The physical mechanisms that produce the quenching of star formation and the 
build-up of bulges in elliptical, passive galaxies are highly debated. 
It is likely that not all passive galaxies have followed the same path and 
that different mechanisms are efficient in galaxies with different physical 
properties (e.g. mass) and in different environments.

In this talk I will report on the analysis of Sinfoni@ESO Adaptive Optics and 
J and H data with WFC3@HST on a sample of star-forming data at z~2. 
The combination of these data allows us to derive radial profiles for both 
recent star formation and mass with ~kpc resolution. While SFR density 
profiles are very similar at all masses, following a Sersic profile with n~1, 
the stellas mass profiles are always more concentrated and their concentration 
increases with increasing stellar mass.

In particular, the mass profiles of the most massive galaxies in our sample 
are similar to those of local early type galaxies, showing that the bulge 
component of these massive star-forming galaxies is already fully in place, 
with low sSFR, while they are still vigorously forming stars in the external 
regions. With a simple toy model we estimate that the quenching of star 
formation in the external regions will follow with a delay with a few Gyr, 
producing an age and color gradient consistent with what is seen in data for 
passive galaxies.

These data suggest that, at least for these massive galaxies, the quenching 
mechanism is more likely to be internal (AGN? local stellar density?) 
rather than external (large scale shut-down of gas supply or a low 
cosmological accretion rate).