Keywords

1 Introduction

The development of humanized mouse models has dramatically expanded the research landscape by facilitating in vivo modeling of many human diseases [1–23]. Different humanized mouse models have been bioengineered for these studies. Differences between humanized mouse models can include the immunodeficient mouse strain manipulated and whether the human hematopoietic stem cell (hHSC) transplant only or the bone marrow-liver-thymus (BLT) humanization strategy was utilized. To achieve the best experimental outcomes possible, it is critical that key attributes of different pairings of a given humanization protocol (e.g., hHSC only or BLT) with specific strains of immunodeficient mice be considered. As an aid in this process, this review compares lymphoid tissue development and human reconstitution among different human reconstitution strategies performed in mouse strains with or without a mouse interleukin (IL)-2Rγ chain. To best characterize the role of the mouse IL-2Rγ chain in lymphoid tissue development and human reconstitution, particular emphasis has been placed on reviewing manuscripts that include direct comparisons between humanized mice generated in immunodeficient mouse strains with or without a mouse IL-2Rγ chain.

2 Immunodeficient Mouse Strains and Humanization Strategies

A key consideration in the generation of humanized mice is the choice of immunodeficient mouse strain. In the mid-1990s, nonobese diabetic (NOD) mice were crossed with severe combined immunodeficient (SCID) mice to develop NOD/SCID mouse strains [24, 25]. Bone marrow engraftment of NOD/SCID mice with human hematopoietic stem cells (NOD/SCID-hHSC) results in a self-renewing source of human B cells, monocyte/ macrophages, dendritic cells , and natural killer (NK) cells that systemically reconstitute the transplanted mouse [26–28]. When planning experiments with NOD/SCID mice, it is important to consider the facts that the endogenous mouse thymus in these animals is vestigial such that it cannot produce human T cells following hHSC engraftment and that the lifespan of these animals is limited due to their predisposition to develop thymic lymphomas [25, 29]. Despite these constraints, the high and sustained levels of human myeloid, dendritic and B cells in NOD/SCID-hHSC have made this model useful for a variety of applications [6, 17, 28]. The lack of human T cells in NOD/SCID-hHSC mice led to the development of a combination approach, which would permit the development of a more complete human immune system (including human T cells) in humanized mice. BLT humanized mice are bioengineered by a bone marrow transplant of hHSC that is paired with the implantation under the renal capsule of autologous human liver and thymus [30, 31]. In BLT mice, the implanted human thymic tissue develops into a human thymus (also referred to as a thymic organoid) with normal human thymic structure and function [30–33]. T cell progenitors migrate to this thymus, mature into human leukocyte antigen (HLA)-restricted human T cells, exit the thymus, systemically populate the BLT mouse and perform effector functions such as HLA-restricted responses to Epstein–Barr virus (EBV) and HIV [31, 34].

The mouse IL-2Rγ chain is an essential component in the receptors for six different cytokines (IL-2, IL-4, IL-7, IL-9, IL-15, IL-21) [35]. Additional immunodeficient mouse strains have been developed that exhibit improved hHSC engraftment by modifying NOD/SCID [36, 37] and Rag2 knockout [38, 39] mice to prevent mouse IL-2Rγ chain function. Two different NOD/SCID IL-2Rγ chain mice have been developed and are designated as NOG [36] and NSG [37] mice. When compared to NOD/SCID mice, NOG and NSG mice are slightly more radiosensitive, do not develop thymic lymphomas, and consequently have a longer lifespan [37]. Both neonate and adult mice lacking the mouse IL-2Rγ chain have been transplanted with hHSC leading to humanization [40–42]. One of the most important differences between mice lacking a functional mouse IL-2Rγ chain and NOD/SCID mice is that the endogenous mouse thymus in NOG, NSG and Rag2γc mice supports human T cell production leading to systemic reconstitution of these mice with human T cells together with human B cells, human monocyte/macrophages, human dendritic cells, and human NK cells [41–43].

3 Primary and Secondary Lymphoid Tissues

Several reports have discussed the human reconstitution of the bone marrow, thymus, and spleen in humanized mice [23, 44, 45]. However, only four published studies report direct comparisons in the human reconstitution of these tissues between mice with and without a functional IL-2Rγ chain [36, 41, 46, 47]. The approach taken for the comparisons in three of these reports was to transplant similar numbers of hHSC into NOG or NSG mice and mice from the isogenic parent NOD/SCID stain [36, 41, 47]. Ito et al., Hiramatsu et al., and McDermott et al. each used NOD/SCID-hHSC and NOG-hHSC mice for their comparison [36, 41, 47]. McDermott et al. also used NSG-hHSC mice [47]. Their consensus findings were that transplantation of equal numbers of hHSC into NOG and NSG mice yielded significantly higher human reconstitution levels in the peripheral blood, bone marrow and spleen relative to NOD/SCID mice [36, 41, 47]. In addition, the production of human T cells by the mouse thymus was confirmed in the NOG-hHSC mice [41, 47] and NSG-hHSC mice [47].

In the fourth report [46], each humanized mouse was generated utilizing an optimized number of hHSC which is very different from the strategy tested in the other three reports. Specifically, NOD/SCID-BLT mice were transplanted with approximately 10 times more hHSC compared to NSG-BLT, NSG-hHSC, and Rag2γc-hHSC. Under these optimized conditions, the absolute numbers of splenic human T cells were not significantly different between these four types of humanized mice and limited differences were observed in the numbers of human T cell in bone marrow [46]. This paper did report that the human thymus present in BLT mice harbors a greater number of human thymocytes compared to a mouse thymus in hHSC transplant only mice [46]. A key point for consideration in the design and interpretation of experiments utilizing NSG-BLT mice is that human thymopoiesis in these animals occurs in both the implanted human thymus and the endogenous mouse thymus with the majority of human thymocytes present in the human thymus [46]. The fact that NSG-BLT mice have two separate locations with ongoing human thymopoiesis suggests that T cells in these mice could be heterogeneous in their restrictions which could complicate some experimental analyses.

Together, the available literature supports the conclusion that the presence or absence of a mouse IL-2Rγ chain does not impact humanization of bone marrow or spleen regardless of whether the hHSC transplant only or BLT method of humanization is utilized (Fig. 6.1). Regarding the endogenous mouse thymus in these strains, the production of human T cells from a mouse thymus only occurs in mice without a functional mouse IL-2Rγ chain [26–28, 40–43]. The development of a human thymus does not depend on the presence or absence of a mouse IL-2Rγ chain. Rather transplant of human tissues, as in BLT mice, is required for development of a human thymus (Fig. 6.1).

Fig. 6.1
figure 1

Relative human reconstitution in the lymphoid organs of different humanized mice. At the bottom of the figure is a key describing the relative level of human reconstitution assigned to each color. In the mice not implanted with human thymus tissue, that organ is listed as not present. NOD/SCIDγc = NOD/SCID IL2R gamma chain deficient mice, Rag2γc = Rag2null IL2R gamma chainnull mice. NOD nonobese diabetic, SCID severe combined immune deficiency, BLT bone marrow-liver-thymus, hHSC human hematopoietic stem cell

Full size image

The mouse IL-2Rγ chain is a component of the IL-7 receptor [35] . This molecule is important, but dispensable, for lymph node organogenesis [48]. Since many humanized mice are bioengineered in mice lacking an IL-2Rγ chain, we examined the literature and found that the presence of lymph nodes in humanized mice has been reported in at least 44 manuscripts [3, 9, 10, 16, 19, 30, 31, 34, 37, 41–43, 46, 49–79]. Specific lymph nodes identified include: cervical [9, 30, 37, 74, 75, 77], mediastenal [75], axillary [9, 30, 37, 76], brachial [30, 37, 76], renal [30]; iliac [3], inguinal [30, 37, 76], mesenteric [3, 9, 19, 30, 41–43, 46, 65–71, 76, 78], and popliteal [79]. The most commonly reported lymph nodes in humanized mice are the mesenteric, cervical, axillary, and brachial. Why lymph nodes in the caudal region of humanized mice are less common has not been determined. The fact that 39 of these 44 manuscripts utilized mouse IL-2Rγ chain deficient mouse strains and still were able to detect lymph nodes [3, 9, 10, 16, 19, 30, 34, 37, 41–43, 46, 49, 50, 52–54, 56–72, 74–78] confirms that a functional mouse IL-2Rγ chain is dispensable for lymph node organogenesis [48] in humanized mice.

One critical function of secondary lymphoid tissues (i.e., spleen and lymph nodes) is to facilitate the development of humoral immune responses . The broad B cell repertoires [80–83] and antigen-specific human antibody responses [9, 34, 54, 55, 73, 84, 85] that have been reported in different types of humanized mice indicates that these animals have some degree of spleen and lymph node function in this regard, although T cell help has clearly not been optimized in these models [57, 81, 85, 86]. Shultz et al. suggested that human T cells in humanized mice may require thymic selection with HLA in order to provide adequate T cell help to B cells for T-dependent antibody production [87], yet T-dependent antibody production has been reported in humanized mice lacking human thymic stroma for T cell education [9, 88]. Furthermore, there is a question of whether lymphoid follicles and germinal centers, anatomical regions that facilitate T cell help, are present within the secondary lymphoid tissues of humanized mice. Lymphoid follicles and germinal center-like structures have been reported to be both present [16, 30, 31, 41, 43, 89] and absent [8, 37, 82] in humanized mice. Beyond direct observation with immunohistochemistry approaches, B cell development and function has been used as a surrogate marker for the presence of lymphoid follicles and germinal centers in humanized mice. A post-germinal center surface phenotype on B cells in humanized mice has been reported [9, 88] while others have identified B cells as maturing independent of a germinal center [68, 85, 90]. In summary, further work is required to reach a consensus regarding the development of lymphoid follicles and germinal centers in humanized mice and to define any role of the mouse IL-2Rγ chain in these processes.

4 Intestinal Lymphoid Tissues

At least 15 manuscripts have examined the presence of human immune cells within intestines of humanized mice [34, 43, 46, 59, 61, 70, 73, 77, 91–97]. For hHSC transplant only mice, the results reported lack both quantitation and a broad analysis of the different human hematopoietic lineages present in the mouse intestines [43, 70, 91–94, 97]. In contrast, several groups have reported that multiple lineages of human immune cells (including T cells, B cells, macrophages, NK cells, and dendritic cells) are distributed through the intestines of BLT mice [34, 73, 77, 95]. To fully characterize the differential intestinal humanization between various humanize mice, one manuscript reported direct, quantitative comparisons of intestinal humanization between NOD/SCID-BLT, NSG-BLT, NSG-hHSC and Rag2γc-hHSC mice [46]. Quantified intestinal cell populations in this report included the lamina propria lymphocytes (LPL) and intraepithelial lymphocytes (IEL) from both the small intestine (SI) and large intestine (LI): SI LPL; SI IEL; LI LPL; and LI IEL. NOD/SCID BLT mice exhibited the highest numbers of human T cells in each of these four intestinal populations [46]. Regardless of the human reconstitution method (i.e., –BLT or –hHSC transplant only), intestinal T cell levels were consistently lower when the immunodeficient mouse strain used lacked a functional IL-2Rγ chain (Fig. 6.1) [46]. These data indicate that the presence of a functional mouse IL-2Rγ chain is critical for robust intestinal human T cell engraftment.

The intestinal immune system is composed of gut-associated lymphoid tissue (GALT) (e.g., Peyer’s patches and isolated lymphoid follicles) and intestinal lamina propria. Many humanized mice are bioengineered in mice lacking an IL-2Rγ chain (a component of the IL-7 receptor) and IL-7 signaling in lymphoid tissue organizer cells and lymphoid tissue inducer cells is indispensible for GALT organogenesis [98–104]. Therefore, we examined the literature for reports describing the presence of GALT in humanized mice. One manuscript reported undefined lymphoid aggregates (e.g., isolated lymphoid follicles) in the intestines of NOD/SCID-BLT mice [73]. These aggregates contained CD4+ T cells , macrophages, and dendritic cells and were susceptible to HIV-induced pathology [73]. Given that the humanized mouse research field is so heavily reliant upon immunodeficient mice without a functional mouse IL-2Rγ chain and there are translational research questions regarding human GALT pathologies that could be studied if humanized mice exhibit GALT structures, defining the presence or absence of GALT in different types of humanized mice is a very important area of future study.

5 Conclusions

Lymphocyte migration into and egress from lymphoid tissues is the result of a series of poorly understood but highly complex species-specific interactions between cell adhesion molecules, integrins, chemokine receptors, and chemokines. Here we discussed the remarkable observation that human lymphocytes traffic in humanized mice such that lymphoid tissues develop and are populated with human cells in these mice. For the bone marrow, thymus, spleen, and to a lesser extent lymph nodes, the presence of a functional mouse IL-2Rγ chain is dispensable for lymphoid tissue development and human reconstitution. However, utilization of the BLT humanization protocol on mice with and without a functional mouse IL-2Rγ chain has provided compelling evidence that this molecule is important for intestinal human reconstitution. Specifically, NOD/SCID-BLT mice have the highest levels of intestinal humanization relative to NSG-BLT mice, as well as to NSG-hHSC and Rag2γc-hHSC mice. Thus, immunodeficient mouse strains with improved hHSC engraftment may exhibit deficiencies in human reconstitution in certain tissues (e.g., intestines). Understanding these distinctions increases our ability to accurately interpret previously published humanized mouse data and to properly plan future experiments.

6 Future Directions

There are two major research areas for future investigation related to human lymphoid tissue development in humanized mice. First, a clear understanding of the cross-species activity for individual cytokines, chemokines, and adhesion molecules would offer critical insights into the optimal combination of human transgenes for appropriate human lymphoid tissue development. There are ongoing efforts by multiple research groups to apply human growth factors exogenously [65, 105–107] or develop mice that transgenically expressing human development and growth factors [108–111]. Yet a comprehensive description of cross-species activity is still lacking [112]. Second, when humanized mice are created by a transplant of hHSC only, all of the human cells present in these mice are hematopoietic in origin. The mice still lack nonhematopoietic cells (e.g., epithelial cells) that are involved in human immune development and function. Pairing current human reconstitution techniques with previously developed strategies for the implantation into mice of human tissues beyond the human thymus in BLT mice (e.g., human bone marrow and human intestines) could introduce additional nonhematopoietic cells into the humanize mouse models [113, 114]. Successful implementation of such strategies would provide the best possible environment for proper lymphoid tissue development in humanized mice.