Mechanisms of ganglioside inhibition of APC function

Abstract

Gangliosides shed by tumor cells exert potent inhibitory effects on cellular immune responses. Here we have studied ganglioside inhibition of APC function. When human monocytes were preincubated in 50 micro M highly purified ganglioside G(D1a), pulsed with tetanus toxoid (TT), and washed, the expected Ag-induced proliferative response of autologous normal T cells added to these monocytes was inhibited by 81%. Strikingly, there was also almost complete (92%) and selective inhibition of the up-regulation of the monocyte costimulatory molecule CD80, while I-CAM-1, LFA-3, HLA-DR, and CD86 expression were unaffected. Purified LPS-stimulated monocytes that had been preincubated in G(D1a) likewise showed inhibition of CD80 up-regulation (59%) as well as down-regulation of CD40 (54%) and impaired release of IL-12 and TNF-alpha (reduced by 59 and 51%). G(D1a)-preincubated human dendritic cells (DC) were also affected. They had reduced constitutive expression of CD40 (33%) and CD80 (61%), but not CD86, and marked inhibition of release of IL-6 (72%), IL-12 (70%), and TNF-alpha (46%). Even when pulsed with TT, these ganglioside-preincubated DC remained deficient in costimulatory molecule expression and cytokine secretion and were unable to induce a normal T cell proliferative response to TT. Finally, significant inhibition of nuclear localization of NF-kappaB proteins in activated DC suggests that disruption of NF-kappaB activation may be one mechanism contributing to ganglioside interference with APC expression of costimulatory molecules and cytokine secretion, which, in turn, may diminish antitumor immune responses.

Figure 1

Expression of costimulatory molecules CD80 and CD86 by G_D1a-preincubated Ag-pulsed monocytes. Monocytes were incubated with 50 µM G_D1a for 48 h and with TT during the last 24 of the 48 h. The monocytes were washed, and autologous T cells were added. After a further 48-h incubation, the cells were harvested. CD80 and CD86 expression on monocytes was determined by gating CD14⁺ cells. The histogram represents 50,000 cells stained with the indicated mAb (solid lines) or isotype-matched control (dotted lines). The percentage of positively stained cells is indicated on the CD80 histograms. The bar graph represents the level of CD80 and CD86 expression (mean ± SEM of five separate donors) on G_D1a-exposed, Ag-pulsed monocytes relative to Ag-pulsed control monocytes (=100%).

Figure 2

Kinetics of ganglioside inhibition of monocyte CD80 up-regulation. A, PBMC were incubated with TT, and 50 µM G_D1a was added to the culture at various times (0, 24, or 44 h) after TT addition. At the end of the 44 h, CD80 was quantified by FACS on cells gated for CD14. B, Time course of CD80 expression. Monocytes were incubated with 50 µM G_D1a for 48 h and with TT during the last 24 of the 48 h, washed, and cultured with autologous T cells for up to 72 h. The cells were gated on CD14⁺ and 7-amino actinomycin D⁻ for analysis of CD80 expression after 24, 48, and 72 h of coculture with T cells.

Figure 3

Expression of the costimulatory molecules CD40, CD80, and CD86 by G_D1a-preincubated, LPS-stimulated monocytes. Monocytes were incubated with 50 µM G_D1a for 48 h, washed, stimulated with 5ng/ml LPS for 24 h, and recovered. CD40, CD80, and CD86 expression on monocytes was determined by gating CD14⁺ cells. The histogram represents 50,000 cells stained with the indicated mAb (red lines) or isotype-matched control (black lines). The median fluorescence intensity is indicated in the upper right corner of each histogram. The bar graph depicts the median fluorescence intensity of CD40, CD80, and CD86 expression on monocytes exposed to G_D1a and LPS as a percentage of that of the control cells (monocytes and LPS). Results are the mean ± SEM of three different donors.

Figure 4

Expression of costimulatory molecules CD40, CD80, and CD86 on G_D1a-preincubated DC. DC were incubated with 50 µM G_D1a for 48 h and with TT during the last 24 h of G_D1a preincubation, recovered, and stained to determine CD40, CD80, and CD86 expression. The histogram represents 50,000 cells stained with indicated mAb (red lines) or isotype-matched control (black lines). The median fluorescence intensity is indicated in the upper right corner of each histogram. In the bar graph: ■, median fluorescence intensity of CD40, CD80, and CD86 expression on G_D1a-preincubated, TT-pulsed DC of three to five separate donors, expressed as the mean ± SEM percentage of expression by control TT-pulsed DC; , results obtained by incubating G_D1a-treated and control DC as described above and then conincubating them for an additional 24 h with autologous T cells before determining costimulatory molecule expression.

Figure 5

Effect of G_D1a preincubation on LPS-stimulated cytokine production by monocytes. Monocytes were incubated with various concentrations (0–50 µM) of G_D1a for 48 h, washed, and stimulated with 5 ng/ml of LPS for 24 h. Then the culture supernatants were collected, and cytokine production was measured by ELISA (BioSource). The data are the mean ± SEM concentrations in samples from three different donors.

Figure 6

Effect of G_D1a preincubation on cytokine production by TT-stimulated DC. DC were incubated with 50 µM G_D1a for 48 h and with or without TT for the last 24 h of the 48-h G_D1a incubation period. Then the supernatant cytokine concentrations were measured by ELISA. Data represent the mean ± SEM OD of undiluted samples from six donors for IL-6 and TNF-α and three donors for IL-12.

Figure 7

Nuclear translocation of NF-κB proteins in G_D1a-preincubated DC. DC were incubated with 50 µM G_D1a for 48 h and with or without TT for the last 24 h of the 72-h G_D1a incubation period. DC were then recovered and lysed, and protein was extracted from the nuclear pellets. Equal amounts of protein from the DC nuclear extracts were analyzed for NF-κB proteins by Western blot.