CD74 is a member of the regulated intramembrane proteolysis-processed protein family

Abstract

Quite a few regulatory proteins, including transcription factors, are normally maintained in a dormant state to be activated after internal or environmental cues. Recently, a novel strategy, requiring proteolytic cleavage, was described for the mobilization of dormant transcription factors. These transcription factors are initially synthesized in an inactive form, whereas "nesting" in integral membrane precursor proteins. After a cleavage event, these new active factors are released from the membrane and can migrate into the nucleus to drive regulated gene transcription. This mechanism, regulated intramembrane proteolysis (RIP), controls diverse biological processes in prokaryotes and eukaryotes in response to a variety of signals. The MHC class II chaperone, CD74 (invariant chain, Ii), was previously shown to function as a signaling molecule in several pathways. Recently, we demonstrated that after intramembranal cleavage, the CD74 cytosolic fragment (CD74-ICD) is released and induces activation of transcription mediated by the NF-kappaB p65/RelA homodimer and the B-cell-enriched coactivator, TAF(II)105. Here, we add CD74 to the growing family of RIP-processed proteins. Our studies show that CD74 ectodomain must be processed in the endocytic compartments to allow its intramembrane cleavage that liberates CD74 intracellular domain (CD74-ICD). We demonstrate that CD74-ICD translocates to the nucleus and induces the activation of the p65 member of NF-kappaB in this compartment.

Figure 1.

Diagram of tagged CD74 constructs used in this study.

Figure 2.

Arrival to the endocytic compartments is essential for CD74 intramembrane cleavage. (A) Fluorescence microscopy of 293 cells transfected with the GFP FL p31 CD74 chimera. At 24 h posttransfection, cells were incubated in the presence or absence of brefeldin A (BFA; 2.5 μg/ml) and in the presence or absence of NOC (2.5 μg/ml). (B and C) Western blots (B) FL-CD74 was transfected into 293 cells, which were then treated with BFA (2.5 μg/ml) in the presence or absence of NOC (2.5 μg/ml). Total lysates were separated on Tricine gel and analyzed with IN1 rat monoclonal antibodies that recognize the CD74 cytosolic domain. Ratio calculation: The intensity of CD74-ICD band in each treatment was divided by the intensity of a nonrelevant band in each lane. The CD74-ICD ratio in the absence of any treatment was normalized to 1 and the ratio for each treatment was calculated as the intensity of the treatment sample relative to 1. The results presented are representative of three different experiments. (C) Total splenocytes were treated with BFA (2.5 μg/ml), NOC (2.5 μg/ml), LEU (1 mM), and with MON (10 μM) for 3 h; lysates were separated on Tricine gel and analyzed with IN1. The results presented are representative of five different experiments.

Figure 3.

Processing in the endocytic compartments is essential for CD74 intramembrane cleavage. (A) Fluorescence microscopy of 293 cells transfected with GFP FL and GFP FL LI 7,8 AA (green) shown together with the endocytic compartment staining (red). (B) Western blot analysis of 293 cell lysates transfected with FL myc and FL LI7,8AA Myc; at 24 h posttransfection, lysates were separated on Tricine gel and analyzed with IN1. (C and D) FL-myc was transfected into 293 cells, which were treated with MON (10 mM) (C) and with chloroquine (25 and 100 μM; D) for 3 h. Total lysates were separated on Tricine gel and analyzed with IN1. The intensity of the CD74-ICD band after each treatment was divided by the intensity of a nonrelevant band in each lane. The CD74-ICD ratio in the absence of any treatment was normalized to 1 and the ratio for each treatment was calculated as the intensity of the treatment sample relative to 1. The results presented are representative of three different experiments.

Figure 4.

Localization of CD74 cytosolic domain. (A) Fluorescence microscopy of 293 cells transfected with the GFP FL, GFP 1–197, GFP 1–100, and GFP 1–82 chimera (green) together with the endocytic compartment staining (red). The results presented are representative of three different experiments. (B) 293 cells expressing the indicated forms of CD74, were preincubated for 40 min at 4°C with an EGF conjugated to Alexa Fluor 488. Thereafter, cells were incubated for 20 min at 37°C, permeabilized, and stained with an anti-CD74 antibody and detection using a CY-conjugated secondary antibody. The distribution of fluorescent EGF and CD74 is shown, along with merge panels depicting both signals.

Figure 5.

CD74 lumenal domain regulates its intramembrane cleavage in the endocytic compartments and NF-κB activation. (A and B) 293 cells were transfected with CD74 full length, CD74 1–197, CD74 1–160, CD74 1–120, CD74 1–100, and CD74 1–82. Myc tagged constructs were separated on Tricine gels and analyzed with anti CD74. Ratio calculation: the intensity of the CD74-ICD band in each construct was divided by the intensity of a nonrelevant band in each lane. The CD74-ICD ratio in the full-length (A) or 1–120 (B) transfection was normalized to 1 and the ratio of CD74-ICD in each construct was calculated as the intensity relative to 1. (C–E) 1–82 and 1–60 myc constructs were transfected into 293 cells and treated with MON (10 μM; C). 293 cells were transfected with 1–82 myc construct and treated with LEU (1 mM) for 3 h (D). 293 cells were transfected with 1–82 and 1–82 LI 7,8 AA (E). Total lysates were separated on Tricine gel and analyzed with IN1 rat monoclonal antibodies, which recognize the CD74 cytosolic domain. Ratio calculation: the intensity of CD74-ICD band in each treatment was divided by the intensity of a nonrelevant band in each lane. CD74-ICD ratio in the absence of any treatment was normalized to 1, and the ratio for each treatment was calculated as the intensity for that treatment relative to 1. The results presented are representative of three different experiments. (F) Nonsaturating amounts of G4-p65 TA1 (the TA1 activation domain of p65/RelA fused to the GAL4 DBD) and luciferase reporter gene containing the Gal4 binding site were transfected into 293 cells in the presence of 25 ng of FL, 1–160, 1–100, and 1–82 Myc constructs. Luciferase activities were normalized to the activity of cotransfected RSV promoter-driven Renilla reporter luciferase, which was used to correct for differences in transfection efficiencies. Fold activation was calculated as the activity of each CD74 construct relative to the activity of the empty plasmid. The graph represents the average of five independent experiments.

Figure 6.

γ-secretase cleaves CD74. (A) 293 cells were cotransfected with FL-CD74 Myc and a dominant negative construct of presenilin (PS1). Lysates were separated into a particulate and cytosolic fractions, separated on Tricine gel and analyzed with IN1 antibody. The results presented are representative of five different experiments. (B and C) 293 cells were transfected with FL-CD74 Myc (B) or 1–82 Myc (C) constructs and treated with or without DAPT (50 μM; B and C) or (ZLL)2-ketone (10 μM; C) for 5 h. Lysates were separated into a particulate and cytosolic fractions, separated on Tricine gel, and analyzed with IN1 antibody. The results presented are representatives of five different experiments. CD74-ICD ratio in the absence of any treatment was normalized to 1, and the ratio for each treatment was calculated as the intensity for that treatment relative to 1.

Figure 7.

CD74-ICD translocates to the nucleus to activate NF-κB. (A) Immunofluorescence microscopy of 293 cells transfected with 1–82 (1) myc or 1–42 myc (2 and 3). After transfection, cells were incubated 48 h, fixed, permeabilized, and stained with IN1 antibody followed by CY3-conjugated anti-rat antibody (1 and 2). The stained cells were analyzed by fluorescence microscopy. In (3), DNA was visualized using DAPI staining of the same cells (blue). (B) Cells were transfected with 1–42 myc and fractionated 48 h later. Equivalent cell numbers from each fraction were loaded on tricine gel. 1–42 was detected by Western blot analysis with IN1. The blot was then stripped and analyzed with anti-α-tubulin antibody as a marker for cytosolic fraction and with anti-SP1 for the nuclear fraction.

Figure 8.

CD74-ICD activates NF-κB in the nucleus. (A) Schematic illustration of the constructs of CD74 1–42, 1–42 nuc, and 1–42 mito. (B) HEK 293 cells were transfected with 1–42 Myc, 1–42 nuc, or 1–42 mito and the localization of CD74 was analyzed by immunofluorescence as described above. DNA was visualized using DAPI staining of the same cells (blue). (C) Subsaturating amounts of G4-p65 TA1 (the TA1 activation domain of p65/RelA fused to the Gal4 DNA-binding domain) and Gal4 reporter plasmid were transfected into HEK 293 cells along with increasing amounts of 1–42 myc, 1–42 nuc, or 1–42 mito plasmids. Luciferase activity was measured, and activation fold was calculated as described for Figure 5.