In various mammalian cells, the function and subcellular localization of eIF5A have been studied by transfection with eIF5A vector alone [13], [15], [16] and [17]. However, we have found that exogenously produced eIF5A is poorly modified by endogenous DHS and/or DOHH in mammalian
NSC 74859 and remains largely as the unmodified precursor, eIF5A(Lys). As shown in Fig. 1, in HeLa cells transfected with p3XFLAG-CMV-7.1/eIF5A vector alone, there was little labeling of FLAG-eIF5A, even when FLAG-eIF5A protein was highly expressed as shown in the Western blot (Fig. 1A, lanes 2 and 5). The lack of labeling of FLAG-eIF5A upon incubation with [3H]spermidine indicates that FLAG-eIF5A was not being modified by endogenous DHS and DOHH. Only upon cotransfection with the modification enzymes, was the FLAG-eIF5A precursor fully modified to the hypusine form (Fig. 1A, lanes 3 and 6, hypusine analysis data not shown). Similar results were obtained with GFP-eIF5A (Fig. 1B). These findings support a premise that FLAG-eIF5A or GFP-eIF5A, produced upon transfection with an eIF5A vector alone, represents largely the non-hypusinated eIF5A precursor, whereas that produced by cotransfection with all three
vectors (eIF5A, DHS and DOHH) represents the hypusine form.