FLT3-Ligand Treatment of Humanized Mice Results in the Generation of Large Numbers of CD141⁺ and CD1c⁺ Dendritic Cells In Vivo

Dendritic cells (DC) play a pivotal role in the induction and regulation of adaptive immune responses. The DC network is comprised of multiple subsets with specialized functions. In mice, these are divided into plasmacytoid DC (pDC) and myeloid or conventional DC (cDC). The cDC can be further subdivided into lymphoid resident or migratory DC based on their tissue location. DC can also be derived from monocytes under inflammatory conditions (termed inflammatory or monocyte-derived DC [MoDC]). In the mouse, the lymphoid-resident CD8⁺ cDC and migratory CD103⁺ cDC subsets are crucial for the induction of CTL responses against cancers, viruses, and other pathogenic infections (1). These DC are therefore attractive candidates to target for the development of CTL-mediated vaccines, a strategy that has already demonstrated efficacy in mouse models (2). In humans, the DC subset defined by expression of the marker CD141 (blood DC Ag [BDCA] 3) has been identified in both lymphoid and nonlymphoid tissues as the equivalent of mouse lymphoid- resident CD8⁺ cDC and migratory CD103⁺ cDC (3–7). The features that define these DC subsets in both species are expression of C-type lectin-like receptor 9A (CLEC9A), chemokine receptor XCR1, nectin-like protein 2 (NECL2/CADM1), and TLR3 (3–13). They are major producers of IFN-λ (14) following activation with the TLR3 ligand polyinosinic-polycytidylic acid (poly I:C) and are specialized in their capacity to cross-present exogenous Ag for recognition by CTLs.

Despite the similarities, whether CD141⁺ DC are the most effective subset to target for CTL-mediated vaccines in humans is currently under debate. There is clear evidence that other human DC subtypes, including CD1c⁺ DC, MoDC, and pDC can cross-present some forms of Ag and activate CTL under certain conditions (15–19). In contrast to their mouse counterparts, human CD141⁺ DC lack expression of TLR4 and TLR9 (6, 19). These interspecies differences necessitate the preclinical evaluation of both the Ag and adjuvant components of vaccines on human DC subsets. This is currently limited by the availability of human CD141⁺ DC and appropriate culture models to study them.

CD141⁺ DC are found in low frequency, constituting 0.03% of mononuclear cells in peripheral blood. The isolation of CD141⁺ DC from human leukapheresis products is currently the most practical method of obtaining CD141⁺ DC, but only ∼3 × 10⁵ cells are obtained following expensive enrichment procedures (20). Human DC subsets, including CD141⁺ DC, develop from CD34⁺ progenitors following in vitro culture with a mixture of cytokines and growth factors, but this strategy is also currently limited by low cell yields and purity (7).

Mice reconstituted with human hematopoietic stem cells (humanized mice) are a practical model to study aspects of the human immune system and bridge the translational gap (21). Development of human myeloid cells has been described in immunodeficient mice following injection of human CD34⁺ progenitors from cord blood, adult G-CSF–mobilized blood or fetal liver (21). However, human myelopoiesis is not efficient in these models due to a lack of species cross-reactivity of nonhematopoietic cell-derived growth factors including CSF-1, GM-CSF, IL-3, and erythropoietin. This has resulted in poor differentiation and/or function of cells of the monocyte/macrophage lineage unless supported by the addition of human cytokines (22–27). Human myeloid DC (defined as human CD45⁺, lineage⁻, HLA-DR⁺, CD11c⁺) can develop in humanized mice (28–31), but whether fully functional specialized myeloid DC subsets develop has not been addressed. Although DC phenotypically resembling human blood CD141⁺ DC by expression of CD141, CLEC9A, NECL2, and TLR3 were found in spleens of humanized mice, their low frequency has limited functional characterization (7).

FLT3-ligand (FLT3-L) is a hematopoetic cytokine that induces differentiation and survival of DC. FLT3-L administration into humans increases CD1c⁺ DC, pDC, and, importantly, CD141⁺ DC in the circulation (10, 32, 33). Murine FLT3-L preferentially expands the mouse CD8⁺ DC subset (34), but whether human FLT3-L biases expansion of the human CD141⁺ DC equivalent is not known. Moreover, it is not known whether FLT3-L treatment may preferentially increase human DC subset expansion in humanized mice. In this study, we demonstrate that FLT3-L induces large numbers of fully functional human CD141⁺, CD1c⁺ DC, and, to a lesser extent, pDC in humanized mice.

Cord blood was obtained from the Queensland Cord Blood Bank following informed consent in accordance with approval from the Mater Adult Hospital Human Ethics Committee. CD34⁺ progenitors were isolated using a CD34⁺ isolation kit (Miltenyi Biotec). Mice were housed and treated in accordance with approval by the University of Queensland Animal Ethics Committee. Humanized (hu)NOD/SCID mice were generated as previously described (30) with minor modifications. Five- to 6-wk-old female NOD/SCID mice were sublethally irradiated with 325 cGy and then injected with CD122 Ab (BioXCell) to deplete NK cells. Mice were transplanted i.v. with 2 × 10⁵ CD34⁺ cells 48 h later. Engraftment was confirmed after 4 wk by the detection of huCD45⁺ cells in blood (day 0 of FLT3-L treatment). Engrafted mice were then injected s.c. with 50 μg recombinant huFLT3-L (Amgen) on days 1 and 4, and DC were harvested on days 8 to 9.

Single-cell suspensions of crushed bones, spleen, lung and liver were stained with the following mAbs: LIVE/DEAD Fixable Aqua Dead Cell Stain Kit (LIVE/DEAD Aqua, Invitrogen), anti- huCD45 (2D1), anti-mouse (m)CD45 (30-F11), anti–huHLA-DR (g46-6), anti-huCD19 (HIB19), anti-huCD20 (L27), anti-huCD14 (M5E2), anti-huCD123 (9F5), anti-huCD1c (AD5-8E7; Miltenyi Biotec), and anti-huCD141 (AD5-14H12; Miltenyi Biotec). Unless otherwise stated, all Abs were purchased from BD Biosciences. Absolute numbers of DC subsets in a single femur, whole spleen, and per microliter whole blood were quantitated relative to a defined number of TruCOUNT (BD Biosciences) beads using an LSRII flow cytometer (BD Biosciences) and analyzed using FlowJo software (Tree Star) according to the gating strategy shown in Fig. 1.

CD141⁺ DC and CD1c⁺ DC were enriched from bone marrow by an initial depletion using rat anti-human CD14, CD19, CD20, and CD34 Abs and rat anti-mouse Ter-119 and CD45 Abs followed by sheep anti-rat IgG Dynabeads (Invitrogen). Cells were labeled with Abs and sorted using an FACSAria (BD Biosciences) on a live cell gate as mCD45⁻huCD45⁺huHLA-DR⁺CD19⁻CD20⁻ and CD141⁺CD1c⁻ or CD141⁻CD1c⁺. CD14⁺ cells, B cells, and pDC were sorted from unfractionated cell suspensions as mCD45⁻huCD45⁺huHLA-DR⁺ and CD19⁺CD20⁺, CD19⁻CD20⁻CD14⁺, or CD19⁻CD20⁻CD123⁺ events. Cytospins of purified cells were made with a Shandon Cytospin 4 cytocentrifuge (Thermo Scientific) and stained with Leishmans solution. Cell-surface phenotype was examined by staining with anti-CD11c (B-ly6), anti-CD11b (M1/70), anti-BDCA2 (AC144; Miltenyi Biotec), anti–DEC-205 [MMRI-7 (35)], anti–CLEC9A-FITC [23/05-4C6 (8)], CD40 (5C3), or CD83 (HB15e), acquisition on a CyAn flow cytometer (Beckman Coulter), and analyzed using FlowJo software (Tree Star).

Total RNA was isolated using an RNeasy Mini kit (Qiagen), treated with DNase I, and cDNA synthesized using a (dT)₁₅ primer (Roche) and Superscript III Reverse Transcription (Invitrogen). Real-time PCR for TLR3 was performed as previously described (6). Real-time PCR for XCR1, SIRPα, NECL2, TLR4, TLR7, and TLR9 were performed on an Applied Biosystems 7900HT Fast Real-Time PCR System (Applied Biosystems) with SuperArray RT2 qPCR Primers (Qiagen) and RT2 SYBR Green qPCR Master Mix (Qiagen). Ubiquitin C was used for normalization of cDNA input. Conditions used were: 95°C for 10 min, 45–50 cycles at 95°C for 15 s, and 60°C for 60 s. Data analysis was performed using SDS software (v2.3; Applied Biosystems) and the ΔΔ threshold cycle method. TLR3 PCR data analysis was performed using Rotor-Gene 6.0 software (Qiagen).

Purified DC were cultured in RPMI 1640 supplemented with 10% human AB serum, 100 U/ml penicillin, 100 μg/ml streptomycin, 2 mM l-glutamine, 1 mM sodium pyruvate, 0.1 mM nonessential amino acids, 10 mM HEPES buffer solution (all from Invitrogen), and 50 μM 2-ME (Sigma-Aldrich) in the presence or absence of poly I:C (10 μg/ml; Invivogen), CpG2216 (2 μM; Invivogen), or LPS (100 ng/ml Invivogen). Supernatants were harvested after 24 h. ELISA kits were used to detect IL-12p70 (BD Biosciences) and IFN-λ (IL-29; R&D Systems). IFN-α, TNF, and IL-6 were quantitated using FlowCytomix bead arrays (eBioscience).

For Ag uptake, DCs were pulsed with 50 μg/ml OVA–Alexa 488 for 1 h at 37°C and then washed and analyzed by flow cytometry. For Ag processing, HLA-A*0201⁺ DC isolated from huNOD/SCID mice were cultured alone, with the HLA-A*0201–restricted pp65_495–503 peptide (NLVPMVATV) from human CMV pp65 protein (ProImmune), or with 38.5 μg/ml rCMVpp65 protein (Miltenyi Biotec), in the absence or presence of adjuvants for 2 h. DCs were cocultured with a pp65_495–503-specific CD8⁺ T cell line. Culture supernatants were collected after 24 h and IFN-γ production assessed by ELISA (eBioscience). For in vivo targeting of Ab, huNOD/SCID mice were injected i.v. with Alexa 488–conjugated anti-human CLEC9A, DEC-205, or isotype control Ab. Bone marrow and spleens were harvested after 2 h and analyzed by flow cytometry.

huNOD/SCID mice were developed by reconstituting 5- to 6-wk-old NOD/SCID mice with human cord blood CD34⁺ progenitors (Fig. 1A). Reliable engraftment of human cells (in 94% of mice; n = 69), was obtained after 4 wk, with an average of 56% of human CD45⁺ cells in bone marrow and 34% in spleen (Fig. 1B). B cells were the most prevalent human leukocyte subset, comprising 47% of human CD45⁺ cells in bone marrow and 50% in spleen. DC subsets were identified as mouse CD45⁻ or huCD45⁺HLA-DR⁺ CD19/CD20⁻ events and were further subdivided by expression of CD141, CD1c, CD123 (pDC), and CD14 according to the gating strategy shown in Fig. 1. CD141⁺ DC and CD1c⁺ DC were found in similar proportions, comprising 0.44 and 0.35% of huCD45⁺HLA-DR⁺ cells in bone marrow, whereas pDC and CD14⁺ cells comprised 2.1 and 4.47%, respectively (Fig. 1C). All of these subsets could also be detected in spleen, although at much lower frequency compared with bone marrow (Fig. 1C).

FLT3-L is a potent growth factor for the expansion of large numbers of mouse and human DC in vivo including CD141⁺ DC, even in hematopoietic cell transplant patients (10, 32, 34, 36). We postulated that human FLT3-L administration would lead to the expansion of human DC subsets in huNOD/SCID mice. CD141⁺ DC, CD1c⁺ DC, pDC, and CD14⁺ cells could be detected in bone marrow, spleen, and blood of FLT3-L–treated mice (Fig. 2A) as well as in the liver and lungs (not shown). FLT3-L increased overall numbers of huCD45⁺HLA-DR⁺ by an average of 2-fold in femur, 8-fold in spleen, and 4-fold in blood (Fig. 2B). Although FLT3-L expanded CD141⁺ DC, CD1c⁺ DC, pDC, and CD14⁺ cells in all tissues examined, the largest expansion was found within the CD1c⁺ DC compartment, in which numbers increased 15-fold in bone marrow, 58-fold in spleen, and 23-fold in blood (Fig. 2B). CD141⁺ DC numbers increased on average 8-fold in bone marrow, 43-fold in spleen, and 18-fold in blood, whereas increases in pDC numbers were more modest, being 6-fold in bone marrow, 12-fold in spleen, and 10-fold in blood. There was also some expansion of the CD14⁺ compartment (3-fold in bone marrow, 31-fold in spleen, and 10-fold in blood). These data demonstrate that human FLT3-L administered to huNOD/SCID mice can dramatically enhance reconstitution of human DC subsets, with particular emphasis on expansion of the CD1c⁺ DC and CD141⁺ DC subsets.

We developed a procedure to isolate human DC from bone marrow of huNOD/SCID mice, which resulted in highly pure populations (mean purity >96%) of each subset. Using this procedure, we could isolate a mean of 3 × 10⁵ CD141⁺ DC from the total crushed bones of the humanized mice, similar to the numbers obtained from the costly leukapheresis procedures from human donors (Fig. 3A). The isolated subsets had morphological features similar to those isolated from human blood (Fig. 3B). Similar to human blood (3, 4, 6, 17, 37), CD141⁺ DC from huNOD/SCID mice expressed CD11c, DEC-205, CLEC9A, XCR1, NECL2, and TLR3, but not CD11b, SIRPα, BDCA2, TLR4, or TLR9 (Fig. 3C, 3D). Similar to their human blood counterparts, CD1c⁺ DC in FLT3-L–treated huNOD/SCID mice expressed CD11c, CD11b, DEC-205, SIRPα, and lower levels of TLR3 compared with CD141⁺ DC (Fig. 3C, 3D). pDC in humanized mice expressed BDCA2, DEC-205, and TLR9, whereas the CD14⁺ population resembled human monocytes and expressed CD11c, CD11b, DEC-205, SIRPα, and TLR4. These phenotypes were also observed in the non–FLT3-L control mice, except for CD1c expression, which was upregulated on CD141⁺ DC in response to FLT3-L (Fig. 3E) and HLA-DR expression, which was decreased on CD1c⁺ DC after FLT3-L treatment (Fig. 3F). Thus, DC and CD14⁺ cells that develop in FLT3-L–treated humanized mice closely resemble those circulating in human blood in morphology and phenotype.

We next examined the function of human DC subsets from FLT3-L–treated huNOD/SCID mice in response to TLR ligands. CD141⁺ DC and CD1c⁺ DC were stimulated with the TLR3 agonist poly I:C; pDC were stimulated with the TLR9 ligand CpG2216; and CD14⁺ cells were stimulated with the TLR4 agonist LPS. None of the freshly isolated DC or CD14⁺ cells expressed costimulatory molecules CD40 or CD83 (Fig. 4A). CD141⁺ DC spontaneously upregulated CD40 and CD83 following culture in the absence of exogenous stimuli. CD40 was further upregulated following poly I:C in some donors but not others, and no further upregulation of CD83 was observed (Fig. 4A). In contrast, CD40 was not expressed by CD1c⁺ DC, pDC, or CD14⁺ cells in the absence of exogenous stimuli. CD40 was upregulated on CD1c⁺ DC and pDC following activation with poly I:C and CpG2216, respectively, but was not expressed by CD14⁺ cells stimulated with LPS. CD83 was upregulated by CD1c⁺ DC, but not by pDC and CD14⁺ cells in the absence of activation and expressed by CD1c⁺ DC and pDC, but not CD14⁺ cells, after activation.

CD141⁺ DC isolated from FLT3-L–treated humanized mice were the major producers of IFN-λ in response to poly I:C (Fig. 4B). Poly I:C induced equivalent levels of IL-12p70 in both CD141⁺ DC and CD1c⁺ DC. As expected, pDC were the major producers of IFN-α in response to CpG2216 but produced little IFN-λ. CD141⁺ DC and CD1c⁺ DC also produced IFN-α in response to poly I:C; however, the levels were substantially lower (in the range of 100 pg/ml) compared with CpG-activated pDC (∼16 ng/ml). Activated DC subsets also produced modest levels of IL-6 and TNF; however, LPS-activated CD14⁺ cells were inefficient producers of these cytokines (Fig. 4B). These data demonstrate that CD141⁺ DC, CD1c⁺ DC, and pDC from huNOD/SCID mice are functional, whereas the CD14⁺ cells are defective in their response to LPS.

We next examined the capacity of DC from huNOD/SCID mice to present protein Ag. All three DC subsets and CD14⁺ cells were comparable in their ability to take up soluble OVA protein (Fig. 5A). To examine the ability of the DC subsets to directly present peptide Ag to CD8⁺ T cells, HLA-A*0201⁺ DC were pulsed with the HLA-A*0201–restricted CMV pp65_495–503 peptide and used to stimulate a pp65_495–503-specific CTL line. All subtypes could directly present peptide to Ag-specific CD8⁺ T cells, although the CD14⁺ population was slightly less effective compared with the DC subsets in this regard (Fig. 5B). We next investigated the ability of the DC subsets to cross-present protein Ag. rCMVpp65 matrix protein was added to the DC, and we examined their ability to process and present the pp65_495–503 epitope to a pp65_495–503-specific CTL line. Consistent with their human blood counterparts, all DC subsets isolated from humanized mice could weakly cross-present pp65 protein in the absence of exogenous stimuli, as read out by IFN-γ production by the CD8⁺ T cells (Fig. 5C). Cross-presentation was substantially enhanced when CD141⁺ DC were activated with poly I:C. Activation of pDC with CpG increased cross-presentation by this subset, but they were not as effective as CD141⁺ DC. Activation of CD1c⁺ DC and CD14⁺ cells did not enhance their cross-presenting capacity. These data demonstrate that, like human blood DC, poly I:C–activated CD141⁺ DC are superior in their capacity to cross-present protein Ag (Fig. 5C).

Targeting of vaccine Ag to mouse CD8⁺ DC in vivo using Abs against DEC-205 or CLEC9A is an effective vaccine strategy for inducing CTL responses in tumor and infectious disease models. However, it is unknown whether specifically targeting the human counterpart of mouse CD8⁺ DC (i.e., huCD141⁺ DC) or whether targeting multiple human DC subsets will be the most effective approach in humans. Consistent with broad expression of DEC-205 by multiple subsets, injection of anti-human DEC-205 into humanized mice efficiently targeted all human DC subsets and CD14⁺ cells in bone marrow within 2 h of uptake (Fig. 6). By contrast, injection of anti-huCLEC9A mAb specifically targeted only CD141⁺ DC. These data show that DC can be targeted in humanized mice in vivo.

In this study, we developed a mouse model in which human CD1c⁺ DC and CD141⁺ DC develop and are phenotypically and functionally similar to their equivalents in human blood. Earlier in vivo studies showed that administration of FLT3-L into healthy volunteers can dramatically increase numbers of both CD11c⁻ (corresponding to pDC) and CD11c⁺ DC subtypes into the circulation, with the greatest fold expansion observed within the CD11c⁺ compartment (32, 33). Although previous studies showed that the majority of the CD11c⁺ DC expanded by FLT3-L were CD1c⁺ (32), our data showed that the CD141⁺ DC subset also expressed the CD1c molecule in response to FLT3-L. Hence, whether FLT3-L preferentially enhances CD141⁺ DC in blood and other tissues, similar to the effect of mouse FLT3-L on their CD8⁺ DC counterparts (34), remained an open question.

In the current study, we demonstrated that FLT3-L dramatically increases numbers of CD1c⁺ DC, CD141⁺ DC, and pDC in blood, bone marrow, and spleen of humanized mice. pDC were expanded to a lesser extent than the cDC subsets, consistent with previous studies (32, 33). Importantly, the proportions of the CD1c⁺ and CD141⁺ DC subsets were expanded to a similar extent in blood and spleen, with a preferential expansion of the CD1c⁺ DC subset in the bone marrow. This concurs with our previous report showing that injection of human FLT3-L in mice resulted in equally large increases in both CD8⁺ and CD8^- DC and highlights an important interspecies difference in the ability of mouse or human FLT3-L to expand DC subsets (34).

The DC subsets that developed in humanized mice displayed phenotypic markers characteristic of their counterparts in human blood. FLT3-L did not alter the phenotype of the DC, with the exceptions that the CD1c molecule was upregulated on CD141⁺ DC and the CD1c⁺ DC subset expressed lower levels of MHC class II molecules in the bone marrow, which may indicate a less mature phenotype. In our study and in a similar model (38), CD141⁺ DC outnumber CD1c⁺ DC 1.5–2-fold in the blood of humanized mice in the absence of FLT3-L, which contrasts with human blood, in which the CD1c⁺ DC/CD141⁺ DC ratio is ∼8:1 (17, 37). Although the relative proportions of DC did not alter in the blood in response to FLT3-L, the ratio of CD1c⁺/CD141⁺ DC in the spleens of humanized mice increased to similar proportions found in human spleen (∼4:1) (17). This suggests that CD1c⁺ DC might be especially dependent on FLT3-L for their development. Increasing the frequency of FLT3-L administration and or coadministration with GM-CSF may be required to further increase the CD1c⁺ DC proportion in the blood of humanized mice to that in human blood (34).

Following FLT3-L administration, we could isolate similar numbers of highly purified CD141⁺ DC from a single humanized mouse as compared with a human leukapheresis product. CD141⁺ DC that developed in humanized mice, like their equivalents in human blood, uniquely expressed CLEC9A, XCR1, and NECL2 and highly expressed TLR3. Importantly, these DC were fully functional, as evidenced by their ability to upregulate costimulatory molecule expression and produce modest levels of IL-12p70 and substantial levels of IFN-λ after poly I:C activation. They also excelled at cross-presentation following activation with poly I:C. Collectively, these features align CD141⁺ DC isolated from humanized mice with the CD141⁺ DC described in human blood (3, 4, 6, 14). Moreover, the data identify bone marrow not only as a primary site for human hematopoiesis in immunodeficient mice, but also as a major reservoir of functional human CD141⁺ DC. Therefore, this is a practical model that can be used to assess CD141⁺ DC function in vitro or in vivo.

CD1c⁺ DC also phenotypically and functionally resembled their human blood counterparts by expression of CD11c, CD11b, DEC-205, and SIRPα and low levels of TLR3 (17, 37). Poly I:C stimulation increased expression of costimulatory molecules CD40 and CD83 and induced IL-12p70, TNF, and IL-6 production by CD1c⁺ DC. Production of IL-12p70 by poly I:C–activated CD1c⁺ DC was comparable to that of CD141⁺ DC, but differed from human blood CD1c⁺ DC, which are poor producers of IL-12, even after additional T cell–derived stimuli (6). This is unlikely due to FLT3-L because CD1c⁺ DC from non–FLT3-L–injected huNOD/SCID mice produced similar levels of IL-12 (not shown). Our data also showed that pDC in the huNOD/SCID model were phenotypically and functionally similar to those found in human blood (28–30). They expressed CD123, BDCA2, TLR7, and TLR9 and were major producers of IFN-α in response to CpG. CD14⁺ cells with morphologic and phenotypic characteristics of monocytes also developed in the huNOD/SCID model, but were only partially functional. They could take up soluble protein and present peptide Ag; however, they were poor producers of TNF in response to LPS. These data support other studies in humanized mice demonstrating the requirement for human cytokines such as CSF-1 for the development of fully functional myeloid cells of monocyte/macrophage lineage (22). Our study extends these observations and suggests that unlike monocyte/macrophages, CD141⁺ DC and CD1c⁺ DC do not require G-CSF or human GM-CSF to acquire full functional capacity.

During the course of our study, three independent groups have generated human CD141⁺ DC using different humanized mouse models (38–40). Meixlsperger et al. (38) identified CD141⁺CLEC9A⁺ DC in the bone marrow and spleens of NOD/SCID γc^−/− mice 4 mo after engraftment of newborn mice with human fetal liver. In another study, CD141⁺ DC-expressing CLEC9A were identified in the lungs of NOD/SCID B2m^−/− mice 4–6 wk after transplantation of adult G-CSF–mobilized CD34⁺ progenitors (40). Our data demonstrate that the CD141⁺ DC generated in our model exhibit the phenotypic and functional qualities characteristic of those in human blood and their mouse CD8⁺ DC counterparts. The use of cord blood as a source of progenitors alleviates the ethical and logistical issues associated with using fetal tissue or G-CSF–mobilized adult leukapheresis products. Combined with the relatively short time for engraftment (4 wk) and FLT3-L treatment, this is a practical and versatile model for generating large numbers of human CD141⁺ DC, CD1c⁺ DC, and pDC. Our unpublished data suggest the model can be readily transferred to other immunodeficient mouse strains, such as the NOD/SCID γc^−/−–A2 strain that facilitates the development of human T cells. This is therefore a powerful model to enhance our understanding of huDC subset biology and validate new vaccine constructs that specifically target huDC, particularly in which expression of uptake and or adjuvant receptors are markedly different between human and mouse DC.

We thank the collection and laboratory staff of the Queensland Cord Blood Bank at Mater Hospital, a member of AusCord, the Australian National Cord Blood Collection and Banking Network, for provision of cord blood samples for the study. We also thank Prof. Derek Hart for early discussions.

The authors have no financial conflicts of interest.