DPH1 and DPH2 variants that confer susceptibility to diphthamide deficiency syndrome in human cells and yeast models

Abstract

The autosomal recessive diphthamide deficiency syndrome presents as intellectual disability with developmental abnormalities, seizures, craniofacial and additional morphological phenotypes. It is caused by reduced activity of proteins that synthesize diphthamide on human translation elongation factor 2. Diphthamide synthesis requires seven proteins (DPH1-DPH7) with clinical deficiency described for DPH1, DPH2 and DPH5. So far, just a limited set of variant alleles from syndromic patients were functionally analyzed, but databases (gnomAD) list additional so far uncharacterized mutations of human DPH1 and DPH2. Because DPH enzymes are conserved among eukaryotes, their functionality can be assessed in yeast and mammalian cells. Our experimental assessment of known and so far uncharacterized DPH1 and DPH2 missense alleles showed that six mutations are tolerated despite of inter-species conservation. 10 additional human DPH1 (G113R, A114T, H132P, H132R, S136R, C137F, L138P, Y152C, S221P, H240R) and two DPH2 variants (H105P, C341Y) showed reduced functionality and are hence deficiency-susceptibility alleles. Structure models indicate that some variants locate close to the active centers of the enzymes (may directly affect catalysis), others more distant (possibly relevant for electron transfer mediated enzyme activation).

Functional Integrity of Radical SAM Enzyme Dph1•Dph2 Requires Non-Canonical Cofactor Motifs with Tandem Cysteines

“…The sequence of Pyrococcus horikoshii (Ph) Dph2 (UniProt-ID O58832) was aligned to Dph1 and Dph2 from S. cerevisiae (Sc) (UniProt-IDs P40487 and P32461), Arabidopsis thaliana (At) (UniProt-IDs Q8RWW3 and A0A1I9LRW3), Drosophila melanogaster (Dm) (UniProt-IDs Q9VTM2 and Q9VFE9), Mus musculus (Mm) (UniProt-IDs Q5NCQ5 and Q9CR25), or Homo sapiens (Hs) (UniProt-IDs Q9BZG8 and Q9BQC3) using ClustalOmega (https://www.ebi.ac.uk/Tools/msa/clustalo/ accessed on 5 February 2024) and illustrated using Jalview (https://www.jalview.org/development/archive/Version-2_11_2_7/ accessed on 5 February 2024). AlphaFold/ColabFold [29,30]-based models of Dph1•Dph2 were guided by the solved structure of PhDph2 (PDB:3LZD) [7,12], as described in [10,21] (https://colab.research.google.com/github/sokrypton/ColabFold/blob/main/AlphaFold 2.ipynb accessed on 5 February 2024). Structure visualization used PyMOL (https://pymo l.informer.com/1.3/ accessed on 5 February 2024), and the cysteine residues of the atypical radical SAM and/or Fe-S binding motifs were curated manually.…”

“…Lämmli samples were run using SDS-PAGE (12% [w/v] polyacrylamide) and blotted onto PVDF membranes (Millipore, Burlington, MA, USA). These were probed overnight at 4 • C with the anti-eEF2(pan) and anti-eEF2(no diphthamide) antibodies and developed using anti-rabbit secondary antibody horseradish-peroxidase conjugates (Dianova, Hamburg, Germany; working concentration: 1/5000) and Lumi-Light Western blotting substrate (Roche, Basel, Switzerland), as described [10,21,32]. Protein loading was controlled in parallel with anti-Cdc19 antibodies (kindly donated by Prof. Jeremy Thorner, University of California, Berkeley, CA, USA), recognizing yeast pyruvate kinase.…”

“…The latter antifungal stalls ribosomes and blocks protein synthesis in a fashion unrelated to DT but also dependent on the diphthamide décor on eEF2 [38][39][40]. (A) Simplified diphthamide pathway with DPH1 and DPH2 genes required to initiate ACP formation [21]; subsequent DPH products and steps for the completion of diphthamide synthesis [14][15][16][17] are not detailed. Diphthamide can be hijacked to induce cell death (skull-crossbones) either by DT for lethal ADP-ribosylation of eEF2 or sordarin (sor), which in a complex with the décor on eEF2 stalls ribosomes [40].…”

“…Diphthamide is formed via a dedicated synthesis network (i.e., Dph1-Dph8) on a conserved histidine of eEF2 in eukaryotes (His-699 in budding yeast Saccharomyces cerevisiae and His-715 in Homo sapiens) and EF2 in some archaea [14][15][16][17]. Importantly, human patients deficient in diphthamide-modified eEF2 display symptoms of a neurodevelopmental disorder known as diphthamide deficiency syndrome (DDS) [18][19][20][21][22] and diphtheria toxin (DT) from Corynebacterium diphtheriae attacks the diphthamide décor on eEF2 by ADP-ribosylation to block mRNA translation in infected host cells [23]. Yeast diphthamide mutants (i.e., dph1∆-dph8∆) are resistant to growth inhibition by DT, which is why the toxin is a valuable molecular tool to tell diphthamide proficiency from deficiency in vivo [17,24].…”

“…The yeast Dph1•Dph2 heterodimer contains one Fe-S cluster per subunit, yet both Dph1 and Dph2 lack the consensus (CX 3 CX 2 C) typical of classical radical SAM domains [9,10]. Instead, Fe-S cluster coordination relies on atypical Cys-based motifs in Dph1 (CX 105 CX 129 C) and Dph2 (CX 21 CX 234 C), which upon mutation cause Fe-S cluster loss and diphthamide defects [9,10] and, importantly, are verified triggers of DDS [21]. Intriguingly, there are other cysteines with ill-defined roles in protein function and stability: Dph1Cys-134 and Dph2Cys-106.…”