Receptor-mediated Endocytosis in the Procyclic Form of Trypanosoma brucei* and

In Trypanosomatids , endocytosis exocytosis occur exclusively at the flagellar pocket, a deep invagination of the plasma membrane where the flagellum extends from the cell. Both bloodstream and procyclic trypanosomes are capable of internalizing macromolecules. However, structures resembling coated vesicles were only identified in bloodstream form and not in procyclic form trypanosomes. Due to the apparent absence of coated vesicles in procyclics, the significance of recep-tor-mediated endocytosis in procyclic trypanosomes has been considered of minimal importance. We show that the flagellar pocket associated cysteine-rich acidic transmembrane protein (CRAM) may function as an high density lipoprotein receptor in the procyclic form trypanosome. Using anti-CRAM IgG we have characterized the process of CRAM-mediated endocytosis in procyclic form trypanosomes. The wild type procyclic trypanosome binds and internalizes anti-CRAM IgG but not the non-immune IgG in a saturable and time-depen-dent manner; the binding and uptake of 125 I-labeled an-ti-CRAM IgG are inhibited by excess unlabeled anti-CRAM IgG. Uptake and degradation of anti-CRAM IgG do not occur at 4 °C. At 28 °C, the internalized anti-CRAM IgG were efficiently degraded through a process that is inhibited by incubation at 4 °C and sensitive to the presence of chloroquine. The uptake and degradation of anti-CRAM IgG does not occur in the CRAM null mutant cell line. These results suggested that the uptake

African trypanosomes are flagellated protozoan parasites, causing sleeping sickness in man and related diseases in livestock. Trypanosoma brucei has a biphasic life cycle alternating between the mammalian host and the insect vector tsetse fly. By undergoing antigenic variation of the variant cell surface glycoprotein, the parasites survive in the bloodstream of the mammals (1)(2)(3). T. brucei relies on its host for nutrients needed for its growth and development. In Trypanosomatids, endocy-tosis and exocytosis occur exclusively at the flagellar pocket (4 -10). The flagellar pocket is an invaginated plasma membrane where the flagellum extends from the cell. The plasma membrane of the cell body is supported by closely spaced subpellicular microtubule networks which prohibit active pinocytosis through the plasma membrane. This subpellicular microtubule network does not extend beneath the flagellar pocket membrane. Thus the pocket is the only part of the cell surface that supports endocytosis and exocytosis in trypanosomes. Macromolecules enter the flagellar pocket by passing through the adhesion zone between the plasma membranes covering the cell body and the flagellum, then macromolecules in the lumen of the pocket will be internalized by receptor-mediated endocytosis or fluid-phase endocytosis.
It has been shown that transferrin and low-density lipoprotein enter the bloodstream form of trypanosome via receptormediated endocytosis (5,(11)(12)(13)(14)(15). Ultrastructural analysis demonstrated that endocytosis occurs in the bloodstream form by the formation of spiny-coated vesicles of 100 -150 nM diameter which pinch off from the flagellar pocket membrane (5,6,16). The lumenal side of these large endocytic vesicles is covered by the variant cell surface glycoprotein coat and the cytosolic side by an electronic-dense material, which may be functionally equivalent to clathrin in higher eukaryotic cells (17,18). However, molecules that are immunologically cross-reactive with mammalian clathrin cannot be identified in the purified coated vesicles of trypanosomes (18). The flagellar pocket volume varies between different life stages; the fractional volume for the bloodstream form is ϳ3-fold larger than that in the procyclic form trypanosome (19). Since the large coated endocytic vesicles are absent in procyclic form trypanosomes, it was generally assumed that little or no endocytosis occurs in procyclic form trypanosomes. It has been shown that procyclic form trypanosomes can take up fluid-phase markers but with a rate much slower than that in bloodstream form trypanosomes (5). The different rates of endocytosis between different stages of parasites may be due to differences in morphological and functional properties of the flagellar pocket in different stages of the parasite.
Receptor-mediated endocytosis is a selective concentrating mechanism enabling cells to ingest large amounts of specific ligands and usually results in the transfer of extracellular molecules to lysosomes. In higher eukaryotes, it seems that ligand binding alters the conformation of the receptor protein that it is now recognized by a protein component of the coated pit. This signaling process can be mimicked by specific antibodies that recognize the receptor (20 -22). The efficient endocytosis of transmembrane receptor proteins in higher eukaryotes requires a signal sequence in the cytoplasmic domain of the protein to promote clustering into coated pits.
A cysteine-rich acidic trans-membrane (CRAM) 1 protein, is abundantly expressed in procyclic form trypanosomes and localized at the surface of the flagellar pocket and in vesicles underneath the pocket (23). CRAM has a predicted molecular mass of 130 kDa but its molecular mass measured in total cell protein extracts of procyclic trypanosomes is ϳ200 kDa, due to a high level of glycosylation. The extracellular domain of CRAM contains a large region of a 12-amino acid cysteine-rich repeat which is highly homologous to the complement-like cysteine-rich repeats present in the apolipoprotein-binding domain of the human LDL receptor (23,24). Based on this homology, we hypothesized that CRAM may function as a lipoprotein receptor of trypanosomes. Very little is known about receptors and receptor-mediated endocytosis in procyclic form trypanosomes. In this paper, we demonstrate that CRAM can function as a receptor and that receptor-mediated endocytosis occurs in the procyclic form trypanosome. Comparison of binding and uptake of high density lipoprotein (HDL) particles in the wild type procyclics and the CRAM null mutant cell lines demonstrated that CRAM potentially functions as an HDL receptor and multiple pathways exist for the uptake of HDL in the procyclic form trypanosomes. The presence of multiple pathways for the uptake of HDL in trypanosomes has hampered the analysis of mechanisms involved in the CRAM-mediated endocytosis using HDL as a specific ligand. Subsequently, using anti-CRAM IgG as a specific ligand, we have characterized the process of CRAM receptor-mediated endocytosis and have demonstrated that the cytoplasmic extension of CRAM plays an important role in the processing of internalized materials.

Trypanosome Strains
The procyclic form of T. brucei stock 427-60, originally obtained from Dr. Brun, was maintained in SDM-79 medium at 25-28°C (25). For all binding and uptake of IgG, early to mid log phase trypanosomes at a cell density at ϳ0.5 ϫ 10 7 cells ml Ϫ1 were used. Prior to the incubation with different IgGs, trypanosomes were incubated in SDM-79 medium without fetal calf serum for ϳ2 h and thus trypanosomes were devoid of the IgG molecules present in fetal calf serum. Cell line CRAM 4 -29 is a CRAM null mutant cell line (26). The CRAMB2 cell line contains only one allele of the CRAM gene and the CRAM-40 cell line contains also only one allele of the CRAM gene and encodes the CRAM protein that is truncated by 40 amino acids from its C-terminal end. 2

Serum Lipoprotein Purification
Different lipoprotein fractions from bovine serum were isolated by sequential flotation ultracentrifugation (28 (29). Lipoproteins were radiolabeled using Bilheimer's modification of the McFarlane method (30).

HDL Binding and Uptake
For each assay, 1-ml cells of 1 ϫ 10 7 cells ml Ϫ1 was used. Binding Assay-125 I-HDL was incubated with trypanosomes in SDM-79 medium (without serum), 1% bovine serum albumin (BSA) in the absence or presence of 1 mg ml Ϫ1 unlabeled HDL, at 4°C for 2 h. Following the incubation, trypanosomes were washed three times with 1 ϫ PBS, 0.2% BSA. Then, trypanosomes were transferred into a new set of tubes, spun down, and the amount of radioactivity associated with cell pellets was measured in a ␥-counter.
Uptake Assay-The procedure is as described for the binding assay except that the incubation of trypanosomes with 125 I-HDL was performed at 28°C for 4 h. Total and nonspecific counts were referred to the measurement of reactions performed in the absence and presence, respectively, of 1 mg ml Ϫ1 unlabeled HDL. Specific counts were obtained by subtraction of nonspecific counts from total counts.

Purification of IgG
IgG molecules from immunized rabbit serum were purified by using the Affi-Gel Protein A maps II kit from Bio-Rad. Then the purified IgG were concentrated using Centricon YM-30 from Millipore. Protein concentrations were measured using the bicinchoninic protein assay reagent from Pierce. Labeling of IgG with 125 I was performed using IODO-GEN Iodination reagent and pre-coated tubes from Pierce. Nonimmune rabbit IgG was purchased from Sigma.

Binding, Uptake, and Degradation of IgG
For each assay, 1 or 0.5 ml cells of 1 ϫ 10 7 cells ml Ϫ1 was used. Binding Assay-125 I-IgG was incubated with trypanosomes in SDM-79 medium (without serum), 1-3% bovine serum albumin (BSA) in the absence or presence of 20-fold excess amount of unlabeled IgG, at 4°C for 2 to 4 h. Following incubation, trypanosomes were washed three times with SDM-79 medium, 0.2% BSA. Then, trypanosomes were transferred into a new set of tubes, spun down, and the amount of radioactivity associated with cell pellets was measured in a ␥-counter.
Uptake Assay-The procedure is as described for the binding assay except that the incubation of trypanosomes with 125 I-IgG was performed at 28°C for 2-4 h.
Degradation Assay-Following the incubation of trypanosomes with 125 I-IgG at 4 or 28°C, trypanosomes were spun down and the supernatants were collected for determination of the amount of trichloroacetic acid (10% w/v)-soluble 125 I-labeled products. The amount of [ 125 I]iodine was removed from the trichloroacetic acid-soluble fraction by precipitation with 5% (w/v) silver nitrate. The amount of radioactivity in the trichloroacetic acid soluble, non-iodide fraction was referred to as the amount of degradation. Total and nonspecific counts were referred to the measurement of reactions performed in the absence and presence, respectively, of 20-fold excess of unlabeled IgG. Specific counts were obtained by subtraction of nonspecific counts from total counts.

Immunofluorescence Assay of Uptake
Live trypanosomes were first incubated with anti-CRAM IgG in serum-free SDM-79 medium at 28°C for 2 h. Following incubations, trypanosomes were washed two times with SDM-79 medium, 0.2% BSA and once with PBS, pH 7.4. Then cells were suspended in 3.7% formaldehyde in PBS for 10 min in ice and neutralized with 0.1 M glycine for 10 min. Following fixation, cells were spun down, washed with PBS, dotted on slides, and fixed in cold methanol and cold acetone for 5 min each. After rehydration in PBS for 5 min, slides were incubated with the first antibody in the presence of 3% bovine serum albumin and 0.05% Tween 20 in PBS for 1 h. Following the first antibody reaction, slides were washed three times with PBS and then reacted with the fluorescein isothiocyanate-and/or rhodamine-conjugated goat derived antirabbit or anti-rat IgG (Cappel) for 1 h. After the last wash, 4,6-diamino-2-phenylindole was applied in water at a concentration of 0.1 g/ml for 1 min at room temperature, after which the slides were briefly rinsed with water. The cells were mounted under a coverslip with a mounting medium containing 90% glycerol in PBS and 25 mg/ml of 1.4diazabicyclo[2.2.2]octane (Sigma). Cells were viewed and photographed under a Leica fluorescence microscope using a ϫ100 objective. Some of the images were directly captured by a CCD camera and analyzed by the MetaMorph program from Universal Imaging Co.

Binding and Uptake of HDL Particles in Wild Type Procyclic
Trypanosomes and CRAM Null Mutant Cell Lines-We have previously shown that the procyclic form trypanosome is capable of taking up HDL and LDL and uptake of LDL was significantly less efficient than uptake of HDL (31). Further kinetic analysis indicated that the HDL uptake in procyclics can occur through the process of receptor-mediated endocytosis. To determine whether CRAM can function as an HDL receptor in procyclic trypanosomes, the binding and uptake of 125 I-HDL particles in the CRAM null mutant cell line was compared with wild type procyclic trypanosomes. The CRAM null mutant showed a ϳ60% reduction in its maximal ability to bind 125 I-HDL particles; its ability to bind HDL was not completely abolished ( Fig. 1A; two independent cell lines gave similar results). Control cell lines containing the Neo and Hph markers brane protein; HDL, high density lipoprotein; LDL, low density lipoprotein; VLDL, very low density lipoprotein; BSA, bovine serum albumin. 2  integrated into a mini-chromosome showed a similar binding ability to that of the wild type procyclic trypanosome (data not shown). Thus neomycin or hygromycin treatment had no obvious effect on HDL binding in procyclic trypanosomes. The reduced HDL binding capacity in the CRAM null cell line suggested that CRAM may be involved in only one of several mechanisms for the uptake of HDL in trypanosomes and additional processes may exist for the uptake of HDL. At 28°C, the CRAM null mutant cell line exhibited a 20% reduction in its ability to internalize HDL, when compared with wild type trypanosomes, demonstrating that the CRAM null mutant remained capable of acquiring significant amounts of HDL (Fig.  1B).
Deletion of the CRAM gene drastically reduced the ability of procyclic trypanosomes to bind HDL particles at 4°C and moderately reduced a significant level of the ability to take up HDL at 28°C. These results strongly suggested that CRAM may function as an HDL receptor and the uptake of HDL in trypanosomes can occur via alternative pathways that operate in addition to the CRAM-mediated endocytosis. It is interesting to note that biochemical machinery required for de novo synthesis of fatty acids is present in trypanosomes, although synthesis is reduced when exogenous fatty acids are available (32-35).
Since the CRAM null mutant is viable, alternative mechanisms apparently exist, which complement the function of CRAM. The presence of multiple pathways for the uptake of HDL in the procyclic trypanosomes has interfered with the analysis of CRAM-mediated endocytosis using HDL as a specific ligand. Thus following the observation that interaction of ligand binding to the receptor protein can be mimicked by specific antibodies that recognize the receptor (20), we investigated whether anti-CRAM antibody can be used as a specific ligand to characterize the CRAM-mediated endocytosis. The anti-CRAM antibody was raised against the extracellular cysteinerich 12-amino acid repeat domain which presumably functions as the ligand-binding domain.
Specific Uptake and Degradation of Anti-CRAM IgG by the Wild Type Procyclic Trypanosome-We first determined whether the anti-CRAM IgG can serve as a specific ligand for the study of CRAM-mediated endocytosis in wild type procyclic form trypanosomes. Procyclic trypanosomes were incubated with 125 I-rabbit derived anti-CRAM IgG (50 g/ml) and 125 Inon-immune rabbit IgG (50 g/ml), respectively, at 28°C for different periods of time. Then, cell associated counts representing the amount of internalized IgG and trichloroacetic acid soluble/silver nitrate unprecipitable materials in supernatants FIG. 2. Time-dependent uptake and degradation of anti-CRAM in procyclic form trypanosomes. Procyclic trypanosomes (1 ϫ 10 7 cells ml Ϫ1 ) were incubated with 50 g/ml 125 I-anti-CRAM IgG (squares) and 50 g/ml rabbit non-immune IgG (circles), respectively, in SDM-79 serum-free medium and 3% BSA, at 28°C for different periods of time. Following the incubation, trypanosomes were spun down, separated from the supernatant, and washed three times with SDM-79 serum-free medium. Then, the cell associated counts, referred to as the amount of uptake, were measured (A). The supernatants were collected for the measurement of trichloroacetic acid soluble and silver nitrate unprecipitable materials which represent the degraded and secreted IgG molecules (B). The nonspecific uptake and degradation at 0 incubation time was subtracted from each experiment. Results represent the average of triplicate determinations.
FIG. 1. Binding and uptake of 125 I-HDL to the CRAM null mutant and the wild type procyclic trypanosome. Trypanosomes at a growing density of ϳ8 ϫ 10 6 cells/ml were used for this comparison. Aliquots of 10 7 cells were incubated at 4°C for 2 h or 28°C for 4 h with increasing concentrations of 125 I-HDL in the presence or absence of 1 mg/ml unlabeled HDL in trypanosome media containing 1% BSA (1 ml/tube). After which cells were washed and pellets were counted for radioactivity. The amount of specific binding or uptake were calculated as the difference between total (observed in the absence of unlabeled HDL) and nonspecific binding or uptake (observed in the presence of unlabeled HDL). A, binding. Results are represented as the mean Ϯ S.E. (n ϭ 4) of specific binding at 4°C. B, uptake. The average results of three independently performed experiments were plotted. The relative amount of uptake was calculated as the ratio of each specific uptake to that of the wild type procyclic trypanosome incubated with 100 mg/ml 125 I-labeled HDL. Closed squares, the wild type procyclic form trypanosome; open squares, the CRAM null mutant cell line.
representing the degraded product were measured. Significant uptake of non-immune IgG cannot be detected ( Fig. 2A, circles); this result indicated that procyclic form trypanosomes do not randomly internalize rabbit non-immune IgG. However, the wild type procyclic trypanosomes are capable of taking up 125 I-anti-CRAM IgG in a time-dependent manner; the amount of internalized 125 I-anti-CRAM IgG increased with the incubation time and reached a steady state after 2 h and remained constant for at least upto 4 h ( Fig. 2A, squares). The significance of the uptake of 125 I-anti-CRAM IgG was further confirmed by the amount of degraded 125 I-anti-CRAM IgG meas-ured in the incubation media (Fig. 2B). A significant amount of degradation was observed at 30 min after incubation and the amount of degradation product drastically increased with the incubation time. It appeared that the amount of degraded 125 I-anti-CRAM IgG accumulated in the media is ϳ4 -5-fold higher than the steady state level of cell associated 125 I-anti-CRAM IgG (refereed to as the uptake). Apparently, the procyclic form trypanosome is not just capable of taking up 125 I-anti-CRAM IgG but also degrades the internalized 125 I-anti-CRAM IgG which was subsequently released into the culture media. A significant amount of 125 I-anti-CRAM IgG degradation was only observed when the incubation time was longer than 30 min and the amount of degraded 125 I-anti-CRAM IgG increased with the increase of incubation time (Fig. 2B, squares). Degradation of non-immune-IgG could not be observed even after 4 h incubation with procyclic trypanosomes (Fig. 2B, circles). These results, taken together with the high level expression of the CRAM protein in procyclic form trypanosomes, suggested that procyclic form trypanosomes most likely efficiently take up anti-CRAM IgG, via the CRAM-mediated endocytic process, leading to proteolytic degradation of the anti-CRAM IgG molecules. Along with the recycling of transport vesicles, the degraded 125 I-anti-CRAM IgG was exported to the outside.
We further determined whether the uptake and/or degradation of 125 I-anti-CRAM IgG in procyclic trypanosomes is sensitive to the drug chloroquine, a lysosomotropic amine, which elevates the pH of lysosome resulting in its inhibitory effect. It appeared that following addition of 25 M chloroquine, the uptake and degradation of 125 I-anti-CRAM IgG was reduced to ϳ60 and ϳ30% of normal levels measured in the absence of chloroquine, respectively (Fig. 3). After addition of 50, 100, or 200 M chloroquine, the uptake of 125 I-anti-CRAM IgG was reduced to a steady state of ϳ30% of the normal level (Fig. 3A) and the level of degraded 125 I-anti-CRAM IgG was almost completely abolished (Fig. 3B). We confirmed that addition of Ͻ300 M chloroquine does not affect cell viability for at least 20 h (data not shown). These results indicated that degradation of the anti-CRAM IgG molecules in procyclic trypanosomes most likely occurs at the acidic lysosomal compartment. It is possible that chloroquine may also inhibit the recycling efficiency of the CRAM receptor resulting in the reduced level of uptake. Since chloroquine inhibits the release of the anti-CRAM IgG from the CRAM protein, the anti-CRAM IgG⅐CRAM complex may continuously cycle between the surface and lysosomes. The reduced level of uptake may reflect the steady state recycling of anti-CRAM IgG⅐CRAM complex due to chloroquine.

Requirement of CRAM for Uptake and Degradation of 125 I-Anti-CRAM IgG in Trypanosomes-
We further compared the kinetics of uptake and degradation of 125 I-anti-CRAM IgG in wild type procyclic trypanosomes and in the procyclic CRAM null mutant cell line (CRAM 4 -29). Trypanosomes were incubated with different concentrations of 125 I-anti-CRAM IgG at 28°C for 2 h. Then the amount of internalized and degraded 125 I-anti-CRAM IgG was measured. In wild type procyclics, the amount of internalized 125 I-anti-CRAM IgG increased with increasing concentrations of the 125 I-anti-CRAM IgG (Fig. 4A) and saturation was reached at ϳ50 g ml Ϫ1 of the 125 I-anti-CRAM IgG. The uptake of the 125 I-anti-CRAM IgG was efficiently competed by the addition of unlabeled anti-CRAM IgG; increasing concentration of unlabeled anti-CRAM IgG decreased the uptake of 125 I-anti-CRAM IgG (Fig. 5). However, the CRAM null mutant cell line exhibited no specific uptake of the 125 I-anti-CRAM IgG at all, i.e. the cell associated counts measured in the presence and absence of an excess amount of unlabeled anti-CRAM IgG were similar (Fig. 4C). This result indicated that the uptake of the 125 I-anti-CRAM IgG in wild type procyclic trypanosomes is dependent on the presence of CRAM. The CRAM mediated uptake of the 125 I-anti-CRAM IgG in procyclic trypanosomes lead to a significant amount of degradation of the 125 I-anti-CRAM IgG and the amount of degradation increased with increasing concentrations of the 125 Ianti-CRAM IgG (Fig. 4B); while no degradation product can be detected in the incubation medium derived from CRAM null trypanosomes (Fig. 4D). Additionally, late log phase trypanosomes showed a significant reduction in their rate of anti-CRAM uptake and degradation compared with that in early log phase (data not shown). Thus the density to which trypano-somes were grown contributed somewhat to a slight variation in the values of uptake and degradation. It is therefore important to maintain cell lines at the same growth phase for comparison.
Specific Binding of 125 I-Anti-CRAM IgG at 4°C-When trypanosomes were incubated at 4°C with increasing concentrations of the 125 I-anti-CRAM IgG, the amount of specifically bound 125 I-anti-CRAM IgG increased and saturation was reached at ϳ50 g ml Ϫ1 125 I-labeled anti-CRAM IgG (Fig. 6A). Half-maximum binding was achieved with 20 -30 g ml Ϫ1 125 Ianti-CRAM IgG. Addition of an excess amount (20-fold) of unlabeled anti-CRAM IgG efficiently competed with 125 I-labeled anti-CRAM IgG leaving a residual nonspecific binding (Fig. 6A). The binding of 125 I-anti-CRAM IgG at 4°C appeared to be of high affinity and saturable. Since the anti-CRAM is raised against the extracellular 12-amino acid repeat domain, we assumed that anti-CRAM is a monospecific and calculated the binding site for 125 I-anti-CRAM IgG with an equilibrium dissociation constant (K d ) of ϳ10 Ϫ10 M. We further determined the amount of degradation of 125 I-anti-CRAM IgG occurred at 4°C. As shown in Fig. 6B, no significant specific degradation of 125 I-anti-CRAM IgG at 4°C can be detected, indicating that the degradation process is inhibited by incubation at 4°C.
Localization of Internalized Anti-CRAM IgG in the Endocytic Vesicles-To directly visualize the process of internalization of anti-CRAM and to further confirm that internalized materials are indeed delivered into endocytic compartments (endosome and lysosome) in wild type procyclic trypanosomes, we performed subcellular localization of internalized anti-CRAM IgG by indirect immunofluorescence assay. Both wild type and CRAM null mutant cell lines were incubated with 60 g/ml rabbit derived anti-CRAM IgG at 28°C for 2 h. Following washes, cells were fixed in 3.7% formaldehyde and then the subcellular location of internalized anti-CRAM in these fixed trypanosomes was examined using fluorescein isothiocyanateconjugated goat derived anti-rabbit IgG. Obvious staining was only observed in wild type trypanosomes and no staining was observed in the CRAM null mutant cell line. To determine the location of internalized anti-CRAM, we simultaneously per- form co-staining with monoclonal antibody against the endosome/lysosomal membrane glycoprotein p67 (provided by Dr. J. Bangs; 36) and with the DNA specific dye, 4,6-diamidino-2phenylindole staining (Blue). Fig. 7 represents the superimposition of the fluorescence staining with anti-CRAM antibody (green), anti-p67 (red), and 4,6-diamidino-2-phenylindole (blue). The large and small blue dots locate the position of nucleus and kinetoplast, respectively. The green staining locates the internalized anti-CRAM IgG in small vacuoles distributed in between the kinetoplast and the nucleus. These anti-CRAM containing vacuoles either colocalized with the p67 or adjacent to the area identified by p67. This result indicates that the internalized anti-CRAM IgG are indeed distributed into endocytic compartments (endosome and lysosome). We observed that roughly 60% of labeled cells contained one (or two) vacuole containing anti-CRAM in each cell (Fig. 7, panel a  and b) and 40% contained multiple vacuoles (Fig. 7, panel c). Treatment of trypanosome with 100 or 200 M chloroquine increased the population of cells containing multiple vacuoles labeled with anti-CRAM to 64%.
Sequences Required for Efficient Internalization of CRAM-It has been shown in higher eukaryotes that the efficient endocytosis of transmembrane receptor proteins requires a signal sequence in the cytoplasmic domain of the protein to promote clustering into clathrin-coated pits through an interaction with plasma membrane adaptor protein complex. We investigated whether the sequence of the cytoplasmic extension of CRAM is required for the efficient internalization of receptor or receptor-ligand complex in trypanosomes. We compared the uptake and degradation of 125 I-anti-CRAM IgG in wild type procyclic trypanosomes and two previously established CRAM mutant cell lines: cell line CRAMB2 and cell line CRAM-40. 2 The cell line CRAMB2 contains only one functional CRAM gene and the other allele of CRAM was deleted (26). Cell line CRAM-40 contains also only one CRAM gene and this CRAM gene encodes the CRAM protein which was shortened by 40 amino acids from the C terminus (a termination codon was introduced immediately downstream of the sequence encoding the transmembrane domain of CRAM in cell line CRAM-40). We showed that a significant amount of the CRAM protein in cell line CRAM-40 is located at the flagellar pocket. 2 Trypanosomes were incubated with 60 g/ml 125 I-anti-CRAM IgG at 28°C for 2 h and then the cell associated counts and the amount of trichloroacetic acid soluble/non-iodide (silver nitrate unprecipitable) counts in the culture media were measured (Fig. 8). It appeared that 1) cell lines CRAM B2 and CRAM-40 exhibited ϳ33 and 40% reduction, respectively, of their ability to take up the 125 I-anti-CRAM IgG compared with that of wild type trypanosomes; 2) the level of degraded 125 Ianti-CRAM IgG produced by the CRAMB2 cell line is reduced to 60% of that of wild type procyclic trypanosomes; and 3) the CRAM-40 cell line exhibited only a very small residual background level of its ability to degrade 125 I-anti-CRAM IgG (6% of that of wild type procyclic trypanosomes). Thus, following deletion of one allele of CRAM, Ͼ50% of the ability to internalize and degrade 125 I-anti-CRAM IgG was maintained. However, deletion of the cytoplasmic extension of CRAM in cell line CRAM-40 (expressing one allele of CRAM) did not significantly affect the anticipated level of cell associated 125 I-anti-CRAM IgG but abolished the degradation ability. It seems that cell line CRAM-40 may have lost the ability to internalize and transport the CRAM⅐ 125 I-anti-CRAM IgG complex to the lysosomal compartment, although its ability to bind 125 I-anti-CRAM IgG remains. These results indicated that the cytoplasmic extension of CRAM most likely plays an important role in the process of receptor-mediated endocytosis. DISCUSSION Both bloodstream and procyclic form trypanosomes are capable of internalizing macromolecules from their environment. Ultrastructural analysis has shown that structures resembling coated vesicles were only identified in bloodstream form trypanosomes but not in procyclic form trypanosomes (5,6). Due to the apparent absence of coated vesicles in procyclic trypanosomes, the significance of receptor-mediated endocytosis in procyclic trypanosomes has been generally neglected. Most studies on endocytosis by T. brucei have been conducted on the bloodstream form trypanosome. Very little information is available on the process of receptor-mediated endocytosis in procyclic form trypanosomes. Several laboratories have shown that procyclic form trypanosomes endocytose ferritin and peroxidase via fluid-phase endocytosis (5,15,19). We have previously shown that procyclic form trypanosomes endocytose HDL and possibly LDL particles, most likely via a receptor(s) located at the flagellar pocket (31). In this paper, we show that CRAM can function as an HDL receptor and indeed, receptor-mediated endocytosis occurs in procyclic form trypanosomes. By ultrastructural analysis, Langreth and Balber showed that procyclic form trypanosomes can take up macromolecules in smooth vesicles associated with the flagellar pocket (5). It seems that the endocytic machinery in procyclic form trypanosomes is structurally distinct from that in bloodstream form trypanosomes.
We have previously shown that serum lipoproteins are es- sential for optimal growth of procyclic trypanosomes in axenic culture (31). The procyclic form trypanosome has a higher capacity to bind and internalize HDL than LDL. The uptake of HDL particles can occur via receptor-mediated endocytosis in procyclic form trypanosomes. Comparison of the binding and uptake of HDL in the CRAM null mutant and wild type procyclic trypanosomes revealed that deletion of the CRAM gene significantly reduced the ability to bind HDL at 4°C but did not drastically affect the ability to internalize HDL at 28°C. This result indicates that CRAM may play a role in the binding of HDL. However, multiple pathways may be involved in the uptake of HDL in procyclic trypanosomes. Our current data strongly supports the possibility that CRAM may be an HDL receptor. We favor two models to explain our observations. 1) CRAM may be a receptor for the uptake of lipoproteins, however, this receptor is not essential for the survival of the parasite. 2) CRAM may be part of a lipoprotein receptor complex. It has been difficult to distinguish between these possibilities because of the presence of different pathways for the uptake of lipoproteins (37). To distinguish these options, we are currently expressing CRAM in different mammalian cell lines including those lacking LDL receptors. Thus the function of CRAM as a lipoprotein receptor can be evaluated.
The significant amount of HDL uptake observed in the CRAM null cell line must occur through CRAM independent pathways. The presence of multiple pathways for the uptake of HDL has interfered the analysis of CRAM-mediated endocytosis using HDL as a specific ligand. Instead, we used the antibody against the extracellular repeat domain of CRAM as a specific ligand to study CRAM-mediated endocytosis in the procyclic form of T. brucei. This is the first time that receptormediated endocytosis in procyclic form trypanosomes has been well characterized with a defined receptor and ligand. We show that the wild type procyclic trypanosomes, which express a relatively high level of CRAM, can efficiently take up anti-CRAM IgG via CRAM receptor-mediated endocytosis, while the CRAM null mutant exhibited no significant uptake of anti-CRAM IgG. The process of anti-CRAM IgG uptake occurs at 28°C in a time-and concentration-dependent manner in wild type procyclics. The internalized anti-CRAM IgG molecules were degraded in a process which is sensitive to chloroquine and inhibited by incubation at 4°C. Our results indicate that a significant amount of the internalized anti-CRAM IgG is most likely processed in a lysosomal compartment, and we hypothesized that via the recycling of vesicles, the processed products are transported to the flagellar pocket and released to culture media. The treatment with chloroquine also reduced the level of uptake of anti-CRAM IgG which may be due to the reduced efficiency of the CRAM receptor recycling. The binding of anti-CRAM IgG at 4°C appeared to be of high affinity and saturable, and the uptake of anti-CRAM IgG at 28°C reaches saturation with increasing concentrations of 125 I-anti-CRAM IgG. It seems that the process of receptor-mediated uptake is slower in the procyclic form than in the bloodstream form. The uptake of anti-CRAM IgG in procyclic form trypanosomes reaches a steady state only after 2 h of incubation, while the uptake of transferrin in bloodstream form trypanosomes reaches a steady state within a 30 -60-min incubation period (38,39). The different efficiency of receptor-mediated endocytosis between two stages of the parasite could be due to the difference of the receptor number and the difference in the structure of receptorligand complex and the endocytic machinery involved. For example, CRAM is a transmembrane protein, while the transferrin receptor complex is GPI anchored (38 -40); the fractional volume of the flagellar pocket in the bloodstream form is larger than that in the procyclic form trypanosomes (19); and the large coated endocytic vesicles are only present in the bloodstream form and absent in the procyclic form trypanosomes (5,6).
Comparison of the uptake and degradation of anti-CRAM IgG incubated with the wild type procyclic trypanosomes and the mutant cell line CRAM-40 demonstrate that deletion of the cytoplasmic domain of the CRAM drastically reduced the level of degraded 125 I-anti-CRAM IgG observed in the culture media and did not significantly affect the level of cell-associated 125 I-anti-CRAM IgG as expected for the expression of one allele of CRAM only. This result indicates that the cyto- FIG. 7. Subcellular localization of internalized anti-CRAM IgG in wild type procyclic trypanosomes. Trypanosomes were incubated at 28°C for 2 h with 60 g/ml rabbit-derived anti-CRAM IgG in serum-free SDM-79 medium containing 1% (w/v) BSA. After which, cells were washed, fixed in 3.7% of formaldehyde, and spotted on slides. Following fixation in acetone and methanol, these slides were first incubated with mouse-derived monoclonal anti-p67, followed by reaction with fluorescein isothiocyanate-conjugated goat anti-rabbit IgG and rhodamine-conjugated goat anti-mouse IgG. Then cells were stained with 4,6-diamidino-2-phenylindole at a concentration of 0.1 g/ml for 1 min. The images were captured using a CCD camera and analyzed by MetaMorph program from Universal Imaging Co. The images are superimposed and presented in pseudocolors: green for fluorescein isothiocyanate-labeled anti-CRAM; red for rhodamine-labeled p67; blue for 4,6-diamidino-2-phenylindole staining which identifies the nucleus and the kinetoplast. a, b, and c show three representative images. plasmic domain of CRAM may play an important role in the processing of internalized anti-CRAM IgG. We previously compared the cytoplasmic domain sequences of CRAM, the human LDL receptor and the LDL receptor-related protein and found that they share amino acid sequence homology in the domain responsible for internalization of the receptor in humans (the region spanning the tyrosine-based internalization signal; 41): residues 801 to 812 in human LDL receptor, residues 4482 to 4493 of the LDL receptor-related protein and residues 931 to 942 (residues Ϫ4 to Ϫ15 from C terminus) of CRAM are identical in 5 out of 11 amino acids ( Fig. 9; 23). The significance of this homology is not completely clear yet, although deletion of amino acids from Ϫ8 to Ϫ19 from the CRAM C terminus resulted in localization of CRAM in the ER. 2 The studies in higher eukaryotic systems indicated that the cytoplasmic internalization signals direct both rapid internalization and other sorting processes such as targeting to lysosomal/endosomal compartments (42,43). Extensive evidence indicated that the internalization signals first bind directly to the clathrin adaptor protein complexes (AP) and then clathrin binds to membrane-bound AP complexes for the formation of coated pits (27,44,45). In trypanosomes, mechanisms and machinery involved in the receptor-mediated endocytosis is unclear and protein components involved in the coated vesicles have not yet been isolated. The characterization of CRAM-mediated endocytosis provides excellent opportunities for further identification of machinery involved in the receptor-mediated endocytosis pathway in trypanosomes. FIG. 9. Alignment of residues in the short cytoplasmic domain of the LDL receptor related protein (LRP), the human LDL-receptor (LDLR), and CRAM. Asterisks indicate identical residues. The tyrosine in the human (residue 807) had been replaced by a threonine (residue 937 in T. brucei) and this protein retained its capability for internalization, although with a reduced efficiency (41) .   FIG. 8. Comparison of the uptake and degradation of the 125 I-anti-CRAM IgG between the wild type procyclic trypanosome and procyclic CRAM mutant cell lines. Procyclic trypanosomes (1 ϫ 10 7 cells ml Ϫ1 ) were incubated at 28°C for 2 h with 60 g/ml 125 I-anti-CRAM IgG in serum-free SDM-79 medium containing 1% (w/v) BSA. After which, cells were washed and the cell associated counts were measured and referred to as the amount of uptake (panel A). The supernatants were collected for the measurement of degradation products (panel B). The results represent the amount of specific uptake and degradation, calculated as the difference between total (in the absence of unlabeled anti-CRAM IgG) and nonspecific uptake and degradation, respectively (in the presence of 20 fold unlabeled anti-CRAM IgG). Results represent the average of triplicate determinations. Black bars, wild type procyclic trypanosomes; shaded bars, the procyclic trypanosome cell line CRAM B2; white bars, the procyclic trypanosome cell line CRAM-40.