Structural heterogeneity and functional convergence of transposable elements Authors: Gleb Yu. Kosovsky, Galina V. Glazko, Tatiana T. Glazko Published: 2025 ABSTRACT Almost half the mammalian genomes consist of transposable elements (TEs) and their derivatives. The distribution density of TEs can be associated with genomic regions of chromosomal rearrangements in different mammalian species and with the genomic localization of protein-coding genes that differ in length and function. To evaluate these characteristics at the local genomic level, an analysis of the distribution of various TEs (retrotransposons and DNA transposons) was performed in three mammalian species (human, cattle, and domestic rabbit) in genes with different functions and chromosomal localizations and their flanking regions. In humans and rabbits, melanophilin (MLPH) and myostatin (MSTN) are syntenic, but not in cattle. In the latter, MLPH and the leptin receptor (LEPR) are syntenic, but not in humans and rabbits. The alpha-thalassemia gene is always located on chromosome X. The results indicate that the frequencies of different TEs are species-specific and do not depend on the length of genes, their function, or chromosomal localization. There were also species-specific differences in the ratio of “ancient” and “young” short interspersed nuclear elements (SINEs) and long interspersed nuclear elements (LINEs). There was a statistically significant positive correlation between ancient SINE + LINE and LTR-ERV (p < 0.01) and a significant negative correlation between young SINE + LINE and DNA transposons (p < 0.05). Competitive relationships between TEs are probably defined by the presence of identical regulatory motifs in different TEs, associated with the reliance of TE amplification on the host’s own regulatory systems. Introduction Recent studies have demonstrated that transposon-derived regulatory network elements are associated with phenotypic diversity (Fueyo et al., 2022; Du et al., 2024; Zhao et al., 2023). These studies are particularly important in the era of genome editing, as the structural and functional organization of potential gene editing targets requires careful study to assess possible negative side effects (Kosovsky et al., 2024). Transposable elements (TEs) contain sequences necessary for their expression and transposition. These sequences can also regulate the host’s transcription, even when the TE itself has lost transposition function because of accumulated mutations. The mechanisms of host transcriptional regulation by TEs include providing enhancer and promoter sequences, targets of three dimensional chromatin organization, transcription factor (TF)-binding sites, and new regulatory elements including non-coding RNAs, the modification of nearby genes’ methylation states, and chromosomal structural modifications (Du et al., 2024). In mammalian genomes, there are gene blocks that are evolutionarily conserved and gene blocks that are frequently subjected to chromosomal rearrangements (homologous synteny blocks (HSBs) and evolutionary breakpoint regions (EBRs)). The differences between HSBs and EBRs are that the latter have high frequencies of TEs and microsatellite loci, more genes with a low number of base pairs, and they are frequently located in regions without topologically associated domains (TAD). HSBs are enriched with genes involved in anatomical morphogenesis and the development of the central nervous system (Damas et al., 2022). Accumulated data suggest that when searching for genomic elements associated with phenotypic diversity, non-coding sequences with TEs make the greatest contribution, since they can regulate the expression profiles of adjacent genes (Du et al., 2024). TEs of different origin with the same TF-binding sequences are frequently located in syntenic regions. If one of the TEs carrying the corresponding TF-binding nucleotide motif is damaged, then its function can be performed by another TE variant carrying the same motif (Du et al., 2024). The density of TEs carrying TF-binding sites varies significantly between different genomic regions and depends on the localization of genes involved in different pathways (the lowest density is in the regions of genes associated with the early stages of embryonic development) (Damas et al., 2022; Nikitin et al., 2018; Nikitin et al., 2019). Many studies are focusing on influencing phenotypic characteristics by editing TE elements (e.g., Buttler et al., 2023; Simpson and Chuong, 2023). However, the high evolutionary rate of TEs, as well as their diversity and multifunctionality, make this research avenue difficult. Mammalian genomes are dominated by TEs and their derivatives that can comprise approximately half of the genome (Farmiloe et al., 2023). They are divided into two groups based on their mobilization intermediates. Class I elements, also known as retrotransposons, mobilize as an RNA intermediate and include autonomous retrotransposons (long interspersed nuclear elements (LINEs)and endogenous retroviruses (ERVs)), as well as non-autonomous short interspersed nuclear elements (SINEs). Class II elements, also known as DNA transposons, mobilize as a DNA intermediate. The diversity and abundance of TEs can be highly different, even for closely related species. One explanation for this pattern is that TEs with lower genomic abundance more strongly influence closely linked host genes and host activity to block TE transcription and insertion at other positions in the genome (Platt et al., 2018). The difficulties in understanding TE effects on phenotypic diversity are not exclusively related with their extreme diversity and differences in copy numbers; there could also be species-specific TEs influencing the same phenotypic characteristic. For example, exhaustive studies support the hypothesis that different ERVs could be evolutionary driving forces for trophoblast cell fusion and species-specific placental structures (Sakurai et al., 2023). Thus, the same phenotypic characteristic can be formed with convergent participation of different TEs. In order to elucidate the potential relationship between the function of protein-coding genes and the distribution of various TEs, we here perform a comparative analysis of TEs and the structural and functional features of the genes encoding melanophilin (MLPH, a transporter of melanosomes, in particular), myostatin (MSTN, an intracellular signaling system; a ligand of the transforming growth factor beta), leptin receptor (LEPR, an extracellular signaling system), and alpha-thalassemia (ATRX, involved in cell division mechanisms) in three mammalian species belonging to different orders: Oryctolagus cuniculus (domestic rabbit), Bos taurus (cattle), and Homo sapiens (human). Considering the associations between the functions of protein-coding genes and the distribution of TEs (Damas et al., 2022; Nikitin et al., 2018; Nikitin et al., 2019), one would expect that the lowest TE density would be in the ATRX gene, which is always located on chromosome X in mammals. The analysis of TE distribution in genes and their flanking regions (1 million base pairs long, 1 Mb) includes LINEs, SINEs, endogenous retroviruses (ERVs, long terminal repeats; LTR in what follows), DNA transposons (DNA in what follows) and microsatellites. LINEs/L2 and SINEs/MIRs were considered separately because they are classified as evolutionarily “ancient” repeats compared to LINEs/L1 and SINEs/Alu, tRNA (Smalheiser and Torvik, 2005; Jurka et al., 2007). The subdivision of SINEs and LINEs into “ancient” and “young” transposons is associated with different frequencies of LINEs/L2 and SINEs/MIRs in mammals (monotremes, marsupials, and placentals). Estimates of the monotreme–theria divergence time range between 160 and 210 M years ago; platypus, placed with the echidnas into the Monotremata taxon, has mostly ancient LINEs/L2 and SINEs/MIRs (Warren et al., 2008). The 2.3 Gb platypus genome contains 1.9 and 2.75 million copies of LINE2 and MIR, respectively, compared to significantly lower numbers in other mammals (Warren et al., 2008). It has been shown that ancient LINE/L2s and SINE/MIRs are gradually replaced by young LINE and SINE repeats in mammalian genomes (Buc et al., 2018). All four genes have distinct functions and belong to gene families of varying sizes and syntenic conservation. Only the ATRX gene is consistently located on the X chromosome across mammalian species. Comparative analysis reveals significant differences in the distribution of LINE, SINE, LTR, and DNA transposons across genes and species regardless of the localization and function of the genes considered. Materials and methods For comparative analyses, O. cuniculus, GCF_009806435.1, B. taurus, GCF_002263795.3, and H. sapiens, GCF_000001405.40 (https://www.ncbi.nlm.nih.gov/, GenBank) genome assemblies were used. MLPH, MSTN, LEPR, and ATRX coordinates were uploaded from GenBank (https://www.ncbi.nlm.nih.gov/). H. sapiens genome assembly: GRCh38.p14 (GCF_000001405.40). B. taurus genome assembly: ARS-UCD1.3 (GCF_002263795.3). O. cuniculus genome assembly: mOryCun1.1 (GCF_964237555.1). The gene neighborhoods were obtained with Gene Advanced Search Builder (https://www.ncbi.nlm.nih.gov/) using the “sort by chromosome” option and “chromosome & organism” query. Evolutionary conservation of genetic linkage analysis for genes in the neighborhoods of MLPH, MSTN, LEPR, and ATRX was performed in genomic segments approximately 2 × 106 bp long, with the reference gene located in the middle of the segment. Non-annotated loci were excluded. The frequencies and distribution of dispersed and microsatellite repeats were found with RepeatMasker software tool (RepeatMasker.org, https://repeatmasker.org/cgi-bin/AnnotationRequest). Genomic TE annotations were obtained from RepeatMasker (https://repeatmasker.org/; H. sapiens—RepeatMasker version 406; B taurus—RepeatMasker version 405; O cuniculus—RepeatMasker version: 405), and correlation coefficients were estimated with https://www.statskingdom.com/correlation-calculator.html. To estimate the evolutionary conservation of genetic linkage for the reference genes, the distribution of neighborhood genes was analyzed in 2 Mb segments for H. sapiens, B. taurus, and O. cuniculus genomes. Neogale vison (GCF_020171115.1) was also included in this analysis but was excluded from the analysis of dispersed repeats distribution because this species is not represented in the RepeatMasker database. Results To estimate the evolutionary conservation of melanophilin’s genetic linkage, the distribution of neighborhood genes was analyzed in 2 Mb segments for the genomes of H. sapiens, B. taurus and O. cuniculus. The length (1 Mb) for 5′ and 3′ flanking regions was selected because, on average, the length of each chromatin loop is approximately 1 Mb long. Generally, the length of topologically associated regions varies from 100 kbp to 1 Mb (Cao et al., 2023; Glaser and Mundlos, 2022). We have previously examined the evolutionary conservation of genetic linkage for MSTN and LEPR genes in various species, including reptiles (Kosovsky et al., 2024). The results of genetic linkage for MLPH are presented in Table 1. There is consistent genetic linkage between MLPH and COPS8, COL6A3, PRLH, and RAB17 genes. PRLH (enables neuropeptide hormone activity and prolactin-releasing peptide receptor binding activity) and RAB17 (a key gene regulating melanocytic filopodia formation in melanocytes in mast cells (Beaumont et al., 2011; Babina et al., 2022)) are genetically linked with MLPH even in birds (Bed’hom et al., 2012). All three genes participate in exosome formation and, eventually, in intercellular transport. Two other genes, genetically linked with MLPH, COL6A3 (encodes the alpha 3 chain of type VI collagen, a protein that plays a crucial role in the extracellular matrix), and COPS8 (one of the eight signalosome complex subunits that interact with ubiquitin ligase and is involved in histone modifications) are presumably involved in fundamental processes of intercellular interactions and the differential regulation of gene expression. It is known that E3 ligase triggers extensive gene expression reprogramming by changing global levels of H3K27me3 (a histone modification associated with gene repression) in plants (Wang et al., 2024), bees (Lowe et al., 2022), and in some cellular populations in tumorigenesis (Marine, 2012). TABLE 1 a Comparative analysis of genetic linkage for MLPH gene flanking regions in different mammalian species.HumanAmerican minkCattleDomestic rabbit236486410–238555322, Chr 243484351–45484351, Chr 3115959554–118005443, Chr 31702392–3739474, Chr 3 IQCA Motif containing with AAA domain AGAP1 GTP-binding and GTPase-activating protein43000000–43282943 IQCA1 Motif containing with AAA domain TWIST2 Twist family bHLH transcription factor 2 ACKR3 Atypical chemokine receptor GBX2 Gastrulation brain homeobox 2 ACKR3 Atypical chemokine receptor ASB1 Ankyrin repeat and SOCS box-containing 1 COPS8-DT COPS8 divergent transcript ASB18 Ankyrin repeat and SOCS box-containing 18 LOC100335382 Histone–lysine N-methyltransferase SETMAR-like TRAF3IP1 TRAF3-interacting protein 1 RNU6-1051P RNA, U6 Small Nuclear 1051, Pseudogene IQCA1 Motif containing with AAA domain COPS8 COP9 signalosome subunit 8 PER2 Period circadian regulator 2 COPS8 COP9 signalosome subunit 8 ACKR3 Atypical chemokine receptor COL6A3 Collagen type VI alpha 3 chain HES6 Hes family bHLH transcription factor 6 COL6A3 Collagen type VI alpha 3 chain COPS8 COP9 signalosome subunit 8 LOC513039 40S ribosomal protein S16 pseudogene ILKAP ILK-associated serine/threonine phosphatase MLPH 237486410–237555322 COL6A3 Collagen type VI alpha 3 chain MLPH 116959511–117005449 ERFE Erythroferrone PRLH Prolactin-releasing hormone MLPH 44484351–44528469 RAB17 Member RAS oncogene family ESPNL Espin RAB17 Member RAS oncogene family PRLH Prolactin-releasing hormone PRLH Prolactin-releasing hormone SCLY Selenocysteine lyase RAB17-DT RAB17 divergent transcript RAB17 Member RAS oncogene family LRRFIP1 LRR-binding FLII-interacting protein 1 UBE2F Ubiquitin-conjugating enzyme E2F (putative) LRRFIP1 LRR-binding FLII-interacting protein 1 LRRFIP1 LRR-binding FLII-interacting protein 1 RBM44 RNA-binding motif protein 44 RAMP1 Receptor activity-modifying protein 1 RBM44 RNA-binding motif protein 44 RBM44 RNA-binding motif protein 44 RAMP1 Receptor activity-modifying protein 1 RBM44 RNA-binding motif protein 44 RAMP1 Receptor activity-modifying protein 1 RAMP1 Receptor activity-modifying protein 1 MIR2902 TNFR1 pathway and apoptosis modulation and signaling LRRFIP1 LRR-binding FLII-interacting protein 1 LOC124900521 Small nucleolar RNA SNORD55 UBE2F Ubiquitin-conjugating enzyme E2F (putative) UBE2F Ubiquitin-conjugating enzyme E2F (putative) RAB17 RAB17%2C member RAS oncogene family UBE2F-SCLY UBE2F-SCLY readthrough (NMD candidate) SCLY Selenocysteine lyase SCLY Selenocysteine lyase PRLH Prolactin-releasing hormone UBE2F Ubiquitin-conjugating enzyme E2F (putative) ESPNL Espin-like ESPNL Espin-like MLPH 2702392–2739474 RNU6-1333P RNA%2C U6 small nuclear 1333%2C pseudogene KLHL30 Kelch-like family member 30 KLHL30 Kelch-like family member 30 COL6A3 Collagen type VI alpha 3 chain SCLY Selenocysteine lyase CAMKMT Calmodulin-lysine N-methyltransferase ERFE Erythroferrone COPS8 COP9 signalosome subunit 8 ESPNL Espin-like LINC01833 Long intergenic non-protein-coding RNA 1833 ILKAP ILK-associated serine/threonine phosphatase LOC138849152 YLP motif-containing protein 1-like KLHL30 Kelch-like family member 30 SIX3-AS1 SIX3 antisense RNA 1 HES6 Hes family bHLH transcription factor 6 ACKR3 Atypical chemokine receptor 3 ERFE Erythroferrone SIX3 SIX homeobox 3 PER2 Period circadian regulator 2 IQCA1 IQ motif containing with AAA domain 1 ILKAP ILK-associated serine KRTCAP2P1 KRTCAP2 pseudogene 1 TRAF3IP1 TRAF3-interacting protein 1 ASB18 Ankyrin repeat and SOCS box-containing 18 LINC02610 Long intergenic non-protein-coding RNA LOC124907760 Keratinocyte-associated protein 2-like ASB1 Ankyrin repeat and SOCS box-containing 1​​ TARDBPP3 TARDBP pseudogene 3 SIX2 SIX homeobox 2​​​​ HES6 Hes family bHLH transcription factor 6 LINC01121 Long intergenic non-protein-coding RNA 1121​​​​ PER2 Period circadian regulator 2 SRBD1 S1 RNA-binding domain 1​​​​ a First table raw indicates species; second, coordinates of the analyzed segment (2 Mb in length). Color highlights the four genes with evolutionary conserved genetic linkage for the reference species. a Comparative analysis of genetic linkage for MLPH gene flanking regions in different mammalian species. First table raw indicates species; second, coordinates of the analyzed segment (2 Mb in length). Color highlights the four genes with evolutionary conserved genetic linkage for the reference species. TEs within MLPH, MSTN, LEPR, and ATRX genes Figures 1–3 present comparative analysis of dispersed and microsatellite repeat distributions in MLPH, MSTN, LEPR, and ATRX genes in humans, cattle, and rabbits. SINE and LINE repeats are classified as “young” and “ancient”. There are young primate-specific SINE/Alu, and in all species—SINE/tRNA, young LINE/L1, and ancient SINE/MIR and LINE/L2. This classification is subjective because these groups can also be subdivided into relatively recent and relatively ancient repeats, as well as endogenous retrovirus repeats (LTR group) and DNA transposons (DNA group). SINE/Alu repeats are present in primate genomes, while SINE/MIR can be found even in non-mammalian species (Jurka et al., 2007). LINE/L1 is younger than LINE/L2 and actively transposes, while L2 repeats are ancient. Number of the interspersed (SINE: Alu, MIR; LINE: L1, L2; LTR; DNA) and tandem (STR) repeats in human genes MLPH, MSTN, LEPR, and ATRX. Number of the interspersed (SINE: tRNA-C, MIR; LINE: L1, L2; LTR; DNA) and tandem (STR) repeats in cattle genes MLPH, MSTN, LEPR, and ATRX. Number of the interspersed (SINE: tRNA-C, MIR; LINE: L1, L2; LTR; DNA) and tandem (STR) repeats in rabbit genes MLPH, MSTN, LEPR, and ATRX. In humans, the length of selected genes is very different, the longest being ATRX, then LEPR, MLPH, and the shortest, MSTN. The frequency of repeats in LEPR and ATRX is also different: approximately twice for LTR, four times for SINE, and almost twice for LINE and microsatellites (STR) (Figure 1). The ratio of ancient and young repeats is approximately the same for three genes; however, for ATRX, young LINE L1 is more frequent. LEPR (220908 bp) and ATRX (281336 bp) have similar lengths; however, the latter has more SINEs and STRs (Figure 1). The shortest gene, MSTN, has the lowest frequency of repeats (Figure 1). Interestingly, the frequency of young (SINE/Alu) and ancient (SINE/MIR and LINE1/2) repeats is approximately the same for all genes except ATRX, where LINE/L1 was three times more frequent than LINE/L2. The same comparative analysis of dispersed repeats’ frequency and distribution was performed for cattle (Figure 2). For cattle, gene lengths are much smaller than for humans (except ATRX); however, the number of repeats is approximately the same, except that MSTN has more repeats than humans. ATRX has more MIR, L2, LTR, and DNA in cattle than in humans, despite very similar length (281336 bp in humans and 286037 bp in cattle), unlike MLPH (68913 bp in humans and 39463 bp in cattle) and LEPR (220908 bp in humans and 104345 bp in cattle). MLPH, LEPR, and ATRX in cattle, humans, and rabbits also have different exon numbers. MLPH has 20 exons in humans, 17 in cattle, and 16 in rabbits; LEPR has 20 exons in humans, 21 in cattle, and 27 in rabbits. ATRX has 35 exons in humans, 38 in cattle, and 41 in rabbits (GenBank, ncbi.nlm.nih.gov). For all species, MSTN had only three exons (Figures 1–3). In human, MLPH and LEPR are in different autosomes (Chr 2 and 1), and in cattle they are in the same autosome, Chr 3. In human and rabbit genes, MLPH and MSTN are in the same autosomes (Chr 2 in human and Chr 3 in rabbit), but in different autosomes in cattle. MLPH and MSTN are shorter in rabbits and cattle than in humans, reflected in their exon numbers (MLPH: human, cattle, rabbit – 20, 17, and 16, respectively). Despite ATRX being shorter in rabbits than in humans and cattle, it has more exons (41 in rabbits vs. 35 in humans and 38 and cattle). LEPR (27 exons) and ATRX (41 exons) in rabbits have more exons than these genes in humans and cattle (LEPR: 20 and 21, ATRX: 35 and 38, in humans and cattle, respectively). In addition, in these two genes, rabbits have more SINEs than other species, except ATRX in humans (Figures 1–3). This observation is in agreement with the recent finding that SINE TEs are more prevalent in the domestic rabbit genome compared to other livestock species (Zhao et al., 2023). Furthermore, the number of exons in these genes (except MSTN with just three exons) was different for all species. MLPH has 20, 17, and 16 exons in humans, cattle, and rabbits; LEPR has 20 exons in humans, 21 in cattle, and 27 exons in rabbits; ATRX has 35 exons in humans, 38 in cattle, and 41 in rabbits. The number of different repeats varies from gene to gene, independently of chromosome localization, and depends only on the given gene and species (Table 2). The number of young (SINE/Alu, LINE/L1) and ancient (SINE/MIR, LINE/L2) ТЕs in humans is similar for all genes, except for more LINE/L1 in ATRX. In contrast, this gene in cattle has more young repeat SINE/tRNA, while rabbits have more repeat SINE/tRNA not only in ATRX but also in LEPR (Table 2). TABLE 2Number of repeats in human, cattle, and rabbit genes (melanophilin (MLPH), myostatin (MSTN), and ATRX). Numbers indicate number of repeats. STR stands for microsatellites (https://repeatmasker.org/cgi-bin/AnnotationRequest).SpeciesGeneSINELINE​​​Gene’s length, bpAlu, SINE/tRNA-CMIRL1L2LTRDNASTRHumanGenome/assembly: human - December 2013 - hg3868913 MLPH 1712691917167030 MSTN 0011012220908 LEPR 50388776433653281336 ATRX 2333026921184268CattleGenome/assembly: cattle – October 2011 – bostau739463 MLPH 327251551086627 MSTN 6215101104345 LEPR 49222432162122286037 ATRX 133649859918351RabbitGenome/assembly: rabbit – April 2009 – OryCun237082 MLPH 27420011084909 MSTN 2210031213705 LEPR 132206029451160201272 ATRX 142279915282652 Number of repeats in human, cattle, and rabbit genes (melanophilin (MLPH), myostatin (MSTN), and ATRX). Numbers indicate number of repeats. STR stands for microsatellites (https://repeatmasker.org/cgi-bin/AnnotationRequest). Adelson et al. (2009) found a statistically significant correlation between the ancient SINE/MIR and the ancient LINE/L2 in a whole-genome analysis of cattle. They suggested that this is because non-autonomous SINE/MIR uses the autonomous LINE/L2 for transpositions. Gene- and species-wide differences in their ratio, as well as in the ratio between young SINE and LINE (Table 2), suggest that there are other factors influencing these ratios. These factors could be the presence of motifs that are included in regulatory networks—for example, the motif for transcriptional regulatory factor CTCF (responsible for chromatin higher order structure) (Choudhary et al., 2020; Sharif et al., 2023; Hansen et al., 2024) and G4 quadruplexes associated with regulatory elements (Zhang et al., 2024) that are found in different ТЕs. Recently, it was shown that one TE with CTCF motif can be exchanged for another TE but with the same motif, and the exchange does not disrupt chromatin loop organization (Ivancevic et al., 2018). It was found in the same study that orthologous chromatin loops in mice and humans employ TEs of different structural and evolutionary origin (SINE/B2 in mice, LTR-ERV and DNA endogenous viruses in human), with only motifs for CTCF binding being in common (Choudhary et al., 2020). Currently, there are many data demonstrating that similar regulatory elements are present in different transposons involved in regulatory networks which could influence the association or balance between the numbers of different TE repeats (Du et al., 2024). In summary, among all genes considered, MLPH and LEPR are almost twice longer in humans than in cattle, and there are virtually no differences in length for MLPH between cattle and rabbits and in LEPR between humans and rabbits (Table 2). The frequencies of repeats were independent of gene length and chromosomal localization; however, their distribution and the ratio of young and ancient repeats (SINE and LINE) were species-specific. It might be expected that there are several factors that influence the structure of genes: natural selection in genomes shaping the number of exons, and the presence of different repeats with similar regulatory elements participating in regulatory networks. The distribution of repeats depends more on gene structure than localization on the same chromosome. TEs within flanking regions of genes MLPH, MSTN, and LEPR The ATRX gene was excluded from this analysis because the mammalian X chromosome has unique features, particularly increased enrichment of LINE/L1 (Buc et al., 2018). For each flanking region, a 1 Mb segment length was selected because it is the upper limit of chromosomal loop length, organizing genes in the same structural and regulatory unit (Cao et al., 2023; Glaser and Mundlos, 2022). The results are presented in Figures 4–6. Number of the interspersed (SINE: Alu, tRNA-C, MIR; LINE: L1, L2; LTR; DNA) repeats in 5' (1 Mb) flanking regions of human, cattle, and rabbit genes MLPH, MSTN and LEPR (Supplementary Materials, Supplementary Table S1, S2, S3). Number of the interspersed (SINE: Alu, tRNA-C, MIR; LINE: L1, L2; LTR; DNA) repeats in 3' (1 Mb) flanking regions of human, cattle, and rabbit genes MLPH, MSTN and LEPR (Supplementary Materials, Supplementary Table S1, S2, S3). Number of “young” (SINE/Alu and SINE/tRNA + LINE/L1), “ancient” (SINE/MIR + LINE/L2), LTR, DNA transposons and tandem repeats STR in 5'+3' (2 Mb) flanking regions of human, cattle, and rabbit genes MLPH, MSTN and LEPR (Supplementary Materials, Supplementary Table S4). MLPH and MSTN human gene flanking regions contain more young SINE and LINE (Figures 4, 5). LEPR has more ancient than young repeats in 5′ flanking region (1_mb_1), but in 3′ flanking region (1_mb_2) their frequencies are almost identical. LEPR is located in an evolutionarily conserved gene neighborhood and is probably more insulated from repeat insertion than MSTN, which does not belong to an evolutionarily conserved gene cluster (Kosovsky et al., 2024). In the flanking regions of all three genes are sequences, homologous to bovine autonomous retrotransposon LINE/RTE-BovB (Supplementary Tables S1–S3) involved in horizontal gene transfer (Ivancevic et al., 2018), and in humans to DNA Mutator transposable element that was first described in corn and was later found in mice as an element involved in meiotic recombination (Dupeyron et al., 2019; Underwood and Choi, 2019). MLPH and MSTN cattle genes in both flanking regions contain more young than ancient retrotransposons (Figure 6), except for the cattle MLPH 2MB_5’+3′ flanking region. Sometimes there were 250–300 more young repeats per 1 Mb than ancient SINE and LINE. For cattle gene LEPR flanking regions, the differences between ancient and young repeat frequencies were less drastic than in humans (Figure 6). Unlike the human LEPR gene, the cattle gene had more young repeats in flanking regions, but the differences from the ancient were less than for other genes at only approximately 100 repeats (Figure 6). Additionally, SINE/MIR frequencies for cattle were generally less than in humans (Figure 6). That is, there is still the tendency in cattle that young and ancient SINE and LINE frequencies are gene-specific. For rabbits, repeat frequencies for ancient SINE and LINE were lower than for young repeats. This is probably expected because of the multiple chromosomal rearrangements in the O. cuniculus genome (Zhao et al., 2023). Bovine LINE/RTE-BovB repeat frequencies in rabbits were also less than in humans (Supplementary Tables S3). For some repeats the overall distribution and frequencies were also species-specific (Figure 6). SINE and LINE frequencies were higher for all flanking regions in cattle, and LTR frequencies were higher in MLPH and MSTN flanking regions than in humans (Figure 6). However, for all flanking regions in cattle, the frequencies of DNA transposons were lower than in humans (Figure 6). Rabbit LEPR flanking regions contain more young SINE and LINE repeats than cattle and humans but less LTR and DNA transposons. The higher frequency of young SINE and LINE repeats in the rabbit than in the cattle genome is in agreement with the results of genomic comparisons in different agricultural species, including rabbits and cattle (Zhao et al., 2023). Lagomorpha and Rodentia belong to the same clade (Glires) and presumably diverged from the last common ancestor approximately 70–65 million years (Fueyo et al., 2022). This allows comparison of the frequencies of young and ancient SINE and LINE repeats in human, cattle, and rabbit genomes with their frequencies in the house mouse genome. The results are presented in Table 3 and suggest that, based on the repeat frequencies, there are two groups of species. In the human and cattle group, the frequencies of ancient SINE/MIR and LINE/L2 repeats (2.9 and 3.7 in human and 2.7 and 2.7 in bovine, respectively) are higher than in the rabbit and mouse group (1.6 and 1.8 in rabbits and 0.6 and 0.4 in mice, respectively). The accumulated data suggest that at the genomic level, the differences in the frequencies of young and ancient DNA repeats and the replacement of ancient repeats with young could be related to differences in litter size (e.g., the size of rabbit and mouse litters is far larger than in humans and cattle), as well as the rate of generational change. TABLE 3Number of dispersed (non-LTR) and tandem repeats in Homo sapiens, Bos taurus, Oryctolagus cuniculus, and Mus musculus genomes (%) (https://www.repeatmasker.org/genomicDatasets/RMGenomicDatasets.html).Repeats/speciesSINE/Alu, tRNASINE/MIRLINE/L1LINE/L2STR Homo sapiens 10.52.917.53.71.5 Bos taurus 9.42.726.22.70.9 Oryctolagus cuniculus 18.41.614.91.81.4 Mus musculus 7.20.619.90.43.1 Number of dispersed (non-LTR) and tandem repeats in Homo sapiens, Bos taurus, Oryctolagus cuniculus, and Mus musculus genomes (%) (https://www.repeatmasker.org/genomicDatasets/RMGenomicDatasets.html). The frequencies of ancient and young SINE and LINE repeats in MLPH flanking regions were similar in humans and cattle. This is probably because MLPH genetic linkage with neighborhood genes is more evolutionarily conserved in humans and cattle (Table 1). There were co-occurrences in all three species of some TEs from different families in MLPH, MSTN, and LEPR flanking regions. To better understand the dependencies between different transposons, we estimated Spearman’s rank correlation coefficients between them. Intragroup correlations Here, for simplicity, we consider conventional p < 0.05 as a threshold to indicate “statistical significance”, although we understand it is a controversial and subjective matter (Wasserstein and Lazar, 2016). There were no statistically significant Spearman’s correlations between all ancient and all young repeats, ancient and DNA, young and LTR repeats for MLPH, MSTN, and LEPR flanking regions (ATRX was excluded from this analysis because the mammalian X chromosome has unique features that could bias the analysis). However, there were significant correlations between the presence of MIR + L2 and LTR repeats (positive correlation, r = 0.862, p = 0.003) and L1 and DNA repeats (negative correlation, r = −0.767, p = 0.026). Pairwise correlation analysis in nine flanking regions (1 Mb each) for different transposons in three genes resulted in only one statistically significant correlation: between young SINE (SINE/Alu, SINE/tRNA-Core-RTE) and LTR (negative correlation, r = −0.883, p = 0.004). It should be noted again that the classification into “young” and “ancient” repeats is rather subjective because L1 and MIR groups consist of repeats originating in different times, with different evolutionary histories and species-specific and functional peculiarities. Next, for the 5′ and 3′ flanking regions (1 Mb length), repeat frequencies were estimated separately (Tables 4, 5). For both the 5′ and 3′ flanking regions, the frequencies of young SINE, ancient SINE, and ancient LINE/L2 were significantly correlated. There were no correlations between the frequencies of young LINE/L1, as well as endogenous retroviruses (LTR). The correlation between DNA transposons for both flanking regions was almost significant (r = 0.680, p = 0.058). TABLE 4Distribution of dispersed repeats in 5’ (1 Mb) flanking regions of MLPH, MSTN, and LEPR genes for human, cattle, and rabbit species.SpeciesGeneSINELINELTRDNAHumanAlu, SINE/tRNA-CMIRL1L2 MLPH 413162253129188157 MSTN 333187387178244221 LEPR 345420251314232239Cattle MLPH 353186442361138235 MSTN 343133324251250156 LEPR 412189292406223175Rabbit MLPH 486532893513577 MSTN 566287185131130172 LEPR 549734556113064 Distribution of dispersed repeats in 5’ (1 Mb) flanking regions of MLPH, MSTN, and LEPR genes for human, cattle, and rabbit species. TABLE 5Distribution of dispersed repeats in 3′ prime (1 Mb) flanking regions of MLPH, MSTN, and LEPR genes for human, cattle, and rabbit species.SpeciesGeneSINELINELTRDNAHumanAlu, SINE/tRNA-CMIRL1L2 MLPH 216182278151325198 MSTN 175145300163260136 LEPR 265288375257233185Cattle MLPH 383324341493294213 MSTN 375137363382182133 LEPR 472158266479232191Rabbit MLPH 641713046915388 MSTN 515313246167190163 LEPR 68011429791153115 Distribution of dispersed repeats in 3′ prime (1 Mb) flanking regions of MLPH, MSTN, and LEPR genes for human, cattle, and rabbit species. For the 5′ flanking regions (for all genes considered), there were two statistically significant correlations: negative between young SINE and LTR (r = −0.930 p < 0.001) and positive between ancient LINE/L2 and LTR. In addition, there was almost significant correlation between SINE/MIR and LINE/L2 (r = 0.690, p = 0.076). For the 3′ flanking regions, there was almost significant correlation between SINE/MIR and LINE/L2. However, statistically significant correlations were found between SINE/MIR and LTR (r = 0.690 p = 0.038), SINE/MIR and DNA transposons (r = 0.850 p = 0.007), and LTR and DNA transposons (r = 0.860, p = 0.003). Similarly, for 5′ flanking regions, negative correlation between young SINE and LTR (r = −0.780 p = 0.014) was also found in the 3′ flanking regions. Despite the limited sample size, structural complexity, and different evolutionary origin of every TE group, the straightforward intragroup correlations that we observed suggest potential cooperative and exclusive associations between the TE groups. Even in the limited genomic regions for different species, we found almost statistically significant positive correlations between SINE/MIR and LINE/L2, presumably because the replication machinery of non-autonomous SINE/MIR ancient retrotransposons depends on the reverse transcriptase of autonomous ancient retrotransposon LINE/L2. Statistically significant negative correlations were also found for two groups: widespread active autonomous retrotransposons LINE/L1 and DNA transposons. It appears that young SINEs and LINEs, due to their relative activity, disrupt the balance between ancient SINEs, LINEs, and LTR, and DNA TEs and genome sizes. It should be noted that a certain competition between SINEs, LINEs, and ERVs has been observed in the genomes of South American rodents (Erickson et al., 2011), as well as between SINEs, LINEs, and DNA transposons in bats (Ricci et al., 2023). It can be expected that such interactions may be due to the functional similarity of different TEs in the presence of regulatory motifs and purifying selection against functionally defective TEs, as well as limits to genome expansion. Discussion The MLPH, MSTN, LEPR, and ATRX genes of all three species have variable length, number of exons, and dispersed repeat frequencies (non-autonomous retrotransposons SINE, autonomous retrotransposons LINE, endogenous retroviruses LTR-ERV, DNA transposons, and tandem microsatellite repeats, STR) inside the genes and their flanking regions. The number of TEs inside the genes was significantly different: MSTN (extensively used for successful gene editing in livestock) had the least amount of TEs and LEPR and ATRX had the most. There were also significant differences in the ratio of “young” and “ancient” SINE and LINE repeats. For humans and cattle, the balance between SINE and LINE inside LEPR gene and in its flanking regions was shifted toward ancient repeats compared to MLPH and MSTN genes (Figures 1–6; Tables 4, 5). In this context, it should be noted that only LEPR and ATRX have gene insertions inside: LEPR has the gene LEPROT (its sequence overlaps with the LEPR gene itself) and human ATRX has FABP5P15 pseudogene (fatty acid binding protein 5 pseudogene 15, GRCh38.p14, NC_000023.11), as well as long non-coding RNA (lncRNA). In addition, the high frequencies of young SINE and LINE in LEPR and ATRX suggest they are less insulated from DNA insertions than MLPH and MSTN (specifically MSTN, Figures 1–3). Moreover, LEPR belongs to an evolutionarily conserved gene cluster (Kosovsky et al., 2024), and ATRX in mammals is always located on the X chromosome. Thus, TEs can also target evolutionarily conserved gene blocks, not only the evolutionarily “fragile” chromosomal regions. However, the relatively high number of ancient (SINE/MIR + LINE/L2) versus young TEs (SINE/Alu and SINE/tRNA + LINE/L1) in the flanking regions of LEPR, compared to the MLPH and MSTN flanking regions in humans and cattle (Figure 6), could be associated with LEPR being in a genetically-linked evolutionarily conserved gene cluster. Questions about the origin, evolutionary mode, and functions of dispersed repeats in the genome are still largely open, despite more than two decades of active research (Jordan et al., 2003). The origin of dispersed repeats is complex; for example, the SINE/Alu repeat originated from the 7SL RNA sequences, a non-coding RNA that is a key component of the signal recognition particle and is essential for protein secretion and translocation into the endoplasmic reticulum. Another example is SINE/tRNA, which originated from tRNA genes; some other repeats are recombination products and overlap, so their origin is untraceable (Sharma et al., 2024). Domestic rabbits have two families of dispersed repeats—OcuSINEA and OcuSINEB—with their central parts originating from LINE and not tRNA, demonstrating that SINE can coopt LINE’s endonucleases (Yang et al., 2021). However, based on the similar frequencies of SINE/Alu in human and SINE/tRNA in cattle and rabbit genomes and the presence of similar regulatory elements in both repeats, it is reasonable to assume that these repeats could have some exchangeable functionalities. For example, SINE/Alu, other SINEs, and some LTR and DNA transposons (e.g., DNA/hATCharlie) contain the same motif for transcriptional regulatory factor CTCF (responsible for chromatin higher order structure). Even without the CTCF binding motif, the clusters of SINE/Alu and SINE/MIR influence local chromatin organization (Gunsalus et al., 2023). In addition, SINE/MIR, LTR, and SINE-VNTR-Alu form a “transposon code”, providing particular patterns of transcription factor binding sites for regulatory networks (Testori et al., 2012; Trizzino et al., 2017). It should be noted that transcription factors that bind heavily depend on the DNA methylation of their binding sites (Rimoldi et al., 2024), and the methylation itself is an epigenetic silencing mechanism of TEs transcription that has both species-specific and common characteristics in vertebrates (Carotti et al., 2023; Ilık et al., 2024). In addition, SINE, LINE, ERV, and DNA transposons are the major sources of regulatory network dynamics, regulating transcription factors, providing promoter, enhancer, and silencer sequences, histone and microsatellite codes, methylation and imprinting, and different non-coding RNAs (Mandal, 2024). For example, in different mammalian species, not only do all TE groups contain transcription factor-binding sites but also many genes encoding these factors contain different TE groups (Du et al., 2024). In summary, the accumulated evidence suggests that all transposons, even from different classes, despite their structural differences and evolutionary origin, have common characteristics that are related to the mechanisms of their replication, transposition, genomic integration, involvement in 3D genome organization, and interactions with host genes. The interaction could be at the level of neighboring genes and could also be at varying distances. Thus, the common trait of all different TEs is their ability to participate in the host’s regulatory networks by providing a multitude of different regulatory mechanisms to them. Their integration and distribution are defined not by external factors but by instant need inside the genome organization that can be solved by any available source of regulatory element, at any evolutionary moment, thus optimizing survival and successful reproduction. In other words, TE repeat integration is defined not by group membership but by the availability of a TE with necessary regulatory elements. For example, some TEs play a supporting function in the interphase nucleus architecture (Jordan et al., 2003; Falk et al., 2019). This “internal” genomic adaptation was first defined by the architectural term “spandrel” by Gould and Lewontin (1979) and was later adopted by many others (Koonin, 2011). It can be expected that the accumulation of TEs and their turnover is a consequence of the non-adaptive accumulation and exchange of various “junk” genomic elements, functionally convergent in the presence of regulatory motifs (Koonin, 2016). In the area of finding the most effective markers associated with phenotypic diversity, this approach could shift attention from finding SNPs in causative loci or finding specific genomic elements toward estimating polymorphism for the specific members of regulatory networks.