In this contribution we propose an innovative scheme for distributed coding of two correlated variables X and Y. The scheme comprises three steps: 1) processing of source outcomes via a generalized modulo-reduction folding operation. In a symmetric scheme, this reduction is operated on vectors drawn both from X and Y by independent coders, producing vectors of variables (X,Y); 2) independent lossy coding of the quantized reduced variable vectors (X,Y); 3) optimal decoding of the pair (X,Y) from quantized (X,Y). A block diagram of the proposed coder is shown in Fig. 1.a. The block marked " mod ? " operates a reduction modulo an isometry group ? which generalizes the lattice modular reduction used in Nested Lattice Codes.