Intraspecific genetic variation can affect community structure and ecosystem processes (Bolnick et al. 2011). It can also influence phenotypic expression by genotypes within other species to produce genotype-by-genotype (G × G) interaction (Falconer & Mackay 1996). Evolution of one species drives correlated evolution of others when it causes G × G for fitness (Thompson 2005). However, the mechanisms by which species interact also influence evolutionary outcomes (Kummel & Salant 2006; Golubski & Klausmeier 2010; Akçay & Simms 2011; Grman et al. 2012). To identify genes and putative functional mechanisms underlying G × G interactions, Heath et al. (2012) analysed natural variation in the symbiotic transcriptome of the mutualistic nutritional symbiosis between a legume host Medicago truncatula and the facultative endosymbiotic rhizobium Sinorhizobium meliloti. Using twelve microarrays, the authors simultaneously measured host and symbiont gene expression in root nodules from four factorial pairings of host and symbiont genotypes that produced G × G in host fitness (Fig. 1, upper panel). Rhizobium gene expression was influenced by rhizobium and plant genotype and the G × G interaction (Fig. 1, lower panel), whereas plant gene expression was influenced primarily by plant genotype. The authors identified rhizobium genes that might contribute to G × G in host plant fitness. Heath et al. (2012) have moved beyond the constraints of single organism analysis towards a more realistic understanding of plants and bacteria as organisms inextricably linked with symbioses that affect even basic patterns of gene expression.