A point mutation produced a class 3 aldehyde dehydrogenase with increased protective ability against the killing effect of cyclophosphamide

Abstract

Cyclophosphamides are pro-drugs whose killing agent is produced from an aldehyde that is formed by the action of a P450 oxidation step. The mustard from the aldehyde can destroy bone marrow cells as well as the tumor. Aldehyde dehydrogenase (EC 1.2.1.3) can oxidize the aldehyde and hence inactivate the cytotoxic intermediate but bone marrow has little, if any, of the enzyme. Others have shown that over-expression of the enzyme can afford protection of the marrow. A T186S mutant of the human stomach enzyme (ALDH3) that we developed has increased activity against the aldehyde compared to the native enzyme and HeLa cells transformed with the point mutant are better protected against the killing effect of the drug. It took threefold more drug to kill 90% of the cells transformed with the mutant compared to the native enzyme (15.8 compared to 5.1mM of a precursor of the toxic aldehyde). Analysis of molecular models makes it appear that removing the methyl group of threonine in the T186S mutant allows the bulky aldehyde to bind better. The mutant was found to be a poorer enzyme when small substrates such as benzaldehyde derivatives were investigated. Thus, the enzyme appears to be better only with large substrates such as the one produced by cyclophosphamide.

Figure 1

Pathway for the conversion of cyclophosphamide to aldophosphamide, the toxic mustard. Mafosfamide (upper right) will spontaneously form aldophosphamide which is the actual substrate for aldehyde dehydrogenase. Cyclophosphamide is enzymatically converted to aldophosphamide.

Figure 2

Molecular models of the native and the threonine to serine mutant form of aldehyde dehydrogenases. The methyl group of the threonine appears to interact favorably with benzaldehyde but could interfere with the binding of aldophosphamide with ALDH3. Mutating the residue to a serine removed both the favorable and the unfavorable interactions, respectfully, consistent with the observation that the mutant was a better enzyme when mafosfamide was the substrate and a poorer one when benzaldehyde was the substrate. In contrast, the binding cavity for ALDH1 was already large so removing the methyl group of the threonine did not make the mutant appear to be a better enzyme for aldophosphamide.

Figure 3

Protection of HeLa cells from the killing effects of mafosfamide. Cells were incubated with the compound and the survivors calculated as described in the Material and Methods section. Open box represents cells transformed with an empty vector. Box with parallel lines represents cell transformed with native ALDH3. Lastly, the box with diagonal lines represents cells transfected with the T186S mutant of ALDH3.

Figure 4

Extracts from HeLa cells that were either transfected with the native [lane 1] or T186S mutant form [lane 2] of ALDH3 were analyzed by Western Blotting with anti-ALDH3 antibodies, to confirm the expression levels of the two were comparable. Untransfected cells showed neither ALDH activity nor a protein band after Western blotting [lane 3]. The same amount of total protein was added to each lane.