A humanized “new-trophil” mouse to study early inflammatory processes

Neutrophils—discovered by Elie Metchnikoff in the late 19th century—are the most abundant leukocyte population in circulation and play a crucial role in the innate immune system. With one of the shortest life spans of any cell type in mammals (4–18 h in the blood), neutrophils rely on constant replenishment from the bone marrow. This has largely prevented studying the role of human neutrophils in early inflammatory responses. To address this challenge, Flavell and colleagues generated a humanized mouse model capable of supporting human myelopoiesis, including the generation of neutrophils, following transplantation of human hematopoietic stem cells (HSCs) (1).

Humanized Mice: A Long and Winding Road…

In the late 1980s, McCune and colleagues demonstrated that severe combined immunodeficiency mice engrafted with donor-matched human fetal liver HSCs, thymus, and lymph nodes supported the differentiation of mature human T and B cells. Although human hematopoietic engraftment was low, these initial humanized mice opened the door for studying human hematolymphoid differentiation and function (2). Notably, such mice engrafted with components of a human immune system (HIS) also supported HIV infection (3), enabling the testing of antiretrovirals. The low human immune reconstitution was partly due to the residual immune functions of early xenorecipient strains. Indeed, using more severely immunocompromised mice considerably enhanced hematolymphoid chimerism (4). Repopulation was further enhanced when destruction of human cells was mitigated by blocking the “eat-me” signaling relayed to murine macrophages due to incompatibilities between human CD47 and a murine signal regulatory protein (SIRPα) (5, 6). These second-generation humanized mouse models found considerable utility as challenge models for various pathogens, both viral and bacterial (reviewed in ref. 7).

However, immune responses were weak, possibly due to various factors, including improper thymic selection, the distorted architecture of secondary lymphoid organs, lack of human major histocompatibility complex expression and the scarcity or total lack of many critical immune cell subsets, especially those of the erythromyeloid lineages. Hematopoiesis is driven by a combination of cell–cell contacts and cytokines, many of which do not cross-react between mice and humans. To foster reconstitution with a more diverse array of human immune cell lineages and improve immune function, transient expression or injection of non–cross-reactive human cytokines in mice successfully led to increases in natural killer (NK) cells, dendritic cells, and macrophages (8). To attain more physiological expression of human cytokines, the Flavell group systematically humanized multiple murine cytokine gene loci via targeted insertions of human macrophage colony-stimulating factor (M), interleukin-3/granulocyte-monocyte colony-stimulating factor (I), SIRPα (S), and thrombopoietin (T) on the Rag2^−/−Il2rg^−/− (RG) background. The resultant “MISTRG” mice support differentiation of human monocytes, macrophages, and NK cells derived from HSCs (9). Remarkably, additional knock-in of human erythropoietin fosters human erythropoiesis, further substantiating the general approach of cytokine gene knock-in to improve human hematolymphoid system mice (10).

These and other humanized mice have found broad utility in many fields, including infectious disease and cancer biology. Potential therapeutics, such as broadly neutralizing antibodies to treat HIV, have also been assessed in such mice (11). Most recently, humanized mice coengrafted with lung tissue and components of a HIS revealed gene signatures correlated with protection against SARS-CoV2 (12). Humanized mice have also served as a unique platform for investigating the role of human macrophages in local and systemic inflammation upon SARS-CoV-2 infection (13, 14).

A “New-Trophil” Humanized Mouse Model

Neutrophils are at the forefront of our immune defenses against pathogens through their phagocytic activity, capacity to induce tissue inflammation and stress responses, and ability to form chromatin-based extracellular traps. However, current humanized mouse models do not produce these critical immune cell types, limiting their recapitulation of critical host–pathogen interactions that drive the course of disease. Furthermore, as conventional humanized mice retain most murine myeloid cell populations, functional differences between mouse and human neutrophils (15) can interfere with the development of proper human immune responses.

In this paper, Zheng et al. aimed to address this shortcoming by developing a novel derivative of their MISTRG mouse (1). Neutrophil differentiation depends on granulocyte colony-stimulating factor (G-CSF) signaling (16). Therefore, the authors humanized the murine G-CSF locus, yielding MISTRGG. However, while MISTRGG exhibited increased granulopoiesis in the bone marrow, the frequencies of circulating neutrophils were unchanged from MISTRG. Genetic ablation of the murine G-CSF receptor (G-CSFR) to yield MISTRGGR boosted peripheral neutrophil numbers, highlighting that competition between human and mouse G-CSFR interfered with neutrophil differentiation in the bone marrow. Consistently, disruption of mouse G-CSF signaling resulted in increased numbers of neutrophil precursors in the bone marrow of MISTRGGR vs. MISTRGG.

“Flavell and colleagues generated a humanized mouse model capable of supporting human myelopoiesis, including the generation of neutrophils, following transplantation of human hematopoietic stem cells.”

Neutrophils from MISTRGGR displayed functional features similar to polymorphonuclear neutrophils (PMNs) isolated from healthy human donors. MISTRGGR neutrophils were capable of phagocytosing Pseudomonas aeruginosa or Escherichia coli and responded by the characteristic production of ROS and expression of tissue-resident chemokine receptors. Consistently, these neutrophils chemotaxed ex vivo, a critical feature to improve the modeling of human-mediated tissue inflammation in humanized mice. Logically, the authors then assessed neutrophil recruitment in vivo upon local inflammation. While human neutrophils from MISTRG failed to respond to lipopolysaccharide-induced lung inflammation, MISTRGGR neutrophils were robustly recruited into the lung vascular and interstitial compartment. The most significant infiltration—especially of mature hyper-segmented neutrophils—was in the alveolar compartment. Mouse neutrophils were also recruited to the lung of MISTRGGR upon LPS exposure, albeit to a lesser extent than in MISTRG. Consistently, human neutrophil infiltration was considerably higher in the lung of MISTRGGR vs. MISTRG upon intranasal P. aeruginosa infection. Enhanced lung recruitment correlated with an increased ability of MISTRGGR to control bacterial infection as compared with MISTRG. However, although human PMNs are critical drivers of P. aeruginosa clearance in humans, MISTRGGR mice could not clear the bacterial infection.

The skin is one of the first barriers to pathogens and is a major recruitment site for neutrophils upon inflammatory signals and/or microbial exposure (17). The authors validated that MISTRGGR mice recapitulate neutrophil infiltration in the cutaneous compartment upon local inflammatory stimuli. As skin neutrophil infiltration is dependent on integrin-mediated adhesion, unlike in the lung, this experiment suggested that neutrophil recruitment in MISTRGGR is not lung-specific and can be modeled in the context of inflammatory scenarios involving other tissues.

A Humanized Mouse Model to Rule Them All?

MISTRGGR represents a remarkable advancement in the development of humanized mouse models that more faithfully recapitulate human immunological processes. This model has the potential to jumpstart various lines of investigation regarding the role of neutrophils in human disease, from cancer and autoinflammation to infectious diseases. Notably, this mouse model could prove transformative for our understanding of how these cells regulate early innate immune responses to pathogens and can even contribute to disease, as in mosquito-mediated arbovirus infection.

Despite its remarkable potential, the MISTRGGR model still suffers from several drawbacks. Humanized mouse development was initially driven by the need to outcompete mouse hematopoiesis to enhance human hematopoietic engraftment, but the field now faces the opposite challenge. The improvement of humanized mice has come at the price of disrupting mouse hematopoiesis, resulting in homeostatic imbalance and diminished health and life span. The MISTRGGR model is an excellent example of this tradeoff. Disruption of mouse G-CSF signaling increases immunodeficiency and thus, susceptibility to opportunistic pathogens, requiring extremely clean conditions for colony maintenance. More importantly, their limited lifespan (up to 14 wk post engraftment) restricts their potential use for long-term immunological studies such as vaccine or trained immunity studies. Of note, murine neutrophils still developed even when G-CSF signaling was disrupted. As they are recruited to inflamed tissues, residual murine neutrophils may still interfere with accurately modeling human neutrophil-mediated responses to inflammatory stimuli. Antibody-mediated depletion of mouse neutrophils, or similar measures, could alleviate such concerns but may render these mice even more hypersusceptible to their environment.

Although superior to that of MISTRG, the limited ability of MISTRGGR to clear P. aeruginosa infection is a second limitation. As the phagocytic activity of human neutrophils from MISTRGGR was like those of PMNs, ineffective clearance in vivo may reflect functional interference between human and mouse neutrophils in the bone marrow or infected lungs. Alternatively, as the roles of other granulocyte lineages in bacterial clearance and the functional plasticity between neutrophils and eosinophils gains more traction, G-CSF–mediated skewing of human granulopoiesis toward neutrophils could contribute to differential regulation of neutrophil functions in vivo that may affect their phagocytic activity. The absence of human thymus and proper T cell education in MISTRGGR could also impair neutrophil recruitment or function in vivo through partial T cell-mediated signaling. Indeed, neutrophils and Th17 cells can engage in reciprocal recruitment (18). As G-CSF also promotes T cell tolerance and Th2 responses (19), it is unclear how this cytokine may impact T cell-mediated immune crosstalk in MISTRGGR.

These limitations of MISTRGGR may dampen their use for investigating neutrophil functions in a specific disease or immunological context or in the context of multilineage inflammatory responses. This is yet another example of why each biological question ultimately calls for a specifically tailored humanized mouse model. In other words, no humanized mouse model is likely to rule them all.

What’s Next?

Regardless of its limitations, MISTRGGR represents a unique opportunity to further our understanding of neutrophil functions in vivo. Beyond its potential for better modeling immune responses to pathogens, this model could serve as an invaluable resource to i) understand the genetic basis of neutropenia, ii) identify novel neutrophil subsets using single-cell technologies, iii) increase our understanding of the short-lived nature of these cells to develop a renewable source of neutrophils, and iv) delineate the role of neutrophils in cancer. Fig. 1

Fig. 1.

The MISTRGGR mouse model to investigate human neutrophil functions in healthy and disease states. Figure created with Biorender.com.

A key improvement of MISTRGGR would be to recapitulate neutrophil recruitment in a human tissue context, as cross-species recruitment mechanisms could undermine human neutrophil functions. Indeed, as immunodeficient mice can be engrafted with human liver cells, skin explants or lung, coengrafting MISTRGGR could facilitate studying neutrophils in tissue-specific responses. Other avenues of improvement could focus on improving the reconstitution of different granulocyte lineages and/or evaluating the impact of proper T cell education on neutrophil functions, such as through coengraftment of a human fetal thymus.