Understanding the structure of swine leukocyte antigen class I molecules

Abstract

Abstract

Swine leukocyte antigen class I (SLA-I) molecules are swine orthologs of human MHC class I molecules and are encoded by three classical loci, SLA-1, SLA-2, and SLA-3. By engaging T-cell receptors (TCRs), these cell-surface proteins present antigenic peptides, whose binding specificity is largely dictated by the architecture of the peptide-binding groove (PBG) and its constituent pockets. Although early SLA-I research primarily emphasized molecular and functional features, recent advances in structural biology have yielded an increasing number of crystal structures of SLA-I–peptide complexes, providing critical insights into the principles of peptide presentation and T-cell recognition in pigs. In this review, we examine the structural features of SLA-I molecules, focusing on the PBG and binding pockets that accommodate peptide anchor residues. Structural studies revealed that the SLA-I PBG, while architecturally conserved, exhibits pronounced allelic polymorphism and plasticity, predominantly within its six binding pockets (A–F). A key insight is the profound functional impact of micropolymorphisms; for instance, single-residue variations in pockets such as D can dramatically alter the peptide-binding specificity and repertoire. Finally, we highlight current knowledge gaps and future research directions to facilitate the exploitation of SLA-I structural features for targeted vaccine design and immunotherapeutic development, thus addressing the challenges of porcine immune recognition.

Graphical abstract

Introduction

Major histocompatibility complex (MHC) molecules are a family of transmembrane glycoproteins expressed on the surfaces of vertebrate cells and play pivotal roles in immune responses, antigen presentation, and tissue transplantation [1–3]. MHC molecules are classified into three distinct categories: MHC class I (MHC-I), class II (MHC-II), and class III (MHC-III). Classical antigen presentation theory posits that MHC-I molecules primarily present endogenous antigens to CD8⁺ T lymphocytes via their T-cell receptors (TCRs), thus initiating cytotoxic T-cell-mediated immune responses [4, 5]. In contrast, MHC-II molecules predominantly facilitate the presentation of exogenous antigenic peptides to CD4⁺ T cells, leading to the activation of helper T-cell responses and subsequent humoral immunity [4, 6]; moreover, MHC-III molecules encode mainly complement system components rather than directly participating in antigen presentation [7, 8]. Structurally, MHC-I molecules were first characterized by Cresswell et al. (1973), who identified them as heterodimers consisting of a glycosylated heavy chain (30–31 kDa) and a noncovalently associated light chain (11–12 kDa) [9]. Subsequent work by Poulik et al. [10] revealed that the light chain corresponds to β2-microglobulin (β₂m), a short polypeptide, while the heavy chain is a longer glycosylated subunit. Genomic sequencing of the murine MHC-I molecule (H-2 kb) revealed that the H-2 kb gene comprises eight exons and encodes multiple functional regions, including a signal peptide, extracellular regions, a transmembrane segment, and an intracellular region [11]. The extracellular portion itself is subdivided into three functional domains, initially designated as extracellular domains 1–3 and later redefined as the α1, α2, and α3 domains [12, 13].

However, the tertiary structure of MHC-I molecules remained elusive until Bjorkman et al. [14] successfully determined the crystal structures of human leukocyte antigen (HLA)-A2 and HLA-A28 using X-ray diffraction analysis. Their structural data revealed a twofold-symmetric organization in HLA-A, with the homologous domains paired as α1-α2 and α3-β₂m. Subsequently, Bjorkman et al. [13] elucidated the HLA-A2 structure, demonstrating that the αl and α2 domains form a peptide-binding groove (PBG) consisting of an eight-stranded antiparallel β-sheet platform flanked by two α-helices. This groove accommodates antigenic peptides that are typically 8–11 amino acids in length, as later corroborated by independent studies [15, 16]. While the α3 domain and β₂m are spatially adjacent, they lack covalent interactions. Both domains adopt an immunoglobulin-like β-sandwich fold, each comprising two antiparallel β-sheets, one with four β-strands and the other with three strands stabilized by an intramolecular disulfide bond. Following these pioneering studies, the tertiary structures of MHC-I molecules were resolved in other species, including mice [17], rhesus macaques [18, 19], cattle [20], horses [21], chickens [22], dogs [23], bony fishes [24], cats [25], bats [26], giant pandas [27], Xenopus laeviscats [28], rabbits [29], lizards [30], sharks [31], and marsupial brushtails [32]. Notably, despite their evolutionary divergence, these MHC-I molecules exhibit a highly conserved overall structure.

The swine MHC complex, commonly referred to as the swine leukocyte antigen (SLA) complex, represents one of the most gene-dense regions in the porcine genome, spanning approximately 2.4–2.6 Mb [33]. Initial characterization of SLA began with the studies of Vaiman et al. in 1970 [34], followed by the first isolation of SLA molecules from boar sperm in 1978 [35]. The SLA complex is organized into three primary gene clusters corresponding to SLA classes I, II and III (SLA-I, SLA-II, and SLA-III, respectively) [33]. This genomic region spans the centromere of swine chromosome 7, with a distinct spatial arrangement: the class II region occupies the long arm, whereas class I and III regions are located on the short arm, with class I positioned telomeric to class III [36]. Owing to limited genomic synteny between SLA-I and HLA-I genes, the porcine loci have been systematically designated as SLA-1, SLA-2, and SLA-3 to prevent confusion with their human counterparts. Functionally, SLA-I molecules are ubiquitously expressed on nucleated cells and play a critical role in CD8⁺ T-cell-mediated cellular immunity [37, 38].

Structural elucidation of SLA-I molecules was first determined by Zhang et al. [39] in 2011, with the crystal structure of SLA-1*04:01. This was followed by Fan et al. [40], who resolved the SLA-3 structure. Although the structural characterization of SLA-2 was reported later, it has since been extensively investigated, yielding detailed structural information [41–45]. Comparative structural analyses revealed that SLA-I molecules share fundamental architectural features with HLA-A and other mammalian MHC-I molecules. Nevertheless, potential structural variations or unique domain characteristics in SLA-I molecules warrant further investigation to fully understand their functional implications. Although multiple crystal structures of SLA-I molecules have been reported, a systematic synthesis of their structural diversity and functional implications remains limited. This review will present a detailed overview of the research on the structure of SLA-I molecules as follows.

SLA-I genome organization

The human major histocompatibility complex, also referred to as human leukocyte antigen (HLA), contains 220 genes in 4.6 Mb of DNA on chromosome (Chr) 6p21.3 [46]. The human MHC is subdivided into three main regions: class I (MHC-I), class II (MHC-II), and class III (MHC-III). Among them, MHC-I is larger than the other two. It spans approximately 2000 kb and includes classical HLA class I (HLA-A, HLA-B, and HLA-C) genes and nonclassical class I (HLA-E, HLA-F, and HLA-G) genes, with 224 identified gene loci [47], including some pseudogenes in this region, which can be explained by the large series of duplications and deletions associated with the evolution type of a birth-and-death model of gene duplication and loss [48] or by pathogen-driven balancing selection [49].

The SLA complex corresponds to a genomic region of 2.40–2.6 Mb on chromosome 7 that spans the centromere (Fig. 1). SLAs are also classified into three groups: class I (SLA-I), class II (SLA-II), and class III (SLA-III). SLA-I and SLA-III are located on 7p1.1, and SLA-II is located on 7q1.1. The split of the swine MHC genome at the centromere is unique among mammals studied to date [33]. The SLA-I region spans approximately 1.01–1.06 Mb, with 54–55 loci identified. The gene encoding the heavy chain α of SLA-I molecules is located on chromosome 7 [50], whereas the gene encoding the light chain β₂m is located on chromosome 1q17 [51]. The SLA-I heavy chain α genes are categorized into three types (Ia, Ib, and Ic) [52], encompassing 12 loci. Among these, there are seven classical class I loci, namely, SLA-1, SLA-2, SLA-3, SLA-4, SLA-5, SLA-9, and SLA-11. However, only SLA-1, SLA-2, and SLA-3 have been confirmed to be functionally expressed with intact coding regions [36]. Notably, SLA-2 differs from SLA-1 and SLA-3 by three amino acid residues in the signal peptide. SLA-1, SLA-2, and SLA-3 are highly polymorphic [33, 53]. Currently, a total of 252 classical SLA class I alleles (100 SLA-1, 105 SLA-2, and 47 SLA-3) have been designated by the SLA Nomenclature Committee of the International Society for Animal Genetics (ISAG) in the Immuno Polymorphism Database (IPD) MHC SLA sequence database [54]. At these three functional loci, pathogen-driven balancing selection and gene recombination appear to be the primary drivers of their diversity, especially for SLA-1 and SLA-2 [55–57]. The polymorphisms of SLA-1, SLA-2, and SLA-3 are also associated with SLA haplotype diversity, which was generated by allele/gene conversion and recombination [57]. Tanaka et al. [58] identified an additional functional SLA-I gene, SLA-12, although its structure and functional characteristics remain to be elucidated. The SLA-4, -5, -9, and -11 genes have traditionally been annotated as pseudogenes owing to the lack of intact coding regions [33], indicating that a birth-and-death evolution model of gene duplication and loss is taking place, in accordance with the views that MHC genes are subject to the birth-and-death model of evolution because some Ia genes are deleted or become nonfunctional after gene duplications [59, 60]. In addition to the classical loci, three nonclassical class I loci, Ib (SLA-6, SLA-7, and SLA-8), that share homology with classical SLA-I genes have been identified, but their biological roles remain uncharacterized [33, 59–62]. SLA-6 should be under positive selection pressure, despite its highly conserved, low-polymorphism features [57]. Furthermore, the class I-related genes (Ic) in swine comprise only two loci, MIC1 and MIC2, which are orthologs of human MICA and MICB, respectively; however, only MIC2 is transcribed [33, 36, 50].

Fig. 1.

Genomic structure of the swine leukocyte antigen. The SLA complex is on chromosome 7. The genomic region is bounded by myelin oligodendrocyte glycoprotein (MOG) upstream of the class I region and by Ring finger protein 1 (RING1) downstream of the class II region. The nucleotide positions refer to the swine genome assembly Sscrofa11.1 (Ensembl Release 96). The anchor genes that delimit the three regions, referred to as class I, class III, and class II are highlighted in gray. These regions span 1.06 Mb, 0.67 Mb, and 0.46 Mb, respectively, within the 2.66 Mb SLA region (the centromeric region accounts for 0.47 Mb). The functional MHC genes predicted to encode proteins are highlighted. Two copies of SLA-1 and SLA-5 were found only in Hp-28.0 and in the Sscrofa11.1 assembly [33]. The figure was drawn with Microsoft Office PowerPoint (Version 2021)

Structures of SLA-I molecules

Crystallographic investigations of MHC-I molecules have focused primarily on heterodimers comprising three extracellular domains of the heavy chain (α1–α3) and β₂m. These complexes are typically obtained either by papain cleavage of cell-surface proteins or by recombinant expression of the extracellular domains. The determination of the first crystal structures of the classical SLA-I molecule, SLA-1*0401, in complex with peptides derived from S-OIV or the Ebola virus provided critical insights into the structural basis of SLA-I–mediated antigen presentation [39]. These structures established a foundational framework that elucidates both the peptide-binding motifs and the general architectural principles governing SLA-I molecules (Fig. 2A). To date, 24 SLA-I molecules have been structurally characterized, comprehensively covering the products encoded by all three SLA-I loci (Table 1). As in other mammalian species, the SLA-I heavy chain consists of three distinct structural domains. Together, the α1 and α2 domains form an eight-stranded β-sheet platform that supports two antiparallel α-helices, creating the peptide-binding groove. The similarities in integral spatial structure among different SLA-I molecules and among other animal MHC-I molecules suggest a possible basis for functional conservation [39, 40, 43], although extensive structural data further demonstrate that the polymorphic, plastic peptide-binding groove accommodates diverse peptide repertoires.

Fig. 2.

Overall structure and peptide-binding groove of SLA-I. A Representative structure of the SLA-1*04:01 peptide complex. B Top-down view of the peptide-binding groove, highlighting the six peptide-anchor pockets (A–F). C Close-up view of the peptide-SLA-I anchoring interface. The peptide (orange cartoon) is anchored by its primary residues P2, P3, and PΩ (sticks), which fit into pockets B (light pink), D (deep salmon), and F (slate) of the SLA-I binding groove. Structural visualizations and molecular graphics were generated using the Open-source PyMOL Molecular Graphics System, version 3.1.0 (Schrödinger, LLC) [95]

Table 1.

Summary of SLA-I molecules available in the Protein Data Bank (PDB)

Locus	Alleles	Organism	Derived protein	Sequence	PDB ID	Refs.
SLA-1	SLA-1*0101	ASFV	p54	AAAIEEEDI	7emb	[64]
		ASFV	p72	YGDFFHDMV	7emd
		ASFV	p72	SLDEYSSDV	7em9
		ASFV	p72	YSSDVTTLV	7ema
		FMDV	VP1	ATEIRELLV	7emc
	SLA-1*0401	H1N1	NA	NSDTVGWSW	3qq3	[39]
		EBOV	VP35	ATAAATEAY	3qq4
		H1N1	SU	RVEDVTNTAEYW	6lf8	[65]
		FMDV	VP2	MTAHITVPY	6kwk	[82]
	SLA-1*0401 R156A mutant	FMDV	VP2	MTAHITVPY	6kwl	[82]
	SLA-1*1301	EIAV	SU	RVEDVTNTAEYW	6lf9	[69]
		H1N1	NA	ESDTVGWSW N1E mutant	6kwo	[82]
	SLA-1*1301 F99Y mutant	H1N1	NA	NSDTVGWSW	6kwn
	SLA-1*1402	H1N1	NP	EFEDLTFLA	5npz	[45]
		H1N1	NP	DFEREGYSL	5nq0
	SLA-1*1502	PRRSV	NSP9	TMPPGFELY	5ylx	[94]
	SLA-1*Z0301	PRRSV	GP3	ALLSSKTSV	9ul1	[75]
SLA-2	SLA-2*HB01	FMDV	VP1	ALLRSATYY	8gqv	[43]
		FMDV	VP1	ALLRTATYY	8gqw
	SLA-2*040202	FMDV	VP2	MTAHIVVPY	6a6h	[44]
	SLA-2*1001	ASFV	MGF_110-13 L	YMNCSLPTY	8jv0	unpublished
		ASFV	MGF 505-2R	TMYSLGYIF	8w6l
	SLA-2*1104	H1N1	NP	IAYERMCNI	5nq2	[45]
SLA-3	SLA-3*hs0202	H1N1	HA	KMNTQFTAV	5h94	[40]

The membrane-proximal α3 domain, alongside β₂m, belongs to the immunoglobulin superfamily and is contiguous with the antigen-binding platform (α1/α2). In SLA-I molecules, β₂m interacts noncovalently with both α3 and the base of the α1/α2 platform. Structural analyses indicate considerable variation in the orientation and interactions between α3 and β₂m, observed both across species (e.g., human versus mouse) and among classical SLA-I alleles, suggesting functional and evolutionary divergence in peptide presentation. The functional implications of such conformational variation are often species-specific. A notable example is the murine system, where the use of human β₂m enhances the stability of peptide-loaded complexes and improves the performance of MHC-I–peptide complexes, an effect attributed to key structural differences in the β-strand of human β₂m that interact with the binding platform [63]. In contrast, such a functional benefit is not observed in the porcine system. Comparative studies indicate that replacing porcine β₂m with human β₂m in SLA-I–peptide complexes does not confer substantial structural or functional advantages [45]. This suggests that the architecture of the swine heavy chain is intrinsically optimized for interaction with its cognate porcine β₂m.

Peptide-binding mechanism of SLA-I molecules

The conformation of peptides bound to SLA-I molecules is not static but varies significantly, primarily depending on the peptide length and amino acid sequence. These peptides typically adopt a conserved backbone conformation at their N- and C-termini, which are anchored deeply within the pocket A and F of the peptide-binding groove, respectively (Fig. 2B). Current research indicates that for canonical nonapeptides presented by SLA-I molecules, the P2, P3, and P9 positions serve as anchor residues for specific binding to the peptide-binding groove of the MHC class I molecule, ensuring the stable presentation of the peptide to the TCR [39, 64] (Fig. 2C). However, the central region of the peptide (approximately positions P4–P6) displays considerable structural plasticity. Peptides containing bulky or β-branched residues in this central region often arch upward, forming a prominent bulged conformation that serves as a primary contact point for the TCR. In contrast, peptides with small, flexible residues in the center can lie flatter within the groove, resulting in an extended conformation. Furthermore, peptide length is a critical determinant; longer peptides (e.g., > 10-mer) frequently accommodate the fixed distance between the anchor pockets by kinking or bulging, whereas canonical 9-mer peptides are more likely to adopt a straighter backbone [65]. These conformational nuances, shaped by the polymorphic nature of the SLA peptide-binding groove, collectively form the composite surface that the TCR scrutinizes, thus providing the structural basis for the exquisite specificity of cytotoxic T-cell responses in swine.

As noted above, the structural characterization of the SLA-I–peptide complex has provided substantial insights into the antigen presentation mechanisms employed by SLA-I molecules. Notably, these structural and functional features have since been corroborated in MHC-I molecules across diverse mammalian species, including humans and mice. A common observation is that in mammalian MHC-I–peptide structures, even identical peptides can adopt distinct conformations when bound to the same MHC class I molecule [66]. This plasticity has been frequently reported in SLA-I molecules and has been observed across multiple alleles encoded by different loci [39, 40, 43]. Notably, these conformational variations often occur outside the canonical anchor residues and terminal binding groove regions but instead involve intermediate peptide residues that induce structural adjustments. However, the functional implications of such peptide plasticity remain experimentally unexplored. The observed conformational disparity in the peptide backbone between the two molecules in the asymmetric unit may be due to crystal-packing constraints. Differential lattice contacts could selectively stabilize one conformation over the other, representing a localized crystal artifact rather than a biologically relevant state. Although the heterogeneity might also reflect intrinsic peptide flexibility within the binding groove, the influence of the crystalline environment cannot be ruled out. Therefore, caution is warranted when attributing functional significance to these structural differences without additional biochemical validation.

Through the structural superposition of currently available SLA-I–peptide complexes, a high degree of architectural conservation is observed in the overall fold of the heterotrimeric assembly. However, a detailed comparative analysis revealed that significant conformational deviations are primarily localized to nonfunctional flexible regions (Fig. 3A). These variations are most prominent in the solvent-exposed loops connecting the β-strands and in the distal regions of the α3 domain, which exhibit higher B-factors and greater structural plasticity. Beyond these flexible regions, the most critical structural divergences are driven by the high degree of polymorphism within the PBG. Specifically, the amino acid substitutions at key polymorphic sites significantly remodel the internal architecture of the A–F pockets. These localized conformational changes can range from subtle side-chain rearrangements to reshaping of the pocket volume and electrostatic potential, ultimately dictating the peptide-binding motif and the selection of the immunodominant repertoire. Consequently, while the SLA-I framework remains stable, the interplay between distal flexibility and groove polymorphism provides the structural basis for the broad yet specific antigen presentation observed across different swine leukocyte antigen alleles. This structural orchestration establishes the spatial framework within which the bound peptides must adapt.

Fig. 3.

Overall structural comparison of the peptide-binding grooves (PBGs) in SLA-I. A Superposition of resolved SLA-I crystal structures, shown as a ribbon representation. The PBG region is highlighted to illustrate peptide conformations. The enlarged view shows the superimposed conformations of nonamer peptides from the resolved SLA-I structures, with peptides from SLA-1, SLA-2, and SLA-3 colored green, light blue, and red, respectively. The black circles indicate the relative positions of the peptide residues (epitope sites). The arrows denote the orientations of side chains at different positions, illustrating the characteristic “M-shaped” topology and the distribution of solvent-exposed versus anchor residues. Structural visualizations and molecular graphics were generated using the PyMOL Molecular Graphics System, version 3.1.0 (Schrödinger, LLC) [95]. B Heatmap analysis of pocket volumes within the PBG of resolved SLA-I crystal structures. Pocket volume calculations were performed using the CASTpFold server (https://cfold.bme.uic.edu/castpfold/) [96]. C Heatmap analysis of the average pocket hydrophobicity within the antigen-binding cleft of resolved SLA-I crystal structures. Hydrophobicity was calculated using the Kyte–Doolittle scale [97]. Higher scores indicate greater pocket hydrophobicity, suggesting a preference for hydrophobic side-chain binding

Given this intricate groove architecture, SLA-I can constrain antigenic peptides into a conserved “M-shaped” conformation (Fig. 3A). Structural surveys across resolved SLA-1, -2, and -3 peptide complexes reveal a universal presentation logic. The nonamer peptides are anchored via a rigorous P2-P3-P9 triad, where P2 and P9 serve as canonical anchors and P3 acts as a pivotal primary anchor, a feature that distinguishes many SLA alleles from their human counterparts. In this structural arrangement, P1, P4, and P8 are oriented toward the solvent, forming the essential topographical interface for TCR recognition. In contrast to the rigid anchoring of the termini, the central segment comprising P5 and P6 functions as an “adaptive bridge.” The conformation of this bridge is not fixed but is dynamically tuned by the physicochemical environment of the center of the groove. This spatial coordination ensures that while the primary anchors (P2, P3, P9) provide the requisite binding energy, the central bulge (P5, P6) and the TCR-facing residues (P1, P4, P8) provide fine-grained specificity. This “M-shaped” paradigm highlights a remarkable evolutionary convergence in which the porcine MHC maintains a highly standardized structural language for immune communication despite the underlying diversity of its scaffold and allelic variants.

Quantitative assessment of the PBG across resolved SLA-I–peptide complexes reveals a distinct physicochemical hierarchy among the six pockets (A–F). A, B, and F pockets exhibit significantly larger volumes and more pronounced, deeply recessed morphologies (Fig. 3B), providing the primary spatial framework for peptide anchoring. In contrast, the C, D, and E pockets exhibit relatively constricted volumes, with the centrally located pocket C being the most spatially limited (Fig. 3B). This spatial arrangement is complemented by a striking gradient of hydrophobicity, where pocket B emerges as the most hydrophobic region, followed by A and F (Fig. 3C). In contrast, the intermediate pockets (C, D, and E) display significant variability, ranging from hydrophobic to hydrophilic, across different SLA-I alleles. Furthermore, the structural plasticity of these pockets is exemplified by alleles such as SLA-1*0101 and SLA-2*1001, which undergo localized conformational adjustments to accommodate diverse peptide ligands (Fig. 3B). This architectural flexibility, coupled with the varied physicochemical properties of the internal pockets, suggests that the SLA-I groove is not a rigid template but a dynamic environment tuned for selective yet versatile antigen presentation.

The micropolymorphism in the pockets (A-F) of SLA-I molecules

The concept of “pockets,” first introduced by Wiley and colleagues, delineates specific regions within the peptide-binding groove and has been consistently used to characterize SLA-I molecules [67, 68]. In SLA-I molecules, six pockets (A–F) have been identified (Fig. 2B). Their size, shape, and chemical and electronic properties are determined by the distinctive side chains of each SLA-I molecule. Within the SLA-I–peptide complex, solvent-exposed peptide residues directly engage the TCR, whereas buried residues indirectly influence engagement by altering peptide conformation and dynamics. It is widely accepted that pocket A, a highly conserved core structure within MHC class I molecules, functions primarily as a universal “N-terminal anchor” that stabilizes the N-terminus of antigenic peptides via its conserved architecture. In SLA-I molecules, pocket A is structurally regulated by two distinct regions, often described as “gates,” which collectively determine its spatial dimensions. The first gate comprises residues at positions 59 and 167. Tyr59 is highly conserved across mammals, whereas position 167 is typically occupied by Ser or Trp. These residues modulate pocket accessibility through steric effects determined by their side-chain volumes. The second gate consists of residues Glu55 and 170, with Glu55 being generally conserved in mammals, whereas position 170 is predominantly occupied by Glu, Arg, or Lys. In contrast to the steric control exerted by the first gate, the second gate primarily influences the openness of pocket A through electrostatic interactions. On the basis of these structural features, specific residue combinations give rise to two distinct conformational states of pocket A: an open state and a closed state (Fig. 4A, B). In the open state, the first gate is dominated by small side-chain residues, whereas in the second gate, electrostatic repulsion between Glu55 and Glu170 favors an open conformation, resulting in low selectivity for residues at the P1 position (Fig. 4A). However, this open state might favor binding long peptides (> 10-mer) [69]. In contrast, the closed state is characterized by large side-chain residues at the first gate, which impose steric constraints. Concurrently, electrostatic attraction between Glu55 and Lys170 in the second gate leads to occlusion of the N-terminal segment of the peptide-binding groove. This closed conformation is likely permissive only to residues with smaller side chains. Therefore, this state is more likely to bind short peptides with 8–10 amino acids [65, 69]. Interestingly, certain alleles, such as SLA-1*0401, increase peptide binding via an N-terminal extension (P-1 residue) beyond the primary anchor residue at P1 [65] (Fig. 4C). This mechanism facilitates the binding of longer peptides and thus expands the peptide length constraints typically imposed by SLA-I molecules.

Fig. 4.

Structural characteristics of pocket A of SLA-I molecules. A and B Two conformational states of pocket A. The open (A) and closed (B) states of pocket A are shown. Key residues involved in the conformational change are depicted as sticks, colored by residue type: acidic residues in red, basic residues in blue, and other residues in green. C Structural details of pocket A in pSLA-1*0401_RW12. Pocket A is depicted as a molecular surface. The peptide residues R-P-1 and V-P1 (pink sticks) form hydrogen bonds (yellow dashed lines) with key pocket A residues E63 and Y159 (green sticks). Structural visualizations and molecular graphics were generated using the Open-source PyMOL Molecular Graphics System, version 3.1.0 (Schrödinger, LLC) [95]

The principal variation among MHC alleles occurs in the central region of the peptide-binding groove, where the size and conformation of solvent-exposed side- chains, especially those on the groove floor, determine the architecture of the bound peptide, governing whether it lies flat and extended or forms a central bulge. These structural differences directly specify allele-specific peptide-binding motifs. Although the overall backbone architecture is conserved, the chemical and steric characteristics of pockets B–E further refine the specificity and configuration of the binding cleft. In contrast to pocket A, pocket B in SLA-I molecules likely functions as the initial peptide selector, facilitating the loading of optimal peptides while excluding those with low affinity [64]. In general, pocket B has 7–10 residues and binds the P2 peptide side chain that defines SLA binding motifs; pocket B in multiple SLA-I molecules exhibits considerable plasticity, as observed for alleles such as SLA-2*HB01 and SLA-3*hs0202 [40, 43]. These alleles display a pronounced preference for hydrophobic residues with extended side chains, thus imposing a defining constraint on peptide binding specificity.

As the third major restrictive pocket after B and F, pocket D has peptide binding selectivity that is highly dependent on microvariations among its constituent residues, such as those at positions 99, 114, and 156 [39, 40, 64, 69] (Fig. 5). For instance, position 156 has been reported to be a pivotal “functional switch.” In alleles such as SLA-1*0401, the presence of Arg156 exerts a “one-ballot veto” function by imposing rigorous steric and electrostatic constraints on the plasticity of pocket D. It acts as a primary gatekeeper for peptide selection (Fig. 5A). Furthermore, SLA-1*0101 displays distinct peptide-binding properties that differentiate it from other MHC-I molecules. Central to its peptide selectivity is residue Arg114 within pocket D of the binding groove, which preferentially accommodates negatively charged residues at peptide positions P5 to P7 (Fig. 5B). This electrostatic interaction induces the formation of a salt bridge between Arg114 and the acidic side chain, leading to characteristic “M-shaped” bulged peptide conformations that vary according to the position (P5, P6, or P7) of the acidic residue. In the absence of such negative charges in this region, peptides adopt a more conventional N-shaped conformation. Conformational diversity governed by Arg114-mediated electrostatic interactions thus promotes various modes of T-cell receptor recognition.

Fig. 5.

Structural characteristics of pocket D of SLA-I molecules. A Structural details of pocket D in pSLA-1*0401_3qq3. B Structural details of pocket D in pSLA-1*0101_7emb. The peptides are shown as orange and yellow sticks, with the corresponding residues highlighted in pink. Pocket D is shown in surface representation, with its constituent residues depicted as sticks. Three key polymorphic residues (Y99, R114, and R156) are highlighted with red labels. Schematic illustrating how micropolymorphisms at these key positions (Y99, R114, and R156) alter the physicochemical properties of pocket D, thus determining its peptide binding selectivity. Structural visualizations and molecular graphics were generated using the Open-source PyMOL Molecular Graphics System, version 3.1.0 (Schrödinger, LLC) [95]

Micropolymorphisms significantly influence the structural and functional characteristics of pocket D in SLA-I molecules. Even single amino acid substitutions at position 99 (e.g., Tyr99 in SLA-1*0401 to Phe99 in SLA-1*1301), which represent a canonical form of micropolymorphism, can markedly modulate both the strength and specificity of interactions with peptide anchor residues by altering the pocket’s steric conformation, charge distribution, or hydrophobic properties [69]. Consequently, the functional constraint of pocket D is not determined solely by its macrostructural architecture but is precisely regulated by micropolymorphisms at these critical positions, ultimately shaping the peptide-binding repertoire of SLA-I molecules. This suggests that adaptive evolution of pocket D, driven by micropolymorphisms, may be a key mechanism conferring optimized antigen-presenting capability on SLA-I molecules, thus providing a selective advantage against pathogenic pressures.

The available structural data consistently demonstrate that pocket F of the SLA-I molecule, an essential functional region formed by both conserved and polymorphic residues, anchors the C-terminus (PΩ) of bound peptides (Fig. 6A). This pocket is characterized by a deep, hydrophobic architecture lined with conserved aromatic residues (e.g., Tyr84, Tyr118, Tyr123, Lys146, and Trp147) (Fig. 6A, B), which creates a pronounced chemical preference for the hydrophobic and aromatic side chains of the peptide’s C-terminal anchor residue [39, 40, 43, 64] (Fig. 6C). Beyond these conserved elements, ten polymorphic residues (73, 74, 77, 80, 81, 95, 97, 114, 116, and 143) within pocket F further modulate the steric and chemical restriction of the PΩ residue. For instance, unlike most SLA-1 alleles, SLA-2*11:04 and SLA-3*hs02:02 possess pocket F that preferentially accommodates hydrophobic residues with shorter side chains, likely because of specific polymorphic variations that constrain the pocket volume (Fig. 6B, C). By fine-tuning the pocket’s dimensions and electrostatic landscape through precise interactions, these polymorphic sites collectively determine the breadth and specificity of the antigenic repertoire presented by SLA-I molecules. Notably, such structural plasticity allows the SLA-I system to balance the stable presentation of diverse peptides with the refined selection of species-specific epitopes, thus maximizing the host’s immunosurveillance capabilities against evolving pathogens.

Fig. 6.

Structural characteristics of pocket F of SLA-I molecules. A Overall view of pocket F (surface) and ten polymorphic sites (73, 74, 77, 80, 81, 95, 97, 114, 116, and 143, shown as pink surfaces). Five conserved residues (Y84, Y118, Y123, K146, and W147) are depicted as slate sticks and surfaces. B Sequence alignment of the pocket F-constituting residues in representative SLA-I alleles. Visualized and analyzed using Jalview version 2.11.5.1. [98]. C Sequence characteristics of PΩ positions based on identified nonamer peptide-binding motifs from reported SLA-I–peptide complexes. Amino acids are shown in single-letter codes, with colors indicating their specific physicochemical properties. Structural visualizations and molecular graphics were generated using the Open-source PyMOL Molecular Graphics System, version 3.1.0 (Schrödinger, LLC) [95]

Functional implications based on the structure of SLA-I molecules

SLA-I molecules are expressed in almost all nucleated cells [50]. As with human MHC-I, SLA-I molecules likely present mainly endogenous antigens. When viral particles or nucleic acids intrude into host cells, viral RNA or DNA takes advantage of the host’s system of duplication, transcription, and translation to express endogenous antigens. Endogenous antigens are subsequently ubiquitinated and degraded by the proteasome to form a series of peptides. Peptides with 8–16 amino acids are transported into the endoplasmic reticulum (ER) by transporters associated with antigen presentation transporters (TAPs). In the ER, the peptides are further trimmed by ER aminopeptidase (ERAAP) and then assembled with the free heavy chain of MHC-I and the β₂m light chain into an MHC-I complex. The MHC-I complex is further processed and modified in the Golgi apparatus for correct refolding, after which it is presented to the cell surface to interact with the CD8⁺ TCR to trigger cellular immunity [70, 71] (Fig. 7).

Fig. 7.

Mechanism of MHC class I-mediated endogenous antigen presentation. The solid arrows represent replication and transcription in the cytoplasm after an RNA virus or DNA vaccine infects a cell, the translation of all viral nucleic acids and DNA, and antigen processing and presentation after all endogenous antigens are produced. The single dotted line represents viral nucleic acid replication and transcription in the cell nucleus after a DNA virus infects a cell (except for poxviruses). The text labels for each protein in this figure are shown in the same color as the corresponding protein. ER lumen, endoplasmic reticulum lumen; HC of MHC-I, heavy chain of MHC-I. The figure was created with biorender.com (https://www.biorender.com/)

The structural evolution of MHC-I is characterized by a complex interplay between sequence conservation and functional diversification. A fundamental paradigm in immunogenetics is that the distribution of polymorphisms within MHC-I molecules is nonstochastic; instead, it follows a precise architectural logic defined by the spatial segregation of variation. This phenomenon is robustly substantiated in the SLA-I molecule.

This principle of regional constraint is most evident in the intraspecies differentiation of the classical porcine loci: SLA-1, -2, and -3. Although these lineages have followed distinct genetic trajectories, their lineage-specific motifs are predominantly sequestered within the nonfunctional backbone rather than the peptide-binding region (Fig. 8A, B). Such structural “canalization” indicates that the diversification of SLA lineages is not driven primarily by shifts in peptide-binding logic but rather by subtle modifications to the molecular scaffold. These framework-based variations likely serve as evolutionary signatures that modulate ancillary porcine-specific properties, including protein thermal stability, chaperone-mediated assembly (e.g., interaction with the TAP complex), and differential engagement with innate immune receptors.

Fig. 8.

Structural and evolutionary landscape of SLA-I. A Sequence alignment and variability analysis of the extracellular domains of resolved SLA-I structures. Sequence alignments were visualized and processed using Jalview, version 2.11.5.1 [98]. The consensus sequence is highlighted with yellow boxes, while the white boxes denote conserved SLA-I identity. Species-specific residues for HLA-I are shaded in pink. Lineage-specific motifs for SLA-1, -2, and -3 are highlighted in green, blue, and orange, respectively. The nine conserved residues constituting the PBG in SLA-1, -2, and -3 are indicated by asterisks. Secondary structure elements are indicated above the sequence alignment, with arrows representing β-strands and cylinders representing α-helices. The variability at each position is quantified below the sequence using a Wu-Kabat plot [99]. B Structural mapping of species-specific and hypervariable residues. Lineage-specific residues for SLA-1, -2, and -3 are represented in green, blue, and orange, respectively. Hypervariable sites are indicated by red spheres and identified by their Wu-Kabat scores, which are highlighted numerically to illustrate their clustering within the functional pockets. C and D Spatial distribution of HLA-I species-specific sites compared with SLA-I. The MHC-I skeleton is shown as a white ribbon. Species-specific residues are highlighted as pink spheres within the α1, α2, and α3 domains. Structural visualizations and molecular graphics were generated using the Open-source PyMOL Molecular Graphics System, version 3.1.0 (Schrödinger, LLC) [95]

In sharp contrast to the conserved scaffold, the functional plasticity of SLA-I is sequestered within a discrete set of hypervariable residues that define the physicochemical environment of the PBG. Structural mapping revealed a clear hierarchy of critical residues, with position 156 emerging as a pivotal determinant of allelic diversity, followed by positions 67, 97, 151, 152, 70, 77, 95, 114, 155, and 66 (Fig. 8B). Unlike lineage-specific motifs, these allelic polymorphisms are strategically concentrated within the six major pockets (A–F). This spatial segregation reflects a refined evolutionary strategy: the framework remains constrained to ensure molecular integrity and fundamental cellular interactions, while the PBG undergoes rapid diversification under balancing selection. This architectural dichotomy maximizes the breadth of the antigenic library, providing a robust immunological defense mechanism against rapidly evolving pathogens such as African swine fever virus (ASFV) and porcine reproductive and respiratory syndrome virus (PRRSV) [64, 72, 73]. Moreover, it seems that SLA-I molecules attempt to maintain intraspecies balanced selection; for example, it has been reported that the PBG of SLA-I contains six pockets consisting of 33 key amino acid residues, while nine of 33 residues, 5(L), 7(Y), 59(Y), 84(Y), 123(Y), 146(K), 159(Y), 160(L), and 171(Y), are completely conserved in all the SLA-1, SLA-2, and SLA-3 loci identified in different pig breeds or porcine cell lines [39, 40, 57]. This can explain why some conserved cytotoxic T lymphocyte (CTL) epitopes can be bound and presented by different SLA-I molecules across pig breeds [74, 75], indicating the functional conservation of SLA-I molecules, although SLA-I–restricted viral epitopes still dominate. Shared or conserved CTL epitopes are considered the key components in the development of an epitope vaccine [43, 74, 75]. This kind of vaccine is composed of multiple epitopes, including CTL epitopes, B-cell epitopes, and T-helper epitopes, which can provide more comprehensive immunity for animals against different serotypes or strains of swine-origin viruses [43, 76, 77]. Recently, novel bioinformatics tools, such as netMHCpan, PigMatrix, and EpiCC, have been used to screen and design CTL epitopes with high affinity for SLA-I molecules to develop multipe epitope vaccines [78–81]. Moreover, immunopeptidomics based on LC‒MS/MS and peptide elution experiments has been adopted to isolate and identify experimental SLA-I–binding peptides from the SLA-I complex or cells, thus improving the positive rate of designed SLA-I–binding peptides identified by bioinformatics tools and greatly benefiting the development of multiepitope vaccines against swine-origin viruses [82, 83].

This strategic segregation of variability is not only confined to allelic diversification but also manifests at the trans-species level, where amino acid variations between SLA-I and HLA-I exhibit a distinct spatial bias that spares the functional core. The molecular architecture of SLA-I is characterized by a sophisticated balance between functional stasis and structural divergence. At the macroevolutionary level, amino acid variations between SLA-I and HLA-I exhibit a distinct spatial bias. The species-specific substitution sites are predominantly sequestered within the structural scaffold, with the most (23) in the α3 domain of HLA-I, the next (11) in the α2 domain, and the least (6) in the α1 domain (Fig. 8C, D), but they are quite different from the distribution of intraspecies-specific substitution sites in SLA-I molecules, i.e., the most (11) in α2 domain, the second (7) in α1 domain, and the least (3) in α3 domain (Fig. 8A, B). This evolutionary spatial bias has profound implications for xenotransplantation: the α3 domain might determine critical molecular incompatibility between humans and pigs. The high conservation of the PBG architecture, especially in pockets A and F between SLA-I and HLA-I [84], suggests that SLA-I molecules may possess a cross-species capacity to present porcine endogenous or exogenous antigens in a manner “recognizable” by human TCRs, potentially triggering robust cellular rejection [85–87]. Recently, some gene-edited cell lines [88, 89] and gene-edited pig models [90, 91] have been established to treat xenotransplantation-related diseases in humans, with the aim of using pig tissues that present SLA-I-associated epitopes.

Conclusion and prospects

The spatial structure of SLA-I molecules, including SLA-1, SLA-2, and SLA-3, has been extensively elucidated through crystallographic studies, revealing a canonical MHC-I fold. Built upon this conserved scaffold, a conserved integral binding mode is observed, with the peptide termini anchored deeply into pockets A and F, causing the primary anchor residues to be typically P2 and P9, with P3 also serving as a key determinant in many SLA-I molecules, which is different from HLA-I molecules with P2 and P9 as anchor residues. This is also because of the polymorphism of the three SLA-I molecules and the micropolymorphism within the peptide-binding groove (PBG), which confer distinct peptide-binding specificities to each SLA-I allele, as the central region (P4–P6) exhibits considerable conformational plasticity, allowing peptides to adopt extended or bulged conformations that shape the TCR interface, thus producing strict SLA-I–restricted epitope presentation. Similarly, the A-pocket’s “gate” mechanism and the F-pocket’s hydrophobic architecture further refine the peptide repertoire. Comparative structural analyses of SLA-I molecules by intraspecies and trans-species revealed that the conserved structure of PBG allelic variants is in favor of shared or conserved viral epitopes produced in the peptide presentation of SLA-I molecules, which have the potential to be used to develop multi-epitope vaccines against different serotypes or strains of swine-origin viruses and to cause cross-reaction to human MHC-I and TCR in xenotransplantation, indicating the functional conservation of SLA-I molecules to some extent.

The existing structural repertoire, while comprehensive in its coverage of the three classical loci, highlights the need for further investigation of peptide conformational dynamics to fully elucidate the molecular rules of SLA-I–restricted immunity. Nevertheless, research on SLA-I–restricted CTL epitopes remains underdeveloped relative to that on these model systems, largely because of the lack of robust cellular platforms, such as the well-established T2 cell lines used for rapid MHC-I epitope screening in humans and mice [92, 93]. Therefore, a swine CTL epitope screening platform at the cellular level should also be developed so that more SLA-I–presented epitopes can be used to study the comprehensive spatial structure of SLA-I molecules, thereby increasing our understanding of the structure and function of SLA-I molecules.

Abbreviations

β₂m

β2 microglobulin

CTL

Cytotoxic T lymphocyte

ER

Endoplasmic reticulum

ERAAP

ER aminopeptidase

FMDV

Foot and mouth disease virus

HLA

Human leukocyte antigen

MHC

Major histocompatibility complex

MOG

Myelin oligodendrocyte glycoprotein

PBG

Peptide-binding groove

RING1

Ring finger protein 1

SLA

Swine leukocyte antigen

TCR

T-cell receptor

TAP

Transporters associated with antigen presentation transporter

Author contributions

Conceptualization: Y.Y.G. and F.S.G.; Writing—original draft preparation: Y.Y.G. and F.S.G. Writing—review and editing: Y.Y.G., H.Y.L., S.Y.Y., C.J.S., Y.D.C., and Y.T.; supervision: G.X.H., Z.B.L. and F.S.G. All authors have read and agreed to the published version of the manuscript.

Funding

This research was supported by the National Natural Science Foundation of China (Grant number 32273022), the Innovative Development Project in the Department of Education of Liaoning Province (LJ242511258002), and the Dalian University Discipline Construction Special Project (grant number DLUXK-2025-FX-003).

Data availability

No datasets were generated or analyzed during the current study.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.