I will present results from The X-Shooter Lens Survey (XLENS). With XLENS we are unambiguously separate the stellar from the dark-matter content in the internal region of lens early-type galaxies (ETGs) to understand their interplay and to probe directly their formation and dynamical evolution. We combine precise strong gravitational lensing and dynamical constraints on the mass distribution with high signal-to-noise spectroscopy in the entire rest-frame visible to the NIR.In this talk I will present results obtained on a sample of very massive lens ETGs from the SLACS Survey, with velocity dispersions greater than 250 km/s and redshift>0.1.

First I will show how to constrain the low mass end of the Initial Mass Function (IMF)directly from galaxy optical spectra using a new set of non-degenerate optical spectroscopic indices which are strong in cool giants and dwarfs and almost absent in main sequence stars (Spiniello et al., 2014a). I will present unambiguous evidence that the low-mass end of the IMF is not universal. Then, I will demonstrate that the combination of this SSP modelling with a fully self-consistent joint lensing+dynamics analysis (Barnabè et al. 2012) allows us to disentangle IMF slope variations from internal dark-matter variations and, for the first time ever, to contemporary put constrains on the IMF cutoff mass (Barnabè et al., 2013, Spiniello et al., in prep).