Immunization of mice with GRP94, the endoplasmic reticulum (ER) Hsp90, elicits cytotoxic T lymphocyte (CTL) responses to chaperone-bound, source cell-derived peptides. Elicitation of a CTL response requires that GRP94-associated peptides be transferred onto major histocompatability complex (MHC) class I molecules, a process that is postulated to accompany GRP94 internalization by antigen presenting cells, such as macrophages (MΦ) and dendritic cells (DC). In studies of GRP94 uptake in elicited MΦ, we report that MΦ display specific cell surface binding of GRP94, and that surface-bound GRP94 can be internalized via receptor mediated endocytosis. GRP94 internalized by this pathway co-localized predominately with transferrin-positive early endosomes. At time periods of up to 20 minutes, little trafficking of GRP94 to the lysosomal compartment was observed. When GRP94 was present in the medium, and thus accessible to both receptor-mediated and fluid phase internalization pathways, internalization was modestly inhibited in the presence of yeast mannan, a competitive inhibitor of mannose/fucose receptor activity, and substantially inhibited by dimethylamiloride, an inhibitor of macropinocytosis. GRP94 internalized via macropinocytosis did not display prominent co-staining with the lysosomal marker LAMP-2. These data identify multiple pathways of GRP94 internalization and indicate that receptor-dependent uptake of GRP94 is not dependent upon its high mannose oligosaccharide moiety. Most significantly, these data demonstrate the existence of cell surface receptor(s), apparently unique to antigen presenting cells, that function in the binding and internalization of the ER chaperone GRP94.

In antigen presenting cells, such as macrophages (MΦ) and dendritic cells (DC), the endoplasmic reticulum (ER) chaperone GRP94 (gp96) can access the endogenous pathway for antigen presentation and thereby elicit cytotoxic T cell (CTL) responses to chaperone-bound peptides (Arnold et al., 1995; Nicchitta, 1998; Srivastava et al., 1994, 1998). The endogenous pathway for antigen presentation functions to allow immune recognition of intracellular proteins (Heemels and Ploegh, 1995; York and Rock, 1996). In this pathway, cytosolic peptides, generated primarily through proteasome-mediated degradation of intracellular proteins, are transported via the transporter associated with antigen presentation (TAP) into the lumen of the ER. In a series of reactions requiring TAP, tapasin and the ER chaperones calreticulin and calnexin, appropriate peptides are loaded onto nascent major histocompatability complex (MHC) class I/β2microglobulin (β2m) molecules (Ortmann et al., 1997; Sadasivan et al., 1996; Solheim et al., 1997). Formation of the structurally stable class I/β2m/peptide ternary complex allows its exit from the ER, and trafficking to the cell surface. Given the established subcellular compartmentation of the reactions associated with peptide loading onto class I molecules, it would appear that elicitation of a CTL response by GRP94 requires the internalization and subsequent trafficking of either GRP94-peptide complexes or the GRP94-derived peptides, to a subcellular compartment in which peptide loading or exchange onto class I molecules can occur.

An essential role for phagocytic cells in the generation of GRP94 elicited CTL responses was first reported by Udono et al. (1994), who observed that depletion of such cells severely compromised the immunogenicity of GRP94-peptide complexes. A role for phagocytic cells in this process was subsequently recapitulated in vitro, in experiments demonstrating that murine macrophages pulsed with GRP94 purified from cells expressing the nucleocapsid protein of the vesicular stomatitis virus (VSV), presented VSV-derived peptides on cell surface MHC class I molecules (Suto and Srivastava, 1995). Similarly, immature, bone marrow-derived dendritic cells, pulsed with GRP94 purified from E.G7-OVA cells, a murine thymoma cell line transfected for the gene encoding chicken ovalbumin, were recognized for lysis by an OVA-specific CTL line, and thus also displayed chaperone derived peptides on cell surface class I molecules (Nair et al. 1999). At present, it is not known whether phagocytic cells, such as macrophages and dendritic cells, are unique in their ability to internalize GRP94.

The capacity of exogenous GRP94 to elicit a CTL response is a noteworthy example of an exogenous antigen gaining access to the endogenous antigen processing pathway, a phenomenon commonly referred to as cross-priming, or cross-presentation (Bevan, 1995; Brossart and Bevan, 1997; Carbone et al., 1998; Harding, 1996; Matzinger and Bevan, 1977; Moore et al., 1988; Watts, 1997). Cross-priming is critical for CTL surveillance of peripheral antigens and represents the primary physiological pathway for antigen presenting cells, in the absence of infection, to process exogenous antigens for MHC class I expression (Bevan, 1995, 1987; Carbone et al., 1998; Watts, 1997). The cellular trafficking pathway(s) responsible for directing exogenous antigens to the endogenous pathway have not been defined, although a number of pathways have been postulated. Prominent among these models are two themes: a vacuole to cytosol transport process (Harding, 1996; Kovacsovics-Bankowski and Rock, 1995; Reis e Sousa and Germain, 1995; Srivastava et al., 1994) and/or vacuolar proteolytic degradation followed by either discharge of peptides to the extracellular space, and binding to empty class I molecules, or binding of the peptides to class I molecules during endocytic internalization and recycling of class I molecules (Harding, 1996; Harding and Song, 1994; Pfeifer et al., 1993). With regard to ER resident chaperones such as GRP94, additional pathways may contribute to the intracellular trafficking of internalized chaperone (Nicchitta, 1998). For example, the capacity of cells to capture proteins bearing a C-terminal KDEL ER-retention in the trans-Golgi network, and traffic them in a retrograde manner to the ER lumen (Miesenböck and Rothman, 1995) could be exploited by antigen presenting cells to deliver exogenous GRP94 to the ER. This pathway is utilized by various toxins, such as Shiga toxin and ricin, to gain access to the cytosol (Sandvig and Van Duers, 1996). At present, however, neither the mechanism of ER chaperone uptake nor the cellular trafficking pathways accessed by internalized ER chaperones in antigen presenting cells are known.

We have investigated the pathways for internalization of GRP94 into murine MΦ and report that MΦ display receptor mediated and fluid phase uptake pathways for GRP94. Although GRP94 bears high mannose oligosccharides, internalization of GRP94 by the receptor mediated pathway exhibits low sensitivity to mannan, and thus is unlikely to occur via the mannose receptor. Receptor-mediated uptake yields localization to the endosomal compartment, with little progression to the lysosomes at time points of up to 20 minutes. Internalization of GRP94 through macropinocytosis was also readily evident, and as with receptor-mediated uptake, did not yield prominent co-localization with lysosomal markers. The identification of a receptor-mediated GRP94 uptake pathway in antigen presenting cells indicates that in such cells, the ER chaperone GRP94 can access multiple compartments of the secretory pathway, and provides mechanistic insight into the cellular basis for chaperone-based cross-presentation.

Media and reagents

All tissue culture was performed in RPMI 1640, supplemented with 25 mM Hepes, 2 mM L-glutamine, 1% non-essential amino acids, 1 mM pyruvate, 50 μM 2-mercaptoethanol and 5% fetal bovine serum (Gibco BRL, Grand Island, NY). GRP94 was purified from porcine pancreas as described by Wearsch and Nicchitta (1996). GRP94 and BSA (Sigma Chemical) were labeled with fluorescein isothiocyanate (FITC) (Molecular Probes Inc., Eugene, OR) to a ratio of 2-3 mol FITC:mol GRP94. Labeling was performed according to the manufacturer’s instructions. FITC-GRP94 was used at a concentration of 80 μg/ml (ca. 440 nM) unless otherwise specified. FITC-mannose-bovine serum albumin (FITC:mBSA) (16.1 mol mannose:2.5 mol FITC: mol albumin), yeast mannan and 5-(N,N-dimethyl) amiloride (DMA) were purchased from Sigma (St Louis, MO). FITC-mBSA was reconstituted from a lyophilized powder to 1 mg/ml with HEPES-buffered saline (HBS). Mannan was reconstituted from a lyophilized powder to a 50 mg/ml stock with HBS and used at a final concentration of 1 mg/ml. DMA was reconstituted from a lyophilized powder to a 10 mM stock with DMSO and used at final concentration of 100 μM. Texas Red (TR) labeled dextran (TR-DX) (Mr=40,000) and TR-transferrin (TR-Tfn) were purchased from Molecular Probes, Inc. (Eugene, OH). TR-DX was reconstituted with water from a lyophilized powder to yield a 25 mg/ml solution. TR-DX was used at 1 mg/ml final concentration. TR-Tfn was reconstituted from a lyophilized powder to a 5 mg/ml stock with distilled water supplemented with 0.005% thimerosal and used at 100 μg/ml final concentration. Anti-LAMP-2 rat monoclonal antibody (clone ABL-93) tissue culture supernatant was the generous gift of Dr Chris Norbury (NIH, Bethesda, MD). Lissamine rhodamine (LR) goat anti-rat secondary antibody was obtained from Jackson ImmunoResearch Laboratories, Inc. (West Grove, PA). Phycoerythrin-labeled anti-F4/80 antibody (MΦ marker) was obtained from Caltag Laboratories (Burlingame, CA) and used at a concentration of 5 μg/ml.

Cells

Peritoneal elicited macrophages (MΦ) were used in all experiments. Macrophages were elicted in C57BL/6 mice (Jackson Laboratory, Bar Harbor, ME, or Charles River, Wilmington, MA) by intraperitioneal injection of 1 ml of a 4.2% (w:v) solution of Brewer’s thioglycollate broth. Mice were sacrificed 4-5 days post-injection and MΦ obtained by peritoneal lavage with DMEM, 10% FCS. Macrophages were selected by adherence for 1 hour at 37°C on 12-well plates (Corning Glass Works, Corning NY) for flow cytometry experiments or 18 mm #1 glass coverslips (VWR Scientific, Inc., Media, PA) for laser scanning confocal microscopy studies.

Confocal microscopy: internalization studies

For studies on receptor-mediated internalization, coverslips with adherent MΦ were placed in 12-well plates on ice, washed 4× with ice-cold RPMI and incubated at 4°C for 30 minutes in 500 μl complete RPMI containing FITC-GRP94 (80 μg/ml). To remove free FITC-GRP94, cells were subsequently washed 4× with ice-cold RPMI. After the final wash, the medium was exchanged with 0.5 ml prewarmed complete RPMI and plates incubated at 37°C. To investigate fluid phase and receptor mediated entry, FITC-GRP94 (80 μg/ml) was present in the medium throughout the time course of the experiment. Where indicated, the medium was supplemented with TR-Tfn, a marker of cell surface receptors and early endosomes or TR-DX, a fluid phase marker. At the indicated time points, coverslips were placed into ice-cold Hepes binding buffer (HBB: 0.8% NaCl, 25 mM K-Hepes (pH 7.2) containing 1 mM CaCl2 and 0.5 mM MgCl2) to block transport. Cells were rapidly washed 2× in HBB and fixed for 15 minutes on ice in 4% paraformaldehdye/PBS. The fixation solution was aspirated, coverslips washed twice with ice-cold PBS and unreacted paraformaldehyde quenched by incubation for 5 minutes with 0.05 M NH4Cl at 4°C. Coverslips were again washed three times with PBS and mounted onto slides in mounting medium (10% PBS, 90% glycerol, 1 mg/ml phenylenediamine).

For lysosomal staining, the cells were fixed as described above and permeabilized at 4°C in PBS supplemented with 0.1% saponin for 15 minutes. Permeabilized cells were washed in room temperature (RT) PBS and blocked with 1% BSA for 30 minutes at RT. All subsequent steps were performed at RT. Cells were stained with anti-LAMP-2 mAb supernatant, diluted 1:25 in complete RPMI, for 1 hour, washed 4× 5 minutes in PBS, stained with the LR-secondary antibody (1:500) for one hour at RT, washed and mounted as described above. All images were obtained on a Zeiss LSM 410 laser scanning confocal microscope (Thornwood, NY). Digital image capture was performed with Zeiss LSM version 3.95 software. All image size and contrast adjustments were performed with PhotoShop (version 4) software (Adobe Software, Palo Alto, CA).

Quantitation of endocytosis by flow cytometry

MΦ on 12-well plates (1×106 cells/well) were washed 4 times with 37°C PBS and incubated with 0.5 ml of complete RPMI containing either FITC-mBSA or FITC-GRP94. Where indicated, the medium contained DMA or mannan along with the fluorescent probe. Uptake of the probe was allowed to proceed for 30 minutes at 37°C. At the end of 30 minutes, the plates were place on ice and internalization was stopped by aspiration of the medium and rapid washing of the cells 4 times with ice-cold HBB. Cells were then immediately fixed in 0.6 ml HBS containing 1% paraformaldehyde (PFA) and removed from the plates with a cell scraper (Costar Corporation, Cambridge, MA). Cells were stored in fixative at 4°C and fluorescence quantitated by flow cytometry (Becton Dickinson, San Jose, CA). The provided fluorescence values represent the geometric mean fluorescence, an average fluorescence weighted to the number of cells emitting at each fluorescence intensity in the range of 100-105 fluorescence units. Cell-derived autofluorescence was subtracted from all values.

Receptor-mediated internalization of GRP94

Previously, it was demonstrated that peritoneal Mφ can mediate the transfer of peptides, derived from soluble, extracellular GRP94, into the endogenous (MHC class I) antigen presentation pathway (Suto and Srivastava, 1995). To investigate the pathway(s) for internalization of GRP94 into murine macrophages (Mφ), GRP94, purified as described by Wearsch and Nicchitta, (1996), was labeled with FITC and cell surface binding and internalization of FITC-GRP94 investigated by laser scanning confocal microscopy. To study cell surface binding, adherent Mφ were incubated for 60 minutes at 4°C in medium supplemented with 50 μg/ml FITC-GRP94 (ca. 270 nM), the cells subjected to repeated washes with culture medium lacking FITC-GRP94, and the staining pattern determined (Fig. 1A). In the images presented in Fig. 1A, cells were fixed and subsequently stained with antibodies directed against the lysosomal marker LAMP-2. Clearly evident in these images is the presence of cell surface staining, as detailed in a cross-sectional focal plane. Equally evident is the observation that well-defined cell surface FITC-GRP94 staining of adherent Mφ displayed cell-to-cell variability. The biological basis for this variability remains to be defined, although it is presumed to relate to previous observations indicating that in Mφ, access of exogenous antigens to the class I processing pathway may be limited to a subpopulation of MΦ (Kovacsovics-Bankowski and Rock, 1995; Suto and Srivastava, 1995).

Fig. 1.

Receptor-mediated endocytosis of GRP94 in peritoneal macrophages. Adherent peritoneal macrophages were incubated for 30 minutes at 4°C in the presence of 50 μg/ml FITC-GRP94, washed extensively and either fixed (A) or warmed to 37°C to allow internalization (B and C). In the images depicted in B, Texas Red-transferrin (100 μg/ml) was present in the medium during the 10 minute incubation period. In C, uptake was allowed to proceed for 20 minutes at 37°C, the cells washed and fixed, and, subsequently, the lysosomal compartment stained with antisera directed against LAMP-2. All images are representative laser scanning confocal microscopy sections obtained at a single focal plane.

Fig. 1.

Receptor-mediated endocytosis of GRP94 in peritoneal macrophages. Adherent peritoneal macrophages were incubated for 30 minutes at 4°C in the presence of 50 μg/ml FITC-GRP94, washed extensively and either fixed (A) or warmed to 37°C to allow internalization (B and C). In the images depicted in B, Texas Red-transferrin (100 μg/ml) was present in the medium during the 10 minute incubation period. In C, uptake was allowed to proceed for 20 minutes at 37°C, the cells washed and fixed, and, subsequently, the lysosomal compartment stained with antisera directed against LAMP-2. All images are representative laser scanning confocal microscopy sections obtained at a single focal plane.

To determine the early trafficking fate of cell surface-associated GRP94, FITC-GRP94 was allowed to bind to adherent Mφ for 30 minutes on ice, the cells extensively washed, and the incubations continued at 37°C in medium supplemented with 100 μg/ml of the endosomal marker transferrin, conjugated to Texas Red (TR-Tfn). When warmed to 37°C, surface bound GRP94 was observed to gain entry into the cell (Fig. 1B). After a 10 minute internalization period (37°C), punctate FITC-GRP94 staining was observed within the cells, predominately localized in a region proximal to the cell surface. This region coincided with that stained by TR-Tfn and thus FITC-GRP94 staining (green), TR-Tfn staining (red) and co-staining (yellow) were observed. After a more extended interval at 37°C (20 minutes), Mφ internalized FITC-GRP94 staining remained decidedly punctate and again, predominately localized proximal to the cell surface (Fig. 1C). In the image depicted in Fig. 1C, cells were also co-stained for the lysosomal marker, LAMP-2 (red). LAMP-2 staining was observed throughout the cell (compare Fig. 1A and C) and was distinct from that observed for FITC-GRP94 (Fig. 1C). Note, however, that some apparent co-staining (yellow) can be observed, suggesting a small degree of trafficking of the internalized FITC-GRP94 to lysosomes at the 20 minute time point. Extensive optical sectioning studies, however, are necessary to determine whether this small degree of co-staining is due to a proximal location of separate compartments, or is representative of a true co-localization.

Specificity of GRP94 binding to murine MΦ

The cell population used in these studies is derived from a peritoneal lavage and subjected to a single round of adherence selection. Although substantially enriched in MΦ, this population of cells is not homogeneous. To determine whether cell surface binding of GRP94 could be attributed to the MΦ component of the cell population, a two channel flow cytometry study was performed in which cell surface binding of GRP94 and expression of the MΦ marker F4/80 were coincidentally analyzed. In viewing the FACS data presented in Fig. 2A, it is evident that the vast majority of cells displaying cell surface binding of fluorescein-labeled GRP94 are also F4/80 positive, and thus MΦ? As expected from the data depicted in Fig. 1A, also present in the MΦ-enriched cell population were small numbers of cells that appeared to express F40/80, yet not bind FITC-GRP94, cells that bound FITC-GRP94, and did not express F4/80, and cells that neither bound FITC-GRP94, nor expressed F4/80.

Fig. 2.

GRP94 binding to the cell surface of peritoneal macrophages is specific. (A) Peritoneal macrophages were selected by adherence for 60 minutes at 37°C and stained at 4°C with FITC-GRP94 and phycoerythrin labeled anti-F4/80. The cells were subsequently washed, fixed, recovered by scraping and analyzed by two-channel flow cytometry. (B) Subconfluent monolayers of adherence selected peritoneal macropahges, the macrophage-derived cell line RAW 264.7, HepG2, BRL, COS-7 and CHO cells were labeled with 50 μg/ml FITC-GRP94 for 60 minutes at 4°C. Cells were extensively washed, fixed, recovered by scraping, and the presence of surface-bound GRP94 assayed by flow cytometry. The geometric mean fluorescence obtained in analysis of approximately identical quantities of cells is depicted. Background fluorescence was determined for each cell type, and subtracted from the determined values. (C) Cell surface binding of FITC-GRP94 and FITC-bovine serum albumin to peritoneal macrophages was performed as described in the legend to B. Conjugate binding was assayed by flow cytometry and data derived from both conjugates, and cell autofluorescence is depicted.

Fig. 2.

GRP94 binding to the cell surface of peritoneal macrophages is specific. (A) Peritoneal macrophages were selected by adherence for 60 minutes at 37°C and stained at 4°C with FITC-GRP94 and phycoerythrin labeled anti-F4/80. The cells were subsequently washed, fixed, recovered by scraping and analyzed by two-channel flow cytometry. (B) Subconfluent monolayers of adherence selected peritoneal macropahges, the macrophage-derived cell line RAW 264.7, HepG2, BRL, COS-7 and CHO cells were labeled with 50 μg/ml FITC-GRP94 for 60 minutes at 4°C. Cells were extensively washed, fixed, recovered by scraping, and the presence of surface-bound GRP94 assayed by flow cytometry. The geometric mean fluorescence obtained in analysis of approximately identical quantities of cells is depicted. Background fluorescence was determined for each cell type, and subtracted from the determined values. (C) Cell surface binding of FITC-GRP94 and FITC-bovine serum albumin to peritoneal macrophages was performed as described in the legend to B. Conjugate binding was assayed by flow cytometry and data derived from both conjugates, and cell autofluorescence is depicted.

Questions regarding the cell specificity of GRP94 surface binding were further investigated in a flow cytometry-based analysis of FITC-GRP94 binding to elicited murine MΦ, the macrophage-derived cell line RAW 264.7, the human hepatic cell line HepG2, BRL (Buffalo rat liver), COS-7, and CHO cells. The results of this study are shown in Fig. 2B. Comparison of the geometric mean fluorescence values from this experiment indicate that significant cell surface binding of FITC-GRP94 was observed in elicited murine MΦ, and to a lesser degree in RAW 264.7 cells. No binding was observed in HepG2, BRL, COS-7 or CHO cells. These data support the conclusion that cell surface binding of FITC-GRP94 to MΦ is indicative of interaction with a cell surface receptor(s), and not a circumstantial interaction, as might be predicted to arise through non-specific interactions between GRP94 and cell surface components.

To further investigate the specificity of this binding interaction, additional experiments were performed to determine the contribution of the conjugate fluorescein moiety to the observed binding. In these experiments, MΦ surface binding of FITC-labeled GRP94 and FITC-labeled bovine serum albumin (BSA) were compared. To control for effects of fluorophore coupling stoichiometry, fluorescein conjugates of GRP94 and BSA were prepared at near-identical coupling stoichiometries (ca. 2 mol fluorescein: mol protein). As shown in Fig. 2C, whereas cell surface binding of FITC-GRP94 was clearly evident, the fluorescent signal obtained in the FITC-BSA samples coincided with autofluorescence. In summary, the data presented in Fig. 2A-C indicate that elicited murine MΦ display cell surface binding of FITC-GRP94 and that this binding activity is cell-type specific, and does not arise through fluorescein-dependent interactions with a MΦ cell surface component.

Receptor-mediated and fluid phase pathways for internalization of GRP94 into pMφ

The experiments depicted in Fig. 1 identify a receptor mediated pathway for internalization of GRP94 into Mφ. As there exist multiple pathways for the internalization of both receptor-bound and fluid phase ligands, studies were performed to further characterize the contribution of receptor-mediated and fluid phase endocytic pathways to the internalization of GRP94. Since GRP94 contains a single, high mannose oligosaccharide (Wearsch and Nicchitta, 1996; Welch et al., 1983), it would be expected to serve as a ligand for the mannose/fucose receptor (Stahl et al., 1980). To characterize the contribution of the mannose receptor to the internalization of GRP94, FITC-GRP94 uptake was assayed as a function of FITC-GRP94 concentration in the presence or absence of yeast mannan, an established inhibitor of the mannose/fucose receptor. In these experiments, uptake was allowed to proceed for 60 minutes at 37°C, the cells washed, and uptake determined by flow cytometry. As a positive control, parallel studies were performed with FITC-mannose BSA (FITC-manBSA), a ligand known to be internalized by the mannose/fucose receptor (Sallusto et al., 1995; Stahl et al., 1980). The results of these studies are depicted in Fig. 3. For both FITC-GRP94 and FITC-manBSA, uptake by both mannan sensitive and mannan insensitive pathways could be clearly discerned. Subtraction of the uptake observed in the presence of mannan from that observed in its absence yielded the mannan-sensitive internalization component (Fig. 3). For both FITC-GRP94 and FITC-manBSA, this component of the internalization process appeared saturable, with maximum uptake being observed at approximately 440 nM (80 μg/ml) FITC-GRP94 and 100 μg/ml FITC-manBSA.

Fig. 3.

Identification of mannan-sensitive and mannan-insensitive pathways for uptake of GRP94. Uptake of FITC-GRP94 or FITC-mannose BSA by adherent peritoneal macrophages was assayed as a function of FITC-GRP94 or FITC-mannose BSA concentration. Uptake studies were performed for 30 minutes at 37°C, the cells washed, fixed, retrieved with a cell scraper, and cell-associated fluorescence determined by flow cytometry. Where indicated, yeast mannan was present at a final concentration of 500 μg/ml. To determine mannan-sensitive uptake, the geometric mean fluorescence values obtained in the presence of mannan were subtracted from those obtained in the absence of mannan. All values were corrected for cell-derived autofluorescence obtained in the absence of labeled ligand. Quantitation of these data is presented in Table 1.

Fig. 3.

Identification of mannan-sensitive and mannan-insensitive pathways for uptake of GRP94. Uptake of FITC-GRP94 or FITC-mannose BSA by adherent peritoneal macrophages was assayed as a function of FITC-GRP94 or FITC-mannose BSA concentration. Uptake studies were performed for 30 minutes at 37°C, the cells washed, fixed, retrieved with a cell scraper, and cell-associated fluorescence determined by flow cytometry. Where indicated, yeast mannan was present at a final concentration of 500 μg/ml. To determine mannan-sensitive uptake, the geometric mean fluorescence values obtained in the presence of mannan were subtracted from those obtained in the absence of mannan. All values were corrected for cell-derived autofluorescence obtained in the absence of labeled ligand. Quantitation of these data is presented in Table 1.

Macropinocytosis, a fluid-phase uptake pathway constitutively present in Mφ, dendritic cells, and growth factor-stimulated epithelial cells (Swanson, 1989; Swanson and Watts, 1995; Watts, 1997), can deliver soluble antigens to both the MHC class I and class II antigen presentation pathways (Brossart and Bevan, 1997; Norbury et al., 1995; Sallusto et al., 1995; Swanson, 1989; West et al., 1989). As macropinocytotic uptake can be distinguished from clathrin-dependent uptake pathways by its sensitivity to amiloride (West et al., 1989), the contribution of amiloride-sensitive macropinocytosis to FITC-GRP94 uptake was assayed. In the experiment depicted in Fig. 4, either FITC-manBSA (A,B,C) or FITC-GRP94 (D,E,F) was continuously present in the medium, and internalization into pMφ assayed at 37°C, for 60 minutes, in the presence or absence of mannan (A and D), dimethylamiloride (B and E) or both (C and F). Under the described assay conditions, internalization of both FITC-GRP94 and FITC-mannose BSA displayed a mannan-sensitive component. In this experiment, mannan-sensitive uptake represented 66% of the total FITC-mannose BSA, and 24% of the total FITC-GRP94, internalized. Over the course of repeated experiments, the relative contribution of the mannan-sensitive pathway to FITC-mannose BSA and FITC-GRP94 uptake into Mφ has proven variable, with FITC-mannose BSA uptake being inhibited by 60-70%, and FITC-GRP94 uptake by 20-35% in the presence of excess mannan. As shown in B and E, FITC-mannose-BSA uptake was insensitive to the presence of the macropinocytosis inhibitor DMA (100 μM), whereas FITC-GRP94 uptake was markedly suppressed. When internalization was assayed in the combined presence of both inhibitors, the effects were approximately additive (C and F; see Table 1). In summary, these data confirm that mannose-receptor dependent uptake does not represent the primary pathway for the internalization of GRP94 into MΦ, and, as expected, macropinocytosis can serve as a prominent GRP94 uptake pathway in macrophages.

Table 1.

Quantitative analysis of the effects of yeast mannan and dimethylamiloride on the internalization of FITC-mannose BSA and FITC-GRP94 by murine macrophages

Quantitative analysis of the effects of yeast mannan and dimethylamiloride on the internalization of FITC-mannose BSA and FITC-GRP94 by murine macrophages
Quantitative analysis of the effects of yeast mannan and dimethylamiloride on the internalization of FITC-mannose BSA and FITC-GRP94 by murine macrophages
Fig. 4.

Internalization of FITC-mannose BSA and FITC-GRP94 by murine macrophages: inhibition by mannan and dimethylamiloride. Adherent peritoneal macrophages were incubated in the presence of 80 μg/ml FITC-mannose BSA (A-C) or FITC-GRP94 (D-F) for 30 minutes at 37°C. As indicated, incubations were performed in the presence and absence of yeast mannan (500 μg/ml) (A and D), 100 μM dimethylamiloride (B and E) or both (C and F). At the completion of the uptake period, cells were fixed, scraped, and fluorescence determined by flow cytometry. Autofluorescence is depicted in A and D, and was subtracted from geometric mean fluorescence values in all determinations.

Fig. 4.

Internalization of FITC-mannose BSA and FITC-GRP94 by murine macrophages: inhibition by mannan and dimethylamiloride. Adherent peritoneal macrophages were incubated in the presence of 80 μg/ml FITC-mannose BSA (A-C) or FITC-GRP94 (D-F) for 30 minutes at 37°C. As indicated, incubations were performed in the presence and absence of yeast mannan (500 μg/ml) (A and D), 100 μM dimethylamiloride (B and E) or both (C and F). At the completion of the uptake period, cells were fixed, scraped, and fluorescence determined by flow cytometry. Autofluorescence is depicted in A and D, and was subtracted from geometric mean fluorescence values in all determinations.

Subcellular localization of FITC-GRP94 internalized by fluid-phase and receptor-mediated pathways

Confocal microscopy studies were performed to characterize the intracellular localization of FITC-GRP94 internalized by both fluid-phase and receptor-mediated pathways (Fig. 5). As previously noted, incubation of Mφ with FITC-GRP94 at 4°C yielded cell surface staining (Fig. 5A). Upon warming to 37°C in FITC-GRP94 supplemented medium, internalization began, yielding, after 5 minutes, abundant labeling of vesicular compartments in close apposition to the cell surface (B). In the image depicted in C, uptake studies were performed in the presence of FITC-GRP94 (green) and the fluid phase marker, TR-DX (red), for 10 minutes at 37°C. Abundant co-staining (yellow) was observed, indicating co-localization with this fluid phase marker. In parallel studies performed under identical conditions with FITC-GRP94 and Texas Red-transferrin, substantial colocalization was also observed (data not shown). In D, cells were stained for the lysosomal marker LAMP-2 (red), following internalization of FITC-GRP94 for 20 minutes at 37°C. Under these conditions, and as seen in the experiments depicted in Fig. 1, prominent co-staining with lysosomal markers was not observed. The intensity of the FITC-GRP94 staining observed after 20 minutes at 37°C was observed to significantly decrease in experiments addressing both receptor-mediated internalization or combined receptor-mediated/fluid phase internalization. Because of the fluor concentration, volume dilution and focal plane constraints that complicate interpretation of this diminished confocal microscopy signal, identical experiments were performed and the cell-associated fluorescence determined by flow cytometry. The results of these experiments are shown in Fig. 6, where it is evident that at time periods up to 20 minutes only a negligible decrease in cell-associated fluorescence was observed. When the incubations were extended to 60 minutes, total cell-associated fluorescence decreased to 65% of that seen immediately upon internalization. It appears, therefore, that the decrease in fluorescence intensity seen at later time points in the confocal microscopy studies is not a consequence of release of the probe back into the medium. Rather, this decrease more likely represents the loss in signal intensity that would accompany dilution of the probe into a relatively large volume compartment, such as the cytosol, Golgi complex, and/or the ER.

Fig. 5.

Receptor-mediated and fluid-phase uptake of FITC-GRP94 in murine macrophages. Adherent peritoneal macrophages were incubated at 4°C for 30 minutes (A) or 37°C for 5 minutes (B), 10 minutes (C) or 20 minutes (D) in medium supplemented with 80 μg/ml FITC-GRP94. In the experiment depicted in C, the medium was additionally supplemented with 1 mg/ml Texas Red dextran. In the experiments depicted in A and D, cells were fixed at the start or completion, respectively, of the uptake period and the lysosomal compartment stained with antibodies directed against LAMP-2. All images depict representative laser confocal sections and staining patterns. FITC-GRP94 is displayed as green, Texas Red-dextran or LAMP-2 staining as red, and co-staining as yellow.

Fig. 5.

Receptor-mediated and fluid-phase uptake of FITC-GRP94 in murine macrophages. Adherent peritoneal macrophages were incubated at 4°C for 30 minutes (A) or 37°C for 5 minutes (B), 10 minutes (C) or 20 minutes (D) in medium supplemented with 80 μg/ml FITC-GRP94. In the experiment depicted in C, the medium was additionally supplemented with 1 mg/ml Texas Red dextran. In the experiments depicted in A and D, cells were fixed at the start or completion, respectively, of the uptake period and the lysosomal compartment stained with antibodies directed against LAMP-2. All images depict representative laser confocal sections and staining patterns. FITC-GRP94 is displayed as green, Texas Red-dextran or LAMP-2 staining as red, and co-staining as yellow.

Fig. 6.

Time course of GRP94 internalization: Analysis of cell-associated FITC-GRP94 by flow cytometry. Peritoneal macropahges were selected by adherence and FITC-GRP94 binding performed for 60 minutes at 4°C in the presence of 50 μg/ml FITC-GRP94. Unbound FITC-GRP94 was removed by washing, and cells warmed to 37°C for the indicated times. At the completion of each incubation, cells were rinsed with ice-cold PBS, and fixed on ice. Cells were recovered by scraping, and cell-associated FITC determined by flow cytometry.

Fig. 6.

Time course of GRP94 internalization: Analysis of cell-associated FITC-GRP94 by flow cytometry. Peritoneal macropahges were selected by adherence and FITC-GRP94 binding performed for 60 minutes at 4°C in the presence of 50 μg/ml FITC-GRP94. Unbound FITC-GRP94 was removed by washing, and cells warmed to 37°C for the indicated times. At the completion of each incubation, cells were rinsed with ice-cold PBS, and fixed on ice. Cells were recovered by scraping, and cell-associated FITC determined by flow cytometry.

Soluble, exogenous GRP94 can be internalized by murine peritoneal macrophages through receptor-dependent and receptor-independent pathways. GRP94 uptake into Mφ displayed moderate sensitivity to inhibition by yeast mannan, a competitive inhibitor of the mannose/fucose receptor (Stahl et al., 1980), and substantial sensitivity to dimethylamiloride, an inhibitor of macropinocytosis (West et al., 1989). Internalization of GRP94 by either pathway initially yielded a predominant localization to an endosomal (transferrin positive) compartment, with little co-staining of the lysosomal compartment (LAMP-2 positive) at time periods of up to 20 minutes. The ultimate fate of internalized GRP94 is the subject of ongoing study, and may include transport into the cytosol, and/or transport into a non-endosomal compartment, such as the Golgi apparatus or the endoplasmic reticulum.

A trafficking pathway supporting the transfer of exogenous, soluble antigens into the MHC class I antigen presentation pathway of antigen presenting cells has been postulated to exist, and, in fact, is considered critical for cytotoxic T cell surveillance of antigens residing outside the secondary lymphoid compartments (Bevan, 1987, 1995; Carbone et al., 1998). Appropriately, such a pathway has been observed to function in macrophages (Harding and Song, 1994; Norbury et al., 1995) and dendritic cells (Brossart and Bevan, 1997; Mitchell et al., 1998; Norbury et al., 1997). Although the subcellular compartments accessed in this pathway are not understood in substantial molecular detail, it is noteworthy that in both Mφ and DC, the process of macropinocytosis has been demonstrated to support transfer of soluble, exogenous antigens into the cytosol, and presentation on MHC class I molecules (Brossart and Bevan, 1997; Kovacsovics-Bankowski and Rock, 1995; Norbury et al., 1995, 1997). As is demonstrated herein, GRP94, when presented as a soluble, exogenous protein to Mφ can be internalized by a fluid phase pathway that by virtue of its sensitivity to amiloride, likely represents macropinocytosis.

What is the immunologically relevant pathway by which GRP94 elicits CD8+ T cell responses in vivo? As it is generally accepted that internalization via receptor-mediated endocytosis is more efficient then a fluid phase process, it could be presumed that the receptor-dependent pathway would prove, from an immunological standpoint, most relevant. However, the extremely high rates of macropinocytosis observed in antigen presenting cells, such as DC, may, on a time-averaged scale, compensate for the necessarily reduced efficiency of a fluid-phase process (Sallusto et al., 1995; Swanson and Watts, 1995). Nonetheless, antigen presenting cells, such as macrophages, display cell surface receptors capable of binding and internalization of the ER chaperone GRP94. In light of evidence supporting the existence of a cell surface receptor capable of binding a protein (GRP94) whose access to the extracellular space would likely be limited to conditions in which cells undergo lysis (e.g. necrosis/apoptosis/viral lysis), it would indeed appear that a receptor dependent GRP94 uptake pathway is likely to be of immunological relevance. Delineation of the relevant pathway for GRP94-dependent presentation of peptide antigens will, however, require knowledge of the relative rates and efficiencies of GRP94 internalization by both receptor-mediated and fluid phase pathways, and the efficiency by which each pathway can deliver peptide antigens to the relevant cellular compartment, for exchange onto MHC class I molecules.

In mammals, all cells with the exception of erythrocytes, express GRP94 (Nicchitta, 1998). Furthermore, as noted, GRP94 exists in stable association with peptides, some of which are suitable for assembly on class I molecules (Srivastava et al., 1998). In view of the described data supporting the existence of receptor dependent and receptor-independent pathways for the internalization of GRP94 by antigen presenting cells (Mφ), it is reasonable to consider that in vivo, GRP94, and/or other ER-resident chaperones, may perform an immunologically relevant function in cross presentation (Bevan, 1995; Matzinger, 1994). Should this prove to be true, the scope of GRP94 function will expand to include a role in the regulation of immune responses to tissue damage.

The authors acknowledge the valuable support and assistance of Dr Michael Cook and Ms Lynn Martinek of the Duke University Comprehensive Cancer Center Flow Cytometry Facility and Dr Richard Fehon and Ms Suzanne Ward of the Duke University Comprehensive Cancer Center shared confocal microscope facility. The authors are very grateful to Dr Eli Gilboa (Center for Genetic and Cellular Therapy, DUMC) and Dr Blanche Capel (Dept of Cell Biology, DUMC) for generously providing the mice used in this study. We also recognize the valuable contributions of Benjamin B. Edell, a recipient of the Duke University Comprehensive Cancer Center Summer on the Edge Program and dedicate this work in memory of his father. C.D. was a recipient of a Howard Hughes Medical Institute Research Training Fellowship. This work was supported by NIH grant DK53058 (C.N.).

Arnold
,
D.
,
Faath
,
S.
,
Rammensee
,
H.-G.
and
Schild
,
H.
(
1995
).
Cross-priming of histocompatability antigen-specific cytotoxic T cells upon immunization with the heat shock protein gp96
.
J. Exp. Med
.
182
,
885
889
.
Bevan
,
M. J.
(
1987
).
Antigen recognition. Class discrimination in the world of immunology
.
Nature
325
,
192
193
.
Bevan
,
M. J.
(
1995
).
Antigen presentation to cytotoxic T lymphocytes in vivo
.
J. Exp. Med
.
182
,
639
641
.
Brossart
,
P.
and
Bevan
,
M. J.
(
1997
).
Presentation of exogenous protein antigens on major histocompatibility complex class I molecules by dendritic cells: pathway of presentation and regulation by cytokines
.
Blood
90
,
1594
1599
.
Carbone
,
F. R.
,
Kurts
,
C.
,
Bennett
,
S. R. M.
,
Miller
,
J. F. A. P.
and
Heath
,
W. R.
(
1998
).
Cross-presentation: a general mechanism for CTL immunity and tolerance
.
Immunol. Today
19
,
103
109
.
Harding
,
C. V.
and
Song
,
R.
(
1994
).
Phagocytic processing of exogenous particulate antigens by macrophages for presentation by class I MHC molecules
.
J. Immunol
.
53
,
4925
4933
.
Harding
,
C. V.
(
1996
).
Class I MHC presentation of exogenous antigens
.
J. Clin. Immunol
.
16
,
90
96
.
Heemels
,
M.-T.
and
Ploegh
,
H.
(
1995
).
Generation, translocation and presentation of MHC Class I restricted peptides
.
Annu. Rev. Biochem
.
64
,
463
491
.
Kovacsovics-Bankowski
,
M.
and
Rock
,
K. L.
(
1995
).
A phagosome-to-cytosol pathway for exogenous anatigens presented on class I MHC molecules
.
Science
267
,
243
246
.
Matzinger
,
P.
and
Bevan
,
M. J.
(
1977
).
Induction of H-2-restricted cytotoxic T cells: in vivo induction has the appearance of being unrestricted
.
Cell. Immunol
.
33
,
92
100
.
Matzinger
,
P.
(
1994
).
Tolerance, danger and the extended family
.
Annu. Rev. Immunol
.
12
,
991
1045
.
Miesenböck
,
G.
and
Rothman
,
J. E.
(
1995
).
The capacity to retrieve escaped ER proteins extends to the trans-most cisterna of the Golgi stack
.
J. Cell Biol
.
129
,
309
319
.
Mitchell
,
D. A.
,
Nair
,
S. K.
and
Gilboa
,
E.
(
1998
).
Dendritic cell/macrophage precursors capture exogenous antigen for MHC class I presentation by dendritic cells
.
Eur. J. Immunol
.
28
,
1923
1933
.
Moore
,
M. W.
,
Carbone
,
F. R.
and
Bevan
,
M. J.
(
1988
).
Introduction of soluble protein into class I pathway of antigen processing and presentation
.
Cell
54
,
777
785
.
Nair
,
S.
,
Wearsch
,
P.
,
Mitchell
,
D. A.
,
Wassenberg
,
J. J.
,
Gilboa
,
E.
and
Nicchitta
,
C. V.
(
1999
).
Careticulin displays in vivo peptide binding activity and can elicit cytotoxic T lymphocyte responses against bound peptides
.
J. Immunol
.
162
,
6426
6432
.
Nicchitta
,
C. V.
(
1998
).
Biochemical, cell biological and immunological issues surrounding the endoplasmic reticulum chaperone GRP94
.
Curr. Opin. Immunol
.
10
,
103
109
.
Norbury
,
C.
,
Hewlett
,
L. J.
,
Prescott
,
A. R.
,
Shastri
,
N.
and
Watts
,
C.
(
1995
).
Class I MHC presentation of exogenous soluble antigen via macropinocytosis in bone marrow macrophages
.
Immunity
3
,
789
791
.
Norbury
,
C.
,
Chamber
,
B. J.
,
Prescott
,
A. R.
,
Ljunggren
,
H.-G.
and
Watts
,
C.
(
1997
).
Constitutive macropinocytosis allows TAP-dependent major histocompatability complex class I presentation of exogenous soluble antigen by bone marrow-derived dendritic cells
.
Eur. J. Immunol
.
27
,
280
288
.
Ortmann
,
B.
,
Copeman
,
J.
,
Lehner
,
P. J.
,
Sadasivan
,
B.
,
Herberg
,
J. A.
,
Grandea
,
A. G.
,
Riddell
,
S. R.
,
Tampe
,
R.
,
Spies
,
T.
,
Trowsdale
,
J.
and
Cressell
,
P.
(
1997
).
A critical role for tapasin in the assembly and function of multimeric MHC class I-TAP complexes
.
Science
277
,
1306
1309
.
Pfeifer
,
J. D.
,
Wick
,
M. J.
,
Roberts
,
R. L.
,
Findlay
,
K. F.
,
Normark
,
S. J.
and
Harding
,
C. V.
(
1993
).
Phagocytic processing of bacterial antigens for class I MHC presentation to T cells
.
Nature
361
,
359
362
.
Reis e Sousa
,
C.
and
Germain
,
R.
(
1995
).
Major histocompatability complex class I presentation of peptides derived from soluble exogenous antigen by a subset of cells engaged in phagocytosis
.
J. Exp. Med
.
182
,
841
851
.
Sadasivan
,
B.
,
Lehner
,
P. J.
,
Ortmann
,
B.
,
Spies
,
T.
and
Cressell
,
P.
(
1996
).
Roles for calreticulin and a novel glycoprotein, tapasin, in the interaction of MHC class I molecules with TAP
.
Cell
5
,
103
114
.
Sallusto
,
F.
,
Cella
,
M.
,
Danieli
,
C.
and
Lanzavecchia
,
A.
(
1995
).
Dendritic cells use macropinocytosis and the mannose receptor to concentrate macromolecules in the major histocompatability complex class II compartment: Downregulation by cytokines and bacterial products
.
J. Exp. Med
.
182
,
389
400
.
Sandvig
,
K.
and
Van Duers
,
B.
(
1996
).
Endocytosis, intracellular transport and cytotoxic action of Shiga toxin and ricin
.
Physiol. Rev
.
76
,
949
966
.
Solheim
,
J. C.
,
Harris
,
M. R.
,
Kindle
,
C. S.
and
Hansen
,
T. H.
(
1997
).
Prominence of beta 2-microglobulin, class I heavy chain conformation and tapasin in the interactions of class I heavy chain with calreticulin and the transporter associated with antigen processing
.
J. Immunol
.
158
,
2236
2241
.
Srivastava
,
P. K.
,
Udono
,
H.
,
Blachere
,
N. E.
and
Li
,
Z.
(
1994
).
Heat shock proteins transfer peptides during antigen processing and CTL priming
.
Immunogenetics
39
,
93
98
.
Srivastava
,
P. K.
,
Menoret
,
A.
,
Basu
,
S.
,
Binder
,
R. J.
and
McQuade
,
K.L.
(
1998
).
Heat shock proteins come of age: primative functions acquire new roles in an adaptive world
.
Immunity
8
,
657
665
.
Stahl
,
P.
,
Schelsinger
,
P. H.
,
Sigardson
,
E.
,
Rodman
,
J. J.
and
Lee
,
Y. C.
(
1980
).
Receptor-mediated pinocytosis of mannose glycoconjugates by macrophages: characterization and evidence for receptor recycling
.
Cell
19
,
207
215
.
Suto
,
R.
and
Srivastava
,
P. K.
(
1995
).
A mechanism for the specific immunogenicity of heat shock protein-chaperoned peptides
.
Science
269
,
1585
1588
.
Swanson
,
J. A.
(
1989
).
Phorbol esters stimulate macropinocytosis and soute flow through macrophates
.
J. Cell. Sci
.
94
,
135
142
.
Swanson
,
J. A.
and
Watts
,
C.
(
1995
).
Macropinocytosis
.
Trends Cell Biol
.
5
,
424
428
.
Udono
,
H.
,
Levy
,
D. L.
and
Srivastava
,
P. K.
(
1994
).
Cellular requirements for tumor-specific immunity elicited by heat shock proteins: Tumor rejection antigen gp96 primes CD8+ T cells in vivo
.
Proc. Nat. Acad. Sci. USA
91
,
3077
3081
.
Walden
,
P. R.
and
Eisen
,
H. N.
(
1990
).
Cognate peptides induce self-destruction of CD8+ cytotoxic T lymphocytes
.
Proc. Nat. Acad. Sci. USA
87
,
9015
9019
.
Watts
,
C.
(
1997
).
Capture and processing of exogenous antigens for presentation on MHC molecules
.
Annu. Rev. Immunol
.
15
,
821
850
.
Wearsch
,
P. A.
and
Nicchitta
,
C. V.
(
1996
).
Purification and partial molecular characterization of GRP94, an ER resident chaperone
.
Prot. Express. Purif
.
7
,
114
121
.
Welch
,
W. J.
,
Garrels
,
J. I.
,
Thomas
,
G. P.
and
Feramisco
,
J. R.
(
1983
).
Biochemical characterization of the mammalian stress proteins and identification of two stress proteins as glucose- and Ca2+-ionophore regulated proteins
.
J. Biol. Chem
.
258
,
7102
7111
.
West
,
M. A.
,
Bretscher
,
M. S.
and
Watts
,
C.
(
1989
).
Distinct endocytic pathways in epidermal growth factor-stimulated human carcinoma. A431 cells
.
J. Cell Biol
.
109
,
2731
2739
.
York
,
I. A.
and
Rock
,
K. L.
(
1996
).
Antigen processing and presentation by the class I major histocompatability complex
.
Annu. Rev. Immunol
.
14
,
369
396
.

Receptor mediated endocytosis of GRP94 (gp96) in antigen presenting cells has recently been reported by Arnold-Schild et al. (1999) J. Immunol. 162, 3757-3760.