The Protease Inhibitor Cystatin C Is Differentially Expressed among Dendritic Cell Populations, but Does Not Control Antigen Presentation

Mice

C57BL/6, Cat S^−/− (6) and CyC^−/− (27) mice were bred in the animal facilities of the Walter and Eliza Hall Institute (Melbourne, Australia) according to institutional guidelines and used at 6–8 wk of age. As described previously, neither of the knockout strains used in this study show major developmental or health defects.

Purification of splenic DC

Splenic mouse DC were purified as described elsewhere (24). The protocol used excluded B220⁺ “plasmacytoid DC” and yielded 70–80% pure preparations of CD11c⁺ DC, in which the contaminating cells were mostly macrophages. Segregation of the CD4⁺ and CD8⁺ DC populations was conducted by two different methods. For biochemical experiments in which relatively large numbers of cells were needed, the DC preparation was first incubated with the anti-CD4 rat mAb GK1.5, followed by anti-rat magnetic beads (Dynabeads; Dynal Biotech, Oslo, Norway), to retrieve the CD4⁺ cells. The CD8⁺ cells were then purified sequentially using the rat anti-CD8 mAb 53-6.7 and anti-rat magnetic beads. The remaining cells consisted of autofluorescent macrophages and CD4⁻CD8⁻ DC. For the Ag presentation and confocal microscopy experiments in which the presence of the magnetic beads was undesirable, the splenic DC preparation was labeled with mAb against CD11c, CD4, and CD8, and the CD4⁺ and CD8⁺ DC populations were purified by preparative FACS. Where indicated, the splenic DC were cultured after preparative cell sorting in RPMI 1640 supplemented with 10% FCS and 200 U/ml recombinant mouse GM-CSF (PeproTech, Rocky Hill, NJ).

Culture of DC from BMDC and D1DC precursors

BMDC and D1DC were generated in culture as described elsewhere (22). The results shown in Fig. 5⇓ were obtained with BMDC grown in the absence of IL-4; similar conclusions were obtained with cells grown in the presence of 1, 2, and 5 U/ml IL-4. Maturation of the BMDC and the D1DC was induced by addition of 1 μg/ml LPS (Sigma-Aldrich, St. Louis, MO) to the culture medium 20 h before cell harvesting.

Analytical and preparative flow cytometry

The Abs used for preparative or analytical FACS were hamster mAb N418 (anti-CD11c), rat mAb YTS (anti-CD8), rat mAb GK1.5 (anti-CD4), rat mAb M5/114 (anti-MHC II), and rat mAb GL1 (anti-CD86). All the Abs were purified and conjugated to fluorochromes (FITC, PE, Texas Red, or Cy5) in our laboratory. Samples were analyzed using a FACS II, FACStar^Plus, FACScan, or LSR instrument (BD Biosciences, San Jose, CA). Preparative FACS was performed using a Mo-Flo (Cytomation, Fort Collins, CO), a DiVa, or a FACStar^Plus (BD Biosciences).

Pulse-chase, immunoprecipitation, and Western blot

Pulse-chase analysis, immunoprecipitations with the mAb N22, and Western blot of N22 immunoprecipitates were conducted as described elsewhere (5, 22). Normalization of [³⁵S]met/cys incorporation was conducted by pipetting 10 μl of cell lysate on filter paper, washing the paper five times in 5% TCA (Sigma-Aldrich), and counting the amount of radioactivity precipitated on the paper in a scintillation counter. The amount of cell lysate used for the immunoprecipitation was then adjusted accordingly. For Western blot of cell lysates, cells were subjected to 5–7 cycles of vortexing and heating at 95°C in 2× SDS-PAGE loading buffer. The anti-heat shock protein 70 (Hsp70) mouse mAb N6 was the kind gift of Dr. R. Anderson (Peter MacCallum Cancer Research Institute, Melbourne, Australia). CyC detection and immunoprecipitation was conducted with anti-human CyC rabbit sera from Upstate Biotechnology (Lake Placid, NY) or from DAKO (Glostrup, Denmark); the two antisera gave identical results.

Immunofluorescence confocal microscopy (ICM)

ICM analysis was conducted exactly as described elsewhere (22). The primary Abs used were: biotinylated anti-MHC II mAb N22, purified or biotinylated rat mAb 1D4B (anti-Lamp), a rabbit serum specific for the cytosolic portion of MHC II I-Aα, and the anti-CyC rabbit sera described above. The secondary Abs were: streptavidin conjugated to Cy5 (Amersham, Castle Hill, Australia), goat anti-rat serum conjugated to FITC (Caltag Laboratories, Burlingame, CA), and sheep anti-rabbit conjugated to FITC (Amrad, Boronia, Australia). Where indicated, 4′,6′-diamidino-2-phenyindole (DAPI) (Sigma-Aldrich) was also included in the second incubation to label the nuclei. The confocal microscope used was a Leica STS SP2 (Leica, Wetzlar, Germany).

Ag presentation assays

The T cell hybridomas used in this study have been described elsewhere (28, 29); their specificity is indicated in Fig. 5⇓. For the assays, splenic DC were purified by preparative FACS and incubated for 20 h in RPMI 1640 supplemented with 10% FCS and 200 U/ml recombinant mouse GM-CSF (PeproTech) with 10⁵ hybridoma T cells in the presence of 0.1 mM OVA or hen egg lysozyme (HEL) (Sigma-Aldrich). The amount of IL-2 released by the hybridomas was determined using a standard proliferation assay with the IL-2-dependent cell line CTLL-2, or by ELISA.

CyC is differentially expressed among splenic DC populations

Mouse spleens contain in the steady state three distinct populations of DC: CD4⁺ DC, CD8⁺ DC, and CD4⁻CD8⁻ DC (Fig. 1⇓A) (24). After purification, these DC are highly endocytic, cannot activate naive T cells in vitro unless they are induced to mature (26), express low levels of surface MHC II and CD86 (Fig. 1⇓B) (30), and accumulate MHC II molecules in Lamp⁺ compartments (Fig. 1⇓C). These features define the splenic DC as immature (18). When put in culture, the three splenic DC undergo the phenotypic and functional changes that define DC maturation (18): they down-regulate endocytosis (data not shown), acquire the capacity to stimulate naive T cells (26), increase their surface expression of CD86 by 100-fold (Fig. 1⇓B) (30), and their MHC II molecules accumulate on the cell surface whereas their Lamp⁺ compartments cluster in a perinuclear region (Fig. 1⇓, B and C).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Splenic DC populations and maturation in culture. A, A DC preparation from mouse spleens was stained with anti-CD11c, anti-CD4, and anti-CD8 mAb, and analyzed by FACS. The plot shows the expression pattern of CD4 and CD8 in the CD11c⁺ cells (70–80% of the DC preparation, the rest being mostly macrophages). CD45RA⁺ “plasmacytoid DC” were excluded during the purification protocol. Mouse spleens contain three DC populations distinguished by their expression of CD4 and CD8: CD4⁺ DC, CD8⁺ DC, and CD4⁻CD8⁻ DC. The result is representative of multiple experiments. B, Splenic DC mature in culture, increasing the surface expression of MHC II and CD86 by 4- to 5- and 100-fold, respectively. The expression of MHC II and CD86 was analyzed by FACS in the splenic DC immediately after purification (thin continuous line, F) or after culture overnight (thick line, o/n). The dashed line corresponds to nonlabeled cells (background, B). Profiles for the total DC preparation are shown; the results for each of the three splenic DC populations were comparable. The result is representative of multiple experiments. C, MHC II molecules accumulate in the Lamp⁺ compartments of both CD8⁻ and CD8⁺ freshly isolated (immature) splenic DC, but on the plasma membrane in their cultured (mature) counterparts. DC were analyzed by ICM immediately after purification (Fresh) or after overnight culture using Abs for Lamp (green), MHC II (red) and the nuclear dye DAPI (blue). The result is representative of multiple experiments.

It has been proposed that the cysteine protease inhibitor CyC may play a major role in the control of MHC II expression in BMDC (23). We were interested in testing whether CyC plays such a role in vivo, so first we determined whether the expression of CyC in splenic DC was similar to that in other APC such as B cells. Indeed, Western blot analysis showed that CyC was expressed at much higher levels in freshly isolated spleen DC than in total splenocytes (Fig. 2⇓A). However, the expression of CyC varied drastically among the three splenic DC populations. The CD4⁺ and the CD4⁻CD8⁻ DC expressed very little, whereas the CD8⁺ DC contained virtually all the CyC detected in the unfractionated DC preparation (Fig. 2⇓A). The CD4⁺ and the CD4⁻CD8⁻ DC are located in different areas of the spleen compared to those where the CD8⁺ DC are found (31), so it could be that the CD8⁺ DC had simply captured CyC that had been secreted by other neighboring cells. Therefore we assessed, by metabolic labeling and immunoprecipitation, whether the CD8⁺ DC synthesized CyC themselves. The CD8⁺ DC synthesized 10-fold more CyC in 30 min than the CD4⁺ or the CD4⁻CD8⁻ DC (Fig. 2⇓B). We then assessed by Western blot whether the pattern of CyC expression was maintained after DC maturation. The relative difference in CyC content between the CD8⁻ and the CD8⁺ DC did not vary after overnight culture (Fig. 2⇓C). Therefore, differences in the expression levels of CyC among the three DC types were constitutive rather than linked to their maturation status.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

Expression pattern and subcellular localization of CyC in splenic DC. A, Cell lysates from total splenocytes, freshly purified splenic DC, and each of the three splenic DC populations, were loaded in a 12.5% SDS-PAGE and transferred to a polyvinylidene difluoride membrane. The membrane was cut transversally into three pieces, and each piece was analyzed by Western blot with Abs for Hsp70 (molecular mass 70 kDa, top), MHC II α-chain (molecular mass 35 kDa, middle), and CyC (molecular mass 15 kDa, bottom). For normalization purposes, three different amounts of cell lysate from each sample were loaded in adjacent lanes; the results shown correspond to the lanes in which the signals for Hsp70 were comparable, and are representative of multiple experiments. B, Freshly isolated CD4⁻CD8⁻ DC, CD4⁺ DC, and CD8⁺ DC were metabolically labeled with [³⁵S]met/cys for 30 min. The cells were lysed, and CyC was immunoprecipitated from equal amounts or radiolabeled cell lysate, quantitated by TCA precipitation. The immunoprecipitate was run in a 12.5% SDS-PAGE and visualyzed by autoradiography. Sequential immunoprecipitation of MHC II showed that the three DC populations had synthesized comparable amounts of this molecule (data not shown). The result is representative of multiple experiments. C, CD4⁺ and CD8⁺ splenic DC were purified by FACS and frozen immediately (Fresh) or after overnight culture (o/n). Cell lysates were prepared and analyzed by Western blot as in A. D, Freshly purified CD8⁺ DC were analyzed by ICM using Abs for CyC (green), and either MHC II (top, red), or Lamp (bottom, red). DAPI was included to label the nuclei (blue).

In immature BMDC, CyC is located in MHC II⁺Lamp⁺ compartments (22, 23) where it might interfere with MHC II peptide loading by inhibiting the cysteine protease Cat S (23). Similarly, in CD8⁺ spleen DC, CyC colocalized strongly with MHC II and Lamp (Fig. 2⇑C). Consistent with the Western blot and the immunoprecipitation experiments shown above, CyC was not detected by microscopy in the CD8⁻ DC (data not shown).

Because the subcellular distribution of MHC II, before and after maturation, was comparable between CD8⁺ and CD8⁻ DC, regardless of their expression of CyC, these results did not support a role for CyC in controlling MHC II expression. However, it could be that the CD8⁺ DC and the CD8⁻ DC used two different mechanisms of control, one dependent and the other independent of CyC (32). If so, Ii degradation and MHC II peptide loading might follow different kinetics in the CD8⁺ and the CD8⁻ DC. This possibility was assessed using biochemical methods.

Normal processing of Ii in CD8⁺ DC

If CyC was inactivating Cat S in the CD8⁺ DC, it would be expected that these DC would not break down Iip10, or would do it more slowly than the CyC-negative CD4⁺ DC (Fig. 3⇓A). To test this hypothesis, we compared Ii processing and MHC II peptide loading in both DC types by pulse-chase and immunoprecipitation. Freshly purified CD4⁺ and CD8⁺ DC were metabolically labeled with [³⁵S]Met/Cys for 30 min, and then chased for 60 and 240 min. The cells were lysed, and their MHC II molecules immunoprecipitated with the mAb N22. The immunoprecipitates were divided into two, incubated at room temperature (nonboiled) or at 98°C (boiled), respectively, and then analyzed by 12.5% SDS-PAGE (Fig. 3⇓B). This protocol allows following the conversion of αβ-Ii into αβ-Iip10, and of this into αβ-peptide (Fig. 3⇓A). The lanes loaded with the boiled samples show the composition of the MHC II molecules immunoprecipitated at each time point; those loaded with the nonboiled samples also allow to visualize the formation of peptide-loaded MHC II complexes (SDS-stable dimers), which are resistant to dissociation at room temperature (33).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

Both CD4⁺ and CD8⁺ DC process Ii and load their MHC II molecules with antigenic peptides in a Cat S-dependent manner. A, Schematic representation of the stepwise conversion of newly syntyhesized MHC II αβ-Ii complexes into αβ-peptide (1 ). The chaperone Ii (thick line) is first cleaved by one or more noncysteine proteases to generate Iip10, which is then processed by Cat S to yield CLIP. The CLIP is then exchanged for antigenic peptides, generating αβ-peptide complexes. CyC might act as an inhibitor of Cat S. The depicted complexes have a variable resistance to denaturation in SDS-PAGE at room temperature (32 ), which has been indicated by one to three plus signs (+) under each complex. The αβ-CLIP complex has been drawn in a lighter color to represent that unless the chaperone H-2DM is absent, it is quickly converted into αβ-peptide and therefore not detectable by pulse-chase analysis and immunoprecipitation (48 ,49 ). B, Analysis of Ii degradation and MHC II peptide loading in splenic CD4⁺ and CD8⁺ DC. Freshly isolated CD4⁺ and CD8⁺ DC were metabolically labeled for 30 min, washed, and then cultured in normal medium for 60 and 240 min. The MHC II molecules were immunoprecipitated from each sample with mAb N22, loaded in a 12.5% SDS-PAGE without boiling (NB) or after boiling (B), and visualized by autoradiography. The positions occupied by the free MHC II α and β polypeptides, full length Ii, the Ii spliced variant Iip41, the degradation intermediate Iip10, and the SDS-stable complexes αβIi, αβIip10, and αβ-peptide, are indicated on the left. C, Western blot analysis of MHC II immunoprecipitates obtained from freshly isolated splenic DC. Cat S^−/− CD4⁺ DC and wild-type CD4⁺ and CD8⁺ DC were lysed immediately after purification. Their MHC II molecules were immunoprecipitated with mAb N22, loaded in a 12.5% SDS-PAGE without boiling, and detected by Western blot using a rabbit serum that recognizes the cytoplasmic portion of the MHC II α-chain. Most of the MHC II molecules contained in the Cat S^−/− DC were associated with Iip10, whereas in both the CD4⁺ and the CD8⁺ wild-type DC, most of the MHC II molecules were forming αβ-peptide complexes. D, Both CD4⁺ and CD8⁺ splenic DC require Cat S to fully degrade Ii. The two DC populations were purified from the spleens of Cat S^−/− mice and analyzed by pulse-chase and immunoprecipitation as in B. Degradation of Iip10 was impaired in both DC types, causing the accumulation of αβ-Iip10 complexes, instead of αβ-peptide, after a 240-min chase. Two additional intermediates of Ii degradation can be observed above the position occupied by Iip10 (9 ,13 ).

The newly synthesized MHC II molecules were associated with full-length Ii, forming complexes that were mostly labile at room temperature (Fig. 3⇑B, pulse) (32). At the 240 min timepoint, most of Ii had been degraded (note that the carbohydrates contained in the MHC II α- and β-chains are modified in the Golgi, causing the β-chain precipitated at this timepoint to run at the same position as full-length Ii (32)), and most of the MHC II molecules were forming SDS-stable αβ-peptide complexes (Fig. 3⇑B, 240 min). The amount of αβ-peptide complexes generated in the CD4⁺ and the CD8⁺ DC were comparable. The Iip10 fragment, the substrate of Cat S, could be seen at the 60 min timepoint as a free polypeptide in the boiled lanes and as part of a relatively stable αβ-Iip10 complex in the nonboiled ones (Fig. 3⇑B, 60 min) (5, 12, 32). Both DC types generated similar amounts of αβ-Iip10 at the 60 min timepoint, and both converted this complex into αβ-peptide. Thus, the presence of CyC in MHC II⁺Lamp⁺ compartments did not impair Ii processing or MHC II peptide loading in the CD8⁺ DC.

The pulse-chase experiment was conducted in vitro immediately after DC purification. Because culturing the DC triggered their maturation (26), and mature DC may down-regulate the inhibition of Cat S by CyC (23), it could be argued that the newly synthesized MHC II molecules that we analyzed by pulse-chase encountered endocytic compartments with normal Cat S activity. To address this possibility, we analyzed the structure of the MHC II moleculescontained in the CD4⁺ and CD8⁺ DC immediately after purification. Because the purification protocol did not include incubations at 37°C, if the CD8⁺ DC were accumulating αβ-Iip10 complexes in vivo, such complexes would still be present in the fresh DC preparations. MHC II was immunoprecipitated with mAb N22 from the lysates of purified CD4⁺ and CD8⁺ DC and loaded in a 12.5% SDS-PAGE without boiling. An immunoprecipitate from CD4⁺ DC obtained from a Cat S^−/− mouse was also included as a control. The MHC II complexes were then detected by Western blot using a rabbit serum against the MHC II α-chain (Fig. 3⇑C). Most of the MHC II molecules immunoprecipitated from the Cat S^−/− DC were forming αβ-Iip10 complexes (7, 9, 13, 22), but the MHC II from both the CD4⁺ and the CD8⁺ DC of normal mice were forming αβ-peptide complexes. The same immunoprecipitates were also loaded in SDS-PAGE after boiling and analyzed by Western blot using Abs against the cytoplasmic portion of Ii, showing the presence of Iip10 in the immunoprecipitate obtained from the Cat S^−/− cells, but not in those obtained from normal CD4⁺ or CD8⁺ DC (data not shown). We conclude that degradation of Ii and MHC II peptide loading proceeds normally in CD8⁺ DC.

Cat S is necessary to degrade Iip10 in both CD4⁺ and CD8⁺ splenic DC

The results shown above suggested that the constitutive expression of CyC in the endosomal compartments of CD8⁺ DC did not inhibit Cat S activity, but it could be that CD8⁺ DC degraded Iip10 by a mechanism insensitive to CyC. Ii can be eliminated in early endosomal compartments independently of Cat S or other cysteine proteases (32), so such mechanism could be enhanced in the CD8⁺ DC. Alternatively, other proteases capable of cleaving Iip10, and not inhibited by CyC, could be selectively expressed in CD8⁺ DC. To address these possibilities, we analyzed Ii processing and MHC II peptide loading in splenic CD8⁻ and CD8⁺ DC from Cat S-deficient mice.

The spleens of Cat S^−/− mice contained similar amounts of CD4⁺ and CD8⁺ DC compared with normal mice (data not shown). Both DC types were analyzed by pulse-chase and immunoprecipitation of MHC II, as described above for normal DC. As shown in Fig. 3⇑D, maturation of MHC II was blocked at the αβ-Iip10 stage in both the CD4⁺ and CD8⁺ Cat S^−/− DC so that no MHC II-peptide complexes were generated after 240 min of chase. Therefore, the CD8⁺ DC depend on Cat S to degrade Iip10, just as CD4⁺ DC and splenocytes do.

MHC II expression and subcellular localization are not affected by the lack of CyC

The results above showed no correlation between CyC expression and control of MHC II Ag presentation in CD8⁺ and CD8⁻ splenic DC. However, this did not formally exclude that the complete absence of CyC might have an impact on MHC II expression or trafficking. Therefore, we analyzed DC from CyC-deficient mice (Fig. 4⇓A) (27). The subcellular distribution of MHC II in freshly isolated (immature) and cultured (mature) CD8⁺ DC was unaffected by the lack of CyC (Fig. 4⇓B).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

Analysis of CyC-deficient CD8⁺ DC. A, Freshly isolated CD8⁺ splenic DC from wild-type or CyC^−/− mice were metabolically labeled for 60 min. The cells were lysed, and CyC was immunoprecipitated from equal amounts or radiolabeled cell lysate, quantitated by TCA precipitation. MHC II was immunoprecipitated sequentially from the same lysates, showing that both cell samples had synthesized comparable amounts of this molecule. The immunoprecipitates were analyzed by 12.5% SDS-PAGE and autoradiography. B, CyC-deficient CD8⁺ DC were analyzed by ICM immediately after purification (Fresh, top) or after overnight culture (o/n, bottom) as in Fig. 1⇑ (except that DAPI was not included). C, The indicated number of purified CD8⁺ DC from normal or CyC^−/− mice were used in assays of Ag presentation to the hybridomas shown. The specificity of each hybridoma is indicated. Note that the Ags used were OVA or HEL protein, not the antigenic peptides.

The role of CyC in the CD8⁺ DC could be to interfere with processing of exogenous Ags rather than Ii degradation. To examine this possibility, we compared the capacity of normal and CyC-deficient CD8⁺ DC to present exogenous OVA and hen egg lysozyme (HEL) to T cell hybridomas. The lack of CyC did not affect the Ag presentation efficiency of the CD8⁺ DC (Fig. 4⇑C). These results exclude a role for CyC in controlling MHC II Ag presentation in splenic DC.

The study that suggested a role for CyC in controlling MHC II Ag presentation was performed on BMDC generated in vitro (23), so it could be that CyC only plays a role in these DC or other DC types obtained by in vitro culture. Therefore, we analyzed BMDC and D1DC (21) from CyC^−/− and wild-type mice. The surface expression of MHC II and CD86, before and after maturation, were not affected by the lack of CyC in either BMDC or D1DC (Fig. 5⇓, A and C, and data not shown). The MHC II molecules of immature CyC^−/− BMDC and D1DC accumulated in Lamp⁺ compartments, as they did in wild-type cells, and they redistributed to the cell surface after maturation (Fig. 5⇓, A and B).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5.

CyC does not control MHC II expression or peptide loading in BMDC or D1DC. A, BMDC were generated from wild-type or CyC-deficient animals and analyzed by FACS (top) or ICM (bottom) 5 days after initiation of the cultures. The samples labeled LPS (mature) received LPS at a final concentration of 1 μg/ml 20 h before cell harvesting. For the FACS analysis, cells were labeled with Abs for CD11c, CD86, and MHC II; the results shown were gated on live (propidium iodide⁻) CD11c⁺ cells. The numbers indicate the percentage of MHC II^highCD86^high cells. For the ICM analysis, CD11c⁺ cells were sorted by preparative FACS, plated on microscopy coverlips, and labeled with anti-MHC II mAb N22 (red), anti-Lamp1 mAb 1D4B (green) and the nuclear dye DAPI (blue). Only the merged images are shown. B, D1DC were generated from the spleens of wild-type or CyC-deficient animals and analyzed by FACS (data not shown) or ICM. The cells labeled LPS (mature) were treated with 1 μg/ml LPS for 20 h before harvesting. Cells were plated directly on microscopy coverslips (preparative cell sorting was not necessary because the D1DC cultures were over 90% CD11c⁺ (data not shown)). ICM analysis was performed with an anti-MHC II α rabbit serum (green), anti-Lamp 1 mAb 1D4B (red), and the nuclear dye DAPI (blue). Only the merged images are shown. The micrograph of the mature cells was taken at lower magnification than the control. C, Wild-type, CyC-deficient, and Cat S-deficient CD11c⁺ BMDC were purified by preparative FACS. Analysis of the surface expression of CD86 and MHC II confirmed that virtually 100% of all the DC were immature (right histograms, the dashed line corresponds to the background level in the wild-type BMDC). The MHC II molecules contained in each cell preparation were immunoprecipitated and run in 12.5% SDS-PAGE before (top) or after boiling (middle and bottom). The immunoprecipitates were then analyzed by Western blot with Abs specific for I-Aα (top and middle), and the N terminus of the Ii (bottom). The regions containing the SDS-stable αβ-peptide complexes (around 50 kDa, top), the free I-Aα-chain (35 kDa, middle), and Iip10 (10 kDa, bottom) are shown.

Finally, we determined whether the lack of expression of CyC had an impact on the generation of peptide-loaded MHC II complexes in BMDC. We immunoprecipitated MHC II from immature BMDC grown from normal, CyC-deficient, or Cat S-deficient mice (Fig. 5⇑C). The immunoprecipitates were incubated at room temperature or 95°C, and analyzed by SDS-PAGE followed by Western blot (Fig. 5⇑C). As expected, the MHC II molecules of the Cat S^−/− BMDC coprecipitated with Iip10 and were forming few αβ-peptide SDS-stable complexes. In contrast, most of the MHC II molecules contained in the normal and the CyC^−/− BMDC were forming αβ-peptide complexes and were not associated with Iip10. These results demonstrate that as in immature D1DC (22) and lymphoid organ DC (Fig. 3⇑C), Cat S is active in immature BMDC, enabling the constitutive generation of MHC II molecules loaded with antigenic peptides (22, 34, 35). Therefore, it is not surprising that the CyC deficiency did not have a deleterious effect on Iip10 degradation or MHC II-peptide complex formation. The BMDC shown in these experiments were generated in culture medium supplemented with GM-CSF, but without IL-4; similar results were obtained with BMDC generated in culture medium containing GM-CSF plus 1, 2, or 5 U/ml IL-4 (data not shown).