One of the proposed solutions for food security is to adopt a local food production that is more sustainable and affordable to meet nutritional requirement. Therefore, a small-scale production of native chickens, e.g., NT chickens, has been suggested as one of the approaches for increasing local food availability and improving food security [
12]. However, due to their slow growth rate, the production of native chickens requires a longer rearing period of time and additional resources (e.g., feed, water, and land) for the birds to reach their market weight. Still, the meat yield is lower than that of the fast-growing high-performing commercial broilers. Hence, medium-growing chickens, yielded from a crossbreeding between high-performing and slow-growing strains may overcome such issues of their parents [
13].
To define biological pathways associated with thermal stress response in BR, NT, and H75 chickens, transcriptome of P. major muscle (breast muscle) of thermal-stressed group was compared with their control counterparts. The findings of each breed will be discussed in detail below.
Fast-growing commercial broilers
The results highlighted the KEGG pathways associated with glucose homeostasis and energy production, including glycolysis/gluconeogenesis, glucagon signaling and HIF-1 signaling, as the enriched altered pathways in the BR breast muscles (
Supplementary Table S4). The findings were agreed with metabolic shifts in the birds exposed to thermal stress [
10]. This was not beyond our expectation. Indeed, it is extensively reported that upon an exposure to heat stress, chickens reduce their feed intake to avoid metabolic heat production [
13,
14] and to reserve resources for cellular adaptation to maintain homeostasis [
14]. Glucose is a primary source of energy supply. Increased levels of plasma glucose [
3] and glucocorticoids [
15] were reported among commercial broilers exposed to heat challenges. Awad et al [
3] suggested that the elevated plasma glucose was a consequence of increased plasma glucocorticoids in order to reserve glucose for the brain during stress situations. In fact, upon exposure to heat stress, muscle glycogen utilization increased with no change in glucose uptake by the muscle and decreased carbohydrate oxidation [
3,
14,
15]. This might lead to reduced glycogen and increased lactate in the breast muscle of the stressed BR (
Figure 4). The breakdown of glycogen and accumulated lactate could lead to an acidic condition within the poultry breast muscle [
16]. Such conditions ultimately impacted meat quality of the stressed BR [
10]. In addition, we observed that
LDHA (log
2FC = −1.3 to −3.4), fructose-bisphosphate aldolase A (
ALDO, log
2FC = 1.2) and alpha-enolase isoform X3 (
ENO1, log
2FC = −2.5) were mapped into glycolysis/gluconeogenesis and HIF-1 signaling (
Supplementary Table S4). The findings suggested the closed interplay of those pathways in response to the cyclic thermal stress within the breast muscle of BR.
The KEGG biosynthesis of amino acids was among the enriched altered pathways impacted by the cyclic thermal stress in the BR breast muscles (
Supplementary Table S4). Differential expressions of genes encoding glycolytic enzymes, including
ALDO,
ENO1, and bisphosphoglycerate mutase isoform X1 (
PGAM, log
2FC = −1.1), were also mapped into biosynthesis of amino acids. This could be due to the interrelationships between glucose and protein as the precursor pools for energy production. Additionally, the carbon chains for some amino acids are the intermediates of glycolysis. The current results are consistent with previous studies showing an increased muscle protein degradation upon exposure to thermal stress [
7–
9]. The catabolism of protein and amino acids has been hypothesized to reflect an energy demand for the animals to counter heat stress [
7,
9].
Changes in gene expression patterns in the stressed BR also suggested the oxidative stress condition within the breast muscle. Cystathionine-β-synthase (
CBS, log
2FC = 2.2), glutathione S-transferase (
GSTA4, log
2FC = 1.4) and ribose-5-phosphate isomerase (
RPIA, log
2FC = 1.3) were up-regulated in the thermal-stressed BR muscles (
Supplementary Table S1;
Table S4). The CBS enzyme catalyzes metabolisms of sulfur-containing amino acids, i.e., methionine and cysteine, through an irreversible conversion of homocysteine to cystathionine, the upstream process of glutathione formation [
17]. Increased
CBS and
GSTA4 abundance in the stressed BR might imply an autocorrective function of CBS to oxidative condition through glutathione system [
17]. As for RPIA, the enzyme catalyzes the process of histidine biosynthesis and formation of nucleotides or glycolytic intermediates depending on body condition [
18]. The crucial role of RPIA in mediating ROS level, apoptosis and autophagy was addressed in lung cancer cells [
18]. Furthermore, differential expression of gene encoding glutamine-fructose-6-phosphate aminotransferase (isomerizing) 2 isoform X2 (
GFPT2, log
2FC = −1.1) suggested the altered hexosamine pathways in the thermal-stressed BR. The pathway is activated in response to an intracellular oxidative stress linked to hyperglycemia and insulin resistance [
19]. The end product of hexosamine pathway, uridine diphosphate N-acetyl glucosamine, can further be converted to O-linked N-acetyl glucosamine, a crucial intracellular signal transductor that enhances the cyto-protective effects in stressed cells [
19].
PI3K/Akt signaling pathway, phagosome and focal adhesion are also among the top enriched altered pathways for BR samples. Differentially expressed integrin alpha 11 (
ITGA11, log
2FC = −1.3), cartilage oligomeric matrix protein (
COMP, log
2FC = −2.0) and death domain-containing membrane protein isoform X3 (
NRADD, log
2FC = −1.2) were mapped into those pathways. The proteins encoded by those genes facilitate cell communication in different manners.
COMP encodes a secretory non-collagenous extracellular matrix pentamer glycoprotein, also known as thrombospondin-5, that activates PI3K/Akt signaling pathways when it binds with an integral membrane receptor. Such signal has shown to suppress apoptosis and promote proliferation and migration in cancer cells. Integrins and NRADD are transmembrane receptors activated by the binding of their cognate ligands to the extracellular signals [
20]. Down-regulation of
ITGA11 was found in chickens exposed to oxidative stress via heat stress and toxic minerals [
4]. NRADD, belonged to nerve growth factor receptor (NGFR) superfamily, allows the cells to receive apoptotic signals, hence activating cell death when the cells are exposed stress [
20]. Increased abundance of major histocompatibility complex (MHC) class II beta chain (
BLB1, log
2FC = 3.1) in the phagosome pathway may imply the clearance process of cellular debris upon the oxidative damages and apoptosis. The role of nicotinamide adenine dinucleotide (NAD) phosphate (NADPH) oxidase 2 (NOX2), encoded by cytochrome b-245 beta chain (
CYBB, log
2FC = −1.2), in regulating apoptotic cell removal alkalization of phagosome has been reported [
21].
The current RNA-Seq findings for BR breast muscles un derlined metabolic adaptation as the key altered pathways in response to the cyclic thermal stress. The occurrence of cellular damage, potentially from oxidative stress, and cellular debris clearance was emphasized in the BR breast muscles. Similar molecular alteration was recently underlined and associated with the impact of chronic heat stress on reduced breast meat quality in Arbor Acres broilers [
10].
Slow-growing Thai native chickens
Previous report indicated that indigenous chickens can tolerate and adapt to high ambient temperature to a greater extent than fast-growing commercial broilers [
5]. No adverse effects of heat stress on production performance and meat quality are extensively reported. Nevertheless, previous studies showed that thermal challenges stressed the slow-growing native birds as well [
4,
13]. Deviated plasma triglyceride, plasma glucose and serum total protein were also reported in the 21-day-old slow-growing chicks exposed to heat stress (38°C, 4 h daily) for 3 days [
22]. Thermoregulatory capabilities of slow-growing chickens indeed vary among the different strains [
23]. In this study, decreased glycogen content in the breast muscle of stressed NT but no change in lactate content was detected (
Figure 4).
Based on the current RNA-Seq (
Supplementary Table S5), glycolysis/gluconeogenesis, glucagon signaling, metabolisms of proteins and amino acids, and HIF-1 signaling were the top enriched pathways identified for the NT samples. In addition to those pathways, metabolisms of fructose, mannose, pyruvate, and nucleotides were mediated within the NT breast muscles. Besides, as observed in the BR, the signaling pathways (
Figure 3), including PI3K/Akt, focal adhesion, MAPK, and Ras signaling, were altered in NT breast muscles. Even though the list of DETs (
Supplementary Table S5) mapped into those pathways was slightly different from that of BR samples, the results suggested some similarities in metabolic shifts and the cellular signals between BR and NT birds when they were facing the cyclic thermal challenge.
Interestingly, oxidative phosphorylation is among the en riched altered pathways for NT chickens but not for the BR group (
Figure 3). The DETs mapped into this pathway include ATP synthase F0 subunit 6 (
ATPeF0A, log
2FC = 5.7), cytochrome c oxidase subunit 2 (
COX2, log
2FC = 2.7) and COX subunit 3 (
COX3, log
2FC = 1.7). In addition, NADH dehydrogenase subunit 1 (log
2FC = 1.6), subunit 2 (log
2FC = 2.8), subunit 3 (log
2FC ranging from 1.4 to 3.2) and subunit 4 (log
2FC ranging from 1.4 to 3.5) along with NADH-ubiquinone oxidoreductase chain 5 isoform X2 (log
2FC = −1.4) were differentially expressed between thermal-stressed and control NT birds. Those genes encode subunits of the Complex I of the mitochondrial electron transport chain (ETC). Oxidative phosphorylation is the final step in cellular respiration occurred in mitochondria. Up-regulation of those genes suggested that the Thai NT chickens modified mitochondrial ATP synthesis in response to the cyclic thermal challenge.
Furthermore, vascular endothelial growth factor (VEGF) D isoform X1 (
VEGFD, log
2FC = 1.2) were up-regulated within the thermal-stressed NT. VEGF is a potent angiogenic factor, regulating a wide range of responses including metabolic homeostasis, cell proliferation and migration, as well as formation and maintenance of blood vessel structure. Transcript abundance of fibroblast growth factor (FGF) 16 isoform X1 was also increased in the stressed NT (log
2FC = 1.2). Together, VEGF and FGF signaling might promote angiogenesis, improving oxygenation in the stressed NT breast muscle [
24]. In addition, increased
VEGF and integrin subunit beta 3 (
ITBG3) with decreased myosin light chain isoform 9 (
MYL9) were associated with immune suppression of Jersey cattle under thermal stress [
24]. In this study, we found differential expression of myosin light chain 1 skeletal muscle isoform X1 (
MYL1, log
2FC ranging between −1.4 to 9.0, recognized as MYL9 by KEGG) and integrin alpha-D isoform X13 (
ITGAD, log
2FC = 2.5) in the breast muscle of NT chickens. Additionally, serine/threonine-protein kinase isoform X1, encoding the p21 activated kinase (PAK) 6 was up-regulated (
PAK6, log
2FC = 1.4). PAK6, a member of the downstream effectors of Ras-related Rho GTPase Cdc42 and Rac, facilitates cytoskeletal organization, cell motility and apoptosis in response to stress. Such transcriptional modifications of those genes in association with toll-like receptor and T cell receptor signaling pathways might attenuate the adverse effects [
25], particularly from oxidative stress, in the NT muscle. Therefore, the properties of NT breast meat did not significantly differ between the control and the thermal-stressed groups [
4].
Medium-growing crossbred H75
The numbers of DETs and the altered pathways identified for H75 were higher than that of BR and NT. Nonetheless, the top altered metabolic pathways and signaling pathways identified for H75 were similar to those found for BR and NT (
Supplementary Table S6). In addition to metabolic shifts, mitochondrial
ETC genes, including, NADH-ubiquinone oxidoreductase (log
2FC ranging between 1.1 and 1.6), cytochrome b (
CYB, log
2FC = 1.1),
COX2 (log
2FC ranging between 1.1 and 1.3),
COX4 (log
2FC = 1.5), were differentially expressed between stressed H75 and their control counterparts (
Supplementary Table S3;
Table S6). Changes in the abundance of those genes might imply adaptive mitochondrial ATP synthesis in the stressed H75 breast muscle in a similar manner as observed in the NT breasts. No differences in either muscle glycogen or lactate were observed in the breast muscles between control and stressed NT (
Figure 4).
Interestingly, adipocytokine signaling, insulin signaling, autophagy, mammalian target of rapamycin (mTOR) and NF-κB signaling pathways were specifically listed for the H75 (
Figure 3). Alteration of those pathways was identified in broilers exposed to chronic heat stress [
7]. Those pathways were shown to involve in fat deposition in broilers [
26]. Zhang et al [
26] addressed that phosphoenolpyruvate carboxykinase 2 (
PCK2), acetyl-CoA carboxylase 1 alpha and beta (
ACACA and
ACACB), and
AMPK gene family are the key genes regulating fat deposition in muscle and fat tissues of broilers. In this study, we observed an increased
ACACB (log
2FC = 1.6) and decreased 5′-AMP-activated protein kinase catalytic alpha subunit (
PRKAA2, log
2FC = −1.1), a catalytic subunit of AMPK, in the stressed H75. However, in our previous study, no significant differences in abdominal fat and fat content were observed in the carcass and breast meat, respectively, between control and H75 chickens [
4]. The score for adipose infiltration of stressed H75 breast muscle was not statistically different from that of their control counterparts. The reason for the discrepancy needs further investigation. The results, however, suggested the role of lipid metabolism in metabolic modification in H75 exposed to the cyclic thermal stress. In addition to that, adipocytokines released from adipose tissues are often altered in obese subjects, of which the levels of proinflammatory cytokines and ROS are increased. Signals from adipocytokines might assist inflammatory responses and maintenance of systemic energy metabolism under the oxidative stress condition within the muscle of the stressed H75.
A dynamic interplay between autophagy and mTOR sig naling pathways also regulates a recycle of damaged organelles and macromolecules for energy and building blocks for normal growth. Such action is dictated by nutrient and energy status with the coordination of AMPK signaling. Recently, Tang et al [
27] showed that heat stress (35°C±2°C, 8 h daily, 7 and 14 days) decreased autophagic activity in the liver of broilers, corresponding with liver inflammation. In addition, similarly observed in NT, toll-like receptor and T cell receptor signaling pathways were altered in H75 (
Figure 2B). In this study, those altered pathways along with NF-κB signaling were linked together by differential expression of phosphatidylinositol 3-kinase regulatory subunit alpha isoform X1 (
PIK3R1, log
2FC = −1.4), NF-κB inhibitor alpha (
NFKBIA, log
2 FC = 1.5), dual specificity mitogen-activated protein kinase kinase 4 (
MAP2K4, log
2FC = −1.2),
MAP2K6 (log
2FC = −1.1),
PAK2 (log
2FC = −1.1) and novel protein kinase C theta type (
PRKCQ, log
2FC = −1.2) (
Supplementary Table S3). PAK2 involves in heat shock-induced apoptotic cell death process in mammalian cells. Incorporated signals of those pathways are essential for precise regulation of the immune system against infections and inflammatory diseases. Hence, those unique altered pathways might be the key system assisting the adaptation and maintenance of the H75 chickens upon exposure to the cyclic thermal challenge, leading to less detrimental effects on breast meat quality compared with BR [
4].
Comparing differential transcriptome profiles among breeds
Although different lists of DETs among the three breeds were observed, the overall data ascertained modification of metabolic homeostasis, potentially under cellular oxidative stress condition, among all breeds [
7]. Carbohydrate metabolisms were impacted in all breeds with some differential mechanisms. Glycolysis/gluconeogenesis was enriched in all breeds. It has been reported that, in poultry, thermal stress led to increased mobilization of glucose to supply energy [
10]. The muscle glycogen would undergo breakdown at a greater rate in the stressed chickens leading to lower muscle glycogen and higher lactate [
10,
15,
16]. Our results underlying decreased muscle glycogen and increased lactate in the breast muscle of the stressed BR (
Figure 4) agreed with previous studies [
10,
15]. Among NT, reduced muscle glycogen was also detected when the birds were under the cyclic thermal stress (
Figure 4A). However, no differences in lactate was observed between the NT control and stressed groups. The results agreed with the current transcriptome analysis suggesting an adaptive mechanism, particularly the promoted angiogenesis, within the NT that attenuated muscle lactate accumulation. On the contrary, no differences in muscle glycogen and lactate were found between control and stressed H75. The explanation required further investigation.
Altered metabolisms of protein and amino acids were highlighted. The analysis of enriched KEGG pathways revealed the association of the DETs in regulation of actin cytoskeleton along with enriched GO term related regulation of muscles. The results suggested the increased muscle protein degradation upon exposure to thermal stress [
7–
9]. The cyclic thermal stress also affected signal transduction, inflammatory response, immune system, programmed-cell death as well as the process of cellular function (e.g., cell growth, cell differentiation and cell migration). The upstream signals for those biological response appeared to be centered around the PI3K/Akt signaling in coordination with mTOR, AMPK, and MAPK signaling (
Figure 5). The PI3K/Akt, MAPK and AMPK pathways are essential for metabolic control and also involved in fundamental cellular processes, e.g. cell growth and cell survival. Crosstalk among those pathways when cells were under stress has been demonstrated [
10]. The findings were in line with previous reports [
7,
10] and suggested the important roles of those pathways in chicken breast muscle in response to the cyclic thermal challenges.
Interestingly, alteration of AMPK signaling was found for BR and H75 but not for NT samples. Only stearoyl-CoA desaturase (
SCD1, log
2FC = −2.2) was listed in the AMPK signaling pathway for BR. Decreased
SCD1 expression was found during the onset of cell death due to cytotoxic stress. SCD1-derived lipokine was shown to attenuate cell damage due to cytotoxic stress by limiting cell apoptosis through inhibition of MAPK signaling and rerouting cell fate to autophagy. Since heat stress was shown to induce oxidative stress which further led to MAPK-mediated cell apoptosis, the results suggested extensive apoptosis in the BR muscle which might negatively affect breast meat quality [
4,
10].
In contrast to BR, more DETs were involved in the AMPK signaling pathways for H75. A similar trend was also found for the regulation of the actin cytoskeleton. AMPK involves in mitochondrial homeostasis by removing the defective mitochondria and stimulating de novo mitochondria synthesis, hence controlling mitochondrial adaptability under stress. Skeletal muscle of meat-type chickens exhibited mitochondrial oxidative phosphorylation of skeletal muscle exhibited a higher efficiency in comparison to that of laying type chickens. However, the rapid metabolic rates in the early life of the fast-growing broilers might exert constraints on oxidative capacity among modern broilers. Therefore, the ability of BR chickens to adapt to the oxidative stress under the cyclic thermal challenge might be to a much lesser extent.
An exposure of skeletal muscle to heat stress has been ex tensively associated with increased expression of heat shock protein (HSP), particularly
HSP70 and
HSP90 [
28]. Those proteins function to protect the cells against the stress through inducing a series of cellular defense mechanisms, including antioxidant system, NF-κB and PI3K/Akt signaling. However, in this study, no significant differences in
HSP abundance were observed in either BR or NT. Rimoldi et al [
29] also observed the lack of increased
HSP70 and
HSP90 in hepatic tissues of broilers exposed to 4-week heat stress. They hypothesized that either the expression levels of those genes had already declined upon a long-term heat exposure or
HSP mRNA was lost due to cell lesion. The speculation of Rimoldi et al [
29] appears to agree well with our observation for BR and NT of which the changes in the HSP downstream pathways, particularly PI3K/Akt and antioxidant enzymes, were observed. However, further experiments are required to support this hypothesis. As for H75, changes in transcript abundance of heat shock cognate protein HSP 90-beta isoform X1 (
HSP90AB1, log
2FC = −1.1), heat shock factor protein 2 (
HSF2, log
2FC = 1.1) and α-crystalline B chain isoform X1 (
CRYAB, log
2FC = 4.2), which is a small HSP, were observed between thermal-stress and their control counterparts. Inconsistent with our transcriptome results, overexpression of
CRYAB was shown to increase survival of myocardial cells subjected to heat stress through the processes of actin cytoskeleton stabilization and prevention of apoptosis.
Comparing blood transcriptomes between chickens ex posed to heat stress and control, Kim et al [
14] also found focal adhesion among the enriched biological pathways. They speculated that focal adhesion, together with Ca
2+ signaling, contributed to an upregulation of heat shock protein genes which further activate MAPK signaling pathways [
14]. However, although focal adhesion was underlined in all breeds, no alteration in Ca
2+ signaling was observed in the BR samples. Taken together, it may be reasonable to speculate that those pathways may contribute to adaptive mechanisms in the H75 breast muscle under the cyclic thermal stress. The BR chickens were able to activate stress-response pathways to a lesser extent in comparison to those of H75 and NT. Further up-close investigation remains to be elucidated.
In this study, the effects of the cyclic thermal stress were focused on the later age (three weeks before market age) of BR, NT, and H75. This is due to the fact that the birds with higher age and weight showed more thermal sensitivity [
5,
6]. Hence, the loss of chickens at this stage would bring about a significant economic loss. The differences in age of those chickens are, however, worth noting. In addition, transcriptome analysis captures a snapshot of the total transcripts presenting in the cells under a certain circumstance [
30]. Either the DETs showed up- or down-regulation in this study might not precisely reflect their immediate responses to the stress. Further investigation remains to be elucidated for such aspects. However, it is certain that the molecular pathways of the DETs involved would be altered by the stress. In addition, inflammatory response is majorly reported in broilers upon heat exposure [
30]. Nonetheless, in this study, only NT and H75 showed pathways associated with inflammation. The discrepancies might be due to the different stress intensity between this study and the previous ones.
In conclusion, the current RNA-Seq analysis provides in sights into biological responses in breast muscles of BR, NT, and H75 chickens exposed to the cyclic thermal stress. The stress triggered metabolic shifts in all breeds with the signals centered around PI3K/Akt signaling, focal adhesion, and MAPK signaling. Different molecular signal transduction patterns were observed. The findings underlined the key roles of AMPK, MAPK signaling and regulation of actin cytoskeleton in H75 and NT chickens as an adaptive system against cyclic thermal stress.