Fig 1.
The glutamate residue at position 10 in the HCF-1PRO repeat inhibits OGT binding.
(A) (Top) Schematic representation of the human HCF-1 protein showing the six HCF-1PRO repeats as yellow triangles. (Bottom) Schematic of the recombinant HCF-1rep1 precursor (HCF-1 residues 867–1071), containing the HCF-1PRO repeat 1 (rep1) and surrounding sequences fused to glutathione-S-transferase (GST). The 26 residues of rep1 are shown and the site of proteolysis is marked at the C/E sequence by a black arrowhead. The 20 amino acid core sequence [14] is shaded gray, and the cleavage and threonine regions are indicated. (B) E10 exhibits inhibitory contributions to OGT binding. Computational estimation of the contributions to the binding free-energy, ΔGbind, of single amino acid residues in the HCF-1PRO repeat (P7-T14) for the association with OGT. Negative and positive values correspond to favorable and unfavorable OGT-association contributions, respectively, in the in silico structural complex (based on the 4N3B structure [22]). (C) An E10 to alanine substitution enhances HCF-1PRO-repeat–OGT binding. In vitro HCF-1rep1–OGT binding assay in the presence of UDP-GlcNAc with HCF-1rep1 constructs containing individual alanine substitutions in the HCF-1PRO repeat (P7A-T14A). HCF-1rep1 binding to OGT was quantified from an immunoblot (S1B Fig) as ratio of OGT-bound HCF-1rep1 to total HCF-1rep1 in the assay. Obtained values are presented as log2 fold change relative to wild-type HCF-1rep1–OGT binding. (D) HCF-1PRO-repeat–OGT binding is influenced by the E10 side-chain properties and exhibits sensitivity to UDP-GlcNAc. In vitro HCF-1rep1–OGT binding assay. Constructs with wild-type (WT), E10 missense mutations (E10A, E10D, E10Q, and E10S), and T17-T22A mutations were incubated in the presence (left panel) or absence (right panel) of UDP-GlcNAc. Detection of OGT and HCF-1rep1 was performed using the indicated antibodies. *, IgG heavy chain. (E) Computational estimation of the contributions to the binding free-energy, ΔGbind, of single amino acid residues at the E10 cleavage site for the association with OGT. Shown are residue contributions of WT (E10), alanine (E10A), aspartate (E10D) or glutamine (E10Q) side-chains in the presence or absence of UDP-GlcNAc in the in silico structural complex (based on the 4N3B structure). (F) Close-up view (4N3C crystal structure [22]) of the OGT active site with the HCF-1PRO repeat and UDP-GlcNAc. The deprotonated E10 oxygen atom exhibits an unfavorable interaction with OGT. (G) Snapshot from a molecular dynamics simulation (based on the 4N3B structure) of the OGT active site in complex with the HCF-1PRO repeat without UDP-GlcNAc. The displayed frame is representative of the average distances sampled along the simulations. In (F) and (G), the E10 side-chain is shown in ball and stick representation (carbons: gray, nitrogen: blue, oxygens: red), and dashed lines indicate distances between atoms.
Fig 2.
HCF-1PRO-repeat cleavage enhancement by a sequence nearby the HCF-1PRO repeat 1.
(A) HCF-1PRO-repeat cleavage is context dependent in vivo. (Top) Schematic of the HCF-1rep1 precursor subdivided into Region I (25 residues, blue), Region II (58 residues, pink), and Region III (60 residues, gray). (Bottom) HEK 293 cells were transfected with expression vectors encoding HCF-1rep1 FL or deletion constructs, either containing or lacking Regions I, II or III. Proteins were immunoprecipitated by an N-terminal HA-tag and assayed for cleavage by visualization by α-HA-tag immunoblot. *, C-terminal precursor truncations. (B) Region II enhances HCF-1PRO-repeat cleavage in vitro. Cleavage efficiency during an in vitro cleavage assay time course of selected HCF-1rep1 constructs. HCF-1rep1 constructs were incubated with OGT for 0 to 8 h and precursor and resulting N-terminal cleavage products were analyzed for cleavage by α-GST-immunoblot. Uncleaved and cleaved products were quantified and cleavage efficiencies determined as cleaved products over total. Shown are the means and standard deviations of three independent experiments. (C) Region II cleavage-enhancement activity is sequence specific. In vitro cleavage assay of HCF-1rep1 FL and Region II constructs containing a scrambled Region II sequence (+II_scrambled) or an inactive HCF-1PRO repeat (+II_T17–22A). Resulting precursor and N-terminal cleavage products were analyzed for cleavage with the indicated antibodies. (D) Region II activates the inactive POUrep2 construct for cleavage. (Left) Schematic of the GST-fusion construct POUrep2 containing HCF-1PRO repeat 2 (rep2), embedded in between the POU-specific (POUS) and POU-homeo domains (POUH) of Oct-1. Region II or Region III were inserted N-terminal of rep2, respectively. (Right) In vivo cleavage activities in HEK 293 cells, transiently transfected with transfection medium (mock) or POUrep2 encoding plasmids. Precursors and cleaved fragments were purified via immunoprecipitation of an N-terminal HA-tag and cleavage assayed using the indicated antibody. In (A), (C) and (D), prominent (●) and faint (⭕) cleavage products are indicated.
Fig 3.
The Region II CEE represents an OGT-binding sequence.
(A) Region II enhances OGT–HCF-1rep1 binding. Full-length (FL) and deletion HCF-1rep1 constructs were tested for OGT binding in the presence of UDP-GlcNAc using an in vitro OGT-directed pull-down assay. Detection of OGT and HCF-1rep1 was performed, using the indicated antibodies. Shown are 100% of OGT pull-down (panels a and b) and 11% of the input (panels c and d). *, IgG heavy chain. (B) HCF-1PRO-repeat-independent OGT–Region II binding. (Left) Schematics of the HCF-1 constructs used in this experiment. (Right) HCF-1rep1 containing Region II and an OGT-binding defective HCF-1PRO repeat (+II_T17–22A), or GST-fusion constructs containing Region II (wild-type or scrambled) alone or Region III alone (II_alone, II_scramb_alone, III_alone) were tested for binding with wild-type (WT) (left panel) or 5N-5A mutant (right panel) OGT. HCF-1 binding was detected as in (A). In (A) and (B), weak (⭕) and effective (●) OGT binding is indicated.
Fig 4.
Region II CEE O-GlcNAcylation and HCF-1PRO-repeat proteolysis are independent OGT activities.
(A) (Left) The full-length (FL) HCF-1rep1 precursor (band a) and the N-terminal cleavage product (band b) were purified from HEK 293 lysates via α-HA-epitope immunoprecipitation and visualized by Coomassie staining. The bands were analyzed for O-GlcNAcylation and phosphorylation sites by LC-MS/MS. (Right) Schematic representation of identified HCF-1rep1 O-GlcNAcylation (squares) and phosphorylation (yellow circles) sites in the uncleaved HCF-1rep1 precursor. The HCF-1 sequences covered by the analysis (residues 867–1071) and the engineered trypsin cleavage sites A933K and M951K are indicated below the diagram. Red and blue squares indicate confident (Mascot score > 23 & probability of localization > 70%) and potential (Mascot score 14–22 or probability of localization 50–70%) O-GlcNAcylation sites, respectively. Squares surrounded in black indicate previously identified sites [9]. The HCF-1 Region II CEE amino acid sequence spanning a peptide sequence used in subsequent analyses (underlined: 901–933K) is shown below the diagram.(B) Analysis of a representative Region II CEE peptide (901–933K sequence shown in A) by LC-MS/MS for proportions of different O-GlcNAcylated forms. The proportions of 901–933K peptides containing 0, 1, 2 or 3 attached O-GlcNAc moieties are given for each sample in percent. HCF-1rep1 constructs were synthesized in HEK 293 cells and peptides were derived from constructs containing wild-type (WT) or mutated (E10A, E10D, E10Q, E10S, T17–22A) HCF-1PRO repeats, or containing a deletion of the HCF-1PRO-repeat sequence (∆PRO). The results with WT precursor, E10A, and E10S were confirmed in a second independent experiment.(C) HCF-1rep1 O-GlcNAcylation is not fundamental for HCF-1PRO-repeat cleavage. In vitro cleavage activities of wild-type OGT (WT) and an O-GlcNAcylation compromised OGT mutant (D554H_H558D) on selected HCF-1rep1 substrates. Cleavage and O-GlcNAcylation activities of constructs containing the full-length HCF-1rep1 sequence (FL), or the Region II CEE (+II) or Region III (+III) sequences were analyzed by immunoblot using the indicated antibodies. We note that the lack of the OGT D554H_H558D O-GlcNAcylation activity results in differential mobility of the HCF-1rep1 cleavage products during electrophoresis. Prominent (●) and faint (⭕) cleavage products are indicated.