. D) The cKD TgISU1-HA mutant was grown for up to 14 days in the presence of ATc and co-stained with early cyst wall marker DBL together with tachyzoite marker SAG1, or intermediate (P18/SAG4), or late (P21) bradyzoite markers. Scale bar represents 10 μm. DNA was labelled with DAPI. E) Measurement of the cyst area size after growing the cKD TgISU1-HA mutant for 7 and 14 days in the presence of ATc and labelling the cyst wall with DBL and measuring the surface of 60 cysts per condition. Mean ±SD is represented. One representative experiment out of three independent biological replicates is shown. **** denotes p ≤ 0.0001, Student’s t-test. https://doi.org/10.1371/journal.ppat.1010096.g009 Stage conversion is a progressive process that happens over the course of several days, as it involves the expression of distinct transcriptomes and proteomes [54]. Markers for specific steps of in vitro cyst formation had been previously described [75], so we have used several of these to check the kinetics of stage conversion in the TgISU1-depleted parasites. We kept the cKD TgISU1-HA parasites for up to 14 days in the presence of ATc and tested for the presence of SAG1 (tachyzoite maker), DBL (early bradyzoite marker), P18/SAG4 (intermediate bradyzoite marker) and P21 (late bradyzoite marker) (Fig 9D and S11B Fig). After 7 days of ATc treatment, the DBL-positive cyst contained parasites that were still expressing SAG1 and not yet P18/SAG4, whereas after 14 days parasites with P18/SAG4 labelling were found, but there was still a residual SAG1 expression; expression of late marker P21 was, however, never detected. As controls, the TATi ΔKu80 parental cell line (derived from the RH type I strain) and the Prugniaud cystogenic type II strain were subjected to alkaline stress-induced stage conversion [76] for a similar duration (S11A and S11B Fig). While a majority of DBL-positive cysts of the type II strain expressed the P18/SAG4 and P21 markers after 14 days of differentiation, this was not the case for the type I TATi ΔKu80 cell line. Cysts containing TgISU1-depleted parasites thus displayed a somewhat intermediate phenotype with the expression of P18/SAG4, but not P21. This suggests stage conversion of these parasites progresses beyond the appearance of early cyst wall markers, but it seems incomplete. In fact, DBL-positive cysts showed a marked decrease in their mean size between the 7 and 14 days timepoints upon TgISU1 depletion (Fig 9D and S11C Fig). Smaller cyst size seemed to be a feature of the parental TATi ΔKu80 cell line when compared with the type II strain (S11C Fig), and noticeably these type I parasites also kept growing largely as tachyzoites and were able to reinitiate invasion cycles, leading to considerable host cell lysis during the course of the pH stress-induced conversion. Decrease in cyst size over time may thus reflect incomplete conversion, and be a consequence of later reactivation/reinvasion events; in the case of the TgISU1 mutant, the markedly smaller cyst size after 14 days of protein depletion (Fig 9E and S11C Fig) may also suggest a lack of fitness in the long term.

[END]

[1] Url: https://journals.plos.org/plospathogens/article?id=10.1371/journal.ppat.1010096

(C) Plos One. "Accelerating the publication of peer-reviewed science."
Licensed under Creative Commons Attribution (CC BY 4.0)
URL: https://creativecommons.org/licenses/by/4.0/


via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/

You are viewing proxied material from magical.fish. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.