First transcriptomic insight into the working muscles of racing pigeons during a competition flight

Background The currently known homing pigeon is a result of a sharp one-sided selection for flight characteristics focused on speed, endurance, and spatial orientation. This has led to extremely well-adapted athletic phenotypes in racing birds. Methods Here, we identify genes and pathways contributing to exercise adaptation in sport pigeons by applying next-generation transcriptome sequencing of m.pectoralis muscle samples, collected before and after a 300 km competition flight. Results The analysis of differentially expressed genes pictured the central role of pathways involved in fuel selection and muscle maintenance during flight, with a set of genes, in which variations may therefore be exploited for genetic improvement of the racing pigeon population towards specific categories of competition flights. Conclusions The presented results are a background to understanding the genetic processes in the muscles of birds during flight and also are the starting point of further selection of genetic markers associated with racing performance in carrier pigeons. Supplementary Information The online version contains supplementary material available at 10.1007/s11033-024-09566-7.


Introduction
The currently known homing pigeon is a result of crossing many lines of pigeons and a sharp one-sided selection for flight characteristics focused on speed, endurance, and spatial orientation.The purpose of selection focuses on breeding an extraordinarily motivated to get a home bird, that in the shortest possible time in various weather conditions at an average speed greater than 70 km/h will cover a certain distance to the loft [1].
For over 200 years of sports competition, breeders developed a highly specialised breed of pigeon called Racing Homer, paying great attention to improving their ability to approximate the direction to home from foreign locations defeating hundreds of kilometers, avoiding hazards and coping with unexpected weather conditions [2].The competition aims to compare bird individuals' flight performance by the speed of returning to the loft.Contests are held at various distances from, short ranging from 95 km to long exhaustive marathons with more than 700 km to cope [3].When displaced to an unfamiliar location, homing pigeons apply a spatial navigation system and outstanding physiological adaptations to returning to the loft [4] including cardiorespiratory properties [5], energy expenditure supported by anaerobic and aerobic metabolic pathways with efficient circulatory system and oxygen transport [6,7].
Avian flight is powered primarily by large pectoralis muscles (m.pectoralis pars thoracicus which accounts for up to 11% of total body mass and generates up to 95% of the power used for flight.The large m.pectoralis extends from the sternum, clavicle, and ribs to the humerus, and consists of two anatomical parts, the sternobrachialis, and the thoracobrachialis, separated by an aponeurotic central tendon [8].The fibre type composition contains mainly IIb type, referred to as fast oxidative (~ 85% in pigeons) adapted to anaerobic glycolytic metabolism [11].Fibers possessed representative sarcomere structures, however, with shorter resting sarcomere lengths compared to mammalians [8].
The avian flying muscles introduce the most energetically expensive muscle work with the highest mass-specific metabolic rates in vertebrates, in comparison to exercising mammals, flapping is energetically more costly than running [6,9].To cope with these demands, several mechanisms have been described.Fuel selection of avian muscles during locomotion supports lipid oxidation, with minimum changes in blood glucose concentrations.Pigeons flying at their maximum rate of O 2 uptake (VȮ 2 , max ) and a respiratory quotient (RQ) at 0.73 indicating dependence on fat oxidation [10] whilst mammals exercising at similar VȮ 2 , max uptake reach RQ at 0.9 reflect dependence on carbohydrates and finally induce fatigue, with low muscle glycogen and blood glucose [11].However, birds maintain a very high plasma glucose concentration (1-2 times) compared to mammals of equal body mass but with no harmful physiological effect.It is believed that endogenous antioxidant mechanisms such as free radical scavenging, DNA protection and uric acidmediated inhibition of lipid peroxidation help with homeostasis maintenance [12].
The dominant role and large size of the m.pectoralis, enable an assessment of adaptation and muscle function tailored to meet the mechanical and energetical requirements of exercising flight, compared to exercising mammals of different body masses and aerobic capabilities [13].
Several studies have been undertaken to find molecular pathways modified in skeletal muscles during exercise for example in humans and horses.It has been established that repeated sets of exercises lead to new basal levels of gene expression [14][15][16].The molecular mechanism underlying the genetic adaptation during pigeon flight and training remains poorly understood.Since the sport and breeding of racing pigeons is a profitable business covering areas such as nutrition, supplementation, and genetic markers, the aim of the present study was transcriptome profiling of pigeons m.pectoralis muscle, collected from untrained birds and trained birds after competing 300 km competition flight with the use of high throughput RNA-sequencing.

Animals and study design
The present study was performed on 13 muscle samples of m. pectoralis collected from 13 racing pigeons (Columba livia).Samples were collected from adult birds never trained for racing (k; n = 5) and birds after competing in a 300 km race (f; n = 8), who had undergone earlier flight training.All birds were bred, and raised on a private loft owner, at the same location, with the same environmental and feeding conditions.The E group was basketed the day before and transported to the place of release (300 km away from the loft).
The racing group was released at 6 a.m.Upon their return, both groups were sedated and euthanized using the cervical dislocation method.The tissues were secured in liquid nitrogen and stored at −80 °C for further analysis.
The Animal Care and Use Committee of the Institute of Pharmacology, Polish Academy of Sciences in Cracow reviewed the experiment protocol.All procedures were conducted following the guidelines for animal care and by qualified staff.

m.pectoralis whole transcriptome sequencing
The total RNA was isolated using TriReagent (Ambion, Life Technologies) and according to the method described by Chomczyński [17].The muscle samples were homogenized with the use of zirconium oxide beads (diameter 0.5 mm) in BulletBlender homogenizer (Next Advance Inc., USA).The concentration and quality of RNA were estimated on TapeStation 2200 using RNA Screen Tape (Agilent Technologies, Warsaw, Poland).The samples with RIN (RNA integrity number) ranging from 7.7 to 10.0, were sent to sequencing by a commercial company (CeGat gGmbH) on NovaSeq 6000 (Ilumina) apparatus, and 100 bp (± 50 bp) paired-end reads were generated.

Data analysis
Demultiplexing of the sequencing reads was performed with Illumina bcl2fasq (2.20).Adapters were trimmed with Skewer version 0.2.2 [18].The FastQC software (v0.11.5) was applied to quality control (QC).The raw reads were aligned to the Columba livia genome (assembly 2.1_Colliv) as a reference, and the whole procedure was followed by alignment parameters from ENCODE3's STAR-RSEM pipeline.The quantification of genes and transcripts was achieved using htseq-count software.The quality control metrics across aligned reads per sample were obtained using RNA-SeQC software [19].To detect the differentially expressed genes (DEGs) and to show contrasts between the considered conditions, Bioconductor/R-project package, DESeq2 has been applied [20].The differentially expressed genes were those which reached adjusted p-value < 0.05 (with the use Benjamini/Hochberg method) and fold change ≥ 1.2.
Significantly expressed genes were subjected to functional annotation and pathways enrichment analysis using KOBAS server v3 with the corrected p-value < 0.05 with the C.Livia reference.The enriched analysis of independent GO aspects (molecular function, biological process, cellular component) was performed using the Gene Ontology enrichment analysis tool with implemented PANTHER databases.The significant GO terms were those which exceed FDR (False Discovery Rates) < 0.05 [21].

Validation of RNA-seq with qPCR
cDNA was synthesized from 0.5 μg of the same total RNA used in RNA sequencing using the High-Capacity RNAto-cDNA™ Kit (Applied Biosystems) as per the manufacturer's protocol.qPCR reactions were run on a QuantStudio7 (Applied Biosystems, Life Technologies) in a 10-μl reaction containing 0.5 μl of cDNA template, 5 μl of 2 × SYBR Green Master Mix, 0.3 μl of each primer (10 μmol/μl) and 4.4 μl nuclease-free water.The amplification program consisted of one cycle at 95 °C for 10 s, followed by 40 cycles at 95 °C for 15 s and 55 °C for 34 s.The qPCR reactions for HSF2BP, ADIPOQ and PPARD gene were performed with three biological replicates.Relative gene expression was normalized to the expression of pigeon ACTB and calculated with the 2 − ΔΔCT method [22].Primer pairs for these genes were designed using Primer3 version 4.0.0 and are listed in Supplementary Table S1.The expression levels of the genes obtained in RNA-seq and qPCR were compared with the Pearson correlation coefficients.

Sequencing data
The detailed next-generation sequencing statistics for each library are shown in Supplementary Table S2.On average, 14,564,488 mapped read pairs were identified in the pectoralis muscle of racing pigeons and an average of 689,230 were paired with multiple alignments.An average of 10,954,222 read pairs were annotated to the genome (2.1_ Colliv).The presented study was submitted to the international functional genomics database-GEO database (Gene Expression Omnibus; NCBI) and assigned GEO accession number-GSE222537.

Identification of differentially expressed genes, functional annotation, and pathway analysis
The muscle transcriptomes of both groups were compared to detect differentially expressed genes under a 300 km competition race condition.The differentially expressed genes (DEGs) selected for further analysis were those which reached adjusted p-value < 0.05 (with the use of the Benjamini/Hochberg method) and fold change ≥ 1.2.With the threshold of padj < 0.05, we identified a total of 502 differentially expressed protein-coding transcripts, of which 103 were identified as novel genes.330 were up-regulated and 171 were down-regulated (Fig. 1; Table 1).
Next, the significantly expressed genes were subjected to functional annotation and pathways enrichment analysis using KOBAS server v3 with the corrected p-value < 0.05 [21].The analysis of all significantly expressed DEGs allowed for the detection of 13 significantly deregulated pathways.Among others, the top five significantly overrepresented pathways were autophagy animal, mTOR signalling pathway, lysosome, metabolic pathways, and necroptosis.For top underrepresented pathways were those involved in focal adhesion, ECM-receptor interaction, and Gap junction, respectively (Table 2).
After a 300 km flight, the gene ontology (GO) enrichment analysis with the use Gene ontology database powered by PANTHER v.17.0 [23], indicates that binding; molecular function and G protein-coupled receptor activity for GO molecular function.GO cellular component processes analysis shows lewy body core, perichromatin fibrils, vacuolar proton-transporting V-type ATPase, V1 domain; and components of vacuolar membrane.The Go biological results pinpoint terms such as chemical homeostasis, regulation of inflammatory response; negative regulation of pri-miRNA transcription by RNA polymerase II; positive regulation of transport, and negative regulation of programmed cell death.

Validation of DEGs with qPCR
The differential expression of HSF2BP, ADIPOQ, PPARD and ACTB genes was validated by qPCR.Pearson's correlation coefficient calculated between normalized counts obtained after the RNA-seq method and Relative Quantification range 0.43 to 0.97, thereby validating the RNA-seq results.

Discussion
A sport involving racing pigeons is a highly profitable industry with services such as breeding, feeding, genetic testing, equipment and the starts of pigeons.Pigeons are the only birds competing in sprint flights and analogically to mammalian athletes trained to achieve sufficient fitness.Several studies approached transcriptomic profiling of skeletal muscles before and after different training schedules such as in horses or dogs [16,24,25], but to date, no reports reveal results obtained by the use of high throughput methods in birds, under one-sided selection for sport flight performance.
The mammalian skeletal muscle response to exercise stimuli results in adaptation through hypertrophy and increasing metabolic capacity manifested by gene expression changes.In this study, samples of the pectoralis major muscle from both untrained and trained racing pigeons, collected after a 300 km flight, were sequenced using Next Generation methods.This approach aimed to explore differentially expressed genes (DEGs) that may reflect not only the effects of flight Triglyceride lipase in the liver increases the levels of intracellular fatty acids derived from the hydrolysis of newly formed triglyceride stores and plays a role in very lowdensity lipoprotein assembly but also the effects of training.The functional analyses of gene expression profiles in skeletal muscle have identified suites of genes and deregulated molecular pathways that are enriched for functions in energy metabolism, autophagy, mitophagy, necroptosis, lipid and fatty acid metabolism, the mitochondrion, and a wide range of signalling pathways indicate adaptation to prolonged flight at several levels.
In the present study, we found several upregulated transcripts of neural implications such as NREP (neuronal regeneration-related protein), SH3TC2 (SH3 domain and tetratricopeptide repeats-containing protein 2), ARTN (artemin), PRG4 (proteoglycan 4) illustrating adaptation of the neuromuscular junction to exercise [26,27].Recent studies have shown that genes involved in the formation of neuromuscular junctions are under positive selection according to the analysis of quantitative traits related to performance in racing pigeons [1,28,29].SH3TC2 encodes a protein expressed in Schwann cells which are the major glial cell type in the peripheral nervous system.SH3TC2 interacts with Rab11 GTPase to regulate the recycling of membranes and receptors to the cell surface [30].The NREP gene plays several roles in endocrine signalling, myogenesis, transformation or morphogenesis of neural, muscle and fibroblast cells [31].NREP gene regulates genes associated with lipid synthesis, significantly increases intracellular cholesterol and triglyceride levels, and increases intracellular lipid droplets, supporting the hypothesis of the metabolic shift toward fatty acid utilization during prolonged flight [32].PRG4 gene encodes Proteoglycan 4 a protein present at the articular cartilage surface and can regulate the inflammatory response and pain through toll-like receptors (TLRs) during tissue injury [33].Flying is one of the most energetically challenged endurance exercise among vertebrates.The energetical cost is compared to running or swimming [6].The respiratory chain capacity of the pigeons' pectoralis muscle is adequate for the simultaneous maximal rate of respiration with the main use of fatty acids.To cope with fatty acid utilization as the main substrate of energy the entire mechanism for inventory, transportation, and disposal had to be set up.Within the obtained results, the several genes and metabolic pathways involved in fuel selection towards fatty acids phenotype were significantly expressed.In the presented report, the most significantly downregulated identified gene is SLC25A30, encoding member of mitochondrial carrier protein in humans, involved in a shift from carbohydrate to lipid metabolism, and protection from oxidative damage when mitochondrial metabolism increases [34].Most recent research indicates that the exact role of SLC25A30 is to transport H 2 S degradation products from the mitochondria thus modulating levels of hydrogen sulfide.Whereas higher concentrations inhibit the electron transport chain and lower maintain mitochondria's bioenergetic state [35].Additionally, we observed the upregulation of SLC15A4 a proton-coupled amino-acid transporter that mediates the transmembrane transport of L-histidine and di-and tripeptides whose activity is pH-dependent and maximized in the acid environment.This strongly supports the switch to fatty acid oxidation.The acidic environment within muscles during the flight is possibly maintained by a set of upregulated genes, involved in the inhibition of lactate dehydrogenase-A (LDHA) activity such as; FNIP2; FNIP1; ATP6V1A; ATP6V1C1, which are involved in the regulations of folliculin.In turn, folliculin acts as a critical link between mTORC1 with PPARGC1A-driven mitochondrial biogenesis, FLCN deficiency increased PPARGC1A expression and increased mitochondrial function and oxidative metabolism suggests that FLCN may play a role in cellular energy and nutrient sensing through interaction with the AMPK-mTOR pathway [36].Furthermore, FLCN controls the movement of the LDHA active-site loop regulating its enzyme activity [37].In the obtained study, no LDHA gene was differentially expressed, which indicates other regulations of anaerobic glycolysis.
One of the most up-regulated DEGs within the set of expressed genes is RGN (regucalcin).The RGN gene encodes a protein with unique calcium-binding without containing the classic EF-hand motif of the calcium-binding domain.The overexpression of RGN enhances glucose utilization and lipid production, which regulates insulin's effect [38].Furthermore, has a suppressive impact on protein phosphatase activity and participates in lipid metabolism and insulin resistance, however, in avian skeletal muscle insulin does not affect muscle glucose transport, mainly due to the lack of the SLC2A4 gene.Obtained results however indicate the highly upregulated glucose transporter SLC2A11 might compensate for the GLUT4 loss.These findings, however, need to be addressed further.
The skeletal muscle tissue is characterized by effective adaptation to changes in metabolic demand.Our results indicate the autophagy pathway as one of the most significant upregulated during intensive flight, which is crucial for cell homeostasis, maintained by the balance of protein synthesis and degradation.This evolutionarily conserved process of controlled degradation of cellular components, in muscles, is regulated by nutrient availability and energy demand.In turn, another significantly upregulated pathway, the mTOR signalling is a key regulator in the anabolic and catabolic signalling of skeletal muscle mass, resulting in the modulation of muscle hypertrophy and muscle wastage [39].Within our results, both pathways are merged via, the upregulated RRAGD (Ras Related GTP Binding D) gene.RRAGD plays, through recruitment of the mTORC1, a crucial role in the cellular response to amino acid availability, regulating anabolic pathways in response to nutrients and physical activity [35].

Conclusions
The next generation of high-throughput sequencing techniques made it possible to undertake research aimed at understanding the genetic basis of pigeon m. pectoralis performance during flight.In this study, for the first time, samples of m.pectorallis from untrained and trained, after 300 km flight, racing pigeons were collected and sequenced with the use of the Next Generation method.The presented results are the skeleton for the next stages of research to understand the processes in the muscles of both trained and untrained birds and also are the background for a further selection of genetic markers associated with certain traits.

Fig. 1
Fig. 1 The results of differential gene expression analysis of DEGs.Volcano plots showing the fold change of transcripts between samples and their corrected p values (vertical axis).Dashed grey lines represent the nominal significance threshold (Padj = 0.05)

Table 1
Top significantly upregulated and downregulated DEG's before vs 300 km flight

Table 2
Significantly overrepresented pathways included DEGs modified in both groups of pigeons