Glycolipid-activated NKT cells support the induction of persistent plasma cell responses and antibody titers

Abstract

NKT cell activation with CD1d-binding glycolipid alpha-galactosylceramide (alpha-GC) enhances antibody responses to co-administered T-dependent antigen. The efficacy of alpha-GC relative to other CD1d-binding glycolipids and adjuvants is not known. There is little information on how NKT cells affect antibody production beyond initial booster-stimulated recall responses. We therefore tested the hypothesis that alpha-GC stimulates induction of plasma cells and antibody responses as effectively as Th1- and Th2-skewing variants of alpha-GC and several other adjuvants. C57BL/6 and CD1d-/- mice were immunized with nitrophenol-conjugated keyhole limpet hemocyanin (NP-KLH) plus alpha-GC or NP-KLH plus adjuvants before administration of an NP-KLH booster and assessing antibody responses and plasma cell frequency. alpha-GC boosted long-term antibody responses as efficiently as all other agents tested and induced plasma cells that were detected in bone marrow 13 weeks after immunization. We then determined whether NKT cells were required in the presence of other adjuvants. CD1d-/- mice had a reduced induction of plasma cells in response to NP-KLH/Alum as compared to C57BL/6 mice. However, NKT cells were not required for the continued presence of those cells that were induced. Although NKT cells are capable of inducing persistent plasma cell responses, they may not play a major role in supporting longevity post-induction.

Figure 1

NKT activation enhances primary and recall Ab responses. (a) C57BL/6 and CD1d^−/− mice were untreated (naive) or immunized s.c. with 10 µg NP-KLH, NP-KLH adsorbed to Alum or NP-KLH plus 4 µg α-GC. Mice were bled on day 14. (b) Mice were boosted on day 28 with 10 µg NP-KLH and bled on day 33. (c) In a separate experiment, C57BL/6 mice were immunized with NP-KLH or NP-KLH plus α-GC and blood samples drawn at times indicated (left panel). Mice then received a booster (NP-KLH) and blood samples were drawn at times indicated (right panel). (d) C57BL/6 mice were immunized with NP-KLH or NP-KLH plus α-GC and bled on days indicated before receiving an NP-KLH booster on day 56. Blood samples were collected on days 61 and 66 and anti-NP IgG1 titers assessed by ELISA. For (a–d) ELISA were performed to determine the endpoint anti-NP IgG1 Ab titer. Data in (a–d) show the geometric mean titers ±95% confidence intervals for five mice per group. Significantly different titers from 'NP-KLH only' controls are shown in asterisks. (e) Ratios for Ab binding to NP-(3)-BSA versus NP-(30)-BSA-coated plates were assessed. Data show mean±SD ratio for five mice per group. Data are representative of two independent experiments. Significantly different titers from NP-KLH only controls are shown by asterisks.

Figure 2

Dose-dependent effects of α-GC in vivo. (a) C57BL/6 mice were immunized s.c. with PBS (naive) or α-GC. After the time indicated, spleens were harvested and analyzed for CD1d tetramer/TCRβ-positive cells, indicated by the box R2. (b) Graph shows effects of dosage and time on the splenic NKT population. Similar results were observed for lymph nodes (not shown) and are representative of at least three independent experiments. (c) Spleen and lymph node were harvested 16 h after immunization and NKT cells enumerated. (d) Splenocytes were cultured with or without α-GC at a 100 ng/mL final concentration. After 20 h, supernatants were collected and concentrations of IL-4 and IFN-γ measured by Bio-Plex assay.

Figure 3

Th1-skewing α-GC enhances short-term Ab recall responses. (a) C57BL/6 mice were bled (naive), then immunized s.c. with 10 µg NP-KLH alone, adsorbed to Alum or mixed with the α-GC molecules indicated. Dosages are detailed in the Materials and methods. Mice were bled on day 14 and ELISA performed to determine the endpoint anti-NP IgG1 Ab titer. (b) Mice were boosted on day 28 with 10 µg NP-KLH and bled on day 33. Data show the geometric mean titer ±95% confidence intervals for five mice per group [with exception of Alum and α-GC in (a) where n=3 and 4, respectively]. Asterisks indicate titers significantly different from NP-KLH only control.

Figure 4

Lack of co-operation between α-GC and TLR ligands in Ab responses. (a) C57BL/6mice were bled (naive), then immunized s.c. with 10 µg NP-KLH alone, or mixed with 0.5 µg α-GC, 6.25 µg poly I:C, 6.25 µg Imiquimod, or a combination of α-GC and TLR agonist. Mice were bled at day 14 and ELISA were performed to determine the endpoint anti-NP IgG1 Ab titer. (b) Mice were boosted on day 28 with 10 µg NP-KLH, and then bled on day 33. Data in (a) and (b) show the geometric mean titer ±95% confidence intervals for five mice per group (n=3 for naive group). Asterisks indicate titers significantly different from NP-KLH only control.

Figure 5

NKT cells are required for induction of PC (antibody secreting cells, ASC) but not persistent responses. C57BL/6 and CD1d−/− mice were immunized as indicated on day 0, and then boosted with NP-KLH on week 4 (day 28). On week 5 (day 35) (a) and week 9 (day 63) (b), mice were killed and splenocytes and bone marrow cells assessed by ELISPOT assay. Data show mean ± SD frequency of NP-specific PC for three mice per group, each sample being performed in triplicate. Data are representative of three separate experiments. In (b), the endpoint anti-NP IgG1 Ab titer is also shown. Asterisks indicate titers significantly different from NP-KLH only control in (a) or between C57BL/6 and CD1d^−/− mice in (b). In (c), C57BL/6 and CD1d−/− mice (five per group) were immunized as indicated on day 0, and then boosted with NP-KLH on week 4 (day 28), and bone marrow PC were enumerated on week 5 (day 35) and week 13 (day 91).

Figure 6

NKT activation leads to persistent PC (antibody secreting cells, ASC) responses. C57BL/6 mice were immunized as indicated on day 0 and then boosted with NP-KLH on week 4 (day 28). On week 13 (day 91), mice were killed and splenocytes (a) and bone marrow cells (b) assessed by ELISPOT assay. Data in (a) and (b) show mean ± SD frequency of NP-specific PC for three mice per group, each sample being performed in triplicate. (c) Sera were obtained 13 weeks post immunization and anti-NP IgG1 endpoint titers measured by ELISA. Data show geometric mean titers ±95% confidence intervals for three mice per group.