Enhanced catalytic efficiency of aminoglycoside phosphotransferase (3')-IIa achieved through protein fragmentation and reassembly.

Many monomeric proteins can be split into two fragments, yet the two fragments can associate to make an active heterodimer. However, for most locations in a protein such a conversion is not feasible, presumably due to inefficient assembly or improper folding of the fragments. For some locations, this can be overcome by fusion of the fragments to dimerization domains that facilitate correct assembly. A variety of heterodimers of aminoglycoside phosphotransferase (3')-IIa (Neo) were created in which the Neo fragments required fusion to a pair of leucine zippers for activity in vivo. However, the ability of these heterodimers to confer kanamycin resistance to Escherichia coli cells was impaired compared to wild-type Neo, primarily due to poor production of soluble protein. The mutations R177S and V198E restored the kanamycin resistance to wild-type levels while maintaining the dependence on leucine zippers for activity. These mutations restored high levels of kanamycin resistance not through an improvement in the production of soluble protein but rather by conferring a large improvement in k(cat)/K(m), surpassing that of Neo. Furthermore, whereas R177S and V198E served to improve k(cat)/K(m) 60-fold in the context of the heterodimer, the same mutations in the context of wild-type Neo had a ninefold negative effect on k(cat)/K(m). This demonstrates the possibility that enzymes with improved catalytic properties can be created through a process involving fragmentation and fusion to domains that facilitate assembly of the fragments.