Efficient differentiation and function of human macrophages in humanized CSF-1 mice

Abstract

Humanized mouse models are useful tools to understand pathophysiology and to develop therapies for human diseases. While significant progress has been made in generating immunocompromised mice with a human hematopoietic system, there are still several shortcomings, one of which is poor human myelopoiesis. Here, we report that human CSF-1 knockin mice show augmented frequencies and functions of human myeloid cells. Insertion of human CSF1 into the corresponding mouse locus of Balb/c Rag2 ؊/؊ ␥c ؊/؊ mice through VELOCIGENE technology resulted in faithful expression of human CSF-1 in these mice both qualitatively and quantitatively. Intra-hepatic transfer of human fetal liver derived hematopoietic stem and progenitor cells (CD34 ؉ ) in humanized CSF-1 (CSF1 h/h ) newborn mice resulted in more efficient differentiation and enhanced frequencies of human monocytes/macrophages in the bone marrow, spleens, peripheral blood, lungs, liver and peritoneal cavity. Human monocytes/ macrophages obtained from the humanized CSF-1 mice show augmented functional properties including migration, phagocytosis, activation and responses to LPS. Thus, humanized mice engineered to express human cytokines will significantly help to overcome the current technical challenges in the field. In addition, humanized CSF-1 mice will be a valuable experimental model to study human myeloid cell biology. (Blood. 2011; 118(11):3119-3128) IntroductionThe development of animal models to study human diseases has significantly advanced our understanding of the underlying mechanisms of several diseases, including cancer. To date, animal models, particularly in the laboratory mouse, have proven to be excellent candidates for the evaluation of the efficiency and efficacy of drugs and therapy options. While the utilization of these surrogate models to study human biology and diseases can be largely justified, because of ethical and technical constrains on the conduct of experimental therapies in humans, much study has highlighted the potential limitations of extrapolating data from mice to humans. 1 It is increasingly evident that there are several differences between the physiology of humans and mice and so, therapies that have been proven successful in mice may not work well in humans.To overcome these issues, there has been a long-standing interest in developing a "humanized mouse" model and intensive work by several groups has successfully demonstrated the feasibility of studying human biology and diseases in mice. Since, a functional and effective immune system in the recipient mouse will result in the elimination of the transplanted tissues/cells of human origin, use of genetic mutants that lack cells of the adaptive immune system such as T, B, and NK cells has significantly contributed to the success of the "humanized mouse" model. Accordingly, the most effective candidates of humanized mouse models include the NOD-SCID and the Balb/c strains that lack genes including recombination activating genes (RAG), common ␥ chain (␥C), beta2 microglobin (B2M), ...

Induction of Functional Human Macrophages from Bone Marrow Promonocytes by M-CSF in Humanized Mice

“…Supporting this hypothesis, expression of human M-CSF in humanized mice by hydrodynamic injection of cytokine-encoding plasmid significantly improved the reconstitution of human monocytes in the peripheral blood (16). Similarly, knockin of human M-CSF gene into the corresponding mouse locus in the recipient mice resulted in a significant increase in the circulating human monocytes, although the increase in tissue macrophages was much less (17). In GM-CSF/IL-3 knockin mice, the level of human macrophages in the lungs was selectively increased without apparent increase in human monocytes in the circulation or macrophages in other tissues, consistent with the requirement for GM-CSF and IL-3 for alveolar macrophage development (18).…”

“…Human GM-CSF, IL-3, and M-CSF genes have been used to replace the corresponding mouse genes in BALB/c Rag2 2/2 gc 2/2 mice. In M-CSF knockin mice, the level of human CD14 + monocytes in the circulation was increased ∼6-fold, reaching 30% of human CD45 + cells, whereas the highest level of tissue macrophage reconstitution was still under 5% of human CD45 + cells (17). In GM-CSF/IL-3 knockin mice, the level of human macrophages in the lungs was increased without apparent increase in human monocytes in the circulation or macrophages in other tissues (17,18).…”

“…In M-CSF knockin mice, the level of human CD14 + monocytes in the circulation was increased ∼6-fold, reaching 30% of human CD45 + cells, whereas the highest level of tissue macrophage reconstitution was still under 5% of human CD45 + cells (17). In GM-CSF/IL-3 knockin mice, the level of human macrophages in the lungs was increased without apparent increase in human monocytes in the circulation or macrophages in other tissues (17,18). In comparison, following expression of M-CSF, circulating monocytes were increased ∼10-fold in humanized mice, and the significant levels of human macrophages were also detected in the spleen, liver, and lungs, reaching 20-70% of human leukocytes in these tissues.…”

“…Following LPS stimulation, low levels (∼300 pg/ml) of human IL-6 and TNF-a were induced in the sera of humanized M-CSF knockin mice (17). In humanized GM-CSF/IL-3 knockin mice, the level of human IL-6 reached ∼6000 pg/ml (18).…”

“…We further show that M-CSF-induced macrophages are mature as they express CD68 and have similar distribution pattern as endogenous mouse macrophages in the liver and lungs, whereas monocytes in the blood did not express CD68 in M-CSF-treated mice, as expected. Compared with the cytokine gene knockin mice where human cytokines were ∼50 pg/ml in the mouse serum (17), the more efficient induction of human monocytes and macrophages following hydrodynamic expression of M-CSF is probably due to the high level of human cytokine that was produced (2.5 ng/ml in serum at 24 h postinjection).…”