[R-sig-ME] Autocorrelated and Heteroscedastic Random-Slopes

[Sun, John]

Dear All,

I am writing to ask how to obtain autocorrelated and heteroscedastic random slopes. I know the single pipe | makes the random-slopes cross-correlated with the random-intercept. 
Are the random-slopes already autocorrelated and heteroscedastic by default in common uses of lme4-nlme?
I also want to try making the random-slopes have a Toeplitz, Unstructured or AR-1 structure with itself.

I am not talking about the correlation between the random-slopes and intercepts.
I am talking about the covariance between random-slopes with random-slopes with different periods-itself. 

Best regards,
Kpjm