The review, authored by Jae Young Kim of Gachon University, examines sialic acid molecules, the terminal sugars that decorate nearly every cell surface, as active regulators of innate immunity, blood-cell homeostasis, complement regulation, and thromboinflammation. The central argument is deceptively simple: deleting the genes that make pig-specific sugars, such as GGTA1, CMAH, and B4GALNT2, removes the major antibody targets but does not automatically produce a human-compatible glycome. When CMAH is knocked out, the pig loses N-glycolylneuraminic acid, or Neu5Gc, and its cell surfaces shift to a Neu5Ac-dominant profile. But the type of sialic acid is only one variable. The α2,3, α2,6, and α2,8 linkages that attach these sugars, the chemical modifications they carry, the glycoproteins and glycolipids they ride on, their surface density, and their spatial presentation all remain governed by the pig’s own glycosylation machinery, not by the human template.

This distinction matters because the immune system does not read sialic acid in isolation. A family of receptors called Siglecs, sialic acid-binding immunoglobulin-like lectins, regulates the activation thresholds of macrophages, neutrophils, and other innate immune cells. Most CD33-related Siglecs are inhibitory, recruiting SHP phosphatases through ITIM motifs to dampen inflammatory signaling. Siglec-1, by contrast, is a capture receptor without an inhibitory motif, primarily mediating adhesion and uptake of sialylated targets. The review emphasizes that which direction the interaction runs, donor receptor binding a recipient cell or a recipient receptor binding a donor cell, changes the entire interpretation of the experiment and its clinical relevance.

The best-established Siglec-dependent xenogeneic interaction runs in the graft-versus-host direction. Porcine Siglec-1, expressed on Kupffer cells and alveolar macrophages, captures Neu5Ac-rich human erythrocytes, particularly through α2,3-linked sialic acid motifs. Blocking studies showed that an anti-Siglec-1 antibody or sialidase treatment of human red blood cells markedly reduced this binding, and during extracorporeal pig liver perfusion with human blood, anti-Siglec-1 treatment partially prevented erythrocyte sequestration and the associated drop in hematocrit. Crucially, human red blood cells, being Neu5Ac-dominant, bind porcine Siglec-1 more strongly than red blood cells from Old World monkeys, which carry Neu5Gc. This means pig-to-baboon models may underestimate the erythrocyte sequestration that could occur in humans, a sobering caveat for the preclinical pipeline.

The reciprocal direction has recently gained direct molecular evidence. Recombinant human Siglec-1 binds CMAH-deficient gene-edited pig red blood cells, including DKO, TKO, and QKO cells, more strongly than it binds wild-type pig or human red blood cells. Binding was abolished by sialidase treatment and competed away by free Neu5Ac, confirming its sialic acid dependence. Intriguingly, although human red blood cells carry the highest total Neu5Ac content, they bind human Siglec-1 less avidly than the edited pig cells, a reminder that Siglec-1 does not count sialic acid molecules but prefers specific α2,3-linked motifs presented in multivalent clusters on particular carriers. Whether this receptor-level recognition translates into recipient macrophage phagocytosis and in vivo clearance of donor cells remains unresolved, and the review is careful to grade this pathway as moderate evidence rather than established causation.

Blood platelets tell a parallel story with a different receptor system. During ex vivo perfusion of pig livers with human blood, platelets are rapidly consumed through lectin-dependent recognition by liver sinusoidal endothelial cells and Kupffer cells. Deleting the donor asialoglycoprotein receptor ASGR1 reduces this consumption, providing one of the clearest demonstrations that a defined glycan-dependent interaction can cause clinically relevant thrombocytopenia. On platelets themselves, sialic acid regulates lifespan and activation: desialylation exposes terminal galactose residues that are cleared by the hepatic Ashwell-Morell receptor, and cis interactions between sialylated glycoproteins and platelet Siglec-9 appear to restrain activation. The review cautions, however, that the ASGR1 pathway demonstrated in the liver cannot be assumed to operate identically in heart or kidney xenografts, since glycan-receptor interactions are strongly organ- and cell-type specific.

Complement regulation adds another layer of complexity, and here the review identifies what may be its most consequential conceptual point: a fundamental functional trade-off in sialome engineering. Factor H, the key plasma regulator of the alternative complement pathway, protects host surfaces by recognizing self-associated markers, and its C-terminal domain preferentially binds α2,3-linked Neu5Ac while showing little affinity for α2,6 or α2,8 linkages. The same α2,3-linked sialoglycans that recruit protective Factor H are also the preferred ligands for Siglec-1-mediated capture. Conversely, increasing α2,6-linked sialylation, for example by overexpressing the sialyltransferase ST6GAL1, may reduce macrophage xenoreactivity and enhance engagement of inhibitory Siglecs, but at the cost of weakening Factor H recruitment. Neither direction is uniformly protective, and engineering one linkage class can simultaneously create benefit and harm.

The review extends this evidence-graded framework to other candidate pathways. The CD24-Siglec-10 axis, a checkpoint that restrains inflammatory responses to damage-associated molecular patterns released during ischemia and reperfusion, is biologically plausible but validated only in non-xenotransplantation settings. Siglec-9 on neutrophils suppresses oxidative burst and NETosis, the release of neutrophil extracellular traps that can amplify microvascular thrombosis, but its relevance to porcine graft surfaces has not been directly tested. The reviewer’s proposed solution is a hierarchy: glycan-receptor pathways should be evaluated by directionality, evidence depth, and functional trade-off rather than by ligand abundance or donor genotype alone. An evidence chain linking a glycan change in the actual target organ to receptor binding, receptor-dependent immune-cell function, and in vivo outcome would allow candidates to be prioritized for engineering.

This reframing also casts doubt on the reliability of non-human primate models as proxies for human compatibility. Old World monkeys retain functional CMAH, treat Neu5Gc as self, and produce natural antibodies against some structures exposed on Neu5Gc-deficient pig cells that humans generally lack. Consistent with this, baboon kidney studies found that Neu5Gc-negative DKO/TKO grafts survived for shorter periods with more antibody-mediated rejection than Neu5Gc-positive GTKO grafts. The same CMAH deletion that helps compatibility with human serum can paradoxically worsen compatibility in baboons, so the review argues for human serum, blood-cell, and receptor-based functional assays to complement NHP survival data, together with ex vivo organ perfusion as a testing and intervention platform.

Looking forward, the reviewer proposes precision glycoengineering rather than indiscriminate gene stacking. Organ- and cell-type-restricted expression of sialyltransferases, guided by endothelial-specific promoters, could modify only the glycocalyx that recipient blood actually encounters. Quantitative intact glycopeptide profiling by mass spectrometry could verify target-antigen depletion and reveal post-editing glycoform redistribution that lectin staining misses, and transient inhibition of neuraminidase NEU1 during organ preservation might protect existing sialylation from ischemia-reperfusion damage. Donor modification strategies, including human CD47, CD200, thrombomodulin, and humanized von Willebrand factor, remain complementary tools, but the guiding principle is minimal, mechanistically justified engineering matched to each organ and recipient. Until defined glycan-receptor interventions are shown to improve clinically meaningful outcomes, the sialic acid-Siglec axis is best viewed not as an alternative explanation for rejection but as a candidate regulatory layer that could account for part of the innate immune and thromboinflammatory incompatibility that survives even the most sophisticated xenoantigen removal.

Subject of Research: Sialic acid-dependent glycan recognition in pig-to-human xenotransplantation

Article Title: Neu5Gc and beyond: sialylation at the innate immune and thromboinflammatory interface of xenotransplantation

Article References: Kim, J. Y. (2026). Neu5Gc and beyond: sialylation at the innate immune and thromboinflammatory interface of xenotransplantation. Frontiers in Transplantation, 5, Article 1968710. https://doi.org/10.3389/frtra.2026.1968710

Image Credits: AI Generated

DOI: 10.3389/frtra.2026.1968710

Keywords: xenotransplantation, sialic acid, Neu5Gc, Siglec, glycoengineering, complement, Factor H, thromboinflammation, pig organ transplant, platelet clearance, innate immunity, gene editing