Introduction

Cell types arise by stepwise acquisition of specific gene expression programs during development and have stable cellular phenotypes throughout organism adulthood1. The transcriptome of any given cell type is the culmination of contextual interplay between intrinsic factors (e.g., transcription factors—TFs, non-coding RNAs) and extrinsic factors from the extracellular milieu under the context of species-specific variation that may have arisen during evolutionary trajectories2,3,4,5,6.

Several mammalian species have similar transcriptomes across multiple organs7,8,9, thus suggesting that few loci account for species-specific organ function and response to environmental cues or disease states10. Hence, cross-species gene expression analysis becomes an attractive strategy to assist unraveling evolutionary-driven differences in the gene expression regulatory landscape8. Although this comparative transcriptomic approach—largely based upon RNA sequencing—has been informative, most of them account for analysis carried out with whole organs7,9. Therefore, cell-type specific interrogations and cross-species comparisons of relative gene expression have remained mostly unexplored.

Unbiased cross-species gene expression analysis may pinpoint differences in species-specific relative gene expression at a given locus10,11,12. Recently, one such report has found that the expression of two TFs (TLR4 and ZFX) were up-regulated in Bos taurus (B. taurus) fibroblasts compared to Ovis aries (O. aries) counterparts using rigorous reverse transcription quantitative PCR (RT-qPCR) normalization12. The dissection of such differences between closely related species may be fruitful due to the expected similarities in their gene expression regulatory landscapes11,12.

Species-specific differences in gene expression suggest the rewiring of gene regulatory landscapes. The evolutionary-driven divergences in such networks are byproducts of cis-regulatory elements (CRE) turnover13,14 and sequence variation in TFs15. The CREs are DNA sequences that contribute to the regulation of gene expression, usually containing several TF-binding sites (TFBS) and may include sequence elements that affect mRNA stability and translation15. The CRE evolution is the result of transposable elements acquiring new regulatory roles or sequence evolution within chromatin regions with regulatory potential13,14. TFs are also source of novel gene regulatory landscapes, thus resulting in TF orthologs with different functions or target genes13. The interplay between TF and CRE variation may limit the conservation of TFBS4,16,17. For instance, OCT4 (also known as POU5F1) and NANOG share few (~ 5%) homologous TFBS in mouse and human embryonic stem cells17. Therefore, integrative analysis coupling gene expression analysis and TFBS mapping should contribute toward elucidating evolutionary-driven divergences in gene regulatory landscapes.

C-MYC is an interesting TF to investigate species-specific gene expression patterns because of its contribution to several cellular processes. C-MYC forms a unique transcriptional network in pluripotent cells18,19, regulates over one-tenth of the transcriptome20,21, elicits chromatin architecture modulation22, participates in DNA replication and cell cycle control23,24, modulates apoptosis25, and may increase overall transcription in a given cell26. In the context of cellular reprogramming, C-MYC enhances the conversion of somatic cells into induced pluripotent stem (iPS) cells and accelerates this process in both human and mouse systems19,27.

C-MYC levels have profound effects in cellular physiology28,29. Therefore, this locus is regulated by autoregulation30 and several other factors (RONIN also known as THAP11, RXRβ, TCF3, ZFX, among others) at the transcription, post-transcription, and post-translation levels28,29,30,31,32,33,34,35,36,37. The analysis of regulatory sequences may identify potential sources of species-specific C-MYC relative expression. The aim of this work was to determine the relative expression of C-MYC, CDK9, and C-MYC regulators between O. aries and B. taurus fibroblasts. In silico tools were further applied to identify potential sources of C-MYC relative expression between species at the mRNA, protein, and CRE levels.

Results

Analysis of C-MYC cis-regulatory elements in Ovis aries and Bos taurus genomes

The C-MYC locus is located on the O. aries chromosome 9 (genome version oar_rambouillet_v1.0) and on chromosome 14 of the B. taurus genome (genome version ARS_UCD1.2). The C-MYC gene is transcribed by the minus strand of the DNA in both O. aries and B. taurus (see Fig. 1A). The B. taurus C-MYC locus has six exons and O. aries has three (as found in the mouse and humans). An unbiased prediction of conserved TF binding sites (TFBS) at the C-MYC locus and 2.5 kb upstream the transcription start site between O. aries and B. taurus retrieved one site for Hunchback and three overlapping sites for MYF (see Fig. 1B; Supplementary Fig. S1). Since Hunchback is a Drosophila melanogaster-specific TF, only MYF TFBS are potentially conserved between O. aries and B. taurus (see Fig. 1B). The TFBS prediction for C-MYC, RXRβ, and TCF3 demonstrated species-specific variation (see Fig. 1C; Supplementary Fig. S2), thus reinforcing the notion of evolutionary-driven re-wiring of CREs. The current genome assemblies suggest different number of C-MYC mRNA isoforms between species, thus implying B. taurus-specific alternative splicing (see Fig. 1D). The Ensembl genome browser displayed two C-MYC mRNA isoforms in the current B. taurus genome assembly ARS_UCD1.2 and one C-MYC isoform in O. aries (see Fig. 1E), although based on a previous sheep genome version (oar_v3.1).

Figure 1
figure 1

Cross-species analysis of the C-MYC locus. (A) Genomic context of O. aries (reference genome oar_rambouillet_v1.0) and B. taurus (reference genome ARS_UCD1.2) C-MYC gene orthologs retrieved from genome data viewer (NCBI). (B) Prediction of conserved transcription factor binding sites (TFBS) between O. aries (blue lines) and B. taurus (green lines) using ConSITE in the C-MYC locus and 2.5 kb upstream sequence of the transcription start site (TSS). The gray boxes show the non-conserved DNA sequences between C-MYC loci. (C) Representation of TFBS for C-MYC (yellow), RXRβ (blue), and TCF3 (red) in the C-MYC locus and 2.5 kb upstream sequence of the TSS. Exons were outlined as black boxes and the TSS was indicated with an arrow. (D) Schematic representation of predicted alternative splicing in the B. taurus C-MYC locus by analysis of reference mRNA sequences. (E) Prediction of alternative splicing in the B. taurus C-MYC locus in the Ensembl genome browser.

Full size image

Relative expression of C-MYC, CDK9, and C-MYC-regulators in Ovis aries and Bos taurus fibroblasts

Primer efficiency was determined using O. aries and B. taurus fibroblast cDNA (see Table 1). All primers reached qPCR efficiency threshold, which fluctuated from 94.52 to 109.33%. The correlation coefficient of such qPCR reactions varied from − 0.990 to − 0.998, while their respective slopes ranged from − 3.12 to − 3.46 and Y intercepts from 32.49 to 38.40 (see Table 1). Hence, the RT-qPCR normalization relied on a set of previously validated reference genes (RGs) using five software (GeNorm, Normfinder, BestKeeper, Delta CT method, and RefFinder) under identical experimental conditions12.

Table 1 Primer efficiency (E), coefficient correlation (R), slope, and Y intercept derived from the standard curve of each transcript in the study via RT-qPCR.
Full size table

The RT-qPCR assay determined the C-MYC relative expression in unmodified mammalian fibroblasts (see Fig. 2A). C-MYC relative expression was 1.82–2.23 fold higher in B. taurus in comparison to O. aries (P < 0.001; see Fig. 2B). Further, the C-MYC downstream target TBX3 was up-regulated in B. taurus by 3.07–3.84-fold (P < 0.001). In spite of different RGs, C-MYC and TBX3 relative expression differed between species (see Fig. 2B).

Figure 2
figure 2

Relative expression of C-MYC, CDK9, and C-MYC regulators in mammalian fibroblasts. (A) Experimental design for cross-species gene expression analysis. (B) Relative expression of CDK9 (Yellow), C-MYC (red), and TBX3 (Orange). (C) Relative expression of RONIN (red), RXRβ (Orange), and TCF3 (Yellow). The relative expression was determined as expression fold (x) of the relative expression of the B. taurus relative to O. aries ortholog. Standard error range (+ /−) was calculated by REST52.

Full size image

To infer if mammalian fibroblasts may have variable global transcription output, the RT-qPCR determined the relative expression of the transcript elongation factor (CDK9). However, CDK9 relative expression was similar between species (P = 0.64; see Fig. 2B). To explore potential sources of C-MYC relative expression between species, gene expression analysis measured the relative expression of C-MYC trans-acting regulators. The relative expression of the C-MYC negative regulators did not differ between O. aries and B. taurus fibroblasts (RONIN: P = 0.40; RXRβ: P = 0.75; and TCF3: P = 0.43; see Fig. 2C). Moreover, alternative RGs did not affect the results of the gene expression analysis (see Fig. 2C).

Analysis of C-MYC mRNA regulatory sequences and N6-methyladenosine in Ovis aries and Bos taurus orthologs

The C-MYC locus is under rigorous transcriptional and post-transcriptional regulation. The analysis of mRNA sequence conservation between C-MYC orthologs may point out sources of variable relative expression between species. Based on current genome assemblies, the O. aries has one reference mRNA sequence while B. taurus has three isoforms (see Supplementary Fig. S3). The annotation of characterized C-MYC mRNA regulatory sequences across mammals supports high sequence conservation between O. aries and B. taurus (see Fig. 3A). The 5ʹ untranslated region (UTR) of both O. aries and B. taurus orthologs display most of the sequence variation between these species (see Supplementary Fig. S3). In contrast, coding sequences (CDS) displayed high sequence conservation, including the coding region instability determinant (CRD) in the last 249 nucleotides of the CDS, adjacent to the ribosomal pausing site. The 3ʹ UTR sequence of the four O. aries and B. taurus C-MYC orthologs also have high sequence conservation (see Supplementary Fig. S4) and share multiple regulatory sequences (see Fig. 3A), such as AU-rich octamer, AU-rich stem loop, AU-rich elements, five-nucleotide motifs, polyadenylation sites, and the AU-rich sequence element (see Supplementary Fig. S4).

Figure 3
figure 3

In silico analysis of C-MYC mRNA orthologs. (A) Representation of O. aries and B. taurus C-MYC mRNA isoforms highlighting the regulatory sequences. (B) Total number of predicted RNA-binding proteins (RBPs) in C-MYC mRNA isoforms. (C) Distribution of predicted RBPs in C-MYC mRNA isoforms. (D) Schematic representation (5ʹ to 3ʹ direction) of predicted binding sites for the non-coding RNAs GNAS1 (Orange) and UBE3A (Yellow). (E) Total prediction of N6-methyladenosine (m6A) sites. (F) m6A sites with very high predictive scores and their distribution relative to the coding sequence (CDS). ARE: AU-rich elements. ASE: AU-rich sequence element UUUN [A/U] U. IRES: internal ribosome entry site. O: 3ʹ UTR octamer. PA: polyadenylation site. Red diamonds: AUUUA motifs. SL: Stem loop/AU-rich sequence. UTR: untranslated region.

Full size image

Several reports showed that RNA-binding proteins (RBPs) contribute to C-MYC mRNA stability. The prediction of C-MYC-binding RBPs showed that C-MYC mRNAs have multiple putative binding sites (see Fig. 3B). C-MYC mRNA orthologs showed similar number of binding sites for RBPs, albeit more frequent in the CDS (see Fig. 3C). The functional annotation of predicted RBPs most covered by TFs (12%), although few of these TFs were characterized to bind RNA. Further, mRNA motif predictions suggest B. taurus-specific binding sites for the non-coding RNAs GNAS1 and UBE3A (see Fig. 3D).

The N6-methyladenosine (m6A) is an epigenetic modification of RNA that has important regulatory roles. The M6A prediction was also investigated in O. aries and B. taurus C-MYC mRNA orthologs. The O. aries C-MYC mRNA displayed 31 potential m6A sites, while B. taurus C-MYC mRNA isoforms had 29 to 33 m6A potential sites (see Fig. 3E). Further, both O. aries and B. taurus C-MYC mRNA did not display differences in m6A sites according to very high (P = 0.51), high (P = 0.78), moderate (P = 0.22) or low (P = 0.97) predictive scores (see material and methods; see Fig. 3E). A schematic representation demonstrates that highly predicted m6A sites were mostly found in the CDS and similarly distributed among O. aries and B. taurus C-MYC mRNAs (see Fig. 3F).

Comparative analysis of Ovis aries and Bos taurus C-MYC protein sequences and potential sites for post-translational modification

C-MYC is under complex regulation by post-translational modification (PTM). The alignment of O. aries and B. taurus C-MYC protein orthologs showed a sequence conservation of 98.63% (433/439 residues; see Supplementary Fig. S5). Further, distinct amino acid residues between O. aries and B. taurus species were found in the transactivation domain (TAD; N-terminus) but enriched in the ‘leucine zipper’ motif in the C-terminus (see Fig. 4A). In comparison to the Mus musculus (mouse) C-MYC protein, O. aries and B. taurus proteins share 91.14% (413/440 residues) and 90.45% (398/440 residues) homologies, respectively (see Fig. 4B; see Supplementary Fig. S5). C-MYC domains showed high conservation between O. aries and B. taurus (see Fig. 4C). The calpain cleavage site and the bHLH DNA-binding domain were identical among species, thus leading to overlapping E-BOX DNA binding motifs. The motif rich in proline [P], glutamic acid [E], serine [S], and threonine [T] (PEST) domain and the nuclear localization signal were also identical between O. aries and B. taurus (see Fig. 4C). Further, C-MYC orthologs display substitutions in only two residues in the TAD (i.e., within C-MYC degron site and the auto-repression sequences). The sequence variation in C-MYC protein orthologs did not affect the secondary structure prediction of key domains (see Fig. 4D). Hence, the fourth leucine of this leucine-rich motif was not found in both O. aries and B. taurus, although it was not within the amino acid residues expected for the C-MYC/MAX heterodimer formation (see Fig. 4C).

Figure 4
figure 4

In silico analysis of C-MYC protein orthologs. (A) Representation of C-MYC of both O. aries and B. taurus C-MYC orthologs and their non-conserved residues (black spheres). (B) C-MYC boxes (I–IV) outlined in red. Conservation of C-MYC protein among selected mammalian species. (C) Motif analysis between O. aries and B. taurus C-MYC orthologs. (D) Secondary structure prediction of O. aries and B. taurus C-MYC orthologs describing the transactivation domain (TAD; gray boxes), the basic helix-loop-helix DNA-binding domains (bHLH; large pink boxes), and low complexity region (small pink boxes), and intron sites (*). B: basic region. C: calpain cleavage site. NLS: nuclear localization signal. PEST: motif rich in proline [P], glutamic acid [E], serine [S], and threonine [T]. ZIP: leucine zipper motif. Light blue box: Conserved amino acid residues (RR) required for C-MYC/MAX heterodimer formation.

Full size image

Most PTM sites in the Homo sapiens (human) C-MYC protein were conserved residues in O. aries and B. taurus orthologs (see Supplementary Fig. S6). The exceptions to this rule are three substitutions of threonine to alanine (i.e., residues 8, 78, and 343), which could be (albeit uncharacterized) phosphorylation sites (see Supplementary Fig. S6). Nonetheless, no predicted PTM differed between O. aries and B. taurus orthologs.

Based upon the results above, a scheme was described with potential evolutionary-driven differences between O. aries and B. taurus C-MYC gene orthologs (see Fig. 5A). An outline was also presented describing the relative expression of C-MYC, CDK9, and C-MYC-regulators between B. taurus and O. aries, their interactions to C-MYC transcription, and potential cellular effects that may arise from such species-specific gene expression signatures (see Fig. 5B).

Figure 5
figure 5

Scheme highlighting potential sources of evolutionary-driven C-MYC relative expression in mammalian fibroblasts. (A) Representation of the C-MYC locus in O. aries and B. taurus highlighting potential evolutionary-driven divergences in their regulatory landscape. (B) C-MYC gene transcription-based circuitry investigated in this study. C-MYC repressors and activators outlined by “T” and “arrows”, respectively. Non-significant cross-species gene expression was described in yellow and B. taurus upregulated genes in green. Potential cellular effects of B. taurus-specific C-MYC and TBX3 upregulation highlighted in gray boxes. Moura et al. described the data on ZFX in a recent report12. TF: transcription factor.

Full size image

Discussion

The prediction of TFBS at the C-MYC locus for its encoded protein (autoregulation) and some of its negative regulators revealed substantial variation in TFBS numbers and locations. Genome-wide analyses have demonstrated that TFBS were poorly conserved across mammalian genomes4,38,39, perhaps due to CRE re-wiring caused by transposable elements17,38,39,40. In turn, relative expression of known C-MYC-negative regulators RONIN, RXRβ, and TCF331,32,33 was similar between O. aries and B. taurus. This fact suggests that expression levels of these TFs may not affect species-specific gene expression patterns in mammalian fibroblasts. A next logical step is to investigate the conservation of TFBS for ZFX at O. aries and B. taurus C-MYC locus and its impact on TBX3 gene expression, due to their connection to this proto-oncogene41,42. The re-wiring of TF networks must be a significant driving force in the regulatory modes of the C-MYC locus across species and variation at CREs should explain some (if not most) of the gene expression variation due to the evolutionary trajectories.

Bos taurus fibroblasts express C-MYC to a greater extent than O. aries. The species-specific C-MYC relative expression motivated the exploration of evolutionary-driven divergences in the regulatory landscape at the genomic, mRNA and protein levels. The B. taurus C-MYC locus contains six exons, in contrast to the common three-exon structure found in Ovis aries and other mammals. Further, B. taurus C-MYC locus is expected to carry out alternative splicing, in contrast to O. aries. The creation of new exons and emergence of alternative splicing are clear indications of locus evolution, since these processes tend to avoid negative pleiotropy of evolving TFs15. In silico analyses suggested more sequence variation in C-MYC mRNA orthologs. Rather remarkably, most mRNA sequence variation was found in the 5ʹ UTR of O. aries and B. taurus C-MYC mRNA, although its significance remains elusive. In other species, the C-MYC 5′ UTR was found to contain an internal ribosome entry site43, to contribute to cancer-associated cellular phenotypes44 and to translational efficiency45. Binding sites for non-coding RNAs were another potential source of regulatory variation between C-MYC mRNA orthologs. There is extensive evidence of long non-coding RNAs and microRNAs regulatory hubs at the C-MYC locus in both mice and humans36,37. The evidence of non-coding RNAs as potential regulators of the C-MYC locus in O. aries and B. taurus remains preliminary due to the limited availability of in silico tools for these species and should be focus of further work. Nonetheless, cross-species gene expression analysis coupled with bioinformatics identified a small subset of CREs that may explain species-specific C-MYC relative expression. Future research should focus on interrogating the role of each CRE on C-MYC transcription across mammals. A detailed characterization of C-MYC enhancers in O. aries and B. taurus may be another fruitful endeavor because these CREs evolve more rapidly than promoters and gene expression14,46.

The higher C-MYC relative expression in B. taurus than O. aries may reflect greater differences in the transcriptome of fibroblasts from these species. C-MYC interacts with CDK9 from the elongation complex P-TEFb (formed by CDK9 and cyclins T1 and T2) and releases hundreds of transcripts from RNA polymerase II-mediated transcriptional elongation pausing26. However, CDK9 transcript abundance was similar between O. aries and B. taurus. The pharmacological CDK9 down-regulation may lead to increased C-MYC expression in human cancer cells47,48, although the results described here do not accommodate such compensatory mode. Gene expression analysis using identical cell numbers or single-cell analysis coupled with pharmacological CDK9 modulation should resolve such discrepancies. Nonetheless, the results showed that C-MYC had variable gene expression levels across mammalian fibroblasts and this difference may affect the transcriptome of such cells.

These facts on C-MYC species-specific gene expression are paramount for understanding biological processes at an evolutionary level, for modeling human conditions in animal models, and improved understanding of cellular reprogramming. C-MYC acts pleiotropically in mammalian cells, particularly as a pro-survival factor and inducing cell proliferation23,24. For instance, C-MYC overexpression increases the efficiency of cellular reprogramming in mice and humans19,27. Higher endogenous C-MYC and TBX3 expression or their overexpression during reprogramming was associated with germ-line contribution of iPS cell lines19,27,41. It would be informative to determine if species-specific C-MYC levels correlate with iPS reprogramming efficiency and kinetics. Alternatively, the cross-species analysis of C-MYC could guide the development of improved animal models of C-MYC-driven cancer, such as mice expressing human C-MYC protein49, by focusing on adjusting oncogenic C-MYC expression between mammalian models and patient samples.

In conclusion, mammalian fibroblasts display evolutionary-driven C-MYC relative expression, most probably due to rewiring of CREs, which becomes instructive for understanding and modeling C-MYC-related developmental processes and associated diseases.

Methods

Somatic cell culture

Both O. aries (sheep) and B. taurus (cattle) ear skin fibroblasts were derived from adult males and cultured in high glucose DMEM (Gibco) supplemented with 10% fetal bovine serum as described by Moura et al.12. Fibroblasts cultures were passaged by 0.25% trypsin/EDTA (Gibco) when dishes became confluent (passage zero) and subject to 1:3 splits within seven-day intervals. Fibroblasts samples (~ 1.0 × 106) were dissociated by 0.25% trypsin/EDTA, washed twice in 500 µL Phosphate Buffered Saline (PBS) by centrifugation at 500g for 5 min. and cell pellets were resuspended in 200 µL PBS without calcium and magnesium (Gibco). Further, cell suspensions were snap-frozen in N2 (− 196 °C), and stored at − 80 °C. Total RNA extraction used confluent dishes of early passage fibroblasts (passages two and three).

Total RNA extraction and cDNA synthesis

Total RNA extraction was carried out using Reliaprep RNA Cell Miniprep (Promega), following the manufacturer instructions. Total RNA was quantified using Nanodrop 2000C (Thermo Scientific) to determine 260/280 and 260/230 ratios and further quantified using Qubit (Thermo Scientific) for cDNA synthesis. The RNA was evaluated by electrophoresis with 1.0% agarose gels in 0.5× TBE buffer under 80 V and 120 A for 40 min.

The reverse transcription (RT) reaction was performed after total RNA extractions using 1.0 µg of total RNA per sample. The procedure was performed with QuantiTect Reverse Transcription Kit (Qiagen). Firstly, residual genomic DNA was removed by the gDNA elimination reaction (7× gDNA wipeout buffer, 1.0 µg total RNA, and ultra-pure H2O; 14 µL of total reaction) at 42 °C for 2 min., and transferred to 4 °C. Secondly, the previous reaction mixed to the RT reaction (4 µL 5× Quantscript RT buffer, 1 µL RT primer mix, and 1 µL Quantiscript RT) was kept at 42 °C for 30 min. Finally, samples incubated at 95 °C for three min., and stored at − 20 °C.

Reverse transcription quantitative PCR (RT-qPCR)

The experiment followed the Minimum Information for Publication of Quantitative Real-Time PCR Experiments (MIQE) guidelines to increase both the transparency and reliability of the RT-qPCR data50.

Primers were designed using the strategy outlined by Moura et al.51. Briefly, reference mRNA sequences were retrieved from GenBank (https://www.ncbi.nlm.nih.gov/genbank/) from selected mammalian species (Capra hircus, O. aries, and B. taurus) and used as templates to design multi-species qPCR primers (see Table 2). Primers were designed using Primer-BLAST (https://www.ncbi.nlm.nih.gov/tools/primer-blast/) and further selected using Primer3plus (https://www.bioinformatics.nl/cgi-bin/primer3plus/primer3plus.cgi). Primer amplification efficiency (E = 10−1/slope), correlation coefficient (R2), and interception (y) were determined using the standard curve method using cDNA serial dilutions, i.e., 100—non-diluted samples, 10−1, 10−2, 10−3, and 10−412.

Table 2 Primers used in the study for reverse transcription quantitative PCR (RT-qPCR).
Full size table

The RT-qPCR reactions were carried out using SYBR green system in a Line Gene 9660 FQD-96A real-time PCR (Bioer). The analysis relied on three biological replicates and three technical replicates. The reaction was composed of 1.0 µL cDNA, 2.5 µM primers, 1× Go Taq qPCR Master Mix (Promega), and ultra-pure H2O. The reactions were performed under strict conditions (initial denaturation at 95 ºC for 2 min., 40 PCR cycles at 95 ºC for 15 s, 58 ºC for 30 s, and 72 ºC for 30 s). Melting curves were analyzed in 65–95 ºC for 20 min. after the PCR cycles. The RT-qPCR data normalization was carried out using the RGs ATP1A1, RPL19, and UBB12. Expression levels of candidate genes evaluated were based upon the number of cycles required for reaching a fixed threshold (quantification cycle—Cq) during the exponential phase of the PCR assay. The relative gene expression levels were evaluated with the REST tool (version 2.0.13) that relies on the 2(−ΔΔCT) method52. The supplementary information section contains the raw RT-qPCR data (see Supplementary Table 1) and the results from the REST analysis (see Supplementary Table 2).

Bioinformatics

The DNA sequences of C-MYC gene orthologs were retrieved from GenBank in July 2019. Exons were annotated manually and a 2.5 kb upstream sequence from the transcription start site of each C-MYC gene orthologs. Conserved TFBS in C-MYC gene orthologs (the complete coding sequences and the additional 2.5 kb DNA upstream sequence) obtained from ConSITE53 using the ORCA alignment method and selecting the “all transcription factor profiles” option (https://consite.genereg.net/cgi-bin/consite). Additional TFBS were predicted in C-MYC loci using PROMO 3.054. The TFBS search in PROMO was limited to C-MYC [T00140], RXRβ [T01332], and TCF3 [T02857] using version 8.3 of TRANSFAC (https://alggen.lsi.upc.es/recerca/menu_recerca.html).

The identification of C-MYC mRNA sequences (isoforms) was based on the most recent reference genome assemblies of B. taurus (ARS_UCD1.2)55 and O. aries (oar_rambouillet_v1.0; https://www.hgsc.bcm.edu/other-mammals/sheep-genome-project) using the genome data viewer (https://www.ncbi.nlm.nih.gov/genome/gdv/). The number and identity of B. taurus and O. aries C-MYC isoforms were retrieved from the Ensembl genome browser (m.ensembl.org). C-MYC mRNA sequences obtained from GenBank (RefSeq). The mRNA sequence alignment was carried out using MUSCLE (https://www.ebi.ac.uk/Tools/msa/muscle/)56. The mRNA regulatory sequences/motifs were annotated manually. The prediction of RBP binding sites and non-coding RNA binding sites were performed using RegRNA 2.057 with the search option for all available RNA motifs (https://regrna2.mbc.nctu.edu.tw/detection.html). The N6-methyladenosine (m6A) messenger RNA methylation prediction was carried out using SRAMP58. The m6A predictive scores (very high, high, moderate, and low) were calculated by SRAMP using the full transcript mode and the generic (default) model for tissue choice (cuilab.cn/sramp/).

C-MYC reference protein sequences were retrieved from GenBank and aligned using MUSCLE. Predicted post-translational modifications (PTM) were retrieved from PhosphoSitePlus (phosphositeplus.org). The Venn diagram was prepared using Venny 2.1 using the default option (https://bioinfogp.cnb.csic.es/tools/venny/index.html). Protein secondary structure was predicted with SMART59 using the default setting but including the PFAM domain, signal peptide, and internal repeat options (https://smart.embl-heidelberg.de/). Protein domains were obtained from the literature and annotated manually.

Ethical approval

The research project was approved by the Ethics Commission on Animal Experimentation (CEUA) under the license 031/2016 at the Federal Rural University of Pernambuco (UFRPE). All experiments were performed in accordance with institutional and national guidelines.