Decidual natural killer (dNK) cells are the dominant leukocyte of the first-trimester human decidua, and one subset — CD3−CD56brightCD25+ dNK — was reported to accumulate at the maternal-fetal interface, display a regulatory-like profile, and be recruited via the CXCL12/CXCR4 chemokine axis after trophoblast-mediated instruction of peripheral NK cells. Whether this population is a genuine regulatory NK subset in vivo, whether any definitive surface marker can define regulatory NK cells, and how its TGF-β output divides responsibility for maternal-fetal tolerance with IL-10-producing regulatory T cells (Tregs) remain unresolved.

Background

Natural killer (NK) cells are innate lymphocytes that conventionally kill stressed or infected cells, but at the human maternal-fetal interface they adopt a non-cytotoxic, secretory role. In early pregnancy the uterine mucosa becomes the decidua, populated by fetal placental trophoblast cells that invade maternal tissue and remodel the uterine spiral arteries into low-resistance vessels; ~70% of first-trimester decidual leukocytes are CD56brightCD16− decidual NK (dNK) cells — phenotypically distinct from the mainly cytotoxic CD56dim NK cells of blood — which produce angiogenic factors and cytokines that support trophoblast invasion and vascular remodeling. CD56 (neural cell adhesion molecule) marks the immunomodulatory, poorly cytotoxic NK subset; CD25 is the interleukin-2 receptor alpha chain, classically used to identify regulatory T cells (Tregs, a suppressive CD4+ T lineage sustained by the transcription factor Foxp3). Tregs and dNK cells are thought to cooperate in maternal-fetal tolerance — the mother's acceptance of the semi-allogeneic fetus — partly through the anti-inflammatory cytokines transforming growth factor beta (TGF-β) and interleukin-10 (IL-10). Against this background, mouse studies had proposed regulatory NK populations (e.g., a DX5+CD3−CD25+Thy1.2brightc-kitdim subset), and Tao et al. (2015) identified the human analogue: CD3−CD56brightCD25+ dNK cells enriched at the interface (~75% of CD25+ decidual immune cells), more activated and cytokine-producing than CD25− dNK cells, inducible from peripheral NK by trophoblast coculture, and recruited through the CXCL12/CXCR4 chemokine axis (CXCL12 being the ligand of the CXCR4 receptor). The authors themselves flagged what remained unknown: no definitive surface marker defines regulatory NK cells; in vivo regulatory function of this population was unestablished; the relative contributions of CD25+ dNK-derived TGF-β versus IL-10-producing Tregs to tolerance were undissected; and whether trophoblast instruction of NK cells is soluble or contact-dependent was untested. Later work reshaped the landscape without closing these questions: single-cell transcriptomic atlases (Vento-Tormo et al. 2018) divided dNK cells into dNK1/dNK2/dNK3 subsets with distinct immunomodulatory and chemokine profiles and predicted many tolerogenic ligand-receptor interactions, and the same group's follow-up (Tao et al. 2021) showed that a related CXCR4+CD56bright dNK subset promotes a Th2-skewed (anti-inflammatory) environment and reduces embryo resorption after adoptive transfer into NK-deficient mice. In parallel, the broader NK literature proposed several 'regulatory NK' phenotypes — CD73-expressing NK in tumors (Neo et al. 2020), CD1dhiPD-L1hiCD27+ NK in atopic dermatitis (Min et al. 2022), CD38+ NK with disease-opposite functions in rheumatoid arthritis versus colorectal cancer (Chang & Fang 2026), and CD49a+CD25+ liver NK cells (Martrus et al. 2017) — all context-specific, with a 2025 review of regulatory immune cells still treating NKregs as an emerging, incompletely characterized population. Most strikingly, Frutoso et al. (2026) showed that anatomically adjacent decidual and decidual-placental-interface NK populations at term are functionally distinct in ways not predictable from their surface phenotype or transcriptome, underscoring that defining a regulatory NK subset at the interface requires functional — not merely phenotypic — criteria.

Problem Statement

Does a bona fide regulatory NK cell subset exist and operate in vivo at the human maternal-fetal interface, and what defines it? For this task, "regulatory" is used in a functional sense — an NK cell whose role at the interface is to promote tolerance and restrain effector immune responses rather than to kill — so that the existence and marker questions are adjudicated on that criterion. Concretely, as posed by the source work and left open since: (1) determine whether the CD3−CD56brightCD25+ decidual NK population — or a refined phenotype — truly functions as a regulatory NK subset in vivo, including how trophoblasts generate it (soluble factors versus cell-contact instruction of peripheral NK cells) and how the CXCL12/CXCR4 axis and possibly other chemokine axes drive its accumulation at the interface; (2) identify a definitive surface marker (or marker combination/signature) that reproducibly defines regulatory NK cells, as opposed to the current collection of context-specific candidates; and (3) causally dissect the relative contributions of CD25+ dNK-derived TGF-β versus IL-10-producing Treg cells to maternal-fetal tolerance — with the linked functional question of how this subset's IFN-γ and IL-8 output feeds into spiral artery remodeling. These three deliverables jointly serve one objective: establishing the existence, identity, and tolerogenic mechanism of regulatory NK cells at the maternal-fetal interface. Deliverables (1)–(3) are the mandatory sub-questions; the IFN-γ/IL-8-to-spiral-artery-remodeling question is a linked sub-question of (3), to be addressed insofar as it bears on this subset's tolerogenic mechanism. A well-supported negative answer to (1) resolves deliverable (1): the CD25+ candidate is not a bona fide regulatory NK subset in vivo. It does not by itself resolve (2), which asks whether any marker reproducibly defines regulatory NK cells as a general class and must therefore be assessed independently of (1); it does resolve (3) in the negative for this candidate, since no CD25+ dNK-derived TGF-β output then remains to be separated from that of Tregs.

References

  1. Tao Y, Li Y, Piao H, Zhou W, Zhang D, Fu Q, Wang S, Li D, Du M (2015). CD56brightCD25+ NK cells are preferentially recruited to the maternal/fetal interface in early human pregnancy. Cellular & Molecular Immunology 12(1):77-86. doi:10.1038/cmi.2014.26. https://doi.org/10.1038/cmi.2014.26
  2. Tao Y, Li Y, Zhang D, Xu L, Chen J, Sang Y, Piao H, Jing X, Yu M, Fu Q, Zhou S, Li D, Du M (2021). Decidual CXCR4+CD56bright NK cells as a novel NK subset in maternal-foetal immune tolerance to alleviate early pregnancy failure. Clinical and Translational Medicine 11(10):e540. doi:10.1002/ctm2.540. https://doi.org/10.1002/ctm2.540
  3. Vento-Tormo R, Efremova M, Botting RA, Turco MY, Vento-Tormo M, Meyer KB, Park J, Stephenson E, Polański K, Goncalves A, Gardner L, Holmqvist S, Henriksson J, Zou A, Sharkey AM, Millar B, Innes B, Wood L, Wilbrey-Clark A, Payne RP, Ivarsson MA, Lisgo S, Filby A, Rowitch DH, Bulmer JN, Wright GJ, Stubbington MJT, Haniffa M, Moffett A, Teichmann SA (2018). Single-cell reconstruction of the early maternal-fetal interface in humans. Nature 563:347-353. doi:10.1038/s41586-018-0698-6. https://doi.org/10.1038/s41586-018-0698-6
  4. Liu L, Zhang Z, Huang Q, Liu B, Wu H (2026). dNK3 cells in normal pregnancy and recurrent pregnancy loss: from molecular identity to functional imbalance. Frontiers in Immunology 17:1822205. doi:10.3389/fimmu.2026.1822205. https://doi.org/10.3389/fimmu.2026.1822205
  5. Neo SY, Yang Y, Record J, Ma R, Chen X, Chen Z, Tobin NP, Blake E, Seitz C, Thomas R, Wagner AK, Andersson J, de Boniface J, Bergh J, Murray S, Alici E, Childs R, Johansson M, Westerberg LS, Haglund F, Hartman J, Lundqvist A (2020). CD73 immune checkpoint defines regulatory NK cells within the tumor microenvironment. Journal of Clinical Investigation 130(2):1185-1198. doi:10.1172/jci128895. https://doi.org/10.1172/jci128895
  6. Min KY, Koo J, Noh G, Lee D, Jo MG, Lee JE, Kang M, Hyun SY, Choi WS, Kim HS (2022). CD1dhiPD-L1hiCD27+ regulatory natural killer subset suppresses atopic dermatitis. Frontiers in Immunology 12:752888. doi:10.3389/fimmu.2021.752888. https://doi.org/10.3389/fimmu.2021.752888
  7. Chang X, Fang K (2026). CD38+ NK cells: novel players in immunoregulation. Frontiers in Immunology 17:1757958. doi:10.3389/fimmu.2026.1757958. https://doi.org/10.3389/fimmu.2026.1757958
  8. Martrus G, Kautz T, Lunemann S, Richert L, Glau L, Salzberger W, Goebels H, Langeneckert A, Hess L, Poch T, Schramm C, Oldhafer KJ, Koch M, Tolosa E, Nashan B, Altfeld M (2017). Proliferative capacity exhibited by human liver-resident CD49a+CD25+ NK cells. PLOS ONE 12(8):e0182532. doi:10.1371/journal.pone.0182532. https://doi.org/10.1371/journal.pone.0182532
  9. Chen Z, Tong L, Neo SY, Li S, Gao J, Schlisio S, Lundqvist A (2023). CD25bright NK cells display superior function and metabolic activity under regulatory T cell-mediated suppression. OncoImmunology 12(1):2175517. doi:10.1080/2162402X.2023.2175517. https://doi.org/10.1080/2162402X.2023.2175517
  10. Horwitz DA, La Cava A (2025). CD2 targeted nanoparticles containing IL-2 mimic fetal-maternal tolerance of pregnancy by inducing synergistic TGF-β-producing NK cells and Tregs. Frontiers in Immunology 16:1692547. doi:10.3389/fimmu.2025.1692547. https://doi.org/10.3389/fimmu.2025.1692547
  11. Frutoso M, DeJong CS, Shree R, McCartney SA, Prlic M (2026). Phenotypically similar but functionally distinct NK cell populations within the human maternal-fetal interface. ImmunoHorizons 10(3):vlag001. doi:10.1093/immhor/vlag001. https://doi.org/10.1093/immhor/vlag001
  12. Mikhailova V, Grebenkina P, Selkov S, Sokolov D (2025). JEG-3 trophoblast cells influence ILC-like transformation of NK cells in vitro. International Journal of Molecular Sciences 26(8):3687. doi:10.3390/ijms26083687. https://doi.org/10.3390/ijms26083687
  13. Shi P, Yu Y, Xie H, Yin X, Chen X, Zhao Y, Zhao H (2025). Recent advances in regulatory immune cells: exploring the world beyond Tregs. Frontiers in Immunology 16:1530301. doi:10.3389/fimmu.2025.1530301. https://doi.org/10.3389/fimmu.2025.1530301
  14. Liu Y, Gao S, Zhao Y, Wang H, Pan Q, Shao Q (2021). Decidual natural killer cells: a good nanny at the maternal-fetal interface during early pregnancy. Frontiers in Immunology 12:663660. doi:10.3389/fimmu.2021.663660. https://doi.org/10.3389/fimmu.2021.663660
  15. Guan D, Chen Z, Zhang Y, Sun W, Li L, Huang X (2025). Dual role of natural killer cells in early pregnancy: immunopathological implications and therapeutic potential in recurrent spontaneous abortion and recurrent implantation failure. Cell Proliferation 58(9):e70037. doi:10.1111/cpr.70037. https://doi.org/10.1111/cpr.70037
  16. Li Y, Zhou W, Tao Y, Wang S, Jiang Y, Zhang D, Piao H, Fu Q, Li D, Du M (2016). The Galectin-9/Tim-3 pathway is involved in the regulation of NK cell function at the maternal-fetal interface in early pregnancy. Cellular & Molecular Immunology 13(1):73-81. doi:10.1038/cmi.2014.126. https://doi.org/10.1038/cmi.2014.126