Figure 1.
CADA selectively and reversibly down-modulates human CD4 in a dose-dependent way.
(A) Structure of CADA, molecular weight (MW) = 618. (B) U87 cells stably expressing hCD4 were treated with a serial 1∶10 dilution of CADA. After 24 h, cell lysates were analyzed by immunoblotting with anti-CD4 or anti-clathrin antibodies. A non-transfected control is included (first lane). Molecular mass is in kDa. (C) Dose-response curve showing the effect of CADA on hCD4 expression in the human T-lymphoid cell line SupT1 expressing CD4 naturally (circles; IC50 is 0.55 µM), in the stably transfected human U87 glioblastoma (crosses, IC50 is 0.32 µM) and Hela epithelial cervical cancer cells (triangles, IC50 is 0.27 µM). Surface CD4 expression levels were normalized to non-treated controls as determined by flow cytometry (n≥4). (D) Dose-response curve showing the selective effect of CADA on primate CD4. Primary T-cells isolated from the blood of humans, macaques, or mice were stimulated with 2 µg/ml phytohaemagglutinin (PHA) and treated with CADA for 3 days. Next, cells were stained for surface CD4 and analyzed by flow cytometry. Data represent mean values from two different donors. (E) Human T-lymphoid MT-4 cells were treated with control medium or 3.2 µM CADA. After 24 h, cell surface proteins were quantified by flow cytometry. DC-SIGN expression was determined in stably DC-SIGN-transfected CEM cells. Protein expression levels were normalized to non-treated controls (three independent experiments of 10,000 analyzed cells each). *p<0.01. (F) Flow cytometry analysis of surface proteins in transfected HEK293T cells treated with CADA for 48 h. Protein expression levels were normalized to non-treated controls (three independent experiments). *p<0.01. (G) CD4 down-modulation by CADA is reversible. Primary human T-cells isolated from the blood of healthy donors were stimulated with 2 µg/ml PHA and treated with CADA for 3 days. Next, cells were washed and given control medium for another 6 days. Cells were harvested at specified time points and analyzed for surface CD4 expression by flow cytometry. Data represent mean values from two different donors. (H) Kinetics of CD4 down-modulation showing a slow but steady removal of surface CD4 with CADA, and a fast but transient decrease in CD4 with PMA. Graph represents flow cytometric surface CD4 analysis of CHO cells stably expressing hCD4 either treated with 16 µM CADA or 0.16 µM PMA, or treated simultaneously with CADA and PMA. Surface CD4 expression levels were normalized to non-treated controls (n≥2).
Figure 2.
CADA specifically inhibits the biogenesis of human CD4.
(A, B) CADA inhibits the biosynthesis of CD4. CD4+.CHO cells were washed and kept in methionine and cysteine-free medium in the presence or absence of 16 µM CADA for 45 min before exposure to [35S]methionine/cysteine (Met/Cys) for 30 min. Pulsed-labelled cells were then washed, lysed, and analyzed directly (A) or incubated in normal medium for up to 4 h (chase) in the presence or absence of 16 µM CADA (B). At specified time points cell lysates were immunoprecipitated for CD4. The flow through fraction (FT) of the CD4-immunoprecipitated samples is also presented. Note that the weaker CD4 bands in the control samples at longer chase time points are the result of the high turnover of hCD4 in CHO cells. Molecular mass is in kDa. (C–F) CD4 negative and stably CD4-YFP transfected CHO cells were pretreated with CADA (5 µM) or DMSO for 1 h before starvation in Met/Cys free medium with CADA, DMSO, or 50 µg/ml CHX. Cells were pulsed for 30 min, washed, and incubated in fresh medium without serum for 90 min. After collection of supernatant proteins (Media) cells were first permeabilized with digitonin buffer to obtain the cytosolic cell fraction before lysis in NP-40 buffer to collect the membrane proteins. Membrane fractions were further incubated with Concanavalin A (ConA) agarose beads (Glycosylated). Molecular mass is in kDa. (D) Quantification of 35S incorporation in (C) by scintillation counting (n = 4). NS, not significant; *p<0.01. (F) ConA fraction from a repeat experiment with CHO.CD4+ cells treated with DMSO or CADA. Note that the expression of CD4-YFP (∼80 kDa) is clearly reduced by CADA-treatment (indicated by arrow).
Figure 3.
CADA inhibits the expression of human CD4 in a signal peptide-dependent way.
(A) Schematic representation of the hCD4 C-terminal deletion constructs used. Numbers represent the corresponding amino acids of the pre-protein. The N-terminal SPs are indicated by a black rectangle. In mutant hCD4-426 the intracellular C-terminal domain has been removed. Construct hCD4-CD8 is composed of the D1 and D2 domains of hCD4 fused to the membrane-bound alpha chain of hCD8. CD4− A2.01 human T-cells stably expressing hCD4WT, hCD4-426, or hCD4-CD8 were treated for 48 h with control medium or CADA (8 µM) and stained with anti-CD4 antibody to determine surface expression levels of WT or mutant CD4 by flow cytometry. Graph shows relative receptor expression levels of CADA-treated cells as compared to non-treated controls. (B) Schematic representation of the human/mouse chimaeric constructs used. Numbers represent the corresponding amino acids of the pre-protein. The N-terminal SPs are indicated by a black rectangle. The human/mouse chimaeric constructs are C-terminally fused to YFP for ease of flow cytometric detection. Construct hCD4-mD1 is composed of hCD4 in which the D1 region is replaced by the corresponding D1 domain of mouse CD4. It still contains the SP of hCD4 and the first seven residues of mature hCD4. Construct hCD4-mD2 is composed of hCD4 in which the D2 is replaced by the corresponding D2 domain of mouse CD4. HEK293T cells were transiently transfected with the expression plasmids and left either untreated or treated with CADA for 24 h. YFP fluorescence was determined and normalized to non-treated controls.
Figure 4.
CADA dose-dependently inhibits co-translational translocation of human CD4.
(A) Schematic representation of the constructs used. Numbers represent the corresponding amino acids of the pre-protein. The N-terminal SPs are indicated by a black rectangle. The star represents the N-glycosylation site in full length hCD4. (B) HEK293T cells were transiently transfected with the expression plasmids and left either untreated or treated with CADA for 48 h. Receptor expression levels were determined by flow cytometry and normalized to non-treated controls (n≥3). NS, not significant. (C) Transcripts encoding full length (hCD4 WT) and truncated (hCD4-201) hCD4 were translated in the presence of [35S]methionine and, where indicated, RMs and/or CADA (15 µM). Equal aliquots of the translated material were left untreated or treated with PK in the presence or absence of Triton X-100 (Triton). Samples were separated by SDS-PAGE and analyzed by autoradiography. The positions of precursor (open arrowhead), signal-cleaved (solid arrowhead), and glycosylated (asterisk) species are indicated. Of note is the lower appearance of the PK-protected glycosylated form of hCD4 WT (lane 5) compared to the non-PK form (lane 4), which is due to proteolytic cleavage of the 38 amino acid cytoplasmic domain, resulting in a smaller and faster migrating fragment. (D) As in (C), except that a 1∶10 serial dilution of CADA was used. (E) Translocation data of (D) quantified by phosphorimager analysis. The translocation efficiencies, at 15 µM CADA, normalized to control are indicated (mean values of at least three experiments). (F) Cytosolic degradation of precursor hCD4 in CADA-treated cells. HEK293T cells were transiently transfected with V5-tagged hCD4, treated with CADA (10 µM) and/or the proteasome inhibitor MG132 (200 nM). Cells were lysed in NP-40 buffer and analyzed by immunoblotting for CD4 with anti-V5 antibody. For the cell loading control, an anti-actin antibody was used. The predicted molecular mass for the SP-cleaved sCD4 (V5sCD4) was 22 kDa, whereas that of the precursor form ([hCD4]-V5sCD4) was 25 kDa. Molecular mass is in kDa. Similar data were obtained with 2 µM CADA. One representative experiment out of two is shown.
Figure 5.
CADA binds to the signal peptide of human CD4 and interacts primarily with its hydrophobic H-region.
(A) Binding of CADA to the SP of human but not to mouse CD4. Chemically synthesized peptides composed of the signal sequence and eight N-terminal residues of mature CD4, plus a PEG11 linker and a biotinylated lysine at the C-terminus, were captured on a streptavidin C1 chip for SPR analysis. The chip density was 84 and 153 resonance units (RUs) for human and mouse SP, respectively. Graph shows interaction of CADA (160 µM) to hCD4 SP (green) but not to mCD4 (blue). (B) As in (A), but for the binding of SRP to the SPs. The chip density was between 120 and 130 resonance units (RUs). Graph shows dose response of SRP (nM) to hCD4 SP (green). Similar binding of SRP (25 nM) was observed for mouse CD4 SP (blue). (C) The inactive CADA-analog MFS105 (MFS, red line) did not bind to the hCD4 SP (up to 1,000 µM), whereas for CADA a dose-dependent binding was measured (black lines). The chip density was 95 RUs. For clarity of the figure, only the 500 µM line is shown. Comparable data were obtained in an independent experiment shown in Figure S4D. (D) Schematic representation of the constructs used. The N-, H-, and C-regions of the SP of hCD4 were exchanged for those of mouse CD4. Also the first seven N-terminal residues of the mature protein were swapped as indicated. The residues of hCD4 are depicted in green, whereas those of mouse CD4 are in blue and underlined. (E) Flow cytometry analysis of HEK293T cells transiently transfected with the expression plasmids from (D) and left either untreated or treated with CADA for 48 h. Cells were collected and stained for hCD4. CD4 expression levels were normalized to non-treated controls (n≥3). The IC50 values of CADA for each construct are included.
Figure 6.
CADA inhibits co-translational translocation of hCD4-pPL at a post-targeting step.
(A) Schematic representation of the truncated nascent chains of [hCD4]-(7)-pPL. The nascent chains remain attached to the ribosome through a peptidyl-tRNA bond. The (approximately) 30 residues at the C-terminus of the polypeptide that are buried inside the ribosome are indicated by a grey rectangle. The SP cleavage site (at residue 25) is also marked. (B) In vitro translation, targeting, translocation (puromycin-release), and PK digestion of [35S]methionine-labelled [hCD4]-(7)-pPL 80-mers (left panel) and WT pPL 78-mers (right panel). Intact peptidyl-tRNA bands (NC-tRNA, arrow) are indicated and represent the targeted and PK protected RNCs. Note that the presence of microsomes (RM) releases some nascent chains (NC) from tRNA in an unproductive way (compare lanes 1 and 2, open arrowhead). The positions of released precursor (open arrowhead) and signal-cleaved (solid arrowhead) polypeptide chains are indicated. The SP of hCD4 but not pPL was also detected (open circle). Molecular mass is in kDa. (C) Dose response of CADA (µM) on translocation of RNCs as in (B). Graph shows translocation fractions of the autoradiogram quantified by phosphorimager analysis. (D) Time of addition of CADA. [hCD4]-(7)-pPL nascent chains of 80 residues (80-mers) were synthesized in the absence of membranes before exposure to RM for targeting. Nascent chains were left untreated (C, control), or were treated with CADA (15 µM) either administered at the beginning of the synthesis in the translation mixture (R, ribosomes), administered to membranes for pretreatment (M, microsomes), or applied to the RNC/RM mixture 15 minutes after initiation of targeting but before puromycin release (P, post-targeting). Graph shows translocation data of three experiments quantified by phosphorimager analysis. NS, not significant. *p<0.01.
Figure 7.
CADA locks the hCD4 SP in an intermediate inversion conformation in the translocon.
(A) Line diagram of the constructs with an N-terminal extension of the SP lacking (no-Glyc) or containing a diagnostic glycosylation site (N-Glyc). For ease of comparison, the numbering refers to the non-extended WT hCD4 SP-containing polypeptide. The N-glycosylation site is underlined and the star indicates the N-glycosylated form of the protein. (B) Autoradiogram of truncated in vitro translated [35S]methionine-labelled [hCD4]-(7)-pPL RNCs with the glycosylation site-containing extension (N-Glyc). A non-glycosylation control (no-Glyc) is included (every first lane). Nascent chains were synthesized in the presence of RM, with or without CADA (15 µM). Samples were separated by SDS-PAGE and analyzed by autoradiography. Intact RNCs (NC-tRNA) and N-terminal glycosylated RNCs (asterisk) are indicated. Notice that CADA strongly inhibited the glycosylation of the 17+71-mer. (C) Same as in (B), but for the puromycin-treated samples. Equal aliquots of the translated material were left untreated or treated with PK. Released non-processed nascent chains (NC, open arrowhead), N-terminal glycosylated forms (asterisk) and signal-cleaved C-terminal translocated species (solid arrowhead) are indicated. (D) Fraction of glycosylated nascent chains in the absence or presence of CADA as in (C), quantified and plotted against chain length (n≥3). *p<0.05; **p<0.001. (E) Fraction of SP-cleaved C-terminal translocated nascent chains in the absence or presence of CADA as in (C), quantified and plotted against chain length (n≥3). *p<0.01; **p<0.001. (F) Schematic of ribosome/translocon complexes showing probable positioning of the SP of hCD4 in absence or presence of CADA. Star represents N-glycosylation.