Reconstitution of Allogeneic Hemopoietic Stem Cells: The Essential Role of FcRγ and the TCR β-Chain-FCp33 Complex¹

The transplantation of purified allogeneic hemopoietic stem cells (SC)³ decreases the risk of graft-vs-host disease (GVHD), yet fails to establish reliable reconstitution capable of rescuing MHC-disparate allogeneic recipients from radiation-induced aplasia (1, 2, 3). These early studies suggested that additional cell types are required for SC alloengraftment across complete MHC barriers. Cotransplantation studies using purified allogeneic SC have demonstrated that the donor bone marrow (BM)-derived facilitating cell (FC) population permits reliable SC reconstitution across MHC-disparate barriers, results in donor-specific transplantation tolerance, and does not induce GVHD (3, 4, 5).

Immunoprecipitation studies of biotinylated FC lysates using the TCR β-chain (TCRβ) or CD3ε mAb revealed a unique 33-kDa protein named FCp33, together with TCRβ and CD3ε proteins in a single disulfide-linked complex on the FC cell surface (6). Recent studies have demonstrated that the majority of cells in the FC population express markers consistent with a plasmacytoid precursor dendritic cell (p-preDC) phenotype (5). Given that murine BM p-preDC have not been shown to express CD3ε and TCRβ on the cell surface, the importance of expression of these proteins on the FC has been questioned. However, several studies using knockout mice have demonstrated loss of facilitating function by the FC population when CD3ε^−/−, TCRβ^−/−, or RAG^−/− strains are used as FC donors (6, 7). These findings provide compelling evidence that CD3ε and TCRβ-chain gene expression are required for FC function and/or development.

TCRβ-associated signal transduction occurs via the transmembrane immunotyrosine receptors CD3ζ and FcRγ-chain (FcRγ) (8, 9, 10, 11, 12), in association with the CD3γδε core complex expressed on T cells. In the setting of TCR expression, CD3ζ and FcRγ are expressed as homodimers (ζζ or γγ) or heterodimers (ζγ), and are associated with Ag receptor signal transduction in T cells and NKT cells (13, 14, 15, 16, 17, 18, 19). CD3ζ is the dominant TCRβ-associated signal transduction molecule used by mature T cells. However, transfection of FcRγ into CD3ζ^−/− cells has demonstrated that FcRγ can function in TCRβ signal transduction and induce IL-2 transcription in vitro, yet evidence of a distinct physiologic role for TCRβ-FcRγ signaling in mature T cells has not been described in vivo (14, 20, 21, 22, 23). Although originally described as the γ-chain within the IgΕ receptor complex, FcRγ is the signaling molecule for several other receptors including FcγRIII (CD16) and the αβ and γδ TCRs (24).

Unlike mature alloreactive T cells, allogeneic FC transplantation is associated with the absence of GVHD and the induction of donor-specific tolerance, suggesting that TCRβ signaling differs between mature T cells and the FC (3, 4). Although mature T cells preferentially express and use the CD3ζ homodimer, differences in CD3ζ or FcRγ gene expression are noted during various stages of thymocyte development and presumably result in the use of alternative signaling pathways with specific downstream effects (12, 25, 26, 27, 28). These differences suggest that although both CD3ζ and FcRγ are capable of functioning in TCRβ-chain signaling, there is selective association with TCRβ during cell development. Given the loss of function in FC derived from TCRβ^−/− and CD3ε^−/− donors, we hypothesized that FC function uses a TCRβ-associated signaling pathway that is distinct from conventional alloreactive T cells responsible for GVHD.

The current study firmly establishes the molecular and biochemical expression of TCRβ and CD3ε within the FC population at the time of transplantation and investigates the mechanism of FC function in allogeneic reconstitution by assessing the expression and in vivo function of the TCRβ-signaling molecules, CD3ζ and FcRγ within the FC population. These findings further characterize the FC population and the unique TCRβ-FCp33 complex. Notably, this is the first evidence for a physiologic and potentially clinically relevant role for a TCRβ-FcRγ complex in allogeneic hemopoietic SC reconstitution in vivo.

Six- to 8-wk-old female C57BL/6J (B6), B10.BRSgSnJ (B10.BR) CD3ζ^−/− (B6.129S4-CD3ζ^tm1Lov), CD16^−/− (B6.129-FcγR3^tm1Sjv), and TCRβ^−/− (B6.129P2-TCRβ^tm1Mom/J) mice were purchased from The Jackson Laboratory. FcRγ^−/− (B6.129P2-Fcer1g^tm1Rav/J) mice were purchased from Taconic Farms. All animals were housed in a specific pathogen-free facility at the Dana-Farber Cancer Institutes under the guidelines of the National Institutes of Health for care and use of laboratory animals.

The following mAbs were used to sort CD8⁺TCR⁻ FC or CD8⁺CD3ε⁺ TCR⁻: anti-CD8α (53-6.7) PE, anti-TCRβ (H57-597) FITC, anti-TCR γδ chain (GL3) FITC, and anti-CD3ε biotin (145-2C11) with secondary streptavidin (SA) PE-Cy5. To sort ScaI⁺c-kit^dim/int SC: Ly6A/E (Sca-1) PE, c-kit biotin with secondary SA PE-Cy5, and a mixture of FITC-conjugated anti-lineage (Lin⁻) mAbs: B220, CD8α, GR-1, CD11b, αβTCR, and γδTCR were used. To sort B220⁺CD11b⁻CD11c^dim/int p-preDC: anti-B220 allophycocyanin, anti-CD11b FITC, and anti-CD11c biotin, with SA PE-Cy5 were used. All sort mAb and the hamster IgG isotype control for CD3ε were from BD Biosciences. Anti-FcRγ Ab and rabbit IgG isotype control were purchased from Upstate Biotechnology, anti-CD3ζ (3F67) from U.S. Biological, and murine anti-PDCA-1 Ab was purchased from Miltenyi Biotec.

BM preparation was performed as previously described (3, 4). Briefly, BM was isolated from the long bones of mice by flushing with cold HBSS (Invitrogen Life Technologies). After washing with HBSS, BM cells were resuspended in sterile cell sort medium (CSM; Invitrogen Life Technologies) (CSM:HBSS without phenol red, 2% FCS, 2 μg/ml HEPES buffer and 30 μg/ml gentamicin).

SC and FC from donor BM were sorted as previously described (3, 4). Briefly, BM cells were incubated for 30 min at 4°C with mAbs listed previously, to isolate murine SC as ScaI⁺ c-kit^dim/int Lin⁻. After incubation with primary Abs, the cells were washed twice with CSM and incubated with SA PE-Cy5 for 30 min. BM-derived FC were similarly isolated as CD8⁺αβγδTCR⁻ or as CD8α⁺CD3ε⁺αβγδTCR⁻. Spleen T cells were isolated as CD8α⁺αβγδTCR⁺ and p-preDC were purified as B220⁺CD11b⁻CD11c^dim/int phenotype. Cells were washed and resuspended to a final concentration of 2 × 10⁸ cells/ml in CSM before multiparameter sterile live cell sorting within the conventional lymphoid gate on a MoFlo flow cytometric cell sorter (Cytomation). Postsort purity was determined with respect to forward and side scatter parameters and the designated cell surface markers. Postsort purity for all experimental samples was ≥90%.

For allogeneic BMT, B10.BR recipients were lethally irradiated (950cGy) followed by transplantation with 50,000 FC from donor BM of normal B6, CD3ζ^−/−, FcRγ^−/−, TCRβ^−/−, or CD16^−/− mice and 10,000 purified donor SC obtained from normal B6 mice or control B10.BR mice received 10,000 donor SC alone. The syngeneic controls (B6 mice) received 2,000 purified donor SC obtained from B6 mice. To maintain homogeneity within a given experiment, recipients were of the same age, from same shipment of mice and underwent transplantation with donor populations. Allogeneic SC engraftment in the recipients was confirmed by flow cytometric PBL typing of donor MHC Ag expression 28 days posttransplant.

Sulfo-N-hydroxy succinimidester-LC-biotin (Pierce) was used for surface protein biotinylation, as described previously (6). Sorted FC were washed in cold PBS (pH 8) and resuspended in 200 μl of 0.5 mg/ml sulfo-N-hydroxy succinimidester LC-biotin in PBS. Cells were rotated end-over-end for 30 min at room temperature. After the incubation, cell suspensions were diluted to 1 ml with cold PBS and centrifuged at 2000 × g for 10 min, followed by solubilization in lysis buffer (20 mM Tris-HCl (pH 7.6), 150 mM NaCl, and 1% digitonin with proteinase inhibitors PMSF, iodoacetamide, aprotinin, and leupeptin) for subsequent immunoprecipitation.

Surface biotinylated or nonbiotinylated cell pellets were solubilized in lysis buffer and immunoprecipitations were performed by preincubating anti-CD3ζ, anti-FcRγ, or anti-TCRβ Ab with protein G-Sepharose beads (Amersham Biosciences). Equal numbers of cells were compared in each experiment. Reduced SDS-PAGE was conducted as described previously (29) using reducing SDS sample buffer (4% SDS, 20% glycerol, 125 mM Tris-HCl (pH 6.8), 0.025% bromphenol blue, and 10% 2-ME). Protein Sepharose pellets of coprecipitates in sample buffer were separated by either 11.5 or 12.5% PAGE. The proteins were transferred to ECL-polyvinylidene difluoride membrane (Amersham Biosciences) in Tris-glycine transfer buffer (pH 8.3) (25 mM Tris, 192 mM glycine, and 20% methanol) and blocked with 5% milk in PBS-T (PBS (pH 7.4) with 0.05% Tween 20). The membranes were blotted with primary Ab (anti-CD3ζ or anti-FcRγ) followed by blotting with HRP-conjugated secondary Abs. Biotinylated samples were blotted with SA-HRP. All blots were visualized using the Supersignal West Pico detection system (Pierce).

BM cells were harvested as previously described and resuspended in PBS. The cells were first stained extracellularly with surface Abs for CD8α (CD8α-PE) and TCRαβ and TCRγδ (αβγδTCR-FITC) for 30 min at 4°C. The cells were washed with PBS-BSA buffer (PBS with 0.05% BSA) then fixed with 2% paraformaldehyde in PBS for 7 min at room temperature, followed by incubation with 300 μg/ml mouse IgG in PBS-BSA for 15 min. The cells were permeabilize with 0.05% Triton X-100 in PBS-BSA for 10 min at room temperature. Primary Abs for intracellular staining (anti-FcRγ or isotype control rabbit IgG) were incubated for 45 min at 4°C. Goat anti-rabbit-Cy5 was used as the secondary Ab for intracellular FcRγ staining (Jackson Immunological) and was incubated for 40 min at 4°C. Cells were washed twice in PBS-BSA and resuspended in 2% paraformaldehyde in PBS.

Total RNA was isolated from FC, p-preDC, and Spleen T cells using the RNeasy System (Qiagen) and converted to cDNA (Ambion), as recommended by the manufacturers. Primer sequences for FcRγ, CD3ζ, CD3γ, CD3δ, CD3ε, β-actin, and TCRβ were selected from published sequences for conventional PCR amplification (7, 15, 30). β-actin, FcRγ, CD3γδε, and CD3ζ cDNAs were amplified for 35 cycles using cycling conditions 1 min each at 94°C, 57°C, and 72°C. TCRβ cDNA was amplified for 40 cycles with cycling conditions 1 min at 94°C, 2 min at 60°C, and 3 min at 72°C with a final extension of 10 min at 72°C. PCR products were visualized on a 1.5% agarose gel stained with ethidium bromide (EtBr).

Real-time PCR was performed using 1–3 μl of FC cDNA mixed with diethyl pyrocarbonate-treated water, SYBR green PCR master mix (Applied Biosystems) and the primer pair for CD3ζ (sense, CAATCCTGTGCCAGCGTCTT and antisense, TGGCCATGGACTCCACAGA) or FcRγ (sense, CAAGAT CCAGGTCCGAAAGG and antisense, GCATCTGCTTTCTCACGGCT) in a 20-μl reaction. The specific primer pairs used were designed with Primer Express software (Applied Biosystems). The specific cDNAs were amplified for 40 cycles using the Gene Amp 5700 Sequence Detection System (Applied Biosystems). Nontemplate controls and dissociation curves were used to detect nonspecific amplification and the formation of primer-dimers. All experiments were run in duplicate and gene expression was normalized to the expression of the housekeeping gene, GAPDH.

TCRβ transcript was amplified by conventional PCR using cDNA from FC, T cells (positive control), or p-preDC (negative control). The PCR products were electrophoresed for ∼1 h at 100 V on a 1.5% agarose gel. Gels were stained with EtBr at 1 μg/ml, rinsed and photographed under UV light. Capillary transfer was performed to transfer the DNA to the membrane (31).

All membranes were briefly rinsed in 3× standard saline citrate phosphate/EDTA, air-dried for 15 min and cross-linked for 2 h at 80–90°C. Prehybridization was performed using 1× Perfect-Hyb Plus hybridization buffer (Sigma-Aldrich) for 1 h at 40°C.

Five micrograms of biotin-conjugated DNA oligonucleotide probe, complementary to an internal region of the mature TCRβ transcript, was added to the solution and all membranes were incubated overnight at 40°C. Membranes were rinsed and washed with 1× PBST for 10 min then blocked with 0.5% nonfat milk in PBST for 30 min followed by 1-h incubation in SA-HRP solution. The membranes were washed for 1.5 h in PBS-T. Chemiluminescence was detected using the Super Signal West Pico detection system.

Although initially characterized as CD8⁺CD3ε⁺αβTCR⁻ and γδTCR⁻ (αβγδTCR⁻), significant controversy remains as to the identity and characteristic phenotype of the FC. The CD8⁺TCR⁻ FC population contains both CD3ε low and negative subsets as demonstrated by flow cytometric analysis of protein expression (6, 7). However, given the low level of CD3ε expression detected on the FC some investigators have questioned whether CD3ε is intrinsic to the FC at the time of transplantation.

In the setting of this controversy, we have characterized the gene transcription and protein expression of CD3ε within the CD8⁺CD3ε⁺αβγδTCR⁻ FC population to establish that CD3ε surface expression was not merely the result of increased background staining. As shown in Fig. 1,A, CD3ε protein expression as established vs an isotype control within the CD8α⁺αβγδTCR⁻ FC population is characterized by a CD3ε⁺ population (7.2 ± 0.27%) and a large CD3ε⁻ population. The CD3ε⁺ FC subset, purified from C57BL/6 (B6) BM using high-speed flow cytometric cell sorting within the lymphoid gate, was analyzed for gene expression of CD3ε and the other components of the CD3 core, CD3γ, and δ using RT-PCR. As demonstrated in Fig. 1 B, the CD3 core transcripts CD3ε, CD3γ, and CD3δ are expressed within the CD3ε⁺ FC population. These results establish that CD3ε gene and protein expression are present within a subset of the FC population at the time of transplantation and corroborates the requirement for CD3ε previously identified in the in vivo knockout studies (7). Given recent reports that demonstrated the FC population contains a p-preDC subset (5) and p-preDC have not been shown to express surface CD3ε, we hypothesized that p-preDC reside in the CD3ε⁻ FC subset. Using flow cytometric analysis, 99.8 ± 0.1% of all FC cells positive for the p-preDC marker (PDCA-1) are contained within the CD3ε⁻ FC subset. These findings support previous reports demonstrating the absence of surface CD3ε on p-preDC.

Similar to FC isolated from CD3ε^−/− donors, the failure of RAG^−/− or TCRβ^−/− FC donors to facilitate allogeneic SC engraftment across complete MHC barriers suggested that TCRβ rearrangement and expression must occur within the FC population. Unfortunately, an Ab capable of Western blotting the murine TCRβ protein has not been identified and therefore coprecipitation studies have been used as evidence that TCRβ protein is present within the FC (6). We hypothesize that the presence of TCRβ transcription in normal FC would suggest that the loss of facilitating function by RAG^−/− and TCRβ^−/− FC donors was due to the need for TCRβ expression on the FC at the time of transplantation. In contrast, the absence of TCRβ transcript in normal FC would suggest that RAG^−/− and TCRβ^−/− FC do not facilitate secondary to a failure in FC development. To determine whether the TCRβ transcript is present within the FC at the time of cotransplantation with allogeneic SC, total RNA was isolated from 20,000 FC, T cells (positive control) and B220⁺CD11b⁻CD11c^dim/int p-preDC (negative control) and analyzed for the presence of mature TCRβ transcript. RT-PCR was performed and the PCR products were analyzed by Southern blotting using an internal probe complementary to the mature TCRβ transcript. As evident in Fig. 2, the TCRβ transcript was detected in both the T cell and FC populations. However, as expected the p-preDC population does not express the TCRβ transcript, corroborating previous reports demonstrating the absence of the TCRβ transcript in murine p-preDC. Importantly, the presence of TCRβ transcript in the FC and the failure of TCRβ^−/− FC to facilitate SC reconstitution demonstrate that the requirement for TCRβ is characteristic of an FC subset separate from p-preDC yet critical to FC-mediated SC reconstitution.

We have now demonstrated that CD3ε and TCRβ transcripts are present within the FC population at the time of transplantation with allogeneic SC. Evidence that FC-mediated reconstitution is dependent on protein expression of CD3ε and TCRβ on the FC, together with the description of the TCRβ-FCp33 receptor (6, 7), provides significant evidence that a functional TCRβ receptor complex is present on the FC and is required for FC-mediated SC reconstitution. We hypothesized that the TCRβ-dependent FC function in FC-mediated SC reconstitution is dependent on one of the two described TCRβ signaling molecules, CD3ζ or FcRγ.

The TCRβ has the ability to associate with and use immunotyrosine receptors CD3ζ or FcRγ. We asked whether CD3ζ or FcRγ is critical for FC-mediated reconstitution of allogeneic SC in vivo and further characterized the nature of FC function using CD3ζ^−/− and FcRγ^−/− mice. Lethally irradiated (950 cGy) B10.BR mice were reconstituted with 10,000 donor SC isolated from normal B6 mice coadministered with 50,000 donor FC obtained from normal B6, CD3ζ^−/−, or FcRγ^−/− mice (Fig. 3). Control B10.BR mice received 10,000 allogeneic SC alone and syngeneic B6 mice served as controls for SC reconstitution and received 2,000 SC alone (Table I). FC were isolated by flow cytometric cell sorting for the standard CD8⁺αβγδTCR⁻ FC phenotype to compare our results with the previously established SC + FC model. Previous studies have shown that donor FC alone do not engraft nor induce tolerance. Furthermore, lethally irradiated B10.BR recipients of SC alone failed to survive longer than 25 days, as allogeneic SC alone do not reconstitute (Table I). As expected, the transplantation of SC plus FC from normal B6 donors into lethally irradiated B10.BR recipients resulted in 100% survival (Fig. 3). Surprisingly, B10.BR recipients reconstituted with SC from B6 donors and FC derived from CD3ζ^−/− donors also survived and engrafted, demonstrating that CD3ζ is not critical for FC function in allogeneic SC reconstitution (Fig. 3). The reconstituted recipients demonstrated >90% donor B6 chimerism. In contrast, all recipients reconstituted with B6 SC and FC from FcRγ^−/− donors failed to engraft resulting in a 35-day mortality of 100% (Fig. 3). Moreover, the failure of engraftment is not due to loss of p-preDC because FcRγ^−/− and normal B6 mice have similar numbers of p-preDC as determined by B220⁺CD11b⁻CD11c^dim/int flow cytometric analysis (4.5 ± 1.2% and 3.6 ± 0.9%, respectively). These findings suggest that FC-mediated SC reconstitution may be critically dependent on the expression of FcRγ within a different FC subset.

Given that FcRγ expression is important in CD16 function, we addressed the possibility that the requirement for TCRβ and FcRγ expression may reflect the coexpression of these molecules, with FcRγ functioning as part of the CD16 receptor independent of the TCRβ. Therefore, lethally irradiated B10.BR mice were reconstituted with SC isolated from normal B6 mice coadministered with FC isolated from the BM of normal B6 mice or CD16^−/− donors. As expected, all SC + FC recipients of normal B6 inocula survived. Similarly, recipients of FC isolated from CD16^−/− donors also exhibited 100% survival at 45 days (Table I). Donor lymphocyte typing of these recipients demonstrated 90.2 ± 6.4% donor chimerism, confirming that FC from CD16^−/− donors are capable of promoting allogeneic donor SC reconstitution. A composite of the in vivo experimental findings for controls and CD16^−/−, CD3ζ^−/−, TCRβ^−/−, and FcRγ^−/− FC donors is shown in Table I.

The current in vivo studies have now shown that FcRγ and TCRβ expression is required for FC function. We next characterized FcRγ expression within the FC at the molecular level. FC were analyzed for gene expression of CD3ζ and FcRγ. Studies of early thymic cell development have demonstrated that FcRγ gene expression is greater than CD3ζ expression within the precursor cells of early maturing thymocytes (12, 27). Therefore, it is likely that differential gene expression could be directly associated with the selective use of one of these proteins within FC, likely FcRγ. To assess differential gene expression within FC, real-time quantitative PCR was performed using equal concentrations of cDNA from FC. Gene expression of CD3ζ and FcRγ in FC was normalized to GAPDH gene expression and is shown as percent of GAPDH. FcRγ gene expression within the FC population is significantly greater than CD3ζ expression (Fig. 4,A). Moreover, the difference in gene expression was further confirmed by analysis of PCR products on an EtBr-stained agarose gel (Fig. 4 B).

We subsequently investigated whether the higher FcRγ gene expression in the FC population results in predominant expression of the FcRγ protein by Western blotting and FACS analysis. Equal numbers of FC and T cells were solubilized and sequentially immunoprecipitated with anti-FcRγ followed by anti-CD3ζ Abs. Subsequent analysis of protein expression by Western blotting revealed that within the FC population FcRγ protein expression is greater than CD3ζ, which was extremely low or undetectable (Fig. 5,A). As expected, CD3ζ was abundant in lysates from T cells. Similar results were seen when CD3ζ immunoprecipitations were performed first. To assess the degree of FcRγ expression in terms of the percentage of FcRγ⁺ cells, intracellular staining of FcRγ within the FC was performed and revealed that 84.13 ± 9% of the FC population expresses FcRγ (Fig. 5 B).

Given the abundance of FcRγ expression on FC and the dependence of FC-mediated SC reconstitution on FcRγ and TCRβ expression, we next examined whether FcRγ associates with the TCRβ complex on the FC in vitro, as part of the TCRβ-FCp33 receptor complex critical for FC-mediated SC reconstitution. Such an association would offer rationale to explain the absence of FC function in TCRβ and FcRγ-deficient donors. To confirm this hypothesis, immunoprecipitation of FC lysates were performed using anti-TCRβ Ab. TCRβ immunoprecipitation of surface biotinylated FC lysates demonstrates the presence of the TCRβ-FCp33 receptor on the FC (Fig. 6,A). Subsequent Western blotting with anti-FcRγ Ab of nonbiotinylated FC lysates following TCRβ immunoprecipitation revealed the 12-kDa FcRγ protein within the TCRβ immunoprecipitated complex on the FC (Fig. 6 B). These results provide evidence that FcRγ is associated with the unique TCRβ-FCp33 receptor complex expressed on an FC subset and previously shown to be critical for FC-mediated SC reconstitution.

Conventional BMT is limited to closely matched donor-recipient combinations by the increased incidence of engraftment failure and GVHD that results in the setting of greater genetic disparity. Recent focus has been on the identification of cell populations that will augment SC engraftment and decrease GVHD. FC are currently characterized by their unique ability to promote the reconstitution of purified allogeneic hemopoietic SC in completely MHC-disparate recipients, without inducing clinical evidence of GVHD. In this murine model, purified SC fail to reliably reconstitute irradiated fully allogeneic recipients, whereas the coadministration of SC and BM-derived FC from the same allogeneic donor results in stable allogeneic chimerism and donor-specific transplantation tolerance. Original flow cytometric studies identified “facilitating” function within the CD8⁺CD3ε⁺αβγδTCR⁻ FC subset (3, 4). Subsequent in vivo studies characterizing FC-mediated allogeneic SC reconstitution revealed the unexpected dependence on both CD3ε and TCRβ expression by FC, as FC derived from CD3ε^−/− and TCRβ^−/− donors fail to facilitate allogeneic SC reconstitution (6, 7). Biochemical analysis has resolved this dichotomy by demonstrating that the TCRβ is present on FC, but not within the conventional αβTCR complex. Instead, TCRβ on the FC is disulfide linked to FCp33, a unique 33-kDa protein that is distinct from previously described TCR proteins including pre-TCRα (6). Furthermore, the TCRβ-FCp33 heterodimer coprecipitates with CD3ε.

In the current study, we provide further molecular and in vivo functional characterization of the FC population and the TCRβ-FCp33 complex. Given that in vivo facilitation is dependent on TCRβ expression (6), evidence for the presence of a TCRβ-associated signaling molecule in FC would provide insight into the mechanisms associated with FC-mediated SC reconstitution. We now show for the first time that the FC function is dependent on the expression of the TCRβ-associated signaling molecule FcRγ, as demonstrated by the failure of FC derived from FcRγ^−/− donors to facilitate allogeneic SC reconstitution (Fig. 3). In contrast, facilitation of alloreconstitution is undeterred when FC from CD3ζ^−/− or CD16^−/− donors are used (Fig. 3 and Table I).

Fugier-Vivier et al. (5) showed that the majority of the CD8⁺αβγδTCR⁻ FC population consists of a B220⁺CD11c⁺ phenotype that resembles plasmacytoid precursor dendritic cells (p-preDC), a DC subset that does not express CD3ε and TCRβ on the cell surface. It is important to note that the FC population in these studies was isolated as CD8⁺αβγδTCR⁻ and thus included a CD3ε⁻ subset in which p-preDC should reside. Fig. 1 demonstrates that CD8⁺αβγδTCR⁻ FC population contains both a CD3ε⁺ and CD3ε⁻ subset. Fig. 2 confirms that cells other than p-preDC are also contained in the FC population as the TCRβ transcript is not present within the p-preDC population, but is clearly identified in the CD8⁺αβγδTCR⁻ FC population. Because facilitation has been shown to be dependent on CD3ε and TCRβ expression, and p-preDC express neither, it is not unexpected that p-preDC do not facilitate SC reconstitution as effectively as the entire CD8⁺αβγδTCR⁻ FC population or the CD3ε⁺ FC subset (3, 4, 5). Together with the present data, this evidence supports the hypothesis that TCRβ, CD3εand FcRγ expression is intrinsic to the FC at the time of transplantation and are required for FC-mediated SC reconstitution. We now show that all three required components are incorporated within the TCRβ-FCp33 receptor complex present within the FC population.

The FC requirement for TCRβ and FcRγ but not CD3ζ expression for facilitation is similar to studies of early thymic cell development in which FcRγ gene expression is greater than CD3ζ expression within thymic precursor cells. Such differential expression suggests that CD3ζ and FcRγ are not merely interchangeable, but in fact demonstrate specificity requirements during the development of self tolerance in thymocytes (12, 15, 25, 27). Similarly, we now demonstrate that FcRγ gene expression is dominant within the FC population (Fig. 4), at a time when the FC is involved in the establishment of SC reconstitution and the induction of transplantation tolerance. We have also demonstrated that the dominance in FcRγ gene expression directly correlates with the predominant expression of FcRγ protein (Fig. 5). Interestingly, the extent of FcRγ expression (84 ± 9%) suggests that at least some p-preDC within the FC population are FcRγ positive. Therefore, it is possible that p-preDC development is inhibited in FcRγ^−/− donors resulting in the loss of FC-mediated SC reconstitution. We have excluded this hypothesis by demonstrating that the percent of BM p-preDC in FcRγ^−/− and normal B6 donors is similar at 4.5 ± 1.2% and 3.6 ± 0.9%, respectively. Thus, the loss of facilitation when using FcRγ^−/− donor is not due to the absence of p-preDC and therefore suggests a difference in the signaling requirement for FC and p-preDC function in facilitation. This is consistent with the absence of CD3ε and TCRβ expression on the p-preDC as demonstrated in this study and the less efficient facilitation of SC reconstitution exhibited by p-preDC as demonstrated by Fugier-Vivier et al. (5).

Given the requirement for TCRβ expression and the predominance of FcRγ-expressing cells within the FC population, we hypothesized that FcRγ associates with the TCRβ-FCp33 complex present on a subset of the FC. This hypothesis was supported by TCRβ-FcRγ coprecipitation studies and subsequently supports the hypothesis that the signaling requirement for the less efficient p-preDC population differs from the FC subset that contains the TCRβ-FCp33-FcRγ (Fig. 6).

Although the functional importance of TCRβ signaling via FcRγ in hemopoietic SC reconstitution in vivo has not previously been identified, the concept that FcRγ plays an important role in hemopoietic SC reconstitution and induction of tolerance has gained support through investigations with NK cells. The CD16 receptor on NK cells plays a pivotal role in fetal thymocyte and NK/NKT development via an FcRγ-signaling pathway, and cross-linking of the CD16 receptor expressed on pre-T cells has been shown to alter development and induces a higher ratio of αβ to γδT cells (32). Demonstration that immature thymocytes that express FcRγ and CD16 contain precursors for both NK cells and αβ T cells, suggests that CD16-FcRγ signaling is essential to early precursor development and thus may also play a role in FC development. The dual dependence of allogeneic SC reconstitution on TCRβ and FcRγ expression suggests that FC use either a common TCRβ-FcRγ complex, as demonstrated in early thymocytes in vitro, or independent TCRβ and FcRγ-signaling pathways, as in NKT cells which use both TCRβ-CD3ζ and CD16-FcRγ-signaling pathways. This latter possibility was effectively excluded by demonstrating that FC derived from CD3ζ^−/− or CD16^−/− donors remain functional and result in reconstitution of allogeneic SC across complete MHC barriers (Table I). Therefore, the demonstration that the FC function requires FcRγ, but not CD3ζ or CD16 expression, suggests the presence of an alternative FcRγ-associated receptor complex. This association was demonstrated by the coprecipitation of TCRβ, FCp33, and FcRγ from the FC population and provides the first evidence that a TCRβ-FcRγ complex is associated with FC-mediated allogeneic SC reconstitution (Fig. 6).

In summary, these studies establish the expression of TCRβ and CD3ε within the FC population at the time of transplantation. Furthermore, the FC population is distinct from conventional T cells and contains a CD3ε⁺ subset separate from the less efficient facilitating p-preDC cells that do not express CD3ε or TCRβ proteins. In contrast to CD3ζ, a protein that does not affect allogeneic SC reconstitution, we have demonstrated that FcRγ expression is critically important for FC-mediated facilitation of allogeneic SC reconstitution in vivo. The coprecipitation studies revealed that FcRγ associates with the TCRβ-FCp33 receptor, known to be required for facilitation of allogeneic SC reconstitution in vivo. Further characterization of the TCRβ-FCp33-FcRγ complex and FcRγ-dependent signaling will be critical to understanding the mechanism by which the FC facilitates allogeneic SC reconstitution and for the identification of potential therapeutic targets for the clinical induction of tolerance in the future.

We sincerely thank the staff at the Redstone Animal Facility/Dana-Farber Cancer Institute (DFCI) for outstanding animal care. We also thank Renee Wright, Rahilya Napoli, and Evan Cohick for technical assistance, and Peter Schow at the DFCI Flow Cytometry Core Facility for cell sorting and technical assistance.

The authors have no financial conflict of interest.