Boost Allograft Survival with Targeted Antigen Presentation

cDC1s are critical for donor antigen plus CoB-mediated cardiac allograft survival. Presentation of donor antigen by professional antigen-presenting cells to host T cells is a crucial component of the anti-donor response after transplantation (21), and CD11c+ DCs are known to internalize donor antigen, especially within the spleen as compared with other locations such as the lymph nodes or liver (22). The field of immunology has also come to appreciate the heterogeneity of DCs and more importantly, the subsequent identification of unique DC subsets, which differ in development, surface phenotype, and function (23, 24), and likewise play disparate roles within an immune response (17, 25). To identify localization of donor antigen to a specific population of antigen-presenting DCs, BALB/c splenocytes were labeled with a membrane fluorophore (PKH-67) and injected into C57BL/6J (B6) WT mice. The distribution of PKH-67+ cells 18 hours after injection was investigated by flow cytometry (Supplemental Figure 1; supplemental material available online with this article; https://doi.org/10.1172/JCI178960DS1), which revealed alloantigen was taken up by and localized predominantly within the XCR1hi CD172lo cDC1 subset (Figure 1A). Given these results, the known efficiency in which cDC1s present antigen to T cells, and recent findings indicating a role for cDC1s in self-tolerance (15–17), we hypothesized cDC1s are involved in the promotion of a tolerogenic response to a solid organ allograft.

cDC1s are required for donor splenocyte transfusion and anti-CD40L costimulation blockade–mediated cardiac allograft survival. (A) Uptake of PKH membrane–labeled CD45.1+ donor splenocytes 18 hours after i.v. infusion by CD45.1– B6 recipient splenic DCs measured by flow. n = 4 per group. **P < 0.01 by 2-tailed unpaired t test. (B) Flow cytometry gating strategy for cDC1 and cDC2 identification in spleens of B6 and cDC1-KO mice. Flow cytometry gating strategy to estimate the amount of splenic cDC1 and cDC2 cells in cDC1-KO and B6 mice as well as the ratio of cDC1/cDC2 cells. n = 6 per group. ****P < 0.0001 by 2-tailed unpaired t test. (C) Heterotopic heart transplantation experimental scheme. (D) Survival of BALB/c cardiac allografts in B6 and cDC1-KO mice as determined by manual palpation with rejection occurring at complete cessation of heartbeat and a palpation score of 0. n = 9–11 per group. ****P < 0.0001 by log-rank test. (E) Mean palpation score of BALB/c cardiac allografts in B6 and cDC1-KO mice. (F) Measurement of fractional shortening of cardiac allografts by echocardiography 42 days after transplantation. n = 3 per group. *P < 0.05 by 2-tailed unpaired t test.

Prior studies have detailed how transfusion of donor splenocyte antigen (DST) in combination with anti-CD40L CoB can induce long-term (>100 days) antigen-specific tolerance to a variety of solid organ allografts, including the cardiac allograft (26–29). We acquired an Irf8 +32–/– cDC1-KO mouse on a B6 background, which was genetically deficient in the cDC1 subset due to deletion of an enhancer 38 kb downstream of the IRF8 transcriptional start site required for cDC1 fate specification and development (30). We confirmed deletion of splenic cDC1s (Figure 1B) and subjected cDC1-KO and B6 mice to full-MHC mismatch heterotopic heart transplantation (31) utilizing BALB/c donors while employing a DST + CoB tolerization strategy whereby recipient mice received i.v. infusion of donor splenocytes and anti-CD40L (DST + CoB) on the day of transplant (day 0), followed by an additional dose (i.p.) of anti-CD40L on day 7 (Figure 1C, detailed in Methods). Recipient allograft function and survival were assessed and scored by manual palpation with allograft rejection identified as complete cessation of pulsation (32). Mice lacking cDC1s exhibited significantly earlier allograft rejection (Figure 1D) and a concomitant decrease in mean palpation score (Figure 1E) compared with immunocompetent B6 mice. We observed no sex differences in allograft survival within genotypes (Supplemental Figure 2). Echocardiography of allografts on day 42 after transplantation provided quantitative corroboration of clinical palpation scores as it revealed a concordant decrease in ventricular fractional shortening of cDC1-KO allografts (Figure 1F). Additionally, in cardiac allografts that were still functional at day 65 after transplantation prior to harvest, cDC1-KO allografts were found to be of increased weight (Figure 2A) and have a disorganized tissue structure with noticeable quantities of cellular infiltrates compared with B6 allografts (Figure 2B). Utilizing flow cytometry, we were able to further confirm and identify a significant increase in CD3+ T cells within cDC1-KO cardiac allografts at this time point (Figure 2C). It has been shown that treatment with anti-CD40L is able to not only prevent induction of alloantibodies but also dissolve established germinal centers (33). Given the growing body of work demonstrating the role B cells and antibody-mediated rejection can play in cardiac transplantation (34), we assessed peripheral blood for donor-specific antibodies and the quantity of CD19+ B cells in the spleen. The results showed no difference in CD19+ B cell numbers in cDC1-deficient transplant recipients (Supplemental Figure 3A) or in anti-BALB/c IgG levels across different dilution ratios compared with controls (Supplemental Figure 3B). Anti-BALB/c antibody levels in cDC1-KO mice were comparable to the B6 mice at day 65 after transplantation, though there was a trend toward decreased IgG in cDC1-KO recipients (Supplemental Figure 3C). These findings confirmed the allograft rejection was not antibody mediated.

cDC1s are critical for CD25+FoxP3+ T cell induction in cardiac allograft in the setting of DST + CoB. (A) Weight of cardiac allografts 65 days after transplantation. n = 3–4 per group. *P < 0.05, by 2-tailed unpaired t test. (B) H&E staining of cardiac allografts 65 days after transplantation. Scale bar: 50 μm. (C) Quantification of CD3+ T cells in cardiac allografts 65 days after transplantation. n = 3 per group. *P < 0.05, by 2-tailed unpaired t test. (D) Quantification of CD4+ and CD4+CD25+FoxP3+ T cells in hearts of naive mice, recipient orthotopic (host) heart, and recipient heterotopic (allograft) heart 7 days after transplantation. n = 4 per group. **P < 0.01 by 1-way ANOVA followed by Tukey’s test. (E) Representative flow plots of CD25+FoxP3+ T cells in B6 and cDC1-KO cardiac allografts 7 days after transplantation. Cells were pre-gated as live single CD3+CD4+ cells. (F) Quantification of CD4+CD25+FoxP3+ T cells in recipient heterotopic (allograft) hearts 7 days after transplantation in B6 and cDC1-KO mice. Data shown as cells/mg allograft tissue and as the percentage of CD4+CD25+FoxP3+ T cells within CD4+ and CD4+FoxP3+ cell populations. n = 4 per group. **P < 0.01, ***P < 0.001 by 2-tailed unpaired t test.

cDC1s are critical for activation of cardiac allograft CD4+FoxP3+ T cells. As the causal importance of T cells in cardiac allograft rejection has been well established (35–37), and cDCs function in the necessary priming of alloreactive conventional T cells or Tregs (8, 12, 38, 39), we next sought to delineate the contribution of cDC1s in the early immune response to the cardiac allograft. We hypothesized it is the early priming by cDC1s of T cells toward a more tolerant versus activated phenotype that is critical for creating an immunological environment to ultimately allow for acceptance of the allograft. Although no difference in the number of CD4+ T cells was observed in the transplant recipients’ orthotopic or heterotopic allograft hearts 7 days after transplantation, we observed a significant decrease in CD4+CD25+FoxP3+ cells in the cardiac allograft of mice deficient in cDC1s (Figure 2, D and E). Closer analysis of the CD4+ T cell population revealed that, though both cDC1-KO and B6 WT mice had similar numbers of CD4+FoxP3+ cells (Figure 2, E and F), the proportion of these cells expressing the activation marker CD25, which is known to be important for the homeostasis and survival of Tregs (40), was significantly lower in cDC1-KO allografts. This remained true when comparing the proportion of CD4+CD25+FoxP3+ cells present within both the overall CD4+ T cell population and the smaller CD4+FoxP3+ population (Figure 2F). This phenotype was also observed within the spleens of cDC1-KO and B6 WT transplant recipient mice (Supplemental Figure 4).

Antigen-specific regulatory CD4+ T cells are induced by cDC1s. Tregs, defined by their expression of CD25 and FoxP3, are of great interest within the transplant community as a result of their functional role in suppressing antigen-specific alloreactive T cells and promoting states of tolerance (41). Additionally, CD4+CD25+ regulatory cells have been shown to be induced by and required for DST + CoB–mediated allograft tolerance (42, 43). A variety of pathways by which peripheral CD4+ T cells become CD25+FoxP3+ Tregs have been described, though mechanistic gaps remain, particularly in the setting of alloantigen and solid organ transplantation (43–46). Given our interest in understanding the broader role cDC1s play in this induction of CD4+CD25+FoxP3+ Tregs that are responsive to alloantigen, we utilized a non–heart transplant model in which B6 WT and cDC1-KO mice received DST + CoB treatment on day 0 followed by persistent exposure to donor antigen through infusion of DST every other day until harvest at day 7 (47) (Figure 3A). Since T cell priming and antigen presentation by cDCs occurs predominately within lymph nodes and secondary lymphoid organs, we focused our studies within the spleen.

cDC1s are necessary for induction of splenic CD25+FoxP3+ T cells after DST + CoB. (A) Persistent antigen stimulation experimental scheme whereby B6 and cDC1-KO mice were treated with CD45.1+ BALB/c DST + CoB infusion on day 0 (i.v.), followed by CD45.1+ BALB/c DST injections (i.p.) on days 2, 4, and 6 before collection of spleens on day 7. (B) Quantification of CD45.1– adaptive immune cell populations (CD3+ or CD19+) in spleens of naive and persistent antigen-treated B6 or cDC1-KO mice. n = 4–5 per group. *P < 0.05 by 1-way ANOVA with Tukey’s test followed by Tukey’s test. (C) Flow cytometry gating strategy for CD45.1–CD4+CD25+FoxP3+ T cell identification in spleens of B6 and cDC1-KO mice. (D) Quantification of CD45.1–CD8+ and CD45.1–CD4+ T cell populations in spleens of naive and persistent antigen-treated B6 or cDC1-KO mice. n = 4–5 per group. *P < 0.05, **P < 0.01 by 1-way ANOVA followed by Tukey’s test. (E) Quantification of CD4+FoxP3+ and CD4+CD25+FoxP3+ T cells in spleens of naive and persistent antigen-treated B6 or cDC1-KO mice. Cells were pre-gated as live single CD3+ CD4+ CD45.1– cells. Data shown as cells/mg splenic tissue and as the percentage of CD4+FoxP3+ cell population. n = 4–5 per group. *P < 0.05, **P < 0.01 by 1-way ANOVA followed by Tukey’s test. (F) Congenic transfer of CD90.1+ OT-II CD4 T cells into B6 and cDC1-KO mice on day –1 followed by persistent membrane-bound ova + CoB and persistent ova antigen stimulation. Representative gating and flow plots of CD3+CD90.1+CD25+FoxP3+ OTII T cells specific to ova antigen in B6 versus cDC1-KO spleens. Cells were pre-gated as live single CD3+ CD90.1+ cells. (G) Quantification of CD90.1+FoxP3+ OTII and CD90.1+CD25+FoxP3+ OTII T cells in B6 or cDC1-KO mice after ova + CoB and persistent ova antigen stimulation. n = 5 per group. *P < 0.05 by 2-tailed unpaired t test.

We observed no differences in the quantity of CD19+ B cells in B6 mice and cDC1-KO naive mice before or after DST + CoB with persistent antigen infusion, but we did observe a decrease in the number of splenic CD3+ T cells in cDC1-KO mice after treatment (Figure 3B). We sought to identify the subset of recipient T cells contributing to this observed difference (gating strategy Figure 3C) and observed a decrease in the number of CD8+ and CD4+ T cells in cDC1-KO mice compared with B6 mice after DST + CoB treatment (Figure 3D). Although the role of cDC1s in antigen presentation and stimulation of CD8+ T cells has been well characterized (7, 12), we were very intrigued to observe this decrease in CD4+ T cells given our prior results in which both cardiac allografts and spleens of cDC1-KO mice had fewer CD4+CD25+FoxP3+ T cells 7 days after transplantation. Indeed, we observed the same results in which no difference in CD4+FoxP3+ T cell numbers was observed after DST + CoB and persistent antigen infusion; however, a significant decrease in the absolute quantity of splenic CD4+CD25+FoxP3+ T cells was seen in treated cDC1-KO mice (Figure 3E). We also characterized cells of the innate immune system in this setting and observed no differences in the quantity of cells in naive mice or mice treated with DST + CoB and persistent antigen infusion (Supplemental Figure 5).

Given that it was the specific subset of CD25+FoxP3+ T cells affected by cDC1 deficiency after both transplantation and alloantigen exposure, and that CD25 is a marker of T cell activation and proliferation in both regulatory and effector T cells (48), we hypothesized cDC1s may be functioning in an antigen-specific manner to promote and activate this regulatory response to alloantigen. Therefore, we transferred congenic CD90.1+ OT-II splenic CD4+ T cells, which recognize ovalbumin (ova) presented by I-Ab MHC class II into B6 and cDC1-KO mice on day –1, a strategy proven to efficiently promote T cell activation. We subsequently treated these mice with our DST + CoB and persistent antigen infusion protocol using splenocytes from mice constitutively expressing membrane-bound ova for each DST infusion. Indeed, when assessing CD90.1+ ova antigen-specific T cells, we observed fewer FoxP3+ and CD25+FoxP3+ antigen-specific cells in cDC1-KO mice (Figure 3, F and G), suggesting that cDC1s are important in the conversion of OTII conventional T cells into Tregs or in the expansion of OTII Tregs after CoB.

Expression of TGF-β1 is increased in cDC1s after exposure to alloantigen in the setting of costimulation blockade. Because our results continued to emphasize the importance of cDC1s in the induction of antigen-specific CD4+CD25+FoxP3+ T cells, we sought to delineate the means by which cDC1s promote this response after exposure to alloantigen. Thus, we infused DST + CoB into B6 and cDC1-KO mice and assessed the phenotype of treated versus naive splenic DCs (Figure 4A). We saw no differences in cDC1 or cDC2 cell numbers after DST + CoB infusion when compared with control injection in WT mice, but did observe a slight increase in the number of cDC2s after DST + CoB in cDC1-KO mice (Figure 4B). As we began to evaluate candidate tolerogenic or regulatory ligands on cDC subsets, we identified a notable increase in cDC1 expression of membrane-bound TGF-β1 after DST + CoB treatment (Figure 4C). Interestingly, this intensity of TGF-β1 expression was higher in cDC1s when compared with cDC2s (Figure 4C), which may indicate a distinct function of this DC subset.

After exposure to allogenic cells, cDC1s increase expression of TGF-β1. (A) Acute DST + CoB experimental scheme whereby B6 and cDC1-KO mice were treated with CD45.1+ BALB/c DST + CoB infusion D0 (i.v.) before collection of spleens after 48 hours. (B) Quantification of cDC1 (Xcr1+CD172–) and cDC2 (Xcr1–CD172+) cell populations in spleens of naive and DST + CoB–treated B6 or cDC1-KO mice. Cells were pre-gated as live single CD11c+MHCIIhi cells. n = 5–7 per group. **P < 0.01 by 1-way ANOVA followed by Tukey’s test. (C) Expression of TGF-β1 in cDC1 and cDC2 splenic cells in naive and acute DST + CoB–treated mice. n = 3–4 per group. *P < 0.05, ***P < 0.001, ****P < 0.0001 by 1-way ANOVA followed by Tukey’s test. (D) In vitro allogenic stimulation experimental scheme whereby Flt3L BM-derived DCs were cultured for 8 days and exposed to CD45.1+ BALB/c DST or saline control for 48 hours. (E) Expression of canonical markers of DC activation (CD40, CD80, CD86) and TGF-β1 of cultured cDC1s following allogenic cell exposure. n = 5–6 per group. **P < 0.01 by 2-tailed unpaired t test. FMO, fluorescence minus one.

Although these results highlight an important in vivo response, it is important to acknowledge that these DCs were likely subjected to a variety of additional environmental signals aside from simple exposure to alloantigen, including crosstalk with other immune cells. Therefore, we wanted to see whether this induction of TGF-β1 expression was cell-intrinsic and could be observed when DCs were isolated and exposed to alloantigen in a controlled cellular environment. It is now well accepted that the receptor tyrosine kinase Flt3 is required for DC development (49); thus, in vitro culture of BM-derived DCs (BMDCs) must utilize Flt3 ligand (Flt3L) to generate DC subsets with biological relevance that mimic an in vivo response (50). As such, we used the Flt3L culture system to generate a high quantity of biologically relevant DCs (Supplemental Figure 6A). After differentiation, BMDCs were exposed to CD45.1+ BALB/c DST or saline and harvested 48 hours later (Figure 4D). Within the cDC1 population, we observed no changes in the expression of canonical markers of DC activation, including CD40 and CD80, but did identify an increased expression of cDC1 CD86 and TGF-β1 (Figure 4E and Supplemental Figure 6B).

Membrane-bound cDC1 TGF-β1 is necessary for antigen-specific induction of regulatory CD4+ T cells. TGF-β1 is known to induce Foxp3 gene expression, to mediate the conversion of CD4+CD25– conventional T cells into a CD4+CD25+ Treg population in vitro, and to be critical for generation of the Treg population and T cell tolerance in vivo (48, 51, 52). Additionally, membrane-bound TGF-β1 on CD4+ T cells and human DCs has been reported to have immunoregulatory functions in contact-dependent suppression of an effector cell population (53–55). Thanks to new genetic tools that allow for specific deletion of genes in the cDC1 cell population (Xcr1Cre/+) (8), we crossed Xcr1Cre/+ mice with mice containing loxP flanked sites on exon 3 of the Tgfb1 gene (TGF-β1fl/fl) to delete membrane-bound TGF-β1. Xcr1cre/+ TGF-β1fl/fl mice retained normal numbers of cDC1 and cDC2 cells in the spleen (Supplemental Figure 7, A and B). Additionally, the levels of CD40, CD80, and CD86 were similar between Xcr1cre/+ TGF-β1fl/fl and TGF-β1fl/fl mice (Supplemental Figure 7C), suggesting that TGF-β1 deficiency did not affect cDC1 development.

We subsequently performed an adoptive transfer of antigen-specific CD4+ T cells with persistent antigen exposure protocol: CD90.1+ OT-II splenic CD4+ T cells were infused into TGF-β1fl/fl and Xcr1Cre/+ TGF-β1fl/fl mice on day –1 followed by treatment with ova mouse DST + CoB and persistent ova antigen infusions for 1 week (Figure 5A). We then assessed the antigen-specific response of the transferred CD90.1+ OTII cells (Figure 5B). Notably, no difference was observed between naive or treated TGF-β1fl/fl and Xcr1Cre/+TGF-β1fl/fl mice in the overall splenic CD3+, CD4+, and CD4+CD25+FoxP3+ T cell populations (Figure 5C), largely because this is an isogenic infusion. Although a similar number of CD90.1+ OTII cells were recovered from spleens of TGF-β1fl/fl and Xcr1Cre/+TGF-β1fl/fl treated mice (Figure 5D), we did, indeed, observe a significant decrease of ova antigen-specific OTII CD25+FoxP3+ T cells in terms of both the absolute cell number and as a percentage of the OTII cell population (Figure 5, E and F). Moreover, Xcr1Cre/+ TGF-β1fl/fl mice indeed rejected their allograft earlier than TGF-β1fl/fl mice (Supplemental Figure 7, D and E). We further confirmed the importance of TGF-β in in vitro induction of naive CD4 T cells into CD4+CD25+FoxP3+ Treg cells by Flt3L BMDC coculture. BMDCs were either stimulated with saline or UV-irradiated BALB/c DST for 48 hours and then cocultured with anti-CD3/CD28 plate-bound naive CD4 T cells supplemented with Il-2 in the presence of saline or anti–TGF-β neutralizing antibody (Supplemental Figure 8A). After a 5-day coculture, DST-stimulated BMDCs had increased induction of CD4+CD25+FoxP3+ Treg cells compared with the saline BMDC-treated control; however, this induction was ablated in anti-TGF-β–treated cocultured cells (Supplemental Figure 8, B and C). Furthermore, anti-TGF-β–treated CD4+CD25+FoxP3+ T cells had reduced expression of PD-1 (CD279) compared with DST-stimulated BMDC coculture alone (Supplemental Figure 8D). Interestingly, PD-1 expression has been shown to be associated with antigen-specific TCR activation of CD25+ Tregs in a manner similar to CTLA-4 expression and to be associated with CD44 expression (56). TGF-β1 has also been observed to induce PD-1 expression in T cells via Smad3 (57). These results indicate a requirement of cDC1 membrane-bound TGF-β1 as a mechanism for induction and function of antigen-specific CD4+CD25+FoxP3+ T cells and allograft tolerance.

cDC1-expressed TGF-β1 is necessary for induction of antigen-specific CD25+FoxP3+ T cells. (A) Antigen-specific persistent antigen simulation experimental scheme whereby TGF-β1fl/fl and Xcr1Cre/+TGF-β1fl/fl mice were injected (i.v.) with CD90.1+ OTII T cells on day –1 and treated with ova DST + CoB infusion (i.v.) on day 0, followed by ova DST injections (i.p.) on days 2, 4, and 6 before collection of spleens on day 7. (B) Flow cytometry gating strategy for identification of splenic antigen-specific congenic CD90.1+ OTII cells from the spleens of ova DST + CoB–treated mice. (C) Quantification of endogenous splenic CD3+, CD4+, and CD4+CD25+FoxP3+ T cells. n = 5–6 per group. Determined no significance (ns) by 1-way ANOVA followed by Tukey’s test. (D) Quantification of splenic CD90.1+ OTII T cells 7 days after ova DST + CoB and persistent ova antigen treatment. n = 5–6 per group. Determined no significance (ns) by 2-tailed unpaired t test. (E) Quantification of splenic CD90.1+ OTII CD25+FoxP3+ T cells. Data shown as cells/mg splenic tissue and as the percentage of OTII cell population. n = 5–6 per group. *P < 0.05, **P < 0.01 by 2-tailed unpaired t test. (F) Representative flow plots of OTII CD25+FoxP3+ T cells in spleens of persistent antigen-treated TGF-β1fl/fl and Xcr1Cre/+TGF-β1fl/fl mice. Cells were pre-gated as live single CD90.1+ TCRVβ5.1+ cells. FMO, fluorescence minus one.

Mitochondrial metabolism is increased in cDC1s after exposure to alloantigen in the setting of costimulation blockade. Much work has been done to understand the downstream signaling by TGF-β1 in a variety of immune cell types and biological contexts, but we continue to seek to understand the means by which TGF-β1 secretion or expression occurs. Though some work has been done to understand how DCs are influenced by and respond to allogeneic antigen in vivo to promote an activating or regulatory cell response (22, 26, 58), questions remain in identifying the cell-intrinsic transcriptional reprogramming that occurs in cDC1s within this unique setting. Therefore, we performed single-cell RNA sequencing of DC-enriched splenocytes of WT B6 mice 48 hours after receiving BALB/c DST + CoB or saline infusion utilizing the 10x Genomics platform (Figure 6A). After quality control, filtering, normalization, and integration of the data, we employed uniform manifold approximation and projection (UMAP) dimensionality reduction analysis combined with unbiased cell type recognition utilizing the ImmGen open-source reference database (59) and SingleR algorithm to confirm cell cluster identity (Supplemental Figure 9). We then reclustered a subset of DC-identified clusters for further downstream analysis. A total of 2,432 individual cells were analyzed between the control and DST + CoB infusion conditions, which clustered into 7 distinct clusters expressing canonical markers for their respective cellular identity including Xcr1, Sirpa, Siglech, Ebf1, and Pax5 (Figure 6, B and C, and Supplemental Figure 10).

Conventional DC subsets display unique transcriptional signatures that are influenced by DST + CoB treatment. (A) Single-cell sequencing experimental scheme whereby B6 mice were treated with CD45.1+ BALB/c DST + CoB infusion (i.v.) on day 0 or saline control before collection of spleens after 48 hours, enrichment for DCs, and sequencing. (B) UMAP projections and identification of DC clusters, color-coded by cluster. (C) Violin plots of DC signature genes specifically expressed in their respective DC cluster. (D) Heatmap showing relative expression of marker genes across DC clusters. (E) Volcano plot of differentially expressed genes (DEGs) within Xcr1hi cDC1 cluster following DST + CoB treatment compared with control.

Two clusters of cDC1s were identified, which differed in their relative expression of Itgax and Xcr1. Given that expression of Xcr1 is both selective and specific to cDC1s, we hypothesized that cells with lower levels of Xcr1 expression were less mature and less likely to be participating in the immunoregulatory responses we observed. As such, we performed differential gene expression analysis of the Xcr1hi cDC1 cluster in control and DST + CoB infusion conditions (Figure 6D). We saw a significant increase in genes such as Pfn1 thought to be related to cell migration and motility, as well as those related to cell proliferation, like Btg2 (Figure 6E). We also noticed a number of genes related to antigen presentation (B2m) and metabolism (Lars2, Cox7c, Naaa), which were noticeably increased in Xcr1hi cDC1s exposed to DST + CoB in vivo (Figure 6E and Figure 7A). Therefore, we performed pathway analysis of the differentially expressed genes found within the Xcr1hi cDC1 cluster after DST + CoB by utilizing g:Profiler and Gene Ontology biological processes (60). We observed numerous processes we would expect for DCs exposed to antigen, including translation, antigen processing and presentation, and regulation of myeloid cell differentiation, in addition to numerous metabolic pathways (Figure 7B).

Mitochondrial respiration is increased in cDC1s after exposure to DST + CoB. (A) Violin plots of antigen presentation (B2m) and metabolism-related genes (Cox7c, Lars2, Tomm7) in Xcr1hi cDC1s in control and DST + CoB infusion conditions by single-cell sequencing. (B) Gene ontology analysis of differentially expressed genes in Xcr1hi cDC1s after DST + CoB compared with controls by single-cell sequencing. (C) Mitochondrial respiration of sorted splenic cDC1s 48 hours after DST + CoB treatment. n = 4 per group. A/P, antimycin A and piericidin A; 2-DG, 2-deoxy-D-glucose; OCR, oxygen consumption rate. (D) Quantification of basal OCR, max respiration, and respiratory capacity of naive and DST + CoB–treated cDC1s. n = 4 per group. *P < 0.05, **P < 0.01 by 2-tailed unpaired t test.

The canonical role of mitochondria and cellular metabolism for the generation of energy is well appreciated; however, in the more recent past, mitochondria are now lauded for their role as critical signaling organelles, especially as it relates to control of the innate and adaptive immune response (61). Studies have begun to investigate the role of metabolism in DCs during different biological states, but few experiments have been performed utilizing DCs isolated from in vivo tissue given low cell numbers and challenges with isolation and sorting protocols that may falsely activate or alternatively damage these fragile cells (62). Therefore, we developed a protocol utilizing a low-pressure, cartridge-based microfluidic cell sorter to isolate a highly pure population (>90%) of splenic cDC1s (Supplemental Figure 11). Using this protocol, we isolated cDC1s from spleens of B6 mice 48 hours after infusion of DST + CoB or saline and subjected them to metabolic analysis. As predicted by transcriptional sequencing, cDC1s exposed to DST + CoB displayed a higher oxygen consumption rate (OCR) compared with controls (Figure 7, C and D), which is a measure of mitochondrial oxidative phosphorylation. Likewise, cDC1s exposed to DST + CoB showed higher basal OCR and max respiration, as well as trends of increased respiratory capacity (Figure 7D). Importantly, no difference in mitochondrial mass of cDC1s was observed after DST + CoB treatment (Supplemental Figure 12), implying the observed results are indeed due to an increase in cDC1 mitochondrial activity.