Va. Karginov et al., PROBING THE ROLE OF AN ACTIVE-SITE ASPARTIC-ACID IN DIHYDROFOLATE-REDUCTASE, Journal of the American Chemical Society, 119(35), 1997, pp. 8166-8176
Analogues of E. coli dihydrofolate reductase (DHFR) containing modifie
d amino acids at single, predetermined sites have been prepared. This
was accomplished by the use of the DHFR gene containing an engineered
nonsense codon (TAG) at the positions corresponding to Val-10 and Asp-
27. Misacylated suppressor tRNAs activated with the modified amino aci
ds of interest were employed for the suppression of the nonsense codon
s in a cell free protein biosynthesizing system, thereby permitting th
e elaboration of the desired protein analogues. In this fashion, the a
spartic acid analogues erythro-carboxyproline, cysteic acid, beta,beta
-dimethylaspartic acid, alpha-methylaspartic acid, erythro- and threo-
beta-methylaspartic acid, N-methylaspartic acid, and phosphonoalanine
were incorporated into one or both of the aformentioned positions. Alt
hough a number of these analogues were incorporated only in low yield,
a modification of the strategy has suggested how this might be improv
ed significantly. The derived proteins were purified and then characte
rized by their mobility on polyacrylamide gels in comparison with wild
-type DHFR. Representative DHFRs modified at position 10 were also deg
raded by defined proteolysis with Glu-C endoproteinase; the fragments
containing the modified amino acids were shown to have the same chroma
tographic properties on reverse phase HPLC as authentic synthetic stan
dards. individual analogues were assayed for their abilities to bind t
o the substrate analogue methotrexate and to convert dihydrofolate to
tetrahydrofolate. DHFR analogues containing erythro-and threo-beta-met
hylaspartic acid and beta,beta-dimethylaspartic acid were all shown to
mediate tetrahydrofolate production 74-86% as efficiently as wild-typ
e DHFR under conditions of multiple substrate turnover. Analysis of th
e rates of tetrahydrofolate production in the presence of NADPH and NA
DPD at two pH values suggests that this was due to rate-limiting hydri
de transfer from NADPH bound to DHFR analogues whose active site had b
een altered structurally.