Phosphorylated MAPK monomers can dimerize - generally into MAPK1 and MAPK3 homodimers, as the heterodimer is unstable- but the physiological significance of dimerization is unclear (Khokhlatchev et al, 1998; reviewed Rosokoski, 2012b). MAPKs have both cytosolic and nuclear targets and dimerization may be particularly important for MAPK-dependent phosphorylation of cytosolic targets. Phosphorylation of cytosolic MAPK targets appears to happen predominantly in the context of larger scaffolding complexes, and since the scaffolds and cytosolic MAPK substrates contact the same hydrophobic surface of MAPK, dimerization is necessary to allow assembly of a functional complex (Casar et al, 2008; Lidke et al, 2010; reviewed in Casar et al, 2009). Consistent with this, disrupting either MAPK dimerization or the MAPK interaction with the scaffolding protein abrogated proliferation and transformation (Casar et al, 2008). Note that, for simplicity in this diagram, dimerization is shown as happening between free cytosolic monomers of activated MAPK rather than in the context of the scaffolding complex.
Although predominantly cytoplasmic in resting cells, a proportion of activated MAPK translocates to the nucleus upon stimulation where it activates nuclear targets. Despite early studies to the suggesting that dimerization was required for nuclear translocation, a few recent papers have challenged this notion (Lenormand et al, 1993; Chen et al, 1992; Khokhlatchev et al, 1998; Casar et al, 2008; Lidke et al, 2010; Burack and Shaw, 2005; reviewed in Roskoski, 2012b).