Why no change in log-likelihood function when prior probabilities are same and no impact of changing mean vectors values as well?

What's the reason behind this specific behavior although we have some densities components(class conditional densities) too? But why it is specific to Prior Probabilities?

Thanks a lot