HLA Loci

Key Takeaways

Genetic Organization:

MHC Region on Chromosome 6:
- Contains Class I (HLA-A, -B, -C) and Class II (HLA-DR, -DQ, -DP) genes.
- Class III region encodes other immune proteins (e.g., complement, TNF).
Haplotype:
- A set of HLA alleles inherited together; shows strong linkage disequilibrium.
DRB3/4/5 Specifics:
- Secondary DR β-chain genes present only with certain DRB1 alleles.
- Only one (DRB3, DRB4, or DRB5) is present per haplotype, expanding peptide presentation.

Molecular Structures:

Class I Molecules:
- Heterodimer: Polymorphic heavy chain + invariant β₂-microglobulin.
- Closed peptide-binding groove accommodates 8–10 amino acid peptides.
- Expressed on all nucleated cells; presents endogenous peptides to CD8⁺ T cells.
Class II Molecules:
- Heterodimer: Two polymorphic chains (α and β).
- Open-ended groove allows binding of longer peptides (13–25+ amino acids).
- Expressed on antigen-presenting cells; presents exogenous peptides to CD4⁺ T cells.

HLA Antigen Expression & Epitopes:

Antigens & Serologic Specificity:
- Historically defined as broad antigens (e.g., A10) later split into finer specificities.
Epitopes & Eplets:
- Epitope: Specific binding site for antibodies/TCRs on the HLA molecule.
- Eplet: Minimal amino acid configuration constituting the core of an epitope.
Private vs. Public Antigens:
- Private: Unique to a single allele.
- Public: Shared among several alleles (e.g., Bw4/Bw6), leading to cross-reactivity.

HLA Allele Nomenclature:

Four-Field System:
- Field 1: Allele group (serologic type).
- Field 2: Protein variant (nonsynonymous changes).
- Field 3: Synonymous substitutions.
- Field 4: Non-coding region variations.
Suffixes:
- Indicate null (N), low expression (L), secreted (S), etc.
Clinical Relevance:
- Matching is typically performed at the two-field (four-digit) level; epitope matching refines compatibility.

1. Genetic Organization of HLA Loci

1.1 HLA Genes on Chromosome 6 and Haplotype Inheritance

The human leukocyte antigen (HLA) genes are clustered on a 3–4 Mb region of chromosome 6p21 (the human MHC) () (). This region is divided into three sub-regions: Class I, Class II, and Class III. The Class I region (telomeric end) contains the classic class I genes HLA-A, HLA-B, HLA-C, which encode the heavy chains of HLA class I molecules (). The Class II region (centromeric end) contains the genes for classic class II molecules: HLA-DR, HLA-DQ, HLA-DP (each consisting of an α and a β chain gene) (). The Class III region, located between class I and II, encodes other immune proteins (complement components C2, C4, factor B; cytokines like TNF) but no HLA molecules () (). The table below summarizes the major transplant-relevant HLA genes:

Class I Genes (chrom. 6)	Class II Genes (chrom. 6)
HLA-A, HLA-B, HLA-C (classical) HLA-E, HLA-F, HLA-G (non-classical)	HLA-DRA (DR α chain) HLA-DRB1 (primary DR β chain) HLA-DRB3, DRB4, DRB5 (secondary DR β chains, if present) HLA-DQA1, DQB1 (DQ α and β) HLA-DPA1, DPB1 (DP α and β)

Haplotype: An HLA haplotype is the set of HLA alleles across these loci inherited together on one chromosome. Because the HLA genes are closely linked, they are usually passed on as an intact block from each parent (). Thus, each individual inherits one haplotype from the mother and one from the father (HLA expression is co-dominant). In a family, the segregation of these haplotypes follows Mendelian rules: two siblings have a 25% chance of being HLA-identical (sharing both haplotypes), 50% chance of sharing one haplotype, and 25% chance of sharing none (). In other words, full siblings have a 1 in 4 chance of being perfect HLA matches. HLA haplotypes are often denoted by the allele or antigen names at each locus (e.g. A1-B8-DR3-DQ2 refers to HLA-A01:01, B08:01, DRB103:01, DQB102:01, etc.) (). A haplotype may span a large segment of the MHC; for example, the “8.1 ancestral haplotype” (A1-B8-DR3-DQ2) is ~4.7 million bases long and is found in ~15% of individuals of Northern European descent ().

Linkage Disequilibrium: Certain HLA allele combinations occur together on haplotypes more frequently than expected by chance, a phenomenon known as linkage disequilibrium (LD) () (). For instance, if HLA-A02 and HLA-B07 appear together far more often than the product of their individual gene frequencies, they are in LD (). This is commonly observed in HLA: populations show “popular” haplotypes. For example, HLA-A1, B8, DR3 (DR17) is one of the most common Caucasian haplotypes (~5% frequency), far higher than random association would predict () (). LD arises from evolutionary selection and limited recombination within the MHC. It means that knowing one allele can make the presence of another allele on the same chromosome more likely. Clinically, LD implies that patients often inherit fixed combinations of HLA alleles (haplotypes), which is important in finding matched donors (especially among siblings or in populations with well-characterized common haplotypes) ().

1.2 Unique Organization of HLA-DRB3/4/5 (DR Subregion)

The HLA-DR region has a unique genomic organization compared to other class II loci. HLA-DR is composed of one essentially invariant alpha-chain gene (DRA1) and multiple beta-chain genes (DRB genes). Every human haplotype contains DRB1, the primary β-chain locus encoding the polymorphic DRβ1 chain that defines the HLA-DR antigen (e.g. DR1, DR15, DRB1*03:01 etc.) (). In addition, some haplotypes carry a secondary DRB gene – one of DRB3, DRB4, or DRB5 – while other haplotypes carry none (). Importantly, only one of these secondary loci can be present on a given haplotype (). Thus:

DRB3 (which encodes the serologic antigen DR52) is present on haplotypes that have certain DRB1 allele lineages – notably DRB1 *03, *11, *12, *13, *14 families ().
DRB4 (encodes DR53) is present on haplotypes with DRB1 *04, *07, *09 allele families ().
DRB5 (encodes DR51) is present on haplotypes with DRB1 *15 or *16 (the DR2 serologic family) ().
Haplotypes with DRB1 alleles outside these groups (e.g. DRB1 *01, *08, *10) usually have no secondary DRB gene (only DRB1).

Because of this, an individual can express at most two DRB1 gene products and two DRB3/4/5 gene products (if each haplotype carries one) – up to four distinct DR β-chains in total (). All these pair with the same DRα chain (DRA) on the cell surface. These secondary loci are inherited in tight linkage with DRB1, forming three serologically defined “DR haplotypes”: DR51, DR52, DR53, corresponding to the presence of DRB5, DRB3, DRB4 respectively () (). For example, a person who is “DR15 and DR4” by serology will have a DRB115 allele (which comes with DRB5 = DR51) on one haplotype and a DRB104 allele (with DRB4 = DR53) on the other; this person would express DRβ115, DRβ5, DRβ104, and DRβ4 chains. In contrast, someone who is “DR1 and DR7” (DRB101 and DRB107) would express DRβ101 and DRβ107 plus DRβ4 (since DR7 is associated with DRB4 = DR53) – three DR β-chains total, because DRB1*01 haplotype has no secondary DRB gene. This complex organization is unique to the DR locus – other class II loci (DQ and DP) have a single α and β gene per haplotype that pair to form one molecule each.

Key point: Unlike HLA-A, B, C or HLA-DQ, DP (which have one gene per locus per haplotype), HLA-DR can have two functional β-chain genes on one haplotype (DRB1 plus DRB3/4/5). Which secondary gene (if any) is present is entirely determined by the specific DRB1 allele lineage on that haplotype (). The secondary DRB genes are less polymorphic and expressed at lower levels than DRB1 (), but they broaden the range of peptides presented (since DRB3/4/5 encode additional DR molecules that pair with the same DRα) () ().

2. Molecular Structures of Class I and Class II HLA

2.1 Structure of HLA Class I Molecules

Class I HLA molecules (e.g. HLA-A, B, C) are heterodimers composed of: (1) a polymorphic heavy (α) chain (~44–47 kDa) encoded by a class I HLA gene, and (2) an invariant light chain called β₂-microglobulin (~12 kDa) (). The heavy chain is a single polypeptide with three extracellular domains (α1, α2, α3), a transmembrane segment, and a short cytoplasmic tail () (). The light chain β₂-microglobulin (β₂m) is not encoded in the MHC (it’s on chromosome 15) and non-covalently associates with the α3 domain to stabilize the class I complex ().

The α1 and α2 domains of the class I heavy chain fold together to form the peptide-binding groove (also called the antigen-binding cleft) (). Structurally, this groove is like a platform of eight β-strands with two α-helices forming the sides (). It creates a closed-ended cleft that accommodates a short peptide 8–10 amino acids in length (usually nonamers) () (). The peptide’s amino- and carboxy-termini are anchored deep in pockets at each end of the groove, which is why class I can only bind peptides of limited length (the ends of longer peptides cannot extend past the closed groove) (). Specific amino acids in the peptide (called anchor residues) must match the pockets of a given HLA allele’s groove for stable binding – this gives each HLA allele a characteristic peptide-binding motif.

The α3 domain of the heavy chain, along with β₂-microglobulin, forms an immunoglobulin-like constant region that does not vary much among alleles (). The α3 domain contains the binding site for the CD8 co-receptor on cytotoxic T cells (). In fact, CD8 binds to a conserved region on the α3 domain of class I molecules (away from the peptide groove), stabilizing the interaction between a CD8⁺ T cell and an HLA-I presenting its cognate peptide ().

Summary of Class I structure: an HLA class I molecule is a membrane-bound heavy chain (α1–α3 domains) noncovalently associated with β₂-microglobulin (). The α1/α2 domains form the peptide-binding platform (with polymorphic residues lining the groove), while α3 is a relatively conserved domain that interacts with CD8 and supports the structure () (). A diagrammatic view would show the heavy chain’s α1 and α2 domains on top (with a peptide lying between them), the α3 domain and β₂m beneath them, and a transmembrane segment anchoring the heavy chain to the cell surface. (β₂m is not membrane-anchored; it nestles alongside α3.) The class I groove typically presents endogenously derived peptides (from cytosolic proteins) to CD8⁺ T cells, as described in Section 3.

2.2 Structure of HLA Class II Molecules

Class II HLA molecules (e.g. HLA-DR, DQ, DP) are also heterodimers, but of a different design: they consist of two polymorphic glycoprotein chains, an α chain (~32–34 kDa) and a β chain (~29–32 kDa), both encoded by HLA class II genes (). Each chain has two extracellular domains (α1/α2 on the alpha; β1/β2 on the beta), a single transmembrane region, and a cytoplasmic tail (). The α and β chains associate noncovalently. Notably, both chains are required for stable expression – there is no β₂-microglobulin equivalent in class II.

The peptide-binding groove of class II is formed at the interface of the α1 and β1 domains () (). Like class I, this groove is a cleft lined by α-helices and β-sheet, but an important difference is that the class II groove is open at both ends (). This means peptides bound to class II can extend beyond the length of the groove. As a result, class II molecules accommodate longer peptides (generally 13–25 amino acids or even longer), which often overhang beyond the cleft’s ends () (). The peptide is held by interactions along the groove (especially by a core ~9-mer segment in the middle), but it doesn’t have strict terminal anchor constraints as in class I. This open-ended groove architecture explains why class II presents longer, processed peptides (typically from extracellular proteins that have been endocytosed and degraded).

The membrane-proximal α2 and β2 domains of class II are structurally similar to Ig constant domains and are relatively conserved. The β2 domain contains the binding site for the CD4 co-receptor on helper T cells (). Thus, CD4 on CD4⁺ T cells binds to the nonpolymorphic β2 region of HLA class II, stabilizing the T cell – antigen-presenting cell interaction (). Like class I, the class II molecule’s top surface (the α1–β1 platform with bound peptide) is what the T-cell receptor (TCR) recognizes.

Both class I and II HLA have a similar overall fold – a peptide-binding groove perched atop Ig-like support domains () () – despite being composed of different subunits. The polymorphism (allelic variation) in both classes is concentrated in the peptide-binding domains (α1/α2 for class I, β1 (and α1 to lesser extent) for class II), altering which peptides can be bound (). Table 1 highlights key structural differences and similarities between Class I and Class II HLA:

Feature	HLA Class I	HLA Class II
Polypeptide composition	One heavy α chain (3 domains) + one β₂-microglobulin light chain ([MHC & Antigen Presentation \| Immunopaedia](https://www.immunopaedia.org.za/immunology/basics/4-mhc-antigen-presentation/#:~:text=Feature%20MHC,binds%20to%20the%20%CE%B22%20region))	One α chain (2 domains) + one β chain (2 domains) ([MHC & Antigen Presentation \| Immunopaedia](https://www.immunopaedia.org.za/immunology/basics/4-mhc-antigen-presentation/#:~:text=Feature%20MHC,binds%20to%20the%20%CE%B22%20region))
Peptide-binding groove	Formed by α1–α2 domains of heavy chain ([MHC & Antigen Presentation \| Immunopaedia](https://www.immunopaedia.org.za/immunology/basics/4-mhc-antigen-presentation/#:~:text=Distribution%20All%20nucleated%20cells%20Antigen,of%2013%E2%80%9325%20residues%20or%20more)); closed ends accommodate peptides ~8–11 aa ([MHC & Antigen Presentation \| Immunopaedia](https://www.immunopaedia.org.za/immunology/basics/4-mhc-antigen-presentation/#:~:text=Binding%20site%20for%20T%20cell,of%2013%E2%80%9325%20residues%20or%20more))	Formed by α1–β1 domains ([MHC & Antigen Presentation \| Immunopaedia](https://www.immunopaedia.org.za/immunology/basics/4-mhc-antigen-presentation/#:~:text=Distribution%20All%20nucleated%20cells%20Antigen,of%2013%E2%80%9325%20residues%20or%20more)); open ends accommodate peptides ≥13 aa (typically 13–25 aa) ([MHC & Antigen Presentation \| Immunopaedia](https://www.immunopaedia.org.za/immunology/basics/4-mhc-antigen-presentation/#:~:text=Binding%20site%20for%20T%20cell,of%2013%E2%80%9325%20residues%20or%20more)) ([ The HLA System: Genetics, Immunology, Clinical Testing, and Clinical Implications - PMC ](https://pmc.ncbi.nlm.nih.gov/articles/PMC2628004/#:~:text=extracellular%20portion%20composed%20of%20two,9%20%2C%2031))
Co-receptor binding	CD8 binds α3 domain (heavy chain) ([MHC & Antigen Presentation \| Immunopaedia](https://www.immunopaedia.org.za/immunology/basics/4-mhc-antigen-presentation/#:~:text=Distribution%20All%20nucleated%20cells%20Antigen,binds%20to%20the%20%CE%B22%20region))	CD4 binds β2 domain ([MHC & Antigen Presentation \| Immunopaedia](https://www.immunopaedia.org.za/immunology/basics/4-mhc-antigen-presentation/#:~:text=Distribution%20All%20nucleated%20cells%20Antigen,binds%20to%20the%20%CE%B22%20region))
Distribution	All nucleated cells (ubiquitous expression) ([MHC & Antigen Presentation \| Immunopaedia](https://www.immunopaedia.org.za/immunology/basics/4-mhc-antigen-presentation/#:~:text=Distribution%20All%20nucleated%20cells%20Antigen,binds%20to%20the%20%CE%B22%20region))	Antigen-presenting cells (B cells, dendritic cells, macrophages) and some activated T cells ([ The HLA System: Genetics, Immunology, Clinical Testing, and Clinical Implications - PMC ](https://pmc.ncbi.nlm.nih.gov/articles/PMC2628004/#:~:text=Expression%20of%20HLA))
Function	Display intracellular (endogenous) peptides to CD8⁺ T cells (cytotoxic T lymphocytes)	Display extracellular (exogenous) peptides to CD4⁺ T cells (helper T lymphocytes)

TCR interaction: In both class I and II, the T-cell receptor docks on the peptide–MHC complex, contacting residues of the peptide and the α-helices of the MHC molecule (this requirement for specific MHC interaction is known as MHC restriction ()). The TCR’s recognition of a foreign peptide bound in an HLA molecule is the central event that triggers T-cell activation. Thus, the structural features of the binding groove (which peptides are presented) and the MHC surface (which the TCR sees) directly influence immune recognition. Polymorphic differences between HLA alleles translate to different peptide-binding preferences and different “looks” to T cells, which is why HLA mismatches can provoke strong immune responses in transplantation.

(For clarity, one can envision class I as a tri-domain α chain sitting on β₂m, with a small peptide buried in its top groove; class II as two chains (α and β) each contributing half the groove, holding a longer peptide that can hang out of the ends. Both present peptide antigens to T cells but engage different T cell subsets due to the CD8/CD4 interactions.)

3. HLA Antigen Expression and Definitions

3.1 Cell Surface Expression of HLA Class I vs II

HLA Class I molecules (HLA-A, B, C) are expressed on the surface of almost all nucleated cells in the body (). This broad expression allows nearly any infected or abnormal cell to present internal peptides to CD8 T cells for immune surveillance. Class I expression can be upregulated by interferons during immune responses. In contrast, HLA Class II molecules (HLA-DR, DQ, DP) are constitutively expressed mainly on professional antigen-presenting cells (APCs) – such as B lymphocytes, dendritic cells, macrophages – and on thymic epithelial cells (). Resting T cells and other somatic cells generally do not express class II, but class II can be induced on some cells (e.g. endothelial cells, activated T cells) by cytokines like IFN-γ (). This restricted pattern ensures class II presents peptides primarily in contexts where initiating a helper T cell response is appropriate (e.g. in an APC that has engulfed a pathogen). Both class I and II genes are co-dominantly expressed, meaning all alleles inherited (2 per locus) are expressed on the cell. For example, a person heterozygous at HLA-A will have both variants of HLA-A protein on all cells. Importantly, HLA molecules are highly polymorphic, so an individual expresses a specific set of HLA “antigens” that the immune system recognizes as self.

3.2 “HLA Antigens”: Broad vs Split Specificities

In transplantation immunology (and immunohematology), the term “HLA antigen” refers to the serologically defined specificity of an HLA allele product. Historically, HLA typing was first done with panels of antisera, which identified groups of HLA proteins on white blood cells. Early sera defined broad antigens – each broad specificity actually encompassed multiple finer specificities that were later distinguished as separate “split” antigens. For example:

HLA-A10 was a broad antigen that was later resolved into two splits: HLA-A25 and HLA-A26 (). Initially, antibodies grouped A25 and A26 together as “A10”; with better reagents they were split apart. Similarly, HLA-B16 was split into B38 and B39 ().
In class II, broad serotypes like HLA-DR5 were split into DR11 and DR12, and broad DQ1 into DQ5 and DQ6 ().

Broad antigens (sometimes called “parent” antigens) thus refer to the older, higher-level groupings, whereas antigen (split) refers to the more specific serologic specificity. In modern practice, we usually refer to the split antigens by their numbers (e.g. “HLA-A26” instead of A10). Each HLA “antigen” corresponds to one or more alleles at the DNA level. For instance, the antigen HLA-A26 can be produced by several HLA-A*26:XX alleles. Conversely, alleles that encode very similar proteins often share the same serologic antigen designation. The first field of the allele name typically indicates this (e.g. any allele starting with *26 belongs to the A26 antigen group).

From an immunological perspective, an HLA antigen is essentially a set of surface epitopes on the HLA molecule that are recognized collectively by specific antibodies or T cells. Antigenic determinants on HLA can be categorized by how unique or shared they are:

A private antigen (or private epitope) is an HLA determinant found on a single or very few HLA alleles. It is unique to a particular HLA protein. For example, the serologic antigen HLA-A1 has a unique epitope defined by the amino acid motif 163RG on its α2 domain that is not present on other HLA-A alleles (). Anti-HLA-A1 antibodies are essentially specific for this private epitope (making HLA-A1 a private specificity in that context) (). Another example: HLA-B7 has an epitope 177DK unique to B7 that specific antisera recognize ().
A public antigen (or public epitope) is a determinant shared by multiple HLA proteins, even across different loci or antigen groups. Classic examples are HLA-Bw4 and Bw6, which are public epitopes on HLA-B molecules. Bw4 and Bw6 are present on mutually exclusive subsets of HLA-B alleles (and some HLA-A) – e.g. HLA-B27, B57, B58 share the Bw4 epitope, while HLA-B7, B8, B35 share the Bw6 epitope (). An antibody against Bw6 will react with any HLA-B that carries Bw6, making Bw6 a public antigen. Public epitopes underlie what are known as cross-reactive groups (see below). In the older literature, “public antigen” referred to these shared determinants (often denoted by a “w” in serology, like Bw4/Bw6, or cross-reactive A2/A28, etc.).

Epitopes and Eplets: An epitope is the specific molecular surface on an antigen that an antibody or T-cell receptor binds. HLA molecules, being large (~180 amino acid extracellular α1+α2 domains for class I), have multiple epitopes on their surface. Some epitopes are immunodominant (the main target of alloantibodies for a given antigen), others are subdominant. It is now understood that HLA antibodies recognize discrete epitopes rather than entire antigen wholes (). For instance, when we say “anti-HLA-A2 antibody,” in reality the antibody is binding a particular epitope on A2 that might also exist on other alleles. This epitope-centric view has led to the concept of eplets. An eplet is a defined minimal amino acid configuration on the HLA surface that corresponds to an antibody binding site (a kind of structural “unit” of an epitope) (). Eplets are typically small clusters of polymorphic residues (often within a 3Å radius) on the HLA α-helix or β-sheet that form the contact points for an antibody paratope (). For example, the epitope recognized by anti-A1 serum can be delineated to a specific combination of residues; in the HLA Epitope Registry, this is registered as eplet 163RG (arg163+gly?) on HLA-A1 (). Likewise, anti-B7 corresponds to eplet 177DK unique to B7 (). Many serologically defined HLA antigens have been analyzed and found to correspond to unique eplets (). In other cases, a broad antigen’s recognition can be explained by an epitope that is common to its splits. For instance, A10 (broad) was defined by antibodies recognizing the epitope 149TAH present on both A25 and A26; the difference between A25 vs A26 is an additional residue difference (80I vs 80N) that forms part of distinct epitopes, explaining why some antibodies differentiate them ().

In summary of terminology:

Broad antigen: a high-level HLA specificity grouping multiple splits (e.g. B5 broad includes B51 and B52; DR2 broad includes DR15 and DR16). These are mostly of historical importance, though broad mismatches vs split mismatches can be clinically noted.
Antigen (split): a specific HLA serotype (e.g. A2, A24, B7, DR15) corresponding to a group of alleles with similar serologic reactions. This is the level at which transplant matching is often discussed (“match at HLA-A, -B, -DR antigens”).
Epitope: the precise section of the HLA molecule that an antibody binds – can be a single polymorphic residue or a patch of a few residues. An antigen contains multiple epitopes.
Eplet: a defined string or configuration of polymorphic residues that constitutes the core of an epitope recognized by HLA antibodies (). Eplets are used in computational algorithms (like HLAMatchmaker) to assess donor-recipient compatibility at the epitope level rather than just whole-antigen level ().

3.3 Role in Immune Recognition and Transplant Rejection

These definitions matter because the immune system “sees” HLA mismatches in terms of epitopes. A transplant recipient’s alloantibodies or T cells might not recognize an entire donor HLA protein if parts of it are similar to self, but they can latch onto a few critical polymorphic differences:

If a donor’s HLA has a private epitope the recipient’s immune system has never seen, it may trigger a strong antibody response (since that epitope is completely foreign). This can lead to donor-specific HLA antibodies (DSA) and transplant rejection.
If a donor HLA shares public epitopes with an HLA that the recipient has been exposed to (e.g. via pregnancy, transfusion, or prior transplant), then the recipient might already have cross-reactive antibodies that recognize the donor’s HLA. For example, a patient sensitized to HLA-A2 might harbor antibodies that also react with HLA-A28, because A2 and A28 share a common public epitope (). In transplant terms, A2 and A28 belong to the same “cross-reactive group.”

Clinically, HLA matching and antibody screenings are increasingly considering epitope and eplet mismatches. Research shows that the number of mismatched eplets may predict rejection risk better than the simple count of antigen mismatches (). The immune system’s fine specificity means that even if a donor and recipient differ at an HLA antigen, they might be tolerated if they share enough key epitopes (the concept behind permissive mismatches or epitope-based matching programs). Conversely, even one antigen match could be problematic if a single dangerous epitope is introduced.

4. Private vs. Public Antigens in Transplant Immunology

As introduced above, private and public HLA antigens refer to whether an immune epitope is unique or shared among HLA molecules:

Private antigens/epitopes are unique to a single HLA allele or a very limited set of alleles. They can be thought of as “allele-specific” determinants. An example is the epitope on HLA-A*01:01 (serologic A1) that is not found on any other common allele (). If a recipient lacks A1 but the donor has it, the A1’s private epitope is completely foreign to the recipient’s immune system, likely inducing a strong specific antibody if exposed. In the context of antibody screening, an antibody that reacts only with HLA-A1 and no other antigens is targeting a private epitope of A1.
Public antigens/epitopes are those shared by multiple HLA alleles (often across what we traditionally consider different antigens). These are sometimes called cross-reactive determinants. For example, HLA-Bw4 and Bw6 are public epitopes on HLA-B molecules recognized by certain antibodies (). The Bw4 epitope is present on many HLA-B alleles including B27, B44, B51, etc., whereas Bw6 is present on B7, B8, B35, B41, and others. An antibody against a public epitope like Bw6 will cause a positive reaction with any B antigen that has Bw6 (). Similarly, HLA-A2 and HLA-A28 share a public epitope (they belong to the same cross-reactive group A2/A28) () – historically denoted by a “w” as Aw68/Aw69 subtypes etc. Antisera often showed that certain sets of seemingly different HLA antigens would always cross-react, leading to the identification of these public determinants ().

In transplant practice, knowing which donor HLA are public vs private to the recipient’s sensitization history is crucial. Public epitopes can cause unexpected cross-reactions: for instance, a patient with antibody to “HLA-B62” might also react to B15:01 (B62) and B15:10 (B75) and others that share a public epitope, even though those are different antigens. These are grouped as cross-reactive antigen groups (CREGs) (). On the other hand, if a donor has an HLA allele that differs from the recipient’s by only private epitopes the recipient has never seen, it may still induce a de novo response post-transplant.

In summary:

Private HLA antigens – unique specificities – are often what we try to match for (to avoid entirely novel epitopes to the recipient). A perfectly HLA-matched transplant avoids all private differences.
Public HLA antigens – shared epitopes – explain why even some HLA “mismatches” might be partly tolerated if they share epitopes with the recipient’s HLA, and conversely why a sensitized patient’s antibodies can cover more antigens than initially apparent. In marrow transplantation, there is evidence that matching at the level of DRB3/4/5 (which can be viewed as avoiding certain public epitope mismatches) improves outcomes ().

When screening a highly sensitized patient’s serum, labs often identify antibody specificities in terms of public vs private. For example, they may determine a patient has antibody to a public epitope present on A1, A3, A11 (a public “P” group often denoted as “P01” which includes those antigens ()). This means the patient would likely react to any donor with A1, A3 or A11. Distinguishing whether an antibody is directed to a public epitope or a private one guides what donor antigens are acceptable.

5. Shared Epitopes and Cross-Reactivity

Shared epitopes are the biochemical basis of cross-reactivity in HLA responses. If two different HLA alleles share a significant epitope, an antibody (or T-cell) specific for that epitope will react with both. This is a major consideration in transplantation: a patient might not have antibodies against a donor’s exact HLA allele, but they might have an antibody to a shared epitope that the donor’s HLA carries.

Serologically, HLA experts defined clusters of cross-reactive antigens known as Cross-Reactive Groups (CREGs). These groups reflect shared public epitopes. For example:

The CREG containing A2, A28, A68, A69 exists because A2 and A28 (and the latter’s subtypes A68, A69) share common epitopes, so antibodies often see them interchangeably (). In fact, a “public marker” P02 was assigned to denote the shared epitope on A2/A28 group ().
Another CREG: A1, A3, A11, A36 frequently cross-react (public marker P01) (). These A locus antigens share one or more epitopes, causing antibodies to one to hit the others.
At the B locus, the broad Bw4 vs Bw6 grouping is an example: all Bw4-positive B alleles cross-react with anti-Bw4 antibodies, and similarly for Bw6 (). There are also finer groupings; for instance, the B7 CREG includes B7, B22, B27, B40, B41 that share certain epitopes, while the B8 CREG includes B8, B14, etc., depending on the classification.

Cross-reactivity implications: When selecting donors for sensitized patients, transplant programs pay attention to these shared epitopes. A patient with an anti-A2 antibody should likely avoid donors with A2 or any antigen in the A2 CREG (like A28) because of cross-reaction. Modern virtual crossmatching tools use epitope data to predict cross-reactivity: e.g. if a patient has antibody to the eplet 62GE on HLA class I, any donor allele with that eplet would be considered incompatible, even if the specific allele was not originally tested.

On the T-cell side, shared epitopes (or structurally similar peptide-MHC surfaces) can also cause cross-reactive T cell responses, though this is less predictable than antibody cross-reactivity.

From a compatibility standpoint, understanding shared epitopes allows:

Epitope-based matching: We try to minimize the number of immunogenic eplet mismatches. For example, two HLA alleles might be different, but if they share most surface eplets, the mismatch might be less immunogenic (they are “epitope cross-reactive” with recipient’s own HLA). This is being explored to improve transplant outcomes.
Acceptable mismatches: In highly sensitized patients, sometimes certain donor HLA that would be considered a mismatch at the antigen level are acceptable if the patient has no antibody to the specific epitopes of that antigen. For instance, a patient might have broad sensitization to many B antigens but careful analysis might find a donor whose B allele, while different, does not carry any of the eplets the patient’s antibodies target – thus an “acceptable” mismatch.

In summary, shared HLA epitopes lead to cross-reactive immune responses, which can be a double-edged sword: it means one antibody can protect against or injure multiple targets (e.g. an anti-Bw4 antibody would attack any Bw4-positive graft), but it also means that if donors are chosen wisely (avoiding the epitopes a patient is sensitized to), even some mismatches can be tolerated. The concept of public vs private epitopes (Section 4) ties in: shared (public) epitopes cause cross-reactivity, unique (private) epitopes do not. Cross-reactivity is essentially the immune system “seeing” a different HLA as if it were the one it recognizes, due to structural mimicry.

Transplant immunologists use tools like the HLA Epitope Registry and HLAMatchmaker to catalog which alleles share epitopes/eplets. This information is increasingly used to refine organ allocation (for example, the Eurotransplant program’s Acceptable Mismatch program uses epitope matching to find donors for sensitized patients). Ultimately, minimizing critical epitope mismatches can reduce alloimmune injury and improve graft survival.

6. HLA Alleles and Nomenclature

HLA genes are extremely polymorphic, and each distinct sequence variant is assigned a unique allele name. The HLA nomenclature follows a systematic format of up to four fields (parts) separated by colons (), plus optional suffix letters. The format can be illustrated as HLA-[locus]*[field1]:[field2]:[field3]:[field4][suffix]. Each field conveys specific information about differences at the DNA/protein level ():

Field 1 (Allele Group): The first set of digits indicates the broad allele family or type, often correlating with the serologic antigen specificity (). For example, A02:101 belongs to the “A2” group (because 02 is field1) which historically is serologically identified as HLA-A2. Alleles sharing field1 usually encode HLA proteins that react with the same serologic antibody (i.e. they have the same broad antigen specificity).
Field 2 (Protein Variant): The second set of digits specifies the particular allele within that group, and differences here reflect nonsynonymous nucleotide changes that alter the amino acid sequence of the protein (). In other words, if two alleles differ in field2, their HLA proteins have at least one amino acid difference in the antigen-binding region. For instance, A02:01 vs A02:02 differ in their coding sequence (and thus the protein). Field1 and 2 together (e.g. A*02:01) are often called the “two-field” or “four-digit” allele name, and they define the HLA protein specificity. Most clinical high-resolution matching considers differences at this level.
Field 3 (Synonymous variation): The third field, if present, indicates nucleotide differences that do not change the amino acid sequence (synonymous substitutions) within the coding region (). So alleles that differ only in DNA but encode the same protein get an additional number in field3. For example, B07:02:01 vs B07:02:02 might have a silent mutation in an exon. Field3 helps uniquely identify the allele at the DNA level.
Field 4 (Non-coding variation): The optional fourth field denotes differences outside the coding sequence, such as in introns or untranslated regions (). These differences have no effect on the amino acid sequence but distinguish alleles at the genomic level. So an allele with a variation in an intron compared to an otherwise identical allele would get a different fourth field number.

All alleles have at least a two-field name (e.g. A02:01). A third or fourth field is only assigned when needed to distinguish additional sequence differences (). For example, **HLA-DRB113:01:01:02** represents: DRB1 locus, allele group 13 (DR13 serotype), specific protein variant 01 (the first DR13 allele found), same protein as DRB113:01:01:01 but differs in a non-coding region (hence field4 is 02) (). If an allele has no known synonymous or non-coding differences, it may be reported with just two fields or three fields (e.g. DRB113:01 or DRB1*13:01:01).

Additionally, suffix letters can be appended to indicate special allele properties () ():

N = “Null” allele (does not produce a functional protein) (). For instance, A*24:02N means this allele has a mutation causing no surface expression. Such alleles are important in that an individual might type for an allele but actually lack the antigen on cells.
L = “Low” expression allele (significantly reduced cell surface expression) ().
S = “Secreted” (the protein is secreted or not membrane-bound, e.g. some class I chain variants) ().
C = “Cytoplasmic” (not expressed on cell surface) ().
Q = “Questionable” (expression or function is uncertain) ().
A = “Aberrant” expression (aberrant or abnormal expression, rare) ().

For example, HLA-A*24:09N would indicate the allele in the A24 group, variant 09, that is a null allele (N) (). Another example: some DRB4 alleles have suffix "N" because DRB4*01:03:102N is a null that produces no DR53 chain ().

Four-field resolution refers to typing that identifies all four components (and thus any difference, even silent ones) of the allele name. This is the highest resolution of HLA typing, often needed for research or precise registry records. Two-field (four-digit) resolution (e.g. HLA-B08:01) is commonly used for clinical matching, as it identifies the protein sequence of the antigen. One-field resolution (e.g. HLA-B08, or just the serologic type “B8”) is low resolution and might group multiple alleles together. In contexts like stem cell transplantation, matching is typically at least at the two-field level for key loci (HLA-A, B, C, DRB1, DQB1). In solid organ transplant, antigen-level (approximately one-field) matching is often considered sufficient for allocating organs, but allele-level matching can still improve outcomes.

To illustrate nomenclature, consider this breakdown from the WHO HLA Nomenclature guidelines ():

HLA-DRB1*13:01 – this denotes the allele group DR13 and the specific protein variant 01. This allele encodes the DR13 antigen (split from broad DR6).
HLA-DRB1*13:01:02 – the same DRB1*13:01 protein, but “02” in the third field indicates a synonymous DNA mutation compared to *13:01:01 ().
HLA-DRB1*13:01:01:02 – adds a fourth field “02”, indicating a non-coding difference from *13:01:01:01 ().
So DRB1*13:01:01:01 would be a fully specified allele, and 13:01:01:02 differs only in non-coding sequence (). A laboratory that reports at four-field resolution would distinguish those, whereas at two-field both would simply be “DRB113:01” (no functional difference).

The explosion of HLA allele discovery (thanks to DNA sequencing) has led to thousands of alleles per locus. For example, there are over 4000 HLA-B alleles known. The nomenclature is managed by the WHO HLA Nomenclature Committee, and updates are published regularly. Understanding the nomenclature is important for interpreting high-resolution typing results in transplant lists or donor registries. For practical purposes, a transplant match is often defined by matching at the allele group level (first two fields) for the classical loci, since differences in the third/fourth field usually don’t change the protein and thus are not immunogenic. However, for certain cases (like identifying a rare null or low-expression allele), the higher resolution info and suffix become very relevant.

Key Takeaway: The four-field HLA allele name precisely identifies the HLA variant at the DNA level () (). In transplantation, we often speak of matching “eight out of eight” or “ten out of ten” – this refers to matching alleles at HLA-A, B, C, DRB1 (and maybe DQB1) at least to the two-field level. As HLA typing methods have advanced, we can now readily determine all four fields. This helps avoid ambiguities (for instance, some older methods might only tell you a broad antigen which could be encoded by multiple alleles). Modern reports might list something like “HLA-A*02:01:01G”, where G indicates a group of alleles with identical protein (a shortcut if specific third/fourth field isn’t resolved, but the group is functionally the same).

In summary, HLA nomenclature provides a structured language to describe the incredible diversity of HLA. The four-field system captures different layers of variation: immunologically relevant peptide-binding polymorphisms (fields 1–2) and neutral DNA polymorphisms (fields 3–4) (). Understanding this helps clinicians and researchers communicate exact HLA types and assess compatibility at the appropriate level of resolution.

References:

Janeway CA et al., Immunobiology (on MHC structure and function) – see also NCBI/Immunopaedia resources () ().
WHO HLA Nomenclature Committee, 2010–2020 reports – allele naming conventions () ().
Parham P, The Immune System – HLA genetics and inheritance (example of sibling haplotype segregation) () ().
Rodey and Fuller, Critical Reviews in Immunology 1987 – “Public epitopes and antigenic structure of HLA” (concept of public vs private epitopes) ().
Duquesnoy RJ, Frontiers in Immunology 2016 – HLA epitope and eplet matching in transplantation () ().
Petersdorf EW et al., NEJM 2015 – the role of DRB3/4/5 matching in marrow transplant (illustrates importance of those loci).
Fernandez-Vina M. et al., Human Immunology 2013 – HLA allele and haplotype frequencies (examples of LD and common haplotypes) ().
Lim WH et al., Transplantation Direct 2016 – eplet mismatch vs antigen mismatch in kidney rejection risk (). (Demonstrates clinical relevance of epitope load).