August 29, 2023 /
What are humanized mouse models?
Humanized mouse models are created by implanting human genes, cells, tissues, or organs into a mouse (Allen et al., 2019; Chen et al., 2022). They aim to recreate parts or the whole of the human immune system in a murine host, thereby providing an in vivo preclinical model to investigate human immune conditions and diseases (e.g. HIV) or to test the safety of drugs that target human-specific immunoreceptors. In addition, humanized models are widely used to study the mechanisms of and potential treatments for different types of cancer, including novel immune-modulating anticancer drugs (Chuprin et al, 2023).
Many of these models are based on reconstituting human immune cell types in immunodeficient mice, which possess genetic defects that cause impaired development of the mouse’s own immune system. This reduces the risk of graft rejection and facilitates expression of human immune function in the host.
Development of immunodeficient mouse models
Immunodeficient mice were first developed based on the CB17 mouse strain in 1988, following the discovery of the severe combined immunodeficiency (SCID) mutation, which leads to a near-complete lack of mature B and T cells in host mice. However, these mice were still found to express innate immune responses via natural killer (NK) cells, macrophages, granulocytes, and complement proteins, and were therefore not entirely immunodeficient.
Subsequent research aimed to refine host immunodeficiency further, with the first breakthrough coming in 1990 with the creation of NOD/SCID mice. These were generated by crossing SCID mice with inbred non-obese diabetic (NOD) mice possessing defective NK cells and macrophages, along with reduced complement activity.
In 2000, the final piece of the puzzle was put in place, when NOD/SCID mice were crossed with IL-2 receptor gamma chain (IL-2Rγ) knockout mice showing additional immune defects, including dendritic cell dysfunction. This yielded current the “gold-standard” immunodeficient NOD/SCID/IL2Rγnull strains, including the NOD/Shi-scid/IL-2Rγnull (NOG) mouse and the NOD.Cg-Prkdcscid Il2rgtm1Vst/Vst (NPG) mouse.
NOD/SCID/IL2Rγnull: a gold-standard immunodeficient mouse for humanization
Overall, NOD/SCID/IL2Rγnull model mice are characterized by a lack of T cell, B cell, and NK cell activity, dysfunction of macrophages and dendritic cells, and impairment of the complement system. Therefore, they readily accept xenogeneic (non-host) cells, enabling them to serve as a “blank canvas” onto which human immune cells can be engrafted for research and drug testing.
Humanization of the mouse model occurs by injecting human immune cells (e.g. peripheral blood mononuclear cells), tissues (e.g. thymus), or multipotent stem cells (e.g. hematopoietic stem cells) to establish a human immune cell population in the immunodeficient host.
Humanized models vs. common alternatives for cancer research
In terms of cancer research, it is important to consider whether an alternative model might be a better choice than a humanized mouse model – in particular, syngeneic and xenograft (CDX and PDX) models are widely used alternatives that may be more suitable depending on the purpose of the study.
Syngeneic mouse models
In syngeneic mouse models (also known as allograft mouse tumor systems), tumor cells obtained from an inbred mouse strain (e.g. C57BL/6 or Balb/c) are immortalized in culture, from where they can be repeatedly implanted into immunocompetent mice of the same genetic strain.
These models offer the advantage of consistent replication of tumor cell biology, microenvironment, and host responses across animals, as they display full immunity, as well as typical development of tumor stroma. Therefore, they are especially applicable to studying the whole-system effects of and responses to cancer, for example during metastasis.
However, as the tumor cells in syngeneic models are of murine origin, these models may replicate the mechanisms and genetics of human cancers poorly. There are also a limited number of tumor cell lines available for syngeneic mouse models, meaning that particular types of cancer may not be well represented in available syngeneic model stocks.
Xenograft mouse models
In xenograft models, an immunocompromised mouse model is engrafted with human tumor cells or tissue. This is obtained either from an immortalized cancer cell line, to form a cell-derived xenograft (CDX) model, or from a patient tumor sample, to form a patient-derived xenograft (PDX) model.
Xenograft models offer numerous advantages, including the simplicity of the engraftment method, the ease with which the implanted tumor can be monitored, and the fast turnaround time from model setup to experimental data acquisition. CDX models in particular have been used for decades and are therefore well characterized and are an excellent choice for rapid, low-cost screening of anti-tumor compounds.
However, their major drawbacks are that the heterogeneity, lineage hierarchy, and microenvironment of the tumor are not reliably preserved, nor do they model host immune responses to the tumor. This leads to differences in tumor characteristics and evolution, which has led to difficulties in translating preclinical results from CDX models into clinical outcomes.
Humanized mouse models
Humanized models are a form of xenograft model, based on engraftment of human tumor cells or tissue into an immunodeficient mouse. However, in contrast to non-humanized CDX or PDX models, they mimic human immune function, enabling them to yield more reliable results in studies of human infectious diseases, immunology, and immuno-oncology.
This has proven especially important in investigating novel classes of anticancer drugs that target immune checkpoint proteins, such as PD-1 and CTLA-4. As humanized models replicate the human immune system, they predict responses to drug candidates more accurately, leading to superior translation from preclinical findings to clinical efficacy.
However, they are more complex and time-consuming to establish and use, making them less well suited to high-throughput screening or testing.
Common humanized mouse models
The most widely used humanized mouse models are those based on immune reconstitution with hematopoietic stem cells (HSCs), peripheral blood mononuclear cells (PBMCs), and/or fetal organ tissue from thymus and liver (Allen et al., 2019; Chen et al., 2022).
Each model has different characteristics, such as the ease with which it can be established, the longevity of the humanized mouse, and the degree of immune function exhibited by the mice. These properties should be considered in light of the requirements of the study being conducted.
Below, we discuss in more detail three of the most widely used humanized models — the HuHSC, HuPBMC, and HuBLT mouse models.
HuHSC mouse model
In the HuHSC model, CD34+ hematopoietic stem cells (HSCs) derived from human umbilical cord blood are injected into immunodeficient NOG/NPG mice. HSCs are multipotent stem cells capable of differentiating into either myeloid-lineage cells (e.g. red blood cells, macrophages) or lymphoid-lineage cells (e.g. T cells, B cells, NK cells). Once injected into the host, they give rise to a wide range of human immune cells that can mediate both innate and adaptive immune responses (Chen et al., 2022).
The degree of humanization of the HuHSC model can be determined by flow cytometry quantification of the ratio of cells in blood expressing human vs. mouse CD45, a cell surface marker of leukocytes (white blood cells). Within 16 weeks, NOG/NPG mice show a human CD45+ cell ratio of 40%–60%, well above the minimum of 15% required to establish a valid model, according to dosing studies. Grafted cells develop into a range of immune cell types, including T cells, B cells, NK cells, and myeloid-derived suppressor cells (MDSCs).
The HuCD34 model is one of the most long-living humanized models, as the grafted cells develop to be tolerant of the host. This enables research to continue with this model for over a year before it exhibits graft-versus-host disease (GVHD). Therefore, this model is useful for a broad range of research topics, including longer term studies of immune function in cancer, infectious diseases such as HIV, and the development of the hematopoietic system.
HuPBMC mouse model
The HuPBMC model is the most direct method for creating a humanized mouse model via injection of human peripheral blood mononuclear cells (PBMCs). PBMCs are mononuclear white blood cells isolated by density gradient centrifugation from whole blood, and include lymphocytes (T cells, B cells, and NK cells), monocytes, and dendritic cells.
After injection, human CD45+ cells (leukocytes) can be detected at high levels in the blood and spleen of HuPBMC mice. Human CD3+ cells (T cells) also mature in the host over the course of 7–10 days, making the HuPBMC model well-suited for studying T cell function.
However, this model exhibits incomplete human lymphocyte reconstruction and lacks normal lymphoid tissue structure. In addition, it is short-lived as the graft PBMCs subsequently attack the host in GVHD, typically leading to death within 4–6 weeks (Chen et al., 2022). Therefore, this model is best suited to short-term studies of immunity (e.g. testing human immunosuppressants), infectious diseases (e.g. HIV), immuno-oncology, and graft rejection.
HuBLT mouse model
The bone marrow/liver/thymus (BLT) model is created by injecting CD34+ HSCs obtained from fetal liver or bone marrow into an immunodeficient mouse, in addition to co-transplanting human fetal liver and thymus into the renal capsule (Smith et al., 2016).
This advanced humanized system can recapitulate a wide range of human immune responses, including mucosal immunity, innate immunity, and adaptive immunity. This near-complete immune function is mediated by a multilineage population of T cells, B cells, NK cells, monocytes, macrophages, and dendritic cells. These immune cells show superior development compared to the HuPBMC and HuHSC models, including education of T cells on human MHC within the implanted thymic organoid. Therefore, the BLT model is a highly effective choice for research into diverse topics, such as cancer, immunology, infectious diseases, and stem cell therapy.
However, this model shows some susceptibility to GVHD from 16–20 weeks after reconstitution (Greenblatt et al., 2012). Moreover, a major drawback of this model is the difficulty in obtaining the fetal material required for transplantation. Recent work, however, suggests that by combining hematopoietic stem and progenitor cells from umbilical cord blood with thymic tissue from neonatal surgeries, a novel form of the BLT mouse, dubbed the NeoThy mouse, can be created using more readily accessible neonatal tissue (Brown et al., 2018; Del Rio et al., 2023).
Human immune checkpoint mouse models
In contrast to the HuHSC, HuPBMC, and HuBLT models, which are based on immunodeficient mice, human immune checkpoint mouse models are established in immunocompetent mice by genetically modifying the host to express a partially or fully human version of an immune checkpoint protein. This enables therapeutic compounds targeting these proteins to be tested in a convenient in vivo system.
HuPD-1 mouse model
PD-1 is an immune checkpoint protein involved in suppressing CD8+ (cytotoxic) T cell responses during the later phase of an infection. It is expressed on T cells, and acts to promote apoptosis in antigen-specific T cells, while reducing apoptosis in regulatory T-cells. Much research in recent years has focused on novel therapies for cancer that block PD-1, thereby increasing the anti-tumor response of CD8+ T cells (Seidel, Otsuka, & Kabashima, 2018).
The HuPD-1 mouse model contains a humanized form of the programmed cell death protein 1 (PD-1), containing an entirely humanized extracellular domain. This model shows a normal human expression pattern and physiological regulation, and is therefore ideal for testing immuno-oncology drugs targeting PD-1.
HuCTLA-4 mouse model
Similar to PD-1, cytotoxic T-lymphocyte–associated antigen 4 (CTLA-4) is an immune checkpoint protein expressed on CD8+ T cells. CTLA-4 is involved in downregulating T-cell proliferation during the early phase of an immune response. Therefore, it is also a promising target for immuno-oncology drugs aiming to boost cytotoxic T cell function against tumors. PD-1 and CTLA-4 blockers have complementary effects, and have already been combined successfully to treat cancer (Rotte, 2019).
The HuCTLA-4 mouse model is developed by knockin at the CTLA-4 locus, yielding a completely humanized extracellular domain. The expression pattern and physiological regulation of CTLA-4 match those in humans, making this model well-suited to testing pharmacological agents that target CTLA-4.
Want to see what Humanized Models PharmaLegacy has available? Download our Humanized Model List!
Summary
In summary, humanized mouse models offer a powerful approach to studying human immune function in a laboratory animal. They are applicable to diverse areas of research, including studies of cancer, infectious diseases, immune system development, and immunomodulation. They present distinct advantages and disadvantages relative to alternative syngeneic and xenograft models, being more complex to establish, but yielding results that potentially translate more effectively to clinical settings.
These factors should be considered carefully in choosing the best model system for a given research study, along with the relative merits and drawbacks of the HuHSC, HuPBMC, HuBLT, HuPD-1, and HuCTLA-4 humanized models, which we summarize below:
| Model | Advantages | Disadvantages | Recommended applications |
| HuHSC | Wide range of immune cells reconstituted Long survival time (>1 year) prior to GVHD |
Difficulty obtaining HSCs | Longer term studies of cancer, infectious diseases, immune system development, etc. |
| HuPBMC | PBMCs easy to obtain Simplest model to set up |
GVHD can develop within 4–6 weeks Incomplete immune reconstruction |
Short-term studies of immunity, infectious diseases, cancer, etc. |
| HuBLT | Near-complete human immune system reconstitution Improved immune cell development |
Possible GVHD after 16–20 weeks Difficult to obtain BLT samples to establish model |
In-depth research requiring accurate results, but lower throughput |
| HuPD-1 or HuCTLA-4 | Designed specifically for testing drugs target PD-1 or CTLA-4, respectively | Murine immune system, therefore does not model human immune responses directly | Testing drugs targeting PD-1 or CTLA-4, respectively |
References
Allen, T. M., Brehm, M. A., Bridges, S., et al. (2019). Humanized immune system mouse models: progress, challenges and opportunities. Nature Immunology, 20, 770–774. https://doi.org/10.1038/s41590-019-0416-z
Brown, M. E., Zhou, Y., McIntosh, B. E., et al. (2018). A Humanized Mouse Model Generated Using Surplus Neonatal Tissue. Stem Cell Reports, 10(4), 1175. https://doi.org/10.1016/j.stemcr.2018.02.011
Chen, J., Liao, S., Xiao, Z., et al. (2022). The development and improvement of immunodeficient mice and humanized immune system mouse models. Frontiers in Immunology, 13, 1007579. https://doi.org/10.3389/fimmu.2022.1007579
Chuprin, J., Buettner, H., Seedhom, M. O., et al. (2023). Humanized mouse models for immuno-oncology research. Nature Reviews Clinical Oncology, 20, 192–206. https://doi.org/10.1038/s41571-022-00721-2
Del Rio, N. M., Huang, L., Murphy, L., et al. (2023). Generation of the NeoThy mouse model for human immune system studies. Lab Animal, 52(7), 149-168. https://doi.org/10.1038/s41684-023-01196-z
Greenblatt, M. B., Vbranac, V., Tivey, T., et al. (2012). Graft versus host disease in the bone marrow, liver and thymus humanized mouse model. PLoS ONE. https://doi.org/10.1371/journal.pone.0044664
Rotte, A. (2019). Combination of CTLA-4 and PD-1 blockers for treatment of cancer. Journal of Experimental & Clinical Cancer Research, 38, 255. https://doi.org/10.1186/s13046-019-1259-z
Seidel J. A., Otsuka A., and Kabashima K. (2018). Anti-PD-1 and Anti-CTLA-4 Therapies in Cancer: Mechanisms of Action, Efficacy, and Limitations. Frontiers in Oncology, 8:86. https://doi.org/10.3389/fonc.2018.00086
Smith D. J., Lin L. J., Moon H., et al. (2016). Propagating Humanized BLT Mice for the Study of Human Immunology and Immunotherapy. Stem Cells and Development, 25(24), 1863-1873. https://doi.org/10.1089/scd.2016.0193