Characterization of 4-hydroxyphenylpyruvate dioxygenases, inhibition by herbicides and engineering for herbicide tolerance in crops

4-Hydroxyphenylpyruvate dioxgenase (HPPD) enzymes from rat and from several plants contained only about a single inhibitor-binding active site per dimer which matched the content of iron in the purified Arabidopsis thaliana and Avena sativa enzymes. The dimeric HPPDs were about 10 fold more catalytically active than the tetrameric P. fluorescens enzyme with kcat/KmHPP values ranging from 0.8 to 2.5 s-1 μM-1. Most were also highly sensitive to herbicides with, for example, Ki values for mesotrione ranging from 25 to 100 pM. Curiously HPPDs from cool climate grasses were much less herbicide-sensitive. When likewise expressed in Nicotinia tabacum, Avena sativa HPPD, Ki value of 11 nM for mesotrione, conferred far greater tolerance to mesotrione (CallistoTM) than did any of the more sensitive HPPDs. Targeted mutagenesis of the Avena HPPD led to the discovery of 4 mutations imparting improved inherent tolerance, defined as the ratio of Ki to KmHPP, by about 16 fold without any loss of catalytic activity. The Nicotinia line with the highest expression of this quadruple mutant exhibited substantial resistance even up to a 3 kg/ha post-emergence application of mesotrione. The maximum observed expression level of heterologous plant HPPDs in tobacco was ca. 0.35% of the total soluble protein whereas the endogenous tobacco HPPD constituted only ca. 0.00075%. At such high expression even HPPDs with impaired catalytic activity could be effective. A quintuple mutant Avena sativa HPPD conferred substantial tolerance across a broad range of HPPD herbicide chemistries despite being only ca. 5 % as catalytically active as the wild type enzyme. Testing various wild type and mutant HPPDs in tobacco revealed that tolerance to field rates of herbicide generally requires about two order of magnitude increases in both inherent herbicide tolerance and expression relative to endogenous levels. This double hurdle may explain why target-site based resistance to HPPD-inhibiting herbicides has been slow to evolve in weeds.