Influence of DPH1 and DPH5 Protein Variants on the Synthesis of Diphthamide, the Target of ADPRibosylating Toxins

Abstract

The diphthamide on eukaryotic translation elongation factor 2 (eEF2) is the target of ADPribosylating toxins and -derivatives that serve as payloads in targeted tumor therapy. Diphthamide is generated by seven DPH proteins; cells deficient in these (DPHko) lack diphthamide and are toxin-resistant. We have established assays to address the functionality of DPH1 (OVCA1) and DPH5 variants listed in dbSNP and cosmic databases: plasmids encoding wildtype and mutant DPHs were transfected into DPHko cells. Supplementation of DPH1 and DPH5 restores diphthamide synthesis and toxin sensitivity in DPH1ko and DPH5ko cells, respectively. Consequently, the determination of the diphthamide status of cells expressing DPH variants differentiates active and compromised proteins. The DPH1 frameshift variant L96fs* (with Nterminal 96 amino acids, truncated thereafter) and two splice isoforms lacking 80 or 140 amino acids at their N-termini failed to restore DPH1ko deficiency. The DPH1 frameshift variant R312fs* retained some residual activity even though it lacks a large C-terminal portion. DPH1 missense variants R27W and S56F retained activity while S221P had reduced activity, indicated by a decreased capability to restore diphthamide synthesis. The DPH5 nonsense or frameshift variants E60*, W136fs* and R207* (containing intact N-termini with truncations after 60, 136 or 207 amino acids, respectively) were inactive: none compensated the deficiency of DPH5ko cells. In contrast, missense variants D57G, G87R, S123C and Q170H as well as the frequently occurring DPH5 isoform delA212 retained activity. Sensitivity to ADP-ribosylating toxins and tumor-targeted immunotoxins depends on diphthamide which, in turn, requires DPH functionality. Because of that, DPH variants (in particular those that are functionally compromised) may serve as a biomarker and correlate with the efficacy of immunotoxin-based therapies.

Keywords: pseudomonas exotoxin; biomarker; OVCA1; diphtheria toxin; immunotoxin; targeted therapy.

Figure 1

(A) Functionality of diphthamide synthesis genes determine His-eEF2 modification which, in turn, affects toxin-mediated ADP-ribosylation; (B) Exposure of cell extracts to DT and biotinylated NAD results in the bioADP-transfer and hence biotinylation of diphthamide-containing eEF2. The presence of diphthamide on eEF2 can thereby be detected by Western blot assays that detect biotinylated proteins with enzyme-conjugated streptavidin. Biotinylated cellular proteins are also detected and serve as the internal standard (loading control).

Figure 2

Expression of recombinant (A) DPH1 protein or (B) DPH5 protein in DPH-deficient MCF7 cells. DPH1 and DPH5 proteins are visualized by Western blot. DPH1-deficient cells are negative for DPH1 protein with full functionality of the other DPHs; DPH5-deficient cells are negative for DPH5 protein with full functionality of the other DPHs. DT-mediated ADP-ribosylation was detected by enzyme labeled streptavidin, as described in Figure 1. Cellular toxin sensitivity or resistance was assessed by GFP fluorescence in co-transfection assays (details in Materials and Methods).

Figure 3

Expression of recombinant DPH proteins in DPH-deficient MCF7 cells. DPH proteins are visualized in Western blots in cells that express recombinant wildtype or variant DPH1 or DPH5 proteins. DPH-deficient cells are negative for the respective proteins. (A) expression of DPH1 and variants in DPH1-deficient cells; (B) expression of DPH5 and variants in DPH5-deficient cells. Some truncated DPH isoforms or frameshift variants are not detected due to size or instability (too small) or due to lack of recognition by the applied antibodies.

Figure 4

Capability of (A) DPH1 variants and (B) DPH5 variants to restore diphthamide synthesis deficiency of DPH1ko or DPH5ko cells, respectively. Extracts of cells expressing DPH variants were subjected to DT-mediated ADP-ribosylation followed by the detection of toxin-mediated modification of eEF2, as described in Figure 1. Extracts of MCF7 wildtype cells and of DPHko cells transfected with expression plasmids providing the functional wildtype enzyme (wtDPH1 and wtDPH5) serve as positive controls, extracts of mock-transfected cells and reactions without addition of DT serve as negative and specificity controls. Note that the right panel of (A) is a duplicate image of the S221P and R312* samples on the left, in a contrast enhanced setting to demonstrate the presence of weak but unambiguously detectable eEF2 bands. Additional ‘non-specific’ signals (unaffected by DT-modification) above and below eEF2 are biotin-containing cellular proteins.