CD74 is a regulator of hematopoietic stem cell maintenance

Hematopoietic stem and progenitor cells (HSPCs) are a small population of undifferentiated cells that have the capacity for self-renewal and differentiate into all blood cell lineages. These cells are the most useful cells for clinical transplantations and for regenerative medicine. So far, it has not been possible to expand adult hematopoietic stem cells (HSCs) without losing their self-renewal properties. CD74 is a cell surface receptor for the cytokine macrophage migration inhibitory factor (MIF), and its mRNA is known to be expressed in HSCs. Here, we demonstrate that mice lacking CD74 exhibit an accumulation of HSCs in the bone marrow (BM) due to their increased potential to repopulate and compete for BM niches. Our results suggest that CD74 regulates the maintenance of the HSCs and CD18 expression. Its absence leads to induced survival of these cells and accumulation of quiescent and proliferating cells. Furthermore, in in vitro experiments, blocking of CD74 elevated the numbers of HSPCs. Thus, we suggest that blocking CD74 could lead to improved clinical insight into BM transplant protocols, enabling improved engraftment.


Introduction
Host immunity requires a constant renewal of red blood cells and leukocytes throughout life, as these cells have a restricted life span. Hematopoietic cell turnover is enhanced following acute stress situations, such as infections or irradiation, by the proliferation of hematopoietic stem cells (HSCs) and progenitor cells (HPCs), which respond to these conditions. The hematopoietic stem and progenitor cells (HSPCs) are a small population of undifferentiated cells that reside in the bone marrow (BM). HSCs are defined by their capacity for self-renewal and ability to differentiate into all blood cell lineages. Another distinct feature of these cells is their ability to migrate out of the BM to the peripheral blood. This process is enhanced under stress as a part of the host mechanisms of defense and repair. In addition, HSCs injected to the numbers of BM cells were detected in WT and CD74 −/− mice (Fig 1B), a significant increase in the Lineage marker-negative (CD11b-,Gr-1-,CD3-,B220-and Ter117-; Fig 1C) and LSK ( Fig  1D and 1E) populations was observed in the CD74 −/− animals. Moreover, an elevation was also detected in the early lymphoid-committed precursors (cKit-Sca1+) and common myeloid progenitor populations (cKit+Sca1-) (S1F Fig). Next, the percent of HSCs and HPCs was compared. As shown in Fig 1F and 1G, a significant increase in both the CD34 + and CD34 − populations was detected in mice lacking CD74, with a more significant elevation of the HSC CD34 − population. This accumulation was detected as well in CD74 −/− progeny of newly crossed WT and CD74 −/− mice, relative to CD74 expressing littermates (S1G Fig).
These progenitor populations are also characterized by expression of the signaling lymphocytic activation molecule (SLAM) family members, CD150 and CD48 [10,23,24]. Similarly, we observed an increase in the percent of HSCs and progenitors in CD74 −/− mice when identified by FACS analysis for CD150, CD48, Lin − , ckit + and Sca-1 + (Fig 1H-1J). These results suggest elevated stem and progenitor cell populations in mice lacking CD74. Since CD74 regulates the expression of SLAM receptors [25], we decided to focus on the CD34 marker for stem cell analysis in this study.
To examine the in vitro repopulation potential of CD74-deficient stem cells, the ability of HSPCs to proliferate and differentiate into colonies in vitro was analyzed by a colony-forming unit cell assay (CFU-C assay). As seen in Fig 1K, and as previously demonstrated [26], higher numbers of colonies were generated from CD74 −/− BM when compared to WT cells.
Next, the expression of the CD74 ligand, MIF, and its coreceptor, CD44, were analyzed. Similar intracellular MIF (S1H and S1I Fig) and cell surface CD44 (S1J and S1K Fig) expression levels were detected in CD74-deficient HSPCs compared to WT. To directly determine the role of MIF in HSPC accumulation, total BM cells were extracted from WT and MIF −/− mice, and progenitor populations were analyzed. Higher numbers of Lin neg and HSC populations were observed in the MIF −/− mice compared to WT animals (S2A- S2D Fig). However, the differences were not as pronounced as in the CD74-deficient mice. This could be explained by partial compensation by the MIF homologue, MIF2 [27]. These results suggest that MIF and its receptor, CD74, limit HSPC number.

CD74 −/− HSPCs demonstrate enhanced long-term self-renewal capacity
To determine whether the expansion of HSPCs in CD74 −/− mice results from an effect intrinsic to the cells themselves, or whether the differences are due to extrinsic environmental factors, chimeric mice were generated. Total BM cells from WT or CD74 −/− mice were transplanted into lethally irradiated WT or CD74 −/− recipients. The animals were killed after 16 weeks, and their HSPCs were analyzed. As seen in Fig 2A-2C, elevation in Lin, LSK, and CD34 − populations was detected in mice transplanted with CD74 −/− BM compared to WT donors. Thus, the lack of CD74 in the donor cells rather than in the microenvironment contributed to HSPC accumulation. Taken together, these results indicate that the lack of MIF/ CD74 signaling results in an intrinsic increase in the HSPC population in the BM.
To further evaluate the effect of the stroma and cytokines in the cellular microenvironment on CD74 −/− HSPC numbers, BM cells were cultured in vitro in the presence of the stroma M210B4 cell line or various cytokines, and the proportion of the HSPCs out of live cells was compared. These conditions had only a minor or no effect on the fold of increase in WT and CD74 −/− stem cell numbers (Fig 2D-2F), further demonstrating that the accumulation potential of HSPCs in the absence of CD74 is intrinsic to the stem cells.
Next, we followed the in vivo potential of CD74-negative HSPCs to repopulate and compete with WT cells. To this end, WT (CD45.1) BM cells were transplanted at a 1:1 ratio with either CD74 −/− (CD45.2) or WT (CD45.2) cells into lethally irradiated recipient mice (CD45.1). BM and peripheral blood (PB) populations of the mixed chimeras were analyzed at 6, 16, and 24 weeks after transplantation. While WT CD45.1 and CD45.2 chimera maintained a 1:1 ratio (S3A-S3H Fig), CD74 −/− -derived BM cells exhibited a growth advantage over the WT populations (Fig 3 and S3I-S3L Fig). The dramatic overgrowth of CD74 −/− cells was observed as early as 6 weeks posttransplant and was maintained throughout the experiment. A significant advantage of CD74 −/− total BM cells (Fig 3A and 3B), myeloid (Fig 3C), total B cells (Fig 3D), and HSPC populations (Fig 3E and 3F) was observed at various time points tested. In the B cell lineage, the growth advantage of the CD74 −/− population was detected from early stages of B cell differentiation through the formation of immature B cells in the BM (Fig 3G). This advantage disappeared at the mature stage ( Fig 3H) due to the separate role of CD74 as a survival receptor on these cells [28]. Analysis of PB populations revealed an advantage similar to that observed in the BM. The myeloid (S3I Fig)  To determine whether this accumulation results from differential homing of the cells or their improved quality, the homing of HSPCs to the BM was followed up to 1 week after BM transplantation. WT and CD74 −/− HSPCs showed similar homing potential to the BM, and no differences in the number of WT versus CD74 −/− HSPCs in the BM were detected during the first week after transplantation (S4A- S4E Fig). This suggests that the accumulation of CD74-deficient HSPCs does not result from their induced homing, but rather from their enhanced potential to repopulate the BM compartment. Thus, the CD74-deficient BM cells have an intrinsic advantage in repopulation and accumulation in the BM niche.

PLOS BIOLOGY
Blocking CD74 can lead to enhanced engraftment. a 1:1 ratio to irradiated recipient mice (CD45.1) to generate competitive BM chimera. BM populations of the mixed chimeras were analyzed 6 and 18 weeks following the transplantation. As shown in Fig 4A-4E, a significant advantage of CD74 −/− BM cells was observed at both 6 and 18 weeks postengraftment. These results indicated that LSK lacking CD74 are more efficient in repopulating the host environment, as seen by the significantly higher levels of those cells when compared to the WT (CD45.1).

PLOS BIOLOGY
Blocking CD74 can lead to enhanced engraftment.
CD74 −/− -derived BM cells show an advantage over the WT populations in BM transplantation. The dramatic overgrowth of CD74 −/− cells might result from their higher numbers of progenitor cells, from their higher potential to repopulate their niche, or both. To determine the basis for the enhanced ability of CD74 −/− HSCs to repopulate the BM, we wished to follow the competition under conditions in which the HSC WT and CD74 −/− numbers are similar. We therefore analyzed BM repopulation after serial dilutions (3:1, 7:1, and 9:1). Chimeric mice were generated at a ratio of approximately 3:1 in favor of the WT, which represents the injection of similar numbers of HSCs. As shown in Fig 5A-5F, while the WT (CD45.1):WT (CD45.2) chimeric mice preserved a ratio of 3:1, CD74 −/− -derived BM cells retained an advantage over the WT populations and resulted in a dramatic takeover of CD74 −/− cells as early as 6 weeks posttransplant, which was maintained throughout the experiment (16 weeks). A significant numerical advantage of CD74 −/− total BM (Fig 5A and S4F Fig) myeloid cells (Fig 5B), immature B cells (Fig 5C), LSK (Fig 5D), and CD34-/LSK (Fig 5E) was observed at this time point. The only population that could not compete was the mature B population (Fig 5F), whose survival is dependent on CD74 expression [29]. To gain further insight into the enhanced potential of CD74 −/− HSPCs to repopulate the BM, we followed the population at  (Fig 5G-5L), although even at this dilution, the ratio of CD74 −/− cells (CD45.2) to WT (CD45.1) was higher compared to the ratio of WT CD45.2 to WT CD45.1. Thus, CD74-deficient HSCs have a stronger potential to repopulate the BM and compete for the niches relative to WT HSPCs.
To evaluate the long-term self-renewal and functional properties of CD74 −/− HSCs, a serial BM transplantation assay was performed. BM cells from 6 WT and 6 CD74 −/− mice were isolated, and the cells were serially transplanted into lethally irradiated WT mice. Each host was transplanted with 2 × 10 6 BM cells. During the first 3 first cycles, no significant differences were observed between the WT and CD74 −/− groups, with high survival rates of all mice ( Fig  5M-5O). However, by the fourth transplantation cycle, CD74 −/− -transplanted mice showed a better survival rate compared to the WT mice (57% compared to 33%) ( Fig 5P). Thus, the absence of CD74 in HSCs results in accumulation of cells with a higher potential to repopulate the BM.

CD74 regulates stem cell survival
Next, we wished to identify the molecular mechanism regulating the HSPC accumulation in CD74-deficient mice. The accumulation of stem cells might result from their elevated retention in the BM niche or up-regulation of their proliferation or survival. Since CXCR4 plays a major role in retention of HSCs and HSPCs [8,30,31], we next wished to follow the role of CD74 in CXCR4 expression and function in HSPCs. We previously showed that following activation of CD74 expressed on CLL cells, CD74-ICD binds the chromatin of the CXCR4 promoter ( [32]; S5A Fig). Expression of cell surface CXCR4 was therefore analyzed on HSPCs derived from WT and CD74-deficient mice. As shown in Fig 6A and 6B, a reduction in the expression of CXCR4 on the cell surface was observed on CD74 −/− cells. This result shows that CD74 regulates CXCR4 expression. To determine whether the reduced levels of CD74 affect HSPC retention, the total counts of HSPCs in WT and CD74 −/− PB were compared. As shown in Fig 6C-6E, a dramatic elevation in the number of HSPCs in the circulation of CD74-deficient mice was detected. However, treatment of the mice with AMD3100 induced mobilization of both WT and CD74 −/− HSPCs to the same degree (6-to 7-fold; Fig 6F). All studies performed with CXCR4-deficient HSPCs demonstrated a requirement of this receptor for efficient engraftment. Nevertheless, transient inhibition of CXCR4 had no adverse effects on the engraftment capacity of the HSPCs [8]. CD74 deficiency partially reduces CXCR4 expression; however, our results suggest a negligible role for CXCR4 in the induced mobilization of CD74-deficient cells.
It was recently shown that blocking CXCR4 induces an increase in the cycling activity of the HSPCs [8]. We therefore analyzed the role of CD74 in HSPC proliferation. To determine whether CD74 controls cell proliferation and the cell cycle, Ki67 levels were followed. Higher numbers of both quiescent (Ki67) and cycling (Ki67 + ) stem cells (CD34-LSK) ( Fig 6G) and progenitors (CD34+LSK) (Fig 6H) were detected in CD74 −/− mice. However, the ratio between Ki67 − and Ki67 + populations did not change ( Fig 6I). To further follow cell proliferation in mice lacking CD74, a 5-bromodeoxyuridine (BrdU)-labeling experiment was performed. Mice were fed 0.8 mg/ml BrdU in their drinking water for 3 days, and BrdU incorporation was followed. As shown in Fig 6J and 6K, although higher numbers of both quiescent and cycling cells were detected in CD74 −/− mice, no significant change in the ratio of these populations was observed. These results suggest that although the HSC and HSPC compartments are larger in mice lacking CD74, there is no overproliferation of any specific population, and the proportion of proliferating cells is similar in the WT and CD74 −/− mice.
Electron transfer along the mitochondrial respiration chain induces the formation of reactive oxygen species (ROS) [33]. Emerging evidence shows that oxidative stress and, in particular, ROS content, influences stem cell migration, development, and self-renewal, as well as their progression through the cell cycle [1]. To determine whether oxidative phosphorylation is elevated in the absence of CD74 in stem cells, ROS levels were compared in HSPCs of CD74 −/− and WT cells. As can be seen in Fig 7A, higher numbers of ROS high cells were detected in the CD74 −/− HSPCs compared to the WT. However, no difference was observed in the ratio between ROS high and ROS low expressing cells (Fig 7B), suggesting that level of ROS expression in each subpopulation remains constant. To determine whether excess ROS contributes to the expansion of CD74 −/− stem cells, ROS levels were reduced using the antioxidant, N-acetyl-L-cysteine (NAC) [34]. As shown in Fig 7C and 7D, treatment with NAC for 6 days

PLOS BIOLOGY
partially reduced the levels of HSPCs in CD74 −/− mice. This suggests that ROS levels play a limited role in HSPC accumulation in mice lacking CD74.
To further probe the mechanism of action of CD74 in HSPCs, we wished to determine whether the higher number of CD74 −/− HSPCs results from enhanced cell survival. Therefore, HSPCs cells were analyzed for cell survival using an Annexin V staining assay. As shown in Fig  7E, reduced apoptosis was observed in the CD74 −/− CD34 − LSK cells compared to the WT population. Thus, the higher number of CD74 −/− stem cells might result from an increase in their survival.

PLOS BIOLOGY
Blocking CD74 can lead to enhanced engraftment.
Since CD74 regulates mature B cell survival, and its absence leads to cell death [35,36], we wished to understand the different, and possibly contradictory, roles of CD74 in stem cells. We suggested that hypoxic conditions that exist in the BM environment and especially the perivascular niches where the nondividing HSCs reside control the enhanced survival of CD74 −/− cells [37]. WT and CD74 −/− BM cells were incubated under hypoxia or with normal oxygen levels for 24 h, and cell survival was analyzed by Annexin V staining. As can be seen in Fig 7F, CD74-deficient stem cells survived better under hypoxia. Transfer of the CD74 −/− cells to normoxic conditions reduced the advantage of these cells compared to WT stem cells ( Fig  7G). These results support our suggestion that the BM hypoxic conditions play a role in the control of CD74 function.
A key regulator of adaptation processes induced by hypoxia is the transcription factor, hypoxia-induced factor (HIF), which is active only after heterodimerization of its 2 subunits α and β. Although HIF-1β is constitutively expressed, HIF-1α is induced by hypoxia. It was previously shown that MIF regulates HIF-1α expression in a CD74-dependent manner [38]. We therefore tested HIF-1α protein levels in WT and CD74 −/− CD34-population. As shown in Fig  7H and 7I, lower levels of HIF-1α protein were detected in the CD34-populations, suggesting a role for CD74 in the control of this transcription factor expression.
To further follow CD74 regulated cascades, RNA sequencing (RNAseq) analysis of gene expression in WT and CD74 −/− CD34-LSK cells was performed. RNAseq revealed that CD74 regulated the expression of transcription factors including KLF4, KLF2, and E2F2, and altered the expression of pathways induced by transcription factors such as IRF8, MDB2, and CEBPA, which are known to regulate HSCP maintenance (S5B Fig) [39 -44]. The RNAseq also revealed a regulation of the ITGB2 and ITGB3 pathways. We further studied the ITGB2 gene, which encodes the β2 common leukocyte subunit, CD18 (Fig 7J). Upon binding with an alpha chain, CD18 is capable of forming a complex which plays a significant role in cellular adhesion and immune response [45]. Furthermore, deficiency of this molecule results in a phenotype similar to that described here, in vivo accumulation of HSCs in mouse BM [46]; in humans, downregulation of CD18 expression was associated with a primitive HSC phenotype, and enhanced long-term repopulation in NSG mice [47]. To determine whether CD74 directly regulates CD18 expression, CD74 cytosolic domain binding to regulatory elements on CD18 chromatin was analyzed. As show in Fig 7K, the cytosolic domain of CD74 bound to the CD18 promoter area and to elements in the intron areas. Next, CD18 protein levels were determined in WT and CD74-deficient mice. A significant down-regulation in CD18 cell surface levels was observed on HSPCs and HSCs derived from mice lacking CD74 (Fig 7L and S6A Fig). Furthermore, reduced expression levels of CD18 were detected in the MIF −/− HSCs and HSPCs ( Fig  7M and S6B Fig), further demonstrating that CD74 and its ligand, MIF, regulate CD18 expression.
Next, the effect of MIF inhibitor (ISO-1) on CD18 cell surface levels in HSPCs was analyzed. As shown in Fig 7N and S6C Fig, blocking MIF reduced CD18 cell surface expression on LSK-and CD34-cells. To further follow the role of extracellular MIF in the maintenance of HSPCS and CD18 regulation, WT HSPCs were cocultured with WT or MIF −/− BM cells. As shown in Fig 7O and S6D Fig, MIF secreted from the BM cells was essential for CD18 cell surface expression. In its absence, CD18 expression was reduced. Therefore, we suggest that MIF secreted from HSPCs or the BM environment binds to CD74 and induces CD18 expression; in the absence of MIF or CD74, its expression is down-regulated. To identify the β2 integrin complex regulated by CD74, we analyzed the α subunits that are affected by CD74 deficiency. As shown in the S6E Fig, CD74  Finally, we analyzed the therapeutic potential of CD74 inhibition. First, we analyzed the effect of CD74 blockade on the ability of HSPCs to accumulate in vitro. Blocking CD74 using a specific antibody significantly elevated the percent and counts of LSK cells in WT mice, while

PLOS BIOLOGY
this antibody had no effect on CD74-deficient BM cells. Thus, CD74 has a direct role in the repopulation of the stem cell populations (Fig 8A and 8B, S7B Fig). The role of CD74 in hematological recovery from chemotherapy was followed. WT and CD74 −/− mice were injected weekly with the cell cycle-dependent myelotoxic agent, 5-fluorouracil (5-FU), which kills proliferating cells and thereby stimulates HSC proliferation and the subsequent replenishment of the hematopoietic system [48]. Mice were injected with 150 mg/ml (Fig 8C) or 125 mg/ml ( Fig  8D) 5-FU. After injection of the higher dose, at day 14, about 30% of the WT mice and about 80% of the CD74 −/− mice survived. Injection of 125 mg/ml of the drug resulted in survival of about 10% WT and 60% CD74 −/− mice at day 20. These results further suggest that elevated survival of CD74 −/− HSCs allows better reconstitution of the immune system and increased survival of these mice under hemodepleting conditions. To verify the general nature of the phenomenon, CD74 expression was analyzed in human HSPCs (Gating strategy appears in S7A Fig). CD74 was detected on human CD34+ cells (Fig 8E), and its blocking resulted in down-regulation of CD18 expression (Fig 8F). These results suggest that CD74 regulates a similar pathway in human HSPCs and may provide a useful target for enhancing human stem cell transplantation.

Discussion
Despite the enormous experience in the manipulation and therapeutic use of HSCs, the biology of these cells is still not fully understood. HSCs are useful cells for transplantation and for regenerative medicine. However, it has been not possible to date to expand adult HSCs without losing their self-renewal properties. We suggest that CD74 regulates HSPC maintenance, and its absence or inhibition can result in expansion of this highly potent population.
CD74 is a membrane-associated protein that serves as a cell surface receptor for the cytokine MIF [13]. Here we show that potent stem and progenitor cells are accumulated in the BM of CD74 −/− mice, as well as in chimeric mice comprised of immune cells lacking CD74 and WT nonhematopoietic cells, suggesting that the phenotype of the donor cells and not the microenvironment is responsible for the enhanced CD74 −/− HSC expansion. However, we cannot exclude the possibility that other immune cells contribute to the expansion of CD74 −/− HSPCs in the chimeric mice. The increased quiescent stem cell population in mice lacking CD74 allows enhanced replenishment of the immune system and increased survival following chemotherapy. In the absence of CD74, there is an accumulation of higher numbers of HSPCs, which show a stronger potential to repopulate and compete for the BM niches compared to WT HSPCs. The accumulation does not result from induced stem cell homing or retention. Rather, our results show that the absence of CD74 leads to elevated survival of the HSCs. This results in elevated numbers of HSCs that express similar levels of ROS as wt HSPCs and proliferate at a similar rate. However, due to the elevated numbers of HSCs, the total production of ROS and the number of proliferating cells is increased. This allows the production of elevated numbers of HSPCs, resulting in generation of hematopoietic cells of the different lineages.
CD74 regulates CXCR4 expression [32]. It was recently shown that inhibition of the CXCR4/CXCL12 axis in the HSPC compartment results in an increase in mobilization efficiency. In addition, a concurrent increase in the cycling activity, self-renewing proliferation of the HSPC pool, and expansion of this population in the BM was observed [8]. Our results show that CD74 deficiency down-regulates CXCR4 cell surface expression. However, this regulation has a negligible role in the induced mobilization of CD74-deficient cells and does not affect cell proliferation, but rather supports survival of the cells. Nevertheless, it is possible that this lack of overall change might result from 2 antagonistic trends. CD74 functions as a survival receptor in mature B cells [35,36], while results presented in this study show that the lack of CD74 leads to accumulation of functional HSPCs in the BM, due to their enhanced survival. The cell type-specific regulation of survival by CD74 might result from its association with different cell surface receptors resulting in induction of cell type-specific signals. Alternatively, it is possible that the environment of these 2 cell types controls the different outcomes. While the blood and spleen environment is normoxic, the BM environment is hypoxic. Niches of various stem cell types, including hematopoietic cells, are microenvironments with low oxygen tension, ranging from 1% to 8% O 2 [37]. The hypoxic environment plays a critical role in the regulation of stem cell self-renewal and differentiation [37]. Our results show that the hypoxic microenvironment plays a role in the CD74-induced outcome, providing an advantage for the CD74-deficient HSPCs. Moreover, CD74-deficient CD34-cells express lower levels of HIF-1α protein, suggesting a role for CD74 in the control of expression of this transcription factor, which is a key regulator in the adaptation to hypoxic conditions. Down-regulation of HIF-1α expression was previously shown to inhibit precursor cell differentiation [49,50]. We therefore suggest that the lower levels of HIF-1α might enable the accumulation of more primitive cells.
We suggest that, under normal conditions, CD74 fine-tunes the maintenance of the BM stem cells, inducing a cascade that leads to cell death. It was previously shown that mesenchymal stem cells produce MIF to delay hypoxia-induced senescence [51]. In addition, MIF protects against hypoxia/serum deprivation-induced apoptosis of mesenchymal stem cells by interacting with CD74 to stimulate c-Met, leading to downstream PI3K/Akt-FOXO3a signaling and decreased oxidative stress [52]. Thus, CD74 induces a cell type-specific cascade that leads to different cell death outcomes.
Our results show that CD18 is a target gene of CD74. We demonstrate that CD74-ICD binds to the promoter area and elements in the intron areas of CD18, inducing its transcription and expression. CD18 deficiency affects the expansion of HSCs in the murine BM [46]. In humans, down-regulation of CD18 expression is associated with a primitive HSC phenotype and enhanced long-term repopulating capacity in NSG mice [47]. Thus, CD74 and CD18 deficiency results in a similar HSPCs phenotype. This accumulation does not result from enhanced homing of cells to the BM, but, rather, is caused by a progressive cell-autonomous expansion of the cells. Thus, we can suggest that reduced expression levels of CD74 and its ligand MIF down-regulates CD18 expression, resulting in enhanced stem cell survival.
Several novel HSC transplantation protocols are based on the administration of high numbers of donor cells. The need for large numbers of engraftable cells becomes particularly challenging in the case of cord blood (CB) transplantation and adult HSC gene therapy protocols, because of suboptimal HSC doses available for infusion, and impaired engraftment of the transplanted cells [53]. Furthermore, in patients without a suitable donor, the HLA barrier can be overcome by transplantation of megadoses of highly purified mismatched CD34+ stem cells [54]. Methods to improve maintenance and expansion of HSPCs resulting in increased numbers of CB stem cells may shorten time to engraftment and improve survival in adult recipients. CD74 deficiency results in higher numbers of efficient stem cells, better reconstitution of the immune system, and increased survival of these mice under hemodepleting conditions.
We therefore suggest that in vitro blocking of CD74 expressed on HSPCs or its ligand MIF may lead to expansion of the stem cells and will improve transplantation protocols.
NAC was administered by IP injection (50 mg kg −1 ; Sigma, Germany) for 6 consecutive days before the experiments. 5-FU was administered by IP injection (150 mg kg −1 or 125 mg kg −1 ; ABIC, Teva Group, the Netherlands) once a week. AMD3100 was administered by IP injection (20 mg kg

Cells
Mice. BM cells were obtained by flushing long bones with PBS, and peripheral blood was collected from the eye or heart using heparinized syringes.
Human samples. Human BM cells derived from healthy patients were provided by the hematology institute of the Kaplan Medical Center in compliance with the review board of the hospital. Consent was informed and the samples were obtained by written consent. The Weizmann institute review board approved the study: IRB number 1339-1.

Library construction and sequencing
HSCs (CD34-/LSK) (1 � 10 3 cells) were sorted from WT and CD74 −/− mice (BD FACS Aria III). Sequencing libraries were prepared using the SMART-Seq v4 Ultra Low Input RNA kit (Clonetech, United States of America). Sequencing libraries were constructed containing barcodes. Single-end reads were sequenced on one lane of an Illumina HiSeq2500 machine.
Sequence data analysis: Poly-A/T stretches and Illumina adapters were trimmed from the reads using cutadapt; resulting reads shorter than 30 bp were discarded. Reads were mapped to the M. musculus (GRCm38) reference genome using STAR [56], supplied with gene annotations downloaded from Ensembl release 92 (with EndToEnd option). Expression levels for each gene were quantified using htseq-count [57]. Differentially expressed genes were identified using DESeq2 (version 1.10.1) [58] with the betaPrior, cooksCutoff, and independentFiltering parameters set to False. Raw P values were adjusted for multiple testing using the procedure of Benjamini and Hochberg. The RNA-Seq data discussed in this publication have been deposited in NCBI's Gene Expression Omnibus [59] and are accessible through GEO Series accession number GSE163661 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc= GSE163661).

Colony-forming assay (CFU-C)
BM mononuclear (BM-MNC) cells were isolated by Ficoll separation and were seeded (15 × 10 3 cells/ml) in CFU-C semisolid medium supplemented with EPO, IL-3, GM-CSF, and SCF, as described [60]. CFU-C were scored 7 days after plating, and results are presented as CFU-C per number of seeded cells.

Serial transplantation
For serial transplantation assay, 2 × 10 6 BM cells were obtained from 6 WT donors and 6 CD74 −/− donor mice and transplanted to lethally irradiated WT CD45.1 animals. Each donor population was transplanted to 4 to 5 recipient mice. At 10 to 12 weeks posttransplantation, 1 mouse from each donor served as a donor for the subsequent transplant.

Cell cycle analysis
To analyze quiescent cells, total BM cells were stained for the designated markers (lineage/Sca-1/c-Kit/CD34), fixed and permeabilized using BD Cytofix/Cytoperm Plus kit (BD Bioscience, USA), and stained with the Ki-67 antibody (cat: 556026; BD Pharmingen, USA). To measure proliferation, C57BL/6 and CD74 −/− mice were fed with 0.8 mg/ml BrdU in their drinking water for 3 days. BrdU incorporation was followed in LSK cells from BM using the BrdU flow kit (BD Pharmingen).

ROS analysis
Cellular ROS levels were analyzed by incubation of BM cells with 2 μM hydroethidine (Molecular Probes, USA) for 10 min at 37˚C. Cells were then washed with PBS and stained for lineage/Sca-1/c-Kit/CD34 markers and analyzed by FACS.

Apoptosis analysis
Total BM cells were stained with the appropriate antibodies, using the Annexin V binding buffer (1:10 in ddW; cat: 556454, BD Pharmingen, USA) and mixed with Annexin V (FITC Annexin V, cat: 556419; BD Pharmingen, USA). All samples were incubated for 15 min at room temperature. For measuring Annexin levels under hypoxia, the BM cells were incubated in a hypoxia chamber at 1% O 2 for 24 h before staining.
Human-BM cells from patients (10 6 /in 200 μl) were treated for 48 h with isotype control or with anti-human CD74 (LN2; Biolegend) at 3 μg/ml. Lethally irradiated WT (CD45.1) mice were transplanted with BM derived from WT (CD45.2) mice at a 1:1 ratio. Percent of each population in the BM and PB was analyzed after 6, 16, and 24 weeks. (A) Total BM cells; Data A in S11 Data (B) BM myeloid cells; Data B in S11 Data (C) immature BM B cells; Data C in S11 Data (D) mature BM B cells; Data D in S11 Data (E) BM LSK; Data E in S11 Data (F) PB myeloid cells; Data F in S11 Data (G) PB mature B cells; Data G in S11 Data (H) PB LSK, and 4 h after injection of AMD3100; Data H in S11 Data (I-L) Lethally irradiated WT (CD45.1) mice were transplanted with BM derived from WT (CD45.1) and CD74 −/− (CD45.2) mice at a 1:1 ratio. Percent of donor-derived cells was analyzed in the PB after 6, 16, and 24 weeks in (I) myeloid cells; Data I in S11 Data (J) immature BM B cells; Data J in S11 Data (K) PB LSK and 4 h after injection of AMD3100; Data K in S11 Data (L) PB mature B cells; Data L in S11 Data, n = 6-27. Bars show SEM. Unpaired two-tailed t test � p < 0.05; �� p < 0.01; ��� p < 0.001; ���� p < 0.0001. BM, bone marrow; PB, peripheral blood; WT, wild-type. 2) at a 7:1 ratio, or BM derived from WT (CD45.1) and CD74 −/− (CD45.2) mice at a 7:1 ratio. Mice were analyzed 16 weeks after transplantation. Percent of each population in the BM was analyzed 13 weeks; n = 9-13. (G) Total BM cells; Data F in S12 Data (H) myeloid cells; Data G in S12 Data (I) immature BM B cells; Data H in S12 Data (J) LSK, Data I in S12 Data (K) CD34-/LSK, Data J in S12 Data (L) mature BM B cells, Data K in S12 Data. Bars show SEM.