The placenta sorts by cargo, and that gives antibody engineers a lever

Researchers at the University of Oslo and Oslo University Hospital, with national and international collaborators, report that the placenta's antibody shuttle moves IgG to the fetus while largely leaving albumin behind, and that fusing albumin to an antibody blocks most of the antibody's transfer too ^[1][2][3]. That converts a recurring clinical judgment call, treat the mother and dose the fetus or withhold treatment, into a protein design specification ^[8][9].

The receptor in question, FcRn, is the same one that gives both IgG and albumin their long circulating half-lives, and it binds the two proteins at different sites ^[2][10]. Albumin and IgG are the two most abundant soluble proteins in blood, with albumin acting as a carrier for fatty acids, hormones and waste products ^[7]. In mice, the team led by Jan Terje Andersen found that FcRn carried maternal IgG across the placenta into fetal pups but did not carry albumin ^[4][11]. The work combined conventional and genetically humanized mouse models with an ex vivo human placental perfusion system using placentas donated immediately after childbirth, and the split held across all models: IgG moved efficiently, albumin did not ^[12][13]. The useful implication is that binding FcRn is not by itself enough to get a molecule across the placenta, since albumin does the first without the second ^[18].

The engineering follows from that. Fusing albumin to therapeutic IgG antibodies produced molecules that kept the FcRn-driven long plasma half-life while substantially reducing placental transport ^[14]. Fusing antibody fragments to an engineered albumin variant the authors call QMP, with optimized human FcRn binding, produced a larger effect, which the team presents as evidence that both placental transfer and half-life can be tuned by design rather than accepted as fixed properties of the format ^[15]. In a mouse model of fetal and neonatal alloimmune thrombocytopenia, a potentially life-threatening condition in which maternal antibodies attack fetal platelets, the engineered antibodies produced substantially reduced fetal exposure and fewer associated adverse effects in offspring ^[16].

The paper, "Fusion of IgG antibodies to albumin inhibits transport across the placenta," appears in Science Immunology with Andersen as corresponding author, and concludes that albumin is "an attractive fusion partner for biologics intended to minimize fetal exposure during pregnancy" ^[5][6]. Andersen frames it as a change of question: "Rather than asking whether existing biologic medicines are safe to use during pregnancy, our findings show that we can now design them differently" ^[17].

Why operators in the space should care is the installed base. IgG monoclonals are among the fastest-growing classes of biologics, with approved uses in cancer, autoimmune disease and migraine, and they are actively hauled across the placenta by FcRn, which is precisely what limits their use in pregnancy ^[8]. Evidence supporting safe use of these drugs during pregnancy remains limited, so the current default is a benefit-risk negotiation with thin data on one side ^[9].

What to watch: this is preclinical, in mice, humanized mice and perfused donor placentas, so the numbers that matter next are human pharmacokinetics for an albumin-fused candidate ^[1][12]. Watch also whether the larger reduction seen with fragment-plus-QMP constructs can be reproduced with full-size IgG, since fragments lose the Fc functions many marketed antibodies depend on ^[15].