Humanized immune system mice

Humanized immune system (HIS) mice combine an immunodeficient mouse host with human hematopoietic cells.^[S4][S9] The term covers materially different designs, so a proposal should identify both the host strain and the human-cell source. This article compares those design choices and the limits they impose.

What the humanized compartment does and does not establish

The NSG record describes a severely immunodeficient host carrying Prkdcscid and a null allele of the interleukin-2 receptor common gamma chain; the strain is used to engraft human hematopoietic cells and tissues.^[S1] Engraftment lets investigators measure specified human-cell endpoints in a living mouse context. It does not create a complete human immune system, and it does not make efficacy or safety findings predictive of patient outcomes. Reviews of the field emphasize both progress and continuing challenges.^[S4][S9]

Verified foundational reviews

The two citations in the Japanese original did not match the verified literature records.^[S4][S6] The entries below use the confirmed titles and journals. Their role is to orient study design, not to certify one protocol as standard or clinically predictive.

Platform classification

• {'items': ['Verified paper: “Humanized mice for immune system investigation: progress, promise and challenges.”^[S4]', 'Shultz LD, Brehm MA, Garcia-Martinez JV, and Greiner DL; Nature Reviews Immunology, 2012.^[S4]', 'Related later review: “Humanized mouse models for immuno-oncology research,” Nature Reviews Clinical Oncology, 2023.^[S5]']}

Distinguish hu-HSC or CD34+ HSC-engrafted models from hu-PBL or PBMC-humanized models in the protocol. Reviews addressing immuno-oncology and standardization treat them as different systems and identify graft-versus-host disease (GVHD) as a material constraint for PBMC-based designs.^[S5][S9] Ask for the cell source, donor identifier, conditioning, engraftment criteria, immune-panel schedule, expected observation window, and GVHD stopping rules.

Scope of a disease-model review

• {'items': ['Verified paper: “Humanized Mouse Models of Clinical Disease.”^[S6]', 'Walsh NC, Kenney LL, Jangalwe S, Aryee KE, Greiner DL, Brehm MA, and Shultz LD; Annual Review of Pathology, 2017.^[S6]']}

The verified Walsh paper is an Annual Review of Pathology article about humanized mouse models of clinical disease.^[S6] Its title and bibliographic record do not support the Japanese original’s claim that clinical correlation has been established for checkpoint inhibitors or cell therapies. Use it as a review source, then cite primary studies for any product-specific or outcome-specific assertion.

Cytokine-supported host strains

Some host strains express selected human cytokines to support particular engrafted lineages.^[S2] That changes the experimental system and its liabilities; it does not make the mouse physiologically human. Compare the exact transgenes, target cell populations, survival window, and reported adverse phenotypes before choosing a host.

Cytokine-supported HIS mice

• {'items': ['NSG-SGM3: NOD.Cg-Prkdcscid Il2rgtm1Wjl Tg(CMV-IL3,CSF2,KITLG)1Eav/MloySzJ, JAX Stock No. 013062.^[S2]', 'Corrected citation: Willis et al., Veterinary Pathology, 2024, comparing CD34+ humanization in NOG-EXL and NSG-SGM3 mice.^[S7]', 'Review: Stocks et al., “Advances in human myeloid-engrafting NSG-SGM3 and MISTRG mice,” Trends in Immunology, 2025.^[S8]']}

The JAX record states that NSG-SGM3 expresses human IL3, CSF2 (GM-CSF), and KITLG (SCF).^[S2] A 2024 paper reports improved long-term survival and a less severe myeloid-cell hyperactivation phenotype in CD34+ humanized NOG-EXL mice relative to NSG-SGM3 mice.^[S7] Those are comparative findings under the reported conditions, not proof that one host is generally superior. Match the host to the prespecified cell population and observation window.

Platform records and comparison questions

Do not select a platform from a modality label alone. Compare host genotype, engrafted cell type, donor, conditioning, supported lineages, observation window, GVHD criteria, and the assay that answers the development question. The records below establish strain identity; they do not recommend one supplier or platform.

Designing for donor variation

Treat the donor as an experimental factor. Record donor identity, source material, processing, release criteria, engraftment threshold, allocation across groups, and whether inference is donor-specific or intended to span donors. A standardization review is available to frame the minimum information that should be reported.^[S9] Multiple donors can characterize between-donor variability in the model; they do not prove that the observed range predicts patient responses.

Commissioning checklist

Require the proposal to name the host strain and stock record, human-cell type and donor, conditioning, engraftment and exclusion criteria, immune panel, observation window, GVHD monitoring, primary endpoint, and analysis unit. HIS data establish responses in that combined mouse-host and human-cell system. They do not establish efficacy or safety in patients.