CD40L represents a strong endogenous danger signal that induces pro-inflammatory activation of CD40-expressing cells such as dendritic cells (DC), monocytes, and B cells. However, since CD40 activation alone is insufficient to induce pro-inflammatory cytokines such as IL-12p70, we studied whether CD40-mediated pro-inflammatory activity might be dependent on co-signalling pathways involving JAK/STAT. Using quantitative Western blotting, we demonstrate that JAK/STAT signalling is induced by cytokines such as IL-4, GM-CSF and IFNg, whereas CD40 activation mediates NFkB signalling. CD40L-induced IL-12p70 and IL-10 secretion in human DC, monocytes, B cells, and chronic lymphocytic leukemia (CLL) cells was measured upon complementary JAK/STAT activation by IL-4, GM-CSF and IFNg in the presence and absence of specific inhibitors of JAK2, JAK3, and pan-JAK. Whereas IL-12p70 could not be induced by CD40 ligation or by cytokines alone, IL-12p70 secretion and suppression of IL-10 was reproducibly observed after co-stimulation of CD40L with IL-4, GM-CSF, or IFNg. This effect could be completely reversed by pan-JAK inhibition. Persistence of IL-4/GM-CSF/IFNg-mediated JAK/STAT signalling as late as 12 hours following cellular activation via CD40 was required for IL-12p70 secretion as shown by the effects of delayed JAK inhibition. Similarly, persistence between 12 and 24 hours of IL-12p35 and p40 mRNA expression correlated best with the level of IL-12p70 secretion. Specific inhibition of JAK2 and JAK3 further revealed a context-dependent action of the distinct JAK family members: JAK2 showed a strong co-dominant effect in the setting of IL-4-induced JAK/STAT activity. Both, JAK2 and JAK3 were required for IL-12p70 secretion, whereas JAK2 alone was sufficient to modulate IL-10 secretion. However, in the context of IFNg-induced JAK/STAT signalling in DC, neither JAK2 nor JAK3 inhibition had effects on IL-12p70. Here, only inhibition by the pan-JAK inhibitor involving JAK1 abrogated IL-12p70 secretion, indicating that in IFNg-dependent signalling, JAK2 is apparently sub-dominant to JAK1 and had only a small enhancing effect on IL-10. This context dependence markedly differed in myeloid cells and B cells, as normal and malignant (CLL) B cells maintain a co-dominant JAK2 activity in the context of IFNg-induced JAK/STAT-signalling. In conclusion, complementary JAK/STAT signalling is required for the pro-inflammatory effects of CD40 ligation in humans, with different JAK subset predominance in myeloid and B cells. These results may open new ways of lineage-specific interfering with CD40 signals by modulating JAK/STAT activity using tyrosine kinase inhibitors.

Author notes

Disclosure: No relevant conflicts of interest to declare.