Abstract
Efalizumab (Raptiva) is a humanized CD11a-specific monoclonal antibody that was recently approved for the treatment of moderate to severe psoriasis. In psoriasis patients, the rate of efalizumab clearance from serum is related to T-cell surface expression of CD11a, suggesting a receptor-mediated clearance model for efalizumab (Bauer et al., 1999). However, limited experimental data are available to explain how the interaction with CD11a results in the systemic clearance of efalizumab. The following studies were designed to test the hypothesis that one mechanism of anti-CD11a antibody clearance is mediated in part by cellular internalization. This was tested in vitro using purified mouse and human T-cells as a model to study the cellular uptake and clearance of anti-CD11a antibodies. Data from these studies suggest that anti-CD11a antibodies are internalized by purified T-cells. Upon internalization, the antibodies appeared to be targeted to lysosomes and were cleared from within the cells in a time-dependent manner. CD11a-mediated internalization and lysosomal targeting of efalizumab may constitute one pathway by which this antibody is cleared in vivo.
The mechanism(s) by which antibodies are cleared is of particular interest as antibody-based therapies are developed and marketed for the treatment of multiple indications. One of these therapies is a monoclonal IgG1 antibody named efalizumab, which is specific for the human integrin lymphocyte function associated antigen-1 by targeting the α subunit CD11a (Raptiva; Genentech, South San Francisco, CA) (Werther et al., 1996). Efalizumab binding to human T-cells induces a rapid down-modulation of cell surface CD11a and inhibits T-cell activation, migration, and trafficking in vivo (Werther et al., 1996; Bauer et al., 1999; Gottlieb et al., 2000, 2002; Krueger et al., 2000). In a mouse model of human psoriasis, the antibody was found to significantly reduce the epidermal thickness of psoriatic plaques (Zeigler et al., 2001). Similarly, in psoriasis patients, histological studies demonstrated the utility of efalizumab in reducing the number of T-cells in psoriatic plaques and reducing the overall epidermal thickness of lesions in these patients (Gottlieb et al., 2000; Papp et al., 2001). Efalizumab has demonstrated safety and efficacy and is now approved for the treatment of moderate to severe psoriasis (Gordon et al., 2003; Lebwohl et al., 2003).
When delivered intravenously as a single bolus injection of 0.3 mg/kg, efalizumab was cleared at a rate of 322 ml/day, whereas at higher doses of 3 and 10 mg/kg, the kinetics of clearance dramatically slowed to 11 and 6.6 ml/day, respectively (Bauer et al., 1999). Similar observations were made by Gottlieb et al. (2000) who found that efalizumab saturated CD11a binding sites in human psoriasis patients when delivered intravenously at doses greater than 0.3 mg/kg. Subsaturating doses equal to or lower than 0.3 mg/kg did not fully down-modulate CD11a cell surface expression. These subsaturating doses were also cleared at a faster rate than saturating doses (Bauer et al., 1999; Gottlieb et al., 2000).
By assessing the kinetics of efalizumab clearance as a function of CD11a cell surface expression on CD3-positive T-cells, Bauer et al. (1999) demonstrated a relationship between the level of CD11a expression and the rate of efalizumab clearance that reflected the antibody-induced change in the CD11a receptor population. As remarked upon by the authors, these pharmacokinetic and pharmacodynamic profiles suggested that receptor-mediated clearance of efalizumab may be one mechanism by which efalizumab is cleared in humans. As such, it appeared that efalizumab may be cleared in vivo by both saturable processes, namely CD11a receptor-mediated internalization and clearance and CD11a-independent or nonsaturable processes common to IgG clearance. Given that CD11a is expressed on all circulating and tissue-associated lymphocytes (Desroches et al., 1990), this theoretically constitutes one large pool of cells capable of binding efalizumab. Although hypothesized, this model for efalizumab clearance, as mediated by cellular internalization, has not been further tested.
In most cases, antibodies that modulate the expression of cell surface antigens do so through a process similar to receptor-mediated endocytosis of the entire immunocomplex, where the antibody and antigen enter common or distinct pathways for recycling and/or degradation (for review, Stackpole and Jacobson, 1978; Wileman et al., 1985). The ultimate fate of the immune complex varies considerably and depends on the antibody isotype and structure (De Rie et al., 1988; Goulet et al., 1997), the target antigen (Pesando et al., 1986; Carriere et al., 1989; Pulczynski et al., 1993, 1994), and the extent of antibody-induced receptor cross-linking (Drebin et al., 1985; Morel et al., 1992; Gaietta et al., 1994). Using T-cells as a model, we conducted a series of in vitro studies designed to elucidate the role of purified T-cells in anti-CD11a antibody clearance. Our model predicted that upon binding of anti-CD11a antibody to its cell surface antigen, the antibody would be internalized and cleared from within the cell. To test this model, we used purified human and mouse T-cells to assess cellular binding, internalization, subcellular localization, and the kinetics of total and intracellular clearance of both efalizumab, as well as muM17, a monoclonal murinized rat anti-mouse antibody specific for mouse CD11a.
Materials and Methods
Isolation of Human T-Cells. Human T-cells were isolated from 100 ml of anonymous whole donor blood using the T-Cell Negative Isolation Kit provided by Dynal Biotech (Lake Success, NY). Lymphocyte separation from whole blood was made over a Ficoll gradient (ICN Pharmaceuticals Biochemicals Division, Aurora, OH). Lymphocytes were then suspended in 20 ml of complete media (RPMI 1640, 10% fetal bovine serum, 1% glutamine, 1% penicillin/streptomycin, and 1% sodium pyruvate) in a tissue culture flask and placed in a humidified 37°C tissue culture incubator for 1 h to allow monocytes and macrophages to attach. Following incubation, the unattached cells were removed from the flasks and resuspended in 107 cells per 200 μl of PBS containing 0.1% bovine serum albumin. Twenty microliters of fetal bovine serum per 107 cells was added followed by the addition of 20 μl of Antibody Mix (Dynal Biotech) for 10 min at 4°C. Excess antibody was washed out, and T-cells were negatively separated by the addition of Fc-specific antibodies coated on magnetic beads followed by exposure to a magnetic source. Recovered T-cells were greater than 97% pure in all studies, as determined by flow cytometry using FITC-conjugated anti-CD3 monoclonal antibody (data not shown).
Isolation of Mouse T-Cells. Mouse T-cells were isolated from female C57BL/6 mouse lymph nodes by selecting for major histocompatibility class II negative cells eluted over an LS column following provided protocols (Miltenyi Biotech, Auburn, CA). Briefly, lymph nodes extracted and pooled from three mice were used to prepare a single cell suspension using a 0.44-μm cell separation filter (Sigma-Aldrich, St. Louis, MO). Cells were counted, mixed with 1 μg of magnetically labeled anti-mouse major histocompatibility class II antibody per 106 cells, and then incubated at 4°C for 15 min. Following incubation, the magnetically labeled cells were passed through an LS column attached to a magnetic source enabling the T-cells to flow through while trapping other major cell types. T-cell purity was greater than 97% in all studies and determined by flow cytometery using FITC-conjugated anti-CD3 monoclonal antibody (data not shown).
In Vitro Internalization of Efalizumab and muM17. Efalizumab and muM17 were conjugated to AlexaFluor 488 following procedures supplied by Molecular Probes (Eugene, OR). AlexaFluor 488-conjugated efalizumab (efalizumab-488) or AlexaFluor 488-conjugated muM17 (muM17–488) was added to purified human or mouse T-cells, respectively, at a concentration of 1 μg of antibody per aliquot of 106 cells in 200 μl of PBS for 30 min at 4°C. Following incubation, cells were washed by centrifugation at 800g and resuspended in aliquots of 106 cells in 200 μl of PBS in the presence (cross-linking condition) or absence (noncross-linking condition) of 1 μg of anti-human or anti-mouse IgG antibody (BD Pharmingen, San Diego, CA), then incubated for 30 min at 4°C. At the conclusion of this incubation, cells were again pelleted at 800g and resuspended in full media. Cells were transferred to a 37°C tissue culture incubator, and aliquots of 106 cells were removed at various time points depending on the experiment conducted (see Results). In some experiments, 1 μM concanamycin A (Calbiochem, San Diego, CA) was included in the tissue culture medium during incubation of cells at 37°C to inhibit lysosomal function. At specified time points, cells were immediately placed on ice and stored at 4°C to terminate internalization. One aliquot of cells per time point was resuspended in 500 μl of 2% paraformaldehyde, whereas a second aliquot was first resuspended in 500 μl of a low pH acid wash (double distilled water, 150 mM NaCl, HCl, pH = 2.5) for 5 min at room temperature to remove noninternalized (extracellular-bound) antibody prior to fixing in 2% paraformaldehyde. Fixed cells were then assessed by flow cytometery (FACS Caliber; BD Biosciences, Franklin Lakes, NJ) to quantitate the geometric mean fluorescent intensity (MFI) from each condition. Each condition was replicated a minimum of three times. In each experiment, our hypotheses predicted movement of the data toward a single tail. Hence, the one-tailed Student's t test was performed for all statistical comparisons.
Surface and Intracellular Labeling of CD11a on Human T-Cells. Isolated human T-cells (1 × 106 cells) were incubated in complete RPMI with or without 1 μg/ml acid aggregated efalizumab (acid aggregated with 38% aggregate, lot 32893-83, 24 mg/ml, produced at Genentech) for 24 h at 37°C. Following incubation with efalizumab, the T-cells were washed with ice-cold PBS, centrifuged at 800g, and resuspended in 0.1 ml of PBS + 1% bovine serum albumin. The T-cells were then surface-labeled with PE-anti-CD11a (clone 25.3.1; Immunotech, Marseille, France) which recognizes an epitope distinct from that of efalizumab, FITC-anti-CD8, and allophycocyanin-anti-CD4 (BD Pharmingen, San Diego, CA) for 1 h at 4°C, washed and resuspended in 0.5 ml of PBS + 2% paraformaldehyde in preparation for fluorescence-activated cell sorter analysis. The concentration of antibody used was based on the manufacturer's recommendation. For intracellular labeling of CD11a, the T-cells were first surface-labeled with unconjugated-anti-CD11a (25.3.1), FITC-anti-CD8, and allophycocyanin-anti-CD4 at 10 μg/ml. T-cells were fixed and permeabilized using the intraprep permeabilization reagent kit according to the manufacturer's recommended procedure (Immunotech). The T-cells were then incubated with PE-anti-CD11a (25.3.1) to label the intracellular pools of CD11a receptor. The surface or intracellular expression of CD11a on the T-cells was measured using a three-color analysis on a FACSort cytofluorimeter. T-cells were analyzed by double gating on the forward and side scatter and then on positive staining for anti-CD4 or anti-CD8. The geometric MFI of PE-anti-CD11a was determined from a histogram of 10,000 gated events. The CD11a expression was normalized to the percentage of CD11a expression of untreated T-cells.
Wide-Field Fluorescent Microscopy. Human or mouse T-cells were incubated with efalizumab-488 or muM17–488, respectively, at a ratio of 106 cells per 1 μg of antibody in 200 μl of PBS for 30 min at 4°C. The cells were washed and then incubated in the same volume with 1 μg of anti-human or anti-mouse IgG for an additional 30 min at 4°C. Cells were again washed and incubated at 37°C for a range of time points. At the conclusion of each time point, cells were placed on ice. At the conclusion of all time points, cells were incubated with 5 μM DiI for 5 min at 37°C to label the plasma membrane or 50 nM LysoTracker Red for 30 min at 37°C to label the lysosomes (Molecular Probes). Cells were incubated for 5 min with 100 nM Hoechst 33342 at 37°C to label the nucleus (Calbiochem). Labeled cells were then attached to poly[l]lysine coated plates and assessed for anti-CD11a antibody internalization and subcellular localization using wide-field fluorescent microscopy. In some experiments, following binding and internalization of efalizumab-488, cells were subjected to immunocytochemistry using PE-conjugated anti-human CD11a antibody (25.3.1) to detect cellular CD11a and Hoechst 33342 to label the nucleus. Briefly, cells were plated on poly[l]lysine coated coverslips following internalization of efalizumab-488. Cells were fixed in 3.7% formaldehyde for 20 min, rinsed in PBS, and permeablized with 0.1% Triton X-100 in PBS for 15 min. Cells were again rinsed with PBS before adding 10 μg/ml PE-conjugated anti-CD11a antibody (25.3.1) for 30 min at room temperature. Cells were washed and stained with Hoechst 33342 before imaging.
Transmission Electron Microscopy (TEM). Purified T-cells bound by anti-CD11a antibody followed by secondary anti-human or anti-mouse conjugated 10-nm diameter gold particle secondary antibodies (Ted Pella, Redding, CA) were incubated at 37°C for various time points of internalization. Cells were then fixed in Karnovsky's solution (2% paraformaldehyde, 2.5% glutaraldehyde in 0.1 M cacodylate buffer, pH 7.4) and washed by centrifugation at 800g in 0.1 M sodium cacodylate (pH 7.4) twice for 15 min each wash before postfixing in 1% aqueous osmium tetroxide for 2 h at room temperature. After washing twice in double distilled water for 15 min each wash, the samples were dehydrated using a series of washes in ethanol: 50, 70, and 90% for 15 min each, then 100% twice for 15 min each, followed by propylene oxide twice for 15 min each. The samples were infiltrated with Eponate 12 (Ted Pella) first with 1:1 propylene oxide overnight, followed by 100% Eponate 12 for 8 h. The samples were transferred to fresh resin and polymerized in a 60°C oven overnight, after which ultra-thin (80 nm) sections were cut. The sections were stained with uranyl acetate and lead citrate and were observed on a Philips CM12 TEM. Images were captured with a Gatan Retractable MultiScan camera using Digital Micrograph software (Gatan Inc., Pleasanton, CA) and figures made using Adobe Photoshop, version 7.0.1. (Adobe Systems Inc., San Jose, CA).
Results
Internalization and Clearance of Anti-CD11a Antibody by Purified T-Cells. Human T-cells were used as a model to assess the binding and kinetics of internalization and intracellular clearance of efalizumab-488 from 0 to 72 h. Using this in vitro assay, internalization of efalizumab-488 alone by purified human T-cells was not detected (data not shown). However, in the presence of a secondary cross-linking antibody, internalization was observed (Fig. 1A). At time 0, total cell-associated efalizumab-488 was completely removed from the cells following low pH acid wash. With incrementally greater incubation times at 37°C, the time-dependent intracellular accumulation of efalizumab-488 by purified T-cells was detected with peak intracellular accumulation occurring between 24 and 48 h followed by some evidence for clearance of the antibody from the cells (Figs. 1A and 4B).
In a similar series of studies, mouse T-cell mediated internalization of muM17–488 was assessed from 0 to 24 h. Cross-linking of muM17–488 was not required for internalization (Fig. 1B). At time 0, all bound muM17–488 antibody was removed by the low pH acid wash. Following just 10 min at 37°C, however, a portion of muM17–488 antibody remained associated with the T-cells, suggesting that the antibody was internalized across the plasma membrane. Peak accumulation of the internalized antibody occurred after 2 h followed by evidence for clearance of the antibody from the cells. There was no difference observed in the kinetics of intracellular accumulation between the cross-linking (closed circles) and noncross-linking (open circles) conditions (Fig. 1B). There did appear, however, to be a small but nonsignificant effect of cross-linking on the overall rate of clearance of muM17–488 by isolated T-cells. In all experiments, internalization was defined as the time-dependent increase in MFI, as only antibody that was internalized within the plasma membrane would be retained following acid wash. Cellular clearance of the antibody was defined as the time-dependent decrease in MFI identified when the rate of loss of AlexaFluor488 was greater than its uptake. These data suggest that anti-CD11a antibody is internalized by purified T-cells in a time-dependent manner. muM17–488 appeared to be internalized at a faster rate by purified mouse T-cells relative to efalizumab-488 by purified human T-cells in vitro. Both antibodies were cleared from the cells over time.
Total Cellular Clearance of Anti-CD11a Antibody by Purified T-Cells. In the presence of cross-linking antibody, total cellular clearance of efalizumab-488 from purified human blood T-cells was assessed across 72 h (Fig. 2A). By 4 h of incubation at 37°C, there was an approximated 20% reduction in efalizumab-488 MFI associated with the T-cells. By 24 h, there was an approximated 30% reduction, and by 72 h, nearly a 90% reduction. Similarly, total mouse T-cell bound muM17–488 cleared from the cell in a time-dependent manner (Fig. 2B). In these studies, the clearance of muM17–488 alone from T-cells was detected following 16 h of incubation at 37°C. By comparison, in the presence of secondary cross-linking antibody, muM17–488 clearance from the T-cells was detected as early as 3 h of incubation at 37°C. By 24 h, roughly 10% of the muM17–488 antibody alone was cleared from the cells, whereas roughly 40% of muM17–488 was cleared during this time when in the presence of secondary cross-linking antibody. The difference in the mean clearance of antibody from the T-cells at each time point was compared between these two groups using a one-tailed t test. Statistical significance was reached between these two conditions by 16 and 24 h with p values of 0.018 and 0.0012, respectively.
Efalizumab-Induced T-Cell Surface and Intracellular Down-Modulation of CD11a. Immunocytochemical staining of permeablized human blood T-cells with anti-CD11a antibody revealed the presence of both cell surface CD11a expression in addition to a rather extensive biosynthetic pool of the receptor within the cell (data not shown). Hence, efalizumab-induced down-modulation of cell surface and intracellular pools of CD11a was explored. Results from these analyses are presented in Fig. 3. Cell surface CD11a expression on CD4+ and CD8+ human blood T-cells are depicted with open circles and squares, respectively. Intracellular expression of CD11a by CD4+ and CD8+ T-cells are represented with filled circles and squares, respectively. Over a 48-h incubation period, cell surface and intracellular CD11a was down-modulated by roughly 70%. These data suggest that efalizumab not only strongly down-modulates cell surface CD11a expression, but also depletes the biosynthetic pool of this receptor as well. Between 0 and 2 h following treatment with efalizumab, we observed a spike in intracellular CD11a. Similarly at 2 h, we observed a transient rise in cell surface expression relative to the 1-h time point. These observations may reflect the transient accumulation of intracellular CD11a following efalizumab-induced internalization and perhaps a consequent trafficking of intracellular pools to the plasma membrane during this time.
Inhibition of Total and Intracellular Clearance of Efalizumab by Concanamycin A. Total cellular clearance of efalizumab-488 did appear to be inhibited in the presence of the lysosome inhibitor concanamycin A. In the absence of concanamycin A, there was an approximated 87% reduction in T-cell associated efalizumab-488 MFI by 48 h. This total cellular clearance was reduced to 49% in the presence of concanamycin A over the same time (Fig. 4A). The inhibition of total cellular clearance of efalizumab-488 by concanamycin A was statistically significant (p = 0.0046). Similarly, intracellular clearance of efalizumab-488 by these cells was also inhibited by the inclusion of 1 μM concanamycin A (Fig. 4B). By 24 and 48 h in the presence of concanamycin A, there appeared to be more intracellular accumulation of efalizumab-488 relative to the no concanamycin A condition (p = 0.052 and 0.013, respectively), suggesting that the specific inhibition of lysosomal function blocked intracellular clearance of the antibody by purified human T-cells.
Subcellular Localization of T-Cell Internalized Anti-CD11a Antibody. The fluorescence-activated cell sorter analysis of anti-CD11a internalization suggested that lysosomal localization of both efalizumab and muM17 may be detected within 16 h. Hence, wide-field fluorescent microscopy was used to assess subcellular localization of internalized anti-CD11a antibody across 16 h. At time 0, efalizumab-488 was localized to the plasma membrane (left panel, Fig. 5A). More defined aggregation of the antibody on the plasma membrane was commonly observed following 2 h at 37°C (center panel, Fig. 5A). The internalization of efalizumab was observed at 16 h (right panel, Fig. 5A), during which time there was noticeably less antibody associated with the plasma membrane relative to time 0. Aliquots of these cells incubated in the presence of efalizumab-488 for 16 h at 37°C were labeled with LysoTracker Red and Hoechst 33342 to determine whether the internalized antibody costained with lysosomes. Results from this analysis are presented in Fig. 5B. The left panel represents the internalization of efalizumab-488 by purified human T-cells. The center panel depicts the subcellular location of lysosomes in these cells, whereas the right panel is a merged image of left and center panels, suggesting the costaining of internalized efalizumab-488 with lysosomes.
Similarly, at time 0, muM17–488 was found associated solely with the plasma membrane of purified T-cells (left panel, Fig. 6A). After 15 min at 37°C, however, it was common to see T-cell aggregates at sites of muM17–488 sequestration on the plasma membrane (center panel, Fig. 6A). Following 2 h at 37°C, internalization of muM17–488 was detected (right panel, Fig. 6A). Colocalization of muM17–488 with LysoTracker Red was also assessed (Fig. 6B). The left panel represents muM17–488 internalization by T-cells. The center panel identifies the location of lysosomes in the cytoplasm of these cells. The right panel is a merged image of the left and center panels suggesting that a portion of internalized muM17–488 costained with lysosomes.
Human T-cells were then assessed for costaining of efalizumab-488 with PE-conjugated anti-CD11a antibody to ascertain colocalization of the antibody with its receptor at 16 h following internalization. A representative image from this analysis is presented in Fig. 7. The left panel of Fig. 7 depicts internalized efalizumab-488, the center panel identifies the CD11a receptor, and the right panel represents a merged image of left and center panels. In all images taken, we observed the complete colocalization of efalizumab-488 with the CD11a receptor, suggesting that the entire immunocomplex traffics together upon internalization.
The costaining of internalized anti-CD11a antibody with LysoTracker Red suggested that the antibodies were targeted to lysosomes. To confirm, we utilized TEM to obtain morphological evidence that efalizumab and muM17 were localized within lysosomes. Figure 8 depicts representative images of efalizumab internalization and subcellular localization. At time 0 of internalization (T-0), efalizumab-associated gold particles were detected solely on the plasma membrane. By 2 h (T-2) of incubation at 37°C, the internalized antibody was found associated with the plasma membrane or within cytoplasmic vesicles. At 16 h (T-16), efalizumab was also found in vesicles, but more commonly in structures morphologically similar to lysosomes (Kleijmeer et al., 1997). Similar results were found with muM17 (Fig. 9). With no internalization at 37°C (T-0), muM17 was found solely on the plasma membrane. After 2 h at 37°C, the antibody was primarily contained within cytoplasmic vesicles and in some cells found within lysosomes. Following 16 h of incubation at 37°C, nearly all detectable anti-CD11a antibody was found in lysosomes. These data suggest that at least a portion of the internalized anti-CD11a antibody was targeted to lysosomes.
Discussion
We conducted a series of experiments to assess the kinetics of internalization, cellular clearance, and the mechanism of cellular clearance of anti-CD11a antibody by purified T-cells, which is the population of cells that efalizumab is believed to exert its major therapeutic mechanism of action on in psoriasis patients. Results from these experiments support several points. First, anti-CD11a antibody was internalized in vitro by purified T-cells in a time-dependent manner. muM17 internalization occurred rapidly with peak intracellular accumulation occurring following 2 h of incubation at 37°C. In contrast, peak intracellular accumulation of efalizumab in human T-cells occurred between 24 and 48 h. Second, cross-linking with a secondary antibody was required for internalization of efalizumab, but not muM17 in vitro. Third, internalized efalizumab and muM17 were found in subcellular organelles that stained positive for LysoTracker Red and were morphologically similar to lysosomes (Kleijmeer et al., 1997). Lysosomal localization of muM17 was detected as early as 2 h following incubation at 37°C, whereas efalizumab was found localized to lysosomes following 16 h. Last, treating cells with concanamycin A to retard lysosomal function reduced the total and intracellular clearance of efalizumab providing evidence for the role of lysosomes in clearing the internalized antibody. These studies suggest a mechanism for the previously reported model of efalizumab clearance as mediated through its interaction with the CD11a receptor (Bauer et al., 1999; Gottlieb et al., 2002).
The T-cell mediated internalization of anti-CD11a antibody observed in these studies is consistent with published observations describing antibody-induced internalization of other integrins (Gaietta et al., 1994; Leone et al., 2003). Modulation of α4β1 using the α4-subunit specific antibody TA-2, for example, resulted in internalization of the entire immune complex with peak intracellular accumulation occurring between 24 and 48 h (Leone et al., 2003). By comparison, 20% of total cell surface α6β1 immunocomplexed with GoH3, a monoclonal antibody specific for the α6 subunit, was internalized within the cell following just 30 min at 37°C, which represented peak accumulation of this complex within the cell. Antibody-induced antigenic modulation of α6β4 using this same antibody yielded nearly identical results (Gaietta et al., 1994). Although it appears that cellular internalization may be a common result of antigen/antibody complex formation, the subcellular fate has been less clearly defined for antibodies that modulate integrins.
The muscarinic acetylcholine receptor M1, when bound by a monoclonal antibody specific to an epitope tag, was found to undergo a rapid internalization and recycling back to the plasma membrane (Tolbert and Lameh, 1998). By comparison, Pulczynski (1994) assessed the internalization and subcellular fate of B-cell antigens CD10 and CD19 following modulation with monoclonal antibodies. Both receptors appeared to enter a common pathway for endocytosis. Upon internalization, CD19, more so than CD10, immune complexes appeared to enter a recycling pathway, whereas both CD19 and CD10 immune complexes were targeted for degradation. Antibody-induced antigenic modulation of CD4, CD5, and CD7 with monoclonal antibodies demonstrated differential rates of internalization depending on the antigen targeted. In all cases, however, the immune complexes were targeted to lysosomes (Carriere et al., 1989). Similar to these observations, our data suggest that at least a portion of the internalized anti-CD11a antibody is internalized and targeted to lysosomes.
An additional goal of these studies was to determine the in vitro requirement of cross-linking for antibody internalization and cellular clearance. In our in vitro model, efalizumab alone was not noticeably internalized and consequently did not induce CD11a receptor down-modulation from the cell surface. Only in the presence of secondary cross-linking antibody or aggregated antibody did we observe a time-dependent internalization and subsequent clearance of the antibody by purified human T-cells. However, efalizumab alone very efficiently induces CD11a receptor down-modulation from the plasma membrane in vivo (Bauer et al., 1999; Gottlieb et al., 2000). Hence, there is a fundamental difference in the effect of efalizumab on CD11a receptor down-modulation in vitro relative to in vivo that is not understood at this time. In contrast, muM17 appeared to be rapidly internalized by purified mouse T-cells independent of a secondary cross-linking antibody. We found that cross-linking of muM17 by including the secondary anti-mouse IgG antibody had little effect on the rate of muM17 internalization and intracellular clearance. The most significant influence of cross-linking was observed in the overall clearance of the antibody from the cell, suggesting that the cross-linked protein complex was more efficiently cleared by the mouse T-cells in vitro. Irrespective of cross-linking, muM17 accumulated within mouse T-cells at a faster rate than did efalizumab by human T-cells.
The fate of the CD11a receptor following efalizumab binding and internalization was also explored in vitro. Our studies suggest that the antibody induces not only the clearance of cell surface CD11a, but also the biosynthetic pool of this receptor. Hence, redistribution of the internalized CD11a receptor back to the cell membrane seems unlikely. Rather it seems more likely that during the course of therapy, efalizumab first down-modulates cell surface CD11a (by inducing internalization and lysosomal degradation) followed by the distribution of the biosynthetic pool of CD11a to the cell surface where this pool of receptor is now available for additional binding to efalizumab and therefore subject to further antibody-induced internalization and degradation. Considering CD11a is expressed on all circulating and tissue-associated lymphocytes, this could represent a significant pool of binding sites for efalizumab. In vivo data also support that circulating and tissue-associated lymphocytes contribute to the specific clearance of muM17 in mice (data not shown). Taken together, these data support a model where efalizumab binds the receptor and targets the entire immune complex to lysosomes for degradation. CD11a-mediated internalization and lysosomal degradation is likely to constitute at least one pathway by which efalizumab is cleared from the blood in vivo.
Acknowledgments
We thank Suzie Scales, Wolfgang Dummer, Amita Joshi, Yulia Vugmeyster, Joe Beyer, Cary Austin, William Mallet, and Jean-Philippe Stephan for insightful discussions regarding experimental design, antibody and receptor internalization, and trafficking and Kathy Howell and Anahid Bakshi for technical assistance with flow cytometry.
Footnotes
-
Financial support for this work was provided by Genentech, Inc. (South San Francisco, CA).
-
Article, publication date, and citation information can be found at http://jpet.aspetjournals.org.
-
doi:10.1124/jpet.104.067611.
-
ABBREVIATIONS: PBS, phosphate-buffered saline; FITC, fluorescein isothiocyanate; MFI, mean fluorescent intensity; PE, phycoerythrin; TEM, transmission electron microscopy.
- Received March 5, 2004.
- Accepted June 8, 2004.
- The American Society for Pharmacology and Experimental Therapeutics