Immunoids: Stem Cell Organoids With Immune Systems Could Reshape Disease Research
Scientists are building 'immunoids' — organoids from human stem cells that include functional immune components — to model disease and test therapies.
Summary
Scientists can grow miniature organs from stem cells, but these have usually lacked an immune system. This Cell Stem Cell perspective reviews ten years of work on 'immunoids', organoids and multi-tissue assembloids that contain working immune cells. Immunoids now exist for lung, bone marrow, brain, heart, gut, liver, eye, thymus, blood vessels, and tumors. They have been used to model Alzheimer's-linked APOE4 microglia, heart rhythm problems driven by inflammation, scarring of the lung, sepsis-like liver injury, and myelin repair. The authors describe four ways to build them and explain why the choice matters. They stress that today's immunoids are only partly immune-competent. Their immune cells often resemble fetal rather than adult cells, which limits their use for studying aging. The authors call for shared standards to make these models more reliable.
Detailed Summary
Animal studies often fail to predict how drugs behave in people, because immune regulation differs between species. Regulators have begun to respond: the FDA has issued guidance on new approach methodologies (NAMs), and the NIH has announced an initiative to prioritize them. This perspective focuses on one NAM type, the 'immunoid', an organoid or assembloid built from human pluripotent stem cells that includes at least one functioning immune cell lineage. Because inflammation drives most age-related disease, these models are central to regenerative and disease research.
The authors reviewed ten years of literature and sorted the approaches into four strategies: adding fully mature immune cells to an organoid, adding immune progenitors, mixing organoid and hematopoietic progenitors so they co-differentiate, and letting immune cells arise on their own inside the organoid. Context decides which is best. Many tissue-resident macrophages, including microglia, are seeded in the embryo from the yolk sac, so progenitor-based approaches better mimic development. Mature, activation-competent cells suit models of acute inflammation. Most immunoids so far use the simplest route, adding mature cells to established organoids.
The first immunoid appeared in 2017, when stem cell-derived microglia were added to brain organoids and reacted to needle injury. Immunoids now cover ten tissue types. Alveolar macrophages in lung models captured fibrosis, tested TGF-β inhibitors, and responded to LPS and Mycobacterium tuberculosis. A 2024 bone marrow organoid formed in about 3 weeks with endogenous monocytes, neutrophils, mast cells, megakaryocytes, and dendritic cells, and modeled VPS45 deficiency. Brain models showed that APOE4 microglia take up less amyloid-β and that APOE4 organoids accumulate more amyloid-β aggregation. Microglia also promoted neuronal maturation and myelin repair.
Other organs gave equally specific results. Heart organoids with autologous macrophages developed atrial arrhythmias through NLRP3 inflammasome activation when exposed to inflammatory cytokines. Liver immunoids with Kupffer cells showed sepsis-like dysfunction after LPS and IFN-γ, including more IL-6, lower CYP3A4/1A1/1A2 activity, and impaired ammonia metabolism. One liver model generated CD7+CD5+ T-cell precursors, 12.6% of which were CD4+CD8a+. A thymic immunoid was built by aggregating three cell types in a 1:1:20 ratio over 4 weeks. In it, 44% of cells were CD45+, and 64.3% of those were CD3+. Gut, eye, vascular (21-day culture, SARS-CoV-2 testing), and cancer immunoids are also described.
The authors say the implications are broad: human-specific modeling of development, infection, neuroinflammation, arrhythmia, fibrosis, tumor immune evasion, and drug response. They propose several next steps. These include fluidic bone marrow-on-chip systems linked to distal organoids, perfusable cardiac-vascular models of atherosclerosis, and added adaptive immune cells to model multiple sclerosis, Crohn's disease, and ulcerative colitis.
The caveats are substantial. All current immunoids are only partially immune-competent. Their macrophages and microglia carry canonical markers but retain embryonic or fetal transcriptional, epigenetic, and metabolic states, which limits models of aging, chronic inflammation, and immune memory. Immune cell abundance, location, and ratios are poorly defined, and functional validation is inconsistent across studies. Standardized fabrication criteria will be needed for clinical translation. The full limitations section was truncated in the supplied text.
Key Findings
- Ten years of literature sorted into four immunoid-building strategies; most studies add mature immune cells to established organoids
- First hPSC immunoid reported in 2017 (microglia in brain organoids); immunoids now span 10 tissue types
- 2024 bone marrow organoids formed in about 3 weeks with endogenous monocytes, neutrophils, mast cells, megakaryocytes, and dendritic cells
- APOE4 microglia showed reduced amyloid-β uptake and increased amyloid-β aggregation in APOE4 cerebral organoids
- Heart-macrophage immunoids developed atrial arrhythmias via NLRP3 inflammasome activation under pro-inflammatory cytokines
- Thymic immunoids (1:1:20 cell ratio, 4 weeks) were 44% CD45+, with 64.3% of CD45+ cells CD3+
- Liver immunoid T-cell induction yielded CD7+CD5+ precursors, 12.6% of them CD4+CD8a+; immune cells often retain fetal-like states, limiting aging models
Methodology
This is a narrative Perspective, not a systematic review. The authors surveyed roughly the past 10 years of self-organizing, autologous hPSC-derived immunoids, selecting studies for novelty, impact, recency, and rigor (Table 1, Figures 1-3). They excluded non-autologous, primary-cell, animal, co-culture, and engineered-tissue models. No pooled analysis or statistics were performed; the numbers cited come from the individual primary studies. The supplied text was truncated before the end of the limitations section.
Study Limitations
The authors state that all immunoids to date are only partially immune-competent. Their immune cells often retain fetal-like transcriptional, epigenetic, and metabolic states, which limits models of aging and immune memory. Cell ratios, localization, and functional validation are poorly standardized, and vascularization, adaptive immunity, and scalability remain unresolved. Conflict-of-interest details were not included in the supplied text.
Enjoyed this summary?
Get the latest longevity research delivered to your inbox every week.
Enter your email to subscribe: