DDX3 functions as an eIF4E-inhibitory protein to specifically repress cap-dependent translation, utilizing an N-terminal eIF4E-binding consensus YIPPHLR motif to interact with eIF4E (physical interaction), thereby blocking the formation of pre-initiation complex eIF4F and translation initiation by trapping eIF4E in a translationally inactive complex. DDX3 has been reported to localize into cytoplasmic SGs while overexpressed or after arsenite treatment stress (protein localization). A GST pull-down was conducted with bacterially expressed fusion proteins, indicating a direct physical interaction between DDX3 and PABP1. ShRNA-specific knockdown of endogenous DDX3 (RNAi) in HeLa cells in vivo led to the formation of much fewer eIF4E-positive SG foci on sorbitol treatment (particle size and count by microscopy). Overexpression of FLAG–DDX3 alone could induce the assembly of PABP1-containing SGs without stress stimuli. Overexpression of FLAG–DDX3_DQAD (impaired ATPase and helicase activity) and FLAG–DDX3_AAA (oss of RNA-unwinding activity) mutants induced PABP1- and eIF4E-containing SG assembly without stress treatment (particle size and count by microscopy), and neither affected the ratio of the cells harbouring PABP1-positive SGs nor had any significant effect on the number of PABP1-positive SGs, indicating that enzymatic activities have no critical role in the SG-inducing ability of DDX3. In vivo mutagenesis /truncation studies indicated that, under resting conditions, the SG-induction ability of DDX3 is mapped to its N-terminal region of amino acids 1–100 and that the DDX3–eIF4E interaction is critical. DDX3-silenced cells (RNAi) lost viability significantly after osmotic stress release, indicating that DDX3 down-regulation reduces cell viability (other change in phenotype/functional readout) after osmotic stress. Overexpression of the EIF4E-binding-defective DDX3_L43A mutant failed to rescue cell survival following oxidative stress, as cell viability decreased from 88% (wild-type DDX3) to 53% (DDX3L43A) (PMID:21883093). Both CBP and p300, but not PCAF or Tip60, led to acetylation (other PTMs) of endogenous or overexpressed DDX3X in vivo when expressed in HEK293T cells. Five acetylation sites were detected in the IDR1 of DDX3 (K50, K64, K66, K81 and K118). Presence of DDX3X-K118Ac (immunodetection assay) increased upon treatment of cells with HDAC6-specific inhibitors, but not with other HDAC inhibitors. Acetylation assays (other PTMs) confirmed that stress induces (perturbation of the cell environment to induce phenotypic changes) acetylation of DDX3X and other proteins. Oxidative stress inducers including arsenite, puromycin and sorbitol all significantly increased acetylation of DDX3X-K118, while other stresses did not. When adding polyethylene glycol (PEG) to the in vitro expressed and purified DDX3X IDR1 sample, it became turbid (change in optical properties), and micron-sized droplets (particle size and count) were observed by differential interference contrast (DIC) microscopy. The turbidity (OD600) of unacetylated and acetylated IDR1 during LLPS was quantified. Incubation of purified IDR1 with CBP and acetyl-CoA (other PTMs) led to acetylation of all of its ten lysine residues, including K118 and the turbidity (change in optical properties) of this acetylated IDR1 solution was much lower compared to the solution of unacetylated IDR1. The turbidity was restored upon co-incubation of acetylated IDR1 with the purified catalytic domains of HDAC6 (other PTMs) indicating that lysine acetylation impairs the LLPS of IDR1 and that this effect is reversible. The total area of G3BP foci were quantified in vivo, and in DDX3X KO cells clear defects in G3BP granule formation were observed (morphology, particle size and count by microscopy) under oxidative stress, energy depletion and translation inhibition, indicating that DDX3X is important for SG assembly under these specific stress conditions. The reduced total area of G3BP foci in cells lacking DDX3X is restored to WT levels by re-expression of WT or del-IDR2 DDX3X, an observation indicating the critical role of IDR1 for SGs. Granule growth of the acetyl-mimic mutants starts as for the WT protein, but becomes impaired during the process due to low LLPS efficiency; deacetylation by HDAC6 is then required for SG maturation (PMID:30531905).