Probable molecular and population-genetic mechanisms of microevolution
of the genetic systems controlling interactions of root nodule bacter
ia with legumes, as well as the main pathways of their macroevolution,
are reviewed. It is suggested that synchronization of the rates of ba
cterial and plant genome evolution was the condition necessary for coe
volution of root nodule bacteria and legumes. This could have been ach
ieved via an increase in recombination activity of the bacterial genom
e during the formation of the virulence gene system. The latter proces
s was associated with changes in the pattern of certain genes controll
ing different metabolic functions in nonsymbiotic nitrogen-fixing orga
nisms (origin of Bradyrhizobium and Azorhizobium, the ''primary'' root
nodule bacteria capable of nitrogen fixation ex planta). Increased ge
netic instability could have been to the transfer of main symbiotic ge
nes into plasmids. This, in turn, resulted in the development of a com
plex genetic population structure (origin of ''nonsymbiotic'' subpopul
ations providing for a high frequency of horizontal transfer of symbio
tic genes) and made possible the development of new rhizobia forms via
symbiotic gene transfer to different soil bacteria (origin of Rhizobi
um, the ''secondary'' root nodule bacteria incapable of nitrogen fixat
ion ex planta). Comparing the patterns of genes controlling interactio
ns with plants in Rhizobium and Agrobacterium showed that symbiotic an
d parasitic traits of these related microorganisms developed independe
ntly, although similar mechanisms could be responsible for evolution o
f nodulation in the rhizobia-legume system and for evolution of phytop
arasitic systems. It is suggested that both Rhizobium and Agrobacteriu
m originated from saprophytic soil microorganisms capable of synthesiz
ing certain cell wall molecules (polysaccharides, glucans, etc.) allow
ing them to persist in tissues of higher plants.