Functional divergence among CD103+ dendritic cell subpopulations following pulmonary poxvirus infection

Abstract

A large number of dendritic cell (DC) subsets have now been identified based on the expression of a distinct array of surface markers as well as differences in functional capabilities. More recently, the concept of unique subsets has been extended to the lung, although the functional capabilities of these subsets are only beginning to be explored. Of particular interest are respiratory DCs that express CD103. These cells line the airway and act as sentinels for pathogens that enter the lung, migrating to the draining lymph node, where they add to the already complex array of DC subsets present at this site. Here we assessed the contributions of these individual populations to the generation of a CD8(+) T-cell response following respiratory infection with poxvirus. We found that CD103(+) DCs were the most effective antigen-presenting cells (APC) for naive CD8(+) T-cell activation. Surprisingly, we found no evidence that lymph node-resident or parenchymal DCs could prime virus-specific cells. The increased efficacy of CD103(+) DCs was associated with the increased presence of viral antigen as well as high levels of maturation markers. Within the CD103(+) DCs, we observed a population that expressed CD8alpha. Interestingly, cells bearing CD8alpha were less competent for T-cell activation than their CD8alpha(-) counterparts. These data show that lung-migrating CD103(+) DCs are the major contributors to CD8(+) T-cell activation following poxvirus infection. However, the functional capabilities of cells within this population differ with the expression of CD8, suggesting that CD103(+) cells may be divided further into distinct subsets.

FIG. 1.

Dendritic cells increase in the lung draining MLN following VV infection. C57BL/6 mice were intranasally infected with 10⁷ PFU of VV.NP-S-eGFP. On days 1 to 4 postinfection, MLN were isolated and CD11c⁺, CD90.2⁻, CD49b⁻, and CD19⁻ populations analyzed for expression of CD103, CD11b, CD8, and F4/80. The total number of CD11c⁺ cells (A) and the number present within each DC subset as well as the number of macrophages (B) were calculated based on the total cells recovered. eGFP expression in the populations was analyzed in both the lung (C) and the MLN (D) and is graphed as a percentage of each APC type expressing eGFP. (E) Mice were infected, and 5 h later, CTO was administered intratracheally. Cells were pregated as CD11c⁺, CD90.2⁻, CD49b⁻, or CD19⁻, and subsequently CTO⁺ CD11b⁺ or CD103⁺ DC were analyzed for eGFP⁺ cells on day 2 postinfection. Data in panels A to D reflect the averages from 4 independent experiments. In these experiments, to be considered valid for analysis, the number of eGFP⁺ events in each population had to be 5-fold greater than that observed in mock-infected mice. Significant eGFP⁺ events among the different populations in the lung for individual mice ranged from 19 to 205 for day 1, from 17 to 588 for day 2, from 10 to 598 for day 3, and from 14 to 747 for day 4. The variation in cell number was the result of differences in the sizes of the different APC populations. For the MLN, significant eGFP⁺ events were observed only for CD103⁺ cells. For individual mice, these ranged from 9 to 29 on day 1, from 14 to 32 on day 2, from 16 to 24 on day 3, and from 13 to 39 on day 4. The data in panel E reflect 3 independent experiments, each utilizing between 23 and 25 pooled MLN for each condition. Significance for panels A to D was determined by 2-way analysis of variance (ANOVA) with a Bonferroni posttest comparing values for subsets to values for mock infection. For panel E, Student's t test was used to compare levels of eGFP expression between control and day 2 results within each subset. Error bars represent the standard errors of the means (SEM). *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.005; ns, not significant.

FIG. 2.

Airway-derived CD103⁺ DC are superior to parenchymal DC for priming naive CD8⁺ T cells ex vivo. Mice were intranasally infected with 10⁷ PFU of either VV.NP-S-eGFP or the control virus VV.P. Five hours following infection, mice were given 1 mM Cell Tracker Orange i.t. Two days postinfection, mice were sacrificed and MLN harvested. Recovered cells were gated as CD11c⁺, CD90.2⁻, CD49b⁻, and CD19⁻ and were sorted based on their expression of CTO, CD103, and CD11b (A). Sorted cells were then incubated with CFSE-labeled naive OT-I T cells for 3 days at a ratio of 1:5 DC-OT-I cells. OT-I cells were restimulated for 5 h with 10⁻⁶ M Ova peptide. Cells were analyzed to determine proliferation and IFN-γ production (representative data are shown in panel B and averaged data in panels C and D). The percent divided was calculated using FlowJo software. MLN from 23 to 25 animals were pooled for each sort. Error bars represent the SEM from 2 individual experiments. Significance was determined using Student's t test to compare results from mock infection and day 2. *, P ≤ 0.05; **, P ≤ 0.01. FSC, forward scatter; SSC, side scatter.

FIG. 3.

eGFP⁺ CD103⁺ DC are highly enriched for mature cells. Mice were intranasally infected with 10⁷ PFU of VV.NP-S-eGFP or PBS as a control. On days 1 to 3 postinfection, MLN from animals were assessed for the maturation of CD103⁺ DC. eGFP⁺ and eGFP⁻ cells within the CD11c⁺, CD103⁺, CD90.2⁻, CD49b⁻, and CD19⁻ populations were analyzed for CD86 and CD80 expression. Representative data are shown in panel A. Isotype staining was performed using pooled cells from infected mice. The percentages of cells that were positive for CD80 (B) or CD86 (D) as well as the intensity of staining for CD80 (C) or CD86 (E) within the positive population are shown. Error bars represent the SEM from 4 or 5 independent experiments, each containing 2 to 5 animals per time point. For each graph, significance was determined using 2-way ANOVA with a Bonferroni posttest. For panels B and D, the percentages of eGFP⁺ versus eGFP⁻ cells for each time point were compared. For panels C and E, significance determination was performed by comparing the value at each time point to the value for mock infection as well as comparing eGFP⁺ and eGFP⁻ percentages, as indicated by the brackets. *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.005; ns, not significant. For all data points, the following minimum numbers of eGFP⁺ events were analyzed: on day 1, 18 to 41; on day 2, 239 to 382; on day 3, 64 to 189. In addition, to be considered valid for analysis, the number of eGFP⁺ events had to be a minimum of 5-fold above that for the mock-infected samples, which ranged from 1 to 5. MFI, mean fluorescence intensity.

FIG. 4.

A subset of CD103⁺ cells expressing CD8α⁺ is present in the MLN. MLN from mock-treated or infected (10⁷ PFU of VV.NP-S-eGFP) animals were isolated on the indicated days. CD11c⁺, CD90.2⁻, CD49b⁻, and CD19⁻ MLN cells were analyzed for the expression of CD8α and CD103. Representative data showing the gating strategy and expression of CD103 and CD8α are given in panel A. Averaged data from 4 independent experiments, each containing MLN cells pooled from 1 to 3 animals per time point, are shown in panel B. Error bars represent the SEM.

FIG. 5.

Functional divergence between CD8α⁺ CD103⁺ and CD8α⁻ CD103⁺ DC. Mice were infected intranasally with either VV.NP-S-eGFP or VV.M (10⁷ PFU). On day 2 postinfection, MLN cells were isolated and pooled, and CD11c⁺ cells were enriched by column purification. The enriched population was sorted into subsets based on CD11c⁺, CD90.2⁻, CD49b⁻, and CD19⁻ staining together with expression of CD8α and CD103. Sorted cells were incubated for 3 days with CFSE-labeled naive OT-I T cells at a ratio of 1:4 DC-OT-I cells. Following culture, OT-I cells were identified by staining with CD90.2 and analyzed for CFSE expression. A representative experiment is shown in panel A, and average data from three independent experiments are shown in panel B. Between 22 and 25 mice were used for each group for each experiment. Error bars represent the SEM. Significance was determined using Student's t test. *, P ≤ 0.05; **, P ≤ 0.01; ns, not significant.