The aged bone marrow environment restrains HSC engraftment
Our initial investigation aimed to evaluate the efficacy of CD45-SAP conditioning7 in the context of aged hosts. Following administration of CD45-SAP (3 mg/kg) to young (2 months) and aged (16 months) C57BL/6-CD45.2 mice and analysis 8 days later (Fig. 1a), we observed only marginal changes in overall peripheral blood (PB) white blood cell (WBC) counts in both groups (Fig. 1b). More detailed assessments revealed transient reductions in platelets and hemoglobin of CD45-SAP-treated mice (Extended Data Fig. 1a) and some noticeable changes in WBC distribution, with reduced lymphocyte and elevated myeloid cell counts (Extended Data Fig. 1b).
While splenic cellularity remained relatively constant following CD45-SAP-treatment, more evident reductions were observed in thymic cellularity (Fig. 1c). The overall BM cellularity was also relatively unchanged (Fig. 1d, left), while in agreement with other studies7,23, we observed a pronounced decrease in the numbers of HSCs (Fig. 1d, right). In young mice, the effects on other multipotent and more lineage-restricted BM progenitors varied in a cell-type-specific manner, but many of these changes were less pronounced in aged mice (Extended Data Fig. 1c).
We next transplanted CD45-SAP-treated young and aged mice with HSPCs derived from young mice (Fig. 1e). Donor-derived reconstitution was monitored in PB and by analyzing HSC chimerism in the BM at the experimental endpoint. While young recipients were effectively reconstituted, aged mice presented with only marginal donor-derived reconstitution in the PB (Fig. 1f). This was further reflected in the levels of donor-derived HSCs (Fig. 1g).
To enhance reconstitution levels in aged hosts, we explored additional conditions in which CD45-SAP was co-injected with other selective immunotoxins and antibodies. These included treating aged mice with CD45-SAP in combination with CD8-SAP, CD4-SAP, or a B cell-specific antibody cocktail (Extended Data Fig. 1d). However, neither of these treatments led to any evident enhancement in donor cell engraftment (Extended Data Fig. 1e).
We also expanded our investigation to include a combinatorial treatment with CD45-SAP and low-dose (200 cGy) TBI in young and aged mice (Extended Data Fig. 1f). This aimed to understand whether this synergistic approach could enhance HSC engraftment while minimizing the toxic effects of higher-dose TBI. While this effectively enhanced the reconstitution levels in young recipients to achieve near-complete donor-derived chimerism, this approach was much less effective for aged mice (Extended Data Fig. 1g).
In summary, these results demonstrate that advanced age significantly impairs effective HSC engraftment and transplantation success in C57BL/6 mice.
Ex vivo expanded HSCs effectively reconstitute multilineage hematopoiesis in young CD45-SAP-conditioned recipients
We and others have previously established the efficacy of a polyvinyl alcohol (PVA)-based culture system in promoting murine HSC expansion. Notably, expanded HSCs enable a degree of HSC-derived reconstitution even in completely unconditioned hosts11,12. Given this, we explored the impact of larger quantities of HSCs on in vivo reconstitution outcomes in both non-conditioned and alternatively conditioned hosts.
We expanded HSCs ex vivo for 21 days and transplanted equivalent fractions (EE) derived from either 100 or 500 HSCs into unconditioned young hosts (Fig. 2a). Existing literature suggests that the BM niches available for engraftment in unconditioned hosts are limited, yet they are continuously made accessible through a process of niche recycling9. With this concept in mind, we examined the reconstitution outcome when the EE500 was subdivided into five separate fractions. Each of these fractions was then transplanted at weekly intervals to evaluate the possible advantage of spreading the transplantation over time (Fig. 2a).
In agreement with previous work11, we observed that all non-conditioned recipients of ex vivo expanded HSCs demonstrated durable long-term multilineage engraftment, although the lymphoid chimerism, and in particular for the B cell lineage, was not on par with the myeloid reconstitution (Fig. 2b). EE500 resulted in higher engraftment compared to EE100, demonstrating a linear increase in myeloid lineage chimerism (19.0 ± 2.1 vs. 4.3 ± 2.0) (Fig. 2b). However, dividing the EE500 graft into five weekly doses did not yield better results than a single bolus injection (Fig. 2b).
Subsequently, we evaluated the performance of ex vivo expanded HSCs in young hosts conditioned with CD45-SAP, comparing their behavior with that of hosts subjected to lethal (950 cGy) TBI (Fig. 2c). As anticipated, lethal TBI led to near-complete multilineage donor-reconstitution (Fig. 2d). CD45-SAP conditioning also resulted in prominent multilineage reconstitution, albeit with a lesser contribution to lymphoid lineages (Fig. 2d). Crucially, an examination of HSC chimerism at the end of the experiment demonstrated reconstitution levels equivalent to those observed for myeloid lineages (Fig. 2e), reinforcing that myeloid reconstitution serves as a dependable measure of ongoing HSC activity24.
HSC transplantation into TBI-conditioned hosts detrimentally impacts their capacity to reconstitute secondary hosts25. To ascertain whether this also holds true for CD45-SAP-conditioned hosts, we conducted secondary transplantations of BM cells from the primary transplanted CD45-SAP-conditioned hosts. Two scenarios were considered: a) a non-competitive context where the transplanted cells in secondary hosts competed with the endogenous HSCs from the primary hosts, and b) a situation where the transplanted cells were mixed with an equal number of BM cells from young, untreated mice. BM cells harvested from primary TBI-treated recipients were included for comparison (Fig. 2f). This disclosed that high reconstitution levels observed in primary CD45-SAP-treated hosts were sustained in the non-competitive setting (Fig. 2g). Conversely, the reconstitution levels were markedly decreased upon competitive transplantation, mirroring the reduction in HSCs derived from primary TBI-conditioned hosts (Fig. 2h).
In summary, these results affirm prior studies, underscoring that while unconditioned wild-type (WT) hosts can attain long-term HSC-derived multilineage reconstitution, this necessitates significant quantities of HSCs10. Notably, pairing higher doses of HSCs with CD45-SAP conditioning dramatically enhanced the reconstitution outcomes. However, an intriguing parallel was noted with HSCs transplanted into TBI-conditioned hosts, where the process of transplantation itself appears to impair their potential for serial transplantation.
Engraftment efficiency and functionality of transplanted young HSCs are maintained in aged hosts
The interplay between HSCs and their environment encompasses intricate physiological interactions26. This complexity might be amplified considering the dynamics between transplanted young HSCs and aged host cells. Given our observed barrier in engrafting young HSCs in aged mice (Fig. 1), we entertained either reduced homing of young HSCs in aged recipients and/or an adverse environment in aged recipients for transplanted young HSCs.
To approach these questions experimentally, we harvested HSCs from young mice, expanded them ex vivo, and labeled the expanded grafts with Cell Trace Violet (CTV) dye. The CTV-labeled cells were then transplanted into both unconditioned and CD45-SAP-conditioned young and aged hosts. This allowed for the assessment of engraftment and proliferation of CTV-labeled young HSCs 2–4 weeks post-transplantation (Fig. 3a).
Analyses of unconditioned hosts revealed that young-derived HSCs could be recovered from both young and aged recipients, but with a tendency for lower efficiency in aged hosts (1.5-fold, Fig. 3b). More striking, but in agreement with the well-established expansion of HSCs associated with murine aging2, aged recipients exhibited a significantly increased number of host HSCs in the unconditioned setting (Fig. 3b).
When analyzing CD45-SAP-conditioned young and aged hosts, we observed similar amounts of recoverable young donor HSCs (Fig. 3b). This demonstrated that compromised homing/engraftment in aged mice was unlikely to explain the inefficient reconstitution from young HSCs (Fig. 1). While CD45-SAP conditioning reduced the numbers of host HSCs in both young and aged mice (5.6- and 8.6-fold, respectively), aged recipients still harbored a notably higher number of host HSCs than young recipients (Fig. 3b).
CTV dye dilution revealed that transplanted HSCs exhibited similar proliferation kinetics in both young and aged unconditioned recipients (Fig. 3c and 3d). Even after CD45-SAP treatment, which accelerated HSC proliferation, this similarity persisted across different age environments.
To further evaluate the functionality of young HSCs exposed to an aged environment, we isolated HSCs from young mice, expanded them ex vivo, and labeled them with CTV dye. We then transplanted these cells into unconditioned young and aged mice. Six weeks later, CTV-positive HSCs were extracted from the primary hosts and competitively transplanted into TBI-conditioned recipients (Fig. 3e). These experiments demonstrated efficient long-term multilineage reconstitution from isolated HSCs, irrespective of whether the cells were obtained from young or aged primary hosts (Fig. 3f).
Taken together, these results establish that young HSCs can successfully engraft in an aged environment, which does not inherently harm HSCs or hinder their ability for proliferation or long-term multilineage reconstitution.
Young HSCs support youthful hematopoietic characteristics upon transplantation into aged recipients
Although young HSCs successfully reconstituted aged recipients, the reconstitution was limited (Fig. 3). We hypothesized that residual endogenous HSCs might restrict this process (Fig. 3b). Therefore, more thorough elimination of the host’s aged HSCs could potentially enhance hematopoiesis from transplanted young HSCs.
Recent studies suggest that mobilizing endogenous HSCs could be a viable strategy to coax these cells out of their supportive niches within the BM, thereby creating vacant niches for transplanted HSCs8,27. Therefore, our subsequent experiments evaluated the reconstitution levels of EE100 young HSCs after CD45-SAP conditioning, either alone or in combination with a G-CSF/AMD3100-based mobilization protocol (Fig. 4a).
Examination of PB 18 weeks post-transplantation showed that aged mice undergoing the combined CD45-SAP/mobilization-based conditioning had over a two-fold increase in donor-derived multilineage reconstitution compared to those conditioned with CD45-SAP alone (Fig. 4b). BM HSC analysis revealed nearly four times more recoverable donor HSCs in mice receiving the combined treatment than in those with CD45-SAP conditioning only (Fig. 4c). Consistent with earlier data (Fig. 3b), CD45-SAP significantly reduced host HSC levels, with further reductions observed in mobilized recipients (Fig. 4d).
We next examined the characteristics of young HSC-derived hematopoiesis and its interplay with the host's aged-derived hematopoiesis. We employed multi-parameter flow cytometry to stage hematopoiesis in the BM and, given the notable impact of aging on lymphopoiesis14, conducted detailed examinations of B and T cell compartments in the spleen and thymus. Our analysis also included a small cohort of young recipients undergoing the same transplant procedure.
To explore the relationship between transplanted and host HSCs and their differentiated progeny, we compared the chimerism levels of HSC progeny to those of BM HSCs. This consistently demonstrated that chimerism in progeny from young HSCs was significantly higher than in host-derived cells in aged recipients, but less pronounced in young hosts (Fig. 4e).
For the lymphoid lineages, the contribution to the early stages of differentiation (MPP Ly and CLPs) in aged recipients was almost five times higher than that of the HSCs themselves (Fig. 4e). Further examination of the B cell lineage revealed a slightly lower chimerism at the later B cell progenitor stages, although differentiation into these stages was still considerable higher than that observed for the age-derived cells. By contrast, early B cell progenitors in young recipients were predominantly host-derived (Fig. 4e).
Examination of the thymus revealed higher chimerism across all evaluated T cell subsets compared to cells derived from the host. Importantly, this difference was significantly more pronounced in aged recipients relative to their younger counterparts (Fig. 4e). Additional assessment of splenic B and T cells showed that while the overall chimerism from the young donor was greater than that for BM HSCs, these levels were generally not as high as those observed in the primary hematopoietic organs associated with these lineages (Fig. 4e).
Aging has been reported to correlate with the accumulation of a specific B cell subset known as age-associated B cells (ABCs)28,29. Therefore, we examined the presence and origins of ABCs, in conjunction with traditional follicular and marginal zone (MZ) B cells analysis (Fig. 4f). Although a small proportion of ABCs originated from the young donor, the overwhelming majority of these cells were host-derived (Fig. 4f and 4g). Conversely, donor cells effectively generated follicular B cells, while their contribution to the MZ B cell compartment was similar to the host-derived cells (Fig. 4f and 4g).
Next, we examined the distribution of more mature T cell subsets within the young to aged chimeras. Aging associates with an increased frequency of both central and effector memory cells, alongside a corresponding decrease in naive T cells14. Our analysis revealed that naive CD4 and CD8 T cells from host/aged-derived cells accounted for only about 5% and 15%, respectively, in stark contrast to the roughly 40% chimerism in both subsets derived from young HSCs (Fig. 4h and 4i). This underscores the potential to significantly boost the production of naive CD4 and CD8 T cells in aged mice.
A recent study suggested that selectively depleting aged HSCs through antibody-mediated targeting could mitigate age-related lymphoid deficiencies and potentially enhance immune function in the elderly30. Using CD45-SAP for specific HSC depletion (Fig. 1), we compared mature B and T cell subsets in aged mice with those in unmanipulated aged controls (Extended Data Fig. 2a). We observed significant reductions in host memory CD4 and CD8 T cells, follicular B cells, and a decrease in ABCs post-treatment. Although there was no increase in naive T cell production, this treatment showed promise in boosting immature B cell generation from aged host cells (Extended Data Fig. 2b-d).
Overall, these findings indicate that depleting host HSCs and transplanting young donor cells not only endows their progeny with youthful hematopoietic traits in aged recipients but also that the CD45-SAP treatment contributes to reverting the composition of the aged host's adaptive immune cells to a more youthful-like state.
Molecular stability in early young donor-derived lymphoid progenitors exposed to aging bone marrow
Age-related decline in lymphopoiesis can be linked to reduced production of early lymphoid progenitors, including MPP Ly15,16. Because this subset was effectively regenerated from young donor cells in aged recipients (Fig. 4e), our subsequent analysis examined the molecular features of these cells. For this, we performed RNA-sequencing of donor- and host MPP Ly cells from young and aged recipients (Fig. 4e). Somewhat unexpectedly, principal component analysis failed to separate between donor and host MPP Ly across age groups (Extended Data Fig. 3a) and differential gene expression analysis revealed only 17 upregulated genes in donor MPP Ly from aged mice (Extended Data Fig. 3b). Similarly, analysis of host cells identified merely 16 upregulated genes upon aging (Extended Data Fig. 3c), without association to any MSigDB pathway (Extended Data Fig. 3d). This sharply contrasted with aged HSCs from an independent study31, which, using a similar analytical approach, displayed pronounced differences from their young counterparts, characterized by a distinct aging signature (Extended Data Fig. 3e-f)32.
Together, these results revealed no significant differences in the transcriptomic signatures of donor MPP Ly cells, even when exposed to an aging environment. This corroborates our functional data, which demonstrates that transplanting young HSCs effectively regenerates hematopoiesis with youthful characteristics in aged hosts (Fig. 4).
Non-invasive BM conditioning followed by transplantation mitigates disease progression in a mouse model of myelodysplastic syndrome
In our final experiments, we examined how CD45-SAP conditioning and HSC transplantation affect the development of age-associated hematological malignancies in the NUP98-HOXD13 (NHD13tg) transgenic mouse model, which predisposes to myelodysplastic syndrome (MDS) and acute leukemia33.
Our experimental layout entailed monitoring the disease evolution in NHD13tg mice for their entire lifespan (up to 24 months, n = 20). This group was juxtaposed against a cohort of NHD13tg mice that underwent CD45-SAP conditioning and transplantation with 107 WT BM cells when they were 2 months old (n = 9). We also incorporated a small group (n = 5) of WT littermate mice as a control (Fig. 5a).
Mice subjected to CD45-SAP conditioning and transplantation exhibited high-level donor multilineage reconstitution four months post-transplantation (Fig. 5b). None of the aged WT mice developed hematological malignancies during the 2-year observation period. In contrast, many untreated NHD13tg mice began showing signs of diverse hematological diseases, including both myelo- and lymphoproliferative disorders and acute myeloid and T-cell leukemia, after six months of age. Although not every case could be conclusively diagnosed, many of the unclassified conditions were associated with pronounced thymic hyperplasia.
Overall, the incidence of disease in transplanted mice was significantly lower compared to their non-transplanted counterparts. Among the NHD13tg mice, 75% (15 out of 20) developed hematological malignancies, compared to only 33% (3 out of 9) of those receiving WT cell transplants (Fig. 5c). Furthermore, while 25% (5 out of 20) of NHD13tg mice developed acute leukemia, none of the transplanted mice did (Fig. 5d).