Tissue-resident lymphoid cells (TLCs) span the spectrum of innate-to-adaptive immune function. Unlike traditional, circulating lymphocytes that are continuously generated from hematopoietic stem cells (HSCs), many TLCs are of fetal origin and poorly generated from adult HSCs. Here, we sought to further understand murine TLC development and the roles of Flk2 and IL7Rα, two cytokine receptors with known function in traditional lymphopoiesis. Using Flk2- and Il7r-Cre lineage tracing, we found that peritoneal B1a cells, splenic marginal zone B (MZB) cells, lung ILC2s and regulatory T cells (Tregs) were highly labeled. Despite high labeling, loss of Flk2 minimally affected the generation of these cells. In contrast, loss of IL7Rα, or combined deletion of Flk2 and IL7Rα, dramatically reduced the number of B1a cells, MZBs, ILC2s and Tregs, both in situ and upon transplantation, indicating an intrinsic and essential role for IL7Rα. Surprisingly, reciprocal transplants of wild-type HSCs showed that an IL7Rα−/− environment selectively impaired reconstitution of TLCs when compared with TLC numbers in situ. Taken together, our data defined Flk2- and IL7Rα-positive TLC differentiation paths, and revealed functional roles of Flk2 and IL7Rα in TLC establishment.
Traditional circulating immune cells are typically defined as either myeloid or lymphoid and generated from hematopoietic stem cells (HSCs). Myeloid cells are involved in rapid, broad response innate immunity whereas traditional lymphoid cells are involved in slow, specific adaptive immunity (Bonilla and Oettgen, 2010; Netea et al., 2011; Beaudin and Forsberg, 2016; Cool and Forsberg, 2019). The adaptive immune response relies on the complex B and T cell receptor repertoire generation. Although this dichotomy of immune cells is clear for circulating cells, it is unclear whether tissue-resident immune cells squarely belong to the myeloid or lymphoid lineage, with many recent studies alluding to complex and dynamic origins (Beaudin and Forsberg, 2016; Beaudin et al., 2016; Leung et al., 2019). Unlike circulating immune cells that traffic to non-lymphoid organs upon activation, tissue-resident immune cells reside in non-lymphoid organs, do not recirculate and have specialized functions that span the innate-adaptive spectrum (Davies et al., 2013; Chou and Li, 2018). For example, B1a cells in the peritoneal cavity are considered B cells (Kantor and Herzenberg, 1993). However, they do not undergo the same B cell receptor selection process as circulating B cells and are thus deemed an ‘innate-like’ lymphoid cell. Although recent studies have rapidly propelled the understanding of tissue macrophage specification and function (Blériot et al., 2020; Martin and Gurevich, 2021), it is unclear whether tissue-resident lymphoid cell (TLC) differentiation is orchestrated similarly to circulating lymphoid cells; here, we sought to determine whether their differentiation is regulated by classic lymphoid genes.
We were particularly interested in the cytokine receptors Flk2 (also known as Flt3) and IL7Rα because they are required for traditional adult lymphopoiesis. This finding was demonstrated by impaired lymphopoiesis in both Flk2−/− and Il7rα−/− mice and supported by Flk2-Cre and Il7rα-Cre driven lineage tracing of cells with increasingly restricted lymphoid potential (Peschon et al., 1994; Kikuchi et al., 2008; Beaudin et al., 2014; Leung et al., 2019) (Fig. 1A,B,C). As previously described (Boyer et al., 2011, 2012; Beaudin et al., 2016; Leung et al., 2019; Cool et al., 2020), these models were generated by crossing mice expressing Flk2-Cre (Benz et al., 2008) or Il7r-Cre (Schlenner et al., 2010) to mTmG mice expressing a dual-color fluorescent reporter (Muzumdar et al., 2007), creating the ‘FlkSwitch’ and ‘Il7rSwitch’ models (Fig. 1A). In both models, cells express Tomato (Tom) until Cre-mediated recombination results in the irreversible switch to GFP expression by that cell and all of its downstream progeny (Fig. 1B,C). We previously demonstrated that Il7r-Cre labeling is not constrained to lymphoid cells. Surprisingly, in the ‘Il7rSwitch’ lineage tracing model, fetally derived adult tissue-resident macrophages (trMacs), were highly labeled and IL7Rα was required for early trMac development (Leung et al., 2019), whereas Flk2-Cre did not efficiently label trMacs (Hoeffel et al., 2015; Hashimoto et al., 2013; Epelman et al., 2014; Leung et al., 2019). We have also previously examined lung eosinophils, another myeloid cell type; interestingly, despite the high Flk2-Cre and minimal Il7r-Cre labeling of eosinophils, they depend on cell extrinsic IL7Rα, but not Flk2 (Cool et al., 2020). The development and generation of these tissue-resident myeloid cells evidently does not follow that of other traditional circulating myeloid cells. Similarly, many innate-like lymphoid cells are thought to arise from fetal progenitors (Beaudin and Forsberg, 2016) and it remains unclear whether their development follows traditional lymphoid paths and whether they are functionally regulated by known lymphoid drivers. For example, there is evidence that many TLCs arise via common lymphoid progenitor (CLP)-independent pathways (Ghaedi et al., 2016), and there is differential requirement for Flk2 and IL7Rα amongst different hematopoietic cell types (Sitnicka et al., 2007; Beaudin et al., 2014). To determine whether Flk2 and IL7Rα are involved in TLC development, we employed lineage tracing, germline knockouts and HSC transplantation assays.
Flk2-Cre and Il7r-Cre highly label tissue-resident lymphoid cell populations
To test whether TLCs arise via differentiation pathways similar to circulating lymphocytes, we examined labeling of TLCs in the FlkSwitch and Il7rSwitch lineage tracing models (Fig. 1A-C). We compared reporter expression of traditional peripheral blood B cells with reporter expression of B1a cells isolated from the peritoneal cavity, marginal zone B cells (MZB) from the spleen, type 2 innate lymphoid cells (ILC2) from the lung and regulatory T cells (Treg) from the lung (Fig. 1D-H). As expected, and previously reported, Cre-driven labeling of circulating peripheral blood B cells was greater than 95% in both FlkSwitch and Il7rSwitch mice (Fig. 1D) (Schlenner et al., 2010; Leung et al., 2019). Similarly, Cre-driven labeling was uniformly high for all TLCs examined, a trend we also observed in B1b and B2 cells (Fig. S1A). Together, our data showed that B1a cells, ILC2, MZB and Tregs arise from Flk2- and IL7Rα-positive cells, as do traditional circulating lymphoid cells.
Tissue-resident lymphoid cells are severely reduced in the absence of IL7Rα, but not Flk2
We and others have previously shown that loss of Flk2 results in a reduction of hematopoietic progenitor cells, with a less severe reduction in mature B cells in the peripheral blood (Mackarehtschian et al., 1995; Sitnicka et al., 2007; Jensen et al., 2008; Beaudin et al., 2014). The loss of IL7Rα has also been shown to result in a severe reduction of peripheral blood B cells (Peschon et al., 1994). Here, we observed similar trends in the lung, in which traditional CD19+ B cells, although not significantly reduced in Flk2−/− mice, were significantly reduced in the IL7Rα−/− mice (Fig. 2A,B). Similar trends were observed in peripheral blood CD3+ T cells (Fig. S2A). To determine whether TLCs were similarly affected, we quantified their cellularity in the absence of Flk2 and IL7Rα. Compared with wild-type (WT) numbers of each cell type, only MZBs were significantly reduced in the absence of Flk2 (Fig. 2D), whereas B1a cells, ILC2s and MZBs were all significantly reduced in the absence of IL7Rα (Fig. 2C-E), consistent with previous studies of these cells in the same and other tissues, and peritoneal B1b and B2 cells (Fig. S1B,B′) (Hesslein et al., 2006; Ghosn et al., 2012; Patton et al., 2014; Robinette et al., 2017). Interestingly, despite being as efficiently labeled by both Flk2-Cre and IL7r-Cre (Fig. 1H), Tregs were not significantly reduced in either Flk2−/− or IL7Rα−/− mice, although there was a downward trend (Fig. 2F), as previously seen in thymic and splenic Tregs of IL7Rα−/− mice (Bayer et al., 2008). We hypothesized that Flk2 may compensate for the lack of IL7Rα, and vice versa, allowing for near-normal Treg development in the absence of either receptor. To test this hypothesis, we generated Flk2, IL7Rα double knockout (FIDKO) mice and quantified total Tregs in the lung. We found a significant reduction of Tregs in the lung of FIDKO mice compared with WT mice (Fig. 2F), revealing an overlapping and cell-specific role of Flk2 and IL7Rα. We also quantified B1as, ILC2s and MZBs in the FIDKO mice and found that they recapitulated the severe reductions in numbers that we found in the IL7Rα−/− mice. These data suggest that both Flk2 and IL7Rα do have functional roles in TLC development during steady-state hematopoiesis, although IL7Rα appears to be more important than Flk2, as indicated by the more severe reduction in cell numbers.
Flk2−/− and IL7Rα−/− HSCs have impaired tissue-resident lymphoid cell reconstitution
Hematopoietic stem and progenitor cell (HSPC) differentiation relies on cytokine receptors such as Flk2 and IL7R to receive instructive signals from the environment (Zhang and Lodish, 2008). For example, we have previously shown that Flk2 is intrinsically required on HSPCs for both myeloid and lymphoid reconstitution, as demonstrated by the reduced capacity of Flk2−/− HSCs to generate myeloid and lymphoid progeny, including traditional circulating lymphoid cells, upon transplantation into a WT host (Beaudin et al., 2014). More recently, we demonstrated an extrinsic requirement of IL7Rα in the generation of eosinophils from WT HSCs, as demonstrated by a reduced capacity of WT HSCs to generate eosinophils upon transplantation into an IL7Rα−/− host (Cool et al., 2020). Therefore, we were curious whether the deficiency in TLCs we observed in Flk2−/−, IL7Rα−/− and FIDKO mice (Fig. 2) is due to the cell-intrinsic lack of receptors or due to changes to surrounding cells. To begin to answer this question, we transplanted WT, Flk2−/−, IL7Rα−/− or FIDKO HSCs into sublethally irradiated WT hosts (Fig. 3A). Under such transplantation conditions, host cells are partially, but not completely, depleted (Boyer et al., 2019), leading to robust and long-term HSC engraftment (Fig. S3). We first examined donor chimerism of circulating lymphoid cells. As expected, peripheral blood B cell donor chimerism was significantly lower from Flk2−/− donor HSCs, IL7Rα−/− HSCs and FIDKO HSCs compared with WT HSCs (Fig. 3B); similar results were observed for peripheral blood T cells (Fig. S2B), whereas granulocyte/macrophage (GM) reconstitution was significantly impaired only from Flk2−/− donor HSCs (Fig. S3A). When examining TLCs, we observed only donor chimerism of IL7Rα−/− and FIDKO HSCs was significantly impaired, whereas Flk2 deficiency alone did not result in significant differences (Fig. 3C-F′; Fig. S1C,C″). A limitation of donor chimerism is that this measure relies on host cell numbers and therefore makes it more difficult to determine the intrinsic requirement of these receptors because the number of host cells changes dynamically over time post-conditioning (Boyer et al., 2019). To overcome this limitation, we quantified the absolute cellularity of donor-derived cells with a bead-based method that allows for simultaneous flow cytometric analysis and counting, which is purely a measure of donor cells (Boyer et al., 2019; Rajendiran et al., 2020; Poscablo et al., 2021). Quantification of absolute cellularity of donor-derived cells between WT and knockout HSCs revealed that reconstitution of B1a cells (Fig. 3C′) and MZBs (Fig. 3D′) was also significantly impaired by the loss of Flk2 alone. Regardless of quantification method, deletion of either IL7Rα alone, or both IL7Rα and Flk2, led to impaired reconstitution of all TLCs examined (Fig. 3C′-F′). This was intriguing as we did not observe significant reduction of Tregs in IL7Rα−/− at steady state (Fig. 2F). This is likely due to the stress of transplantation, which reveals more dynamic requirement of IL7Rα – we previously observed a similar trend when we compared GM at steady state in Flk2−/− mice to transplantation of Flk2−/− HSCs (Beaudin et al., 2014). These data suggest that IL7Rα is required cell intrinsically for TLC reconstitution, that Flk2 cannot compensate for this requirement and that IL7Rα is not capable of fully compensating for the loss of Flk2.
WT HSCs have enhanced tissue-resident lymphoid cell reconstitution capacity in an IL7Rα−/− environment
Although the HSC transplantations of Fig. 3 demonstrated a cell intrinsic requirement of IL7Rα for TLC reconstitution, it is also possible that extrinsic IL7Rα may be required, as we previously demonstrated for eosinophil reconstitution (Cool et al., 2020). To test this, we performed transplantation of WT HSCs into IL7Rα-deficient mice (Fig. 4A). HSC engraftment in the bone marrow (BM) was similar between WT and IL7Rα−/− recipients (Fig. S3B), consistent with the lack of IL7R expression and function in HSCs. Surprisingly, we observed a significantly greater number of donor-derived B1a cells, MZBs, ILC2s and Tregs (Fig. 4B-E′), as well as B1b and B2 cells (Fig. S1D,D″), but not circulating T (Fig. S2C) or myeloid cells (Fig. S3C), in the IL7Rα−/− hosts compared with WT hosts. This raised the question of whether the greater number of donor-derived cells in the IL7Rα−/− mice was reflective of an overall greater number of cells, or whether the reconstitution by WT HSCs only compensated for the lower cell numbers of cells in the IL7Rα−/− host. Therefore, we determined the total number of host- and donor-derived cells and compared these numbers between WT hosts and IL7Rα−/− hosts. We found no significant differences between WT and IL7Rα−/− recipients of total B1a (Fig. 4B′) and MZB (Fig. 4C′) cells, B1b and B2 cells (Fig. S1D′,D‴) or peripheral blood T cells (Fig. S2C′). Interestingly, we observed significantly greater total numbers of ILC2s (Fig. 4D′) and Tregs (Fig. 4E′) in the IL7Rα−/− recipients compared with the WT hosts. Designation of each individual across experiments failed to reveal evidence of sex-specific trends or differences (Fig. S4A-C′), despite a previous report of differential KLRG1 receptor expression in ILC2s in males and females (Kadel et al., 2018). Thus, the enhanced reconstitution of tissue-resident T lymphocytes in an IL7Rα−/− environment compared with consistent reconstitution of tissue-resident B lymphocytes suggests differential dependence on IL7/IL7R signaling in the development of these cell types (Schluns et al., 2000; Guimond et al., 2009; Osborne et al., 2011).
WT HSCs transplanted into IL7Rα−/− mice are not fully capable of rescuing the impaired lymphoid phenotype
The majority of TLCs are thought to arise pre/perinatally, with little contribution from adult progenitors. Although we observed enhanced reconstitution of some TLCs by adult WT HSCs in IL7Rα−/− hosts, it remained unclear whether the lymphoid deficiencies observed in the IL7Rα−/− mice (Fig. 2) were fully rescued by transplantation of adult WT HSCs. Therefore, we compared steady-state numbers of circulating lymphocytes and TLCs in both WT and IL7Rα−/− mice (Fig. 2, solid bars) with total numbers after long-term hematopoietic reconstitution post-transplantation of WT HSCs (Fig. 4B′-E′). In WT and IL7Rα−/− mice transplanted with WT HSCs, we found no significant difference in total peripheral blood B cells or T cells, or peritoneal B1b and B2 cells, compared with WT mice at steady state (Fig. 5A; Figs S2D, S1E,E′). Similarly, we observed no difference between total numbers of MZBs in WT mice at steady state and WT mice transplanted with WT HSCs (Fig. 5C, solid gray bar compared with gray lined bar). Therefore, WT HSCs were indeed capable of fully reconstituting these cells. In fact, although the difference in total MZBs was not significantly different between WT and IL7Rα−/− recipients (Fig. 4C′), we observed significantly more MZBs in the IL7Rα−/− recipient mice compared with WT steady state (Fig. 5C, solid gray bar compared with red lined bar). Interestingly, this was not the case for the other cell types examined. We observed significantly fewer B1a cells, ILC2s and Tregs in mice transplanted with WT HSCs compared with steady-state WT mice (Fig. 5B,D,E, solid gray bars compared with gray striped bars), consistent with reports by us and others of limited TLC potential of adult progenitors (Kikuchi and Kondo, 2006; Beaudin et al., 2016; Schneider et al., 2019). Despite the enhanced reconstitution of ILC2 and Tregs observed in the IL7Rα−/− environment (Fig. 4D′-E′), we were surprised to find that this was not sufficient to return cell numbers to those observed in WT steady-state mice; rather, the total number of cells was not greater than the total numbers observed at steady state in the IL7Rα−/− mice (Fig. 5D,E, solid red bars compared with red lined bars). Taken together, these data suggest that adult WT HSCs transplanted into IL7Rα−/− mice are not fully capable of rescuing the impaired lymphoid phenotype caused by IL7Rα deletion and that there is differential requirement for IL7Rα on the development of tissue-resident B and T cells.
Fetal HSCs outcompete adult HSCs in TLC reconstitution
Our results in Fig. 3 clearly demonstrate that adult HSCs can contribute robustly to all four subsets of TLCs that we investigated. As referenced above, there is evidence for a fetal origin of some TLCs (Mold et al., 2010; Beaudin et al., 2016; Schneider et al., 2019; Azevedo Portilho et al., 2021) and the relative contribution of fetal and adult HSCs to TLCs is unclear. There is particular controversy over the capacity of adult HSCs to contribute to B1a cells (Beaudin and Forsberg, 2016), as studies have reported opposing findings (Montecino-Rodriguez et al., 2016; Ghosn et al., 2016; Kristiansen et al., 2016; Sawai et al., 2016). Therefore, we sought to determine the reconstitution capacity of both fetal and adult HSCs to generate TLCs. We performed competitive transplants of equal numbers of fetal and adult HSCs to allow for direct comparison of TLC reconstitution in the same host (Fig. 6A). We were surprised that fetal HSCs almost completely outcompeted adult HSCs, contributing significantly more to donor chimerism of B1a cells and MZBs (Fig. 6B) despite demonstrating reliable adult HSC reconstitution of B1a cells and MZB cells when transplanted without competing fetal HSCs (Fig. 3C,D). These findings were also observed for peritoneal B1b and B2 cells (Fig. S5A), and peripheral blood GM, B and T cells (Fig. S5B). Interestingly, ILC2 and Treg donor chimerism was not as drastically different between fetal and adult HSC compared with B1a and MZB (Fig. 6B). To determine whether the difference we observed in TLC reconstitution was due to HSC engraftment, we analyzed HSPCs in the bone marrow of recipient mice. We found no significant differences in donor chimerism of fetal-derived HSCs compared with adult derived HSCs, or common myeloid progenitors (CMPs) (Fig. 6C). In contrast, we did observe significantly greater donor chimerism of fetal-derived CLPs compared with adult-derived CLPs. These data suggest that fetal reconstitution of B1a and MZB might rely heavily on CLP-dependent pathways of differentiation, whereas adult reconstitution of ILC2 and Tregs may differentiate via alternate, CLP-independent (but IL7Rα-dependent) pathways, as suggested previously (Ghaedi et al., 2016).
Cell-intrinsic IL7Rα is required for differentiation of tissue-resident lymphoid cells
In this study, we aimed to expand our understanding on the roles of the cytokine receptors Flk2 and IL7Rα beyond traditional, circulating lymphocytes by investigating their roles in TLC development and reconstitution. Using lineage tracing, knockout mouse models and transplantation assays, our data demonstrated a clear role of IL7Rα in the differentiation of tissue-resident lymphoid cells across multiple tissues. B1a cells, MZBs, ILC2 and Tregs all displayed very high labeling in both the FlkSwitch and Il7rSwitch lineage tracing models; however, only IL7Rα−/− resulted in severely impaired TLC development, whereas the effects of Flk2−/− were minimal. Furthermore, IL7Rα−/− HSCs were incapable of reconstituting TLCs in WT hosts, whereas WT HSCs were capable of TLC reconstitution in IL7Rα−/− hosts, revealing a primarily cell-intrinsic requirement of IL7Rα for TLC differentiation. Interestingly, although adult HSCs were clearly capable of reconstituting all TLCs investigated, reconstitution of TLCs by fetal HSCs was notably greater compared with adult HSCs upon competitive transplantation.
Flk2 is less essential for tissue-resident lymphocytes than for circulating lymphocytes
Based on the high labeling in the FlkSwitch model (Fig. 1) and our previous finding that multipotent, myeloid and lymphoid progenitors, and peripheral blood B cells were reduced in Flk2−/− mice (Beaudin et al., 2014), we expected to find significantly reduced numbers of TLCs in Flk2-deficient mice. We were surprised that only MZBs were significantly reduced in the Flk2−/− compared with WT mice (Fig. 2; Fig. S2A). We have also previously shown that multilineage reconstitution was impaired upon transplantation of Flk2−/− HSCs in WT recipients, likely due to the significant decrease in HSC differentiation into Flk2+ multipotent progenitors (MPPs), and CLPs (Beaudin et al., 2014). Despite demonstrating that B1a cells, MZBs, ILC2 and Tregs differentiate via a Flk2-positive stage (Fig. 1), we were surprised that, amongst TLCs, only B1a and MZB reconstitution was impaired by the loss of Flk2 (Fig. 3). This may be because of a more stringent reliance on Flk2 for innate-like B cells rather than innate-like T cells, or reported CLP-independent differentiation of ILCs and T cells (Ghaedi et al., 2016). The notion of partial CLP-independence for ILC2 and Treg differentiation is also supported by the fetal and adult HSC reconstitution data, as we observed no significant differences in fetal and adult contribution to ILC2s and Tregs (Fig. 6B) despite a significant reduction in adult-derived CLPs (Fig. 6C). Interestingly, previous studies have shown that Tregs increase in number in response to the Flk2 ligand in mice and humans (McGee et al., 2010; Klein et al., 2013), making it all the more surprising that Tregs are the least affected by the loss of Flk2. These data suggest that IL7Rα is able to compensate for the loss of Flk2 in a cell-type specific manner but not vice versa, and therefore plays a more essential role in TLC reconstitution.
Absolute cell quantification and competitive HSC transplantation revealed unexpected reconstitution capability
In all transplantation experiments, we employed an absolute cell quantification method previously developed in the lab (Boyer et al., 2019; Leung et al., 2019; Cool et al., 2020; Poscablo et al., 2021). This method was crucial to our interpretation of WT HSC reconstitution capability in IL7Rα−/− recipients because it cannot be gleaned from percent donor chimerism alone. As the IL7Rα−/− recipients were already devoid of any of the cells we were examining, donor-derived cells would be constituting 100% donor chimerism. By quantifying absolute cell numbers, we found that not only did WT HSCs generate TLCs in an IL7Rα−/− recipient, but that the overall numbers of ILC2 and Tregs were greater in the IL7Rα−/− recipients (Fig. 4). We suspect this is owing to the abundance of IL7 in the IL7Rα−/− mice (Martin et al., 2017; Cool et al., 2020) and lack of host competition for IL7, which may in turn enhance the ILC2 and Treg reconstitution. Therefore, IL7Rα may have a cell non-autonomous role in regulating TLC subsets that rely more heavily on IL7 signaling by controlling IL7 levels (Martin et al., 2017; Cool et al., 2020). In addition, in our previous examination of eosinophil reconstitution in IL7Rα−/− mice, we discovered that the concentrations of other cytokines involved in immune cell survival are different in IL7Rα−/− compared with WT mice (Cool et al., 2020). Therefore, we speculate that differences in cytokine signals may play a significant role in the differences we observed in Fig. 4. Testing of a number of candidates did not uncover significant upregulation of lymphoid-promoting factors, leaving the previously documented increase in IL7 levels the sole current candidate for the increased reconstitution in IL7Rα−/− mice over WT recipients.
We were also able to compare total TLCs reconstituted by WT HSCs to their total numbers in situ at steady state. With the exception of MZBs in the IL7Rα−/− host, WT HSCs in either the WT host or the IL7Rα−/− host did not fully rescue TLCs to WT steady-state levels (Fig. 5). This could conceivably be because we performed these transplants with adult HSCs and at least some types of TLCs are thought to originate primarily from fetal/neonatal progenitors. Recent work from Locksley and colleagues reported that ILC2s from the lung originate from perinatal progenitors (Schneider et al., 2019), and for decades it has remained controversial to what extent adult HSCs contribute to B1a cells. Here, we found robust B1a cell reconstitution from adult HSCs when transplanted alone (Fig. 3B), but this capacity was drastically overshadowed by fetal HSCs when transplanted competitively (Fig. 6B). Importantly, the fetal versus adult HSC competitiveness did not segregate by cell types often designated as fetal- versus adult-derived, as fetal HSCs also contributed significantly more to other, adult-derived peritoneal and circulating immune cells (Fig. S5). Mechanistically, this was not because of differences in HSC engraftment (Fig. 6C), but more likely related to greater proliferative and lymphoid capacity of fetal relative to adult HSCs, also manifested by significant differences in CLP reconstitution and a trend towards more fetal-derived CMPs (Fig. 6C). These results could be explained by the presence of developmentally restricted HSCs (drHSCs) within the fetal, but not adult, HSC compartment (Beaudin et al., 2016).
Although co-transplantation allowed us to more directly compare HSC reconstitution, our results show that competitive strategies may also obscure certain potential because one HSC type can outcompete the other in one or more lineages. This calls for close attention to the conditions of the assays used to claim reconstitution potential or lack thereof. The absolute quantification comparison may be a better indicator of the ability of adult HSCs to restore homeostasis, as these results revealed impairment in putative fetal-derived B1a, ILC2 and Tregs (Fig. 5B,D,E) but not adult-derived peritoneal or peripheral blood cells (Fig. 5A; Figs S1E,E′, S2D). These new perspectives will be valuable for more completely unraveling the cellular and molecular mechanisms that control TLC potential of fetal and adult HSCs; these clearly deserve further investigation, considering that HSCs are used in the clinic to treat a multitude of immune deficiencies.
IL7Rα is required for both innate and adaptive immunity
Despite some TLCs displaying immune properties reminiscent of myeloid cells (Martin et al., 2001; Gunn and Brewer, 2006; Ha et al., 2006; Drake and Kita, 2014; Panda and Colonna, 2019), we demonstrated here that they nevertheless rely on main regulators of traditional circulating lymphoid cells (Fig. 7). The regulation of immune cell development by Flk2 and IL7R is highly complex, driven by partially overlapping expression and function of these receptors on hematopoietic progenitor cells (Fig. 7A) (Mackarehtschian et al., 1995; Christensen and Weissman, 2001; Schlenner et al., 2010; Boyer et al., 2011). In addition, we found that HSC competition can greatly influence the perceived ability to reconstitute hematopoiesis: when transplanted alone, adult HSCs provided robust TLC contribution (Fig. 3), which was almost completely overshadowed by fetal HSC-derived TLCs in competitive transplantations (Fig. 6). This may lead to reassessment of ‘fetal-derived’ cells and how to test and interpret the reconstitution potential of progenitor cells. Interestingly, IL7Rα regulation of TLC generation appeared to be predominantly cell intrinsic, similar to that of fetal-specified trMacs (Leung et al., 2019), but in contrast to the solely cell extrinsic role we discovered for IL7Rα in adult myeloid eosinophil regulation (Cool et al., 2020). The combination of cell intrinsic and extrinsic regulation by Flk2 and IL7R, and the dynamic secretion of their corresponding cytokines Flk2L and IL7 under different conditions, support a model of cell interdependence and complex, tissue-specific feedback (Fig. 7B) (from Figs 1-5, Figs S2,S3 and Beaudin et al., 2014; Leung et al., 2019; Cool et al., 2020). This interconnectedness can be further unraveled by future tissue-specific deletions of Flk2, Flk2L, IL7Ra, IL7 and/or other cytokines that selectively promote immune development. Collectively, these data indicate broad and temporally dynamic functional roles of both Flk2 and IL7Rα beyond the typical lymphoid ascribed roles.
MATERIALS AND METHODS
All animals were housed and bred in the Association for Assessment and Accreditation of Laboratory Animal Care International (AAALAC)-accredited vivarium at University of California, Santa Cruz (UCSC) and group housed in ventilated cages on a standard 12 h light/12 h dark cycle. All procedures were approved by the UCSC Institutional Animal Care and Use (IACUC) Committees (OLAW assurance A3859-01; USDA Registration 93-R-0439, customer number 9198). Il7r-Cre (Schlenner et al., 2010) and Flk2-Cre (Benz et al., 2008) mice, obtained under fully executed Material Transfer Agreements, were crossed to homozygous Rosa26mTmG females (JAX Stock #007576, The Jackson Laboratory) (Muzumdar et al., 2007) to generate ‘switch’ lines, all on the C57BL/6 background. WT C57BL/6, Rosa26mTmG, UBC-GFP (JAX Stock #004353, The Jackson Laboratory) (Schaefer et al., 2001) and KuO (Hamanaka et al., 2013) mice were used for controls. FIDKO mice were generated by crossing Flk2−/− (Mackarehtschian et al., 1995) and IL7Rα−/− (Peschon et al., 1994) mice to homozygosity. Adult male and female mice (8-12 weeks old) were used randomly and indiscriminately, with the exception of the FlkSwitch line, in which only males were used because many female mice do not carry a Cre allele.
Tissue and cell isolation
Mice were sacrificed by CO2 inhalation. Adult lung and spleen were isolated, weighed, then placed into 1.5 ml of digestion buffer [1× PBS(+/+) with 2% fetal bovine serum, 1 mg/ml (lung) or 2 mg/ml (spleen) collagenase IV (Gibco) with 100 U/ml DNase1] containing Calibrite APC-labeled (BD Biosciences) counting beads and manually disassociated into 2 mm×2 mm pieces using surgical scissors. The tissues were then incubated at 37°C for 1 h (lung) or 2 h (spleen). Following incubation, all tissues were passaged through 19 g then a 16 g needle 10 times, and then filtered through a 70 µM filter and 25 ml of cell staining buffer (1× PBS with 2% fetal bovine serum and 5 mM EDTA) was added to quench the digestion enzymes.
Flow cytometry and cell analysis
Cell labeling was performed on ice in 1× PBS with 5 mM EDTA and 2% fetal bovine serum. Fluorescence-activated cell sorting (FACS) analysis was performed on a BD FACS Aria III (BD Biosciences) at UCSC, and analyzed using FlowJo (Treestar) (Boyer et al., 2019; Smith-Berdan et al., 2019; Martin et al., 2021; Poscablo et al., 2021). Cells were defined as follows: B1a=Lin− (Ter119, CD11b, CD3) IgM+ CD5+ CD11bmid; MZB=Lin− (CD3, CD4, CD8, Ter119, Gr1) B220+ IgM+ CD21+ CD23−; ILC2=Lin− (CD3, CD4, CD5, CD8, NK1-1, CD19, Ter119, F4/80, FcεRIα) CD45+ KLRG1+ Sca1+ CD25+; Treg=Lin− (Ter119, Gr1) CD3+ CD4+ TCRB+ CD25+ FOX-P3+.
Transplantation assays were performed as previously described (Beaudin et al., 2014; Smith-Berdan et al., 2015; Ugarte et al., 2015; Beaudin et al., 2016). Briefly, double sorted HSCs were isolated from bone marrow of either WT (non-fluorescent, mTmG, UBC-GFP or KuO), Flk2−/−, Il7Rα−/− or FIDKO 8-12 week old mice and from embryonic day (E)14.5 fetal livers. HSCs were defined as KLS (cKit+Lin−Sca1+), CD150hiFlk2−. Lineage markers were CD3, CD4, CD5, CD8, B220, Mac1, Gr1 and Ter119. Mac1 antibodies were excluded from the lineage cocktail when sorting fetal progenitors (Morrison et al., 1995). Flk2 could not be used to identify HSCs in Flk2−/− and FIDKO HSCs and so they were sorted on KLS, CD150hi. Recipient mice aged 8-12 weeks were sublethally irradiated (750 rad, single dose) with a Faxitron CP-160. Under isofluorane-induced general anesthesia, sorted cells were transplanted retro-orbitally. Recipient mice were bled 4, 8, 12 and 16 weeks post-transplantation via tail vein and peripheral blood was analyzed for donor chimerism by means of fluorescence profiles and antibodies to lineage markers (Table S1). Long-term multilineage reconstitution was defined as chimerism in both the lymphoid and myeloid lineages of >0.1% at 16 weeks post-transplantation and only mice that displayed long-term reconstitution were used for post-transplantation analysis.
Absolute cell number quantification
A known volume of peripheral blood was mixed with an antibody solution, (PBS, 5 mM EDTA and 2% fetal bovine serum) containing a known quantity of Calibrite APC beads before flow cytometry analysis (Boyer et al., 2019; Cool et al., 2020; Rajendiran et al., 2020; Poscablo et al., 2021). For tissues, a known quantity of beads was added at the very beginning of tissue preparation before antibody staining and analysis. The number of beads counted by flow cytometry was used to calculate the number of mature cells per ml of blood or within each tissue.
Quantification and statistical analysis
Number of experiments, n, and what n represents can be found in the legend for each figure. Power analysis to determine the minimum size of experimental groups was performed with a confidence level of 0.95 and desired power of 0.8 using an online tool (https://epitools.ausvet.com.au/). Statistical significance was determined by two-tailed unpaired Student's t-test when comparing two groups and by one-way ANOVA and Dunnett multi-parameter test when comparing across multiple groups. In Figs 2-4, 6 and Figs S2A-C′, S3-S5, all comparisons are being made to WT controls, and not across all groups despite being graphed together. All data are shown as mean±s.e.m. representing at least two independent experiments.
We thank Dr I. Lemischka for the Flk2−/− mice; Drs H.-R. Rodewald and S. M. Schlenner for the IL7Rα-Cre strain; Dr T. Boehm for the Flt3Cre strain; Bari Nazario and the UCSC Institute for the Biology of Stem Cells for flow cytometry support.
Conceptualization: A.K.W., A.E.B., E.C.F.; Methodology: A.K.W.; Formal analysis: A.K.W.; Investigation: A.K.W., T.C., D.M.P., A.H.; Writing - original draft: A.K.W.; Writing - review & editing: A.K.W., T.C., D.M.P., E.C.F.; Visualization: A.K.W.; Supervision: E.C.F.; Funding acquisition: E.C.F.
This work was supported by a National Institutes of Health/National Institute of Diabetes and Digestive and Kidney Diseases award (R01DK100917) and an American Asthma Foundation Research Scholar award to E.C.F.; by Tobacco-Related Disease Research Program (TRDRP) Predoctoral Fellowships to A.K.W. and T.C.; by a National Heart, Lung, and Blood Institute F31 Predoctoral Fellowship to A.K.W.; by American Heart Association and Howard Hughes Medical Institute Gilliam Fellowships to D.M.P., and by California Institute for Regenerative Medicine Facilities awards CL1-00506 and FA1-00617-1 to UCSC, RRID:SCR_021149. Open Access funding provided by University of California, Santa Cruz. Deposited in PMC for immediate release.
Peer review history
The peer review history is available online at https://journals.biologists.com/dev/article-lookup/doi/10.1242/dev.200139
The authors declare no competing or financial interests.