using function instead of formula in plm

Hi all,

I am trying to use plm to estimate coefficients in a model consisting of a system of equations. So far I used mle2 from the package "bbmle", but now I need to test for autocorrelation and mle2 does not provide for the necessary tests. mle2 needs a function as input that might as well consist of many different equations. plm however requires an object of class formula that needs to be included in the plm-call. Does anyone know if there is a way to make R understand that it should do something like

result <- plm(function,data = paneldata,...) 

and what the object "function" then has to look like?

Thanks in advance,
Erich
--