Dimerization seems to be present in the E. Coli chemotaxis pathway. The receptor family that mediates chemotaxis in e. coli has the "same basic topology" as a bunch of eukaryotic TM receptors, consisting of an extrasensory domain connected to a cytoplasmic signalling domain; the connection consists of an -helical sequence 20 amino acids long. The aspartate receptor involved in chemotaxis, Tar, is a symmetric homodimer that binds to aspartate, and is of this topology type. Binding of ligand promotes receptor dimerization. In addition to the dimerization that occurs on the extracellular receptor domain, dimerization is necessary to the internal signalling domain.
There has been fusion of Tar to EnvZ such that aspartate no longer induces chemotaxis but instead osmolarity responses. Also, the same type of experiment was done with the insulin receptors's signaling domain and it has insulin-like tyrosine kinase activity in response to aspartate. Dimerization of Tar, however, may not be sufficient to cause the signalling cascade, but less is known (at least when the paper was written) about precisely what else is necessary. However, dimerization seems to be a pre-requisite to function.