Host-parasite oxidative arms race: who will win?

Avian haemosporidian parasites are globally widespread with a broad repertoire of hosts. When infected, the host can either reduce the parasite burden (resistance) and/or limit the severity of parasitaemia (tolerance). Oxidative stress is known to play a pivotal role in the host’s resistance and tolerance as well as its detrimental endpoints. The rationale behind this paradox lies in the dual role of reactive oxygen species (ROS): they can have both beneficial and detrimental effects for the host, while being largely harmful to the parasite. Thus, it is in the parasite’s interest to maintain a reduced environment within the host’s cell, whereas the host needs a fine-tuned balance between generating ROS to eliminate the parasites and maintaining sufficient antioxidant levels to protect its own tissues. This dynamic is what we refer to as the host-parasite oxidative arms-race. In this study, Eurasian siskins (Spinus spinus) were experimentally infected with Plasmodium ashfordi to investigate how the fundamental antioxidant system – the glutathione redox-system – responds to infection over time compared to control birds. By combining physiological measures and gene expression data of key glutathione related genes from both the parasite and the host at different time points, we provide evidence of this oxidative arms race. The gene expression data show that the parasite actively maintains reduced intracellular environment and eliminates ROS through high expression of superoxide dismutase (SOD), glutathione reductase (GR), and glutathione synthetase. In contrast, the host upregulates glutathione S-transferases (GSTs) and glutathione peroxidases (GPX), which reduce the physiologically active levels of the key antioxidant, glutathione. Although, the parasite seems to win the race in terms of the oxidative state of the cell, the marked decrease in parasitaemia from day 21 (45%) to 31 (15%) suggests that the host’s strategy by lowering the physiological glutathione levels is sufficient to defeat the parasite. Host-parasite oxidative arms race: who will win?

Avian haemosporidian parasites are globally widespread with a broad repertoire of hosts. When infected, the host can either reduce the parasite burden (resistance) and/or limit the severity of parasitaemia (tolerance). Oxidative stress is known to play a pivotal role in the host’s resistance and tolerance as well as its detrimental endpoints. The rationale behind this paradox lies in the dual role of reactive oxygen species (ROS): they can have both beneficial and detrimental effects for the host, while being largely harmful to the parasite. Thus, it is in the parasite’s interest to maintain a reduced environment within the host’s cell, whereas the host needs a fine-tuned balance between generating ROS to eliminate the parasites and maintaining sufficient antioxidant levels to protect its own tissues. This dynamic is what we refer to as thehost-parasite oxidative arms-race . In this study, Eurasian siskins (Spinus spinus ) were experimentally infected with Plasmodium ashfordi to investigate how the fundamental antioxidant system – the glutathione redox-system – responds to infection over time compared to control birds. By combining physiological measures and gene expression data of key glutathione related genes from both the parasite and the host at different time points, we provide evidence of this oxidative arms race . The gene expression data show that the 1 Posted on 22 Apr 2025 — The copyright holder is the author/funder. All rights reserved. No reuse without permission. — https://doi.org/10.22541/au.174536476.67136425/v1 — This is a preprint and has not been peer-reviewed. Data may be preliminary. parasite actively maintains reduced intracellular environment and eliminates ROS through high expression of superoxide dismutase (SOD), glutathione reductase (GR), and glutathione synthetase. In contrast, the host upregulates glutathione S-transferases (GSTs) and glutathione peroxidases (GPX), which reduce the physiologically active levels of the key antioxidant, glutathione. Although, the parasite seems to win the race in terms of the oxidative state of the cell, the marked decrease in parasitaemia from day 21 (45%) to 31 (15%) suggests that the host’s strategy by lowering the physiological glutathione levels is sufficient to defeat the parasite.

A vian malaria, glutathione, Eurasian siskins, Plasmodium ashfordi, infection, gene expression

Avian haemosporidian parasites of the generaPlasmodium, Haemoproteus , and Leucocytozoon are globally widespread, with a broad repertoire of avian hosts (Clark et al. 2014). The impact of avian haemosporidians on wild birds has been difficult to estimate, especially in areas with high prevalence where the hosts and parasites have co-evolved over a long time. However, long-term and experimental studies show evidence of negative effects on both survival and reproductive success even in high-prevalence areas (Opplinger et al. 1996, Knowles et al. 2009, 2010, Asghar et al. 2011). On naive hosts such as island species and captive birds in zoos, the impacts can be devastating (LaPointe et al. 2011). Upon infection, the host has two ways by which they can defend themselves against pathogens; decrease the parasite burden (i.e., resistance) and/or limit the severity of infection (i.e., tolerance) (Read et al. 2008, Schneider and Ayres 2008, R˚ aberg et al. 2009). One of the mechanisms known to play a pivotal role in the host’s resistance and tolerance and paradoxically also in its fatal endpoints - is oxidative stress (Becker et al. 2004). The rationale for this paradox has to do with the positive and negative effects of reactive oxygen species (ROS) for the host, as well as the negative impacts of ROS for the parasite. Thus, from the host’s perspective, a fine-tuned balance between ROS generation and detoxification by antioxidants is needed. The redox-state of infected cells is regulated by both the host and the parasite, causing what we refer to as host-parasite oxidative arms-race . The detailed logic for this arms-race is outlined below, but first, the consequences of having a high oxidative stress level, i.e. high ROS generation, need to be clarified. ROS, such as superoxide (O2-) and hydrogen peroxide (H2O2), are pro-oxidants and part of all aerobic organisms’ life. In moderate concentrations, ROS are crucial for cellular functioning by acting as signalling substances and regulating gene expression of, for example, enzymes and proteins in the MAP kinases and NF- κ B families, which are both involved in stress response and immune defence (Becker et al. 2004). However, too high concentrations of ROS are toxic and cause oxidative damage to DNA, lipids, and proteins, which can accumulate in the cell and increase the rate of cellular senescence or cause immediate cellular death, i.e. apoptosis (Halliwell and Gutteridge 2002). To protect the cell from too high levels of ROS – oxidative stress – organisms have evolved a large repertoire of antioxidants, including, for example, glutathione (GSH), thioredoxin, superoxide dismutase (SOD), glutathione-S-transferase (GST) and catalase, which scavenge, quench and detoxifies ROS. During parasite growth and multiplication, ROS are produced as by-products of parasite metabolism. For ex- ample, when blood parasites digest haemoglobin to obtain amino acids, it releases toxic free haem (FP-FeIII) and O2- which immediately are dismutated by SOD to H2O2 and O2 (see Fig. 1). Detoxification of FP-FeIII is via FP-specific proteins or by degradation, while detoxification of H 2O2 is dependent on the antioxidant GSH, which is the most abundant antioxidant within erythrocytes (Becker et al. 2004). Interestingly, the parasitic generated pro-oxidants are toxic for both the host and the parasite, and depending on which phase of the infection and the virulence of the parasite, the host’s immune defence can further increase the cellular ROS via a process called oxidative burst (or respiratory burst). During oxidative burst, NADPH oxidases in host’s phagocytes produce and release O2- which is followed by a cascade of free radical generation which aims to attack the pathogen ( Halliwell and Gutteridge 2002 ). Consequently, the immune system uses the toxic effects of ROS as a central mechanism for resistance. But for the oxidative burst to be 2 Posted on 22 Apr 2025 — The copyright holder is the author/funder. All rights reserved. No reuse without permission. — https://doi.org/10.22541/au.174536476.67136425/v1 — This is a preprint and has not been peer-reviewed. Data may be preliminary. efficient it needs to overwhelm its own as well as the parasites’ antioxidant system, which is likely to increase the host’s cellular oxidative damage (Perc´ ario et al. 2012). Tolerance, on the other hand, is (potentially) less costly to the host (if the host survives) and harmless to the parasite (Read et al. 2008). The unavoidable increase of ROS by the parasite is in this scenario detoxified by a joint effort by the host’s and the para- sites’ antioxidant system, particularly by the GSH redox-system, avoiding generation and accumulation of oxidative damages. In the present study, Eurasian siskins (Spinus spinus ) were experimentally infected withPlasmodium ashfordi to investigate its effect on the glutathione redox-system of red blood cells at different stages of infection and what consequences this has on cellular health in terms of oxidative state. In addition, the birds and parasites of this experiment have previously been used for full genome transcriptomics (see Videvall et al. 2015, 2017). Here, we have extracted the expression of genes related to the cellular redox pathway from both the hosts and parasites to include these results in a new framework and in relation to the host-parasite oxidative arms-race . More specifically, at the physiological level we measured 1) the concentration of total glutathione (tGSH, i.e., the sum of the active form (GSH) and the oxidized form (GSSG)) and 2) GSSG (see Fig. 1). The antioxidant function of GSH is catalyzed by the enzyme glutathione peroxidase (GPX), then GSH can donate electrons to H2O2, resulting in the end-products of water and GSSG (Sies 1986). The ratio between the reduced and the oxidized form of glutathione (GSH/GSSG) is a good indicator of cellular redox balance and is commonly used as an indicator of cellular health (Halliwell and Gutteridge 2002, Isaksson et al. 2005) – the lower ratio, the poorer health and the higher oxidative stress. During infection, the proportion of GSSG to GSH can be chronically high (low quotient of the ratio) to increase the pro- oxidative environment of the infected cells. However, this could be very damaging for the host. Alternatively, the ratio is low in the initial phase of infection and then GSSG is reduced back to GSH with the help of NAPDPH and glutathione reductase (GR) to increase the cell’s antioxidant capacity again (Griffith 1999). In humans, malaria infected erythrocytes show more than a 10-fold decrease in GSH/GSSG ratio (e.g.,Atamna and Ginsburg 1997 ), suggesting that the cell is exposed to increased ROS levels caused by oxidative burst by immune cells or/and by the parasite itself by digesting the oxygen carrying haemoglobin. The decline in GSH in response to P. falciparum and P. vivax infection is an overall trend in humans, as shown by a meta-analysis using over 2000 records (Kotepui et al. 2023). To date, there has been only one study on avian malaria and the GSH metabolism (Isaksson et al. 2013). This study was conducted on wild great tits ( Parus major ) during spring, and they concluded that GSH redox system forms a part of the immune system against avian malaria, but that the response depends on Plasmodium species and the host’s age and sex. However, independent of Plasmodium species, parasitaemia was positively associated with oxidative damage, suggesting an increased ROS challenge and cellular cost for the host during the peak of infection. As outlined above, we can predict different responses of both the physiological and molecular expressions in glutathione and its associated enzymes depending on whether birds follow the resistance or the tolerance strategy and depending on time, e.g. if it is during the peak of infection, prior to or after the peak. If the birds show resistance, they consequently produce high levels of internal ROS to defeat infection, thus we predict that the birds aim to maintain a high oxidative stress within the cell through the different components in the glutathione system (see Fig.1). This can be done by reducing the overall antioxidant capacity by reducing tGSH and GSH, via conjugation and efflux of GSH catalysed by the enzyme GST or by increased GPX activity to increase the relative GSSG level to GSH. Alternatively, if the birds show tolerance, the opposite is predicted, i.e. less ROS is produced by the host, hence a high tGSH and GSH/GSSG ratio can be maintained via active reduction of GSSG by GR to its active GSH form (see Fig.1). This is in line with parasitic preferences, since it relies on having a low intracellular oxidative stress to survive.

Experimental infections and parasitemia 3 Posted on 22 Apr 2025 — The copyright holder is the author/funder. All rights reserved. No reuse without permission. — https://doi.org/10.22541/au.174536476.67136425/v1 — This is a preprint and has not been peer-reviewed. Data may be preliminary. The experimental study was conducted in 2012 at the Biological Station Rybachy, located in the Kaliningrad region, Russia. The experimental procedures are described in Videvall et al. (2015). Briefly, juvenile siskins were captured at the study site and used as recipients in the experiment. All birds were initially tested for malaria parasites using microscopic blood smear analysis (following the methodology of Valki¯ unas et al. 2008) and polymerase chain reaction (PCR) techniques (as outlined in Hellgren et al. 2004) to confirm the absence of haemosporidian parasites prior to infection. For the experimental infection, Plasmodium ashfordi (mitochondrial cytochrome b gene lineage GRW2), originally obtained from a naturally infected common cuckoo (Cuculus canorus ), was used. The parasite was subsequently propagated by infecting crossbills, which served as the donor, before being used to inoculate the experimental siskins, following the methodology described by Palinauskas et al. (2008). Twenty-two siskins were inoculated with P. ashfordi infected blood solution into pectoral muscles and another 22 siskins were used as controls, which received the same amount of non-infected blood. Birds were kept for 31 days post inoculation (pi) and were monitored on days 0, 11, 21, and 31. A small volume of blood was drawn from the brachial vein to monitor the development of parasitemia. Blood samples for microscopic (Valki¯ unas et al. 2008) and molecular (Hellgren et al. 2004) analysis were collected from all recipients. Simultaneously, a fraction of the blood was snap-frozen in liquid nitrogen and stored in -80 °C until physiological and molecular analyses were performed. For a subset of the birds, blood was collected for host and parasite RNA extraction (for methods see Videvall et al. 2015). Total and oxidized glutathione (tGSH and GSSG) The protocol used follows Baker et al. (1990), adapted for a microplate reader (Isaksson et al. 2005). Four microliters of whole blood were diluted with 16 μL 5% 5-sulfoasalicylic acid (SSA), then centrifuged at 10000rpm for 10 min at 4 °C. 10 μL of the sample was further diluted with 400 μL GSH buffer (143 mM NaH2PO4 and 6.3 mM EDTA, pH 7.4). For GSSG assay, 1 μl of 4-vinylpyridine (4-VNP) was added to 200 μl of the sample and incubated for 1 hour at room temperature during shaking. After incubation, centrifuge at 12000 rpm for 5min at 4 °C. Then the reaction mixture was prepared containing GSH buffer, 5.5’-dithio- bis(2-nitrobenzoic acid) (DTNB) and NADPH. A 96-well plate was prepared for both tGSH (diluted plasma without 4-VNP) and GSSG (diluted plasma with 4-VNP) by adding, 20 μl sample and 200 μl reaction mix, and then add 5 μl of 0.34 U/sample and 0.17 U/sample glutathione reductase (GR) for tGSH and GSSG, respectively. All samples and standards were run in duplicates and the repeated samples from one individual were always measured on the same plate. The absorbance was immediately measured on FLUOstar OMEGA (BMG LABTECH) plate reader at 412 nm for 5 minutes (new read every 30 s). The kinetics were compared with daily made GSH and GSSG standards of known concentrations and the within-assay repeatability was 89% and 88%, respectively. Only birds with both tGSH and GSSG measurement were included in the statistical analyses. The sample sizes with both tGSH and GSSG measurements for infected birds were: nday0 = 18, n day11= 17, n day21 = 18, n day31 = 15. For controls, the sample sizes were: n day0 = 21, n day11 =20, n day21 = 18, n day31 = 20. The different sample sizes across time were either because of too small blood sample to perform both assays, or the assay for either tGSH or GSSG failed (e.g., negative absorbance values). Gene expression analyses During the same experiments, four infected birds were selected for blood sampling in order to conduct transcriptome sequencing of both the host and the parasite’s gene expression profiles. The birds were sampled before being infected to establish a baseline in gene expression and then at day 21 and day 31, i.e., during peak parasitaemia and during decreasing parasitemia levels, respectively. Both these analyses have been published elsewhere; for methods and results related to the host see Videvall et al. (2015), and for the parasite, see Videvall et al. (2017). To evaluate if genes linked to the cellular redox environment were differentially expressed in the host, we searched for the relevant genes list in supplementary material S1 and S2 in Videvall et al. (2015) and evaluated the log2-fold change and the adjusted P-values (q-values) to assess whether the genes where significantly upregulated or down regulated during the different stages of the infections. This is summarized here in Supplementary Table 1. For the parasite, no controls can be used 4 Posted on 22 Apr 2025 — The copyright holder is the author/funder. All rights reserved. No reuse without permission. — https://doi.org/10.22541/au.174536476.67136425/v1 — This is a preprint and has not been peer-reviewed. Data may be preliminary. to establish the baseline expression as the parasite is always in an “infection stage”. Instead, we searched if the genes involved in the cellular redox pathways were found to be expressed in the parasite during the different infection stages. If the genes were found to be expressed, we extracted the associated normalized read counts and compared them against the distribution of normalized read counts for all the expressed genes in the parasite within the individual bird at that specific timepoint. In such way, we were able to evaluate if the genes were in the upper range of expressed genes, as suggested if the genes were upregulated or just following a “housekeeping” gene expression. Expressed genes and normalized read counts from the parasite were extracted from the supplementary tables S1 and S2 found in Videvall et al. (2017) and summarized here in Supplementary Table 2. Data analysis All statistical models were run in JASP 0.18.3 (Intel) software (Love et al. 2019). Four measures of GSSG were considered outliers (i.e., values were above a biologically reasonable level, see Supplementary Figure 1 for Q-Q plots), hence four complete individuals were removed from the physiological models. Since GSH can oxidate to GSSG during the sample preparation, the four high values are most likely a result of human influences during either blood sampling or lab work. The outliers were from different individuals and at different time points, hence there were no consistency in these outliers. However, none of the models including the outliers changed the overall results and conclusions. Despite removal of the outliers the GSSG data showed negative skewness and GSH/GSSG data showed positive skewness. Hence, to fit normal distribution and model assumptions, raw data was transformed using log 10 and square root, respectively. Linear mixed models (LMM) were run when all time points were included in the model, controlling for individuals as a random factor in the model. This includes the model with parasitaemia and models for tGSH, GSSG and GSH/GSSG ratio across all time points. Parasitaemia was only tested against days (0, 11, 21 and 31) among the infected birds. The LMMs of the physiological models included parasitaemia. These models were only done in the infected group. Body mass did not influence the physiology, hence not included in the models or the results. In addition, ANCOVAs were performed to test the difference between controls and infected across time in the physiological biomarkers. Post hoc ANOVAs were then done for GSSG and GSH/GSSG ratio at the peak of parasitaemia and the last time point between controls and infected birds.

Physiology of tGSH and GSSG in relation to infection At the group level, parasitaemia peaked on day 21 (44.96 % ± 3.11, mean ± SE) and subsequently declined by day 31 (15.13 % ± 3.39, mean ± SE, χ2 = 57.68, p < 0.001, see Fig. 2). Parasitaemia levels were low on day 11 post-injection (0.49% ± 3.19, mean ± SE). Among the infected birds, there was no linear relationship between tGSH (F 1,21.67 = 0.138, p = 0.714), GSSG (F 1,20.01 = 2.539, p = 0.127) and GSH/GSSG (F 1,42.20 = 1.326, p = 0.256) and parasitaemia. Across time there was a significant difference between controls and infected birds in GSSG (treatment × day: F3,136 = 3.880, p = 0.011, Fig. 3 and Table 1 for full models). There was no significant interaction for either tGSH or GSH/GSSG (see Table 1). Although the effect was weak, infected birds had an overall higher GSH/GSSG ratio compared to controls (F 1,135 = 3.867, p = 0.051, Tab. 1). However, since parasitaemia peaked on day 21 and declined by day 31, we examined differences between infected and control groups at these later stages. At the peak of infection, there was no significant difference in tGSH between controls and infected birds (F 1,33 = 0.237, p = 0.630, Fig. 3a). The GSSG levels showed a non-significant trend toward being lower in infected birds compared to controls (F 1,33 = 3.751, p = 0.061, Fig. 3b). However, the ratio was significantly higher in infected birds compared to the controls (mean ± SE, infected (n =17): 13.44 ± 1.94 and control (n = 18): 8.41 ± 1.07; F1,33 = 4.603, p = 0.039, Fig. 3c). On day 31, both tGSH and GSSG levels were significantly lower in infected birds compared to controls. For tGSH (mean ± SE), infected birds (n = 13) had 10.50 μM ± 0.948, while controls (n = 20) had 15.00 μM ± 1.08 (F1,31 = 8.500, p = 0.007). Similarly, GSSG levels (mean ± SE) were lower in infected birds (0.96 μM ± 0.20) than in controls (1.607 μM ± 0.136; F 1,31 = 10.672, p = 0.003). This corresponds to a 30% and 40% reduction in tGSH and GSSG, respectively, in infected birds. Since the decline in GSSG was larger than that 5 Posted on 22 Apr 2025 — The copyright holder is the author/funder. All rights reserved. No reuse without permission. — https://doi.org/10.22541/au.174536476.67136425/v1 — This is a preprint and has not been peer-reviewed. Data may be preliminary. of tGSH on day 31, the tGSH: GSSG ratio was significantly higher in infected birds compared to controls (mean ± SE, infected (n = 13): 15.18 μM ± 2.52; and control (n = 20): 9.39 μM ± 0.999; F 1,31 = 4.906, p = 0.034, Fig. 3c). Redox related gene expression of host and parasite The host: Eurasian siskins Among the top genes that were differentially expressed during the peak of parasitaemia compared to before infection were six GST genes (MGST1, GST, MGST3, GSTK1, GSTZ1, and one without a gene name), which were all highly upregulated (see Supplementary Table 1, Fig. 1, and Videvall et al. 2015). For example, GST was expressed 4.8 times higher during the infection compared to when birds where uninfected. In addition, one gene for GPX was upregulated and was 2.3 times higher in infected birds compared to uninfected. Host GR and SOD were not among the top genes of being differentially expressed. After the peak of parasitaemia, 5 of the 6 GST genes, as well asGPX gene, remained significantly higher than before the infection (Supplementary Table 1). In fact, when comparing mean expression levels for those genes between the peak of parasitaemia and after the peak, all genes were relatively similar in expression, only MGST1 had a notable increase (1.3 times) in expression after the peak infection. The parasite: Plasmodium ashfordi Compared to the host, the parasite gene expression was less focused on GST related genes, instead the P. ashfordi showed an overall upregulation of enzymes related to an activation of the whole glutathione redox system (see Fig. 1). For example, during the peak of parasitaemia the parasite had two highly expressed genes for glutathione synthetase (Supplementary Table 2, Videvall et al. 2017), which catalyses the reaction to form new GSH by binding gamma-glutamylcysteine and glycine. Regarding GST, only one gene was also highly expressed in the parasite compared to the 5 and 6 genes in the siskins. In contrast to the host, the parasite showed very high relative expression of three SOD genes and one GR gene, often within the top 20% of expression among all the expressed genes (see Supplementary Table 2). Finally, one more glutathione related gene (lactoylglutathione lyase) was among the top expressed genes in the parasite during both time points. All these extracted cellular redox genes in the parasite are important for maintaining a low level of oxidative stress within the cell.

In the present study Eurasian siskins were infected with P. ashfordi to investigate the response of the glutathione redox system across time of both the host and the parasite. We measured the physiological levels of the key intracellular antioxidant glutathione and its oxidized form. In addition, gene expression data from both the host and the parasite, generated from the same experiment, were sourced from previously published studies (Videvall et al. 2015, 2017). The selected and extracted genes were those related to the glutathione redox cycle, thereby placed in a new context. These two different approaches provide strong support for a “ host-parasite oxidative arms race ”, with the parasite aiming to maintain a healthy, reduced cellular environment, while the host attempts to elevate oxidative stress by diminishing intracellular antioxidant capacity through GSH conjugation, thereby lowering overall GSH levels in red blood cells. The growth rate of P. ashfordi was very slow in the beginning with only 0.49% parasitaemia on day 11. However, between days 11 and 21, the parasites exhibited a rapid increase in growth, with parasitaemia peaking at 45%. After day 21, parasitaemia declined to 15% by day 31. This timeline for parasitaemia is important to keep in mind for the subsequent discussion on the differing oxidative strategies employed by the host and the parasite. The glutathione redox system is known to play a pivotal role in maintaining a reducing environment in the cytosol of any cell (Sies 1986). This is particularly important for the parasite. The human malaria, Plasmodi- um falciparum is known to show intense glutathione activity and metabolism (Becker et al. 2005). However, much less is known about glutathione dynamics in avian malaria species. For the host, there is a trade-off between maintaining a healthy reducing environment – thereby tolerating the parasite – and increasing the 6 Posted on 22 Apr 2025 — The copyright holder is the author/funder. All rights reserved. No reuse without permission. — https://doi.org/10.22541/au.174536476.67136425/v1 — This is a preprint and has not been peer-reviewed. Data may be preliminary. cellular oxidative stress to defeat the infection, which carries the risk of self-inflicted cellular damage and potential apoptosis. Oxidative burst serves as the first line of defence for the host to defeat parasites. This mechanism is particularly important during malaria infection, as the parasite “hides” within red blood cells and thus evades detection by the acquired immune system. Both phagocytic and non-phagocytic host cells release the reactive oxygen species (ROS), superoxide (O 2-) and hydrogen peroxide (H 2O2) to bombard the parasite. However, also the malaria parasite generates hydrogen peroxide during haemoglobin degradation (Becker et al. 2004). The superoxide can be converted to hydrogen peroxide by the parasite (and the host) if they upregulate their enzyme SOD and then hydrogen peroxide can be converted into water by GPX catalysed reaction with GSH (see Fig. 1). Indeed, P.ashfordi have a very high expression of SOD in relation to their overall gene expression levels, whereas the avian host has no significant upregulation during the peak of infection (see also Videvall et al. 2015, 2017). The GPX expression is also relatively high in the parasite; however, the host displays more than a two-fold increase in expression during the peak of parasitaemia, compared to the mean expression of the controls. This is likely a response to the generation of H 2O2 during parasitic haemoglobin degradation, as well as the host’s immune-induced release of ROS. Following the action of GPX, the utilized GSH is converted in its oxidized form, GSSG. Although the interaction between control and infected birds across the time series was not statistically significant, there was a significant reduction in both GSSG and tGSH levels in infected birds after the peak of parasitaemia compared to controls (Fig. 3a & b). The reduction in overall glutathione can be driven by the host through the strong upregulation of the GST enzyme, observed both at the peak of parasitaemia and on day 31. This enzyme conjugates GSH to reactive molecules such as lipid hydroperoxides – oxidative damage products generated by ROS. The resulting conjugates are then excreted, leading to an efflux of GSH from the cell and, ultimately, from the body. Indeed, in a study of wild great tits the level of hydroperoxides increased with Plasmodium parasitaemia and similar to the present study there were no association between tGSH or the GSH/GSSG ratio with parasitaemia (Isaksson et al. 2013). Unfortunately, the plasma volumes were too small to measure hydroperoxides in the present study. The increase of oxidative stress and damage is a general pattern across most studied Plasmodium species e.g., P. falciparum, P. vivax, P. knowlesi, P. berghei (Vasquez et al. 2021). Since lipid hydroperoxides are also a potent ROS, it is in the parasite’s interest to reduce their levels. Indeed, the parasite also shows higher relative expression of GST compared to other genes, but not in the same extent as the avian host. Interestingly, the parasite – but not the host – exhibits high expression of GSH synthetase, suggesting that the parasite aims to maintain elevated intracellular GSH levels and compensate for the GST-mediated GSH depletion. Furthermore, the oxidized GSSG can be converted back to its active form GSH, via the energy-dependent enzyme GR. This process benefits the parasite, which indeed shows relatively high expression of GR – a pattern not observed in the host. This can contribute to the increased GSH/GSSG ratio observed during and after the peak of parasitaemia (Fig. 3c), ultimately enhancing the antioxidant capacity of the red blood cells. Interestingly, in human malaria, key parasitic antioxidant enzymes, such as SOD, GR, GPX, and GST are considered promising targets for antimalarial drug development (Becker et al. 2004; Vasquez et al. 2021). However, actively increasing oxidative stress within the host is a delicate balance, as oxidative damage is a key driver of disease severity and mortality (e.g., Vasquez et al. 2021). To conclude, the present study reveals a host-parasite oxidative arms race , between P. ashfordi and the infected avian host – Eurasian siskin. The gene expression data shows that the parasite aims to maintain a reduced intracellular environment and eliminate ROS, whereas the host lower its antioxidant capacity through the upregulation of GST and GPX, thereby decreasing the availability of active GSH. Although, the parasite seems to win the race in terms of the oxidative state of the cell, the fact that parasitaemia decrease markedly form day 21 to 31 suggest that the host’s strategy of depleting GSH levels is ultimately effective in suppressing the infection. The host’s strategy is risky, given that severe pathology of malaria is generated through oxidative insult on host cells (e.g., Vasquez et al. 2021). Nevertheless, since all birds recovered from the infection, the short-term physiological costs appear to have been outweighed by the long-term benefits of parasite clearance and survival. Our results provide novel insights into the oxidative strategies employed by both host and parasite, highlighting the complex biochemical interplay underlying infection outcomes. 7 Posted on 22 Apr 2025 — The copyright holder is the author/funder. All rights reserved. No reuse without permission. — https://doi.org/10.22541/au.174536476.67136425/v1 — This is a preprint and has not been peer-reviewed. Data may be preliminary. Future studies may build on this work by investigating whether different avian malaria parasites—those that induce high or low parasitaemias, exhibit generalist or specialist host strategies—display similar redox dynamics, and how these may influence host resilience.

Asghar, M., Hasselquist, D. and Bensch, S. 2011. Are chronic avian haemosporidian infections costly in wild birds? – J. Avian Biol. 42: 530–537. Atamna, H. and Ginsburg, H. 1997. The malaria parasite supplies glutathione to its host cell — Investigation of glutathione transport and metabolism in human erythrocytes infected with Plasmodium Falciparum . – Eur. J. Biochem. 250: 670-679. https://doi.org/10.1111/j.1432- 1033.1997.00670.x Becker, K., Rahlfs, S., Nickel, C. and Schirmer, R. H. 2003. Glutathione – Functions and metabolism in the malarial parasite Plasmodium falciparum . - Biol. Chem. 384 : 551-566. https://doi.org/10.1515/BC.2003.063 Becker, K., Tilley, L., Vennerstrom, J.L., Roberts, D., Rogerson, S. and Ginsburg, H. 2004. Oxidative stress in malaria parasite-infected erythrocytes: host – parasite interactions. - Int. J. Parasitol .34 : 163-189. 10.1016/j.ijpara.2003.09.011. Clark, N.J., Clegg, S.M. and Lima, M.R. 2014. A review of global diversity in avian haemospo- ridians (Plasmodium and Haemoproteus: Haemosporida): new insights from molecular data. – Int. J. Parasitol. 44: 329–38. Griffith, O.W. 1999. Biological and pharmacological regulation of mammalian glutathione syn- thesis. - Free Radic. Biol. Med . 27: 922-935. Halliwell, B. and Gutteridge, J.M.C. 2002. Free radicals in biology and medicine. Oxford University Press, Oxford. Hellgren, O., Waldenstr¨om, J. and Bensch, S. 2004. A new PCR assay for simultaneous studies of Leucocytozoon, Plasmodium, and Haemoproteus from avian blood. - J . Parasitol . 90: 797- 802. DOI: http://dx.doi.org/10.1645/GE-184R1 Isaksson, C., ¨Ornborg, J., Stephensen, E. and Andersson, S. 2005. Plasma glutathione and carotenoid coloration as potential biomarkers of environmental stress in great tits. - EcoHealth 2: 138-146 Isaksson, C., Sepil, I., Baramidze, V. and Sheldon, B.C. 2013. Explaining variance of avian malaria infection in the wild: the importance of host density, habitat, individual life-history and oxidative stress. - BMC Ecol. 13: 15. https://doi.org/10.1186/1472-6785-13-15 Knowles, S.C.L., Nakagawa, S. and Sheldon, B.C. 2009. Elevated reproductive effort increases blood parasitaemia and decreases immune function in birds: a meta-regression approach. – Funct. Ecol.23: 405–415. Knowles, S.C.L., Palinauskas, V. and Sheldon, B.C. 2010. Chronic malaria infections increase family inequalities and reduce parental fitness: experimental evidence from wild bird popula- tion. – J. Evol. Biol . 23: 557–569. Kotepui, M., Kotepui, K., Mahittikorn, A. Majima, H.J., Tangpong, J. and Yen, H-C. 2023. Association of reduced glutathione levels with Plasmodium falciparum and Plas- modium vivax malaria: a systematic review and meta-analysis. – Sci. Rep . 13: 16483. https://doi.org/10.1038/s41598-023-43583-z LaPointe, D.A., Atkinson, C.T. and Samuel, M.D. 2011. Ecology and conservation biology of avian malaria. – Ann. N. Y. Acad. Sci . 1249: 211–226. 8 Posted on 22 Apr 2025 — The copyright holder is the author/funder. All rights reserved. No reuse without permission. — https://doi.org/10.22541/au.174536476.67136425/v1 — This is a preprint and has not been peer-reviewed. Data may be preliminary. Love, J., Selker, R., Marsman, M., Jamil, T., Dropmann, D., Verhagen, J., Ly, A., Gronau, Q.F., ˇSm´ ıra, M., Epskamp, S., Matzke, D., Wild, A., Knight, P., Rouder, J.N., Morey, R.D. and Wagenmakers, E.-J. 2019. JASP: Graphical Statistical Software for Common Statistical Designs. - J. Stat. Software . 88: 1–17. https://doi.org/10.18637/jss.v088.i02 Opplinger, A., Christe, P. and Richner, H. 1996. Clutch size and malaria resistance. - Nature 391: 565 Palinauskas, V., Valki¯ unas, G., Bolshakov, C.V. and Bensch, S. 2008. Plasmodium relictum (lineage P-SGS1): effects on experimentally infected passerine birds. – Exp. Parasitol. 120: 372-80. doi: 10.1016/j.exppara.2008.09.001. Perc´ ario, S., Moreira, D.R., Gomes, B.A.Q., Ferreira, M.E.S., Gon¸ calves, A.C.M., Laurindo, P.S.O.C., Vilhena, T.C., Dolabela, M.F. and Green, M.D. 2012. Oxidative stress in malaria. - Int. J. Mol. Sci. 13: 16346-16372. https://doi.org/10.3390/ijms131216346 R ˚ aberg, L., Graham, A.L. and Read, A.F. 2009. Decomposing health: tolerance and resistance to parasites in animals. – Philos. Trans. R. Soc. Lond. B. Biol. Sci. 364: 37–49. Read, A.F., Graham, A.L. and R ˚ aberg, L. 2008. Animal defences against infectious agents: is damage control more important than pathogen control? - PLoS Biol. 6: 2638–2641. Schneider, D.S. and Ayres, J.S. 2008. Two ways to survive infection: what resistance and tolerance can teach us about treating infectious diseases. – Nat. Rev. Immunol. 8: 889-95. doi: 10.1038/nri2432. Sies, H. 1986. Biochemistry of oxidative stress. - Angew. Chem. Int. Ed. Engl . 25: 1058-1071. https://doi.org/10.1002/anie.198610581 Valkiunas, G., Iezhova, T.A., Krizanauskiene, A., Palinauskas, V., Sehgal, R.N. and Bensch, S. 2008. A comparative analysis of microscopy and PCR-based detection methods for blood parasites. –J. Parasitol. 94: 1395-401. doi: 10.1645/GE-1570.1. PMID: 18576856. Vasquez, M., Zuniga, M. and Rodriguez, A. 2021. Oxidative stress and pathogenesis in malaria. - Front. Cell. Infect. Microbiol . 11. DOI=10.3389/fcimb.2021.768182 Videvall, E., Cornwallis, C.K., Palinauskas, V., Valki¯ unas, G. and Hellgren, O. 2015. The avian transcriptome response to malaria infection. – Mol. Biol. Evol. 32: 1255-67. doi: 10.1093/molbev/msv016. Videvall, E., Cornwallis, C.K., Ahr´ en, D., Palinauskas, V., Valki¯ unas, G. and Hellgren, O. 2017. The transcriptome of the avian malaria parasite Plasmodium ashfordi displays host-specific gene expression. – Mol. Ecol. 26: 2939-2958. https://doi.org/10.1111/mec.14085 Figure 1. Glutathione redox-cycle. The vector transmits the parasite to the hosts blood stream. The avian immune system responds by oxidative burst, by which immune cells release superoxide (O 2-*) and hydro- gen peroxide (H 2O2). These two reactive oxygen species (ROS) can attack and destroy the pathogen. However, they can also be detoxified via the glutathione system, originating from either the host or the pathogen. Superoxide can be dismutated by the enzyme superoxide dis- mutase (SOD) to hydrogen peroxide. Reduced glutathione (GSH) can then detoxify hydrogen peroxide to water by being oxidized to glutathione disulfide (GSSG). This reaction is catal- ysed by the enzyme, glutathione peroxidase (GPX). GSSG can be reduced back to its active antioxidant form (GSH) by the energy-demanding enzyme glutathione reductase (GR). Apart from the redox cycle, GSH can also conjugate to toxic compounds by the enzyme glutathione S-transferase. By doing so, there will be an efflux of GSH out from the cell, and tGSH (GSSG + GSH) will be reduced. In the present study, the tGSH and GSSG are measured in whole 9 Posted on 22 Apr 2025 — The copyright holder is the author/funder. All rights reserved. No reuse without permission. — https://doi.org/10.22541/au.174536476.67136425/v1 — This is a preprint and has not been peer-reviewed. Data may be preliminary. blood. The enzymes are quantified by gene expression (see details in Videvall et al. 2020). In birds, the differential gene expression between the peak of parasitaemia and before infection is presented. For the parasite, the gene expression is instead presented as a relative expression of the targeted genes to other genes in the transcriptome. The gene expression results are summarized here by using a bird and vector illustration (iStock) and the magnitude of the effect is illustrated with +. Hosted file image1.emf available at https://authorea.com/users/309348/articles/1289090-host-parasite- oxidative-arms-race-who-will-win Figure 2. Parasitaemia (%) in P. ashfordi -infected Eurasian siskins ( Spinus spinus ). Para- sitaemia is shown for days 0, 11, 21, and 31 post-infection. Data are presented as box plots, with the median indicated within each box and whiskers representing minimum and maximum values. Figure 3. The mean ( ± SE) of the antioxidant system – glutathione across time of either control or P. ashfordi infected Eurasian siskins (Spinus spinus ). The concentrations of a) total glutathione (tGSH), and b) oxidized glutathione (GSSG) of avian red blood cells are shown. In c) the estimated ratio between the reduced and the oxidized glutathione (GSH/GSSG ratio) is shown. Controls are shown in open circles, and infected birds are shown in filled black circles. Sampling days were 0, 11, 21 and 31 days after injection. 1a) Hosted file 10 Posted on 22 Apr 2025 — The copyright holder is the author/funder. All rights reserved. No reuse without permission. — https://doi.org/10.22541/au.174536476.67136425/v1 — This is a preprint and has not been peer-reviewed. Data may be preliminary. image3.emf available at https://authorea.com/users/309348/articles/1289090-host-parasite- oxidative-arms-race-who-will-win 1b) Hosted file image4.emf available at https://authorea.com/users/309348/articles/1289090-host-parasite- oxidative-arms-race-who-will-win 1c) Hosted file image5.emf available at https://authorea.com/users/309348/articles/1289090-host-parasite- oxidative-arms-race-who-will-win Table 1. Summary statistics of glutathione physiology in red blood cells of malaria infected and healthy Eurasian siskins (Spinus spinus ). In a) total glutathione, b) oxidized glutathione (GSSG) and c) the ratio between the reduced and the oxidized glutathione (GSH/GSSG) the results for treatment (infected versus controls), sampling day (0, 11, 21 and 31) and the interaction between treatment and day are shown using ANCOVAs. GSSG is log 10 transformed and GSH/GSSG is squared rooted to reach normal distribution. Significant effects are indicated in bold (p < 0.05). SS = Sum of squares and MS = Mean Square. Factors SS df MS F p-value a) tGSH treatment 75.982 1 75.982 2.794 0.097 day 193.758 3 64.586 2.375 0.073 treatment day 123.945 3 41.315 1.519 0.212 Residuals 3698.753 136 27.197 b) GSSG treatment 0.389 1 0.389 6.646 0.011 day 1.464 3 0.488 8.340 < .001 treatment day 0.681 3 0.227 3.880 0.011 Residuals 7.960 136 0.059 c)GSH/GSSG treatment 4.621 1 4.621 3.867 0.051 day 5.128 3 1.709 1.431 0.237 treatment day 5.215 3 1.738 1.455 0.230 Residuals 161.326 135 1.195 11