fixed effects are not in line with random effects in a GLMM

I have problems wrapping my head around some findings: I found a significant (negative) interaction between the random interactions A:B and A:C. However, one of these fixed effects is 0 on the fixed level. I have difficulties interpreting these findings with regard to my hypothesis that A affects B and C in an antagonistic manner, as I find that only on an individual level. But in my head I have problems bringing that together with the non-zero interaction of A:C. Could anybody help me here?