Fc gamma receptors: Their evolution, genomic architecture, genetic variation, and impact on human disease

Summary

Fc gamma receptors (FcγRs) are a family of receptors that bind IgG antibodies and interface at the junction of humoral and innate immunity. Precise regulation of receptor expression provides the necessary balance to achieve healthy immune homeostasis by establishing an appropriate immune threshold to limit autoimmunity but respond effectively to infection. The underlying genetics of the FCGR gene family are central to achieving this immune threshold by regulating affinity for IgG, signaling efficacy, and receptor expression. The FCGR gene locus was duplicated during evolution, retaining very high homology and resulting in a genomic region that is technically difficult to study. Here, we review the recent evolution of the gene family in mammals, its complexity and variation through copy number variation and single‐nucleotide polymorphism, and impact of these on disease incidence, resolution, and therapeutic antibody efficacy. We also discuss the progress and limitations of current approaches to study the region and emphasize how new genomics technologies will likely resolve much of the current confusion in the field. This will lead to definitive conclusions on the impact of genetic variation within the FCGR gene locus on immune function and disease.

1. FUNCTIONAL IMPORTANCE OF FcγRs

The Fc gamma receptors (FcγRs) are a family of cell‐surface proteins that are expressed on a variety of lymphoid, myeloid, and non‐immune cells, which function by binding the Fc portion of immunoglobulin G (IgG). They play a pivotal role in orchestrating both the humoral and innate immune response, serving as an immunological rheostat to achieve the delicate balance between facilitating appropriate responses to infections and averting autoimmunity. ¹ The FcγRs also have a crucial role in determining the efficacy of monoclonal antibody (mAb) immunotherapy ² , ³ and are implicated in the genetic predisposition to autoimmune, infectious, and inflammatory diseases.

Characterized by their intracellular signaling capabilities and affinity for IgG, the six human FcγRs are FcγRI (CD64), FcγRIIa (CD32A), FcγRIIb (CD32B), FcγRIIc (CD32C), FcγRIIIa (CD16A), and FcγRIIIb (CD16B). FcγRI is the sole high‐affinity receptor, whereas FcγRIIa, FcγRIIb, FcγRIIc, FcγRIIIa, and FcγRIIIb are classified as low‐affinity receptors. ⁴ While both high‐ and low‐affinity FcγRs bind antibody immune complexes (ICs) with high avidity, only the high‐affinity FcγRI stably binds monomeric IgG. In contrast, low‐affinity receptors bind to multivalent Fc regions on IgG‐opsonized cells, pathogens, and within soluble ICs. ² , ⁵ , ⁶ The majority of FcγRs trigger intracellular signaling pathways that elicit immune effector functions to eliminate opsonized targets, ⁷ and are so termed activating FcγRs. In contrast, the sole inhibitory FcγR, FcγRIIb, can suppress the effects of activating receptors by triggering an inhibitory signaling cascade, and competing for ligand binding at the cell surface. ⁸ , ⁹ The simultaneous expression of both activating and inhibitory FcγRs on a single cell facilitates precise modulation of the stimulation threshold, referred to as the activating/inhibitory (A/I) ratio. ¹⁰ Depending on the specific FcγR and cell type, diverse cellular responses are elicited upon reaching the activation threshold, including phagocytosis, cytokine secretion, reactive oxygen burst, and cytotoxic granule release. ⁵ Consequently, in conjunction with the varying affinities of different IgG isotypes for various FcγR and cell context‐dependent intracellular signaling pathways, the tight regulation of FcγR expression and activation ensures an appropriate and coordinated immune response. Affinity, expression profiles, and activities of the FcγRs are summarized in Figure 1, alongside details of how they elicit their downstream effects. A central component of FcγR regulation is the germline‐encoded variation in the FCGR genes, which themselves have arisen during evolution as described below.

FIGURE 1.

Function, IgG affinity, and expression of the six human Fcγ receptors (FcγRs). FcγRs bind to the Fc region of immunoglobulin G (IgG) and help coordinate the humoral immune response. FcγRI, FcγRIIa, FcγRIIc, and FcγRIIIa signal for cell activation through an immunoreceptor tyrosine‐based activation motif (ITAM), located in the cytoplasmic tail of the receptor (FcγRIIa, FcγRIIc), or the noncovalently associated FcR gamma chain (FcγRI, FcγRIIIA). In NK cells, FcγRIIIA can also be associated with the CD3 ζ‐chain. ¹¹ , ¹² , ¹³ The aggregation of activating FcγRs at the cell surface facilitates the phosphorylation of tyrosine residues in the ITAMs by Src family protein tyrosine kinases (PTKs), triggering a series of downstream signaling cascades leading to cellular activation. ¹⁴ FcγRIIb is the sole inhibitory receptor and has an immunoreceptor tyrosine‐based inhibition motif (ITIM). Phosphorylation of the ITIM tyrosine residues by Src family PTKs during aggregation of the inhibitory FcγRIIb mediates binding of several phosphatases that then proceed to dephosphorylate other proteins and lipids, terminating activating signaling, resulting in cellular inhibition. ⁸ This functions to limit responses through the activating FcγRs as well as other stimulatory receptors, like the B‐cell receptor. FcγRIIb can also reduce activity by competing for activating FcγRs for binding to IgG. ⁹ FcγRIIIb lacks both transmembrane and cytoplasmic domains, with membrane attachment occurring through a glycosylphosphatidylinositol (GPI) anchor instead. ^#While sometimes labeled a decoy receptor due to the lack of intracellular domains and direct signaling capacity, FcγRIIIb has demonstrated the capability for intracellular signaling transduction and cell activation through association with other cell‐surface receptors, such as FcγRIIa ¹⁵ , ¹⁶ , ¹⁷ , ¹⁸ , ¹⁹ and integrins. ²⁰ Only FcγRI is considered high‐affinity as it can engage with monomeric IgG, whereas the remaining FcγRs interact solely with multimeric IgG immune complexes (ICs) or opsonized targets and thus are regarded as low‐affinity. Binding affinities for the different human IgG (hIgG) subclasses have been determined using surface plasmon resonance (SPR). Shown above are affinity constants (×10⁵ M⁻¹) for K_A with standard deviation. ⁶ Most effector leukocytes express a combination of activating and inhibitory FcγRs that dictate the induction of IgG‐mediated immune responses through their opposing signaling capacities. The responsiveness of an effector cell to IgG is also influenced by differentiation state and inflammatory stimuli as these regulate FcγR expression. ♮There is uncertainty around the impact and functional capacity of FcγRIIc, with its previous designation as a pseudogene ²¹ and expression controlled by the presence or absence of a polymorphism dictating a stop codon. Expression of FcγRs has been measured by various methods in peripheral effector cells, ²² , ²³ tissue‐resident macrophages, ²⁴ and dendritic cells. ⁷ “+” represents constitutive expression (+<++<+++<++++<+++++), “‐” represents absence of expression, “+/−” represents inducible expression, and * represents that expression is determined by FCGR2C genotype.

2. EVOLUTION OF FCGRs

The FCGR gene family arose in our common ancestor with bony fish during the development of the adaptive immune system, coinciding with the emergence of their IgG ligands. ²⁵ , ²⁶ Investigating different FCGR loci has highlighted how the FcγR functional repertoire has evolved through multiple events of duplication and divergence, enabling adaptation to immunological challenges.

2.1. High‐affinity FCGRs

The human FcγRI receptor is encoded by the FCGR1A gene, located on chromosome 1q21.2 ²⁷ and separate from the low‐affinity locus. The high‐affinity FcγRI is unique as it exhibits a third extracellular Ig‐like domain, ²⁸ distinguishing it from the low‐affinity FcγRs which only have two, that facilitates binding to both monomeric and aggregated IgG. ⁶ , ²⁹ Additionally, there are two pseudogenes, FCGR1B on chromosome 1p11.2 and FCGR1C on chromosome 1q21.1, ³⁰ that are not believed to contribute to the surface expression of FcγRI. ³¹ The evolution of the three FCGR1 genes is thought to have involved duplication of an ancestral gene, likely located on chromosome 1p, ³² producing FCGR1A and a second homolog with a 6 base‐pair (bp) insertion. Further duplication of this gene likely resulted in a third FCGR1 gene. The separation of FCGR1A and FCGR1C from FCGR1B onto the long arm of chromosome 1 is hypothesized to have occurred following human divergence from chimpanzees through a pericentric inversion event. ³⁰ , ³³

The origins of the high‐affinity FCGR genes are less understood than those of the low‐affinity FCGR locus; in particular, the exact point of divergence is yet to be classified. One study has reported the low‐affinity receptors emerged first in mammalian evolution. ²⁵ This was postulated from analysis of the opossum genome, diverging from placental mammals 130 million years ago, as two‐domain FcγR genes were present, yet three‐domain receptors were absent. However, this may be attributed to technical limitations of genome assembly or evolutionary loss in the opossum lineage rather than the actual absence of the receptor. ²⁶

2.2. Low‐affinity FCGRs

The human low‐affinity FCGR genes are clustered in a single locus on chromosome 1q23.3. This cluster is highly polymorphic and contains homologous sequences with high sequence identity. The genes encoding FcγRIIa (FCGR2A) and FcγRIIb (FCGR2B) originated from a duplication of an ancestral FcγRII gene before the divergence of primates. ²¹ A subsequent segmental duplication occurred in the human‐gorilla ancestor approximately 9 million years ago ²¹ , ³⁴ , ³⁵ due to nonallelic homologous recombination (NAHR) between FCGR2A and FCGR2B (Figure 2A). This event gave rise to the FCGR3 genes (FCGR3A, FCGR3B) and FCGR2C, a new hybrid member of the FCGR2 family, originating from the 5′ end of FCGR2B and the 3′ end of FCGR2A. Therefore, the chimeric FCGR2C gene contains the sequence for an extracellular domain almost identical to that of FcγRIIb but with intracellular ITAM‐containing components homologous to that of FcγRIIa. The homology between these genes is indicated in further detail in Figure 2.

FIGURE 2.

Evolution and homology of the low‐affinity FCGR locus. (A) Evolution of the low‐affinity FCGR locus in humans. A nonallelic homologous recombination (NAHR) event between FCGR2A and FCGR2B resulted in the duplication of FCGR4. The unequal crossover created the large segmental duplication that is known today and gave rise to the FCGR3A and FCGR3B genes as well as the emergence of the fusion gene FCGR2C. The NAHR event occurred approximately 9 million years ago and humans share the segmental duplication with chimpanzees and gorillas, but not orangutans or macaques. (B) Alignment of the low‐affinity FCGR locus to itself demonstrates homology derived from the NAHR event. The human reference sequence (chr1:161,500,000–161,700,000, GRCh38 p.12) was aligned to itself using minimap2 (‐P ‐k19 ‐w19 ‐m50 ‐‐cs = short), ³⁶ and alignments between different regions of the locus are indicated as diagonal segments. The alignment between identical coordinates is not included and neither are alignments between short interspersed nuclear elements (SINE) in two locations within 4k bp of the start of FCGR2A and in the intergenic region between FCGR3A and FCGR2C. Distinct alignments are indicated by color, insertion/deletions are indicated as diagonal gaps, and percentage identities are given for each alignment. The largest alignment (gray) indicates the segmental duplication regions, and the smaller alignments indicate further homology between FCGR2A, FCGR2B, and FCGR2C; including exon 3, exons 4–6, and part of the 3′‐UTR. Gene structures and locations are shown for MANE select and canonical transcripts and, in concordance with current annotation, FCGR2A and FCGR2C have a total of seven exons compared with earlier annotations which show eight. This change is due to updated understanding that the sequence previously annotated as exon 6 in FCGR2A and FCGR2C is spliced out. This exon numbering is continued throughout the figures and review. (C) Gene homology and protein features for FCGR2A, FCGR2B and FCGR2C. Multiple sequence alignments highlight the extensive homology between FCGR2A, FCGR2B and FCGR2C. Gene region reference sequences (GRCh38 p.12) were extended to include flanking regions and aligned to each other using the anchoring‐based CHAOS‐DIALIGN aligner. ³⁷ An anchor‐based method is required for alignment of short, high‐homology regions (such as exon 3) within longer sequences of lower homology. Matching nucleotides between aligned genes are indicated by color, clearly showing that almost all of FCGR2C is homologous to either FCGR2A or FCGR2B, or both. Additionally, FCGR2B has a limited region toward the 3′ end that is not homologous to either FCGR2C or both FCGR2A and FCGR2C. Gene and transcript structure coordinates are for MANE select transcripts (Ensembl IDs indicated) and do not include an annotated 3′‐UTR region for FCGR2C. The flanking regions are 1000 bp in both the 5′ and 3′ UTR directions, except for the 3′ end of FCGR2C which was extended by 2675 bp to include the region of homology between FCGR2A and FCGR2C 3′‐UTR. Locations of protein features within gene regions are shown for Uniprot IDs P12318 (FcγRIIa), P31994 (FcγRIIb), and P31995 (FcγRIIc) and genome coordinates are accessible via UCSC Genome viewer data tables. ITAM and ITIM signaling motif coordinates can be determined from the location of consensus sequence motifs. ³⁸ , ³⁹ , ⁴⁰ , ⁴¹ (D) Gene homology and protein features for FCGR3A and FCGR3B. Pairwise sequence alignment of FCGR3A and FCGR3B (GRCh38 p.12) shows the extent of homology between these genes. Gene region sequences were extended to include 1000 bp 3′ and 5′ flanking regions and aligned using ggsearch2seq. ⁴² Matching nucleotides in the aligned sequences are shown in beige and these extended gene regions have 96.8% identity. Protein features on genomic coordinates are indicated for Uniprot IDs P08637 (FcγRIIIa) and O75015 (FcγRIIIb) and accessible via UCSC Genome viewer data tables apart from annotation for extracellular and propeptide features, which can be inferred from signal peptide and GPI anchor coordinates.

This FCGR segmental duplication was also shown to be present in the closest relatives to humans: gorillas and chimpanzees. ³⁴ , ³⁵ However, it is absent in orangutans, the remaining members of the Hominidae family, and more distant primates such as gibbons, macaques, and baboons. ²⁶ , ³⁵ , ⁴³ , ⁴⁴ There have been conflicting observations regarding the evolutionary timing of this duplication, such as the fact that chimpanzees lack the FCGR2C and FCGR3B genes ²⁶ and FCGR3 expression is detected on sooty mangabey neutrophils. ⁴⁴ These discrepancies may have arisen due to copy number variation between different individuals of the same species, or incomplete genome assemblies. The evolution of these genes involved several significant changes, including the insertion of a retroviral element introducing an intracellular exon with a non‐canonical ITAM to FCGR2A, ³⁵ and alterations to the transmembrane ⁴⁵ and promoter ⁴⁶ sequences of FCGR3B, distinguishing its pattern of expression from that of FCGR3A.

Identifying the specific immune challenges that shaped FCGR evolution is challenging, but studies on mammalian FCGR3 orthologs indicate that helminth infections exerted significant pressure on the low‐affinity FCGR gene family ³⁴ supporting the hypothesis that balancing immunological responses to varying parasitic burdens has been a critical selective pressure at the FCGR loci over evolutionary time. The maintenance of the segmental duplication, especially the emergence and retention of FCGR3B, underscores its evolutionary advantage to the species. In contrast, the chimeric FCGR2C lacks a definitive functional role and appears to have accumulated mutations that hinder its expression and functionality, such as the exon 3 stop codon, splicing variants, and a compromised 3′‐UTR compared with the FCGR2A paralogous sequence. ⁴⁷ , ⁴⁸ These alterations, which are discussed later, suggest that the FCGR2C gene produced by the segmental duplication may be evolving into a non‐functional pseudogene with the classical FCGR2C‐ORF haplotype representing the last functional vestige of the original gene.

3. FCGR GENETICS IN MICE

Understanding the similarities and differences between human FCGR genes and those in mice is crucial due to the widespread use of murine models in functional and preclinical studies. Mice differ from humans in both the number of genes and the structure and function of the encoded receptors, including their binding affinities for immunoglobulin.

Initially, it was thought that mice had the same three classes of FcγR as humans—FcγRI, II, and III—each encoded by a single gene. ⁴⁹ In contrast, humans have these receptors encoded by eight genes as described above, suggesting that mice did not undergo the same duplication events as humans. ⁵⁰ In 2005, a fourth receptor, FcγRIV, was discovered in rodents and some nonhuman primates, ⁵¹ showing high sequence homology to human FcγRIIIa. ⁵² In mice, the Fcgr genes are located on chromosome 1 (Fcgr2, Fcgr3, and Fcgr4) and chromosome 3 (Fcgr1), ⁵⁰ , ⁵³ corresponding to human 1q23 and 1q21, respectively.

Mice, like humans, have only one inhibitory FcγR, FcγRII, which is orthologous to human FcγRIIb and exhibits low‐affinity binding to murine (m) IgG1, IgG2a, and IgG2b, ⁵⁴ , ⁵⁵ as well as weak binding to mIgE. ⁵⁶ mFcγRIII and mFcγRIV also display low‐affinity binding to mIgE, ⁵⁷ while none of the human (h) FcγRs have shown affinity for hIgE, ⁶ indicating a clear interspecies difference. Humans possess only one high‐affinity FcγR, hFcγRI, whereas mice have two high‐affinity activating receptors: the orthologous mFcγRI, ⁵⁸ which binds to monomeric mIgG2a, ⁶ , ⁵⁷ and mFcγRIV, which also binds monomeric mIgG2a, albeit at a lower affinity than mFcγRI, as well as mIgG2b. Additionally, FcγRI in mice shows low‐affinity binding to IgG2b and IgG3. ⁵⁴ The other activating receptor in mice, mFcγRIII, orthologous to hFcγRIIa, ³⁵ demonstrates low‐affinity binding to mIgG1, mIgG2a, and mIgG2b, ⁵⁵ making it unique among murine FcγRs in its ability to bind mIgG1. ⁵⁴ , ⁵⁷ Cross‐reactivity between the IgG of each species with the other's FcγRs has been well documented. ⁵⁸ , ⁵⁹ , ⁶⁰

The cell‐specific expression patterns of FcγRs differ between the two species. While both humans and mice express hFcγRIIb/mFcγRII in B cells, ²² , ⁶¹ overall expression in mice is generally broader, with higher levels of the inhibitory receptor expressed on monocytes, macrophages, and neutrophils. In mice, FcγRI expression is restricted to monocyte‐derived dendritic cells (DCs), ⁶² a pattern not replicated in humans. ⁷ Conversely, the absence of FcγRIIa expression on human natural killer (NK) cells ²³ is not mirrored by the murine ortholog, mFcγRIII. ⁶³

It is important to note that, although there is reasonably high sequence homology between human and mouse FcγRs, ⁶⁰ the defined interspecies orthologs only share around 65% identity. ⁵⁸ This disparity is evident in the differences in affinities for immunoglobulins ⁵⁴ and expression patterns. Significant variation in IgG affinities for the various FcγRs between mice and humans presents a challenge for determining the efficacy of human tumor‐targeting antibodies in mice. ⁶⁴ , ⁶⁵ Additionally, there is no widely recognized ortholog of mFcγRIV in humans, ⁵⁷ , ⁵⁸ nor human FcγRIIa, FcγRIIc, or FcγRIIIb in mice, further limiting the applicability of murine models in inferring the precise roles of human FcγRs and the net result of their engagement in disease. An additional complexity in translating observations into the clinic relates to the extensive FCGR sequence variation that exists in humans, both at the level of single‐nucleotide polymorphisms (SNPs) and copy number variation (CNV).

4. SEQUENCE VARIATION AT THE FCGR LOCUS

4.1. FCGR1

Although several FCGR1A genetic variants have been identified, they often lack further investigation, thus functionally relevant FCGR1 variation is not yet well characterized. A SNP was discovered in 1995 within the extracellular domain of FcγRI that replaced an arginine in exon 3 of FCGR1A with a termination codon (p.R92X, rs74315310). ²⁹ , ⁶⁶ This mutation has been associated with undetectable expression of FcγRI on phagocytes, without apparent health defects, suggesting possible physiological redundancy. However, this variant was investigated in only one family, limiting broader conclusions.

The impact of three additional FCGR1A SNPs on FcγRI effector functions has been explored in vitro. ⁶⁷ The nonsynonymous variants p.I301M (rs12078005) and I338T (rs142350980), located in exon 6 and within the transmembrane and intracellular domains of the receptor, respectively, ⁶⁸ were found to reduce FcγRI signaling. ⁶⁷ The p.V39I (rs7531523) variant, encoding a mutation in the extracellular domain, impaired the receptor's ability to bind IgG in immune complexes. However, since p.V39I is not involved in the IgG‐binding site, ⁶⁸ the receptor's ability to bind monomeric IgG remained unaffected. ⁶⁷ Although the clinical relevance of these variants has not been investigated and they have low prevalence in the healthy population, ⁶⁷ , ⁶⁹ as FcγRI has been shown to play an important regulatory role in response to various bacterial and viral infections, ⁷⁰ , ⁷¹ , ⁷² it could be postulated that FCGR1A variants that reduce FcγRI effector functions may be associated with a poorer immune response to infection.

One study utilized a structural prediction model to examine the effect of the four aforementioned FCGR1A SNPs (rs74315310, ⁶⁶ rs7531523, rs12078005, and rs142350980 ⁶⁷ ) in addition to a nonsense variant in exon 5, p.Q224X (rs1338887), which encodes a stop codon in place of glutamine in the third extracellular domain. ⁶⁸ While a clear impact of these variants on FcγRI receptor function was not observed, limitations regarding the accuracy of modeling the transmembrane and intracellular domains were reported.

A recent study by Wu et al. identified three novel human FCGR1A variants with long‐range PCR and Sanger sequencing. ⁷³ One SNP in the promotor region, c.‐131C > G (rs1848781), encodes glycine in place of cysteine and significantly increases promoter activity and FcγRI expression on monocytes. ⁷³ An increase in FcγRI expression has previously been associated with systemic lupus erythematosus (SLE) ⁷⁴ and rheumatoid arthritis (RA) progression. ⁷⁵ Another SNP, p.D324N (rs1050204), located in exon 6 in the region encoding the FcγRI cytoplasmic domain, was found to enhance phagocytic activity and the production of pro‐inflammatory cytokines in vitro. ⁷³ Additionally, the FCGR1A gene was found to harbor an indel variant, c.845‐23_845‐17delTCTTTG (rs587598788), which involves a 6‐bp sequence in intron 5. This insertion or deletion occurs near exon 6's splice acceptor site, with the insertion variant appearing to correlate with higher levels of FcγRI expression on monocytes. All three variants were also evaluated for associations with sarcoidosis, a disease characterized by abnormal granuloma development. ⁷⁶ Wu et al. found a strong association between a specific FCGR1A haplotype (rs1848781C‐rs587598788Del‐rs1050204N) and protection against sarcoidosis development. Furthermore, the c.‐131G and p.324N variants were linked to poorer lung function in sarcoidosis patients.

Evidence indicates that CNV also exists in FCGR1 genomic regions, ⁷⁷ with approximately 2% and 18% of the population possessing deletions or duplications, respectively. However, the precise CNV boundaries are not well‐defined. ³² Preliminary studies suggest that dosage effects of FCGR1A are observed in various cancers, with increased expression associated with better outcomes. ⁷⁸ Hence, further research into FCGR1A CNVs is required to ascertain their effects and clinical relevance.

Despite being considered a pseudogene ³¹ an alternatively spliced variant of human FCGR1B lacking exon 5 has been detected in humans. ⁷⁹ This variant appears to express a functional receptor capable of binding IgG aggregates in mice, suggesting that the FCGR1 pseudogenes may have previously unrecognized functions.

4.2. Low‐affinity FCGRs

Several SNPs have been identified in each of the low‐affinity FCGR genes that have functional and/or clinical significance (Table 1 and Figure 3), including effects on auto‐inflammatory and infectious disease susceptibility, and cancer immunotherapy efficacy. In addition, a spectrum of other SNPs exists with unknown impacts on gene expression, protein function, and disease prevalence. NB: Below, we include the reference to the rs number as this is a consistent, searchable descriptor of a SNP independent of isoform and genome assembly.

TABLE 1.

Overview of genetic variation at the low‐affinity FCGR locus.

	Variant		Nucleotide a	Amino acid b	Impact
	Name	Rs ID
FCGR2A	131H/R	rs1801274	c.497A>G	p.His131Arg	H: Higher affinity for IgG, 6
					Kawasaki disease, 80 , 81 , 82 childhood ITP, 83 , 84 , 85 Guillain Barre Syndrome, 86 , 87 Graves disease, 88 ulcerative colitis 89 , 90 , 91
					R: Lower affinity for IgG, SLE, 92 , 93 sepsis, 94 , 95 COVID‐19 death 96
	62Q/W	rs201218628
		rs9427397	c.184C>T	p.Gln62Trp	Tryptophan reduces Ca2+ signaling 97 , W: KD 98
		rs9427398	c.185A>G
Intergenic	2A/3A intergenic (1)	rs2099684	g.161530340A>G	–	Takayasu arteritis 99 , 100
	2A/3A intergenic (2)	rs10919543	g.161538827A>G	–
FCGR3A	158F/V	rs396991	c.526T>G	p.Phe158Val	F: SLE susceptibility 93
					V: Higher affinity for IgG, Ulcerative colitis, 101 ITP susceptibility, 83 , 85 , 102 Rheumatoid arthritis 103
	CNV	–	–	–	FcyRIIIa expression, ADCC
					SLE 104 , 105
	3A intron 1 enhancer haplotype	rs4656317	g.161549329C>G	–	Alt alleles: Increased FcyRIIIa expression, ADCC 106 , 107 , 108
		rs12071048	g.161549118G>A	–
	Intragenic haplotype
	5′‐UTR	rs56199187	g.161551141G>A	–
	Intron 3	rs77825069	g. 161547895G>T	–	Alt alleles: Enhanced affinity to IgG1, IgG3 and IgG4, Susceptibility to herpes, Epstein–Barr and Varicella zoster 109
	48 L/R/H	rs10127939	c. 197T>[G/A]	p.Leu48Arg/His]
FCGR2C	Promoter haplotypes	rs149754834	−386G>C	–	2B.2 associated with 2C‐ORF
		rs34701572	−120T>A	–
	57X/Q	rs759550223	c.169T>C	p.Ter57Gln	Ref allele: Protein truncated
					Q: Expression of FcyRIIc
					Immune thrombocytopenia, 102 , 110 Kawasaki disease, 98 IgG deficiency 111
	Intron 6 splice sites	rs76277413	Donor c.798+1	A>G	Only classical FCGR2C‐ORF haplotype associated with FcyRIIc expression 47
		rs430178	c.799‐1 Acceptor	C>G
	Thai hap				Variant haplotype: HIV‐1 vaccine efficacy 112
	Intron 2	rs114945036	126 in In2	C>T
	Exon 3	rs138747765	c.353	C>T
	Intron 3	rs78603008	c.391+111	G>A
	CNV	–	–	–	FcyRIIc expression (ORF only)
					Kawasaki disease 113
FCGR3B	HNA haplotype c	rs200688856	c.108C>G	p.Ser36Arg	NA2 associated with SLE susceptibility 93
		rs527909462	c.114T>C	p.Leu38Leu
		rs448740	c.194A>G	p.Asn65Ser
		rs5030738 d	c.233C>A	p.Ala78Asp
		rs147574249	c.244A>G	p.Asn82Asp
		rs2290834	c.316A>G	p.Ile106Val
	CNV	–	–	–	FcyRIIIb expression, IC uptake, basophil count
Rheumatoid arthritis, 114 , 115 , 116 SLE, 115 Sclerosis, 117 Sjogren syndrome, 118 Ankylosing Spondylitis, 119 Ulcerative colitis, 101 Bullous Pemphigoid, 120 ANCA‐associated vasculitis, 121 COPD 122
FCGR2B	2B promoter haplotype	rs3219018	−386G>C	–	2B.4 (C and A) increase transcription of FCGR2B, 2B.4 associated with SLE susceptibility 123 , 124 , 125
		rs780467580	−120T>A	–
	232I/T	rs1050501	c.695T>C	p.Ile232Thr	T: Excludes FcyRIIb from lipid rafts, SLE susceptibility, 93 , 126 malaria protection 127
	106Del	rs755222686	c.315_318del	pAsn106Del	Del: Increased levels of IgG and abolished binding to IgG1 and IgG3

^a Nucleotide numbering reported according to which direction the gene is read in (sense or antisense strand). For example, nucleotides are positive strand from the FCGR2 genes as they are on the forward strand while nucleotides for the FCGR3 genes are on the negative strand as they are on the reverse strand.

^b Inconsistencies in the amino acid numbering exist in the literature due to discrepancies in naming based on their position, with some including the signal peptides and others focusing on the mature protein, excluding signal peptides. While the current official HGVS guidelines recommend representing protein‐coding variants in the complete primary translation product, and not a processed, mature, or functional protein, this table follows the amino acid numbering most historically used and adopted by the research community.

^c NHA haplotype in the variant name colum for FCGR3B. the footnote explains how the 6 SNPs impact on neutrophil function.

^d The SNP that is not involved in distinguishing NA1 from NA2, this SNP is the 1 position that distinguishes NA2 from SH.

FIGURE 3.

Structure, features, and resolution of the low‐affinity FCGR locus. The figure shows the chromosomic position, gene transcripts, CNRs, and SNPs of interest alongside sequencing coverage for the low‐affinity FCGR locus using various technologies. The location of the FCGR low‐affinity locus is shown within the region chr1:161,490,000–161,700,000 as a red vertical line. Transcript structure and locations are shown for MANE select transcripts, or canonical transcripts where there is no MANE select. Copy number regions (CNR1‐5) are shown with black bars indicating high‐confidence regions of duplication/deletion and dotted outlines indicating regions that may be included in the duplication/deletion. Positions and details of well‐characterized SNPs across the region are indicated as “SNPs of interest” and the track below indicates small variants reported in the gnomAD database. The subsequent tracks are all based on sequencing information from a variety of technologies spanning commonly used read lengths. Technology‐dependent loss of signal occurs in regions of high homology, particularly around FCGR2C and FCGR2B, where high‐confidence alignments are lacking for the majority of technological approaches. The bottom five tracks display alignment coverage data of the HG002 benchmarking genome from public data resources, utilizing sequencing alignment files provided. Median and N50 read lengths are shown for long‐read technology datasets and calculated using alignments from the locus only. Small variants from the Genome Aggregation Database: There are 47,801 SNVs and indels recorded across the region based on short‐read sequencing‐by‐synthesis of whole genomes and exomes (gnomAD version 4.1.0). ¹²⁸ Predicted loss‐of‐function, missense, and synonymous variants are shown in yellow and all other variants are indicated in semi‐transparent gray. Overlaps of multiple variants result in darker gray regions. It should be noted that some variants are potentially missed/misassigned due to homology issues. For example, the well‐characterized FCGR2C‐57X/Q variant is not recorded. (Black) Short‐read ATAC‐seq data using sequencing‐by‐synthesis: Data shown is for primary monocytes (Gene Expression Omnibus GSE165997, GSM5059898 ¹²⁹ ) and the single sample coverage track indicates open chromatin regions when either all alignments are included (gray, HISAT2 aligner) or only uniquely aligned alignments (black, mapq = 60). Similarly, consensus peaks for the whole dataset are indicated when all alignments are used (gray) or uniquely aligned only (black). Sequencing was performed using 2 × 75‐bp reads on the Illumina NovaSeq platform. (Light orange) Short‐read, sequencing‐by‐synthesis. Genome‐in‐a‐bottle (GIAB) deposited data from the National Institute of Standards and Technology (NIST) utilizing 2 × 148‐bp reads on the Illumina HiSeq platform. ¹³⁰ The gray coverage track includes all alignments and orange track indicates the coverage of high‐confidence alignments (mapq = 70). Both coverage tracks are downsampled to 25% of the total data. These alignments utilized the Novoalign aligner and all reported alignments are primary. (Dark orange) Long reads based on short‐read sequencing. Demonstration data are available on application from Illumina, project ID 385613228. Coverage from all available alignments is shown in gray (minimap2 aligner, secondary alignments reported), and from unique alignments only in dark orange (mapq = 60). (Blue) Single‐molecule real‐time (SMRT) long reads. GIAB deposited data from NIST/MtSinai ¹³⁰ utilizing PacBio P6‐C4 chemistry and the RS II platform. Coverage when including all alignments is shown in gray (minimap2 aligner, secondary alignments reported), and unique alignments only in blue (mapq = 60). (Green) Circular consensus sequencing (CCS) of single‐molecule real‐time long reads. GIAB deposited data utilizing the PacBio 15‐kb and 20‐kb insert size HiFi reads and Sequel II platform. ¹³¹ Coverage when including all alignments is shown in gray (minimap2 aligner, only primary alignments reported), and unique alignments only in green (mapq60). (Pink) Nanopore long reads. GIAB deposited data by the University of California Santa Cruz (UCSC) utilizing Oxford Nanopore ultra‐long‐read and PromethION platform. ¹³² Coverage when including all alignments is shown in gray (minimap2 aligner, secondary alignments reported), and unique alignments only in pink.

4.3. FCGR2A

One of the most widely studied SNPs in FCGR2A is the A to G transition within exon 4 (rs1801274), resulting in a histidine to arginine substitution (p.H131R) within the IgG‐binding domain. ¹³³ While both alleles exhibit comparable binding affinities for hIgG3 and hIgG4, the FCGR2A‐131H isoform has a higher affinity for hIgG1 and especially hIgG2. ⁶ In contrast to the 131R variant, the FCGR2A‐131H allele has been shown to facilitate the phagocytosis of IgG2‐opsonized entities, ²¹ , ¹³⁴ augment neutrophil degranulation, and phagocytosis, ¹³⁵ , ¹³⁶ and enhance IgG2‐mediated IL‐1β secretion. ¹³⁷ Both H and R alleles have been associated with various inflammatory diseases, highlighting the intricate role FcγRIIa plays at the intersect of multiple immunoregulatory pathways. For example, FCGR2A‐131H has been associated with increased incidence of Graves' disease, ⁸⁸ ulcerative colitis, ⁸⁹ , ⁹⁰ , ⁹¹ childhood immune thrombocytopenic purpura (ITP), ⁸³ , ⁸⁴ , ⁸⁵ and Kawasaki disease (KD), ⁸⁰ , ⁸¹ , ⁸² whereas FCGR2A‐131R has been associated with sepsis ⁹⁴ , ⁹⁵ and SLE. ⁹² , ⁹³

Additionally, two adjacent SNPs (rs9427397 and rs9427398) within exon three of FCGR2A encode a composite SNP (rs201218628) which replaces a glutamine with a tryptophan residue (p.Q62W) in a region of FcγRIIa that is predicted to be a dimerization domain. Initially reported in a common variable immunodeficiency patient cohort, FCGR2A‐62W has been connected to reduced calcium signaling and impaired MAPK phosphorylation. ¹³⁸ However, its precise functional significance remains elusive as it does not affect FCGR2A expression or neutrophil antibody‐dependent cellular cytotoxity (ADCC) capacity. ⁹⁸ , ¹³⁸ Furthermore, a comprehensive meta‐analysis found a significant association with KD, although its elevated disease risk may result from strong linkage disequilibrium (LD) with other SNPs. ⁹⁸

4.4. FCGR3A

FCGR3A also contains a well‐established SNP within exon 4 (rs396991) that affects the receptor's IgG affinity. ¹³⁹ It results in a phenylalanine to valine substitution at amino acid position 158 (F158V) within the second extracellular domain, with the FCGR3A‐158V variant exhibiting enhanced affinity for all IgG subclasses. ⁶ , ¹³⁹ The homozygous FCGR3A‐158V/V genotype has been associated with heightened NK cell ADCC capacity, ¹⁴⁰ ITP susceptibility, ⁸³ , ⁸⁵ , ¹⁰² and increased risk of developing RA for individuals of European descent. ¹⁰³ However, the lower affinity isoform encoded by FCGR3A‐158F has been associated with increased susceptibility for and development of SLE ⁹³ and lupus nephritis. ¹⁴¹ , ¹⁴²

Within exon 3, there are two alternative alleles that have both been reported to affect FcγRIIIa's affinity for IgG (rs10127939). At amino acid position 48 within the membrane‐distal Ig‐like domain, the reference allele encodes a leucine residue, but substitutions to either G or A can result in an arginine or histidine residue, respectively (FCGR3A‐48L/R/H). Both of these variants show increased binding capacity for hIgG1, hIgG3, and hIgG4. ¹⁴³ This triallelic SNP has also been associated with the FCGR3A‐158V variant whereby the presence of an R or H and at least one copy of the 158‐V allele results in enhanced hIgG‐binding abilities. ¹⁴⁴ , ¹⁴⁵

The FCGR3A‐48 L/R/H variant was additionally found to be in LD with two other SNPs and an indel, forming an intragenic haplotype (IH) associated with high FcγRIIIa expression on monocytes and NK cells. ¹⁰⁶ The original FCGR3A‐IH described consisted of a 5′‐UTR SNP‐1405A (rs56199187), an intronic indel +690–691InsC (rs33959719), the triallelic nonsynonymous polymorphism 48‐L/R/H (rs10127939), and a second intronic variant +1842T (rs77825069). However, the entry for the intronic indel no longer exists in public databases, suggesting its previous description was a technical artifact.

Furthermore, a study performed on 101 healthy Japanese individuals demonstrated how two SNPs (rs4656317 and rs12071048) located within a potential enhancer region in FCGR3A intron 1 were in strong LD with the FCGR3A‐158V SNP, and significantly impacted NK cell FcγRIIIa expression and ADCC activity. ¹⁰⁷ It has been suggested this enhancer haplotype (EH) may be responsible for the contrasting findings ¹⁴⁶ , ¹⁴⁷ regarding whether the FCGR3A‐158V variant is accountable for changes in expression. ¹⁰⁷

There have also been two intergenic SNPs identified (rs2099684 and rs10919543) between FCGR2A and FCGR3A that have been associated with the autoimmune disease Takayasu arteritis in Turkish, North American and Chinese Han populations. ⁹⁹ , ¹⁰⁰ These variants, which are approximately 11.5 and 3.0‐kb upstream of FCGR3A, have been shown to be in high linkage disequilibrium with each other and associated with increased FCGR2A expression levels in lymphoblastoid cells. ¹⁴⁸

4.5. FCGR2C

As previously mentioned, FCGR2C is a fusion gene that is comprised from exons homologous to those from FCGR2A and FCGR2B (Figure 2C). Despite some in vitro evidence showing FcγRIIc can bind IgG, induce calcium flux, and activate cytotoxic NK cell responses, establishing its true in vivo functional capacity has proven challenging. ⁴⁷ , ¹⁴⁹ , ¹⁵⁰ This is partly because it was relatively understudied being previously considered as a pseudogene. ²¹ , ¹⁵¹

A significant proportion (>80%) of the human population harbor a C to T substitution within exon 3, resulting in a stop codon (Ter) at position 57 instead of a glutamine within the open reading frame (ORF). ¹⁵² , ¹⁵³ FcγRIIc was first identified on NK cells of individuals with the ORF variant (p.57Gln, FCGR2C‐ORF) instead of the stop codon (p.57Ter, FCGR2C‐Stop). ¹⁴⁹ However, as most individuals are homozygous for the FCGR2C‐57X genotype, ⁹⁸ which encodes the truncated isoform, the termination codon is integrated within the current GRCh38 reference genome sequence with the protein‐coding variant described as a SNP (rs759550223). ¹⁰² , ¹⁵⁴ The FCGR2C‐ORF variant has been linked to increased susceptibility to several autoimmune diseases, including KD, ⁹⁸ ITP, ¹⁰² , ¹¹⁰ SLE, ¹⁵⁵ and systemic sclerosis, ¹⁵⁶ in addition to IgG subclass deficiency. ¹¹¹ It has been reported that the FCGR2C‐ORF variant also dictates the expression of FcγRIIc on B cells, reducing the inhibitory effects of FcγRIIb and modulating the overall level of B‐cell activation. ¹⁵⁵ However subsequent, independent research has failed to reproduce the finding of FCGR2C expression on B cells. ¹⁵⁷

However, only genotyping FCGR2C‐57X/Q is not sufficient to infer FcγRIIc expression as there have been FCGR2C‐57Q carriers identified with undetectable levels of cell‐surface FcγRIIc. Instead, they expressed various noncoding FCGR2C transcripts that predominantly lacked exon 6 and included a 62‐bp intron 6 insertion. ⁴⁷ These aberrantly spliced isoforms were originally attributed to polymorphic donor and acceptor splice sites within intron 6 (rs76277413 and rs430178) that, together with the FCGR2C‐57Q SNP, were termed nonclassical FCGR2C‐ORF haplotypes (Table 2). However, a more recent study by the same authors only included the non‐canonical donor site to characterize the nonclassical FCGR2C‐ORF haplotype. ⁹⁸ The non‐functional splice site sequences are incorporated within the GRCh38 reference while the canonical splice sites are annotated as SNPs.

TABLE 2.

FCGR2C haplotypes that impact expression of FcyRIIc.

	FCGR2C exon 3 (p.57)	Intron 6 donor splice site (c.798+1)	Consequence
Classical FCGR2C‐ORF	Q	G	FcγRIIc expression
Nonclassical FCGR2C‐ORF	Q	A	No/marginal FcγRIIc expression
FCGR2C‐Stop	X	A or G	No FcγRIIc expression

Note: Haplotypes for FCGR2C detailing amino acid and nucleotide changes which have consequences for FcyRIIc expression. ⁴⁷ , ⁹⁸ X represents the termination (Stop) codon. Traditionally only the SNP in exon 3 of FCGR2C was genotyped to determine the presence of an open reading frame (ORF); however, for a more accurate prediction of FcγRIIc expression, it is necessary to also genotype the splice sites in intron 6.

Recently, additional FCGR2C single nucleotide variants (SNVs) have garnered attention for their potential functional effects, although their definitive biological impact has not yet been conclusively determined. Of particular interest is a haplotype comprising three FCGR2C SNVs—c.134‐96C>T within intron 2 (rs114945036), c.353C>T within exon 3 (rs138747765), and c.391+111G>A within intron 3 (rs78603008)—which have been reported to be in complete LD and significantly enhance vaccine efficacy in a cohort of Thai individuals involved in an HIV‐1 vaccine trial. ¹¹² This haplotype has attracted considerable attention due to its investigation in the only HIV‐1 vaccine trial demonstrating any degree of protection. ¹⁵⁸ However, the functional significance of these variants warrants cautious interpretation as the primers were not FCGR2C‐specific, ¹⁵⁹ and a connection with FcγRIIc surface expression could not be elucidated as only one participant in the study harbored the FCGR2C‐57Q SNP.

Subsequent investigation of this haplotype within the Black South African population revealed that only the intron 2 SNV (rs114945036) was polymorphic ¹⁶⁰ and, in contrast to the above, was associated with an increased likelihood of HIV‐1 disease progression. ¹⁶¹ Once again, the FcγRIIc protein was not considered to be involved in the underlying mechanism as only one individual possessed the classical FCGR2C‐ORF haplotype. Instead, the authors presented an in silico analysis which supported previous claims that these variants could result in alternative splicing of FCGR2C transcripts as a putative CTCF binding motif within a weak enhancer was disrupted. Supporting the notion that these FCGR2C variants regulate gene expression, increased levels of the last exon from FCGR2A and/or FCGR2C were detected in lymphoblastoid B‐cell lines harboring the Thai haplotype SNVs. ¹⁶²

The conflicting outcomes in HIV‐1 studies may be due to LD relationships specific to different genetic backgrounds that are not always determined. For example, the intron 2 SNV (rs114945036) was also found to be in LD with a nonsynonymous exon 4 variant and two intron 5 SNVs in the Thai cohort (rs373013207, rs74341264, and rs201984478), ¹¹² whereas was linked to the FCGR2C‐57Q SNP and two intron 1 variants (rs2169052 and rs111828362) in the South African cohort. ¹⁶¹

4.6. FCGR3B

Within exon 3 of FCGR3B, which encodes the membrane‐distal Ig‐like extracellular domain of FcγRIIIb, a cluster of polymorphisms collectively represents the human neutrophil antigen (HNA)‐1 system. The most common variants, HNA‐1a and HNA‐1b, are differentiated by five SNPs that lead to four amino acid substitutions and affect N‐linked glycosylation (Table 3). HNA‐1a and HNA‐1b are encoded by the NA1 and NA2 haplotypes, respectively. These variants alter the receptor's affinity for hIgG isotypes, with HNA‐1a having a higher affinity for hIgG1 and hIgG3, resulting in more effective phagocytosis of opsonized targets. ⁶ , ¹³⁵ , ¹⁶⁵ In vitro experiments have shown that the FcγRIIIb allotype encoded in the NA1 haplotype had approximately double the affinity to monomeric hIgG3 of the NA2 haplotype. ¹⁶⁶ Another reported allotype is HNA‐1c that is encoded by the SH haplotype. The SH variant is identical to NA2 at the aforementioned five SNPs but differs from NA2 at an additional position. The functional consequences of this SNP, and thus difference between HNA1b and HNA1c allotypes, are not well characterized. Several other HNA variants have been identified, although their functional and clinical significance remains to be confirmed. ¹⁶³ , ¹⁶⁷ , ¹⁶⁸

TABLE 3.

Single‐nucleotide‐proteins (SNPs) involved in the FCGR3B haplotypes which determine the allotypic variants of the human neutrophil antigen (HNA)‐1 classification system.

Rs ID	FCGR3B‐NA1		FCGR3B‐NA2		FCGR3B‐SH
	Nucleotide	Amino acid	Nucleotide	Amino acid	Nucleotide	Amino acid
rs200688856	c.108G	Arginine (R)	c.108C	Serine (S)	c.108C	Serine (S)
rs527909462 a	c.114C	Leucine (L)	c.114T	Leucine (L)	c.114T	Leucine (L)
rs448740	c.194A	Asparagine (N)	c.194G	Serine (S)	c.194G	Serine (S)
rs5030738 b	c.233C	Alanine (A)	c.233C	Alanine (A)	c.233A	Aspartic acid (D)
rs147574249	c.244G	Aspartic acid (D)	c.244A	Asparagine (N)	c.244A	Asparagine (N)
rs2290834	c.316G	Valine (V)	c.316A	Isoleucine (I)	c.316A	Isoleucine (I)

Note: These are the three major haplotypes but typing is considered incomplete because rare cases have been revealed of different combinations of these SNPs and/or additional variants. ¹⁶³ , ¹⁶⁴

^a rs527909462 is a synonymous variant.

^b rs5030738 represents the SNP that is not involved in distinguishing NA1 from NA2; it is the sole position that distinguishes NA2 from SH.

Susceptibility to various autoimmune disorders has been linked to these HNA antigens. The higher affinity HNA‐1a has been associated with an increased risk of a condition marked by FcγR‐mediated neutrophil activation, known as anti‐neutrophil cytoplasmic antibody systemic vasculitis, ¹⁶⁸ whereas the lower affinity HNA‐1b isoform has been linked to SLE development. ¹⁶⁹ , ¹⁷⁰ Additionally, the HNA system holds clinical significance in pregnancy; rare instances of potentially fatal neonatal immune neutropenia have been observed when fetal neutrophils expressing paternal HNA antigens are targeted by maternal antibodies. ¹⁷¹ , ¹⁷² , ¹⁷³

4.7. FCGR2B

Unlike the other two FCGR2 family members which have seven exons, FCGR2B has eight exons that are alternatively spliced. The predominant FcγRIIb isoforms, FcγRIIb1 and FcγRIIb2, exhibit distinct localization patterns and functional properties. FcγRIIb1, transcribed from all eight exons, is the only FcγR routinely expressed in B cells and functions to inhibit B‐cell receptor (BCR) signaling, thus limiting B‐cell expansion and activation in the presence of antigen‐specific ICs. ¹⁷⁴ In contrast, FcγRIIb2 has the 57‐bp exon 6 spliced out from between the transmembrane domain and cytoplasmic ITIM. It is the primary transcript expressed in myeloid cells and is mainly involved in endocytosis of ICs. ³⁸ Exon 6 of FCGR2B, present in FcγRIIb1, enhances B‐cell membrane tethering, whereas its absence from FcγRIIb2 allows for rapid internalization. ¹⁷⁵ This exon does not have a corresponding exon in the current annotation for FCGR2A or FCGR2C (Figure 2C).

A well‐studied polymorphism within exon 5 of FCGR2B (rs1050501), involving a T to C substitution, has attracted attention for its impact on FcγRIIb's inhibitory signaling and its potential to disrupt immune system balance. This substitution, occurring at amino acid position 232 within the transmembrane region and resulting in an isoleucine to threonine change (I232T), was initially identified in a Japanese cohort of SLE patients. ¹⁷⁶ The FCGR2B‐232T variant has been shown to interfere with FcγRIIb's ability to attenuate the activating signaling of the BCR by impeding its association with lipid rafts in the plasma membrane. ¹⁷⁷ , ¹⁷⁸ Moreover, as FcγRIIb is also expressed on dendritic cells (DCs), the presence of FcγRIIb‐p.Thr232 can also impact DC maturation and consequently T‐cell stimulation. ¹⁷⁹ , ¹⁸⁰ , ¹⁸¹ In multiple studies across numerous ethnic backgrounds, homozygosity of the FCGR2B‐232T variant has been strongly associated with SLE. ⁸ , ⁹³ , ¹²⁶ , ¹⁷⁶ , ¹⁸² , ¹⁸³ , ¹⁸⁴ , ¹⁸⁵ Conversely, a T/T genotype was also shown to provide a protective effect against severe malaria infection in an East African population. ¹²⁷

Allelic variation can also impact the expression levels of the FcγRs. Notably, two crucial SNPs reside in the promoter regions of both FCGR2B and FCGR2C, positioned at −386 and −120 relative to the translational start site (Table 4). At the −386 position, these SNPs can either be G or C, while at the −120 position, they can be A or T, resulting in four distinct haplotypes named 2B.1, 2B.2, 2B.3, and 2B.4. The 2B.1 haplotype is prevalent in both FCGR2C and FCGR2B promoter regions, while the 2B.2 and 2B.4 haplotypes have only been reported upstream of FCGR2C and FCGR2B, respectively. To date, there has never been reported a case of the 2B.3 haplotype of −386G with −120A. With a luciferase reporter assay, the rarer FCGR2B.4 haplotype was demonstrated to increase FCGR2B transcription in comparison with the more common FCGR2B.1 haplotype. ¹²³ This observation was then supported when transcription factor (TF) binding and FcγRIIb expression were seen to be enhanced in primary B cells with the FCGR2B.4 haplotype. ¹⁸⁶ Augmented transcriptional activity from 2B.4 has also been observed in neutrophils and monocytes. ¹²⁴ , ¹⁸⁷ The 2B.4 haplotype has also been associated with increased SLE susceptibility. ¹²³ , ¹²⁴ , ¹²⁵

TABLE 4.

Promoter haplotypes for FCGR2B and FCGR2C.

Promoter haplotype	c.−386	c.−120	Found upstream of
2B.1	G	T	FCGR2B and FCGR2C
2B.2	C	T	FCGR2C
2B.3	G	A	Not reported
2B.4	C	A	FCGR2B

Note: Promoter haplotypes for FCGR2B and FCGR2C detailing the nucleotide changes and genes effected. ¹²³ Haplotypes comprised of two SNPs at nucleotide positions −386 and −120 relative to the start of translation in homologous promoter regions of FCGR2B and FCGR2C. Specific haplotypes are associated with particular genes and the 2B.3 haplotype has never been reported.

Additionally, a recently described in‐frame deletion of codon AAT within exon 3 (rs755222686, p.Asn106del) was shown to completely eradicate FcγRIIb binding with hIgG1 and hIgG3. ¹⁸⁸ While this non‐functional protein resulted in heightened levels of hIgG, it has only been reported in a very small subset of the Icelandic population so far.

5. STRUCTURAL VARIATION AT THE FCGR LOCUS

5.1. History

CNV involving FCGR genes have been notoriously challenging to characterize due to the genomic complexity of these loci and limitations in available technologies. The first report of CNV at the FCGR locus dates back to the 1990s when a homozygous deletion of FCGR3B was observed in mothers of children with neonatal immune neutropenia. ¹⁷¹ , ¹⁷² , ¹⁷³ This deletion was subsequently found to be linked with FCGR2C CNV. ¹⁸⁹ , ¹⁹⁰ To investigate associations with autoimmune diseases, methods such as array comparative genomic hybridization (aCGH) and restriction enzyme digest variant ratio assays were developed, and large‐scale studies examining global CNVs suggested that FCGR2A and FCGR2B were also candidate genes for duplication or deletion events. ¹⁹¹ , ¹⁹² , ¹⁹³ , ¹⁹⁴

However, in 2008 the multiplex ligation‐dependent probe amplification (MLPA) assay was developed to study the genetic variation at the FCGR locus in a single assay. ¹⁰² It was hypothesized that the variation captured by both CNV and SNPs could collectively encode a “susceptibility phenotype” for autoimmune/inflammatory diseases as the balance between activating and inhibitory signaling was altered by genomic variants. The MLPA assay was conducted on >600 individuals to demonstrate extensive variation at the FCGR cluster, but showed only FCGR3A, FCGR2C, and FCGR3B were subject to CNV—and not FCGR2A or FCGR2B as previously reported. ¹⁰² , ¹⁵⁴

In 2012, a study by Machado et al. utilized Sanger sequencing of PCR amplicons to investigate the NAHR events generating FCGR CNV and localized the breakpoints to distinct hotspots. ³⁴ This analysis involved identifying and utilizing paralogous sequence variants (PSVs), which are single nucleotide differences between the two paralogous repeats of the segmental duplication. However, in 2013, Mueller et al. examined the genetic pathology of FCGR3B deletions and reported that 36% of the candidate PSVs were polymorphic, compromising any breakpoint analysis that used them. ¹⁵⁹ Mueller et al. also refined the boundaries of the FCGR2C/FCGR3B CNV to two highly paralogous 24.5‐kb blocks covering the 3′ ends of FCGR2C and FCGR2B, which they stated were devoid of true PSVs not subject to allelic variation. In 2017, Rahbari et al. reanalyzed the same fosmid sequences and were able to further define the breakpoints within this 24.5‐kb block, attributing the discrepancies to the consideration of the duplications present. ¹¹⁴

5.2. Structure

MLPA analysis of >4000 individuals was conducted in 2015 and four distinct CNV regions (called CNRs) were described, mapping the most common areas of gain or loss within the low‐affinity FCGR locus. ⁴⁸ The PSVs utilized in this MLPA assay included 7 pairs that were designed to distinguish between the two 82‐kb paralogous repeat units, four of which were designed to detect nonpolymorphic PSVs. During this work, the breakpoints of the CNR were then further refined, documenting the generation of functional chimeric genes and the novel fourth variant CNR4 for the first time. This study also provided support for the idea that there is population variation in FCGR copy number profiles. ¹⁹⁴ Briefly outlined below (and depicted in Figure 3) is the current understanding of the structure of the CNRs, where the numbering reflects their prevalence.

• CNR1 includes FCGR2C, HSPA7, and FCGR3B with breakpoints thought to occur within the intergenic region between FCGR3A‐FCGR2C and FCGR3B‐FCGR2B.

• CNR2 has intragenic breakpoints in intron 6 of FCGR2A and FCGR2C. It encompasses the entire genomic sequence of the HSPA6 and FCGR3A genes as well as the last exon of FCGR2A and all but the last exon of FCGR2C.

• CNR3 includes FCGR3A, FCGR2C, and HSPA7 with the breakpoints within the intergenic region between HSPA6‐FCGR3A and HSPA7‐FCGR3B. Like CNR1 and CNR2, both deletions and duplication of CNR3 have been observed.

• CNR4 has intragenic breakpoints and includes the entire genomic sequence for the HSPA7 and FCGR3B genes. Its start includes exon 4 of FCGR2C, and its end contains the proximal part of FCGR2B including exon 3. Only deletions of CNR4 have been reported.

• CNR5, a novel CNR, was reported in 2021 in some individuals indigenous to the Ecuadorian highlands. ¹⁹⁵ The authors showed that the CNR deletion spanned from the third intron of FCGR3A to the third intron of FCGR3B, resulting in the creation of a FCGR3B/A chimeric gene which was associated with FcγRIIIa and FcγRIIIb deficiency.

Considering that FCGR CNV is facilitated by recurrent NAHR between the highly homologous repeats of the segmental duplication at a high mutation rate, ³⁴ in addition to the development of increasingly sophisticated technologies which exhibit enhanced fidelity, it is anticipated that more CNRs will be discovered as the FCGR locus is explored within larger and more diverse cohorts.

5.3. Functional consequences

The CNRs at the FCGR locus can have significant functional consequences for FcγR expression, and thus the response to IgG, through two main mechanisms: (i) changes to gene dosage and (ii) the introduction of chimeric genomic sequences.

Gain of CNRs that result in increased FCGR3B gene copies leads to an augmentation of FCGR3B mRNA levels, ⁹⁸ and protein expression at the cell surface. ¹²⁴ , ¹⁵⁴ , ¹⁸⁹ , ¹⁹⁰ Increased expression of FcγRIIIb has been shown to enhance the ability of neutrophils to bind and uptake immune complexes. ¹⁹⁶ Recke and co‐workers also reported elevated ROS production in donors with FCGR3B CNV. ¹²⁰ Loss of the same CNRs, such as CNR1 or CNR4, have been shown to reduce FcγRIIIb expression and function on neutrophils, ¹⁷⁰ , ¹⁹⁶ associated with the impairment of IgG‐mediated opsonization. ¹⁷¹ , ¹⁹⁷ Similarly, extra copies of FCGR3A (through duplication of CNR2 or CNR3) have been linked with increased NK cell‐surface expression of FcγRIIIa; however, the association is not as clear compared with the FCGR3B effects. ⁹⁸ , ¹⁵⁴ CNV of FCGR2C has also been linked to changes in expression, but only in individuals that harbor the FCGR2C‐ORF SNP. ⁹⁸ , ¹⁰²

As previously mentioned, CNRs do not just result in expression changes due to differences in gene dosage but can also create chimeric gene products and cause the rearrangement of key cis‐acting regulatory elements. For example, deletion of FCGR3B has been reported to generate a chimeric event which results in aberrant regulation of FCGR2B transcription. The deletion causes the FCGR2C regulatory sequence to juxtapose the coding sequence of FCGR2B, placing FCGR2B expression under the control of the 5′ flanking sequence of FCGR2C. ¹⁵⁹ This results in ectopic accumulation of FcγRIIb on NK cells that apparently impairs ADCC, potentially providing an explanation for SLE susceptibility. ⁴⁷ , ¹⁵⁹

Furthermore, deletion and duplication of CNR2 have both been shown to create chimeric gene products. ¹⁸⁴ The novel chimera generated by CNR2 deletion is comprised of the first seven exons of FCGR2A followed by exon 8 and the 3′‐UTR of FCGR2C. ⁴⁸ Interestingly, despite the distal part of FCGR2C being directly derived from FCGR2A during the segmental duplication event, the FCGR2A/2C chimeric gene product was functionally distinct from the wild type FcγRIIa. It is expressed at significantly lower levels on neutrophils and monocytes thus reducing the ability to induce a response to IgG. ⁴⁸ In contrast, CNR2 duplication can create the inverse FCGR2C/2A chimeric product that consists of the first seven exons of FCGR2C followed by exon 8 and the 3′‐UTR of FCGR2A. This can result in increased FcγRIIc expression at the cell surface but only if the FCGR2C‐ORF SNP coexists on the same haplotype. The difference in expression was attributed to the 3′‐UTR of FCGR2A providing increased stability to the mRNA molecule compared with the 3′‐UTR of FCGR2C.

The CNR4 deletion can result in an FCGR2B null variant, resulting in loss of the first three exons of FCGR2B to be replaced with exons 1–3 of FCGR2C—potentially introducing the FCGR2C‐stop codon in exon 3. When this polymorphic stop codon from FCGR2C is used, it effectively silences the inhibitory FcγRIIb gene and causes expression of FcγRIIb to be reduced by half in heterozygous individuals. ⁴⁸

5.4. Disease consequences

The main rationale for the previous studies was to elucidate the structure of the FCGR locus under different copy number states and link variants to cellular responses, pathology, and disease implications. Deletions of FCGR3B have previously been associated with the susceptibility to numerous autoimmune diseases: glomerulonephritis, ¹⁹² SLE, ³⁴ , ¹¹⁵ , ¹²¹ , ¹⁵⁹ , ¹⁷⁰ , ¹⁸⁵ , ¹⁹⁶ , ¹⁹⁸ ulcerative colitis, ¹⁰¹ rheumatoid arthritis (RA), ¹¹⁴ , ¹¹⁶ , ¹⁹⁹ , ²⁰⁰ , ²⁰¹ ankylosing spondylitis, ¹¹⁹ systemic sclerosis, ¹¹⁷ primary Sjögren syndrome, ¹¹⁸ , ²⁰⁰ microscopic polyangiitis, and Wegener's granulomatosis. ¹²¹ The more recently described CNR4 deletion is reported to be very rare, detected in only 0.1% (5/4357) of the study population, and does not necessarily contribute to autoimmune disease as 3 of the 5 individuals with the variant were healthy adults. However, the other two cases were patients with vasculitis and SLE, so it was suggested loss of CNR4 may predispose individuals to autoimmunity. ⁴⁸ Similarly, knockout mice that lack FcγRIIb are viable but are prone to severe, early onset autoimmunity, ²⁰² , ²⁰³ but only in certain genetic contexts that exhibit higher disease incidence. ²⁰⁴

The data available on the association of FCGR3A CNV and autoimmune diseases have discrepancies, with loss of FCGR3A being associated with RA and SLE in Taiwanese cohorts ¹⁰⁴ but not in individuals of European descent. ¹⁹⁹ These contradictions could be the result of true differences between populations, small cohort sizes, or limitations of the methodologies. A recent UK Biobank study, that investigated CNVs, which modify protein products, showed a significant association between FCGR3B copy number and increased basophil count. ¹²² A reduction in FCGR3B gene dosage was linked with a higher risk of developing chronic obstructive pulmonary disease (COPD). However, it was also noted that while the haplotype‐informed analysis could differentiate between the FCGR3A and FCGR3B paralogs, other functional FCGR SNPs/CNRs could be the causal variants for other associations (e.g., lymphocyte, monocyte, and eosinophil counts) at the locus.

6. LINKAGE DISEQUILIBRIUM AND POPULATION VARIATION

Additional considerations for the interpretation of genetic variants at the low‐affinity FCGR locus are (i) the significant variation across different ethnic backgrounds and (ii) the pronounced extent to which LD is present. Due to the variants being in such close physical proximity to each other, they are prone to co‐inheritance and thus certain combinations co‐exist on the same allele more frequently than would be expected by chance.

In 2019, Nagelkerke et al. published a comprehensive overview of FCGR variant frequency and LD patterns, based on data from over 1800 healthy individuals from African, Chinese, and European populations. ⁹⁸ Significant population differences were found in the frequencies of all genotyped SNPs and CNRs, except for the FCGR3A‐158V variant, which was largely similar across groups. Within the European and African populations, further analysis revealed subtle distinctions among Europeans and notable differences between individuals of Antillean, Ethiopian, South African, Surinamese, and West African descent.

A key finding across all populations was the difference in FCGR2C haplotype prevalence. For instance, the nonclassical FCGR2C‐ORF haplotype was more common in African populations than in Europeans, while the classical FCGR2C‐ORF variants were present in 11% of European alleles but were uncommon in those of African descent and almost entirely absent in the Chinese population. When examining the LD landscape in a CNR‐neutral context, the classical FCGR2C‐ORF variants were shown to be in LD with numerous other FCGR SNPs. In Europeans, they were almost completely in LD with the FCGR2C 2B.2 promoter haplotype, in strong LD with the FCGR2A‐27W and FCGR2B 2B.4 variants, and in weak LD with the higher affinity FCGR3A‐158V SNP. Similar, albeit weaker, LD relationships were also identified in the African and Chinese cohorts carrying the classical FCGR2C‐ORF haplotype.

These findings ⁹⁸ were largely in agreement with previously reported FCGR variant frequencies and LD patterns, ³⁴ , ¹¹⁰ , ¹⁶⁰ , ¹⁸⁴ , ²⁰⁵ , ²⁰⁶ yet expanded the range of variants and ancestry‐specific effects considered. For example, the LD detected between the high‐affinity FCGR2A‐131H and FCGR3A‐158V alleles were consistent with previous conclusions made in African and European cohorts, ¹⁸⁴ , ²⁰⁵ yet were actually inverted in the Chinese population with LD between the higher affinity FCGR2A‐131H and the lower affinity FCGR3A‐158F alleles.

LD was also explored between the SNPs and CNV, revealing that CNR1 was in strong LD with the nonclassical FCGR2C‐ORF haplotype in both European and African populations, and with the FCGR3B‐SH haplotype in Europeans. FCGR2A‐H131R, FCGR3A‐V158F, and FCGR2B‐I232T also displayed association with CNR1 CN changes. LD was only observed for CNR2 between FCGR2B‐I232T in Europeans, and no statistically significant LD was found for CNR3.

While this research provided insight into LD and population variation, the field still lacks an unbiased understanding of the low‐affinity FCGR locus as only previously studied and confirmed as functionally relevant variants were genotyped. The MLPA assay, which is widely used and regarded as the most effective high‐throughput method for determining FCGR variants, ²⁰⁷ only examines eight of the previously detailed SNP/haplotypes (along with CNR1‐4) which, although important as associated with autoimmune/infectious diseases, does not offer a complete picture. For instance, the FCGR3A‐48 L/R/H polymorphic site described above was not assessed but is a good example of why detailed knowledge of all FCGR LD is critical for accurate interpretation. The triallelic FCGR3A variant was first recognized in 1996 where it was observed that FCGR3A‐48R and FCGR3A‐48H variants had higher binding capacity for hIgG compared with the more prevalent wildtype FCGR3A‐48L allele. ²⁰⁸ However, this difference in hIgG binding was quickly ascribed solely to FCGR3A‐158F/V alleles as a consequence of LD. ²⁰⁹ In 1997 the same authors demonstrated a distinct linkage between the lower affinity FCGR3A‐158F and FCGR3A‐48L wildtype alleles, and that NK cells with the FCGR3A‐158V variant bound significantly more IgG irrespective of the FCGR3A‐48L/H/R genotype. ²⁰⁹ However, this claim was challenged when revisited in 2014 and FcγRIIIa expression was appropriately controlled for. ¹⁴⁵ It was shown that although FCGR3A‐158V is strongly linked with both FCGR3A‐48H and R variants, the FCGR3A‐48L/H/R genotype can independently influence IgG affinity. Specifically, the 158V/48R haplotype exhibited the highest IgG‐binding affinity, and both the 158V/48R and 158V/48H haplotypes displayed significantly greater binding capacity compared with the 158V/48L combination. Given that both the biallelic FCGR3A‐158F/V and the triallelic FCGR3A‐48L/R/H polymorphic positions can independently influence FcγRIIIa hIgG binding, alongside alterations in expression levels resulting from CNR linkage and the remaining FCGR3A‐IH variants, it is imperative that appropriate genetic controls and stratified groups are implemented to incorporate all known factors.

7. INFLUENCE OF FCGR GENETICS ON COVID‐19

In addition to the multiple associations that the FCGR genes have with autoimmune diseases, they also have important roles in regulating responses to infection. Most recently this has been demonstrated in their influence on SARS‐CoV‐2 control and disease outcome, ²¹⁰ , ²¹¹ , ²¹² whereby failure to coordinate efficient binding to SARS‐CoV‐2‐specific Abs and subsequent effector functions has been correlated with increased mortality. ²¹³ Considering the growing concerns about SARS‐CoV‐2 variants impacting vaccine efficacy by evading antibody neutralization, there is an expanding interest in maintaining and augmenting protection mediated by FcyR effector functions. ²¹⁴ , ²¹⁵ , ²¹⁶ , ²¹⁷ While it has been reported that FcyR effector activities, such as viral uptake into phagocytic cells and increased inflammation, may also heighten the pathophysiology of the virus through antibody‐dependent enhancement of infection (ADE), ²¹⁸ , ²¹⁹ , ²²⁰ there is currently no evidence from preclinical or clinical studies that SARS‐CoV‐2 vaccines elevate the risk of ADE. ²²¹ , ²²² , ²²³ Furthermore, a single‐cell RNA‐seq analysis of immune cells identified a distinct absence of global induction of the interferon‐stimulated genes (ISGs) program in severe COVID‐19 patients, in contrast to its presence in mild cases, which was attributed to FcyRIIb‐mediated antagonism of interferon receptor signaling. ²²⁴ This altered myeloid transcriptional state, which serves as a COVID‐19 severity hallmark, ²²⁵ implies the clinical blockade of FcyRIIb with antibodies may be a potential means to reverse the inhibition of ISGs and diminish severe responses. ²²⁴

Among factors such as antibody isotype, glycosylation, and vaccine regimen, FCGR polymorphisms have also been reported to influence host immunological response to SARS‐CoV‐2. For example, the homozygous FCGR3A‐158V/V genotype was shown to be overrepresented in COVID‐19 patients that were hospitalized or subsequently died, whereby NK cells expressing the higher affinity (FCGR3A‐158V/V or V/F) demonstrated significantly higher pro‐inflammatory ADCC responses that may have contributed to immunopathogenesis. ²²⁶ However, another study by López‐Martínez et al. evaluated the impact of both FCGR2A‐131H/R and FCGR3A‐158F/V SNPs on COVID‐19 severity, only found the FCGR2A‐131R allele to be significantly associated with an increased risk of mortality. ⁹⁶ This polymorphism, encoding the FcγRIIa isotype with lower IgG affinity, had also previously been associated with more severe SARS‐CoV‐1 infection. ²²⁷

8. IMPACT OF FCGR GENOTYPES ON IMMUNOTHERAPY

It has been proposed that FCGR genotyping can be leveraged for precision medicine in predicting response to mAb immunotherapy. The two FCGR variants that have gained the most attention in this regard are the FCGR2A‐131H/R and FCGR3A‐158F/V SNPs which alter FcγR hIgG affinity as detailed above. In the autoimmune disease setting, the FCGR3A‐158V variant has been associated with improved rituximab response in ITP, ²²⁸ SLE ²²⁹ , ²³⁰ and RA. ²³¹ , ²³² , ²³³ , ²³⁴ , ²³⁵ Similarly, the FCGR2A‐131R variant was associated with favorable responses to adalimumab in RA ²³⁶ , ²³⁷ and with better outcomes for liver transplantation patients treated with rituximab. ²³⁸

Similar associations with efficacy were previously made in the treatment of malignancies. However, as the influence of FCGR variants on mAb efficacy has been more extensively investigated, the literature has become less clear and in places directly conflicting. In part, this discordance may reflect both the nature of the studies undertaken (size, real‐world versus clinical trial etc.), as well as the burgeoning number and type of mAb immunotherapy agents and greater diversity in their mechanisms of action.

With respect to direct targeting or cytotoxic antibodies, it is well established that the FcγRs are key determinants of success, whereby FcγR expression profile (A/I ratio), mAb isotype, and FcγR interaction influence therapeutic efficacy. ³ , ²³⁹ , ²⁴⁰ Here, the primary effector mechanisms are thought to be ADCC and antibody‐dependent cellular phagocytosis (ADCP).

Initial studies supported that the alleles encoding the high‐affinity alleles FCGR2A‐131H and FCGR3A‐158V significantly augment cancer patient responses and outcomes to mAb‐based regimens. Firstly, in lymphomas treated with rituximab, both genotypes have independently been associated with higher response rates and longer progression‐free survival (PFS) in follicular lymphoma (FL) (Weng and Levy 2003) and the FCGR3A‐158V variant has been associated with favorable clinical response rates in FL, ²⁴¹ , ²⁴² mantle cell lymphoma (MCL), ²⁴³ diffuse large B‐cell lymphoma (DLBCL), ²⁴⁴ , ²⁴⁵ and non‐Hodgkin lymphoma (NHL) in general. ²⁴⁶ These latter findings supported the greater ADCC activity seen for NK cells with the FCGR3A‐158V variant, with the mechanism for the FCGR2A‐131H being less apparent (for hIgG1 therapeutics for which the affinity is relatively similar to FCGR2A‐131R).

Among solid tumors, these findings were extended to trastuzumab in breast ²⁴⁷ , ²⁴⁸ , ²⁴⁹ , ²⁵⁰ and gastric ²⁵¹ cancers, farletuzumab for ovarian cancer, ²⁵² as well as cetuximab for colorectal cancer ²⁵³ , ²⁵⁴ , ²⁵⁵ , ²⁵⁶ , ²⁵⁷ , ²⁵⁸ , ²⁵⁹ and head and neck squamous cell carcinoma (HNSCC). ²⁶⁰ Of interest, in renal cell carcinoma patients the high‐affinity genotypes (plus the FCGR2C‐57Q variant) were associated with significantly increased tumor shrinkage and overall survival in response to a non‐antibody immune therapeutic, high‐dose aldesleukin (HD‐IL2). ²⁶¹

However, several other studies have cast doubt on the prognostic significance of these variants, finding no association between high‐affinity genotypes and clinical outcomes in colorectal cancer, ²⁵³ , ²⁵⁵ , ²⁶² , ²⁶³ FL, ²⁶⁴ , ²⁶⁵ , ²⁶⁶ DLBCL, ²⁶⁷ , ²⁶⁸ , ²⁶⁹ and CLL. ²⁷⁰ , ²⁷¹ Moreover, beyond the relatively small‐scale, retrospective studies, the lack of association with clinical impact was also reproduced in studies of clinical trials with much larger cohorts for FL, ²⁷² , ²⁷³ , ²⁷⁴ DLBCL, ²⁷⁴ , ²⁷⁵ and renal cell carcinoma. ²⁷⁶

To add to the complexity, there are also a limited number of studies concluding the lower affinity genotypes are associated with better responses to mAb immunotherapy, including within colorectal cancer, ²⁶² , ²⁷⁷ , ²⁷⁸ , ²⁷⁹ DLBCL, ²⁷⁵ multiple myeloma, ²⁸⁰ and breast cancer. ²⁸¹

Due to these highly variable findings, the utility of FCGR polymorphisms as predictive biomarkers remains inconclusive. The discordance between studies may be attributed to numerous factors, involving differences in disease subset, clonal evolution, treatment regimen, study design, genotyping technique, and sample sizes. Furthermore, a substantial limitation of the aforementioned studies is the majority typically only genotype the two SNPs FCGR2A‐131H/R and FCGR3A‐158F/V—completely overlooking CNV and other FCGR SNPs (which they may be in LD with) that also have the potential for impacting treatment response as detailed above.

Beyond direct targeting mAbs, immunomodulatory mAb therapies—such as agonist antibodies and checkpoint inhibitors—are also impacted by FcγR engagement. ²⁸² , ²⁸³ In preclinical mouse models, therapeutic efficacy of mAbs directed to immunomodulatory receptors such as GITR, OX40, CD25, and CTLA4 in part are reported to stem from their ability to eliminate intratumoral T regulatory cells (Tregs), thereby unleashing antitumor immunity driven by CD8 T cells. ²⁸⁴ , ²⁸⁵ , ²⁸⁶ , ²⁸⁷ , ²⁸⁸ , ²⁸⁹ In advanced melanoma patients, the FCGR3A‐158V variant has been linked to improved responses to the anti‐CTLA4 mAb ipilimumab. ²⁹⁰ The FCGR3A‐158V SNP was found to be associated with significantly improved overall survival in one dataset, ²⁹¹ while a meta‐analysis also reported an association with higher response rates, ²⁹⁰ although it should be noted both studies are small and require further validation.

9. ROLE OF FCGR2B IN CANCER AND ASSOCIATIONS WITH TREATMENT RESISTANCE

Although mAb immunotherapy has revolutionized cancer treatment, ²⁴⁰ patients often experience intrinsic or acquired resistance, thus relapse is still a significant obstacle. The mechanisms of action and hence resistance vary with different types of therapeutic antibody and are incompletely resolved, but deeper understanding is central to improved clinical outcomes. ²⁸² B‐cell malignancies are routinely treated with anti‐CD20 mAbs such as rituximab, which engage FcγR‐mediated effector functions to elicit their therapeutic effects. ²⁹² One mechanism of resistance to rituximab involves high FCGR2B expression. ²⁹³ , ²⁹⁴ , ²⁹⁵ Being the predominant IgG receptor expressed on normal and malignant B cells, FcγRIIb facilitates internalization of CD20:rituximab:FcγRIIb complexes through a cis configuration resulting in loss of antibody from the cell surface, heightened mAb consumption, and abrogation of all FcγR‐mediated effector functions. ²⁹⁶ , ²⁹⁷ Retrospective analyses found that MCL patients treated with rituximab immunochemotherapy had shorter PFS when their tumor biopsies were FcγRIIb‐positive, ²⁹³ with high FcγRIIb expression on FL also associated with lower response rates with rituximab monotherapy. ²⁹⁴

Further support for the negative association of FcγRIIb and treatment success arises from genetic studies. Investigations into genes impacted by 1q21‐23 rearrangements, a common aberration in hematological malignancies associated with poor prognosis, ²⁹⁸ , ²⁹⁹ identified FCGR2B as a gene targeted by the rearrangement leading to deregulated expression. ³⁰⁰ , ³⁰¹ Despite involving different chromosomal partners, the translocations examined in the separate studies resulted in overexpression of the FcγRIIb2 isoform, relative to the expected FcγRIIb1 isoform, on the tumor B cells. The overexpression of FcγRIIb2, the isoform that more readily undergoes internalization, potentially compromises the regulatory function of FcγRIIb, leading to enhanced evasion of antitumor mechanisms and may represent a secondary genetic event that provides additional growth advantage. ³⁰¹ FcγRIIb2 was also identified as enriched relative to FcγRIIb1 in CLL cells, ²⁹³ although signaling through the intracellular tail was shown to be redundant for the rituximab internalization process. ³⁰²

A recent study also identified FCGR2B as a potential novel oncogene in a germinal center B‐cell (GCB) subgroup of DLBCL patients. ³⁰³ Somatic focal amplifications were identified that deregulated expression, resulting in significantly elevated FCGR2B levels that were associated with inferior response to rituximab‐containing immunochemotherapy. ³⁰³ This study also observed increased frequencies of somatic SNVs and indels within FCGR2B introns, and suggested these alterations may result in a truncated protein by facilitating intron retention. ³⁰³ Given the numerous preclinical studies ²⁹³ , ²⁹⁶ , ²⁹⁷ , ³⁰⁴ and smaller clinical trials ²⁹³ , ²⁹⁴ indicating high FCGR2B expression on tumor cells adversely affects rituximab‐based regimens, a series of large clinical cohorts were examined to investigate whether elevated FCGR2B levels in DLBCL imparts resistance to anti‐CD20 mAbs. ²⁹⁵ Results demonstrated that high expression of FCGR2B serves as a robust, independent prognostic biomarker that is associated with shorter PFS in patients treated with rituximab immunochemotherapy, but not with obinutuzumab immunochemotherapy. ²⁹⁵ Obinutuzumab does not internalize rapidly as it is a Type II anti‐CD20 mAb and is less impacted by FcγRIIb, supporting the earlier preclinical observations. ²⁹³ , ²⁹⁷ It was proposed, pending additional validation, that FCGR2B expression could be utilized to identify patients at an early stage who would likely derive the greatest benefit from obinutuzumab‐based treatment. Taken together, enhanced internalization capability—whether through the overexpression of the FcγRIIb2 isoform, or the ability of type I mAbs like rituximab to cluster and reorganize CD20 into lipid rafts—can diminish the availability of the mAb for its therapeutic action, potentially leading to poorer clinical outcomes. Of relevance, the absence of FcγRIIb‐mediated internalization of the type II mAb obinutuzumab was shown to enhance the phagocytosis of CLL target cells. ³⁰⁵

It has also been shown that non‐hematopoietic tumors can ectopically express FcγRIIb, with one group demonstrating 40% of metastatic melanoma samples expressed the “B‐cell” FcγRIIb1 isoform. ³⁰⁶ , ³⁰⁷ , ³⁰⁸ This suggests that ectopic expression of FcγRIIb may represent a progression factor that behaves as a decoy receptor to enable escape from FcγR‐dependent effector mechanisms in the presence of antitumor mAbs. ³⁰⁹ , ³¹⁰ Utilizing murine in vitro and in vivo models, FcγRIIb1 was shown to enhance tumor formation and progression when ectopically expressed on nonlymphoid cells. ³¹¹ Additionally, it has been shown that FcγRIIb expression is significantly increased on monocytes under high density, ³¹² and on both monocytes and macrophages during hypoxia. ¹²⁹ Hypoxic conditions frequently occur in specific tissue niches, like bone marrow, ³¹³ and in poorly vascularized tumor microenvironments typical of advanced cancers. Hypoxia results in FcγRIIb upregulation in these areas, potentially limiting the effectiveness of direct targeting antibodies such as rituximab. ¹²⁹ Upregulation of FCGR2B has also been observed in macrophages that reside in the adipose tissue of breast cancer, ³¹⁴ as well as in macrophages and monocytes of colorectal‐, non‐small‐cell lung‐ and renal cell carcinomas with single‐cell RNA sequencing. ⁶⁵

Preclinical studies have demonstrated that high FcγRIIb expression within the tumor microenvironment also hinders the depletion of intratumoral Tregs by anti‐CD25 mAb, thereby reducing its effectiveness against established tumors. ³¹⁵ However, the therapeutic efficacy of anti‐CD25 mAbs can be restored by augmenting their binding to the activating FcγRs, through isotype switching to isotypes with higher A:I ratios, such as mIgG2a. Accordingly, anti‐CD25 mIgG2a overcame the high FcγRIIb expression in the TME to effectively deplete tumor‐infiltrating Tregs and enhance the control of established tumors. ³¹⁵

FCGR2B was also observed to be upregulated on a subset of highly differentiated, tumor‐infiltrating effector CD8+ T cells in both a melanoma mouse model and patients with metastatic melanoma. ³¹⁶ This upregulation was associated with an activated yet potentially exhausted CD8+ T‐cell phenotype and played a suppressive role in antitumor immunity. It was subsequently shown that FCGR2B mediated suppression of these CD8+ T cells, limiting the efficacy of mAb to both checkpoint inhibitors PD‐1 and CTLA‐4. ³¹⁷

10. REGULATION OF THE LOW‐AFFINITY FCGR LOCUS

Understanding how the genetic sequence, including variants, influences the regulation of FCGR expression has proven to be highly challenging, with limited research uncovering the underlying mechanisms.

Studies have demonstrated that genetic variants in the promoter and intronic regions of FCGR2B modulate TF binding and expression levels. A genetically determined down‐regulation of FcγRIIB1 in GCBs was linked to increased hyper‐IgG and IgG autoantibodies. ³¹⁸ Two substitutions and two deletion sites were identified within the promoter region that was thought to result in lower FcγRIIB1 expression and subsequently provided a mechanism by which B cells escape negative regulatory signals. ³¹⁸ Polymorphisms within the third intron were also identified which correlated with increased IgG antibody responses and are shared among autoimmune‐prone mouse strains. ³¹⁹ , ³²⁰ Two unmethylated regions in the promoter and third intron have been characterized for their activities in the control of mFcγRIIb gene transcription in a cell type‐specific manner. ³²¹ These polymorphisms were reported to be involved in impairing the binding of the TF Activator protein 1 (AP‐1) and subsequently leading to enhanced germinal center formation and excessive autoantibody production, ³²⁰ , ³²² and it was suggested similar alterations in FCGR2B regulatory regions could have implications in human autoimmune susceptibility.

The aforementioned upregulation of FCGR2B under hypoxic conditions, ³²³ , ³²⁴ which subsequently diminishes the FcγR A:I ratio and impairs phagocytosis of mAb‐opsonized targets was shown to be transcriptionally driven. ¹²⁹ TFs AP‐1 and members of the hypoxia‐inducible factor (HIF) family, HIF‐1α and HIF‐2α, were identified as critical upstream regulators of this increase in expression. Early research showed AP‐1 to bind a non‐canonical motif in the FCGR2B promoter region in cell lines, ³²⁵ which was supported by a more recent study that utilized primary material to confirm binding at a non‐canonical AP‐1 motif 339‐bp upstream of FCGR2B TSS. ¹²⁹ Furthermore, nine hypoxia response elements (HREs) were identified within 15‐kb upstream of FCGR2B TSS, with the nearest canonical HRE motif at position −3916 upstream of TSS. ¹²⁹ During investigations of how to recalibrate the immunosuppressive FcyR signature of tumor‐associated macrophages, it was shown how activation of stimulator of interferon genes (STING) results in an increase in activating receptors FcyRIIa/FcyRIII while simultaneously decreasing expression of FcyRIIb. ³²⁶ An equivalent increase in the A/I ratio in mice (increased FcyRIII/IV and decreased FcyRII) was also linked to a significant enhancement in mAb‐mediated depletion of malignant targets in the face of tumor‐mediated FcyR A/I suppression. ³²⁶

Of the three polymorphic positions that make up the FCGR3A‐IH—which is overrepresented in individuals with elevated FcγRIIIa expression on NK cells and monocytes ¹⁰⁶ —the 5′‐UTR ‐1405G>A SNP is reported to be situated within a ~200‐bp putative silencer region. It has been speculated that the SNP promotes increased expression by disrupting repressor binding to the silencer element. ¹⁰⁶ While further validation is required, luciferase reporter assays demonstrated deletion of this region resulted in more than a twofold increase in activity. ¹⁰⁸

One study has highlighted the potential role of FcγRIIa epigenetic regulation in KD; five key CpG sites within the FCGR2A promoter were associated with an increased risk of disease susceptibility and pathogenesis, with significant hypomethylation observed within the KD cohort compared with the control group. ³²⁷ Furthermore, all five sites exhibited lower methylation levels in patients resistant to the intravenous immunoglobulin (IVIg) treatment relative to those who were responsive to IVIg. ³²⁷ The KD cohort also demonstrated markedly elevated levels of FCGR2A mRNA, and reporter gene assays revealed that the five CpG sites were sufficient to regulate expression. ³²⁷ A follow‐up study found FCGR2A expression levels were also significantly higher in IVIg‐resistant KD patients, and those with coronary artery lesions (CALs) compared with those without CALs. ³²⁸

11. LIMITATIONS IN FCGR RESEARCH AND BENEFITS OF LONG‐RANGE GENOMICS

11.1. Current limitations

Investigation of the FCGRs has been hindered by the high intra‐region homology (Figure 2B–D), resulting in technical difficulties that lead to ambiguity and inconsistency of results. Genetic association studies involving FCGR variants are often unreliable due to technical variation, inconsistent nomenclature, high LD, and genetic complexity from segmental duplication. ³²⁹ , ³³⁰

As traditional methods for studying the FCGRs, including MLPA, Taqman genotyping, Haloplex, and whole exome/genome sequencing (WES/WGS), produce short units of data and heavily depend on the current human genome assembly, their capacity to characterize this highly homologous region is significantly restricted and can result in ambiguous data. ³² Within the coding region of FCGR2B, there have been further nonsynonymous variants detected, but none have had their biological relevance, functional validation, or involvement in disease pathogenesis ascertained as yet. ⁸ For instance, a stop codon within exon 3 (rs10917661) has been associated with SLE and ankylosing spondylitis ¹⁸³ , ³³¹ but due to the sequence homology, primers cannot distinguish between FCGR2C and FCGR2B. Therefore, it remains to be determined whether studies are reporting on a stop codon in FCGR2B or the already established Q57X variant within FCGR2C. Discrepancies in reports of the consequences of FCGR2B promoter haplotypes further illustrate these challenges. ¹²⁵ Due to genotyping of the low‐affinity FCGR locus being complex, it was noted that commonly used databases (such as NCBI BLAST and Ensembl) are not always accurate in the distinguishing between FCGR SNPs and genuine PSVs. ³³⁰

Caution has also been recommended regarding ascertaining the true abundance of FCGR structural variants (SVs) when using the commonly used targeted techniques. For example, the frequency of FCGR3B CNV has previously been reported to range from 5% to 40% when using SYBR Green qPCR but with MLPA/PRT assays it has been defined as lower and much more consistent at 5%–8%. ¹²¹ , ¹⁸⁴ , ¹⁹² , ¹⁹⁴ , ¹⁹⁶ , ²⁴⁰ , ³³² A meta‐analysis also reported that data across numerous studies that describe FCGR3B CNV were confounded by the type of experimental methodology that was utilized to assess copy number. ¹¹⁵ However, one recent study compared qPCR, digital PCR (dPCR), and droplet digital PCR (ddPCR) for determining FCGR3B CNV and found complete concordance among the three platforms, including in rare cases with zero and four copies. ³³³ While dPCR generally offers higher sensitivity and precision, ³³⁴ this study concluded all three methods can quantify FCGR3B CN and claimed qPCR is still a viable option. ³³³ Further highlighting the difficulties of accurately characterizing variation at the FCGR locus, an optimization and then evaluation of genomic analysis demonstrated that even MLPA and PRT assays produce discordant results when performed on the exact same cases with nearly 9% conflicting. ²⁰⁷

These discrepancies then have serious consequences when interpreting findings in downstream analyses; for instance, qPCR has linked both low and high FCGR3B copy numbers with the incidence of SLE, ²⁰⁰ while a triple PRT assay identified only the loss of FCGR3B as a contributing factor. ¹⁸⁵ Furthermore, with qPCR only, European SLE cohorts were documented to have a higher frequency of low FCGR3B copy number patients and no association was shown in different Chinese cohorts ¹⁹⁶ , ³³⁵ and yet the PRT assay found a strong association with SLE in individuals from Hong Kong. ¹⁸⁵ Similarly, low FCGR3B has previously been associated with vasculitis ¹²¹ but this has also been shown not to be the case when different groups used the same ¹⁹⁶ or different methodology. ¹⁸⁴ A recent study described a method that takes advantage of the high‐throughput data generated from whole‐genome aCGH. ¹¹⁴ However, it was only able to distinguish heterozygous deletions that involved FCGR3A or FCGR3B with reasonable accuracy and thus could not differentiate between CNR2 and CNR3, or CNR1 and CNR4. An effort has also been made to use intensity values from an Immunochip platform assessing 18,039 individuals to define CNV. This study concluded it could identify 0, 1, and 2 copies but suspicion has been expressed on its reliability due to (i) not validating the findings with the “gold‐standard” MLPA/PRT techniques ³²⁹ and (ii) not being able to demonstrate a relationship between FCGR3B CNV and FcγRIIIb expression (Nagelkerke et al. 2019a). ⁹⁸

One significant gap in knowledge within FCGR genomics is determining precise CNR breakpoints. Sequence homology confounds CNV analysis, making it notoriously difficult to ascertain whether the event is taking place on the proximal or distal repeat unit; impeding the accurate characterization of molecular events and their functional consequences. ¹⁵⁹ Many of the aforementioned studies utilize PSVs to differentiate the first paralogous block from the second—regardless of whether this is for assays involving qPCR, PRT, or MLPA. Relying on these PSVs means these technologies are limited due to their scarcity and ambiguity—that is, PSVs are infrequent and can often be polymorphic—leading to uncertainty in the origin of their products and thus impeding the precise localization of CNR boundaries. ³⁴ , ¹⁵⁹

FCGR analyses using high‐throughput short‐read sequencing are also significantly restricted as they face issues with multimapping. This issue is demonstrated in Figure 3 (light orange track) by the non‐uniform coverage of alignments for 148‐bp paired‐end genomic sequencing. The fact the four CNRs were only described in detail relatively recently is not too surprising considering that the novel chimeric events are undetectable with high‐throughput short‐read sequencing methods. ⁴⁸ The FCGRs high sequence identity prevents confident alignment thus WES is inadequate for detecting CNRs. It has also been speculated that unexplained cases of early onset severe autoimmunity may result from homozygous deletion of CNR4 because the FCGR2B‐stop variant would not be detected by NGS techniques as all FCGR2B‐stop reads would be mapped to the FCGR2C gene instead. ⁴⁸

In order to fully understand the clinico‐biological importance of the FCGRs and their variants in both normal and disease settings, it is crucial to have a high‐quality reference genome for any alignment‐based analyses. However, this is not the case for the FCGR locus and in the first ever human genome reference there was simply an empty gap where the FCGR genes reside. ³³⁶ Moreover, the hg16 reference genome released in 2003 depicted only four of the low‐affinity FCGR genes in a region of just 90 kb, representing only one of the paralogous blocks. Since then, despite all the improvements to sequencing technologies, assembly algorithms, and updates to create the current GRCh38 reference genome, there are still deficiencies and uncertainties in accurate genome builds for the FCGR region. ³² For example, the FCGR2C‐Q57X variant was initially annotated to be in the FCGR2B gene instead of FCGR2C in GRCh38. Following this, the bacterial artificial chromosomes (BACs) used to build the initial human reference have been sequenced with newer technologies and it has been determined that the paralogous region upstream of FCGR2C had been misassembled and swapped with the homologous region on FCGR2B. ³⁰³

Some of the greatest technical challenges with short‐read sequencing technologies are caused by repetitive DNA: sequences that are identical or similar to other regions in the genome, including segmental duplications. ³³⁷ , ³³⁸ Accordingly, it has been demonstrated that more than 70% of structural variation in the human genome could be undetectable with reads less than 300 bp in length. ³³⁹ Consequently in highly homologous regions like the FCGR region where the two paralogous blocks have been reported to share >95% homology, ³⁵ issues are encountered when trying to accurately align or assemble the reads. The repetitive nature creates technical challenges for the alignment and assembly software, leading to difficulties in conclusively resolving which genomic location a read originated from and introduces ambiguities from a computational perspective in the form of biases and errors when interpreting the results. As demonstrated in Figure 3 (black and orange tracks), short reads often have uneven coverage, regions where reads do not align, or do not map uniquely or with high quality. This causes issues with downstream analysis, especially when trying to study regulation with techniques that utilize very short reads, such as the 25‐75‐bp read length that is common in ChIP or ATAC‐seq studies. However, the rise of third‐generation sequencing technologies that produce long reads may offer solutions to these problems, potentially advancing our understanding of the genomic structure and regulation of the FCGRs.

11.1.1. The potential benefits of increased sequence read length

For more than a decade, high‐throughput short‐read sequencing has dominated genomics research with wide‐ranging applications, enriching our understanding of evolution and disease through the study of SNPs, CNV, and indels. ³⁴⁰ , ³⁴¹ , ³⁴² Its high throughput has also facilitated the study of a multitude of other biological phenomena, including gene expression, chromatin accessibility, and TF binding. Yet regardless of the ability to produce highly accurate (>99.9%) sequencing reads, the sequence‐by‐synthesis process results in a limited read length of only 50–300 bp, ³⁴³ leading to the issues described above. Therefore, alternative, complementary methods, mainly involving long‐read approaches such as nanopore‐ and single‐molecule real‐time (SMRT) sequencing‐based technologies, have been developed.

Nanopore sequencing generates reads that typically range from one to >100 kilobases (kb) ³⁴⁴ , ³⁴⁵ , ³⁴⁶ and excels in producing ultra‐long reads, achieving up to 4.1 megabases (mb). ³⁴⁷ SMRT sequencing approaches can generate two different types of long‐read data that differ in both read length (5–60 kb) and accuracy (ranging from 87% to >99%). ³⁴⁸ Optical genome mapping (OGM) is an additional long‐range technology that can aid in detecting SVs. It generates 50–100‐mb genomic maps by de novo assembly of fluorescently labeled DNA molecules. Comprehensive SV analysis can be performed by inputting these molecules into haplotype‐aware assemblers and comparing the final consensus maps to the reference genome. However, unlike sequencing platforms, optical mapping cannot provide bp resolution and is limited to detecting SVs of at least 500 bp in length. ³⁴⁹

Human genetics research has already benefited from these advances in long‐read sequencing, improving genome assembly, variant discovery, disease association, and population genomics. ³⁵⁰ , ³⁵¹ , ³⁵² , ³⁵³ It has provided tools with the capacity to resolve some of the most complex, challenging, and dynamic regions of the human genome thus dramatically improving the understanding of human heritability, diversity, mutational mechanisms, and the genetic basis of disease.

In addition to the significant improvement to individual read lengths, long‐read sequencing can also result in a more even distribution of coverage across the genome as it is not as sensitive to imbalanced sequence composition—unlike short‐read platforms which often leave regions of high GC content with low or no read depth. ³⁵⁴ A systematic analysis showed that over 15mb of the human genome (from >35,000 different regions) are left in the “dark” when only using short‐read sequencing, hence preventing researchers from detecting disease‐relevant polymorphisms. ³⁵⁵ These dark regions, where short‐read data cannot be adequately aligned or assembled, can either be “dark by depth” with no or few mappable reads or “dark by mapping quality.” Sequences that are inherently challenging to sequence at the molecular level, like those with high GC content, give rise to dark by depth regions. In contrast, dark by mapping quality regions arise due to bioinformatic difficulties and ambiguities. This is particularly true of duplicated regions where mapping a short‐read to a unique position is impossible due to insufficient confidence, thereby creating “camouflaged” regions. Consequently, long‐read technologies have begun to resolve the sequence of difficult‐to‐analyze areas. A capture panel targeting medically relevant dark genomic regions for SMRT sequencing has been generated. ³⁵⁶ While it encompasses nearly 400 complete genes, only FCGR1A, FCGR2B, and FCGR3A are included with respect to the FCGR genes.

Now that read lengths are of a sufficient size to traverse the most complex and repetitive genomic structures, long‐read sequencing has also improved the de novo assembly of genomes, achieving greater contiguity compared with short‐read and Sanger‐based methods. ³⁴⁸ Moreover, combining complementary datasets generated with orthogonal technologies provides superior genome assembly. For example, the telomere‐to‐telomere (T2T) consortium utilized different long‐ and short‐read technologies in their effort to construct the very first, truly complete assembly of a human genome. The T2T project has so far released the complete X chromosome, ³⁴⁵ closed the gaps of the first autosome chromosome 8, ³⁵⁷ and presented the entire 3.06 billion bp sequence of all human chromosomes (except Y): resolving complex regions, improving variant calling accuracy, and introducing nearly 200 million more bp of sequence to be explored. ³⁵⁸ Building on the success of the T2T project, the Human Pangenome Reference Consortium (HPRC) was launched with the objective to construct complete T2T reference genomes for every population, striving to promote equitable healthcare for every patient. ³⁵⁹ This initiative is striving to address the lack of global representation in the current human reference genome—which despite being the most widely utilized resource in human genetics, is comprised of collapsed haplotypes from primarily a single individual. By employing a wide range of platforms, the HPRC aims to create high‐quality, contiguous diploid genomes that represent the entire scope of diverse human variation. ³⁶⁰

Long‐read sequencing has also advanced the phasing of genomic sequences, revealing insights into both maternal and paternal haplotypes. This enhances accurate variant detection and is particularly valuable for resolving SV, as differences between haplotypes in short‐read analyses can often result in collapsed or hybrid representations that do not reflect true haplotype architecture. ³⁶¹ , ³⁶² Long‐read technology enables the assembly of complete haplotypes and improves our understanding of the comprehensive spectrum of genomic variants. ³⁶³ , ³⁶⁴ , ³⁶⁵ For example, one study that implemented a multiplatform approach on the same individuals revealed that 46% of deletions and 78% of insertions went undetected by short‐read WGS, despite employing 11 different variant‐calling software tools. ³³⁹ Furthermore, despite impacting the greatest proportion of nucleotides within the genome than any other category of sequence variant, SVs have typically been understudied compared with SNVs. ³⁴² , ³⁶⁶ , ³⁶⁷ It has previously been reported that SV detection approaches using short reads perform poorly in some regions, demonstrating low sensitivity, ³⁶⁴ , ³⁶⁸ often misinterpreting complex SVs ³⁶⁹ and exhibiting high rates of false positives. ³⁶⁶ , ³⁷⁰ , ³⁷¹ Notably, the homology and complexity of the FCGR region have confounded accurate SV annotation, with long‐read technologies now making important progress as discussed below.

A particular strength of nanopore sequencing is its ability to analyze all nucleotides, including modified DNA and RNA bases. This allows for the detection of epigenetic modifications, with reports thus far of base‐callers trained to detect 4‐methylcytosine, 5‐methylcytosine, 5‐hydroxymethylcytosine, N6‐methyladenine, and 8‐oxoguanine. ³⁷² , ³⁷³ , ³⁷⁴ , ³⁷⁵ , ³⁷⁶ , ³⁷⁷ In addition to aiding in the study of allelic‐specific epigenetic variation, it has been used to study the distinctive epigenomic landscape and differential methylation of structurally variant regions associated with cancer. ³⁷⁸

11.1.2. The application of long‐read technologies to the FCGR locus

Long‐read sequencing now facilitates more accurate phased assemblies, enhanced mapping quality, and reduced bioinformatic uncertainties for the study of the FCGR locus. ³⁷⁹ As discussed above, previous analysis of the FCGR locus was significantly limited by short reads and alignment‐dependent methods, particularly due to the issues with the existing human reference genomes when assembling segmental duplications such as the FCGR region with short (<1000 bp) reads. However, while long‐read technologies are overcoming many of these issues aiding advanced variant detection and disease understanding, this has often only occurred subsequent to a high‐quality assembly of the region of interest. ³⁴⁸

To date, a comprehensive characterization of the low‐affinity FCGR locus that accurately depicts all variants has not been published. While four CNRs have been identified, short‐read methods have not precisely resolved their boundaries or CNV‐generating breakpoints. Moreover, it is known large segmental duplications encourage additional unequal cross‐overs between highly homologous sequences via further NAHR events ³⁸⁰ , ³⁸¹ , ³⁸² ; thus, novel CNV could exist that earlier short‐read, unphased approaches could not detect. The same limitations, regarding previously unexplored variants and their impact, also apply to FCGR SNPs—particularly those that are involved in expression regulation as nonsynonymous SNPs within exons have historically been the primary focus. For example, a 2018 study into the full‐length of FCGR3A with nanopore sequencing revealed it is more polymorphic than the literature had previously described with many novel variants residing in noncoding regions. ³⁸³ As shown in Figure 3, long‐read technologies are now beginning to overcome many of these issues, increasing coverage, depth, and resolution.

12. SUMMARY AND FUTURE DIRECTIONS

In summary, the FcγRs are crucial regulators of a well‐coordinated and effective immune response, influencing pathogen defense, susceptibility to autoimmunity, and the success of mAb therapies. Encoded by highly homologous genes shaped by evolutionary duplication events, the FCGRs exhibit significant genetic variation, including SNPs and SVs, which can impact their affinity, expression, function, and clinical relevance. The extensive polymorphic nature of the FCGR loci coupled with the high sequence similarity presents challenges in analysis. While some associations between genetic variants and disease have been identified, further investigation is necessary to address existing difficulties and confounding LD. Advancements in long‐read sequencing offer opportunities to overcome these challenges, providing a more comprehensive understanding of the genetic sequence, variants, regulation, and role in disease.

Due to the limitations of previously used technology, some caution should be exercised when examining historical genetic studies. Careful consideration is needed as much of the research does not account for the full spectrum of genetic variants or have adequately stratified controlled groups, potentially leading to incomplete or misleading conclusions. To accurately identify causative variants at the FCGR locus, comprehensive genotyping of all functional genetic variants is important. The necessity for accurate variant identification has led to multiple revisions of the MLPA assay since its introduction, allowing for the determination of independent risk markers through multiple logistic regression analysis. ⁹⁸ , ³³⁰ However, the challenges in genotyping this locus mean that some functionally relevant variants may still remain undetected and thus focus should not be limited to nonsynonymous SNPs as all variants, including those in noncoding regions, need to be considered as they can play a critical role in gene regulation, expression, and splicing. It has also been emphasized that selecting population‐matched control groups in genetic studies is vital for accurate association analyses ⁹⁸ as incongruence between control and patient populations can confound findings and obscure the true genetic contributions to disease.

The strong LD in the region also complicates efforts to determine which genetic variants contribute to specific traits or diseases. To effectively deconvolute these effects, studies should leverage large, well‐characterized cohorts. With larger sample sizes, it becomes possible to assess the influence of different variant combinations, helping to clarify their individual roles in disease risk or therapeutic outcomes. Given the strong LD, it could be more informative to report some of FCGR genetic variation as haplotypes rather than as individual variants, similar to the established approach used for the FCGR3B‐HNA SNPs. ⁹⁸ Future research also requires large cohorts to gain a comprehensive understanding of FCGR variation, including accurately distinguishing between paralogous and allelic variation, and uncovering any additional CNRs that have gone undiscovered.

The current genome reference sequence (GRCh38) is limited by its reliance on a small number of individuals, with around 70% of the sequence derived from just one donor. ³⁸⁴ , ³⁸⁵ This patchwork approach of collapsed haplotypes introduces biases and inaccuracies, particularly in regions that are difficult to assemble like the FCGR locus, and may not reflect the diversity of the general population, consequently impacting the reliability of variant identification and association with disease. ³⁸⁶ , ³⁸⁷ , ³⁸⁸ De novo assembly of the FCGRs across diverse populations promises to create more representative and accurate genomic references that will enhance the detection of novel variants and improve the precision of genetic studies by providing a more comprehensive view of population diversity and variation.

Expansive datasets, such as those provided by national and international Biobanks, are invaluable for unraveling the complexities of genetic contributions to phenotypes. These large‐scale resources not only facilitate the identification of novel genetic variants but also allow for a robust evaluation of previous associations, either confirming or challenging earlier findings with greater precision. Hujoel et al.'s analysis of the UK Biobank's extensive dataset, which includes WES for 470,000 participants, utilized haplotype‐informed CNV detection to link FCGR3B CNV with COPD susceptibility. ¹²² This approach underscores the importance of comprehensive large‐scale datasets in advancing our grasp of complex phenotypes and translating FCGR genetic findings into practical clinical applications.

In conclusion, to fully understand the genetic variation at the FCGR locus, future research must undertake exhaustive analyses of all relevant SNPs, CNRs, and their haplotypes. Such comprehensive studies are crucial for deciphering the complexities of disease associations and optimizing therapeutic strategies, especially in immunotherapy. Achieving these goals will depend on large, well‐characterized cohorts and diverse control groups alongside technology capable of long‐range analysis such as long‐read sequencing. Progress in understanding FcγR genetics could readily advance personalized medicine and we hope to see the development of more effective treatments tailored to individual genetic profiles in the future.

FUNDING INFORMATION

Funding was provided by Cancer Research UK program grants (award numbers: A24721; DRCRPG‐May23/100001) and Cancer Research UK and Experimental Cancer Centre awards C328/A25139 and C24563/A25171.

CONFLICT OF INTEREST STATEMENT

M.S.C. acts as a consultant for a number of biotech companies, being retained as a consultant for BioInvent International, and has received research funding from BioInvent, GSK, UCB, iTeos, and Roche.

Frampton S, Smith R, Ferson L, et al. Fc gamma receptors: Their evolution, genomic architecture, genetic variation, and impact on human disease. Immunol Rev. 2024;328:65‐97. doi: 10.1111/imr.13401

DATA AVAILABILITY STATEMENT

Data sharing not applicable—no new data is generated.