Modulating offspring responses: concerted effects of stress and immunogenic challenge in the parental generation

Development in avian species takes place in environments where certain conditions may be challenging. Depending on whether birds are raised in wild or captive conditions, such challenges could be associated with long-distance migrations, predation risk, extreme weather conditions, social interactions, resources availability, transport and the exposure to potentially pathogenic microorganisms or vaccines (Ashley and Demas, 2017; Babb et al., 2014; Bowling et al., 2018; Buehler et al., 2008, 2010; Calefi et al., 2017; Nazar et al., 2018b; Storm and Lima, 2010). Regardless of the type of environment in which birds are raised, the perception, processing and response to challenges will involve the activation of the immuno-neuroendocrine (INE) interplay (Ashley and Demas, 2017; Babb et al., 2014; Blom and Ottaviani, 2017; Bowling et al., 2018; Nazar et al., 2018b; Saino et al., 2005). This interplay orchestrates responses to a wide spectrum of challenges, ranging from local inflammation to systemic stress responses (Ashley and Demas, 2017; Blom and Ottaviani, 2017; Martin et al., 2008; Nazar et al., 2017).

The hypothalamic–pituitary–adrenal (HPA) axis is one of the main representative components of the INE interplay, having a pivotal role in the interaction with the environment (Koolhaas, 2008; Kudielka and Kirschbaum, 2005). The activation of this axis releases stress mediators to the bloodstream, inducing physiological and behavioural changes aimed at coping with environmental challenges, promoting survival (Nazar and Marin, 2011; Scanes, 2016). The release of systemic stress mediators involves changes in the immune system, downregulating induced cell-mediated response and upregulating induced humoral response (Elenkov, 2004; Murphy, 2009). Reciprocally, activation of the immune system can also entail the production of stress mediators (Dhabhar, 2014), establishing a bidirectional communication.

Immune responses can be classified based on whether the response is constitutive or induced, as well as its specificity (Davison, 2014; Murphy, 2009; Schmid-Hempel and Ebert, 2003). Constitutive responses are the first line of defence and act rapidly, mainly including phagocytic cells and soluble proteins, i.e. macrophages and the complement system, respectively. The activation of constitutive responses induces non-specific inflammation via inflammatory cytokines, thus restricting the dissemination of a wide range of pathogens (Iwasaki and Medzhitov, 2010). Specific responses can be categorized, according to their main components, as cell-mediated or humoral. A subclass of T-helper cells (Th1) is the major effector of cell-mediated responses. A different subclass (Th2), together with B-cells and their soluble derived products, are mainly responsible for humoral responses, which rely on, for example, soluble proteins (immunoglobulins/antibodies) that protect organisms against pathogens through opsonization, complement activation, or neutralization (Elenkov, 2004).

When birds are exposed to any environmental challenge, energy and resources are invested to deal with it and reduce the risks of mortality. The differential allocation of the energy and resources is related to metabolic and physiological trade-offs. A broad definition of trade-off taken from Senner et al. (2015) states that it refers to any instance when resource allocation to one trait occurs at the expense of another. Consequently, fitness-related traits, such as survival, reproduction, parental care and health status, among others, could be comprised (Buehler et al., 2010; Piersma, 2011; Romero et al., 2009; Senner et al., 2015). Mounting an immune response is energetically costly. When birds are exposed to different environmental challenges, in both in wild and captive environments, the immune responses are compromised, with major consequences on health status (Lifjeld et al., 2002; Zuk and Stoehr, 2002). Therefore, it is necessary to establish a balance between diverting resources towards coping with a challenge and developing a proper immune response. In this context, challenges inducing a sustained activation of the HPA axis might, in turn, be detrimental for avian health status, with several metabolic, immune and hormonal consequences (Nazar and Marin, 2011; Nazar et al., 2018b; Shini et al., 2010; Webster Marketon and Glaser, 2008).

In particular, daily exposure to high environmental temperatures during several continuous days is a challenging phenomenon described in the literature as chronic heat stress (CHS) (Akbarian et al., 2016; Nazar et al., 2018a; Xie et al., 2015). This phenomenon activates the HPA axis, altering hormonal concentrations and biochemical variables (Akbarian et al., 2016; Bozkurt et al., 2012; Mashaly et al., 2004) as well as humoral and cellular immune responses (Lara and Rostagno, 2013; Nazar et al., 2018b). The performance of birds reared with productive purposes and the quality of derived products (such as eggs and meat) have also been affected by CHS exposure (Salabi et al., 2011; Lara and Rostagno, 2013; Mashaly et al., 2004; Nazar et al., 2018a,b; Xie et al., 2015; Renaudeau et al., 2012; Star et al., 2008). Similarly to other environmental challenges, dealing with CHS not only demands great amounts of energy in terms of resources (du Plessis et al., 2012) but also alters the individual behavioural time budget (Mack et al., 2013). It increases heat dissipation behaviours and is capable of altering nutrient adsorption by impairing the integrity of the intestinal mucosae in the domestic fowl and other domestic species (Lara and Rostagno, 2013; Song et al., 2013).

Based on the previous reasoning, exposure to an antigen could also be defined as an environmental challenge. To deal with an immune challenge (ICH), an optimal immune response needs to be mounted, considering its associated costs (Klasing, 2007; Lee, 2006). This involves INE interplay activation, directly stimulating immune responses to fight potential threats (Webster et al., 2002). Immune challenges related to infections can produce sickness symptoms such as fever, having a direct impact on the feed intake and the quality of eggs (Star et al., 2008). Particularly, when responding to Salmonella infection, both Th1 and Th2 responses are stimulated, but the cell-mediated immunity has a more important role in responding to this pathogen (van Immerseel et al., 2005).

Coping with simultaneous challenges would demand energy and resources to deal with the more demanding stressor (Romero et al., 2009). Birds undergoing chronic and demanding events, such as long-distance migrations or the exposure to high environmental temperatures, might, on the one hand, find their energetic balance altered (i.e. energy intake versus energy expenditure), and on the other hand, reach metabolic and physiological thresholds, resulting in the aforementioned trade-offs (Piersma, 2011; Romero et al., 2009). In the present study, domestic birds were exposed to an ICH when undergoing a CHS, meaning that energy expended in coping with the stressor could limit the immune potential of birds (Buehler et al., 2010), threatening their survival according to the actual defence model. Buehler et al. (2010) proposed that an animal's actual defence results from the balance between the available resources to invest in immunity (immune potential), and the resource required for pathogen clearance (required response). Taking this into consideration, a first question arises. How would the quail INE interplay respond to an ICH while also dealing with a demanding chronic stressor (CHS)?

Environmental challenges in one generation are known to influence its offspring (Chin et al., 2009; Reynolds et al., 2010; Schmidt et al., 2014; Storm and Lima, 2010). Developmental programming (Champagne, 2008) is the phenomenon that integrates the challenges experienced by the parental generation together with its transgenerational effects in the developing offspring. While some authors associate developmental programming with negative outcomes such as metabolic diseases (Aiken and Ozanne, 2014; Cottrell and Seckl, 2009; Martin-Gronert and Ozanne, 2012), others have found that transgenerational effects can be adaptive. Transgenerational effects on the HPA axis reactivity might, for example, improve the capability of the offspring to adapt to future stressful conditions both in mammalian (Reynolds et al., 2010; Schmidt et al., 2014) and avian species (Zimmer et al., 2013, 2017). Transgenerational effects of immune challenges have also been described. It has been reported that maternal immune activation alters behaviour, growth and immune-related variables in wild and domestic mammals and birds (Broggi et al., 2016; Grindstaff et al., 2012; Ronovsky et al., 2017; Saino et al., 2005).

For mammals, Ronovsky et al. (2017) showed that maternal immune challenge altered the expression of receptors that are key components of the HPA axis throughout two consecutive generations. In birds, studies on transgenerational consequences of challenging environments during early life have shown effects in both wild and domestic species. Effects on offspring behaviour and learning ability of captive zebra finches (Taeniopygia guttata) (Grindstaff et al., 2012) were observed when parental individuals were immune-challenged. Consistent with enhanced immune function, nestlings of immunized mothers had a more robust inflammatory response and higher fledging success in tree swallows (Tachycineta bicolor) (Burness et al., 2018), and developed stronger specific response in free-living house sparrows (Passer domesticus) (Broggi et al., 2016). In the case of domestic avian species, prenatally stressed quails were programmed to downregulate the HPA axis more effectively and showed a diminished stress response and increasing risk-taking, helping them to better cope with stressful environments (Zimmer and Spencer, 2014; Zimmer et al., 2017). Consequently, when offspring from captive or wild species are exposed to challenging conditions during early life, consequences should be interpreted according to specific environmental contexts to elucidate whether they are adaptive (Nederhof and Schmidt, 2012).

Challenging environments influence the INE responses of both the parental generation and their offspring, through developmental programming. Consequently, the aim of the present study was to evaluate whether concerted challenges, in the form of an immune inoculum with Salmonella (ICH) and a CHS exposure on adult Japanese quail, can modulate parental and offspring INE responses. We hypothesized that concomitant exposure would affect INE variables in adults, also having transgenerational consequences on the INE interplay of their offspring. In particular, we expected the specific immune responses of the parental generation to be negatively affected by both challenges (Buehler et al., 2010; Nazar et al., 2018b). Regarding offspring, the programming should then affect the immune functions directly associated to the HPA axis reactivity (Merrill and Grindstaff, 2015). Consequently, we predicted a polarization of the immune response towards a Th1 profile (by enhancing cell-mediated response and decreasing the humoral response) (Zimmer and Spencer, 2014).

Japanese quail [Coturnix coturnix (Linnaeus 1758)] were used as a laboratory animal model for this study. Their short life cycle and high hatch rate (Huss et al., 2008) make them particularly suitable for transgenerational studies.

Husbandry procedures were developed as established in Marin and Satterlee (2004) and Nazar and Marin (2011). A total of 100 chicks were randomly housed in two enriched rearing boxes (50 chicks per box) measuring 90×90×90 cm (length×width×height). Each box had one 90 cm feeder and 16 automatic nipple-like drinkers. A wire-mesh floor (1 cm grid) was raised 5 cm to allow excreta passage; a lid prevented quail from escaping. The floor of the rearing box was covered with corrugated cardboard (50% of the total surface). Structural and contact enrichments consisted of coloured (blue, green, violet) plastic balls and wooden platforms, respectively (Miller and Mench, 2006; Nazar and Marin, 2011). Brooding temperature was set at 37°C during the first week of life, declining 3°C per week until 25°C was reached. Quail starter diet (24% crude protein; 2900 kcal metabolizable energy kg⁻¹) was used for the first 4 weeks of life. Laying diet was provided (21% protein, 2750 kcal ME kg⁻¹). Food and water were provided ad libitum until the end of the experiment. A daily cycle of 14 h light (300–320 lx) and 10 h dark was established throughout the study, starting at 06:00 h.

A total of 80 quail were used to study the parental generation. At 28 days of age (DA), quail were sexed using breast plumage coloration: while females show speckled brown feathers, males have cinnamon brown feathers (Huss et al., 2008). Quails were wing-banded for later identification and randomly re-housed in 40 male–female pairs in 45×20×25 cm (length×width×height) home cages. Structural and contact environmental enrichments were used as previously described (Miller and Mench, 2006; Nazar and Marin, 2011). The parental treatments defined four different prenatal environments for the offspring.

A 2×2 factorial design evaluating the effects of the ICH and a CHS was used. Thus, at 100 DA, 40 parental pairs of one male and one female were randomly assigned to one of four treatments (10 parental pairs each): non-CHS/non-ICH, non-CHS/ICH, CHS/non-ICH and CHS/ICH. The manipulation of the parental generation defined four offspring treatments.

Heat stress was applied chronically throughout nine consecutive days (between 121 and 129 DA) only during daylight, to simulate the daily rising of temperature in a summer month. Room temperature was increased from 24°C to 34°C (0.1°C min⁻¹) starting at 08:30 h and maintained until 16:30 h. For the next 2 h, temperature was gradually lowered, returning to 24°C. This protocol has been used previously in our laboratory and was reported to induce a physiological chronic stress response (Nazar et al., 2018a,b). Non-stressed treatments (non-CHS/non-ICH and non-CHS/ICH) were reared at standard rearing temperature for the ontogenic stage (24±1°C) during the whole experiment.

At 115 DA, an intramuscular inoculation of 0.3 ml of formalin-inactivated Salmonella ser. enteritidis suspension (10⁸ bacteria quail⁻¹) in saline solution was injected into the quails’ right breast according to previous treatment assignation (Deguchi et al., 2009). This was carried out 6 days before the CHS protocol started, because individuals develop maximum antibody titres approximately 14 days post-inoculation. Thus, we were aiming to have maximum responses to this inoculum challenge at the same time that the CHS would also be requiring a maximum immune physiological demand (i.e. at the end of the CHS exposure; Nazar et al., 2018a). Non-challenged birds were exposed to the same overall handling as their challenged counterparts. However, the administered control injection contained only a saline solution.

To obtain the offspring, eggs were collected for incubation during 125–130 DA. Even though the CHS protocol finished at 129 DA, eggs were also collected at 130 DA considering that those eggs were influenced by the CHS prenatal condition. All eggs and offspring were individually identified according to their parental treatment identity to ensure later follow up. A total of 21, 17, 26 and 18 newly hatched chicks from the non-CHS/non-ICH, non-CHS/ICH, CHS/non-ICH and CHS/ICH prenatal environments, respectively, were evaluated, defining prenatal non-CHS/non-ICH, non-CHS/ICH, CHS/non-ICH and CHS/ICH treatments. Coloured and numbered leg rings were used for the identification of chicks. There was no significant effect of treatments on the hatching number per parental pair (Generalized linear model Poisson distribution; F_1.36<1.67, P>0.20). The average number of chicks per parental pair was 2.41±1.13.

Parental samples were taken at 130 DA, whereas offspring samples were taken at 32 DA. In both cases, sampling started at 09:00 h and blood was withdrawn from the jugular vein using 1 ml syringes previously treated with ethylenediaminetetraacetic acid (EDTA) to avoid blood coagulation. Because the presence of the experimenter was noticed by the birds, blood samples were taken in less than 120 s to avoid changes in the stress response mediators owing to fear and handling effects (Jones et al., 2005; Romero and Reed, 2005) that could influence the measured variables. One blood drop was used for smear preparation, while the rest was centrifuged at 2000 g for 15 min to obtain plasma samples for later determinations. In both cases, the sampling order was random.

Three INE-related variables were evaluated. First, the specific induced humoral response was analysed via a microagglutination assay against injected sheep red blood cells (SRBC) (Al-Khalifa, 2015; Nazar and Marin, 2011; Nazar et al., 2017). This technique allows us to assess primary antibody production after 1 week since antigen encounter and after inactivating the complement system through heating of the samples (Fair et al., 1999). It is important to point out that, as performed, this technique will capture the information from both natural and induced antibodies recognising SRBC. Second, the inflammatory induced response to phytohemagglutinin-p (PHA-P) injection was used to assess pro-inflammatory potential (Vinkler et al., 2010). This variable has historically been used to assess both constitutive and induced non-specific inflammation and Th1 inflammatory response (Hasselquist et al., 1999; Schmid-Hempel and Ebert, 2003). PHA-P stimulates T-cell proliferation and secretion of cytokines that recruit granulocytes. Swelling is due to the infiltration of circulating granulocytes and oedema; thus, swelling reflects a T-cell-mediated inflammatory response. Third, the heterophil/lymphocyte (H/L) ratio was used as a haematological proxy for chronic stress and, therefore, as an indirect indicator of stress response mediators concentration (Scanes, 2016). Elevated glucocorticoids concentrations leads to elevation in the circulating number of heterophils and a diminished number of lymphocytes (Schat and Skinner, 2014). Consequently, an elevation in the ratio is indicative of elevated glucocorticoids concentration as a final product of the HPA axis activation (Babacanoǧ lu et al., 2013; Shini et al., 2009).

One week before each sampling day, birds received an intraperitoneal injection of 0.3 ml of a 10% SRBC (HEMO-G, Rafaela, Santa Fe, Argentina) in order to induce a specific response to a foreign complex antigen. After sampling days, microagglutination assays were developed once plasma samples were obtained, using ‘U’ bottom microplates and following protocol described in Nazar et al. (2018a,b). In order to ensure that we were analysing the induced response, the innate complement was inactivated by heating at 56°C for 30 min (Moore and Siopes, 2005; Nazar and Marin, 2011). A volume of 30 ml of plasma was serially diluted in 30 ml of phosphate buffered saline (PBS; 1:2, 1:4, 1:8, up to 1:512). Next, 30 ml of a 2% suspension of SRBC in PBS was added to each well. Microplates were incubated at 40°C for 45 min and haemagglutination of the plasma samples was compared with that of the blank (PBS only) and the negative control (wells with no SRCB suspension). Antibody titres were reported as the log₂ of the highest dilution yielding agglutination (Nazar and Marin, 2011; Sever, 1962).

Twenty-four hours before the sampling procedure, the right wing-web was measured (in mm) with a digital calliper. Then, 0.05 ml of a 1 mg ml⁻¹ PHA-P solution in PBS was injected into adults and 0.025 ml of the same solution was injected into chicks. On the corresponding sampling days and after blood was withdrawn, the wing-web was again measured by the same experimenter and with the same calliper to determine percentage of inflammation using the following formula: percentage of inflammation=(wing-web measurement 24 h before sampling day/wing-web measurement on the sampling day)×100 (Nazar et al., 2015; Vinkler et al., 2010).

Smears previously made were stained with May–Grünwald–Giemsa and used to count and differentiate sub-populations of leucocytes (heterophils, basophils, eosinophils, lymphocytes and monocytes), counting 100 cells using a white light optical microscope. H/L ratio was calculated for each sample using the following formula: H/L ratio=(number of heterophils)/(number of lymphocytes) (Davis et al., 2008; Gross and Siegel, 1983).

Animal care was provided in adherence with Institutional Animal Care and Use Committee guidelines. The experiment was approved by the ethical committee at the Instituto de Investigaciones Biológicas y Tecnológicas in compliance with the legislation regarding the use of animals for experimental and other scientific purposes (Acta 27, 09/04/2015). No mortalities were registered during or after the experimental procedures started.

Generalized linear mixed models were used to analyse the effects of CHS and ICH on the described variables on each ontogenetic stage. CHS and ICH were used as fixed effects while home boxes (adults) and rearing boxes (offspring) were included as random effects. Considering the reported effect of sex-specific HPA responses (reviewed in Kudielka and Kirschbaum, 2005), in the analysis of parental data (fully adults) the sex of the quail was also included as a fixed factor. Interactions between fixed factors were also evaluated. In both parental and offspring data, induced humoral response was evaluated using a Poisson distribution while inflammatory response and H/L ratio were evaluated using gamma distribution. Fisher's least significant difference (LSD) test (alpha=0.05) was used to compare means when significant main effects were detected. The analyses were performed with an R (The R Foundation for Statistical Computing) user-friendly interface implemented in InfoStat (www.infostat.com.ar).

No significant main effects of CHS, ICH or their interaction were found on the SRBC antibody titre (F_1.69<1.16, P>0.28 in all cases). A significant sex effect was observed (F₁,₆₉=8.69, P=0.004), with females showing higher antibody titres than males (Fig. 1).

Analysis of the inflammatory response showed a significant main effect of CHS (F_1.72=23.24, P<0.001), with quails submitted to the stress protocol showing a lower induced inflammatory response than their non-CHS counterparts (Fig. 2). No significant ICH, sex or interactive effects were found (P>0.17 in all cases).

Analysis of the H/L ratio showed significant main effects of CHS (F_1.67=3.97, P=0.05) and sex (F_1.67=23.35, P<0.0001). Quail under the CHS protocol exhibited a higher H/L ratio compared with non-CHS quail (Fig. 3). Females exhibited higher H/L ratios than males. No significant ICH or interactive effects were found (P>0.11 in all cases).

Analysis of the SRBC antibody titre showed a significant main effect of the prenatal CHS condition (F_1.67=5.06, P=0.027) and no significant prenatal ICH or interactive effects between prenatal CHS and ICH conditions (P>0.81 in both cases). Thus, regardless of the prenatal ICH treatment, chicks whose parents were exposed to CHS showed a lower antibody production compared with their non-CHS counterparts (Fig. 4).

A significant main effect of prenatal CHS condition was detected on the induced inflammatory response (F_1.74=4.42, P=0.03). No significant prenatal ICH or interactive effects between CHS and ICH were found (P>0.76 in both cases). Post hoc tests showed that individuals whose parents were exposed to CHS exhibited a higher induced inflammatory response compared with non-CHS exposed groups (Fig. 5).

No significant main effects of prenatal CHS and ICH treatments or their interaction were found on the H/L ratio of the offspring chicks (CHS, F_1.74=0.94, P=0.33; ICH, F_1.74=0.57, P=0.45; interaction, F_1.74=0.39, P=0.53; Fig. 6).

Simultaneous effects of CHS and ICH as environmental challenges were evaluated on representative INE variables of mature adult Japanese quail. Moreover, to determine potential transgenerational effects of the mentioned challenges and whether they may programme the INE responses of the next generation, the same variables were analysed in both the parental generation and their offspring.

In the parental generation, exposure to CHS affected the INE interplay. Although the humoral response was unaffected, the swelling response to PHA-P was diminished, and the H/L ratio was elevated. According to previous research, the mechanisms to deal with a stressor of this nature would involve chronically elevated concentrations of stress mediators, evidenced by higher H/L ratios (Bedanova et al., 2007; Gross and Siegel, 1983). This directly influences the pro-inflammatory potential, downregulating the swelling response (Dhabhar, 2009; Elenkov, 2004). Moreover, under a variety of other energetically demanding situations such as chronic stressors (du Plessis et al., 2012), the metabolic and energetic trade-offs would be allocated to cope with the stressor, diminishing the immune responses of birds (Buehler et al., 2010; Piersma, 2011; Romero et al., 2009). Attenuated inflammatory responses might be beneficial in some particular cases, such as the mucosae. In the intestinal mucosa, a controlled and non-inflammatory environment is necessary in order to regulate the microbial communities (Kogut, 2019; Noguera et al., 2018). However, in our particular case, we can then hypothesize that the encounter with a potentially pathogenic microorganism during the last days of CHS might result in birds being more susceptible, owing to a diminished swelling response that depends on the communication and recruitment mechanisms involved in inflammatory potential.

Sex differences in INE traits are well documented. Sex hormones can affect the INE interplay, modulating the physiological responses (Davison, 2014; Duffy et al., 2000; Klein, 2000). Our results are aligned with the meta-analytic evidence provided in Foo et al. (2017) and are coherent with previous studies indicating that female sex hormones tends to modulate specific immune responses towards the Th2 profile, showing stronger humoral responses against potentially harmful threats in birds and mammals (reviewed in Roved et al., 2017). This pattern is also in concordance with what is proposed in Lee (2006), who suggested that females invest more in reproduction than males. Consequently, they tend to develop a Th2 profile that uses less energy both for its use and for development (Lee, 2006). Quail have been artificially selected to lay one egg a day (Padgett and Ivey, 1959), imposing a demand to females with no parallel in males. Considering the reproductive effort implied in this laying regime, the previous explanation might be suitable, leading to sex-dependent trade-offs.

Trade-offs between different induced non-specific immune responses and other physiological and behavioural processes such as reproductive success and breeding effort have been described in pied flycatchers (Ficedula hypoleuca), house sparrows (Passer domesticus) and northwestern song sparrows (Melospiza melodia morphna) (Bonneaud et al., 2003; Ilmonen et al., 2003; Owen-Ashley and Wingfield, 2006; Romero et al., 2009). However, and contrary to what we expected, our results suggest that simultaneous exposure of adult quail to the ICH and the CHS did not evocate any INE trade-offs, at least at the level evaluated. One possible explanation for this would be the absence of resource limitation (Sandland and Minchella, 2003). Ad libitum access to food and water would mean that quail did not invest extra resources satisfying their energy needs while developing the immune response associated with the ICH (Klasing, 2004). These results and their explanation have a similar parallel in long-distance migrant shorebirds (red knots, Calidris canutus islandica) subjected to experimentally manipulated temperature treatments (Buehler et al., 2012). The authors suggest that because birds had free access to food during temperature manipulation, conditions may not have been severe enough to warrant trade-offs between physiological traits.

Quail, as any other animal, have a limited amount of energetic resources to deal with challenges (Piersma, 2011; Romero et al., 2009). Experiencing energetically demanding physiological challenges would be detrimental for the individuals, considering that the resources needed to deal with the stressor would not be available in the long term (Piersma, 2011). Owing to its non-pathological nature and the absence of effects on the evaluated variables, the ICH used might be considered a mild immune challenge (Guibert et al., 2011). Our results also suggest that quail INE interplay would be more susceptible to chronic exposure to elevated temperatures (for which effects were registered) rather than to the attenuated Salmonella inoculum.

In our study, the energetic and physiological balances were already altered by the exposure to the CHS as previously depicted. However, the conditions would not have increased in severity when an ICH was added. This is contradictory with some literature suggesting that immune challenges increase the physiological impairments induced by chronic stressors (Dhabhar, 2009, 2014). In contrast, our results back up those of another study in another domestic species, Gallus gallus, where immune challenge was not processed as a stressor, i.e. Newcastle disease vaccination per se was unable to change corticosterone serum levels (Honda et al., 2015). An attenuated Salmonella inoculum would belong to the types of immune challenge that behave in a non-additive manner with CHS in the domestic model used. Nevertheless, parental birds must be considered to be in reproductive state. Eggs were obtained from every parental pair, finding no laying impairment, as previously reported both for stress and immune challenges (Lara and Rostagno, 2013; Mashaly et al., 2004). Carryover effects occur in any situation in which an individual's previous history and experience explains their current performance in a given situation (O'Connor et al., 2014). Finding no effects of ICH in the time frame evaluated on the sampled variables does not assures the absence of carryover effects associated to be evidenced later on. If these delayed fitness effects occur, they could potentially affect birds' performance in future laying periods (Senner et al., 2015).

Keyhole limpet haemocyanin and lipopolysaccharide have previously been used as immune challenges in the parental generation, with several transgenerational effects being described. For example, alterations in the mediators of the HPA axis (Merrill and Grindstaff, 2015), enhancement of antibody production (Grindstaff et al., 2006), and diminished growth rates and reduced learning performance (Grindstaff et al., 2012) were observed in challenged offspring. In our case, finding no consequences of the ICH on the INE responses might reflect no changes in the activity of the HPA axis, differing with the results obtained by Merrill and Grindstaff (2015). This might be due to the intensity or strength of the challenge they used, keyhole limpet haemocyanin, similar to the CHS evaluated in our study, being strong enough to program the INE function via HPA-axis-dependent effects. In contrast, the enhanced antibody production described in Grindstaff et al. (2006) might be an unspecific response and, therefore, it is still plausible that prenatally exposed chicks would respond better to the exposure of Salmonella and other pathogens later in life. Unfortunately, testing this hypothesis is beyond the scope of our experimental design.

The exposure to high environmental temperatures in the parental generation increased the pro-inflammatory potential and diminished the specific humoral response to SRBC in their offspring, suggesting a transgenerational effect. The H/L ratio can also be considered a constitutive defence estimator when health status is not affected (Lee, 2006; Matson et al., 2006). The fact that we did not find differences between the ratios suggest that pro-inflammatory potential differences are driven primarily by an increment of the previously described mechanisms associated with the induced cell-mediated response involved in the PHA-P swelling. Our experimental design unfortunately does not allow us to deepen into whether the observed traits are actually adaptive (Nettle and Bateson, 2015). Therefore, further studies of the INE traits along the ontogeny would be needed to evaluate the offspring behaviour and responses once they reach adulthood.

The similar H/L ratios observed in the offspring are in agreement with values that are classified as basal in other studies (Nazar and Marin, 2011; Nazar et al., 2018b). Consequently, we could suggest that none of the treatment groups was influenced by the stress conditions associated with the offspring rearing. The programmed phenotype showing increased pro-inflammatory potential and diminished specific humoral response would respond to future immune challenges (such as mycobacterial or similar infections), requiring the elimination of infected cells using the innate branch of the immune system in a more efficient way (Altare et al., 1998; Bowling et al., 2018; Buehler et al., 2010; Elenkov, 2008; Rauw, 2012; Vinkler et al., 2010). However, an impaired humoral response could be disadvantageous when responding to other environmental challenges such as parasites or when needing to create immunological memory (Hanssen et al., 2004; Härtle et al., 2014). Developing and using an induced inflammatory response might be more energetically costly compared with the development and use of a Th2 specific humoral response (Lee, 2006). The energy balance of immune expenditure in the programmed phenotype might imply that their basal costs are different from the other groups.

Overall, new interrogations may arise when considering our results altogether. For example, would chicks whose parents were exposed to challenging environmental conditions cope better if they share the influence of a common environmental challenge? A potential answer could be given combining environmental matching and developmental programming. If a continuous environmental demand in the form of elevated temperatures affects the offspring, the Th1-biased programming observed could help mitigate and/or counteract the polarization towards a Th2 profile that chronic stress responses would classically induce (Elenkov, 2008; Glaser et al., 2001; Shini et al., 2010; Webster et al., 2002). This scenario would be an example of programming in which maternal stress effects adaptively prepare offspring for a stressful or rigorous future environment, increasing the empirical proof to the environmental/maternal-matching hypothesis (Chin et al., 2009; Sheriff and Love, 2013).

Findings suggest that CHS exposure influences the INE interplay, establishing trade-offs within the immune system of adult quail (Lee, 2006). These trade-offs would leave birds particularly susceptible to future immune threats cleared through inflammatory processes, but better prepared to deal with high parasitic loads. CHS induced transgenerational effects, but the adaptability of offspring's programming would depend on the encountered environment. It is important to consider that, unlike the ICH, the response to CHS shares common physiological pathways with a variety of well-studied stressors. Many of them are strictly associated with production environments such as high stocking density, social mismatches or maintenance chores (de Kloet, 2003; Jones et al., 2005; Puvadolpirod and Thaxton, 2000). In contrast, stressors may also be found in ecological contexts in the form of high predation risk, low food availability, territorial competition, long-term migrations, etc. (Buehler et al., 2010; Love et al., 2013; Morales et al., 2019; Sheriff and Love, 2013; Sheriff et al., 2017). Therefore, these results could be expected under both intensive production and ecologically relevant conditions.

We are grateful for the technical assistance of Julia Ortiz, Pablo Prokopiuk and Dario C. Arbelo during the development of this study. We also thank two anonymous reviewers, whose comments helped to improve the quality of this paper.