How organisms respond to environmental changes, such as urbanization, remains a pressing question, particularly during times of unprecedented anthropogenic impacts. On the one hand, urban environments may present novel resources for exploitation, such as new, predictable food sources (Jones and Reynolds, 2008; Oro et al., 2013). On the other hand, anthropogenic changes to the environment, like artificial light at night, air pollution, and anthropogenic noise, along with suboptimal nutrient composition in the diet are likely to challenge organismal health (Bonier, 2012; Isaksson, 2015; Ouyang et al., 2015; Isaksson and Bonier, 2020). Such challenges can be met with behavioral and physiological responses in order to avoid negative consequences (Tuomainen and Candolin, 2011; Sih, 2013; Watson et al., 2017; Ouyang et al., 2018). However, most responses have been evaluated when faced with only one stressor at a time and under controlled conditions. The potential responses to multiple stressors acting simultaneously are much less explored.

Oxidative stress, the imbalance between pro- and anti-oxidants, can occur either due to internally released pro-oxidants such as during high metabolic rate and oxidative bursts to defeat pathogens during an infection, or due to external sources such as inhalation of air pollutants (Isaksson, 2010; 2015). Repeated or chronically enhanced oxidative stress will lead to an accumulation of oxidative damage to biomolecules, which may ultimately cause premature death (Mudway et al., 2020). The oxidative stress status of urban-dwelling organisms has been investigated over the last decades, with a consensus that urban environments indeed lead to an increase in oxidative stress (Isaksson, 2010), but the magnitude of effect varies among studies (Isaksson et al., 2009; Rainio et al., 2013; Giraudeau and McGraw, 2014; Herrera-Duenas et al., 2014; de la Casa-Resino et al., 2015; Amri et al., 2017; Salmón et al., 2018b). Such context-dependency could be either caused by variation in pollutant types or concentrations, but it could also indicate that the response to pro-oxidants is modulated by other factors, such as the availability of important nutrients to support the antioxidant system (Isaksson, 2020) and the immune defense (Jiménez-Peñuela et al., 2022). However, the potential interplay between anthropogenic factors, infection, and nutrient availability remains largely unexplored (Jensen et al., 2023).

Diet quantity and quality are known to be influenced by urbanization (Seress et al., 2018; Jarrett et al., 2020). For some species, anthropogenic food sources may be beneficial as an additional, easy-accessible energy resource, and may even positively influence survival and performance in urbanized environments (Jones and Reynolds, 2008; Robb et al., 2008; Harrison et al., 2010). However, anthropogenic food sources as well as naturally occurring foods in the urban environment can be of poorer nutritional value (Isaksson and Andersson, 2007). Specifically, anthropogenic food often differs from naturally available food items in its composition of polyunsaturated fatty acids (PUFAs) (Simopoulos, 2002; Andersson et al., 2015). Indeed, urban birds have relatively lower circulating levels of ω3-PUFAs in their bloodstream than rural birds, especially during winter (Andersson et al., 2015; Isaksson et al., 2017). PUFAs are involved in several important physiological processes in animals, including, for example, growth during development (Twining et al., 2016b) and immune responses (Craig-Schmidt et al., 1987; Larsson et al., 2004). The short-chained ω6-PUFA linoleic acid (LA, 18:2n-6) and ω3-PUFA α-linolenic acid (ALA, 18:3n-3) are essential to birds and strictly acquired through diet (Larsson et al., 2004; Twining et al., 2021a). From these PUFAs, longer-chained PUFAs can be biosynthesized, although the efficiency of this process seems to vary between bird species (Twining et al., 2017; Twining et al., 2021b). Thus, the dietary availability of both short- and long-chained PUFAs can be imperative for physiological processes.

Immune responses are part of these processes since twenty-carbon PUFAs, like arachidonic acid (AA, 20:4n-6, a ω6-PUFA) and eicosapentaenoic acid (EPA, 20:5n-3, a ω3-PUFA), are pre-cursors to eicosanoids (i.e., prostaglandins, thromboxanes, leukotrienes), which function as immunomodulators. Importantly, eicosanoids show opposing inflammatory effects, depending on which substrate they originate from. Eicosanoids derived from ω6-PUFAs (AA) are considered mostly pro-inflammatory and to increase oxidative stress, while eicosanoids derived from ω3-PUFAs (EPA) exhibit anti-inflammatory properties and have been found to lower the production of ω6-derived eicosanoids by displacement or competitive inhibition (Craig-Schmidt et al., 1987; Tapiero et al., 2002; Larsson et al., 2004). Thus, the dietary balance between ω6-and ω3-PUFAs (i.e., the ω6:ω3-ratio) is critical for which pathway of eicosanoid production is favored (Fritsche et al., 1991) with subsequent health effects (Twining et al., 2016a). Subsequently, a diet that is relatively low in ω3-to ω6-PUFAs, which is likely in an urban environment (Andersson et al., 2015; Isaksson et al., 2017), may lead to a physiological state prone to inflammation or elicit increased oxidative damage due to a prolongation of the inflammatory response (Innes and Calder, 2018). The risk of oxidative stress may be further exacerbated by the increased exposure to air pollutants in cities, but research investigating the modulatory potentials of dietary fatty acids in response to air pollution is still limited (Whyand et al., 2018). Studies in humans have shown that a higher intake of ω6-PUFAs increases the adverse health effects of particulate matter (PM), while a higher intake of ω3-PUFAs reduces symptoms and severity of PM-induced health effects (Romieu et al., 2005; Brigham et al., 2019). Rodents and cell cultures of rodents that have been treated with ω3-PUFAs showed ameliorated responses to PM (Li et al., 2019) and ozone (Snow et al., 2018). In summary, the risk of oxidative stress through air pollution could be enhanced by a diet of anthropogenic food due to its lower content of ω3-PUFAs or be potentiated due to the pro-inflammatory properties of ω6-PUFAs. Investigating the potential interactions between air pollution, diet, and immune response on the avian oxidative status may therefore give mechanistic insights into the effects of anthropogenic environments on wildlife health.

Here, we performed a full-factorial laboratory experiment to investigate if different dietary composition of ω6-and ω3-PUFAs (i.e., ω6-versus ω3-rich diets) modulate the oxidative stress response towards pro-oxidative stimuli: ozone (versus normal air) and a simulated pathogen attack (lipopolysaccharide [LPS] versus control injections), using captive zebra finches (Taeniopygia guttata). We chose ozone as the air pollutant in this experiment since ground-level (tropospheric) ozone is a major component of urban air pollution and avian abundance and population growth rates have been shown to be negatively related to concentrations of tropospheric ozone levels, especially in mountain regions (Liang et al., 2020; Reif et al., 2023). Ozone originates mainly from chemical reactions of emission components of motorized vehicle combustion engines triggered by sunlight (Monks et al., 2015; Zhang et al., 2019). Over the last few years, background ozone levels have increased in urban areas compared to rural areas (Paoletti et al., 2014; Sicard et al., 2020). The effects of ozone exposure are wide-ranging, for example, causing severe health impairments, such as respiratory distress and disease through inflammation and immune activation, both in humans and birds (Rombout et al., 1991; Monks et al., 2015; Sanderfoot and Holloway, 2017; Fleming et al., 2018). The toxicity of ozone stems from its ability to either directly oxidize other molecules or to indirectly create free radicals and form reactive intermediate molecules which lead to further cascading oxidizing reactions (Mustafa, 1990).

Furthermore, injection with LPS simulates a pathogen attack, which triggers a costly acute-phase immune response (Owen-Ashley et al., 2006) and thereby, increases the release of pro-oxidants. To assess the oxidative stress response from the treatments, we measured concentrations of hydroperoxides in plasma as a proxy for overall pro-oxidant exposure and oxidative damage (using dROM assay). To get a broad estimate of the antioxidant defenses, we measured the circulating overall antioxidant capacity in plasma (OXY), as well as the intra-cellular and endogenously synthesized antioxidant glutathione in red blood cells, both the total levels (tGSH) and the oxidized form (GSSG). From these two measurements, the reduced (active form, GSH) can be calculated and the GSH/GSSG ratio can be used to infer the cellular antioxidant potential and their oxidative stress state.

Three main predictions can be outlined. First, we predicted that under elevated ozone levels, birds on a ω6-rich diet will have higher oxidative damage and/or a higher level of antioxidant defense than birds on a ω3-enriched diet. Second, we predict that birds on the ω6-enriched diet will have a stronger reaction when immune challenged, compared to birds on an ω3-enriched diet. This will be reflected in either higher levels of oxidative damage and/or upregulation of the antioxidant defense systems compared to birds on an ω3-enriched diet. Third, we predicted that the combined challenge of elevated ozone levels and an immune challenge would exacerbate the diets’ effects on oxidative damage and/or on the antioxidant defense system.

Urban wildlife needs to deal with several human-introduced factors, such as changed diet composition, traffic-related air pollutants, and changed fauna of pathogens. These factors do not act in isolation but can interact, and thereby either exacerbate or mitigate negative effects of living in urban environments. To investigate this, we performed a full-factorial experiment using captive zebra finches. The main aim was to explore the modulatory role of dietary PUFAs on the oxidative stress response to ozone and/or a simulated pathogen attack—two pro-oxidative agents. Our results show that both a diet differing in ω6:ω3 PUFA ratio, exposure to ozone, and an immune challenge have the capacity to alter the oxidative stress status of birds. However, the modulatory effect of diet was less pronounced than predicted and only one marker revealed such an effect.

The modulatory effect of dietary PUFAs was revealed for the ratio between GSH/GSSG, but only under normal air and not when the birds were exposed to ozone. More specifically, birds fed a diet rich in long-chained ω3-PUFAs had a lower GSH/GSSG ratio compared to birds fed ω6-PUFAs in normal air conditions. For context, it is important to note that tGSH (GSSG + GSH) did not differ between our two diet treatments, meaning the results were not driven by absolute differences in total glutathione. Traditionally, the ratio of GSH/GSSG has been interpreted as a marker of current oxidative stress (i.e., a lower ratio (more GSSG) indicates a higher level of oxidative stress), which would suggest that a ω3-rich diet has a negative impact on the oxidative state within the red blood cells, under normal air conditions, but that this effect decreases when exposed to ozone. Given that another study has shown that ω3-supplementation increases the rate of oxidation of GSH when exposed to ozone (Corteselli et al., 2020), we find it unlikely that the ω3-group here experienced higher oxidative stress in normal air than when exposed to the extremly potent pro-oxidant ozone, or compared to the ω6-birds. Our measurement does not take oxidation rate in account, just the absolute concentration of GSSG and GSH. However, to compensate and minimize the pro-oxidant attack, the glutathione system uses glutathione reductase (GR) for catalyzing GSSG back into its active form (i.e., GSH). The expression of GR is regulated by the presence of pro-oxidants, hence when GR is upregulated the GSH/GSSG ratio will increase (Smith et al., 1996; Schafer and Buettner, 2001; Kim et al., 2010). Possibly, this is the reason for the increased ratio in the ω3-birds when exposed to ozone. Furthermore, the continuous high ratio among the ω6-birds, regardless of exposed to ozone or not, might be related to the well-known pro-inflammatory and pro-oxidative effect of dietary ω6-PUFAs (Innes and Calder, 2018). The two-differential diet regimes started 7 days prior to exposure treatments, and 12 days prior to blood sampling, hence the lack of effect of ozone at day 12 in the ω6-birds might be due to an already upregulated GR in response to the ω6-diet seen in the normal air group. This interpretation is further corroborated by the higher plasma antioxidant activity (i.e., OXY) in the ω6-birds compared to the ω3-birds, independently of ozone treatment. It should be noted that, although our interpretation is physiologically and theoretically sound, we regrettably did not measure GR activity, thus the underlying mechanism behind the observed pattern is not fully clarified by our results.

Moreover, we found that ozone exposure alone decreased the cellular concentration of total glutathione. The glutathione redox system is one of the most abundant and important antioxidant defenses in an organism (Meister and Anderson, 1983). The present exposure level of 109 ± 0.05 ppb for 5 days indicates slow depletion of glutathione from the cellular antioxidant machinery. A reduction of glutathione concentration has important implications on organismal condition, as a depletion of glutathione has been found to lead to cell dysfunction and apoptosis (Franco and Cidlowski, 2009). While a decrease of about 5% over 5 days in the ozone-exposed birds may seem a modest reduction, this begs the question of the consequences of more long-term ozone exposure on avian health. In accordance with the present study, toxicological studies of European starlings (Sturnus vulgaris) also reveal a depletion of glutathione when exposed to a mix of urban air pollutants, both in the wild and in a controlled setting when exposed to vehicle exhaust (North et al., 2017a; 2017b). However, other correlative studies of urban birds and pollution levels reveal no effect on glutathione levels, but note that ozone levels were not measured in any of these studies (Salmón et al., 2018b; Li et al., 2021). Experimental studies of ozone specifically and its effects on antioxidants have primarily been done on mammals, either in vitro or in vivo, and instead of using biomarkers in the blood most of these studies focus on other fluids (such as epithelial lining fluid) or specialized cells in the respiratory system (e.g., Rietjens et al., 1985; Avissar et al., 2000). Hence, comparing studies is difficult due to widely varying tissues/fluids used, biomarkers, species but also due to wide ranges of ozone concentrations (1–800 ppb) and exposure times (4 h—1 week). However, given that blood has been proposed to be an insensitive tissue for ozone exposure (Kadiiska et al., 2011), the fact that we here see a 5% reduction in tGSH in red blood cells raises concern about the oxidative status of other tissues.

In addition to the pro-oxidative ozone exposure, we also immune-challenged half of the birds with an injection of LPS to trigger a release of ROS to combat the presumed infection. In contrast to the ozone, the pro-oxidants released during this oxidative burst is an adaptive mechanism to destroy pathogens. In line with this, the immune-challenged birds suffered from higher levels of oxidative damage. Since accumulation and/or re-occurring oxidative damages over time is directly linked to age-related loss in cellular function (Martin and Grotewiel, 2006) a continued oxidative burst could have a profound long-term consequence for the organism (Costantini and Møller, 2009). Yet, the immediate benefit of this adaptive response must overrides the long-term costs. Interestingly, and in line with this adaptive response, the antioxidant system is not triggered to defeat this internal production of ROS, since neither the antioxidant capacity in the plasma (OXY) nor the antioxidant markers in the red blood cells (tGSH and GSH/GSSG) were upregulated. Increased oxidative stress and subsequent oxidative damage have been shown previously after an immune response to fight invading pathogens (Isaksson et al., 2013; Armour et al., 2020). However, our study suggests that this increased oxidative stress is decoupled from the different antioxidant potential of PUFA intake, as well as additional pro-oxidant exposure via ozone; we found no indication that the effects of oxidative burst were modulated by these two treatments. This could indicate a mechanism where the oxidative burst to defeat pathogens overrides the potentially protective effects of diet and detoxification of pro-oxidants.

Lastly, we found less evidence for cumulative and/or modulatory effects between the three experimental treatments—differential PUFA-diet, ozone exposure, and immune challenge—than expected. It has previously been proposed that the magnifying effect of multiple anthropogenic factors might be overstated since interactions often are assumed to be synergistic (Côté et al., 2016). However, although we were neither able to show cumulative nor synergistic effects of the combined treatments in independent biomarkers, the overall effect on the physiological state when taking all effects into account could potentially worsen the condition of wild urban birds. For example, an increased accumulation of oxidative damage (e.g., after an infection) could be more detrimental to the urban than rural birds, given that they also can have the challenge of a reduced cellular antioxidant defense, as shown here by the lower levels of glutathione in response to ozone exposure. In addition, urban birds have relatively more ω6-to ω3-PUFAs circulating in their body (and presumably also a more ω6-rich diet) compared to rural birds (Andersson et al., 2015). Diet could therefore also influence the overall physiological state of urban birds, since experimental diet-type affected both the total antioxidant capacity and the response of the glutathione redox-system to ozone exposure. Yet, whether any of the two PUFA-diets would worsen or improve the overall condition of the urban birds remains to be unraveled.

In summary, we found that diets differing in ω6-and ω3-PUFA concentrations modulate the overall plasma antioxidant capacity and redox state of the cellular glutathione system depending on ozone exposure. Ozone alone depleted the cells from glutathione, which in the long run may have severe consequences for the organism. However, only an immune challenge caused increased levels of oxidative damage, which is notable since it is an adaptive (yet apparently costly) part of the immune defense. Taken together, ozone exposure, dietary PUFAs, and infection simulation all influenced oxidative stress biomarkers of birds but through different physiological pathways. Disease, diet, and air pollution therefore appear as potential parallel drivers of the increased oxidative stress associated with urban-dwelling organisms. Additionally, our results indicate that the response to a simultaneous change of more than one anthropogenic pollutant may lead to different responses, compared to single stressor exposure. Since diet and disease prevalence co-occur with anthropogenic pollution in urban habitats, experimental studies using multiple stressors is key to understand the overall impact urbanization has on organismal health. Despite the lack of any apparent cumulative or synergistic effect of our stressors, the combined effect and conclusion from the independent biomarkers suggests that urban birds face an overall challenged physiological state.