The body tides response (deformation and gravity) of the Earth is generally computed assuming radial symmetry in stratified earth models, at the hydrostatic equilibrium. We present in this paper numerical experiments with the aim to evaluate the impact of very large mantle heterogeneities of density on body tides. In a companion paper, we have developed a new earth elasto-gravitational deformation model able to take into account the heterogeneous structure of the mantle. We use this model to calculate the theoretical perturbation induced by three types of spherical heterogeneities in the mantle on M2 body tides response. The heterogeneities are: (1) our limit case, a heterogeneity of 1000 km of radius in the lower mantle; (2) a heterogeneity of 500 km of radius at the bottom of the lower mantle and (3) a heterogeneity of 285 km of radius in the upper mantle. The density variation has been set to -50 kg m-3. For each experiment, we first calculate the equilibrium state of the Earth when it contains a heterogeneity, including non-hydrostatic pre-stresses, dynamical topography and lateral variation of density. Then we compute the M2 tidal perturbation. We find that the surface tidal displacement perturbation is smaller than 1 mm, and that the gravity perturbation has a maximum amplitude of 525 nanoGal (nGal). Regarding to the present precision in position measurement, the displacement is too small to be detected. The gravity perturbation should be measurable with superconducting gravimeters, which have a nGal instrumental precision. In experiment 2, the maximum gravity perturbation is about 120 nGal, and in experiment 3, the maximum perturbation is about 33 nGal. Finally, we investigate the maximum theoretical impact of the Pacific and the African superplumes on the M2 body tide. The superplumes have been modelled as two spherical heterogeneities with a radius of 1000 km in the lower mantle. We find that these superplumes induce a maximum perturbation in gravity of about 370 nGal with a large part corresponding to a mean variation of gravity. We conclude that we can expect to have a gravity perturbation of body tide with an order of magnitude of about hundred of nGal induced by the biggest mantle heterogeneities of density. This perturbation in gravity should be measurable with superconducting gravimeters if all other contributions in the signal could be extracted with a sufficient precision.