The immune system constitutes a vital defense mechanism through which multicellular organisms recognize and eliminate pathogenic antigens to maintain homeostasis. This complex biological system is functionally divided into two principal components: the innate (non-specific) immune system and the adaptive (specific) immune system (1). The innate immune system serves as the first-line defense in vertebrates, providing rapid non-specific responses through multiple protective mechanisms. These include mechanical barriers such as mucosal epithelia in the respiratory tract, gastrointestinal barriers, urogenital tract defenses, and molecular antimicrobial factors. In contrast, the adaptive immune system mediates antigen-specific responses characterized by immunological memory. This evolutionarily advanced system, which emerged approximately 500 million years ago through the development of antigen receptor gene rearrangement mechanisms, demonstrates two distinct effector pathways: humoral immunity mediated by antibody-producing plasma cells (differentiated from B lymphocytes) and cell-mediated immunity executed by cytotoxic T lymphocytes (2). The adaptive immune response exhibits antigen-specific recognition, clonal expansion, and long-term immunological memory formation.

Immunoglobulins are composed of four heterodimeric polypeptide chains, comprising two identical immunoglobulin heavy chain (IgH) and two identical light chain (IgL). X-ray diffraction phase analysis revealed that the heterodimers, connected by variable numbers of interchain disulfide bonds, adopt a characteristic Y-shaped configuration corresponding to a single immunoglobulin monomer. Both polypeptide chains contain variable (V) and constant (C) regions (3). The C-terminal domains, characterized by conserved amino acid sequences, constitute the constant region, whereas the N-terminal domains exhibit substantial sequence variation within approximately 110 amino acid residues, defining the variable region. The variable domains are designated as the heavy chains variable region (VH) and light chains variable region (VL), respectively. Structurally, VH comprises one-quarter of the IgH polypeptide, while VL accounts for half of the IgL chain. Both variable domains contain three frame regions (FRs) interspersed with three complementarity determining regions (CDRs) (4–6). The FRs, which are evolutionarily conserved, function primarily in maintaining the structural integrity of the immunoglobulin molecule. In contrast, CDRs demonstrate exceptional sequence diversity, directly mediating antigen recognition specificity and binding affinity. These hypervariable segments are positioned at amino acid residues H31-H35B/L24-L34 (CDR1), H50-H65/L50-L56 (CDR2), and H95-102/L89-L97 (CDR3), with CDR3 exhibiting the highest degree of sequence variation (7). Immunoglobulin isotypes differ in the domain organization of their constant regions. In mammals, five major isotypes have been characterized: IgM, IgD, IgG, IgE, and IgA (8). IgD, IgG, and IgA contain three CH domains, whereas IgM and IgE contain four CH domains. Between the CH1 and CH2 domains of IgD, IgG, and IgA, there is a proline-rich peptide segment approximately 20 amino acid residues in length. This region exhibits a high degree of spatial flexibility and is known as the hinge region (9, 10). Notably, the IgL constant region (CL) contains a single constant domain.

In vertebrate organisms, the germline-encoded immunoglobulin gene repertoire remains relatively fixed, whereas environmental pathogens exhibit continuous evolutionary diversification. To address this immunological challenge, vertebrates have evolved sophisticated molecular mechanisms to generate antibody diversity. Key mechanisms include V(D)J recombination, gene conversion (GCV), somatic hypermutation (SHM), and class switch recombination (CSR), which collectively enable the production of diverse antigen-specific immunoglobulins capable of neutralizing evolving pathogens (11–13). While CSR modifies the constant region of antibodies, the other three mechanisms enrich antibody diversity by increasing the diversity of the variable regions (14).

The V(D)J recombination process is initiated by recombination-activating gene 1/2 (RAG1/RAG2) complexes in conjunction with high mobility group (HMG) proteins. These molecular complexes precisely recognize conserved recombination signal sequences (RSS) flanking V, diversity (D), and joining (J) gene segments. The RAG endonuclease subsequently introduces double-strand breaks (DSBs) at specific RSS sites, followed by DNA repair mechanisms that mediate segment rearrangement through non-homologous end joining (15, 16). Furthermore, combinatorial assembly of two identical IgH with two identical IgL generates additional structural variation through heterodimer pairing.

To neutralize certain highly variable antigens, in addition to the four classical mechanisms of immunoglobulin diversity generation, there exist several non-canonical pathways. These atypical diversity-generating mechanisms hold significant value in both scientific research and medical applications (17). In addition to conventional antibodies composed of paired IgH and IgL, certain species have evolved the capability to produce heavy chain homodimers. Notably, camelids generate heavy-chain-only antibodies (HCAbs) that lack IgL and the CH1 domain in their IgH, while cartilaginous fish possess immunoglobulin new antigen receptors (IgNARs) characterized by a single variable domain followed by five constant domains. These atypical antibodies exhibit distinctive features including compact molecular dimensions, enhanced thermal stability, and distinct paratope configurations, thereby significantly expanding the antibody repertoire diversity in both camelids and elasmobranchs (18, 19). In bovines, a limited subset of rearranged IgH sequences exhibit exceptionally long CDR3H domains exceeding 70 amino acid residues, significantly exceeding those observed in the longest CDR3H regions of camelid IgGs and shark IgNARs. X-ray crystallographic analyses of five bovine antibodies featuring these ultralong CDR3H motifs reveal a distinctive architecture characterized by two structural components: a disulfide-bond-stabilized “knob” domain supported by an elongated β-ribbon “stalk” configuration (20). The ultra-long CDR3H region is generated through the involvement of the DH8 gene segment. The germline-encoded DH8 gene contains an even number of cysteine (Cys) residues, where alternative pairing configurations between different Cys residues significantly enhance immunoglobulin diversity. Notably, the DH8 segment harbors abundant activation-induced cytidine deaminase (AID) hotspot motifs. SHM modifies both the spatial distribution and numerical count of specific Cys residues, thereby further diversifying the immunoglobulin repertoire to a greater extent. The molecular mass of ultralong CDR3H is even lower than that of HCAb or IgNARs, potentially enabling its independent binding to epitopes inaccessible to conventional antibodies (21). This distinctive structural characteristic of ultralong CDR3H therefore provides a novel strategy for engineering antibodies targeting challenging antigenic targets. By substituting the original pivot domain with alternative proteins or peptides (e.g., GCSF, EPO, or CXCR4), functional fusion proteins with desired pharmacological properties can be generated (22–24). Given the structural versatility of bovine ultralong CDR3H, it establishes a robust structural platform for developing next-generation diagnostic tools, therapeutic agents, vaccine candidates, and immunomodulatory compounds.

Bovidae, as one of the primary categories of domesticated livestock in contemporary human society, encompasses major working or meat-producing species such as cattle, yaks, buffalo, goats, and sheep. Previous research on the structural diversity of immunoglobulins within this taxon has been constrained by sequencing technologies and genome assembly methods. To date, comprehensive characterization has only been achieved in Holstein cattle and yaks—revealing a unique evolutionary strategy in cattle and yaks that generates immunoglobulin diversity through ultra-long CDR3H domains, a biological phenomenon specific to domestic cattle among mammals (25). Nevertheless, current understanding of the immunoglobulin gene structure and expression diversity in Bovidae remains incomplete, limiting the comprehension of immunoglobulin evolutionary mechanisms within this family. To supplement evidence for immunoglobulin gene evolution in Bovidae, this study integrates genome alignment strategies with high-throughput sequencing technologies. Employing comparative genomics approaches, we systematically analyze the immunoglobulin gene structure and expression diversity in sheep. This expands the current framework for understanding immunoglobulin structural organization and diversity generation mechanisms in Bovidae, providing an important theoretical foundation for advancing disease prevention and immune intervention strategies in these animals.

All experimental procedures were conducted in accordance with the Regulations on the Administration of Laboratory Animals approved by the State Council of the People’s Republic of China.

In the analysis of the locus structure of sheep immunoglobulin heavy chain, the types and quantities of ultra-long DH and μ genes in the germline were mainly focused on. Regrettably, by locating RSS, only four DH gene fragments were identified in the sheep genome, indicating that the number of DH in the sheep germline was significantly lower than that in cattle and yak. Moreover, no ultra-long DH gene analogous to those found in cattle, yak, swamp buffalo, and river buffalo appears to exist in the sheep genome. The longest DH gene fragment identified in sheep is only 42 bp, while the longest germline DH segments in cattle, yaks, and swamp buffalo measure 149 bp, 193 bp, and 119 bp, respectively (27–30). The characteristics containing (YG)_n repetitive amino acid sequences were found in the DH2 (SYYSGYGYAYGY) and DH4 (SYYSDYGY) sequences of sheep, which were consistent with those of DH4 (SYSGYGYGYSYGY) and DH6 (SCYSGYGYGCGYGYGYDY) in domestic cattle and DH28 (SYSGYGYGGYGCYGYGYGY) and DH34 (SCYSGYGYGYGCGYGYGYDY) in yak (27, 28). The analysis of VDJ recombination preferences across three species revealed distinct preferential selection patterns in DH genes during the recombination process. In sheep, the DH2 exhibited the highest selection frequency, whereas cattle predominantly selected DH4 and DH6, yak predominantly selected DH28 and DH34. These findings suggest that Bovidae members exhibit an evolutionary tendency to preferentially select DH fragments containing (YG)n amino acid during VDJ recombination. The JH gene cluster is relatively conserved among reported bovid species. The JH clusters of sheep (Ovis aries), goat (Capra hircus), water buffalo (Bubalus bubalis), and the two JH clusters in cattle (Bos taurus) exhibit conservation in terms of JH gene number, sequence similarity, and arrangement order (27–31). All four species possess six JH genes within their JH clusters. Among these, JH1, JH2, JH3, and JH5 are pseudogenes, while JH4 and JH6 are functional genes. The primary expression is from JH4 (JH10 in cattle), with low-level expression of JH6 (JH12 in cattle). Notably, cattle possess two JH gene clusters. The JH1–JH6 cluster and the JH7–JH12 cluster display extremely high sequence similarity and identical arrangement order. Although only three JH genes, all pseudogenes, were identified in the yak (Bos grunniens) genome, reads obtained from 5’RACE data indicate that the primarily expressed JH gene in yak shares high identity with bovine JH4 (JH10). This suggests that the yak JH gene cluster may be similar to those of other bovids. Water buffalo harbors eight JH genes. JH1-1 to JH1-6 constitute the first JH cluster, while JH2-1 and JH2-2 potentially represent a second JH cluster. The predominantly expressed JH gene in water buffalo shows high identity with the bovid JH4 gene.

Comparative genomic analysis revealed that sheep, similar to yak, possess only a single μ gene (21, 28). Sequence alignment demonstrated that the ovine μ gene shares 94.5% similarity with bovine μ2 and 93.8% with bovine μ1, whereas the yak μ gene exhibits 98.7% sequence identity with bovine μ2. Notably, previous reported differential VDJ recombination patterns in cattle, showing preferential utilization of μ1 gene segments paired with JH6 elements and μ2 gene segments associated with JH10 during immunoglobulin rearrangement (28). These phylogenetic and functional observations collectively suggest that sheep, analogous to yak, may have undergone evolutionary loss of the μ1 gene through genomic deletion or pseudogenization events. Similar to yaks, sheep were found to possess only one μ gene, lacking the μ2 gene present in domestic cattle. A similarity analysis of the sheep μ gene with bovine μ genes revealed 94.5% similarity with the cattle μ2 gene and 93.8% similarity with the cattle μ1 gene. In contrast, the yak μ gene shows a significantly higher similarity of 98.7% with the domestic cattle μ2 gene. According to the research (28), domestic cattle tend to utilize the μ1 gene with JH6 and the μ2 gene with JH10 during VDJ recombination. These findings collectively suggest that sheep, similar to yaks, may have lost the μ1 gene in their evolutionary history (21, 28).

Although 25 VDJ recombinant types were identified in sheep, the three expressed VH genes shared 94% sequence similarity and clustered within the same gene subgroup. The expression of JH genes and DH genes also had strong expression preference, consistent with patterns observed in other ruminant species including goat, cattle, and yak. These findings suggested that IgH VDJ recombination contributes minimally to antibody repertoire diversity in these species. Conservation patterns were similarly observed in the Jκ gene locus, with both bovine and yak genomes containing six Jκ gene segments demonstrating high sequence homology. Comparative analysis revealed that the sheep genome lacked the Jκ1 segment, retaining only five functional Jκ gene segments. Notably, expression profiling of immunoglobulin recombination events in sheep showed differential utilization patterns, with Jκ3 representing the most highly expressed segment, followed by Jκ2 based on somatic recombination frequency analysis.Our previous investigations demonstrated that approximately 80%-90% of immunoglobulin recombination in domestic cattle and yak utilized the Jκ2 gene segment, while 10%-20% of recombinant sequences contained Jκ4. For sheep Igκ, five Vκ gene segments and three Jκ gene segments were involved in recombination, with a total of 12 recombination types, but Vκ2-8 - Jκ1 (38.9%), Vκ1-4 - Jκ3 (32.9%) and Vκ1-4 - Jκ1 (21.3%) accounted for 93.1% of the analyzed reads. The recombination process of yak Igκ involved six Vκ gene segments and four Jκ gene segments (27). A total of 14 distinct recombination types were identified, among which three predominant modes—Vκ1-1 - Jκ2, Vκ1-1 - Jκ4, and Vκ1-1 - Jκ2—were frequently observed. With genome reassembly and optimization for completeness, more germline Vλ gene segments were discovered. Previous studies had only mapped 32 Vλ gene segments, whereas this study mapped 128 germline Vλ gene segments. This also represents one of the main differences in Igλ between sheep and other bovids: yaks possess only 42 germline Vλ gene segments, swamp buffalo only 29, and goats only 35 (27, 30, 32, 33). There were significant differences in the Igλ locus between yak and sheep. Sheep possess 128 germline Vλ genes, while yak had only 45. Notably, the Jλ and Cλ genes exhibit distinct organizational patterns: sheep display a simple alternating arrangement of 3 Jλ and 3 Cλ gene segments, whereas yaks demonstrate a more complex organization with 9 Jλ and 7 Cλ gene segments arranged in the following sequential pattern: Jλ1-Jλ2-Cλ1-Cλ2-Jλ3-Jλ4-Jλ5-Cλ3-Jλ6-Cλ4-Jλ7-Cλ5-Jλ8-Cλ6-Jλ9-Cλ7. Existing studies have revealed that the Igλ loci in most mammals exhibit a conserved (Jλ-Cλ)_n arrangement pattern. Specifically, the equine Igλ locus demonstrates a Vλ_n-(Jλ-Cλ)₇-Vλ_n configuration, domestic cattle possess a Vλ_n-(Jλ-Cλ)₄ structure, pig display a Vλn-(Cλ-Jλ)₃-λ4 organization, and caprine species feature a Vλ_n-(Jλ-Cλ)₃ arrangement (33–36). Unfortunately, the Vλ5-145 – Cλ3 – Jλ3 – Vλ5-146 loci discovered in the sheep genome with opposite transcriptional orientation do not participate in rearrangement due to the incomplete FR1 of Vλ5-145 and the absence of RSS in Vλ5-146. Therefore, our study was unable to demonstrate whether Vλ-Jλ-Cλ loci with opposite transcriptional orientation can undergo rearrangement. Notably, the sheep Igλ locus conforms to this established genomic organizational principle. IMGT analysis of sheep Igλ rearranged sequences revealed the utilization of 26 Vλ genes during VJ recombination in sheep, compared with 10 Vλ genes employed by yak. Although the diversity of Vλ gene segments participating in recombination appeared considerable, these gene segments were ultimately grouped into four distinct families, exhibiting extremely high sequence similarity among subgroup members. During recombination, Jλ1 was utilized in less than 0.1% of sequences, while Jλ2 constituted the remainder. All yak sequences exclusively employed Jλ3 genes, whereas domestic cattle demonstrated recombination involving both Jλ2 and Jλ3. Sequence similarities exceeding 80% are observed in the Jλ genes utilized in recombination across these three bovid species. Although interspecies variations exist in germline Jλ genes, the recombination preference remains consistent, with all species selecting Jλ genes from this conserved subgroup for VJ recombination. Consequently, the diversity of Igλ generated through VJ recombination proves equally limited as observed in IgH and Igκ chains.

Is junctional diversity during recombination a key factor in the generation of immunoglobulin diversity in sheep? We conducted separate analyses of the junctional regions of IgH, Igκ, and Igλ through five parameters: DH segment length, CDR3 length, N/P nucleotide length, and random deletion lengths at V/J junctions.

In sheep, the longest CDR3H was 66 bp, with a mean length of 44 ± 9.7 bp. In contrast, the longest CDR3H observed in yaks is 129 bp, while in domestic cattle, it reaches 195 bp. The CDR3H lengths in sheep are predominantly concentrated at 45 bp, 48 bp, and 51 bp. Domestic cattle exhibit a predominant clustering of CDR3H lengths at 63 bp, 66 bp, and 69 bp, whereas yaks show the highest frequencies at 57 bp and 60 bp. The sheep genome did not reveal ultra-long DH gene segment, with the longest DH fragment utilized during recombination being 38 bp in length. Consequently, no exceptionally long CDR3H regions analogous to those observed in domestic cattle, yaks and buffalo were identified in the sheep immunoglobulin. The DH gene segments in human and mouse contributed 14.3 ± 5.5 bp and 10.8 ± 4.7 bp, respectively, to the CDR3H region. In cattle, the DH-Cμ1 and DH-Cμ2 segments contributed 27.5 ± 8.5 bp and 43.1 ± 25.3 bp to CDR3H length, while yak DH gene segments contributed 39.9 ± 13.7 bp (37). The average DH-derived CDR3H length in sheep was 12.9 ± 4.3 bp, which was significantly shorter than those observed in both cattle and yak, and only marginally exceeds that of mouse. Previous studies have identified ultra-long CDR3H as a unique diversity-enhancing immunoglobulin feature exclusive to domestic cattle and yak. The primary focus of this investigation was to determine whether this structural characteristic exists in other members of the Bovidae family. While ultra-long CDR3H was not observed in sheep, our analysis revealed that the distinctive (Yx)_n amino acid motif encoded in cattle and yak ultra-long CDR3H sequences was unexpectedly present in sheep CDR3H regions. This finding contradicts our initial hypothesis that the (Yx)n motif was specifically associated with ultra-long CDR3H architecture. Instead, our data suggest that Bovidae species may exhibit an evolutionary preference for utilizing DH gene segments encoding (Yx)n amino acid motifs, independent of CDR3H length characteristics.

SHM occurs in activated B cells, further increasing immunoglobulin diversity beyond the initial repertoire generated by the organism, and is also an indispensable step for antibody affinity maturation (13, 38). The SHM pattern in the sheep IgH exhibits a distinct nucleotide substitution bias, with A→G and G→A transition mutations occurring at the highest frequency, followed by T→G transversions. Among these, A→G mutations are most likely generated during the second phase of SHM: the error-prone Polη tends to A→G in WA motifs (where W denotes A/T) during the repair of DNA damage. Conversely, G→A and C→T mutations likely occur during the first phase of SHM: C→U mutations produced by AID deamination become fixed in subsequent DNA replication. These mutational signatures are evolutionarily conserved across multiple species, including domestic cattle, yak, mouse, and zebrafish, indicating that the SHM mechanism has been retained throughout vertebrate evolution (28, 39). Notably, AID preferentially targets hotspot motifs (RGYW/WRCY) for mutagenesis. AID initiates the process through cytidine deamination, creating U:G mismatches. These mismatches are then channeled into distinct repair pathways via different molecular factors. This mechanistic preference leads to elevated mutation frequencies of C and G within these hotspot motifs. Consequently, significant mutation clustering is observed in FRs as well as the classical CDRs, due to the presence of these hotspots (40). Certain limitations also exist in the SHM analysis process: (1) SHM variant types may exhibit certain variations with genomic updates and the assembly of different sheep genome assemblies. (2) Figure 9E displays not only variant types but also variant frequencies. Although there may be unavoidable errors in the variant types at this locus, it can still demonstrate the richness of SHM at this site.

In summary, different species appear to have evolved distinct Ig gene rearrangement sequences and strategies for selecting functional immunoglobulin repertoires (41). Comparative analysis of immunoglobulin diversity generation mechanisms across species reveals that humans and mouse display extensive V(D)J recombination diversity, enabling them to generate a broad repertoire of immunoglobulin types through recombination to meet diverse antigenic challenges (42). In species with limited V(D)J recombination diversity, such as rabbits, this limitation is compensated for by GCV occurring in specific B-cell lineages combined with high levels of SHM, ultimately producing antibody repertoires with greater diversity than those observed in humans or mice. Structurally, rabbits rely more heavily on the IgL for antigen specificity, utilizing extended CDR3L loops and interdomain disulfide bonds (43). In camelids, B cells produce both conventional IgH/IgL paired antibodies and unique HCAbs. These HCAbs contain distinctive structural features that further enrich the camelid antibody repertoire. Previous studies found that domestic cattle, yak, swamp buffalo, and river buffalo utilize ultra-long CDR3H to enhance immunoglobulin diversity. Sheep appear to enrich their antibody repertoire through recombination diversity and junctional diversity of Igλ. Notably, while the V and J gene segments show strong conservation in bovid, the unique ultra-long CDR3H mechanism is absent in sheep, indicating that the ultra-long CDR3H is not a conserved diversity generation strategy among bovid. The evolutionary drivers behind the specific emergence of this mechanism in domestic cattle and yak require further investigation.