MGAT3 (N-acetylglucosaminyltransferase III, GnT-III)

MGAT3 catalyzes the addition of a bisecting N-acetylglucosamine (GlcNAc) — a single GlcNAc residue in β1,4 linkage to the central mannose of biantennary N-glycans. The bisecting GlcNAc is a branching modifier, not a new antenna: it does not extend a chain but instead sterically inhibits subsequent glycan modifications, most importantly fut8-mediated core fucosylation and MGAT5-mediated triantennary branching. Among the four core IgG Fc glycosyltransferases, MGAT3 carries one of the two strongest GWAS signals for IgG glycan variation (alongside st6gal1) and bisecting GlcNAc rises modestly with age — but it is the least aging-prominent of the four because the functional consequences of bisecting GlcNAc addition are smaller and less consistent than the galactosylation and sialylation shifts.

HGNC correction: The seeder brief listed HGNC 7044; the verified value from UniProt REST API is HGNC 7046. Frontmatter has been corrected.

Identity

UniProt: Q09327 (MGAT3_HUMAN) — confirmed via UniProt REST API 2026-05-20
NCBI Gene: 4248
HGNC: 7046 (verified; brief listed 7044 — corrected)
Ensembl: ENSG00000128268
Mouse ortholog: Mgat3
Length: 533 amino acids (canonical isoform)
EC number: 2.4.1.144

Enzymology

MGAT3 is a type II single-pass transmembrane glycoprotein localized to the Golgi apparatus (trans cisternae), colocalizing with st6gal1 and b4galt1 in cis-Golgi regions. It catalyzes:

UDP-GlcNAc + GlcNAc₂Man₃GlcNAc₂-Asn → 
  bisecting-GlcNAc(β1,4)-Man(GlcNAc₂Man₃GlcNAc₂)-Asn + UDP

The bisecting GlcNAc is added specifically to the β-mannose of the trimannosyl core — the central mannose — in β1,4 linkage. Key features:

Steric consequences of the bisecting GlcNAc:

Inhibits FUT8-mediated core fucosylation. The bisecting GlcNAc occupies a position that clashes with FUT8’s catalytic mechanism, rendering the same glycan inaccessible for core fucosylation. Therefore bisected and fucosylated glycoforms are mutually exclusive at the individual-glycan level (see igg-fc-glycosylation structural table).
Inhibits MGAT5-mediated branching. MGAT5 would add a β1,6-GlcNAc to the α1,6-linked Man arm, creating a triantennary (three-antenna) glycan; bisecting GlcNAc sterically prevents this. The MGAT3 vs MGAT5 competition thus determines whether IgG N-glycans become bisected biantennary or multi-antennary branched — with major consequences for lectin binding (galectins, selectins) and cell adhesion.
Alters Fc conformation slightly, modestly enhancing FcγRIIIA binding (independent of the fucosylation competition effect).

The UniProt annotation notes that MGAT3 “alters not only the composition, but also the conformation of the N-glycan.”

Position in the IgG Fc glycosylation cascade

MGAT3 acts in parallel with (and competitively with) fut8 at Step 3 of the Golgi processing cascade:

Step	Enzyme	Product
1–2	MGAT1, MGAT2	Biantennary complex-type scaffold
3a	fut8	Core fucosylation → G0F scaffold (blocks bisecting GlcNAc)
3b	mgat3	Bisecting GlcNAc → blocks FUT8 (mutually exclusive)
4	b4galt1	Galactosylation
5	st6gal1	Sialylation

The balance between Steps 3a and 3b determines the bisected (G0B, G0FB, G1FB, G2FB) vs fully fucosylated fraction of IgG.

Aging context

Bisecting GlcNAc abundance in serum IgG rises modestly with age in population studies ¹². The Leiden Longevity Study (n=1,967) found that decreased levels of bisecting GlcNAc-containing agalactosylated glycoforms were associated with longevity features in participants under 60, with the association weaker in those over 60 ¹. The biological interpretation is complex: bisecting GlcNAc partly inhibits core fucosylation (→ slight ADCC enhancement) but also alters Fc conformation and lectin-pathway interactions. The net effect on the pro-inflammatory IgG phenotype is smaller and less directionally clear than the G0 rise or sialylation loss.

MGAT3 is not regulated by estrogen and does not show the same inflammation-driven feedback loop as B4GALT1 and ST6GAL1. The hallmarks: field is therefore left empty — bisecting GlcNAc changes are an aging-associated glycome feature but not a primary inflammaging driver.

GWAS — the strongest IgG-specific signal

Despite its modest functional importance in aging biology, MGAT3 carries one of the two strongest GWAS signals for IgG glycan composition (alongside st6gal1) in both the Wahl 2018 and Landini 2022 studies ³⁴. Specifically, Landini 2022 demonstrates that MGAT3 (and ST6GAL1) are IgG-specific loci — they associate with IgG glycan variation but not transferrin glycan variation — confirming that substrate-specific genetic regulatory programs operate in the B cell / plasma cell secretory pathway for IgG ⁴.

The mr-causal-evidence field is partial: GWAS instruments at MGAT3 exist, but a dedicated Mendelian randomization study testing bisecting GlcNAc → aging/inflammatory outcomes via MGAT3 instruments has not been published as of 2026-05-20. needs-replication

Functional consequences of bisecting GlcNAc on effector function

The FcγRIIIA binding enhancement from bisecting GlcNAc is mechanistically linked to its inhibition of core fucosylation: glycans that cannot be fucosylated (because bisecting GlcNAc blocks FUT8) have a higher proportion of the afucosylated form, which binds FcγRIIIA more tightly ⁵. This was one basis for the early MGAT3-overexpression approach to engineering enhanced-ADCC antibodies — a strategy now largely superseded by FUT8 KO cell lines (see fut8). Stanley 2002 demonstrated MGAT3 overexpression enhances IgG effector functions in cell-based models ⁶.

Cancer biology — anti-metastatic context

MGAT3 has been studied extensively in cancer biology outside the IgG context. Bisecting GlcNAc inhibits MGAT5-mediated polylactosamino branching, which is associated with cancer-cell adhesion and metastasis (selectin-mediated extravasation). Multiple studies have found MGAT3 overexpression suppresses metastatic potential in cancer cell lines, consistent with its competitive inhibition of pro-metastatic MGAT5-branched glycan epitopes. Stanley 2002 in mice documented that MGAT3 overexpression retards liver tumor progression in chemically-induced models, while MGAT3 knockout did not show this protection ⁶. This is an important biological context — MGAT3 is simultaneously (i) a modest pro-inflammatory modifier of IgG Fc (via bisecting GlcNAc → slight FcγRIII enhancement in the aging context) and (ii) an anti-metastatic modifier in cancer contexts (via MGAT5 competition). These are not contradictory — they reflect different substrates and pathways. Brief mention only; this page focuses on the IgG-glycome-aging context.

Knockout phenotype (mouse)

Mgat3 KO mice (Stanley 2002; examining both KO and overexpression in C57BL6 background) are:

Viable and fertile — no gross developmental phenotype
Show retarded progression of diethylnitrosamine (DEN)-induced liver tumors — suggesting MGAT3’s bisecting GlcNAc addition normally supports (or is neutral on) liver tumor progression and its loss is partially protective ⁶
No immune phenotype specifically reported for the KO in this study

The mild KO phenotype (compare to the severe FUT8 and B4GALT1 KOs) reflects bisecting GlcNAc’s role as a modifier of glycan composition rather than an essential core structural element.

Pharmacological modulation

No clinical-grade MGAT3 modulator exists. Research-stage approaches include:

Competitive UDP-GlcNAc analog inhibitors (not advanced to clinical use)
Transfection-based MGAT3 overexpression in mAb-producing cell lines for enhanced ADCC (historical strategy; now largely superseded by FUT8 KO)

Druggability tier is set to 3 (research probes; therapeutic cell-line engineering demonstrated; no approved systemic MGAT3 modulator for any indication).

Limitations and gaps

The functional significance of the modest bisecting GlcNAc age-increase for IgG Fc effector biology in aging is not well-characterized relative to the galactosylation and sialylation shifts. needs-replication
Whether MGAT3 expression in B cells changes with age (and in what direction) has not been measured in dedicated studies; the serum glycome trajectory is inferred from population glycomics, not enzyme activity measurements. needs-replication
The molecular determinants of MGAT3 vs FUT8 competition in plasma cells (i.e., what sets the bisected fraction) are not known in vivo. no-mechanism
HGNC ID discrepancy (brief listed 7044; verified value 7046) has been corrected in this page.

Footnotes

doi:10.1371/journal.pone.0012566 · PMID 20830288 · observational (n=1,967; Leiden Longevity Study — 1,287 offspring + 680 partners of nonagenarian sibling pairs) · Ruhaak LR, Uh HW, Beekman M et al. · PLoS One 5:e12566 (2010) · decreased bisecting GlcNAc-containing agalactosylated glycoforms associated with familial longevity in participants <60 years; bisecting GlcNAc increases with age in galactosylated glycoforms; galactosylation decreases with age; sex differences documented · archive: downloaded ↩ ↩²
doi:10.1016/j.cellimm.2018.07.009 · PMID 30107893 · review · Gudelj I, Lauc G, Pezer M · Cell Immunol 2018 · IgG glycome aging review; bisecting GlcNAc modest age-increase context · archive: download failed no-fulltext-access ↩
doi:10.3389/fimmu.2018.00277 · PMID 29535710 · GWAS (~1,800 subjects; KORA + RS) · Wahl A, van den Akker E, Klaric L et al., Gieger C · Front Immunol 9:277 (2018) · replicated GWAS signals for IgG glycan variation at MGAT3, ST6GAL1, B4GALT1, FUT8; novel RUNX3 locus · archive: status pending ↩
doi:10.1038/s41467-022-29189-5 · PMID 35332118 · GWAS (n=2,020; CROATIA KORCULA n=948 + VIKING n=952) · Landini A, Trbojević-Akmačić I, Wilson JF, Klarić L et al. · Nat Commun 2022 · MGAT3 and ST6GAL1 are IgG-specific GWAS loci (strongest signals for IgG glycan variation); FUT8/FUT6 shared with transferrin; distinct genetic architectures for IgG vs transferrin glycome · archive: downloaded ↩ ↩²
doi:10.1074/jbc.m202069200 · in-vitro (Lec13 CHO cells + binding/ADCC assays) · Shields RL, Lai J, Keck R et al., Presta LG · J Biol Chem 277:26733–26740 (2002) · defucosylation enhances FcγRIIIA binding ~50-fold; mechanistic basis for MGAT3/FUT8 competition context · archive: downloaded ↩
doi:10.1016/s0304-4165(02)00404-x · PMID 12417419 · in-vivo (MGAT3 KO and overexpression mice) · Stanley P · Biochim Biophys Acta 1573:363–368 (2002) · bisecting GlcNAc enhances IgG effector functions; MGAT3 KO mice viable + fertile; MGAT3 KO shows retarded DEN-induced liver tumor progression; overexpression enhances IgG ADCC · archive: not_oa no-fulltext-access ↩ ↩² ↩³

🧬 Aging Wiki

Explorer

mgat3